Molecular Biology for the Clinician: Understanding Current Methods

Transcription

1 Pearls of Veterinary Practice Molecular Biology for the Clinician: Understanding Current Methods The basics of molecular biology involve the replication of deoxyribonucleic acid and its transcription and translation into proteins. Biochemical assays such as the Southern blot analysis, polymerase chain reaction (PCR), Northern blot analysis, reverse-transcriptase PCR, microarray technology, Western blot analysis, immunohistochemistry, enzyme-linked immunosorbent assay, and flow cytometry utilize various aspects of molecular biology. To understand these assays requires some basic understanding of the principles of molecular biology. This paper provides basic information on the methodology and techniques used in these assays. J Am Anim Hosp Assoc 2006;42: Heather L. Chandler, BS Carmen M.H. Colitz, DVM, PhD, Diplomate ACVO From the Departments of Veterinary Clinical Sciences (Colitz) and Veterinary Bioscience (Chandler), College of Veterinary Medicine, Ohio State University, Columbus, Ohio Introduction to Basic Molecular Biology The nucleus of all mammalian cells contains a specified number of chromosomes that differs between species. Diploid cells have two sets of chromosomes, with one copy being inherited from each parent cell. Including the sex chromosomes, humans have 46 chromosomes, dogs have 78 chromosomes, and cats have 34 chromosomes. 1 Chromosomes divide by mitosis, and all somatic cells undergo mitosis. Genetic material in chromosomes is composed of deoxyribonucleic acid (DNA). Genes that are on the same chromosome show linked inheritance (i.e., they tend to stay together when DNA undergoes replication). Genes that are on different chromosomes segregate independently (i.e., they recombine at random from one generation to the next). A genome consists of the entire set of chromosomes for any organism. Cells also contain ribonucleic acid (RNA), which is found in abundance in both the cytoplasm and the nucleus. In the nucleus, RNA is concentrated in nucleoli and is attached to the chromosomes. 2 Ribonucleic acid and DNA are constructed of nucleotide building blocks. Each nucleotide contains a phosphate group linked to a sugar group containing five carbon atoms, which in turn is linked to either a purine or a pyrimidine base. Purine bases include adenine (A) and guanine (G). Pyrimidine bases include cytosine (C), thymine (T), and uracil (U). Thymine is only found in DNA, and U is only found in RNA. Deoxyribose is the sugar of DNA, and ribose is the sugar of RNA. In both DNA and RNA, the nucleotides are joined by covalent bonds link- 326 JOURNAL of the American Animal Hospital Association

2 September/October 2006, Vol. 42 Molecular Biological Assays 327 ing the phosphate group of one nucleotide to a hydroxyl group on the sugar of the adjacent nucleotide. These linkages connect the 5 carbon atom of one deoxyribose (or ribose) residue to the 3 carbon atom of the deoxyribose (ribose) in the adjacent nucleotide. 1 The four different bases of DNA are attached in any order along the spine of the helix, giving DNA its specificity. Genetic information is carried in the sequence of the bases, like letters and words making up sentences. The orientation of the sugar-phosphate spine defines the direction of the nucleic acid chains. Nucleic acid sequences are referred to as 5 to 3, because the terminal nucleotide at one end of the chain has a free 5 (amino) group, and the terminal nucleotide at the other end has a free 3 (carboxy) group. The 5 terminus is to the left of the 3 terminus. Chains of DNA and RNA elongate from the 5 end to the 3 end; therefore, transcription begins at the 5 end and terminates at the 3 end. 2 Deoxyribonucleic acid is composed of two strands that form a double helix [Figure 1]. The two strands are connected by hydrogen bonds between the nitrogenous bases. Figure 1 Double helix structure of deoxyribonucleic acid. (Image courtesy of Mr. Timothy Vojt.) The bases pair up in a complementary manner. Guanine hydrogen atoms bond only with C, and A hydrogen atoms bond only with T. The model originally described by Watson and Crick requires that the two chains run in an antiparallel manner, meaning that one strand runs in the 5 to 3 direction and its partner runs in the 3 to 5 direction. 2 Guanine-cytosine base pairs are more stable and stronger than AT base pairs, because GC base pairs have three hydrogen bonds and AT base pairs only have two bonds. Precise replication of genetic material is crucial. Replication of parental DNA strands requires that the strands separate to allow each to become a template for synthesis of a new complementary strand. The separation is transient, and the replication fork is the point where the DNA duplex is disrupted, forming a separated junction. Each of the daughter strands is identical to the original parental strand and consists of one parental strand and one newly synthesized strand. Specific enzymes catalyze synthesis of DNA. For example, DNA polymerase catalyzes DNA synthesis. An interesting limitation is that all polymerases work only from the 5 to 3 direction. The replication fork moves from the 5 to 3 direction on the leading strand but from 3 to 5 on the lagging strand (i.e., the DNA strand on the opposite side of the replication fork that is being replicated in short fragments). Although the lagging strand appears to replicate backward, it actually synthesizes DNA in the appropriate 5 to 3 direction, because RNA primers bind to the lagging strand and synthesize short fragments of DNA called Okazaki fragments. 3 These fragments are later joined together by a series of enzymes to create an intact strand. A problem does exist, however, in that the end of the lagging strand is not fully replicated, leaving an overhanging strand of DNA that is unpaired on the leading strand. This occurs every time DNA undergoes replication; it is called the end replication problem, because the DNA duplex becomes shorter by approximately 50 to 200 base pairs each time it undergoes replication. Continued shortening would be a serious problem if the chromosomes did not have protective caps at the ends, called telomeres. Telomeres protect the genomic DNA from degradation and shortening that could lead to mutations. Telomeric ends are repeated base pairs of TTAGGG that do not encode any genes. 4 Once the telomere becomes critically shortened by continuous DNA replication, an as-yet unknown signal stops cell replication forever. This cell then enters a senescent state. It does not necessarily die at that point and may survive for a certain amount of time. Each cell type has a set number of divisions that it undergoes in its lifetime before it becomes senescent, and this fixed number defines one of the major theories of aging. 5 Cells that require unlimited proliferative capabilities like cancer cells, germ-line cells, and some stem cells have acquired the ability to lengthen their telomeres. One of the telomeric ends has a 3 segment that provides a template for the addition of individual bases. Telomere transferase (telomerase) is a ribonucleoprotein complex that uses an RNA template for synthesizing the telomeric DNA. This complex also has a catalytic subunit (telomerase reverse transcriptase [TRT]) that acts upon its own RNA template. The cells that have telomerase activity can maintain their telomeres and theoretically live forever. 5 Protein Synthesis Deoxyribonucleic acid must be converted to messenger RNA (mrna), which is the working copy of the genetic information, so that it can then be translated into protein synthesis. Accessory proteins, called transcription factors, control the expression of different genes by up-regulating (turning on) or down-regulating (turning off) them. Transcription factors bind to specific areas on genes, called promoters and enhancers, in order to control gene expression. Once mrna is made, it is sent out into the cytoplasm, and the ribosomes (translators) translate the mrna into the amino acid sequences of protein. It was once thought that one gene made one protein; now it is known that one gene can make more than one protein via post-translational modifications. The central dogma of molecular biology is that through DNA transcription and RNA translation, proteins are synthesized.

3 328 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Codons are the groups of nucleotides that code for amino acids. All codons contain three successive nucleotides. Many amino acids are specified by more than one codon. There are a possible 64 combinations of the three bases, and 61 of these are used to code for specific amino acids. The other three combinations (UAA, UAG, UGA) code for stop signals, which act like a period at the end of a sentence. The codon AUG encodes the start codon and also codes for methionine, so it has two potential jobs. The AUG combinations that start polypeptides are all closely preceded by a purine-rich sequence that distinguishes it from the AUG that codes for methionine. 2 Mutations are changes in the coding sequence of a gene. Mutations are important, because they are responsible for inherited disorders and are the source of phenotypic variation (i.e., natural selection). Techniques Common to All Assays Gel Electrophoresis Gel electrophoresis is the most widely used assay method in molecular biology. Photographs of gel electrophoresis are often shown in publications to evaluate, analyze, and demonstrate the results of an experiment. The concept of electrophoresis utilizes the premise that positive charges attract negative charges and vice versa. Both DNA and RNA are negatively charged and move at the same speed toward a positive electrode when dissolved in solution. The sizes of the DNA or RNA molecules can be used to identify them when they are run through a gel matrix. A gel is a semisolid meshwork of cross-linked polymers made of agarose. Gels have gaps in the meshwork that allow DNA or RNA molecules to migrate through them based on size. The smaller molecules migrate more quickly and easily, while larger molecules migrate more slowly. Small fragments of DNA are often run through polyacrylamide or sieving agarose gels. These types of gels have very small pore sizes that allow better resolution of the smaller molecules. Visualization of DNA or RNA molecules is often done after they are run through gels. The most common method of visualizing DNA or RNA is by staining the agarose gel with ethidium bromide, which intercalates itself throughout the nucleic acid structure. 3 Nucleic acids that have ethidium bromide bound to them fluoresce bright orange under ultraviolet light and can be photographed. Radioactive labeling is another method used to visualize nucleic acids. Hybridization The principles of hybridization, also called annealing, are often used in molecular biology to identify DNA or RNA products. Hybridization is the binding of a labeled probe to membrane-bound DNA or RNA. Nucleic acid hybridization is affected by the size of the nucleic acids being hybridized, the overall number of guanine and cytosine molecules (G + C content), the concentration of the nucleic acids, and how close the target matches the probe. Probes are molecules that are tagged with radioactivity, fluorescent agents, or digoxigenin, and they are used to detect another molecule. If the probe used is not 100% compatible with the target, it is said to be mismatched. Numerous molecular techniques use hybridization, including Southern and Northern blot analysis and the polymerase chain reaction (PCR). Techniques for DNA Assays Polymerase Chain Reaction The PCR assay was invented in 1987, and today it is the most widely used molecular technique in all of science. 2 Its applications include clinical diagnostics, genetic and forensic analyses, and genetic engineering. On the most basic level, PCR is used to amplify specific DNA sequences. Most PCR assays are based on known sequences of DNA or segments of genes. A component of a PCR is the template, which is DNA that has the target sequence within it (e.g., the patient s sample). Only a trace amount of template is required for a PCR. A pair of primers is also necessary. A primer is a short segment of single-stranded DNA that matches a portion of the target sequence. Most PCRs use a forward primer, which is located at the 5 end of the target sequence, and a reverse primer, which is located at the 3 end of the target sequence on the complementary strand of DNA. Next, a heat-resistant enzyme drives the reaction. The most commonly used enzyme is derived from Thermus aquaticus and is called Taq polymerase. Other necessary ingredients for the PCR are a large supply of nucleotides for the creation of the DNA strands and an appropriate buffer. The mixture of these ingredients is placed into a thermocycler that is programmed to rapidly change temperature for a preset number of cycles (30 to 50). The mixture is heated to 90 C in order to denature or separate the template s double-stranded DNA. Then the temperature is dropped to between 50 and 65 C, which allows the primers to bind or anneal to the complementary sequence on the template DNA (i.e., the target sequence). Lastly, the temperature is raised to 70 C for 1 to 2 minutes, which allows the Taq polymerase to elongate the new DNA strands beginning at the primers. This elongation always occurs in a 5 to 3 direction. Each cycle doubles the amount of DNA present, and by the end of 30 cycles, there are >1 million copies of double-stranded DNA present. This PCR product is then visualized using agarose gel electrophoresis stained with ethidium bromide [Figure 2]. Cloning In the present context, cloning a gene or a piece of that gene refers to obtaining the DNA that makes up that particular gene, using molecular techniques. The DNA is then used in subsequent experiments such as PCR. 2 For example, if the gene of interest is not known for the species being assayed, then the most common approach used to assess a segment of DNA is to align the sequences of this gene with those of two or three closely related species to compare their homology. In areas that are exact, short

4 September/October 2006, Vol. 42 Molecular Biological Assays 329 Restriction enzymes are made by bacteria, and they cut DNA strands at specific sites. Once the fragmented DNA has undergone agarose gel electrophoresis and detection with ethidium bromide, the DNA from the gel is transferred to a nylon membrane, and then the membrane is incubated with a salt solution containing a labeled DNA probe. The DNA probe only binds to DNA fragments on the membrane that are similar to the probe, and the probe can then be detected using autoradiography. An example of samples evaluated by Southern blot analysis is seen in Figure 3. Figure 2 Agarose gel electrophoresis of 10 polymerase chain reactions (PCRs). The bands seen in lanes 1 and 13 are molecular-weight markers that allow estimation of the size of the PCR products. Bands in lanes 2, 3, and 5 are positive for the reactions performed. The bands in lanes 6, 7, and 10 are positive controls for the samples run in lanes 2, 3, and 9. The reactions run in lanes 4, 8, and 9 are negative. sequences are chosen as the primers in the forward and reverse direction to encompass an area of a few hundred base pairs. The DNA of the species of interest is used as the template, and the chosen primers are added to the mixture, which also includes nucleotides (for making the newly created DNA), Taq polymerase (which drives the reaction), and other necessary ingredients. Most initial experiments use a lower annealing temperature and a higher number of cycles, which allows for promiscuous binding of the primers to areas that are not exactly correct resulting in multiple bands on the agarose gel. Even though a band is expected at a certain point/size, it is also possible to amplify garbage (i.e., material unrelated to desirable material being amplified). In order to decipher the band of DNA product that the assay is seeking, either Southern blot hybridization or sequencing of that DNA product can be used. In a Southern blot analysis, a DNA probe is designed within the expected sequence that does not include either of the primer sequences. This probe is then used to confirm or refute the DNA products amplified by PCR. A more common alternative is to sequence the PCR product. This is explained further later. Southern Blot Analysis Southern blot analysis is used to detect DNA that has been immobilized on a nylon membrane by using a DNA probe. 2 The DNA of interest is separated using agarose gel electrophoresis. The DNA of interest may be a PCR product, a large piece of DNA (e.g., plasmid), or genomic DNA. Plasmid DNA is a circular molecule of double-stranded helical DNA that is used in gene cloning. Genomic DNA is chromosomal DNA isolated from a eukaryotic cell. Since plasmid and genomic DNA strands are very large, they cannot be run through an agarose gel without being cut into smaller pieces, which is performed by restriction enzyme digestion. 2 Figure 3 Illustration of a Southern blot analysis of genomic deoxyribonucleic acid (DNA) extracted from seven samples. The analysis was run on an agarose gel, transferred to a nylon membrane, and then a DNA probe was hybridized with the immobilized DNA on the membrane. The bands indicate hybridization (binding) to various sites on the DNA in each sample. Sequencing of DNA Nucleotides (A, T, C, G) are the letters of the genetic alphabet. A DNA sequence is the specific order in which the nucleotides are spelled out to create genes. Therefore, DNA sequencing is a technique that elucidates the order of

5 330 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 the nucleotides that make up a piece of DNA that has been amplified by PCR or by cloning. Currently, a sequence of 700 nucleotides can be obtained from a sequencing reaction. Larger pieces of DNA must be cut into smaller pieces and then sequenced. Two methods are commonly used to determine DNA sequences. The first is the enzymatic dideoxy (Sanger) method, and the second is the chemical (Maxam-Gilbert) method. 6 Real-Time PCR Real-time PCR is used to quantify the number of copies of DNA or mrna present in a sample. This technology is especially useful in diagnostic medicine, because infectious agents causing clinical disease can be distinguished from normal or background flora. The amplification function of the PCR has a sigmoidal shape that becomes saturated after 20 to 40 cycles, because the conditions become less optimal with each cycle [Figure 4]. The saturation point is the flat plateau of the sigmoid curve, and all samples eventually reach this point even though there are differing amounts of DNA present in each sample. A band seen on an agarose gel following traditional PCR is representative of this plateau (i.e., the entire number of cycles of the PCR). Expression levels of the DNA sequence of interest are compared during the exponential phase of the PCR, long before saturation of the reaction occurs. The way in which real-time PCR monitors the amount of DNA being amplified is by using fluorescent dyes that bind only to double-stranded DNA. The fluorescence increases with each cycle as the DNA amplification increases. A computer program uses an algorithm to analyze the curves for each sample as well as for the standard controls. Following mathematical calculations, data for each sample include the copy number of DNA, which can then be used to achieve a diagnosis. In addition, the copy number of DNA is directly related to cycle number. With a higher copy number, the sigmoid curve reaches the plateau at a lower cycle number. For example, in Figure 4, sample A has a higher expression than sample B, because the cycle number where it crossed the threshold of the plateau (saturation point) is lower. Figure 4 Sigmoidal curves generated by real-time polymerase chain reactions (PCRs). Sample A has a higher expression level than sample B, because the cycle number (i.e., the point where the curve reaches the plateau threshold) is lower.

6 September/October 2006, Vol. 42 Molecular Biological Assays 331 Techniques for RNA Assays Virtually all mature mrnas include a long 3 polyadenosine sequence (AAA repeated 200 to 300 times). This 3 polyadenosine tail is thought to protect the young mrna from being destroyed by RNA degrading enzymes. It is also an exceedingly useful feature that is exploited in some molecular techniques. Northern Blot Analysis Northern blot analysis is the most sensitive technique for detecting mrna expression levels. 7 The RNA is separated by size on a denaturing, formaldehyde, agarose electrophoretic gel. Since RNA has a negative charge, it migrates toward the cathode when a current is applied to the gel. Similar to other forms of electrophoresis, mobility of the molecule in the gel is inversely proportional to its size, such that large fragments do not migrate as rapidly as shorter fragments. After separation by electrophoresis, the RNA is transferred to a nitrocellulose membrane, and the RNA is cross-linked to the membrane by ultraviolet (UV) irradiation. Specific RNA sequences can be detected on the blotting membrane by hybridization (i.e., by binding of a labeled probe to the membrane-bound RNA). The probe can be labeled by coupling to a protein such as digoxigenin, which in turn can be detected using an antidigoxigenin antibody conjugated to alkaline phosphatase that converts a color substrate. This staining provides an amplified signal. Alternatively, the probe can be radiolabeled and detected by autoradiography. The results of hybridization reveal the number, size, and abundance of transcripts, with the quantity of the product being approximately proportional to the amount of the specific RNA present. The concentration of a housekeeping gene, such as 18s-rRNA, can be measured on the same Northern blot to evaluate the relative signal strength. While this technique is time-consuming and requires significant amounts of RNA for analysis, it is still commonly used, because it is extremely quantitative and, therefore, provides an accurate assessment of mrna levels. One example of how this technique has been used in veterinary medicine can be seen in the paper by Green (et al.). 8 The researchers examined the mrna from dogs affected with cardiomyopathy to determine if genetic mutations associated with the disease resulted in abnormal expression of mrna. Reverse-Transcriptase PCR The amount of RNA to be measured is often well below the threshold detected by conventional Northern blot analysis. Reverse-transcriptase PCR (RT-PCR) can be used to assess lower levels of mrna expression in samples by relying on the amplification of a DNA template via PCR. Prior to amplification through PCR, the single-stranded mrna is converted into double-stranded complementary DNA (cdna). To make a DNA copy of the mrna, the enzyme, reverse transcriptase, is used, as well as a primer that binds to the polyadenosine tail of mrna. Once the cdna is made, the RNA is degraded by the enzyme, RNAse H, and DNA polymerases are used to build the complementary strand. The result is a collection of DNA molecules that represents all the specific mrna contained within the cell. Similar to PCR of DNA, primers are designed for a specific gene, and they allow exponential amplification by PCR. The resulting PCR product is subjected to agarose gel electrophoresis to separate the PCR products by size. Ethidium bromide intercalates into the DNA and can be visualized using UV light. An important advantage to RT-PCR is that only a small amount of biological sample is necessary to determine the mrna level. The limitation of this method for determining gene expression is that RT-PCR combined with visualization using ethidium bromide is only semiquantitative, owing to the fact that the PCR is linear only within a range of cycles. While RT-PCR is not well suited for the quantitative comparison of mrna expression, the technique is useful in screening tissues for the expression of a gene. For example, RT-PCR-based assays are used in the clinical setting to detect viruses. Real-Time RT-PCR Real-time RT-PCR is the same technique as real-time PCR; however, the starting material is mrna rather than DNA. If comparisons regarding gene expression levels are made in the saturation phase, large initial differences in mrna expression may be lost; therefore, it is necessary to monitor the time kinetics of amplification. The technology of real-time RT-PCR developed out of the need for a quantitative technique to determine gene expression levels in small samples. Measurement of mrna for a particular gene by realtime RT-PCR is not fundamentally different from RT-PCR. Both techniques require mrna initially to be reverse transcribed to create cdna. The cdna is then subjected to PCR. The main difference is that the instrument where the PCR is carried out is both a thermocycler and a detection device. The reaction mixture contains a fluorescent probe that intercalates into the DNA during synthesis. At the end of each PCR cycle, the amount of fluorescence is assessed. Expression levels of analyzed genes of interest are compared during the exponential phase of the PCR, long before saturation of the reaction occurs. Based on an algorithm, a threshold level of fluorescence must be reached for a mrna to be considered expressed, and the threshold is expressed as a cycle number. Expression of mrna is directly related to cycle number. With higher relative expression, a lower cycle number is required to reach the threshold or plateau [Figure 4]. The copy number of the mrna gene of interest can be standardized against the copy number of a housekeeping gene, or it can be determined from a standard curve. Real-time RT-PCR has been used in the authors laboratory to determine changes in gene expression during corneal wound healing. Different treatments, such as tetracycline, have been found to alter the time and duration of expression of particular genes important for cell migration. 9

7 332 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Microarray A microarray is a tool for analyzing the relative expression level of a gene by determining the amount of mrna present. A microarray can be considered a reverse Northern blot technique. In a Northern blot assay, probes are allowed to search for and bind to the appropriate RNA that has been bound to a membrane. In a microarray, probes for hundreds to tens of thousands of transcripts are immobilized on the surface of a slide or chip, and a labeled RNA solution is then added. The technology allows DNA to be spotted or synthesized directly onto a solid surface in an ordered, predetermined fashion at an extremely high density, facilitating quick and efficient analysis of the expression of thousands of genes in a single experiment. 10 As with both RT-PCR and real-time RT-PCR, mrna is isolated from a specific tissue or cell line and used to generate cdna. Unlike the other techniques, the cdna is labeled with different fluorescent tags (typically red and green) so that the samples can be differentiated upon imaging. Next, the cdna mixture is hybridized with the array, probing the target cdna for the presence of known genetic sequences that are anchored to the array slide. To detect the target cdna that reflects gene expression in an experimental sample, the cdna solution is hybridized with the array, causing the fluorescent-labeled molecules to bind to the sites on the array with DNA sequences that correspond to the genes expressed in each cell. 10 After the hybridization step is complete and the target cdna is bound to the array, the microarray is placed in a scanner that consists of specifically designed lasers, a microscope, and a detector. The fluorescent tags are excited by the laser, and the intensity of the red or green signal corresponds to the amount of DNA present in each sample. Spots that are yellow are equally expressed in the sample. The microscope and detector generate a digital picture of the array, and the intensity of the red-yellow-green spectrum is analyzed for each spot on the microarray. Many genes show low expression levels, and the respective signals may be overwhelmed by background noise. As a result, few conclusions can be drawn; therefore, it is necessary to set a threshold level for transcripts that are included in the analysis. Furthermore, it is important to repeat the analysis with samples that have undergone the same processing (e.g., RNA extraction, cdna synthesis) to ensure that the signal is consistent between samples and experimental runs. It should be noted that while microarray is a good technique to analyze global changes between two samples, it must be validated by other molecular biology techniques. Several major categories of diagnostic questions can be addressed using microarrays as a starting point. These include scanning for the presence of a mutation in a specific disease gene, scanning for expression profiles that differentiate molecular subcategories of disease, scanning largescale changes in DNA copy number that can accompany some acquired diseases, and the identification of markers associated with disease Perhaps the most widely used application of microarrays in animal medicine is the rapid and accurate detection and genotyping of a range of animal pathogens, including bacteria, fungi, and viruses. 14,15 In Situ Hybridization In situ hybridization can be used to localize the expression of different transcripts in thinly sliced tissue sections. 7 The principles underlying this technique are the same as those described above for Northern blotting. In the case of in situ hybridization, DNA or RNA probes (riboprobes) containing either radioactive or nonradioactive labels are used. In general, RNA probes offer increased sensitivity and specificity compared to DNA probes, although RNA probes are more difficult to produce and more labile than DNA probes. The oligonucleotide probe is generated to the mrna of interest and is hybridized against the histological tissue section, which allows the cellular location of the transcript to be visualized. The power of this technique is that the location of mrna expression can be visualized in the context of the whole tissue. In situ hybridization allows for an estimation of the level and location of expression of mrna. An example of the technique is shown in Figure 5. Figure 5 In situ hybridization is illustrated in this figure. Cells were processed appropriately, and a deoxyribonucleic acid probe was used in this experiment. The chromogen used was diaminobenzidine (DAB), so the positive brown staining denotes messenger ribonucleic acid in the nuclei of these cells. Protein Techniques Western Blot Analysis Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) is a technique that is used to separate proteins based on their molecular weight. 7 The detergent SDS has a negative charge, and when added to the sample, it binds hydrophobic amino acid residues. Proteins bind proportionally to the same amount of SDS; therefore, the charge to mass ratio of all proteins is equivalent, and all

8 September/October 2006, Vol. 42 Molecular Biological Assays 333 SDS-protein complexes are negative. Protein samples are prepared either under reducing conditions where ß-mercaptoethanol is added and the sample is boiled, or under nonreducing conditions where ß-mercaptoethanol and heat are omitted. The specific sample preparation conditions are determined by experimental necessity. Once the samples have been prepared for electrophoresis, they are loaded on the gel and an electrical field is applied. The rate of migration of the protein is inversely proportional to the logarithm of its molecular weight meaning, larger proteins move slower through the gel because they encounter more resistance. Protein bands that result from differential migration can be stained for visualization with Coomassie blue or silver, and they can be compared against reference proteins with known molecular weights run on the same gel. While this is a powerful tool that allows examination of global changes in protein levels from samples, it lacks specificity. Western immunoblotting utilizes antibody specificity to ascertain information about the levels of a particular protein in samples. 7 In this technique, after proteins are separated on an SDS-PAGE gel, they are transferred to a membrane by applying a current across the gel and membrane in such a way that the proteins migrate from the gel to the membrane. After the proteins have been transferred, the membrane is incubated with an antibody (primary antibody) manufactured against the protein of interest. Following incubation with the primary antibody, the membrane is incubated with a secondary antibody that is conjugated to horseradish peroxidase and is capable of recognizing the primary antibody. For example, if the antibody against protein X was raised in a rabbit, the second antibody used would be an anti-rabbit horseradish peroxidase-conjugated antibody. This highly specific antibody-to-protein binding is detected using a peroxidase-dependent luminescencegenerating system followed by exposure to X-ray film. The bands that appear reflect the size and relative amount of the studied protein. As seen in Figure 6, specific bands appearing at the appropriate weight/point confirm the protein of Figure 6 In this Western immunoblot of four samples, the first two lanes have lighter bands than the last two lanes. Since the ß-actin and p-akt bands are similar in intensity, the samples in lanes 1 and 2 have less expression of p-akt than the samples in lanes 3 and 4. The ß-actin protein is often used as an assay loading control. interest is present. As an internal control, the blots are stripped and re-probed with an antibody for a housekeeping gene (e.g., ß-actin). If the bands for the housekeeping gene have the same relative intensity, the protein concentrations in each lane are equal. The relative quantity of the protein of interest can then be determined. For example, in Figure 6, all lanes have approximately equal protein concentrations (based on the ß-actin bands), and lanes 1 and 2 have less protein (pakt) present than lanes 3 and 4. Western immunoblotting is an extremely powerful tool for examining the presence or absence of proteins in a sample and for comparing samples ascertained under different conditions. For example, western immunoblotting has been used to determine which proteins are overexpressed during cataract formation The major limitation of Western immunoblotting is that it is semiquantitative. Immunohistochemical Staining Immunohistochemistry is used to localize specific proteins in tissues (immunohistochemistry) or cells (immunocytochemistry). 4 Immunohistochemistry often involves special staining techniques used by pathologists to further characterize tissue specimens. Because immunohistochemistry can be applied to formalin-fixed or frozen tissues and cells, it is easy to use after fine-needle aspiration, biopsy, or necropsy examination. It is not a quantifiable technique; however, if numerous samples are stained at the same time (with appropriate positive and negative controls), these samples can be compared, which makes the technique semiquantitative. Immunohistochemistry has several requirements, including tissue or cells immobilized on a microscope slide, antibody against the protein of interest (primary antibody), appropriate buffers and alcohols, a secondary antibody that has been biotinylated, enzyme-conjugated streptavidin or another conjugate, a chromophore, and a counterstain. As with Western blotting techniques, the secondary antibody is typically immunoglobulins made against the species of the primary antibody. The enzyme conjugate can be horseradish peroxidase, alkaline phosphatase, glucose oxidase, or immunogold. The chromophore is a color-producing substrate system, so it gives the reaction a color. Typically, diaminobenzidine (DAB) or 3-amino-9-ethyl carbazole (AEC) is used as the chromophore. Diaminobenzidine imparts a brown color to positive sites, and AEC imparts a red color to positive sites. Depending on the situation, one chromophore may be better than the other. For example, in samples that contain melanin, AEC is superior, because it provides better contrast against the pigment present in the sample. The counterstain commonly used is hematoxylin, because it stains the nucleus of the cells blue, and nuclear proteins will have a brownish-blue hue. The stained samples can be permanently mounted for future evaluation. Immunofluorescent staining is identical to immunohistochemistry, except that the primary or secondary antibody is labeled with a fluorescent tag. A fluorescent microscope is necessary to evaluate these samples, and

9 334 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 often fluorescence does not last for more than a few weeks to months. In Figures 7A and 7B, one tissue sample is stained using immunohistochemistry (A), and the other is an example of immunofluorescence (B). Enzyme-Linked Immunosorbent Assay Enzyme-linked immunosorbent assay (ELISA) is a biochemical method used to detect the presence of an antigen or antibody in a sample. It has numerous advantages, both techniques are similar. Antibodies become passively attached to the solid phase (microtiter plate) and are called capture antibodies. These antibodies capture or bind the antigens added at the next step. Following incubation and washing, an antibody-antigen complex is attached to the solid phase, and the antigen is said to be trapped. With the direct sandwich ELISA technique, enzyme-conjugated antibodies directed against the antigen are added, and this completes the sandwich. Unbound conjugate is washed Figures 7A, 7B Examples of immunohistochemical (A) and immunofluorescent (B) staining. In 7A, the blue staining is the counterstain, and the brown staining is positive for the protein of interest in the sample. In 7B, the blue color is the counterstain, and the green staining is positive for the protein of interest in the sample. including its simplicity, speed (a few hours), sensitivity, and relatively low cost. Also with ELISA, many samples can be evaluated at one time, there is potential for automation, and numerous kits are commercially available. 19 Two antibodies are typically used; one is specific to the antigen or antibody of interest in the sample, and the other is coupled to an enzyme (enzyme-linked) that causes a signal or color change that is quantified by a spectrophotometer. 19 Three main methods form the basis for all ELISAs: the direct, indirect, and sandwich methods. 5 With the direct ELISA method, antigen is added to a microtiter plate and is passively adsorbed to the plate (solid phase). Following incubation, antibodies specific for the antigen are labeled with an enzyme conjugate and are added to the microtiter plate. Once the enzyme conjugatelabeled antibodies have bound to the antigens on the plate, excess unbound antibodies are washed away, and a substrate or chromophore solution is added that catalyzes the enzyme conjugate to produce a color change. This color change is quantified by a spectrophotometer. The indirect ELISA method is similar to the direct method, but antibodies specific for the antigen bound to the microtiter plate are added to wells and incubated, and then excess unbound antibodies are washed away. Antibodies labeled with enzyme conjugate directed against the species of the original antibody are then added and incubated, and excess is washed away. A substrate or chromophore is added that catalyzes the enzyme conjugate to produce a color change, which is quantified by a spectrophotometer. The sandwich ELISA can be divided into two types: the direct or indirect sandwich ELISA. The first four steps of away, and a substrate or chromophore solution is added that catalyzes the enzyme conjugate to produce a color change, which is quantified by a spectrophotometer. With the indirect sandwich ELISA method, detection antibodies not labeled with an enzyme are then added to the antibodyantigen complex described above. Following incubation and washing, the detection antibodies are themselves identified with an antispecies enzyme conjugate that produces a color change. This color change is quantified by a spectrophotometer. The advantage of ELISA-produced data is that they are generated relatively quickly and can be evaluated easily. This method is often used to compare numerous test and control samples in duplicate or triplicate in a quantitative manner. For example, the ELISA has been used to compare protein alterations in lenses exposed to an oxidative stressor (tertiary butyl hydroperoxide) for various periods of time. 20 All ELISAs must be standardized and validated in order to achieve a level of confidence in the results. Standardization and validation procedures include optimization of the reagents and the protocol, determination of the repeatability of the assay, and determination of the sensitivity and specificity of the assay. 19 Flow Cytometry Flow cytometry is another method of evaluating cells. It is most commonly used to count cells, evaluate cell-surface phenotypic markers, or to analyze cell cycles, cytokine production, apoptosis, and cell viability. 21 It is only useful for cells that are suspended and not clumped together, because

10 September/October 2006, Vol. 42 Molecular Biological Assays 335 the cells are passed within a stream of fluid in front of a light beam. Forward scatter and side scatter, as well as fluorescence detectors, are used in flow cytometry. The molecules to be detected on the cell scatter the light, and fluorescent chemicals in the particle may be excited to emit light at a lower frequency than the light source. The combination of fluorescence and scattered light is detected and analyzed by the machine, resulting in data shown as a onedimensional histogram or as two-dimensional dot plots. An example of a blood sample evaluated by flow cytometry is seen in Figure 8. Newer software programs also allow three-dimensional analysis. Figure 8 A dot plot showing the results of flow cytometry. In the sample, 8% of the cells (upper left quadrant) stain positively for cluster differentiation (CD) 8 with phycoerythrin (PE, red); 23% of the cells (lower right quadrant) stain positively for CD4 with fluorescein-5-isothiocyanate (FITC, green); and 1% of the cells (upper right quadrant) stain positive for both CD4 and CD8. The cells represented in the bottom left quadrant are not staining for either CD4 or CD8. (Image courtesy of Ms. Sarah Leavall and Dr. Mary Jo Burkhard.) Conclusion Molecular biological techniques, especially PCR assays and ELISA, have become powerful diagnostic tools in veterinary medicine. Another interesting application of these techniques is using biological samples from clinical animals to investigate the mechanisms of disease. Information obtained from animals allows the basic research models to be validated as acceptable models for studying specific disease processes. Ultimately, results from translational research may enhance disease prevention and lead to more effective and specific medical treatments. References 11. Clark DP, Russell LD. Molecular Biology Made Simple and Fun. Carbondale IL: Cache River Press, Tagliaferro L, Bloom MV. The Complete Idiot s Guide to Decoding Your Genes. Indianapolis: Alpha Books, Watson JD, Hopskins NH, Roberts JW, et al. Molecular Biology of the Gene. 4th ed. Menlo Park, IL: The Benjamin/Cummings Publishing Co., Blackburn EH. Structure and function of telomeres. Nature 1991;350: Shay JW, Wright WE. Forward: Aging and cancer: are telomeres and telomerase the connection? In: Mattson MP, ed. Telomerase, Aging and Disease. Amsterdam: Elsevier, 2001: Sambrook J, Russell DW. Molecular Cloning A Laboratory Manual. 3rd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, Alberts B, Johnson A, Lewis J, et al. Molecular Biology of the Cell. 4th ed. New York: Garland Science, Green SL, Tolwani RJ, Varma S, et al. Absence of mutations in the survival motor neuron cdna from Labrador retrievers with an inherited myopathy. Vet Rec 2005;157: Chandler HL, Colitz CMH, Kusewitt DF. Role of tetracyclines in healing canine refractory ulcers (abstract). Invest Ophthalmol Vis Sci 2004;45: Fathallah-Shaykh HM. Microarray: applications and pitfalls. Arch Neurol 2005;62: Carter NP, Fiegler H, Piper J. Comparative analysis of comparative genomic hybridization microarray technologies: report of a workshop sponsored by the Wellcome Trust. Cytometry 2002;49: Coussens PM, Nobis W. Bioinformatics and high throughput approach to create genomic resources for the study of bovine immunobiology. Vet Immunol 2002;86: Thomas R, Smith KC, Ostrander EA, et al. Chromosome aberrations in canine multicentric lymphomas detected with comparative genomic hybridizations and a panel of a single locus probes. Br J Cancer 2003;89: Sellon RK. Update on molecular techniques for diagnostic testing of infectious disease. Vet Clin North Am Small Anim Pract 2003;33: Wang RF, Beggs ML, Robertson LH, et al. Design and evaluation of oligonucleotide-microarray method for the detection of human intestinal bacteria in fecal samples. FEMS Microbiol Lett 2002;213: Colitz CMH, Barden CA. Estrogen receptor-alpha expression in canine cataracts (abstract). Vet Ophthalmol 2003;6: Colitz CMH, Barden CA, Chandler H, et al. Phosphorylation of telomerase and estrogen receptor-alpha by Akt in cataracts (abstract). Vet Ophthalmol 2004;7: Colitz CMH, Barden CA, Chandler HC, et al. Telomerase interacts with estrogen receptor alpha in cataractous LEC (abstract). Invest Ophthamol Vis Sci 2004;45: Crowther JR. The ELISA Guidebook. Totowa, IA: Humana Press, Colitz CMH, Vicknair JS. Effect of tertiary butyl hydroperoxide on telomerase activity in canine lens epithelial cells. Invest Ophthalmol Vis Sci 2001;42:E Abstract. 21. Watson JV. Introduction to Flow Cytometry. Cambridge: University Press, 1991.

11 The Effects of Clomipramine Hydrochloride in Cats With Psychogenic Alopecia: A Prospective Study A double-blind, placebo-controlled clinical trial was conducted to determine the efficacy of clomipramine hydrochloride in cats with psychogenic alopecia. Twenty-five cats were randomly assigned to receive clomipramine hydrochloride (0.5 mg/kg orally q 24 hours) or placebo for 56 days. Eleven cats in each group completed the trial. The results of this study showed that clomipramine hydrochloride failed to demonstrate significant changes in the number of grooming bouts, hair regrowth, and the area of alopecia in cats with psychogenic alopecia when compared to a placebo. It was uncertain whether these results reflected a lack of drug efficacy, insufficient treatment duration, or an insufficient number of cases enrolled. J Am Anim Hosp Assoc 2006;42: Petra A. Mertens, DVM, PhD, Diplomate ECVBM, Diplomate ACVB Sheila Torres, DVM, PhD, Diplomate ACVD Carl Jessen, DVM, PhD O From the Department of Veterinary Clinical Sciences, College of Veterinary Medicine, University of Minnesota, St. Paul, Minnesota This study was funded by Novartis Animal Health US, Inc., Greensboro, North Carolina Introduction Grooming behavior serves several purposes, such as cleaning, body temperature control, and ectoparasite removal. 1 Normally, grooming occupies 4% of a cat s 24-hour day. 1 An increase in frequency and intensity of grooming behavior may be precipitated by environmental stressors that cause anxiety, may be a component of a displacement activity, or be secondary to increased pruritus. 2 Neuroendocrine reward mechanisms released during grooming may reinforce and maintain the behavioral pattern, leading to self-inflicted alopecia. 3 Grooming activates serotonergic dorsal raphe nuclei, an area in the central nervous system that has been associated with compulsive disorders. 4 Feline psychogenic alopecia has been compared to obsessive-compulsive disorders in humans and has been suggested as a model for this disease. 5 A compatible history, typical clinical signs, and exclusion of other causes of self-induced, noninflammatory alopecia confirm the diagnosis. 6 Psychogenic alopecia is a well-recognized disorder in cats, although the incidence within the feline population is unknown. Overall et al. reported that 16 of 23 cats that were presented over a period of 11 years for compulsive disorder were self-grooming or self-mutilating. 7 Compulsive disorders in humans and animals are commonly treated with tricyclic antidepressants Case reports and retrospective studies indicate that psychogenic alopecia in cats responds favorably to clomipramine hydrochloride, a tricyclic antidepressant However, no placebo-controlled studies have been conducted to evaluate the efficacy of clomipramine hydrochloride in cats with psychogenic alopecia. Therefore, the purpose of this study was to test the hypothesis that clomipramine hydrochloride significantly reduces the clinical signs of feline psychogenic alopecia when compared to a placebo. Materials and Methods Selection of Cases Cats of any breed, sex, or age referred to the Behavior or Dermatology Service of the College of Veterinary Medicine at the University of 336 JOURNAL of the American Animal Hospital Association

12 September/October 2006, Vol. 42 Feline Psychogenic Alopecia 337 Minnesota from April 2002 through May 2003 with the history and clinical signs of noninflammatory alopecia were recruited for the study. Cases were recruited through referring veterinarians, students, newsletters, staff and faculty within the College of Veterinary Medicine, and word of mouth. To be included in the study, dermatological causes of noninflammatory, self-inflicted alopecia were ruled out. Trichograms confirmed the self-inflicted nature of hair loss and were negative for fungal arthrospores and/or hyphae. Multiple skin scrapings from affected and nonaffected areas and fecal flotation were negative for mites (i.e., Demodex cati, Demodex gatoi, Cheyletiella spp., Notoedric cati). Dermatophyte cultures using a two-section plate containing Dermatophyte Test Media and Sabouraund s dextrose agar a yielded no fungal growth. Cats were enrolled in a food trial and had an intradermal test performed if inflammation was observed on skin biopsies of adjacent nonaffected areas. Cats were excluded if either of these tests indicated food or environmental allergies. Only two cats had inflammation of the nonaffected skin that justified enrollment in a food elimination trial. One cat did not improve during the trial and was subsequently skin-tested with 60 environmental allergens. b The skin test was negative, and this cat was included in the study. The other cat inadvertently received an injection of methylprednisolone acetate during the food trial and was excluded from the study. Cats not enrolled in the study were those that had a previous history of sensitivity to tricyclic antidepressants; those that had received selective serotonin reuptake inhibitors (SSRIs), tricyclic antidepressants, and monoamino oxidase inhibitors (MAOs) within the previous 4 weeks; and cats that had received glucocorticoids within the previous 8 weeks. All cats had a normal serum biochemical profile, urinalysis (UA), complete blood cell count (CBC), serum total thyroxine concentration (T4), and electrocardiogram (ECG) before enrollment in the study. The study was reviewed, approved, and overseen by the Institutional Animal Care and Use Committee. The owners of the cats signed an informed consent prior to participation in the study. Treatment Groups An off-site compounding pharmacist assigned the first cat in the study to receive either a placebo or clomipramine hydrochloride. Thereafter, all subjects were assigned to the treatment groups in an alternating fashion. This process allowed for equivalent numbers of cats in the placebo and clomipramine groups. The investigators were masked to the treatments assigned to the cats. The placebo was composed of microcrystalline cellulose that was identical in appearance to the clomipramine hydrochloride. The owners were allowed to choose among one of three presentation options to administer the placebo or clomipramine hydrochloride: size 3 gel capsules, chicken-flavored liquid, or fish-flavored liquid. Clomipramine hydrochloride c was administered at 0.5 mg/kg orally (PO) q 24 hours. The calculated dosage was placed in one capsule or 0.5 ml of the flavored liquid. Owners who experienced problems medicating their cat were allowed to switch from capsules to a liquid formulation and vice versa. Treatment was not accompanied by specific behavior modification. Owners were asked to avoid reinforcement of the grooming behavior and to refrain from punishing the cats for the undesired behavior. Other prescription drugs for the treatment of this condition, over-the-counter medications, bandages, and Elizabethan collars were not used. Assessments The owners completed a questionnaire on the first day of the study that included questions regarding the cat s age, gender, breed, origin, and total number of animals in the home. The owners were also asked about the cat s medical history, past and present treatment, other illnesses, onset of the problem behavior, and other behavioral problems. Each cat was evaluated daily by the owners throughout the study period. Cats were also examined by a behaviorist and a dermatologist four times at set intervals during the study [see Appendix]. The owners and investigators were masked to the treatments being administered. Evaluations by the owners were performed 7 days before enrollment, which served as the baseline assessment, and daily during the 84 days of the study. Each owner was given a form to record the following: 1) total number of episodes of licking, chewing, and hair pulling observed per day; 2) certain behaviors [e.g., anxiousness, calmness, use of litter pan, interactions with humans and household pets]; 3) special events or environmental changes [e.g., vacation, work noise from construction, unusual number of visitors, stray cats entering the house, etc.]. Additionally, the owners recorded each time the medication was given and noted if any medical problems such as vomiting, diarrhea, or lethargy occurred. In homes with more than one family member, it was recommended that the same person make the observations and notes throughout the study period. The owners were asked to observe the cat in the home environment, following guidelines that were designed to minimize observer effects and to maximize intraobserver reliability (e.g., same observer, same recorder, no variations in time of day and duration of time that the animal was observed). Each cat was evaluated by a behaviorist, and a complete physical examination was performed on days 0, 28, 56, and 84 of the study. Each cat was also evaluated by a dermatologist on days 0, 28, 56, and 84 of the study. A complete dermatological examination was performed, and the extent of the area of alopecia and the degree of hair regrowth were assessed and recorded. An outline of the dorsum, ventrum, and lateral sides of the body were drawn on a graph paper to record the area(s) of alopecia noted at each visit. Changes in the area of the alopecia at each examination were compared to the area recorded on day 0, and a scoring system ranging from 1 to 5 was developed to qualitatively demonstrate percentages in reduction of the alopecic area [Table 1]. A score of 5 indicated no change from day 0. To evaluate

13 338 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table 1 Alopecia Scores Used for Cats Receiving Clomipramine Hydrochloride or Placebo to Manage Psychogenic Alopecia Table 2 Hair Regrowth Scores Used for Cats Receiving Clomipramine Hydrochloride or Placebo to Manage Psychogenic Alopecia Scores * Change in Alopecia Scores * Difference in Hair Shaft Length 1 Reduced by >75% 2 Reduced by >50% to 75% 3 Reduced by >25% to 50% 4 Reduced by 25% 5 New areas developed or no change from d 0 1 Reduced by >75% 2 Reduced by >50% to 75% 3 Reduced by >25% to 50% 4 Reduced by 25% 5 No change from d 0 * Alopecia score is defined as the percent reduction in the area of alopecia compared to d 0. * Hair regrowth score is defined as the percent reduction in the difference of hair shaft length between the affected and nonaffected areas. qualitatively the degree of hair regrowth during the study period, the hair length adjacent to the affected area was measured at the first examination, and the hair length of the affected area was measured at each subsequent visit; then the percent reduction in the difference of hair length between these areas was determined. A scoring system ranging from 1 to 5 was devised to describe these differences [Table 2]. A score of 5 indicated no change from day 0. A CBC, serum biochemical profile, T4, and a UA were performed on days 28 and 56 of the study. One or more tests were repeated on day 84 if any abnormality was noted on day 56. Additionally, an ECG was performed on day 28 and repeated on day 56 if deemed necessary. On day 84 of the study, before the owners were made aware of the cat s treatment, they were asked to gauge the overall changes in their cat s condition during the study period. To verify compliance with treatment administration and daily observations, the owners were required to bring the medication and the daily log with them to each visit. The behaviorist reviewed the log, and a technician counted the number of capsules or the volume of liquid remaining. Cats were removed from the study if they were not returned for all the evaluations, if medications were not administered properly, if the evaluation forms were not completed, or if the cat showed any adverse clinical signs associated with the treatment. Statistical Analysis Two-factor analysis of variance (ANOVA) with repeated measures on one factor (i.e., time) was used to compare the numbers of grooming bouts for the placebo and treatment groups. Geisser Greenhouse-corrected P value was used to adjust for the inevitable violations of compound symmetry that accompanies this experimental design. 12 Kruskal- Wallis tests were carried out to compare the mean alopecia score and the mean hair regrowth score of the placebo and treatment groups at days 28, 56, and 84 of the study. Data from 22 cats were included in the statistical analyses. For all statistical analyses, P values of 0.05 were considered significant. Statistical analyses were carried out using a computer software program. d Results Case Material Twenty-five cats met the selection criteria, were included in the study, and were sequentially assigned to either the treatment (n=12) or placebo (n=13) group. Of these, 22 cats completed the study. Two cats were withdrawn from the study because of insufficient recording of data or improper treatment administration. One cat in the clomipramine group was removed from the study when it developed a urinary tract obstruction on day 18. Ages of the cats at the time of enrollment ranged from 4 to 14 years (mean 8.9 years). Fourteen cats were spayed females, and 11 were castrated males. The cats included domestic shorthairs (n=22), domestic longhairs (n=2), and one Siamese. The placebo and treatment groups each included one domestic longhair. Owners had acquired the cats from a rescue organization or animal humane society (n=13), private source (n=5), breeder (n=1), pet store (n=1), or as strays (n=5). Twentytwo (88%) of the cats were in multicat households, with one to six additional cats. Only two cats had access to the outdoors. Nine (36%) households included one or more dogs, and six (24%) cats lived with one or more children. The onset of the problem was often unknown, and most owners were unable to associate the onset of the behavior with a specific event. Other behavioral complaints recorded by the owners included a display of anxiety or fear in the presence

14 September/October 2006, Vol. 42 Feline Psychogenic Alopecia 339 of loud noises, strangers, animals, or unusual situations (n=15); aggression toward humans or other animals (n=5); and elimination outside the litter box (n=2). On presentation, alopecia was detected on the abdomen of 24 (96%) cats, on the medial thigh of seven (28%) cats, on the carpus of seven (28%) cats, and on the flank or dorsum of three (12%) cats. Assessment Outcomes No significant differences (ANOVA, P=0.13) in the number of grooming episodes were noted between the cats that received clomipramine and cats that received the placebo [Figure 1]. No significant differences were observed in the mean alopecia scores (Kruskal-Wallis test, P=0.44, P=0.06, P=0.39) and mean hair regrowth scores (Kruskal-Wallis test, P=0.19, P=0.48, P=0.34) between the placebo and clomipramine groups at days 28, 56, and 84 [Figures 2, 3]. No abnormalities attributed to clomipramine were observed on physical examination, auscultation, or ECG in any of the cats. Results of all CBCs, T4s, and serum biochemical profiles were within normal limits throughout the study. The UA of one cat showed struvite crystals and a ph of 8.0. This cat was diagnosed with urethral obstruction after 18 days of clomipramine and was excluded from the study. Other side effects attributed to clomipramine included lethargy (n=5), constipation (n=1), and reduced appetite (n=1). Four owners reported reduced interaction with their cats during the treatment period, and two owners had difficulties administering medication. Of the cats that received the placebo, one cat had sporadic vomiting throughout the study (less than once weekly), one cat vomited once during week 2, and one cat sought the owner s attention more often during the first 2 weeks of the study. Seven (64%) of 11 owners of cats that received clomipramine and three (27%) of 11 owners of cats that received the placebo reported that the cat s hair coat and behavior improved by at least 50%. Discussion Feline psychogenic alopecia is a well-recognized, selfinflicted skin disorder that results from excessive grooming, chewing, and/or hair-pulling. This persistent and repetitive grooming behavior has led to the disorder being classified as compulsive. 13 The disease may be triggered or aggravated by various stressors, such as anxiety, a geographic move, changes in social circumstances or relationship, trauma, etc., but the owners of cats in the study reported here were unable to identify any of these stressors with the onset of the problematic behavior. 2 Fifteen cats in the current study showed other fear-related behavioral problems. This observation was consistent with other studies on pets suffering from compulsive disorders. 7 Underlying anxiety that arises from real or perceived environmental stimuli may lead to conflict or frustration in the cat. Displacement activities, Figure 1 Mean number of grooming bouts as observed by the owners from day 0 through day 84. Data were collected from 11 cats receiving clomipramine hydrochloride (0.5 mg/kg orally q 24 hours) and 11 cats receiving a placebo for 56 days.

15 340 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Figure 2 Mean alopecia scores recorded at day 0, treatment days 28 and 56, and posttreatment day 84. Data were collected from 11 cats receiving clomipramine hydrochloride (0.5 mg/kg orally q 24 hours) and 11 cats receiving a placebo for 56 days. No significant differences were observed between the groups (P=0.44, P=0.06, P=39). See Table 1 for score definitions. such as grooming, may then offer the pet a way to decrease this tension. 2 Currently, very little is known about the beneficial effect of clomipramine in treating feline psychogenic alopecia. In a previous report, a cat with psychogenic alopecia that failed to respond to prednisolone, diazepam, and phenobarbital demonstrated significant improvement after treatment with clomipramine. 10 In a different study, three cats with signs of psychogenic alopecia responded very well to clomipramine, and in two of these cats the behavior recurred after the dosage was reduced or the treatment was interrupted. 9 Sawyer et al. reported on the beneficial effect of clomipramine in five cats with psychogenic alopecia. 8 Clinical signs did not recur in four cats after discontinuing clomipramine, and three of these cats received environmental modifications to help maintain control of the condition. The study reported here was designed as a double-blind, placebo-controlled trial to investigate the effects of clomipramine on feline psychogenic alopecia. The diagnosis of psychogenic alopecia was based on history, clinical presentation, and the exclusion of other possible causes of similar signs. All cats underwent extensive testing to rule out other possible diseases, but a food trial, an intradermal allergy test, and a parasiticide response trial were not performed in every cat before inclusion in the study. It is unlikely that the cats enrolled in this study had flea bite hypersensitivity, because they lived in a region where fleas were rare, and none of the cats had clinical signs suggestive of this disorder. Multiple skin scrapings and a fecal flotation were performed to rule out demodicosis or cheyletiellosis, and all cats tested negative; therefore, the likelihood of these diseases causing the cats alopecia was reduced. It would have been preferable, however, to have performed a parasiticide response trial in all the cats before inclusion in the study, because it was possible that some cats with demodicosis or cheyletiellosis were unintentionally enrolled. Histopathological evidence of inflammation in nonaffected skin was used as a criterion to determine if food or environmental allergies were present. This criterion was used because clinical signs of feline allergies rarely present as noninflammatory alopecia. 6 However, it was possible that some of the cats enrolled in this study had food allergies and/or atopic dermatitis. It could be argued that the lack of response to clomipramine in this study was influenced by the inclusion of cats with allergic or parasitic disorders. However, the authors believe that if any of such cases were

16 September/October 2006, Vol. 42 Feline Psychogenic Alopecia 341 Figure 3 Mean hair regrowth scores recorded at day 0, treatment days 28 and 56, and posttreatment day 84. Data were collected from 11 cats receiving clomipramine hydrochloride (0.5 mg/kg orally q 24 hours) and 11 cats receiving a placebo for 56 days. No significant differences were observed between the groups (P=0.19, P=0.48, P=0.34). See Table 2 for score definitions. unintentionally included in the study, they were very few. Moreover, the arbitrary allocation of the cats into the placebo or treatment groups would have reduced any influence that cats with allergic or parasitic disorders might have had on the results. When the areas of alopecia and hair regrowth during and after treatments were analyzed, no significant differences were observed between the clomipramine and placebo groups. Hair regrowth was also noted in both groups. Clomipramine also did not significantly decrease the rate of grooming behavior. Despite these results, >50% of the owners of cats that received clomipramine felt that their pets improved by at least 50%. In contrast, this level of improvement was noticed by <33% of owners of cats treated with the placebo. These findings indicated that from the owner s perspective, clomipramine had a superior effect in reducing the excessive grooming. The validity of the owners comments was limited, however, since other factors may have influenced their responses and expectations. Side effects associated with tricyclic antidepressants are commonly related to their serotonergic, anticholinergic, anti-α 1 -adrenergic, and antihistamine effects Most published reports on the side effects of clomipramine pertain to dogs. In the dog, clomipramine may cause decreased activity, nausea, and a decrease in the circulating levels of thyroid hormones The drug does not cause cardiovascular problems in healthy dogs In cats, reported side effects include sedation and urinary retention. 14 The most common side effects observed in the study reported here were lethargy, constipation, and urinary retention, which may have arisen from the drug s anticholinergic effects. It was uncertain if the onset of these side effects led the owners to surmise their cats received clomipramine and therefore affected their evaluations. The results of this study showed that clomipramine failed to produce significant improvement in feline psychogenic alopecia. It is possible that if more cats were included in the study, a significant difference in skin lesions may have been reached. The level of owner compliance required by the study design and the strict inclusion and exclusion criteria may also have limited the number of cats that were recruited in the study during the 13-month enrollment period. Furthermore, a longer duration of treatment may have allowed more significant effects to be discovered. Conclusion Clomipramine hydrochloride administered orally at 0.5 mg/kg per day failed to demonstrate significant changes in

17 342 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Appendix Algorithm Depicting All Evaluations Performed at Various Times in a Study of Cats With Psychogenic Alopecia Pre-enrollment Evaluation History taking Physical examination Behavior and dermatological evaluations Fecal flotation Skin scraping Fungal culture Skin biopsy Trichogram Biopsy showed signs of inflammation: - Food trial and challenge - Food trial and challenge were negative; performed skin test Day 0: Initiation of Study Physical examination Behavior and dermatological evaluations Serum biochemical profile Complete blood count Urine analysis Serum thyroxine levels Electrocardiogram Treatment assignment Day 28: Midterm Evaluation Physical examination Behavior and dermatological evaluations Serum biochemical profile Complete blood count Urine analysis Electrocardiogram Day 56: End of Treatment Period Behavior and dermatological evaluations Serum biochemical profile Physical examination Complete blood count Urine analysis Day 84: End of Follow-up Period Physical examination Owner was made aware of the cat s assigned treatment Behavior and dermatological evaluations

18 September/October 2006, Vol. 42 Feline Psychogenic Alopecia 343 the number of grooming bouts, hair regrowth, and degree of alopecia in cats with psychogenic alopecia. Further placebocontrolled trials enrolling larger numbers of cats to be treated for longer periods of time are necessary to confirm the results reported here. a K-800 SAB-DUET; Bacti-Lab, Mountain View, CA b Allergenic extracts; Greer, Lenoir, NC c Clomipramine hydrochloride; Gallipot, Inc., Mendota Heights, MN d Statistical package for the Social Sciences, version ; SPSS Inc., Chicago, IL References 11. Eckstein RA, Hart BL. The organization and control of grooming in cats. Appl Anim Behav Sci 2000;68: Virga V. Behavioral dermatology. Vet Clin North Am Small Anim Pract 2003;33: Spruijt BM, van Hooff JARAM, Gipsen WH. The ethology and neurobiology of grooming behaviour. Physiol Rev 1992;72: Fornal CA, Metzler CW, Marruso F, et al. A subgroup of dorsal raphe serotonergic neurons in the cat is strongly activated during oral-buccal movements. Brain Res 1996;716: Dodman NH, Moon-Fanelli A, Mertens PA, et al. Veterinary models of OCD. In: Hollander E, Stein DJ, eds. Obsessive-Compulsive Disorders Diagnosis, Etiology, Treatment. New York: Marcel Dekker, 1994: Scott DW, Miller WH, Griffin GE. Psychogenic skin diseases. In: Scott DW, Miller WH, Griffin GE, eds. Small Animal Dermatology. Philadelphia: WB Saunders, 2001: Overall KL, Dunham AE. Clinical features and outcome in dogs and cats with obsessive-compulsive disorders: 126 cases. J Am Vet Med Assoc 2002;221: Sawyer LS, Moon-Fanelli AA, Dodman NH. Psychogenic alopecia in cats: 11 cases ( ). J Am Vet Med Assoc 1999;214: Seksel K, Lindeman MJ. Use of clomipramine in the treatment of anxiety-related and obsessive-compulsive disorders in cats. Aust Vet J 1998;76: Swanepol N, Lee E, Stein DJ. Psychogenic alopecia in a cat: response to clomipramine. J S Afr Vet Assoc 1998;69: McTavish D, Benfield P. Clomipramine: an overview of its pharmacological properties and a review of its therapeutic use in obsessive compulsive disorder and panic disorder. Drugs 1990;39: Glantz SA, Slinker BK. Primer of Applied Regression and Analysis of Variance. 2nd ed. New York: McGraw-Hill, 2001: Luescher AU. Diagnosis and management of compulsive disorders in dogs and cats. Vet Clin North Am Small Anim Pract 2003;33: Pfeiffer E, Guy N, Cribb A. Clomipramine-induced urinary retention in a cat. Can Vet J 1999;40: Hart BL, Cliff KD, Tynes VV, et al. Control of urine marking by use of long-term treatment with fluoxetine or clomipramine in cats. J Am Vet Med Assoc 2005;226: King JN, Steffan J, Health SE, et al. Determination of the dosage of clomipramine for the treatment of urine spraying in cats. J Am Vet Med Assoc 2004;225: Gulikers KP, Paniciera DL. Evaluation of the effects of clomipramine on canine thyroid function tests. J Vet Intern Med 2003;17: Pouchelon JL, Martel E, Champeroux P, et al. Effects of clomipramine hydrochloride on heart rate and rhythm in healthy dogs. Am J Vet Res 2000;61: Reich MR, Ohad DG, Overall KL, et al. Electrocardiographic assessment of antianxiety medication in dogs and correlation with serum drug concentration. J Am Vet Med Assoc 2000;216:

19 Cardiovascular Dysfunction in Dogs Associated With Critical Illnesses The records of 16 dogs with left ventricular dysfunction associated with severe systemic illness were reviewed. The most common diagnoses in affected dogs were sepsis and cancer. Despite left ventricular dysfunction, no dog presented with signs of congestive heart failure. Fifteen dogs were presented with generalized weakness as a part of their clinical complaint. Twelve (75%) of 16 dogs died or were euthanized within 15 days of admission to the hospital. The average time until death was 3.6 days. J Am Anim Hosp Assoc 2006;42: O. Lynne Nelson, DVM, MS, Diplomate ACVIM (Internal Medicine, Cardiology) Pam A. Thompson, LVT RS From the Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Washington State University, Pullman, Washington Introduction Cardiac dysfunction can be caused by abnormalities in contractility or diastolic stiffness, changes in loading conditions, and heart rate or rhythm or valvular dysfunction. 1,2 A primary cardiac abnormality such as ventricular dysfunction is often suspected when animals are presented with weakness or respiratory compromise. 1,2 However, in some humans, cardiac dysfunction may be a manifestation of other circulatory disturbances. Primary causes of such disturbances are often related to complications of systemic inflammation. 3-8 In addition, acute cardiovascular dysfunction may also occur in humans with primary, stable myocardial abnormalities, who have recently become critically ill from other metabolic disorders. 7 Septic shock and systemic inflammatory responses are common causes of acute depression of ventricular function in humans. 7-9 Serum from people with sepsis has been shown to contain a factor that decreases contractility in isolated muscle strips. 10 Tumor necrosis factor (TNF) and other cytokines are thought to mediate decreased cardiac contractility by activating inducible nitric oxide synthase. 10 In inflammatory disease, activated leukocytes have also been shown to contribute to ventricular hypocontractility by generation of oxygen free-radical damage. 11 Furthermore, clinical studies in critically ill people have shown that cardiac injury is a frequently unrecognized complication. 12 In one study, the incidence of myocardial injury as assessed by troponin-i was unexpectedly high (15% of critically ill patients) and correlated well with increased morbidity and mortality in the intensive care setting. 12 Recognizing acute causes of depressed ventricular function may be more important in critically ill patients, because acute causes are potentially reversible and may affect outcome. 7,8,12 The purpose of this report was to retrospectively evaluate dogs with critical illnesses and left ventricular dysfunction. The hypothesis of the study was that identification of left ventricular dysfunction in critically ill dogs was associated with an unfavorable clinical outcome. Materials and Methods The records of dogs admitted to the intensive care unit (ICU) at Washington State University between July 2000 and January 2003 with 344 JOURNAL of the American Animal Hospital Association

20 September/October 2006, Vol. 42 Cardiovascular Dysfunction 345 critical illnesses and left ventricular dysfunction were reviewed. Dogs that required supportive therapy in the ICU were selected. Fluid therapy and constant-rate infusions, blood pressure and central venous pressure monitoring, serial laboratory evaluations, continuous electrocardiographic monitoring, and oxygen therapy were common supportive therapies in this group of dogs. Critical illness was defined as significant metabolic derangements that required intensive care to sustain life or enhance metabolic stability. Dogs that were housed in the ICU for observation purposes (e.g., neurological signs, trauma, or during anesthetic recovery) were not included. If necropsy evaluation and cardiac histopathology were unavailable, breeds of dogs (e.g., Doberman pinscher, boxer, sporting breeds, giant breeds) with an increased incidence of dilated cardiomyopathy were also excluded from the study because of their greater risk of primary cardiac dysfunction. An exception was made for two boxers, one of which underwent a cardiac evaluation 2 weeks prior to the critical illness and one that had extensive follow-up during the recovery period. Sixteen dogs met the criteria for inclusion in this study. The diagnosis and followup were recorded for each case. Dogs with unproven diagnoses were excluded. All dogs had thoracic radiographs as a part of the evaluation of the critical illness. Routine echocardiograms with M-mode and Doppler examinations were performed in selected cases by the same person, using commercially available equipment. a Four cardiac cycles were averaged to determine the reported value for each measured parameter. Standard normal canine echocardiographic parameters were utilized to evaluate the dogs, as has been previously described Left ventricular systolic dysfunction was defined by determining the percent fractional shortening of the left ventricular cavity by M-mode echocardiography and/or percent ejection fraction as determined by a modified Simpson s rule. 14,15 A percent fractional shortening of <26% and/or a percent ejection fraction of <46% were used to define left ventricular systolic dysfunction. The authors chose to use a more restrictive definition of left ventricular systolic function, because it was felt that many larger, athletic breeds of dogs may normally demonstrate a lower-end percent fractional shortening. The authors realized that a percent fractional shortening of 26% would likely indicate poorer left ventricular function in a small dog versus a large-breed dog. Dogs that had obvious endocarditis were excluded from the study regardless of ventricular function. Left ventricular pre-ejection period and pre-ejection period compared to ventricular ejection time have been infrequently reported in dogs. 16 A prolonged pre-ejection period to ejection time ratio was interpreted as reduced pressure generation by the left ventricle, with delayed opening of the aortic valve and delayed onset of ventricular ejection. 16 The normal pre-ejection period to ejection time ratio used was defined as 0.2 to 0.35 based on data from the authors echocardiography laboratory. This value was similar to results previously reported in dogs. 16 Dogs with preejection period to ejection time ratios of >0.4 were considered to have systolic dysfunction, which was also similar to reported values in humans. 17 Doppler echocardiography was also utilized to assess left ventricular diastolic function by interrogation of the mitral valve inflow velocities. Mitral valve inflow E to A waveform ratios of <1 were interpreted as diastolic dysfunction in dogs with normal sinus rhythm. 18,19 Results The signalment, diagnosis, pertinent echocardiographic results, and clinical outcomes are presented in the Table. Despite left ventricular dysfunction, no dog had signs of congestive heart failure. The most common clinical complaint was generalized weakness (n=15). One dog (case no. 3) had tachypnea and radiographic evidence of a generalized, interstitial pulmonary pattern. This dog had the lowest percent fractional shortening (13%) of any dog in the study. Necropsy of this dog did not reveal a primary myocardial problem, and pulmonary histopathology indicated a nonspecific inflammatory process. No other dog showed symptoms or radiographic changes consistent with possible cardiovascular-related pulmonary pathology. All dogs were thought to be normally hydrated at the time of echocardiographic assessment. Five dogs were judged to be normally hydrated on initial presentation, whereas the remaining 11 dogs received intravenous fluid therapy for a minimum of 24 hours prior to the echocardiographic studies. In addition to reduced percent fractional shortening, other pertinent echocardiographic findings were increased left ventricular pre-ejection period to ejection time ratio (n=6), mitral inflow Doppler E wave to A wave inversion (n=3; 10 dogs were considered tachycardic, creating E/A summation), and gradual aortic valve closure during systole on M-mode echocardiography, suggesting subnormal cardiac output (n=3). The most common diagnoses in dogs with critical illness and left ventricular dysfunction were bacterial sepsis (n=5) and cancer (n=5). Other conditions diagnosed in affected dogs are listed in the Table. Of the 16 dogs evaluated, 12 (75%) died or were euthanized within 15 days of admission to the hospital. The average time until death was 3.6 days (range 1 to 15 days). Dogs that died or were euthanized in the hospital had necropsy evaluations, with the exception of two dogs that had antemortem diagnoses of lymphosarcoma. Necropsy findings did not reveal gross or histological evidence of cardiac disease in any dog. Four of the 16 dogs were discharged from the hospital and had follow-up times ranging from 20 days to 2 years. Of the four dogs, one had a diagnosis of lymphosarcoma and was alive at 20 days and undergoing chemotherapy with the local veterinarian. A boxer dog with diagnoses of immunemediated polyarthropathy, anemia, and hyperglobulinemia was still alive 2 years later. The percent fractional shortening in this dog at the time of admission to ICU was 21%, and 2 years later it was 34%. Two dogs with systemic coccidiomycosis were alive 3.5 and 4 months after discharge and were receiving systemic antifungal therapy.

21 346 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table Clinical Data for 16 Dogs With Cardiac Dysfunction Associated With a Critical Illness Echocardiographic Parameters Case No. Signalment * Diagnoses FS% LVIDd (mm) PEP:ET Other Findings Outcomes 1 2-y-old, 5-kg, CM Sepsis, hepatitis, multiple Gradual AV closure Died, d 7 dachshund organ dysfunction 2 2-y-old, 22-kg, SF Lymphosarcoma Alive, followed for 20 d border collie 3 5-y-old, 26-kg, SF Sepsis, hepatitis, Gradual AV closure Euthanized, d 3 German shepherd coagulopathy, dog cross pulmonary inflammation 4 3-y-old, 22-kg, SF Inflammatory bowel disease, Pleural and pericardial Died, d 2 boxer pancreatitis, acute renal failure effusion 5 7-y-old, 25-kg, CM Immune polyarthropathy, Uniform valve thickening, Alive, followed for 2 y boxer anemia, hyperglobulinemia mild congenital subaortic stenosis 6 4-y-old, 32-kg, SF Mediastinal lymphosarcoma Mitral inflow E/A inversion Euthanized, d 1 Labrador retriever 7 5-y-old, 21-kg, SF Immune-mediated hemolytic Gradual AV closure Died, d 4 Australian cattle dog anemia, pulmonary thromboembolism, pancreatitis 8 4-y-old, 35-kg, SF Lymphosarcoma Died, d 15 German shepherd dog 9 9-y-old, 26-kg, CM Mesothelioma, pleural effusion Euthanized, d 3 border collie (Continued on next page)

22 September/October 2006, Vol. 42 Cardiovascular Dysfunction 347 Table (cont d) Clinical Data for 16 Dogs With Cardiac Dysfunction Associated With a Critical Illness Echocardiographic Parameters Case No. Signalment * Diagnoses FS% LVIDd (mm) PEP:ET Other Findings Outcomes 10 4-y-old, 33-kg, CM Coccidiomycosis Died, d 2 German shepherd dog 11 1-y-old, 2-kg, M Coccidiomycosis Alive, followed for 4 mo chihuahua 12 5-y-old, 38-kg, CM Coccidiomycosis, Mitral inflow E/A inversion Alive, followed for 3.5 mo malamute-cross pleural effusion 13 4-y-old, 12-kg, SF Sepsis, hepatopathy Mitral inflow E/A inversion Died, d 2 beagle 14 4-y-old, 34-kg, SF Chronic severe pyoderma, Died, d 1 Labrador retriever sepsis 15 7-y-old, 21-kg, CM Pulmonary adenocarcinoma Mitral inflow E/A inversion Euthanized, d 2 Samoyed cross 16 8-y-old, 8-kg, SF Sepsis, pancreatitis Died, d 1 miniature schnauzer * CM=castrated male, SF=spayed female, M=male FS%=percent fractional shortening; LVIDd=left ventricular internal dimension in diastole, measured in millimeters; PEP:ET=left ventricular pre-ejection period to ejection time ratio; AV=aortic valve; E/A=mitral Doppler E wave to A wave ratio

23 348 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Discussion Although primary myocardial disease (e.g., occult dilated cardiomyopathy) could not be completely ruled out in this group of critically ill dogs with left ventricular dysfunction, it was not suspected based on their age, breed, or follow-up evaluations. Ten dogs had necropsy examinations, none of which identified gross or histological cardiac abnormalities. Despite having left ventricular dysfunction, no dogs had signs of congestive heart failure. Fifteen dogs had moderate to severe weakness, but it was unclear how much the cardiac dysfunction contributed to the weakness. Many dogs of this report had significant systemic inflammatory conditions, particularly bacterial sepsis. Sepsis and other systemic inflammatory diseases are common causes of acute depression of ventricular function in humans, and these critically ill dogs may have been affected similarly. 3-9 Many mechanisms have been identified to explain how systemic illnesses contribute to ventricular dysfunction in humans. 8-11,20-25 Acute causes of depressed ventricular contractility in people include ischemia, hypoxemia, respiratory and metabolic acidosis, low ionized serum calcium, exogenous toxins, endogenous toxins (e.g., depressant factors generated during sepsis), hypothermia, and hyperthermia. 25 Serum from people with sepsis may contain a circulating myocardial depressant factor. 10 Tumor necrosis factor, other cytokines, and activated leukocytes may also contribute to ventricular hypocontractility. 11 Early in the septic process, hypoperfusion can be confounded by paralysis of arteriolar smooth muscle and vasodilatation. However, later in the course, the decline in cardiac output, leading to death, involves significantly decreased contractility and increased diastolic stiffness. 25 Many similar conditions were found in this group of 16 critically ill dogs. The underlying illnesses of these dogs may also have caused increased oxygen demands as a result of elevations in sympathetic nervous system tone. 25 In both humans and animals, sepsis-associated cardiac dysfunction is characterized by ventricular dilatation, decreased ventricular contractility, and diminished relaxation Hypoxemia and anemia may also significantly reduce myocardial oxygen delivery and cause myocardial oxygenation defects. 20 Such deficits are often encountered in humans with sepsis and may worsen myocardial hypoxia. When the heart becomes progressively hypoxic, the proportion of oxygen extracted by the heart increases but may not be enough to prevent anaerobic metabolism; thus, myocardial lactic acid production may ensue. 21 Contractility becomes depressed in lactic acidosis, and the ventricles may acutely dilate. 21 If oxygen delivery in relation to demand is not corrected quickly, the heart may enter a detrimental positive-feedback loop of decreased contractility, cardiac output, and coronary perfusion that may eventually lead to cardiac arrest. In the experimental canine model, this vicious cycle occurred when arterial saturation fell below 75% (i.e., PaO 2 =40 mm Hg). 20 In addition to sepsis, other systemic inflammatory syndromes may be linked to cardiac dysfunction in humans, such as advanced neoplasia and immune-mediated disorders. 26,27 In the group of dogs reported here, five had cancer, and two were diagnosed with immune-mediated disorders. It is important to identify the potential causes of depressed ventricular contractility in critically ill animals, because many factors may significantly contribute to morbidity and mortality. Such factors may also affect animals with primary cardiac diseases when they develop systemic illnesses. It may be especially important to recognize the acute mechanisms of ventricular dysfunction, because they may be reversible. 7,8,12,25 Differentiating acute versus chronic causes of reduced ventricular function may be difficult, particularly in those dogs in which primary heart disease may be encountered. In critically ill dogs that are identified with reduced ventricular function and for which primary myocardial disease is not expected, acute ventricular dysfunction associated with systemic illness should be considered. In addition, dogs with known myocardial disease that also develop severe systemic disorders should not have their poor ventricular function dismissed as solely deterioration of their myocardial disease. In such animals, correction of acute, reversible contributors to decreased muscle contractility may potentially improve outcome. The treatment regimens for the dogs of this report were quite varied and were not compared. The primary treatment for a critical animal with reduced ventricular function is reversal of the underlying cause, if possible. It would seem prudent that if ischemia and hypoxia are present, then aggressive corrective measures should be instituted. Correction of significant anemia may result in substantial improvement in ventricular function. 25 Ideally, attention should be paid to eliminating factors that increase myocardial oxygen demand. In humans, positive inotropic agents are commonly used to enhance myocardial contractility and cardiac output in the critically ill. 12,25 Choosing the lowest level of inotropic support that produces a therapeutic effect is thought to minimize myocardial oxygen demands. 12,25 Alleviating pain or sedating an overly excited animal may diminish increased sympathetic tone and tachycardia. 25 In animals with pulmonary disease, respiratory acidosis may be addressed with oxygen therapy and ventilatory support. In animals with metabolic disease, current evidence indicates that alkalinization is of no benefit and may be detrimental in a critically ill patient. 31 Treatment of myocardial depression from circulating myocardial depressant factors has been attempted in humans with antibodies to endotoxin and TNF, naloxone infusions, and dialysis. 25 These therapies appear beneficial in some cases but are still considered experimental. 25 Some of the mentioned therapies were utilized in the dogs of this report; however, no published clinical studies have evaluated treatment strategies in dogs with critical systemic illnesses and ventricular dysfunction. The outcomes reported here suggest that identification of left ventricular dysfunction in dogs with critical illnesses may be a poor prognostic indicator, particularly in those animals with sepsis and cancer. A significant limitation to this report was the lack of a control group with which to compare outcomes. In addition, a follow-up echocardiogram

24 September/October 2006, Vol. 42 Cardiovascular Dysfunction 349 was performed in only one of the four surviving dogs. Clinical studies in critically ill humans have shown that cardiac injury is a frequently unrecognized complication, and it correlates well with increased morbidity and mortality. 12 Thus, aggressively treating any systemically ill dog with unexpected cardiac dysfunction, regardless of clinical presentation, may be prudent. The prevalence of cardiac dysfunction associated with critical illness in animals is unknown. If cardiovascular-related signs are not obvious, many critically ill dogs may not receive a thorough cardiac examination. Conclusion This report suggests that left ventricular dysfunction may occur in dogs with critical illnesses. Identification of left ventricular dysfunction in critically ill dogs may be a poor prognostic indicator, particularly in those with sepsis and cancer. Early recognition and treatment of ventricular dysfunction could potentially limit morbidity and mortality in these critically ill dogs. Well-designed, prospective studies to evaluate treatments and outcomes are warranted. a HDI 3000; ATL Ultrasound, Bothell, WA References 11. Hamlin RL. Pathophysiology of the failing heart. In: Fox PR, Sisson D, Moise NS, eds. Textbook of Canine and Feline Cardiology. 2nd ed. Philadelphia: WB Saunders, 1999: Colucci WS, Braunwald E. Pathophysiology of heart failure. In: Braunwald E, ed. Heart Disease: A Textbook of Cardiovascular Medicine. 5th ed. Philadelphia: WB Saunders, 1997: Liu H, Song D, Lee SS. Cirrhotic cardiomyopathy. Gastroenterol Clin Biol 2002;26: Liu H, Lee SS. Cardiopulmonary dysfunction in cirrhosis. J Gastroenterol Hepatol 1999;14: Noutsias M, Pauschinger M, Poller WC, et al. Current insights into the pathogenesis, diagnosis and therapy of inflammatory cardiomyopathy. Heart Fail Monit 2003;3: Dhainault JF, Hayghebaert MF, Monsallier JF, et al. Coronary hemodynamics and myocardial metabolism of lactate, for fatty acids, glucose and ketones in patients with septic shock. Circulation 1987;75: Court O, Kumar A, Parrillo JE. Clinical review: myocardial depression in sepsis and septic shock. Crit Care 2002;6: Krishnagopalan S, Kumar A, Parrillo JE. Myocardial dysfunction in the patient with sepsis. Curr Opin Crit Care 2002;8: Bozza M, Satoskar AR, Lin G, et al. Targeted disruption of migration inhibitory factor gene reveals its critical role in sepsis. J Exp Med 1999;189: Parrillo JE, Burch C, Shelhamer JH, et al. A circulating myocardial depressant substance in humans with septic shock. J Clin Invest 1985;76: Granton JT, Goddard CM, Allard MF, et al. Leukocytes and decreased left ventricular contractility during endotoxemia in rabbits. Am J Respir Crit Care Med 1997;155: Guest TM, Ramanathan AV, Tuteur PG, et al. Myocardial injury in critically ill patients, a frequently unrecognized complication. JAMA 1995;273: Thomas WP. Two-dimensional, real-time echocardiography in the dog. Vet Radiol Ultrasound 1984;25: Schiller NB, Shah PM, Crawford M, et al. Recommendations for quantification of the left ventricle by two-dimensional echocardiography. J Am Soc Echocardiogr 1989;2: Bonagura JD, O Grady MR, Herring DS. Echocardiography: principles of interpretation. Vet Clin North Am Small Anim Pract 1985;15: Atkins CE, Synder PS. Systolic time intervals and their derivatives for evaluation of cardiac function. J Vet Intern Med 1992;6: Kyriakidis M, Antonopoulos A, Georgiakodis F, et al. Systolic time intervals after phenylephrine administration for early stratification of patients after acute myocardial infarction. Am J Cardiol 1994;73: Schober KE, Fuentes VL. Effects of age, body weight, and heart rate on transmitral and pulmonary venous flow in clinically normal dogs. Am J Vet Res 2001;62: Rakowski H, Appleton C, Chan KL, et al. Canadian consensus recommendations for the measurement and reporting of diastolic dysfunction by echocardiography. J Am Soc Echocardiogr 1996;9: Walley KR, Becker CJ, Hogan RA, et al. Progressive hypoxemia limits left ventricular oxygen consumption and contractility. Circ Res 1988;63: Steenbergen C, Deleeuw G, Rich T, et al. Effects of acidosis and ischemia on contractility and intracellular ph of rat heart. Circ Res 1988;41: Walley KR, Lewis TH, Wood LDH. Acute respiratory acidosis decreases left ventricular contractility but increases cardiac output in dogs. Circ Res 1990;67: Teplinsky K, O Toole M, Olman M, et al. Effect of lactic acidosis on canine hemodynamics and left ventricular function. Am Physiol 1990;258:h1193-h Desai TK, Carlson RW, Geheb MA. Prevalence and clinical implication of hypocalcemia in acutely ill patients in a medical intensive care setting. Am J Med 1988;84: Walley KR, Wood LDH. Ventricular dysfunction in critical illness. In: Hall JB, Schmidt GA,Wood LDH, eds. Principles of Critical Care. 2nd ed. New York: McGraw-Hill, 1998: Nihoyannopoulos P, Gomez PM, Joshi J, et al. Cardiac abnormalities in systemic lupus erythematosus association with raised anticardiolipin antibodies. Circulation 1990;82: Wolfe F, Michaud K. Heart failure in rheumatoid arthritis: rates, predictors, and the effect of anti-tumor necrosis factor therapy. Am J Med 2004;116: Natanson C, Eichenholz PW, Danner RL, et al. Endotoxin and tumor necrosis factor challenges in dogs simulate the cardiovascular profile of human septic shock. J Exp Med 1989;169: Parker MM, McCarthy KE, Ognibene FP, et al. Right ventricular dysfunction and dilatation, similar to left ventricular changes, characterize the cardiac depression of septic shock in humans. Chest 1990;97: Suffredijni AF, From RE, Parker MM, et al. The cardiovascular response of normal humans to the administration of endotoxin. N Engl J Med 1989;321: Cooper DJ, Walley KR, Wiggs BR, et al. Bicarbonate does not improve hemodynamics in critically ill patients who have lactic acidosis. Annals Intern Med 1990;112:

25 Outcomes of Cats With Oral Tumors Treated With Mandibulectomy: 42 Cases Medical records of 42 cats treated with mandibulectomy for oral neoplasia at eight institutions were reviewed to determine morbidity, progression-free interval, and survival time. Progressionfree and survival rates at 1 and 2 years were 56% and 49%, and 60% and 57%, respectively. Cats with squamous cell carcinoma had significantly shorter survival than cats with fibrosarcoma or osteosarcoma. Seventy-two percent of cats were dysphagic or inappetent immediately postoperatively, and 12% never regained the ability to eat. Despite acute morbidity in 98% and long-term morbidity in 76% of cats, 83% of the 30 owners providing information were satisfied with the outcome of mandibulectomy. J Am Anim Hosp Assoc 2006;42: Nicole C. Northrup, DVM, Diplomate ACVIM (Oncology) Kimberly A. Selting, DVM, Diplomate ACVIM (Oncology) Kenneth M. Rassnick, DVM, Diplomate ACVIM (Oncology) Orna Kristal, DVM, Diplomate ACVIM (Oncology) Maura G. O Brien, DVM, Diplomate ACVS Gillian Dank, DVM, Diplomate ACVIM (Oncology) Ravinder S. Dhaliwal, DVM, Diplomate ACVIM (Oncology) Shyla Jagannatha, PhD Karen K. Cornell, DVM, PhD, Diplomate ACVS Tracy L. Gieger, DVM, Diplomate ACVIM (Oncology, Internal Medicine) RS Introduction Functional outcome and tumor control are major concerns when considering mandibulectomy as a therapeutic option for tumors of the oral cavity. Squamous cell carcinoma and fibrosarcoma are the most common oral tumors of the cat. 1,2 These tumors are locally aggressive, and recurrence after surgical resection is common. Descriptions of only 21 cats treated with mandibulectomy have been reported. 3-8 Sixteen of these cats From the Comparative Oncology Program (Northrup, Jagannatha, Cornell, Gieger), Department of Small Animal Medicine and Surgery, College of Veterinary Medicine, University of Georgia, Athens, Georgia 30602; the Animal Cancer Center (Selting), College of Veterinary Medicine and Biomedical Sciences, Colorado State University, Fort Collins, Colorado; the Department of Clinical Sciences (Rassnick), College of Veterinary Medicine, Cornell University, Ithaca, New York; Harrington Oncology Program (Kristal), School of Veterinary Medicine, Tufts University, North Grafton, Massachusetts; Veterinary Centers of America (VCA) West Los Angeles Animal Hospital (O Brien), 1818 Sepulveda Boulevard, Los Angeles, California 90025; the Department of Surgery and Radiological Sciences (Dank), School of Veterinary Medicine, University of California-Davis, Davis, California; South Bay Veterinary Specialists (Dhaliwal), 5440 Thornwood Drive #E and H, San Jose, California 95123; and All-Care Animal Referral Center (Dhaliwal), Amistad Street, Fountain Valley, California Doctor Selting s current address is the Department of Veterinary Medicine and Surgery, College of Veterinary Medicine, University of Missouri, 379 East Campus Drive, Columbia, Missouri Doctor Kristal s current address is Kfar Saba 44260, Israel. Doctor O Brien s current address is Orchard Road Veterinary Surgery, 2519 South Shields #106, Fort Collins, Colorado Doctor Dank s current address is Faculty of Agriculture, Food, and Environmental Quality Sciences; The Koret School of Veterinary Medicine, POB 121, Rehovot 76100, Israel. Doctor Jagannatha s current address is Merial Limited, 115 Transtech Drive, Athens, Georgia Doctor Gieger s current address is Department of Clinical Sciences, Cornell University, College of Veterinary Medicine, Ithaca, New York JOURNAL of the American Animal Hospital Association

26 September/October 2006, Vol. 42 Mandibulectomies in Cats 351 were clinically abnormal after mandibulectomy, exhibiting signs such as protrusion of the tongue, ptyalism, dysphagia, and medial migration of the remaining mandible (i.e., mandibular drift). 3-8 Three cats died because of postoperative dysphagia. Two were unable to prehend food, and one aspirated food. 3-5 The outcomes for these cats suggested that mandibulectomy may have a significant negative impact on the oral function and quality of life of treated cats. The majority (15 of 21) of cats described in previous reports of mandibulectomy were treated for squamous cell carcinoma. 3-5,7 Eleven of these cats experienced recurrences, including six of seven cats treated with mandibulectomy followed by definitive radiation therapy. 7 For the eight cats treated with mandibulectomy alone for squamous cell carcinoma, five experienced recurrences, and the median survival time was 5.5 months (range 5 weeks to 12 months). 3-5 Only two of these eight cats survived 10 months, suggesting that even aggressive surgery might be inadequate for local control of oral squamous cell carcinoma in the cat. 3-5 Based on the limited information available on cats treated with mandibulectomy, the locally invasive nature of most feline oral neoplasms, and the small size of the feline mandible, the hypothesis of the study reported here was twofold: mandibulectomy would be ineffective for local control of oral tumors in cats, and a mandibulectomy aggressive enough to successfully control oral tumors in cats would result in permanent loss of oral function and compromised quality of life. The purpose of this retrospective study was to evaluate postoperative survival time, tumor control, and morbidity in a large number of cats treated with mandibulectomy. Materials and Methods Case Material The medical records of all cats treated with mandibulectomy at eight participating veterinary referral hospitals between October 1984 and January 2002 were retrospectively reviewed. Each cat included in this study had a mandibulectomy for oral neoplasia and had follow-up information available on tumor control and survival time. Information collected from the medical records describing treated cats included age, gender, presenting clinical signs, duration of clinical signs prior to presentation, and histopathological diagnosis. Information collected regarding the oral tumor included location, size, presence of osteolysis on mandibular imaging with radiography or computed tomography (CT), completeness of excision based on histopathological evaluation of the margins of excised tissues, results of thoracic radiography, and results of mandibular lymph node cytological examination. Staging tests, including imaging of the mandible, lymph node cytology, and thoracic radiography, were performed at the discretion of the attending clinician. When the information was available in the medical record, clinical tumor stage (based on the diameter of the primary tumor) was classified as T1 (<2 cm), T2 (2 to 4 cm), or T3 (>4 cm) according to the World Health Organization (WHO) Tumor, Node, Metastasis (TNM) system. 9 To further define tumor stage, if imaging of the mandible was performed, the presence of osteolysis was included in the WHO stage description (i.e., substage a [no osteolysis] versus substage b [osteolysis present]). Information obtained regarding treatment included type of surgery performed, whether an enteral feeding tube was placed, duration of the feeding tube, postoperative oral function and complications (e.g., ability to eat, ptyalism, tongue protrusion, malocclusion, palate trauma, mandibular drift), and any additional therapy. The types of surgery performed were described as complete unilateral mandibulectomy, bilateral rostral mandibulectomy, resection of >50% of the mandible (i.e., complete unilateral mandibulectomy combined with partial resection of the contralateral mandible or bilateral rostral mandibulectomy with removal of >50% of the mandible), or segmental unilateral mandibulectomy. 2 The type of mandibulectomy performed was based on the surgeon s judgment of the appropriate procedure for the cat and the tumor. Acute morbidity was defined as complications of surgery noted during the first 4 weeks after surgery, and long-term morbidity was defined as complications that did not resolve during the cat s lifetime or follow-up time. Outcome information retrieved from the medical records included development of local recurrence as determined by physical examination, development of metastasis, progression-free interval (i.e., the time, in days, from mandibulectomy until detection of local tumor recurrence or metastasis), survival status, survival time (i.e., the time, in days, from mandibulectomy until death from any cause), cause of death, and owner satisfaction with outcome. Additional follow-up information was obtained by telephone communication with owners and referring veterinarians. Owner satisfaction with outcome was assessed from written reports in the medical records or from direct communication with the owners. Parameters evaluated to determine postmandibulectomy quality of life and owner satisfaction included the cat s ability to prehend food, the cat s quality of life as perceived by the owner, whether the owners were satisfied with the outcome of the surgery, and whether they would consider mandibulectomy surgery again. Statistical Analysis Median survival time and progression-free interval were estimated using the Kaplan-Meier method. 10 Cats that were alive or disease-free at the time of last follow-up were included until the day of last follow-up and then censored from the respective analysis. Cats that died of causes unrelated to their oral tumors were censored from the analysis of survival time on the day of their death. Survival curves were generated using GraphPad Prism software. a Statistical testing was performed using SAS software. b Logrank tests were used to evaluate the relationship of variables to survival time and progression-free interval if there were sufficient

27 352 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table 1 Clinical Information Concerning Cats Treated With Mandibulectomy for Squamous Cell Carcinoma, Fibrosarcoma, or Osteosarcoma Squamous Cell Carcinoma Fibrosarcoma Osteosarcoma Tumor Type (n=21) * (n=6) * (n=6) * Age (y) 13.8 (5-17.5) 8.8 (4-12) 9.5 (5-18) Tumor location Rostral Entire unilateral mandible Middle unilateral mandible Caudal Not available in medical record WHO tumor stage T T T Not available in medical record Type of mandibulectomy Complete unilateral Bilateral rostral Resection of >50% of mandible Segmental unilateral Surgical margins Complete Incomplete Not available in medical record Previous therapy Radiation therapy Chemotherapy Additional therapy Radiation therapy Chemotherapy Chemotherapy and radiation therapy Local recurrence Complete surgical margins Incomplete surgical margins Unknown surgical margin status Still alive (Continued on next page)

28 September/October 2006, Vol. 42 Mandibulectomies in Cats 353 Table 1 (cont d) Clinical Information Concerning Cats Treated With Mandibulectomy for Squamous Cell Carcinoma, Fibrosarcoma, or Osteosarcoma Squamous Cell Carcinoma Fibrosarcoma Osteosarcoma Tumor Type (n=21) * (n=6) * (n=6) * Died of disease Died of unrelated diseases Lost to follow-up Progression-free rates (%) 1 y y y Survival rates (%) 1 y y y * Number of cats Median (range) WHO=World Health Organization All cats were dead or censored by 3 y. numbers of cats in the variable groups to allow analysis. 11 The relationship of progression-free interval and survival time to the following variables was evaluated: diagnosis (i.e., squamous cell carcinoma versus fibrosarcoma or osteosarcoma), WHO tumor stage, tumor location (i.e., rostral versus other), type of surgery (i.e., bilateral rostral versus complete unilateral mandibulectomy), histopathological evidence of completeness of excision, and use of additional therapies. For all analyses, significance was defined as P Results The inclusion criteria for this study were met by 42 cats undergoing mandibulectomy from October 1984 to January 2002 at Colorado State University (n=21), Tufts University (n=6), University of Georgia (n=4), Cornell University (n=3), Veterinary Centers of America (VCA) West Los Angeles Animal Hospital (n=3), University of California- Davis (n=2), All-Care Animal Referral Center (n=2), and South Bay Veterinary Specialists (n=1). The median age of the cats was 11.2 years (range 1 to 18 years). Twenty-three cats were castrated males, 18 were spayed females, and one was an intact male. Presenting clinical signs included a visible mandibular mass (n=39; one noted by the referring veterinarian at a routine dentistry), ptyalism (n=10; four with hemorrhage from the tumor), dysphagia (n=7), inappetence and/or anorexia (n=5), weight loss noted by the owner (n=3), halitosis (n=3), behavioral changes (n=3), inability to open the mouth (n=1), pain perceived by owner (n=1), pawing the mouth (n=1), and cessation of grooming (n=1). The median duration of clinical signs prior to presentation was 28 days (range 1 day to 6 years). Histopathological diagnoses included squamous cell carcinoma (n=21), fibrosarcoma (n=6), osteosarcoma (n=6; one parosteal), salivary gland adenocarcinoma (n=2), lymphoma (n=2; one with concurrent radiation-induced osteonecrosis of the mandible), undifferentiated carcinoma (n=1), chondrosarcoma (n=1), acanthomatous epulis (n=1), melanoma (n=1), and osteoma (n=1). Clinical data and outcomes of cats with the three most common tumor types are summarized in Table 1. Tumor location on the mandible was rostral (n=19), entire unilateral mandible (n=7), caudal unilateral mandible (n=4), middle unilateral mandible

29 354 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table 2 Acute * (n=40 Cats) and Long-term (n=38 Cats) Morbidity Observed in Cats Treated With Mandibulectomy for Oral Tumors Type of Mandibulectomy Bilateral Complete Resection of >50% Segmental Rostral Unilateral Mandible Unilateral Total for All Cats Acute Long-term Acute Long-term Acute Long-term Acute Long-term Acute Long-term Morbidity Noted (n=12) (n=10) (n=19) (n=19) (n=6) (n=6) (n=3) (n=3) (n=40) (n=38) No complications Dysphagia or inappetence Ptyalism Mandibular drift Tongue protrusion Pain Difficulty grooming Dehiscence Malocclusion with palate injury Temporomandibular joint crepitus Death * Acute defined as 4 wks immediately following surgery Remainder of the cat s life or follow-up time Number of cats with follow-up

30 September/October 2006, Vol. 42 Mandibulectomies in Cats 355 (n=3), and not recorded in nine cats (all treated with complete unilateral mandibulectomy). Primary tumors were clinically staged as T1 (n=8), T2 (n=11), or T3 (n=13) based on visual measurement. Ten did not have tumor measurements documented in the medical record. Results of radiographic or CT examinations were available for 30 cats, and osteolysis was evident in 27 of the mandibles. Cats with available tumor measurements and radiographic assessment of osteolysis were able to be staged as T1a (n=1), T1b (n=4), T2b (n=8), or T3b (n=11). Cytological examination of fine-needle aspirates of the ipsilateral mandibular lymph node (n=11) revealed no evidence of metastasis. Thoracic radiographs revealed no evidence of metastasis (n=39). The only cat with abnormal radiographic findings was diagnosed with mandibular osteosarcoma and had a 1-cm pulmonary nodule identified at the time of mandibulectomy. No further radiographs were obtained, and this cat was euthanized 238 days following mandibulectomy for recurrent local disease. No necropsy was performed. Surgical procedures performed included complete unilateral mandibulectomy (n=21), bilateral rostral mandibulectomy (n=12), resection of >50% of the mandible (n=6), and segmental unilateral mandibulectomy (n=3). No intraoperative complications were noted. One cat developed cardiopulmonary arrest of unknown cause and died 3 to 4 hours after a complete unilateral mandibulectomy for squamous cell carcinoma. Histopathological descriptions of the surgical margins were available for 33 tumors. Seventeen specimens (seven of squamous cell carcinoma, three of osteosarcoma, two of fibrosarcoma, and one each of chondrosarcoma, undifferentiated carcinoma, osteoma, acanthomatous epulis, or amelanotic melanoma) had no microscopic evidence of tumor at the margins. Sixteen specimens (nine squamous cell carcinomas, two fibrosarcomas, two osteosarcomas, two salivary adenocarcinomas, and one lymphoma) had microscopic residual disease. Surgical margins were clean for six of 12 cats that underwent bilateral rostral mandibulectomy, nine of 21 cats that underwent complete unilateral mandibulectomy, and two of six cats that underwent resection of >50% of the mandible. Microscopic residual disease was found in three of 12 cats treated with bilateral rostral mandibulectomy, seven of 21 cats treated with complete unilateral mandibulectomy, three of six cats treated with resection of >50% of the mandible, and all three cats treated with segmental unilateral mandibulectomy. One cat died within hours of mandibulectomy, and another cat had no available information on postoperative morbidity in the medical record. Acute morbidity was observed in 39 (98%) of the remaining 40 cats. Most cats experienced more than one adverse effect in the 4 weeks following surgery [Table 2]. Because it was difficult to distinguish between dysphagia and inappetence based on the information provided in medical records, cats were described as having abnormalities in eating. At the end of the study, 27 cats were dead, 12 cats were alive, and three cats were lost to follow-up at 237, 357, and 963 days. The median follow-up time for the 15 surviving cats was 813 days (range 186 to 2137 days). The date of latest follow-up was August 4, The median survival time for all 42 cats was not reached (range 0 to 2920 days), and 1- and 2-year survival rates were 60% and 57%, respectively [Figure 1]. In the logrank analysis, tumor type was the only variable significantly associated with survival time [Table 3]. A diagnosis of squamous cell carcinoma was associated with shorter survival time (P=0.03) compared to fibrosarcoma and osteosarcoma [Figure 2]. Figure 1 Kaplan-Meier curve depicting survival time (in days) for 42 cats treated with mandibulectomy for oral neoplasia. Small, vertical lines indicate days of last follow-up for living cats and cats dying of unrelated diseases. The median survival time was not reached at 2920 days. Figure 2 Kaplan-Meier curves illustrating survival time (in days) for 33 cats treated with mandibulectomy for oral squamous cell carcinoma (SCC), fibrosarcoma (FSA), or osteosarcoma (OSA). Small, vertical lines indicate days of last follow-up for living cats and cats dying of unrelated diseases. The median survival time was significantly shorter (P=0.03) for cats with squamous cell carcinoma (217 days) compared to cats with fibrosarcoma or osteosarcoma (median survival times not reached).

31 356 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table 3 Survival Rates and Results of Logrank Tests on the Association of Potential Prognostic Variables With Survival of Cats Treated With Mandibulectomy for Neoplasia Survival Rates No. of Variable Cats 1 y (%) 2 y (%) P Value * Tumor type 0.03 Squamous cell carcinoma Fibrosarcoma Osteosarcoma World Health Organization tumor stage 0.32 T T T Location 0.18 Rostral Other Type of mandibulectomy performed 0.09 Complete unilateral Bilateral rostral Surgical margins 0.11 Negative Positive Additional therapy utilized 0.44 Yes No * Statistical significance defined as P Fourteen cats (including eight with squamous cell carcinomas, two with fibrosarcomas, and one each with an osteosarcoma, chondrosarcoma, salivary gland adenocarcinoma, or lymphoma) were euthanized because of progression of local disease. One cat with squamous cell carcinoma was euthanized for lymph node metastasis with no evidence of local recurrence at 217 days following complete unilateral mandibulectomy. Another cat was euthanized for dyspnea from mandibular lymph node and pulmonary metastasis of salivary gland adenocarcinoma at 444 days after segmental unilateral mandibulectomy. It was not possible to determine from the medical record of this cat whether the dyspnea was associated with upper airway obstruction or pulmonary/pleural involvement. Multiple pulmonary nodules were detected 7 months prior to euthanasia, and local recurrence was also present at the time of euthanasia. These two cats were the only ones for which initial tumor progression and death could be attributed to metastatic disease. Three additional cats with squamous cell carcinomas developed lymph node metastasis at the time of or after local tumor recurrence, but they were euthanized because of local progression of the oral masses. Eighteen cats developed local tumor recurrence during the follow-up time of the study. These 18 cats included eight with squamous cell carcinoma, four with fibrosarcoma, one with osteosarcoma, two with salivary gland adenocarcinoma, two

32 September/October 2006, Vol. 42 Mandibulectomies in Cats 357 with lymphoma, and one cat with chondrosarcoma. The median progression-free interval for all 42 cats was 544 days, and 1- and 2-year progression-free rates were 56% and 49%, respectively [Figure 3]. Stages of the tumors that recurred included T1 (n=2), T2 (n=5), and T3 (n=4). Local recurrence was noted in 10 tumors removed by complete unilateral mandibulectomy, four removed by bilateral rostral mandibulectomy, one removed by resection of >50% of the mandible, and all three tumors removed by segmental unilateral mandibulectomy. Local recurrence was noted for nine Figure 4 Kaplan-Meier curves illustrating the progression-free interval (in days) for 33 cats treated with mandibulectomy for oral squamous cell carcinoma (SCC), fibrosarcoma (FSA), or osteosarcoma (OSA). Small, vertical lines indicate days of last follow-up for cats with no local recurrence or metastasis. Differences in progressionfree interval based on diagnosis of squamous cell carcinoma versus fibrosarcoma or osteosarcoma were not statistically significant (P=0.16). Figure 3 Kaplan-Meier curve depicting the progressionfree interval (in days) for 42 cats treated with mandibulectomy for oral neoplasia. Small, vertical lines indicate days of last follow-up for cats with no local recurrence or metastasis. The median progression-free interval was 544 days. tumors with histopathological evidence of residual disease at the surgical margins and for four tumors with no microscopic residual disease. Three of the four tumors that recurred despite apparent complete resection were squamous cell carcinomas, and the fourth was chondrosarcoma. Six of the seven tumors that did not recur despite apparent incomplete resection were squamous cell carcinomas. The cats with incompletely excised squamous cell carcinoma that did not develop recurrence by the end of the study had a median follow-up time of 169 days (range 11 to 813 days), and two of them received adjunctive radiation therapy and chemotherapy. The seventh cat with no recurrence, despite apparent incomplete resection, had an osteosarcoma and was treated with adjunctive radiation therapy. This cat was still alive 1688 days after surgery. In logrank analysis, none of the variables examined correlated significantly with the progression-free interval [Table 4; Figure 4]. Information regarding acute and long-term complications of mandibulectomy was available for 40 and 38 cats, respectively. This information is summarized in Table 2. One cat developed dehiscence after the second week of definitive radiation therapy following complete unilateral mandibulectomy for osteosarcoma. For this cat, tumor excision was incomplete based on histopathological evaluation of margins, and radiation therapy was initiated at approximately 30 days after surgery. Another cat developed a ranula 3 months after complete unilateral mandibulectomy for squamous cell carcinoma. Only nine cats were described as completely normal after recovery. Five cats had severe anorexia and/or dysphagia that prevented them from eating to the time of their death. One of these cats died of dilated cardiomyopathy 11 days after complete unilateral mandibulectomy. Three cats were euthanized at 9, 37, and 68 days postmandibulectomy because of inability to eat. The cat that was euthanized 9 days after bilateral rostral mandibulectomy did not have a feeding tube placed for nutritional support. The cats that survived 37 and 68 days had >50% of the mandible removed, and enteral feeding tubes were placed at the time of mandibulectomy. Another cat treated with resection of >50% of the mandible was euthanized 192 days after surgery for local recurrence; this cat never regained the ability to eat and was maintained with a feeding tube. Notably, three of the six cats with >50% of the mandible removed never regained the ability to eat. Enteral feeding tubes were placed in 17 of the 42 cats. Tubes were placed in five cats that had resection of >50% of the mandible, in 10 cats that had a complete unilateral mandibulectomy, in one cat that underwent a segmental unilateral mandibulectomy, and in one cat that had a bilateral rostral mandibulectomy. Feeding

33 358 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table 4 Progression-Free Rates and Results of Logrank Tests on the Association of Potential Prognostic Variables With Progression-Free Interval of Cats Treated With Mandibulectomy for Oral Neoplasia Progression-Free Rates No. of Variable Cats 1 y (%) 2 y (%) P Value * Tumor type 0.16 Squamous cell carcinoma Fibrosarcoma Osteosarcoma World Health Organization tumor stage 0.96 T T T Location 0.32 Rostral Other Type of mandibulectomy performed 0.12 Complete unilateral Bilateral rostral Surgical margins 0.09 Negative Positive Additional therapy utilized 0.53 Yes No * Statistical significance defined as P tubes were in place for a median of 74 days (range 2 to 192 days). Of 30 owners providing information, 25 (83%) were pleased with the outcome of mandibulectomy, and five (17%) regretted the surgery, citing reasons such as inability of the cat to prehend food and water (n=4), inability to groom (n=2), ptyalism (n=1), rapid tumor recurrence (n=1), and poor quality of life during the postoperative recovery period (n=1). Three cats (one each with squamous cell carcinoma, fibrosarcoma, or lymphoma) were treated with mandibulectomy following local recurrence of their tumor after radiation therapy and chemotherapy (the cats with fibrosarcoma and lymphoma only). The survival times for these three cats were 128 (squamous cell carcinoma), 720 (alive at last follow-up, fibrosarcoma), and 147 (lymphoma) days. Twelve cats received additional treatments for their oral tumors after undergoing mandibulectomy. These adjunctive treatments included chemotherapy (n=5), radiation therapy (n=2), radiation therapy and chemotherapy (n=3), and piroxicam c (n=2). The median progression-free interval and median survival time for cats receiving adjunctive therapy were 280 and 217 days, respectively, compared to 850 days and not reached at 2920 days for cats that received surgery

34 September/October 2006, Vol. 42 Mandibulectomies in Cats 359 alone. There was no significant difference in progressionfree interval or survival time of cats that received additional therapies compared to those treated with mandibulectomy alone (P=0.53 and P=0.44, respectively). Because of the inconsistencies in adjunctive therapies used, no conclusions could be made regarding the effects of these therapies. Discussion The most common reason for performing mandibulectomy in cats is resection of oral neoplasia. Outcomes and owner satisfaction with mandibulectomy for control of canine oral tumors have been well documented A literature search revealed much less information describing the outcome of this surgery in cats. 3-8 Because most feline oral tumors are locally invasive and the amount of tissue that may be resected is small, it was expected that mandibulectomy would not achieve local control of most tumors and that performing a more aggressive mandibulectomy procedure (to improve local control of tumors) would result in permanent loss of oral function and compromised quality of life. Contrary to the hypothesized outcome, tumor control for most of the cats in this study was surprisingly good, with a median progression-free interval of almost 1.5 years. Survival rates at 1 year postmandibulectomy were the same as those at 2 years postmandibulectomy for cats with squamous cell carcinoma, fibrosarcoma, or osteosarcoma suggesting that if a cat lived 1 year, the chance of long-term survival was good. In particular, cats with fibrosarcoma or osteosarcoma enjoyed long progression-free and survival times, and the majority of these cats did not die of tumor-related causes. Despite the small numbers of cats with fibrosarcoma or osteosarcoma in this study, the information regarding their outcome was useful since no studies have been published that specifically describe the treatment for cats with these tumors. Considering the fact that four of six cats with fibrosarcoma had recurrences, compared to only one of six cats with osteosarcoma, fibrosarcoma may be more difficult to control locally than osteosarcoma. A study including more cats with longer follow-up times would be necessary to determine whether this hypothesis is true. Although cats with squamous cell carcinoma had significantly shorter survival times than those with fibrosarcoma or osteosarcoma, superior tumor control and survival times were detected in this study when compared to those reported previously for cats with squamous cell carcinoma treated with mandibulectomy. 3-5 These results may have been a consequence of bias in case selection or perhaps improved patient management, better methods of imaging tumors, surgical techniques, and postoperative care. Although the study reported here did not find improved survival times for cats that received adjunctive therapy, the use of additional therapies in some cats with squamous cell carcinomas may have improved their tumor control and survival. The longest survival time reported to date for cats with oral squamous cell carcinoma was in a study of seven cats treated with mandibulectomy and adjunctive definitive radiation therapy, resulting in a median survival time of 14 months. 7 This small study suggests that curative-intent radiation therapy might extend the survival times of cats undergoing mandibulectomy for squamous cell carcinoma. 7 Because adjunctive treatments varied in nature and were applied inconsistently in the study reported here, conclusions regarding their efficacy were not possible. Unfortunately, because of the small numbers of cats in each variable category, this study was not able to identify prognostic indicators for cats with oral squamous cell carcinoma. The significant association of squamous cell carcinoma with survival time but not progression-free interval was likely the result of the small number of cases and the large amount of censored data in this study, because not all cats died of their oral tumors. Forty-six percent of the treated cats that survived >2 weeks following surgery developed local recurrences, and it is possible that the recurrence rate might have been higher if longer follow-up data were available. This high recurrence rate suggested that the extent of disease was underestimated at the time of surgery and that it was often impossible to obtain adequate margins of normal tissue with mandibulectomy. Consequently, careful case selection and surgical planning that includes advanced imaging may be essential to achieve complete excision of the tumor. Concern regarding the potential for surgical complications that result in impaired oral function and negatively impact the quality of life of affected cats was a major impetus for the initiation of this study. During the 4 weeks following surgery, almost all cats experienced adverse effects. The description of adverse effects in this study likely underestimated the number of cats experiencing side effects, because it was limited to information in the medical records and the recollections of owners and referring veterinarians. Because owners were asked to describe outcomes retrospectively, their recollections may have been biased and may have resulted in decreased reporting of morbidity. In addition, evaluation of surgical morbidity may have been confounded by the use of additional therapies in some cats. Despite these limitations, this study demonstrated considerable acute postoperative morbidity. The nature of the adverse effects observed suggested that hospitalization, analgesics, intravenous fluids, and enteral feeding with a feeding tube may improve the quality of life for cats immediately following mandibulectomy. These adverse effects should be discussed thoroughly with cat owners prior to mandibulectomy. For example, cats that develop mandibular drift and resulting malocclusion may require contralateral mandibular canine tooth extraction or vital pulpotomy to prevent damage of the hard palate. Cats that develop ptyalism or tongue protrusion following surgery often require increased efforts by the owners to keep the area clean and to provide some generalized grooming for the cat. It was surprising that although 76% of the cats of this study experienced adverse effects for the remainder of their lives, 25 (83%) owners were pleased with their cat s quality of life and would choose the procedure again under the

35 360 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 same circumstances. This level of acceptance suggested that the long-term complications did not greatly impact the quality of life for the cats. There were, however, cats that experienced life-threatening morbidity. Five cats never regained the ability to prehend food, and three of these cats had resection of >50% of the mandible. Although the number of cats undergoing resection of >50% of the mandible was too small to allow detection of a difference in survival time, the fact that half of these cats never regained the ability to prehend food suggested that the surgery was too aggressive to allow a functional outcome. Conclusions based on this study were limited by its retrospective and multi-institutional nature. The ability to detect differences in progression-free interval or survival time as they related to selected variables was limited by the small number of cats in each group and the large number of censored individuals. A larger, prospective study with uniform treatments and follow-up times might be better able to define prognostic indicators for cats undergoing mandibulectomy for oral tumors. Possible positive prognostic indicators meriting further investigation based on the results of the current study include a rostral location and complete excision of the tumor. Finally, both the description of adverse effects and the analysis of survival were potentially confounded by the limited and inconsistent application of various adjunctive therapies. To increase the number of cats described, this study included 12 cats treated with additional therapies. It was possible that the additional therapies influenced survival and/or complication rates. Cats receiving additional treatments had lower progression-free and survival rates, possibly owing to a selection bias from additional therapies being recommended more commonly in cats suspected to have a poorer prognosis. Despite these limitations, there were no significant differences in the progression-free interval or survival time of cats receiving additional treatments versus cats treated with mandibulectomy alone, and inclusion of these cats provided valuable additional information. Conclusion This retrospective study describes the outcomes for 42 cats treated with mandibulectomy for oral tumors. Results of the study suggest that mandibulectomy provides reasonable local control and an adequate quality of life for some cats with oral tumors. Although the majority of cats experienced complications from the surgery, most owners reported satisfaction with outcome. A diagnosis of oral squamous cell carcinoma was associated with a poorer prognosis compared to fibrosarcoma and osteosarcoma. Further prospective studies incorporating uniform treatments and prolonged follow-up times are required to better define prognostic indicators for cats undergoing mandibulectomy for oral neoplasia. Acknowledgment The authors thank Antony S. Moore, MVSc, Diplomate ACVIM (Oncology) for his assistance with this manuscript. References 11. Moore AS, Ogilvie GK. Tumors of the alimentary tract. In: Ogilvie GK, Moore AS, eds. Feline Oncology: A Comprehensive Guide to Compassionate Care. Trenton: Veterinary Learning Systems, 2002: Withrow SJ. Cancer of the gastrointestinal tract. In: Withrow SJ, MacEwen EG, eds. Small Animal Clinical Oncology. 3rd ed. Philadelphia: WB Saunders, 2001: Bradley RL, MacEwen EG, Loar AS. Mandibular resection for removal of oral tumors in 30 dogs and six cats. J Am Vet Med Assoc 1984;184: Bradley RL, Sponenberg DP, Martin RA. Oral neoplasia in 15 dogs and 4 cats. Semin Vet Med Surg (Small Anim) 1986;1: Penwick RC, Nunamaker DM. Rostral mandibulectomy: a treatment for oral neoplasia in the dog and cat. J Am Anim Hosp Assoc 1987;23: Kapatkin AS, Patnaik AK, Burk RL, et al. Mandibular swellings in cats: prospective study of 24 cats. J Am Anim Hosp Assoc 1991;27: Hutson CA, Willauer CC, Walder EJ, et al. Treatment of mandibular squamous cell carcinoma in cats by use of mandibulectomy and radiotherapy: seven cases ( ). J Am Vet Med Assoc 1992;201: Dernell WS, Hullinger GH. Surgical management of ameloblastic fibroma in the cat. J Small Anim Pract 1994;35: Owen LM. TNM classification of tumours in domestic animals. Geneva: World Health Organization, Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53: Le CT. Applied Survival Analysis. New York: John Wiley and Sons, 1997: Fox LE, Geoghegan SL, Davis LH, et al. Owner satisfaction with partial mandibulectomy or maxillectomy for treatment of oral tumors in 27 dogs. J Am Anim Hosp Assoc 1997;33: Salisbury SK, Lantz GC. Long-term results of partial mandibulectomy for treatment of oral tumors in 30 dogs. J Am Anim Hosp Assoc 1988;24: White RAS. Mandibulectomy and maxillectomy in the dog: longterm survival in 100 cases. J Small Anim Pract 1991;32: Kosovsky JK, Matthiesen DT, Marretta SM, et al. Results of partial mandibulectomy for the treatment of oral tumors in 142 dogs. Vet Surg 1991;20: Schwarz PD, Withrow SJ, Curtis CR, et al. Mandibular resection as a treatment for oral cancer in 81 dogs. J Am Anim Hosp Assoc 1991;27: a Prism 4; GraphPad Software, Inc., San Diego, CA b SAS Version 8.2; SAS Institute, Inc., Cary, NC c Feldene; Pfizer, Inc., New York, NY 10017

36 The Use of Complementary and Alternative Therapies in Dogs and Cats With Cancer The use of complementary and alternative medical therapies is becoming widespread. The objective of this study was to examine the use of complementary and alternative therapies in dogs and cats with cancer. The types of modalities used, the intended purpose for each modality, sources of information pet owners used, and the level of interest in these modalities were all evaluated. Information was obtained by written survey, and 254 owners agreed to participate. Complementary and alternative therapy use was commonplace, with 76% of surveyed owners reporting some use. When prayer for health reasons was excluded, the number of owners reporting use dropped to 65%. Nutritional supplements were the most commonly used therapy. J Am Anim Hosp Assoc 2006;42: Susan E. Lana, DVM, MS, Diplomate ACVIM Lori R. Kogan, PhD Ken A. Crump, AHT J. Terry Graham Narda G. Robinson, DO, DVM O From the Animal Cancer Center, Department of Clinical Sciences, College of Veterinary Medicine and Biomedical Sciences, Colorado State University, 300 West Drake, Fort Collins, Colorado Introduction The use of complementary and alternative medical therapies is becoming widespread. In a recent study done by the National Center for Complementary and Alternative Medicine and the National Center for Health Statistics, 31,044 adults in the United States were surveyed. 1 This study found that 75% of respondents had used complementary and alternative medicines at some point in their lives, and 62% had used them in the past 12 months. 1 When prayer for health reasons was excluded, 50% of respondents had used these therapies at some time, and 36% had used them in the last year. 1 Richardson et al. reported that 83% of 453 human cancer patients surveyed had used at least one complementary or alternative approach. 2 When spiritual approaches and psychotherapy were excluded, the percentage dropped to 63%. 2 In a report assessing women treated for breast and gynecological cancers, it was found that 48% used some type of herbal or vitamin supplement, yet only 53% informed their physician about the usage. 3 In animals, the use of complementary and alternative therapies has been reported in many species, and their use has been associated with a variety of disease states and conditions, including musculoskeletal, neurological, cardiovascular, dermatological, endocrine metabolic, and behavioral disorders. 4-9 Because this area has undergone rapid change in recent years and its popularity has increased, concern has arisen over the possible interactions of complementary and alternative therapies with traditional treatments Herbal products, in particular, may alter the activity of drug-metabolizing enzymes (e.g., cytochrome p450) or bind to drug transporter proteins (e.g., p-glycoprotein), potentially interfering with the pharmacokinetics of a chemotherapeutic drug and resulting in increased or decreased drug levels or alterations in rates of drug clearance. 10 In some instances, an herb may cause direct toxicity or interfere with laboratory analysis of other compounds To avoid unwanted interactions that negate treatment efficacy or increase unwanted side effects, it is important to know if these therapies are being used concurrently with traditional therapies. Although the true prevalence of use is JOURNAL of the American Animal Hospital Association 361

37 362 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 unknown, it is often assumed that the use of complementary and alternative therapies in dogs and cats mirrors that of humans. The purpose of this study was to examine the use of complementary and alternative therapies in a population of dogs and cats with cancer. Specifically, the types of modalities used, their intended purpose, the sources of information pet owners used to learn about these modalities, and the level of interest in such therapies were evaluated. Materials and Methods Study Subjects Pet owners who presented their animal to a university oncology referral center between August 2004 and January 2005 were invited to participate in the study. The study was approved by the institutional review board for the use of human subjects (i.e., Regulatory Compliance Office, Human Research Committee). While clients waited for their pets appointment, they were asked by a designated research assistant if they were willing to participate in the survey. They were informed of the purpose of the study, told they would remain anonymous, and were given the option of declining participation. Owners were also informed that they could choose to stop and not complete the survey at any time. A cover letter stating the same information was also provided with each survey. If verbal consent was obtained, clients were given the survey to complete. The research assistant remained with the client, and the survey was returned to the assistant when completed. The individual who administered the survey was not involved with the pet s veterinary care or treatment. Clients were given the survey only once and were instructed to provide information relating only to the pet they had brought with them on that day. Owners were surveyed at all types of visits (e.g., initial visit, follow-up visit), and owners of both dogs and cats were eligible to participate. Survey Instrument The survey instrument was five pages in length and was intended to take no more than 15 minutes to complete. It was divided into three sections, with 13 questions total. Space for comments was provided after each question. Information was obtained concerning demographic data of the owner (i.e., age, gender, level of education), types of complementary and alternative modalities used, length of use, sources of information about the modality, and intended purpose. The therapies were classified into 11 major categories: herbs or botanicals, nutritional supplements, vitamins, special diets, acupuncture, massage, chiropractic, homeopathy, Reiki or healing touch, prayer, and flower essences. Participants were asked to select from the following sources of information about the modalities they were using: Internet, book or magazine, veterinarian, friend or family member, or other. Participants were also asked to select from the following intended purposes of the modalities chosen: improve general well-being, reduce pain, cure cancer, improve immune function, reduce treatment toxicity/detoxify, improve appetite, or other reasons. If other was the response chosen, participants were asked to give a specific answer in the comment field. More than one response could be given for each question. For example, more than one modality or reason for usage could be noted. Participants were asked if their regular veterinarian supported the use of complementary and alternative therapies and if they had talked to their veterinarian about the use of such therapies. A final question was designed to determine the level of interest each participant had in these therapies and included the following response options: no interest, average interest, or strong interest. Statistical Analysis Analysis was done with a statistical software package. a Frequencies for each variable were calculated. Statistically significant differences between owners who reported using complementary and alternative therapies and those who did not, based on gender and education level (e.g., high school or Graduate Equivalency Diploma; some college [i.e., <2 years]; 2-year technical schooling; 4-year college degree; advanced or postgraduate degree), were assessed through the use of chi-square. Differences based on age were assessed through the use of unpaired t-tests. Significance level was set at P<0.05. Results Two-hundred-fifty-four surveys were completed. Demographic data of the respondents are provided in the Table. There were more female than male respondents and more than half (n=163 or 64%) had a 4-year college or advanced degree. Complementary and alternative therapies were used by 199 (76%) owners. This number dropped to 164 (65%) when prayer was excluded. Based on the 11 categories of treatment modalities previously described, their usage is summarized in Figure 1. The most commonly used modality was nutritional supplements, with 102 (40%) of the respondents reporting their use. The length of time that owners used each modality ranged from one time or sporadic use to 60 months of continuous use. The resources participants used for information on these modalities varied, with many (46%) owners listing their veterinarian as a resource. This was true for seven categories of modalities, with the exception of massage, healing touch, flower essences, and prayer. For these latter categories, family and friends were the predominant resources. Other resources listed by survey participants included school, breeder, training classes, dog shows, alternative health care provider, health food stores, and television [Figure 2]. The intended purpose of complementary and alternative therapies was varied [Figure 3]. The most common reasons for using these therapies were to improve general wellbeing (34%) and improve immune function (22%). When the specific modalities were considered separately, 10 modalities were most often selected in an effort to improve general well-being. Chiropractic procedures were usually chosen to reduce pain. When asked if their regular veterinarian supported the use of complementary and alternative therapies, 163 (64%)

38 September/October 2006, Vol. 42 Complementary Therapies 363 Table Demographic Data of 254 Owners Surveyed on the Use of Complementary and Alternative Therapies Age Years Median (range) 46 (20 to 81) Mean ± standard deviation 46±12 Gender Number (%) Male 88 (35) Female 166 (65) Education Level of Respondent Number (%) Figure 2 Most common sources of information about complementary and alternative therapies. See text for details on the other category (total respondents=199). High school or Graduate 20 (8) Equivalency Diploma Some college 38 (15) 2-y technical schooling 21 (9) 4-y college degree 86 (35) Advanced college degree 77 (32) Not specified 3 (1) Figure 3 Intended purposes of complementary and alternative therapies. See text for details on the other category (total respondents=199). Figure 1 Eleven categories of complementary and alternative treatment modalities used, in order of decreasing frequency (total respondents=199). owners believed they did, 23 (9%) respondents said no, and 50 (20%) respondents did not know their veterinarians views. Eighteen (7%) respondents did not provide an answer. When asked if they talked to their veterinarian about complementary and alternative therapy use, 89 (35%) owners said yes, 144 (57%) respondents said no, and 21 (8%) did not respond. When asked about level of interest, 141 (57%) owners indicated strong interest, 100 (40%) respondents indicated an average interest, and seven (3%) respondents reported no interest in complementary and alternative therapies. When differences based on the age, gender, or education level of owners were compared to usage, no statistically

39 364 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 significant differences were found. With respect to level of interest, gender of the owner had a significant influence (P=0.003), with female participants more likely to report a strong interest (n=161; 64%) than male participants (n=38; 44%). Additionally, even though 24% of those surveyed reported no use, an overwhelming number (97%) reported at least some interest in these modalities. Discussion The study reported here sought to determine the frequency of use of complementary and alternative therapies in a population of dogs and cats diagnosed with cancer and to further define the type and reasons such modalities were chosen. The high usage rate (76%) was similar to that reported in humans in the United States. 1 Demographically, most human studies involving cancer patients or the general public have found that the individuals who most often seek complementary and alternative therapies are female, better educated, of a higher socioeconomic status, and of a younger age. 14 In the survey population of the present study, no differences were found with respect to the age, gender, or education level of the respondents and frequency of use of these modalities. It is possible that the survey participants were skewed with regard to demographics, because no attempt was made to control the number or type of participants in each demographic category. Demographic data for nonsurveyed pet owners who visited the veterinary teaching hospital during the same time period were not maintained and were not available for comparison. The only difference based on demographic data that was found in the current study was the level of interest in complementary and alternative therapies. More female respondents reported a strong interest than male respondents. This finding mirrored the trend seen in humans. 3 The most common modalities used included nutritional supplements, prayer, diet, and vitamins. When prayer was excluded, the rest of the above modalities accounted for 51% of the responses chosen. These modalities are commonly classified as biological interventions and were often chosen, administered, and controlled by the pet owner. In contrast, the least used modalities were acupuncture and chiropractic procedures. Only 5% of all owners used these modalities, and their low frequency may have been affected by their requirement for a third party or practitioner to provide the services. Veterinarians were listed as the most common resource for information about complementary and alternative modalities. This finding emphasized the opportunity veterinarians have to appropriately educate pet owners concerning the potential pitfalls and benefits of these treatments. Simply providing owners with sources of reliable information would likely be beneficial. The reason for use or intended purpose of each modality was quite varied. It was interesting to note, however, that the most common reason was to improve general well-being and not as a treatment for a specific problem or as a cure for cancer. When participants were asked about their regular veterinarians support of the use of complementary and alternative therapies, most (64%) respondents believed the veterinarians supported their use. However, when asked if they had talked to their veterinarian about the subject, the majority (58%) had not done so. This pattern was similar to reports in humans, where many complementary and alternative therapies have not been disclosed to the physician. 3,16 In some cases, patients were not asked, or the information was not recorded in the medical record. 9,16 In a survey of 165 primary care physicians, only 58% always or often asked their patients about complementary therapy use, and most physicians estimated that <15% of their patients used these modalities. 17 It was concluded that physicians underestimated the rate of use by their patients, and it was suggested that many patients do not report such use to their physicians. 17 These findings reiterate the importance of good communication with pet owners, because some owners may not think these therapies are medically important, and some may be concerned about potential veterinarian objections to their use. This survey, which was intended to provide some general information concerning complementary and alternative therapies in dogs and cats with cancer, raised many additional questions. Additional socioeconomic information (e.g., ethnicity, income) regarding owners was not sought, in an attempt to increase survey participation. No detailed information was obtained concerning the type of supplements used or doses of any particular vitamin or herb. The definition of what constitutes a supplement or botanical was not specifically given, although survey participants were free to ask questions if they wished. Such information might prove useful, especially in animals that have cancer and are receiving more traditional medications. Reports have been published about certain herbs and botanicals and their possible interactions with chemotherapeutic agents and drug metabolism. 10 Recently, the Clinical Center of the National Institutes of Health set new policies concerning supplement use in humans enrolled in clinical trials. All patients are to be screened, and all therapies are to be authorized by the attending physician. 15 This policy does not recommend discontinuation in most cases, but instead stresses the importance of open communication and disclosure of what patients are taking, in order to avoid unwanted drug interactions. 15 The survey reported here was limited to owners of pets seen by a clinical oncology service. To ensure anonymity, no attempt was made to link survey participant responses with animal information (e.g., type of cancer, stage of disease, visit type). It may be beneficial to further define the use of complementary and alternative therapies based on these parameters. Navo et al. reported that disease status can have a significant influence on the use of complementary and alternative therapies in women with gynecological cancer. 3 In that study, newly diagnosed patients were less likely to use these modalities than those with relapses or those currently in remission. 3 Conclusion This study found that the use of complementary and alternative therapies in a group of dogs and cats with cancer was

40 September/October 2006, Vol. 42 Complementary Therapies 365 high and paralleled that previously reported for humans. Pet owners turned to their veterinarians most often for information about these modalities. Additional studies are necessary to better define what effect these modalities might have or what interactions there may be with traditional medicines. Expanded studies that include a variety of animal species, disease states, owner demographics, and geographic areas are warranted to further define the use of complementary and alternative therapies in the general animal population as a whole. a SPSS 13.0 for Windows; Apache Software Foundation, Chicago, IL Acknowledgment The authors acknowledge Carrie Griesedieck, CVT, for her help with survey administration. References 11. Barnes P, Powell-Griner E, McFann K, et al. Centers for Disease Control Advance Data Report #343. Complementary and alternative medicine use among adults: United States, ;May 27: Richardson AM, Sanders T, Palmer JL, et al. Complementary/alternative medicine use in a comprehensive cancer center and the implications for oncology. J Clin Oncol 2000;18: Navo MA, Phan J, Vaughn C, et al. An assessment of the utilization of complementary and alternative medication in women with gynecologic or breast malignancies. J Clin Oncol 2004;22: Lokin T. Alternative therapy of animals - homeopathy and other alternative methods of therapy. Act Vet Scand Suppl 2001;95: Boldt E Jr. Use of complementary veterinary medicine in the geriatric horse. Vet Clin North Am Equine Pract 2002;18: Center SA. Metabolic, antioxidant, nutraceutical, probiotic, and herbal therapies relating to the management of hepatobiliary disorders. Vet Clin North Am Small Anim Pract 2004;34: Hawks D. Alternative medicine: musculoskeletal system. Clin Tech Small Anim Pract 2002;17: Berschneider HM. Complementary and alternative veterinary medicine and gastrointestinal disease. Clin Tech Small Anim Pract 2002;17: Wynn SG, Chalmers S. Alternative therapies for pruritic skin disorders. Clin Tech Small Anim Pract 2002;17: Sparreboom A, Cox MC, Acharya MR, et al. Herbal remedies in the United States: potential adverse interactions with anticancer agents. J Clin Oncol 2004;22: Cassileth BR. Evaluating complementary and alternative therapies for cancer patients. CA Cancer J Clin 1999;49: Norred CL, Brinker F. Potential coagulation effects of preoperative complementary and alternative medicines. Altern Ther Health Med 2001;7: Dasgupta A. Review of abnormal laboratory test results and toxic effects due to use of herbal medicines. Am J Clin Pathol 2003;120: Cassileth BR, Deng G. Complementary and alternative therapies for cancer. Oncologist 2004;9: Sparber A, Ford D, Kvochak PA. National Institutes of Health s clinical center sets new policy on use of herbal and other alternative supplements by patients enrolled in clinical trials. Cancer Invest 2004;22: Cohen RJ, Kirsten E, Pan CX. Complementary and alternative medicine use by older adults: a comparison of self-report and physicians chart documentation. J Gerontol 2002;57A:M223-M Giveon SM, Liberman N, Klang S, et al. A survey of primary care physicians perceptions of their patients use of complementary medicine. Complement Ther Med 2003;11:

41 A Combination of Oral Cabergoline and Double Cloprostenol Injections to Produce Third-Quarter Gestation Termination in the Bitch To assess the efficacy and safety of a combined cabergoline and cloprostenol protocol to terminate third-quarter pregnancy, 22 pregnant bitches that ranged from 35 to 45 days after mating were randomly assigned to a treatment group (n=13) or to an untreated control group (n=9). The animals were monitored for 12 days, and pregnancy termination was confirmed by ultrasound examination. Twelve of the 13 treated bitches aborted within 9 days of the initiation of treatment (mean 4.6 days). Only mild side effects were observed. The control animals had normal gestational courses, as did the bitch that did not respond to the therapy. This combination of drugs appeared to be a practical, safe, and efficient abortifacient when used in third-quarter pregnancies. J Am Anim Hosp Assoc 2006;42: Yanina Corrada, DVM Raúl Rodríguez, MV Mariana Tortora, MV Daniel Arias, MV Cristina Gobello, DVM, Diplomate ECAR O From the Imaging Diagnosis Service (Corrada, Rodríguez, Tortora, Arias) and Small Animal Clinic (Gobello), Faculty of Veterinary Medicine, National University of La Plata, La Plata, Argentina. Address all correspondence to Dr. Gobello. Introduction Nonsurgical termination of unwanted pregnancy is one of the major reproductive difficulties in canine practice. While ovariohysterectomy is the solution in many cases, nonsurgical termination can be difficult because of the unavailability of some abortifacients in certain countries and the side effects associated with some of the more commonly used drugs. Such side effects may require the dogs to be hospitalized for several days, with the resultant significant expense. 1 Prostaglandin F 2 α (PGF) analogs and dopaminergic agonists, as well as their combinations, have been used successfully for pregnancy termination in the dog A synergistic effect may explain how dopaminergic agonists and PGF decrease serum progesterone, the essential hormone for pregnancy maintenance. 9 Prostaglandin F 2 α analogs have a direct luteolytic action, while dopaminergic agonists act indirectly on the corpus luteum by withdrawing the main luteotropic hormone (i.e., prolactin) during the second half of the luteal phase. 2,12-14 Cabergoline and cloprostenol are among the most potent representatives of the dopaminergic agonists and PGFs, respectively. 12,14,15 As a result, the dosages required of these drugs can be reduced significantly in relation to other compounds in the same pharmacological groups, thus diminishing their side effects. As a general rule, the action of combined abortifacient drugs is facilitated (i.e., is more rapid) by advanced gestational age. Protocols initiated later in gestation usually require a shorter treatment period, although they may result in abortion of recognizable fetuses, which some clients may find objectionable. 5,15 In clinical practice, rapidity of effect should be weighed against the abortion of developed fetuses. Third-quarter pregnancy termination (i.e., days 35 to 45) appears to be a good compromise that produces relatively rapid pregnancy termination in a fashion that is acceptable to the owner. Another practical aspect to consider when selecting a pharmacological protocol for pregnancy termination is the number of parenteral applications 366 JOURNAL of the American Animal Hospital Association

42 September/October 2006, Vol. 42 Gestation Termination 367 required. Numerous injections may be objectionable or may cause compliance problems, resulting in failure. The purpose of this study was to assess the efficacy and safety of a combined protocol of cabergoline per os [PO] and two subcutaneous [SC] injections of cloprostenol to terminate pregnancy at 35 to 45 days in bitches. Materials and Methods Study Animals Twenty-two, healthy, mixed-breed and pure-bred bitches that had been referred for pregnancy termination were included in this trial. The dogs ranged in age from 1 to 10 years and in weight from 10 to 45 kg. Pregnancy was confirmed in all the animals using transabdominal ultrasonography. 16 The bitches remained in their normal surroundings at home during the study, except at the times of injections and clinical examinations. Written, informed consent was obtained from all owners. Pharmacological Protocol Between days 35 and 45 from the first mating, the bitches were allocated randomly to either a treatment or control group. 17 Dogs in the treatment group (n=13) received cabergoline a (5 µg/kg PO q 24 hours) for 7 days in food, and cloprostenol b (1 µg/kg SC after tenfold dilution in physiological saline) on days 1 and 3, given at least 8 hours after food. If pregnancy was not terminated by day 8, the same dose of cabergoline was continued until day 12. Dogs in the second group were not treated in any way and were used as controls (n=9). Clinical Follow-up All the dogs were clinically monitored for 12 days after initiation of the trial. Clinical examinations included evaluation of behavioral changes, body temperature, vulvar discharge, evidence of abortion, and appearance of local or systemic side effects. Clinical suspicion of pregnancy termination was considered when vaginal discharge or prepartum behavior developed, body temperature decreased by >1 C (33.8 F), or fetuses were expelled. Owners were encouraged to observe their dogs as often as possible and to record when and if abortion occurred during the study period. Ultrasonographic Monitoring Within 8 hours of the owner suspecting that pregnancy termination had occurred, bitches were monitored ultrasonographically to confirm that termination had taken place. c,16 Success was defined as uterine vacuity. 5,9,17 If no clinical signs of termination were visible externally, ultrasound examination was repeated on day 8. If no evidence of termination was found, ultrasonography was repeated again on day 12. Statistical Analysis The frequency of pregnancy termination in bitches was analyzed by PROC FREC (i.e., chi-square). d Days to pregnancy termination were characterized by PROC MEANS (i.e., descriptive statistics), and results were expressed as least square means ± standard error of the mean (LSM ± SEM). d A statistically significant P value was set at <0.05. Results Significant differences were found between the two groups of dogs. In 12 of the 13 treated bitches, pregnancy termination was confirmed ultrasonographically within 4.6±0.7 days (range 2 to 9 days) after initiation of treatment. For 11 dogs, the entire abortion probably occurred in <8 hours, except for one animal (case no. 5) in which abortion occurred on 2 days, 48 hours apart. This animal was excluded from the descriptive statistics, because treatment response was different from the rest of the group. Seven of the aborting bitches exhibited preparturient-like behavior (e.g., anxiety and nesting) and had a decrease in body temperature before termination. In eight of these 12 dogs, neither the abortion nor the expulsed fetuses were observed by the owners. The nonresponsive bitch (case no. 11) and all the control animals developed no clinical signs compatible with pregnancy termination, and ultrasound examinations on days 8 and 12 revealed normal gestations [see Table]. Side effects were mild, occurred in only three of the treated bitches, and included vomiting, nausea, retching, and panting. Signs were observed 10 to 15 minutes after the cloprostenol injections and lasted for approximately 15 to 20 minutes. Abortion was followed by 5 to 8 days of mucoid sanguineous vulvar discharge in all the dogs. Slight mammary enlargement (n=6) and milk secretion (n=4) were also noted in some aborting bitches. Discussion In designing the present protocol, the availability of the drugs, the practicality of administration, efficacy, and safety were all considered. Both cabergoline and cloprostenol are available either in the veterinary and/or human pharmaceutical market in many countries. This combined protocol was easily administered in an outpatient setting. In the present trial, nine (69%) of 13 treated bitches required a total of three visits (i.e., first pregnancy diagnosis and initial medication, PGF administration, and abortion confirmation) to the clinic. For the other three responding animals, only one additional visit was necessary. Although in combined protocols, gestational age is known to affect the number of days treatment is required, in most clinical settings the duration of pregnancy is estimated from the date of mating. 5,15 In the present trial, the date of mating was used rather than an actual endocrine-proven gestational age. In cases where the owners do not know the breeding date, gestation duration may be calculated by an ultrasound examination. 16 In a previous study, a similar protocol was used, but cloprostenol was given on days 1 and 5 starting 25 days after mating. 9 In that study, resorption of fetuses was achieved in four of five treated bitches at 5 to 12 days after the initiation of the treatment. 9 In the present trial, a higher success rate (92.3% versus 80%) and a more rapid effect were obtained. These differences may simply be explained by the later onset of the pharmacological treatments in the present trial.

43 368 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table Clinical Data on 13 Bitches Treated for Pregnancy Termination and Nine Control Bitches Days From Duration of Case Age Weight First Treatment Cabergoline Pregnancy No. Breed (y) (kg) Mating Protocol * Treatment (d) Termination Abortion 1 Boxer Cabergoline + cloprostenol 7 Yes Yes 2 German shepherd Cabergoline + cloprostenol 5 Yes Yes dog 3 Argentine boar Cabergoline + cloprostenol 9 Yes Yes hound 4 Siberian husky Cabergoline + cloprostenol 5 Yes Yes 5 Mixed-breed dog Cabergoline + cloprostenol 5 Yes Yes 6 Argentine boar Cabergoline + cloprostenol 3 Yes Yes hound 7 Argentine boar Cabergoline + cloprostenol 5 Yes Yes hound 8 German shepherd Cabergoline + cloprostenol 2 Yes Yes dog 9 Labrador retriever Cabergoline + cloprostenol 2 Yes Yes 10 Boxer Cabergoline + cloprostenol 3 Yes Yes 11 Mixed-breed dog Cabergoline + cloprostenol 12 No No 12 English cocker Cabergoline + cloprostenol 3 Yes Yes spaniel (Continued on next page)

44 September/October 2006, Vol. 42 Gestation Termination 369 Table (cont d) Clinical Data on 13 Bitches Treated for Pregnancy Termination and Nine Control Bitches Days From Duration of Case Age Weight First Treatment Cabergoline Pregnancy No. Breed (y) (kg) Mating Protocol * Treatment (d) Termination Abortion 13 English bulldog Cabergoline + cloprostenol 7 Yes Yes 14 Mixed-breed dog Control NA No No 15 Labrador retriever Control NA No No 16 Boxer Control NA No No 17 Mixed-breed dog Control NA No No 18 Basset hound Control NA No No 19 Mixed-breed dog Control NA No No 20 Collie Control NA No No 21 Argentine boar Control NA No No hound 22 English cocker Control NA No No spaniel * Dosages: cabergoline 5 µg/kg per os q 24 h and cloprostenol 1 µg/kg subcutaneously on d 1 and 3. NA=not applicable

45 370 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Considering that resorption is usually better accepted than abortion by clients, owners should be fully informed beforehand about the abortion process and advised about the number of potential clinic visits required. The actual abortion period was probably short, as most of the owners did not observe the abortion or the expulsed fetuses. It is also possible the bitches ingested the expelled material. In the present trial, the proportion of failures was low, but the exact cause of these failures remains difficult to explain. Prolongation of dopaminergic agonist and use of PGF protocols until the desired effect is achieved have been previously recommended and may have altered the outcome of this study. 5 Although rare, split abortions (as occurred in case no. 5) have been previously reported using this combination of drugs. 5 Measuring daily progesterone serum concentrations in the future may reveal the reason for prolongation of the process in some cases. Demonstration of a split abortion emphasized that treatment and evaluation must continue until abortion is deemed complete by ultrasonographic evaluation. Educating the owners about the clinical signs to look for and the ultrasound evaluations is crucial for rapid success of confirmation of pregnancy termination and avoidance of unnecessary prolongation of treatment. Similar to earlier studies, no side effects or only minor side effects were associated with the low dose of PGF used in this study. The fact that the potent PGF cloprostenol was administered at least 8 hours after feeding may have lessened the severity of the side effects. 5,9 The mammary enlargement and lactation that was induced in some of the aborting animals may have occurred from a decline in progesterone-related negative feedback following pharmacological luteolysis, which subsequently resulted in an elevation of serum prolactin. 18 Conclusion Pregnancies were terminated in 12 of 13 bitches treated with a combination of oral cabergoline and two cloprostenol injections, 35 to 45 days after mating. This combination appeared to be a practical, safe, and efficient abortifacient protocol and was suitable for use in an outpatient setting. Future evaluations of this combined protocol in a larger number of dogs are warranted. a Galastop; Ceva-Santé Animate, Libourne, France b Estrumate; Schering Plough, Buenos Aries, Argentina c Toshiba Core Visión Pro; Shimoishigami, Otawara-Shi, Tochigi-Ken, Japan d Statistical Analysis System (SAS); SAS Institute, Cary, NC Acknowledgments The authors thank Ceva-Santé Animate in Libourne, France, for supplying the Galastop, and Schering Plough in Buenos Aries, Argentina, for supplying the Estrumate for the study. References 11. Wanke MM, Romagnoli S, Verstegen J, et al. Pharmacological approaches to pregnancy termination in dogs and cats including the use of prostaglandins, dopamine agonists, and dexamethasone. In: Concannon PW, England G, Verstegen J, et al., eds. Recent Advances in Small Animal Reproduction. Ithaca NY: International Veterinary Information Service ( 2002:A Concannon PW, Hansel W. Prostaglandins F2α-induced luteolysis, hypothermia and abortion in beagle bitches. Prostaglandins 1977;13: Fieni F, Dumon C, Tainturier D, et al. Clinical protocol for pregnancy termination in bitches using prostaglandin F2alpha. J Reprod Fertil 1987;51(Suppl): Fieni F, Fuhrer M, Tainturier D, et al. Use of cloprostenol for pregnancy termination in dogs. J Reprod Fertil 1989;39(Suppl): Gobello C, Castex G, Corrada Y, et al. A comparative study of two prostaglandins and bromocriptine for pregnancy termination in bitches. J Am Vet Med Assoc 2002;220: Jochle W, Arbeiter K, Post K, et al. Effects of pseudopregnancy, pregnancy and interestrus intervals of pharmacological suppression of prolactin secretion in female dogs and cats. J Reprod Fertil 1989;39: Onclin K, Silva LDM, Verstegen JP. Termination of unwanted pregnancy in dogs with the dopamine agonist, cabergoline, in combination with a synthetic analog of PGF2-alpha, either cloprostenol or alphaprostol. Theriogenology 1995;43: Onclin K, Verstegen JP. Practical use of a combination of a dopamine agonist and a synthetic prostaglandin analogue to terminate unwanted pregnancy in dogs. J Small Anim Pract 1996;7: Onclin K, Verstegen JP. Comparison of different combinations of analogues of PGF2α and dopamine agonists for the termination of pregnancy in dogs. Vet Rec 1999;144: Post K, Evans LE, Jochle W. Effects of prolactin suppression with cabergoline on the pregnancy of the bitch. Theriogenology 1988;29: Feldman EC, Davidson AP, Nelson RW, et al. Prostaglandin induction of abortion in pregnant bitches after misalliance. J Am Vet Med Assoc 1993;202: Ferney J. Pharmacologie des prostagladines. In: Prostaglandines et Gestation de la Reproduction chez la Vache. Paris: Coopers Veterinaire, 1985: Okkens AC, Bevers MM, Dieleman S, et al. Evidence of prolactin as the main luteotrophic factor in the cycling dog. Vet Quatrly 1990;12: Fieni F, Verstegen J, Herand V, et al. Phisiologie de la prolactine pharmacologie des antiprolactiniques et applications chez la chienne. Prat Med Chir Anim Comp 1999;34: Gobello C. Dopamine agonists, anti-progestins, anti-estrogens, antiandrogens and GnRH agonists on canine reproduction. Theriogenology 2006(Suppl);65:in press. 16. Mattoon JS, Nyland TG. Ultrasonography of the genital system. In: Nyland TG, Mattoon JS, eds. Veterinary Diagnostic Ultrasound. Philadelphia: WB Saunders, 1995: England GCW, Russo M. Ultrasonographic characteristics of early pregnancy failure in bitches. In: Proceedings, 5th Internat Symp Canine Feline Repro, San Pablo, Brazil, 2004: Gobello C, Dela sota L, Castex G, et al. Diestrous ovariectomy: a model to study the role of progesterone in the onset of canine pseudopregnancy. J Reprod Fertil 2001;57(Suppl):55-60.

46 Subtotal Ear Canal Ablation in 18 Dogs and One Cat With Minimal Distal Ear Canal Pathology A modified technique for performing total ear canal ablations is described. This technique requires less dissection than the standard technique and maintains a portion of the distal vertical ear canal. Subtotal ear canal ablations were performed in 18 dogs and one cat for the treatment of otitis externa or masses of the horizontal ear canal. Animals with otitis externa had minimal involvement of the distal ear canal. Dermatological problems associated with the remaining ear canal and pinnae occurred in eight animals and resolved with medical management. Normal ear carriage was maintained in all animals with erect ears. Further investigation is required before the procedure can be recommended as a treatment for otitis externa not caused by masses or anatomical abnormalities of the horizontal ear canal in dogs with pendulous ears. J Am Anim Hosp Assoc 2006;42: Kyle G. Mathews, DVM, MS, Diplomate ACVS Elizabeth M. Hardie, DVM, PhD, Diplomate ACVS K. Marcia Murphy, DVM, Diplomate ACVD RS From the Department of Clinical Sciences, College of Veterinary Medicine, North Carolina State University, 4700 Hillsborough Street, Raleigh, North Carolina Introduction Total ear canal ablation (TECA) with lateral bulla osteotomy is widely accepted as a viable treatment option for dogs and cats with end-stage otitis externa or masses confined to the ear canal. As described, the procedure involves a circumferential incision around the funnel-shaped auricular cavity (i.e., cavum conchae) and through the auricular cartilage at the base of the pinna. 1-9 A vertical incision is made through the skin overlying the vertical ear canal. The vertical and horizontal ear canals are dissected free of the soft tissues to the level of the tympanic bulla, and a bulla osteotomy is performed. 2 Following completion of a TECA and lateral bulla osteotomy, the tissues are closed in a T-shaped or inverted L- shaped suture line. Amodification of this technique for dogs with erect ears has been described in the Japanese literature. 10 The purpose of this technique is to avoid a change in ear carriage associated with excision of the medial portion of the auricular cartilage. An inverted L-shaped skin incision is made over the vertical canal, just ventral to the auricular cavity, to facilitate exposure. The annular cartilage of the vertical canal is transected at this point, followed by routine dissection and removal of the remaining vertical and horizontal canals. 10 The horizontal portion of the L-shaped skin incision is sutured to the remaining transected end of the vertical canal. The distal portion of the vertical canal is thus preserved, resulting in a stoma just ventral to, and communicating with, the external orifice. In the original study, this procedure was performed in three dogs with chronic otitis externa/media, and no apparent problems were reported at 1-year follow-up examinations. 10 A more recent report evaluated the cosmetic outcome of a modified TECA technique in six cats. 11 Erect ear carriage was maintained by creating a ventrally-based advancement flap after circumferential incision around the external acoustic meatus and removal of the entire vertical and horizontal ear canal. 11 JOURNAL of the American Animal Hospital Association 371

47 372 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 The purpose of this report is to describe a subtotal ear canal ablation technique that was used for the treatment of otitis externa/media or masses of the horizontal external ear canal in dogs and cats. The procedure was initially performed as a method to maintain erect ear carriage, and then it was expanded to include a subset of dogs with pendulous ears. In each case, the medial surface of the pinna and the distal auricular cavity were originally disease-free or minimally affected by the underlying disease process. Materials and Methods The medical records of dogs and cats undergoing subtotal ear canal ablation at the North Carolina State University s College of Veterinary Medicine from 1998 to 2004 were reviewed. Subtotal ear canal ablation was defined as resection of the proximal half of the vertical ear canal and the entire horizontal ear canal. Information retrieved from the medical records included species, breed, gender, body weight, age, computed tomographic and otoscopic findings, duration of clinical signs, preoperative medical management, pathological and microbiological diagnoses, postoperative medical management, follow-up times, and postoperative complications. The referring veterinarians and owners of animals that had subtotal ear canal ablations performed were then contacted by telephone. Surgical Technique A single vertical incision was made overlying the vertical ear canal, from just ventral to the midpoint of the external orifice to ventral to the horizontal ear canal [Figures 1A, 2]. With the aid of retractors, the central portion of the vertical ear canal was freed from its soft-tissue attachments, using blunt and sharp dissection, and then transected [Figures 1B, 1C, 3]. The proximal portion of the vertical canal and the Figure 2 Postoperative photograph of an 11-year-old German shepherd dog (case no. 16), showing the location of the skin incision. Figure 3 Photograph of a 5-year-old basset hound-mix (case no. 6) taken after the distal vertical ear canal has been transected. The pinna is at the top of the photograph. The proximal portion of the vertical ear canal (arrow) and the entire horizontal ear canal will be removed, as done for a standard total ear canal ablation. Figures 1A-1D (A) Line drawing showing the location of the skin incision parallel to, and directly over, the vertical ear canal. (B) The vertical canal is dissected free of its attachments and elevated by passing an instrument (Metzenbaum scissors) behind the canal. (C) The canal is transected and the distal portion closed with cruciate sutures placed in the cut edges of the cartilage while avoiding the epithelium. (D) Following closure of the distal canal, the proximal vertical canal and horizontal canal are followed down to the tympanic bulla, as with a standard total ear canal ablation. horizontal canal were then removed, as done for a standard TECA. Following lateral bulla osteotomy and curettage, the transected end of the distal vertical ear canal was grasped and elevated. The medial and lateral vertical ear canal cartilages were apposed with multiple, simple interrupted or cruciate sutures using absorbable 3-0 to 4-0 monofilament material [Figures 1C, 1D, 4A, 4B]. a Care was taken to avoid penetrating the epithelium of the vertical canal with the sutures. Subcutaneous and skin closure was then performed in a routine manner. The result was a shallow, blind-ended auricular cavity, with preservation of the entire circumference of annular cartilage surrounding the external orifice [Figures 2, 5].

48 September/October 2006, Vol. 42 Subtotal Ear Canal Ablation 373 Table Clinical Data for 19 Animals Treated With Subtotal Ear Canal Ablation Case Operated Follow-up No. Signalment * Side(s) Diagnosis Histopathology Culture (mo) Outcome 11 7-y, 38-kg, SF Left Congenital ear Lymphoplasmacytic and No growth 53 No ear problems; occasional rottweiler canal stenosis histiocytic inflammation of mild bilateral wax buildup the horizontal ear canal; that was removed with cotton cholesterol granuloma balls. within the bulla y, 14-kg, CM Right Otitis NA Heavy growth 49 Right facial nerve paralysis Scottish terrier externa/media Staphylococcus for 2 mo. No ear problems intermedius despite chronic atopy, pyoderma y, 33-kg, SF Left Traumatic Otitis externa with No growth 43 Required cleaning of wax Labrador retriever separation of ear severe, diffuse, chronic from both ears q 2 wks. canals neutrophilic otitis media Seasonal inguinal and and severe, diffuse bilateral aural fibrosis erythema/pruritus y, 11-kg, CM Left Craniomandibular Otitis externa/media No growth 42 Left facial nerve paresis for 3 West Highland osteopathy with myelofibrosis, wks. white terrier osteosclerosis, medial vascular fibrosis 15 7-y, 43-kg, CM Bilateral Otitis externa NA Heavy growth 39 Unilateral incisional infection German shepherd Pseudomonas at suture removal. dog aeruginosa and Intermittent seasonal pruritus Staphylococcus and erythema in right ear. aureus 16 5-y, 25-kg, CM Right Otitis externa NA Moderate growth 32 No right ear problems for 2 y basset hound-mix Pseudomonas postoperatively, then aeruginosa developed generalized pruritus. Contralateral ear developed Pseudomonas spp. otitis externa 14 mo postoperatively. (Continued on next page)

49 374 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table (cont d) Clinical Data for 19 Animals Treated With Subtotal Ear Canal Ablation Case Operated Follow-up No. Signalment * Side(s) Diagnosis Histopathology Culture (mo) Outcome 17 8-y, 53-kg, CM Right Plasmacytoma with Plasmacytoma at junction Heavy growth 25 Atopy diagnosed 1.5 y later, Akita secondary otitis of horizontal and vertical Pseudomonas but no aural involvement externa ear canals (clean margins) aeruginosa occurred y, 33-kg, CM Left Chronic otitis Chronic, moderate, diffuse Heavy growth 23 Developed bilateral standard poodle externa/media otitis externa/media with Pseudomonas Pseudomonas spp. otitis externa pyogranulomatous aeruginosa 8 mo postoperatively. Bilateral inflammation, fibrosis, and aural erythema/pruritus occurred mineralization 20 mo postoperatively y, 27-kg, CM Bilateral Chronic otitis externa Chronic, mild left and No growth 14 Several episodes of aural pain German shepherd moderate right diffuse, and discharge between 1 and 7 dog lymphoplasmacytic mo postoperatively. ulcerative otitis externa Staphylococcus intermedius and group G Streptococcus cultured y, 42-kg, CM Left Otitis externa NA Heavy growth 12 Requires cleaning of mild wax German shepherd Pseudomonas from ears every 1-2 mo. dog aeruginosa; moderate growth Staphylococcus intermedius y, 26-kg, SF Left Otitis media Otitis media with chronic, Heavy growth 12 Dark, waxy buildup in left ear 1 beagle severe, diffuse otitis externa; Escherichia coli mo postoperatively, with cocci granulomatous dermatitis; and yeast identified on cytology. cystic glandular dilatation; neutrophilic adenitis; and diffuse epithelial hyperplasia 12 6-y, 40-kg, CM Left Otitis externa/media Otitis externa/media with Moderate growth 11 No left-sided ear problems; rottweiler marked dermal fibroplasias, Staphylococcus intermittent contralateral otitis moderate sebaceous intermedius; externa. Euthanized for hyperplasia, and apocrine heavy growth group G progressive lymphoma. gland dilatation Streptococcus (Continued on next page)

50 September/October 2006, Vol. 42 Subtotal Ear Canal Ablation 375 Table (cont d) Clinical Data for 19 Animals Treated With Subtotal Ear Canal Ablation Case Operated Follow-up No. Signalment * Side(s) Diagnosis Histopathology Culture (mo) Outcome y, 24-kg, SF Bilateral Chronic otitis externa Chronic, severe, diffuse, No growth 10 No ear problems. German histiocytic nodular otitis shepherd-mix externa; histiocytic, neutrophilic, ceruminous adenitis and osseous metaplasia y, 40-kg, CM Left Otitis externa/media Otitis externa/media; bulla: Heavy growth 9 Developed left Malassezia spp. bullmastiff fibrosis with chronic Pseudomonas otitis externa 3 mo lymphoplasmacytic and aeruginosa postoperatively. neutrophilic inflammation 15 5-y, 11-kg, CM pug Bilateral Otitis externa/media Left chronic Light growth 5 No external ear problems. Left lymphoplasmacytic Staphylococcus head tilt and diminished otitis media; intermedius palpebral reflex (present right moderate preoperatively) persisted. hyperkeratosis of Developed ataxia, anisocoria, bulla epithelium circling to right, facial twitch, and cervical pain 5 mo postoperatively. Diagnosed with encephalitis/meningitis; developed aspiration pneumonia and was euthanized y, 34-kg, CM Left Right otitis media; Right otitis media; Light growth 4 Bilateral facial nerve paresis for German shepherd left otitis externa left chronic, hyperplastic, Staphylococcus 3 d; preoperative head tilt and dog neutrophilic, and intermedius ataxia resolved within 2 wks; lymphohistiocytic otitis right ear pruritus with minimal externa with extensive waxy debris developed 3 mo fibrosis postoperatively y, 9-kg, SF pug Right Otitis externa/media Ulcerative fibronecrotic otitis No growth 4 No ear problems; right corneal externa/media ulcer developed secondary to facial nerve paralysis; palpebral reflex returned 2 mo postoperatively. (Continued on next page)

51 376 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table (cont d) Clinical Data for 19 Animals Treated With Subtotal Ear Canal Ablation Case Operated Follow-up No. Signalment * Side(s) Diagnosis Histopathology Culture (mo) Outcome y, 31-kg, SF Bilateral Otitis externa/media NA Heavy growth 3 Left head tilt with horizontal German shepherd Pseudomonas nystagmus and left facial nerve dog aeruginosa; paralysis developed moderate growth postoperatively; nystagmus group G resolved in 3 d. Head tilt Streptococcus diminished but still present at 3 mo; facial paralysis resolved over 3 mo; developed bilateral Malassezia spp. otitis externa 2 mo postoperatively y, 7-kg, SF Left Ceruminous gland Ceruminous gland adenoma Heavy growth 8 No ear problems. domestic shorthair adenoma with otitis at tympanic membrane with Staphylococcus cat externa chronic, diffuse intermedius lymphoplasmacytic otitis externa, sebaceous hyperplasia, and lamina propria fibrosis * SF=spayed female; CM=castrated male; erect-eared breeds=case nos. 2, 4, 5, 7, 9, 10, 13, 16, 18, 19 Case nos. 1 and 16 also had ventral bulla osteotomies (case no. 1, ipsilateral; case no. 16, contralateral); case no. 11 also had a total ear canal ablation on the contralateral (right) side. NA=not available

52 September/October 2006, Vol. 42 Subtotal Ear Canal Ablation 377 A B Figures 4A, 4B Photographs of the same dog in Figure 3. (A) The distal vertical ear canal is closed with monofilament absorbable sutures that engage only the cartilage. Care is taken to make sure the sutures do not penetrate the epithelium of the ear canal (arrow). (B) Closure of the distal vertical ear canal cartilage is completed, resulting in a blind pouch at the auricular cavity. Figure 5 Photograph of a 12-year-old, spayed female, German shepherd-mix (case no. 13) 3 days postoperatively, showing that erect ear carriage was maintained. Results Clinical Data Subtotal ear canal ablation was performed in 19 animals (24 ears). Nine dogs and one cat had erect ears (14 operated ears). A specific goal in animals with erect ears was to maintain ear carriage. The procedure was also performed in nine dogs with pendulous ears (10 operated ears) that had no or minimal involvement of the distal vertical ear canal. Numerous breeds were represented, including the German shepherd dog (n=5), rottweiler (n=2), pug (n=2), and one each of a German shepherd-mix, Scottish terrier, Labrador retriever, West Highland white terrier, basset hound-mix, Akita, standard poodle, beagle, bullmastiff, and domestic shorthair cat [see Table]. Twelve of the animals were castrated males, and seven were spayed females. Mean body weight of the 18 dogs was 30 kg (range 9 to 53 kg). Median age at the time of surgery was 7 years (range 2 to 12 years). All animals were diagnosed with otitis externa with minimal or no involvement of the distal vertical ear canal based on otoscopic, pathological, and/or microbiological evaluations [see Table]. Computed tomography (CT) revealed partial or complete filling of the horizontal ear canal with soft-tissue density in the 18 cases in which it was performed. Eleven of these 18 cases (23 ears) also had evidence of horizontal canal mineralization. An additional 11 animals had evidence of middle ear involvement on CT. One dog (case no. 7) and one cat (case no. 19) had benign masses within the horizontal ear canal (i.e., plasmacytoma and ceruminous adenoma, respectively). Three dogs had additional diagnoses, such as congenital ear canal stenosis based on intraoperative and CT findings (case no. 1), traumatic ear canal separation based on intraoperative findings and a history of head trauma (case no. 3), and craniomandibular osteopathy based on CT and histopathological results (case no. 4). Other diseases previously diagnosed and treated in the affected animals included hypothyroidism (case nos. 2, 12), occipital hemangiopericytoma (case no. 1), and keratoconjunctivitis sicca and lymphosarcoma (case no. 12). For those animals with clinical signs (n=18), the median duration was 3.5 years (range 1 month to 10 years). Clinical signs were consistent with otitis externa and included aural erythema, pain, pruritus, purulent discharge, and head shaking. Case no. 2 also had a draining tract cranioventral to the right ear canal. Prior treatments were numerous, varied, and frequently consisted of topical and oral antibiotics, antifungal medications, and corticosteroids; topical aural cleaning solutions; and hypoallergenic diets. Intraoperative bacterial cultures were positive in 13 of 19 cases (when bilateral surgery was performed, samples were combined in the laboratory). Bacteria identified

53 378 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 included Pseudomonas aeruginosa (n=7), Staphylococcus intermedius or aureus (n=6), group G Streptococcus (n=2), and Escherichia coli (n=1). Surgical Technique Slight modifications of the surgical procedure were needed in three cases. In case no. 1 (congenital ear canal stenosis), the bulla osteotomy was performed via a ventral approach rather than a lateral one, because the horizontal ear canal tapered to a blind end medially. Continuing with the lateral approach would have jeopardized the facial nerve. Case no. 3 did not require suture closure of the distal vertical ear canal, as it was already closed secondary to traumatic separation from the horizontal canal. Exposure of the tympanic bulla was improved by creating a T at the proximal extent of the skin incision in the cat (case no. 19) that underwent the procedure. The mean duration of surgery was 1.6 hours (range 0.75 to 3 hours). The longest procedure (3 hours [case no. 1]) required more time to reposition and perform a ventral bulla osteotomy. Other Treatments and Follow-up All animals received perioperative analgesic therapy consisting of topical fentanyl patches, local nerve blocks with 0.25% bupivicaine, and intravenous narcotics. In addition, animals were discharged on oral or subcutaneous antibiotics (n=17), oral nonsteroidal anti-inflammatory drugs (n=10), and eye lubricants (n=7). Median time to follow-up was 12 months (mean 21 months; range 3 to 53 months) [see Table]. Telephone contact was made with the owners of 15 animals and with the referring veterinarians of 18 animals. Fourteen of the 15 owners contacted were pleased with the outcome. Postoperative Complications Temporary facial nerve paresis or paralysis (duration range 3 days to 3 months) developed in five dogs and was treated with eye lubricants (n=5) and a temporary tarsorrhaphy (n=1). Case no. 18 also developed a head tilt to the same side (i.e., left) and horizontal nystagmus following bilateral subtotal ear canal ablations. The nystagmus resolved in 3 days, the palpebral reflex returned after 3 months, and the head tilt improved but was still present 3 months postoperatively. One dog with chronic otitis had a left head tilt and diminished left palpebral reflex (case no. 15) at initial presentation. Loss of bone on the medial aspect of the left bulla was noted on CT. Neurological signs persisted for 5 months following surgery, at which time the dog became ataxic, was diagnosed with encephalitis/meningitis and aspiration pneumonia, and was euthanized. Eleven of 19 animals (13 of 24 ears) treated with subtotal ear canal ablation had no dermatological problems associated with the remaining vertical ear canal and pinna [see Table]. The owners of two of these animals (case nos. 1, 10; both had unilateral surgery) intermittently cleaned wax from the ears, although the animals had no clinical signs of otitis. Three dogs were subsequently diagnosed with atopy, and two others developed otitis externa in the contralateral ear. None of these latter five dogs developed any ear-related clinical signs on the side in which subtotal ear canal ablation was performed. Four of 19 animals (five of 24 ears) treated with subtotal ear canal ablation had minor and transient dermatological problems associated with the remaining vertical ear canal and pinna [see Table]. Case no. 3 (left-sided ablation) developed seasonal, inguinal and bilateral aural erythema and pruritus that required sporadic treatment with topical medications. Case no. 5 (bilateral ablation) had a temporary, left-sided incisional infection and intermittent, seasonal right-sided erythema and pruritus that responded to several days of topical therapy. Case no. 6 (right-sided ablation) developed generalized pruritus of the interdigital area, rump, and pinnae 2 years postoperatively, which responded to topical therapy and did not recur. This same dog also developed contralateral (i.e., left-sided) otitis externa 14 months after the right ear was operated. Case no. 11 (leftsided subtotal ablation, right-sided TECA) developed cocci and yeast in the left auricular cavity 1 month postoperatively, which responded to topical therapy and did not recur. Four of 19 animals (six of 24 ears) treated with subtotal ear canal ablation had protracted dermatological problems associated with the remaining vertical ear canal and pinna that required intermittent or long-term therapy [see Table]. Four of the six ears were erect, and two were pendulous. Case no. 8 (left-sided ablation) developed bilateral otitis externa from Pseudomonas spp. 8 months postoperatively that resolved after 1 month of topical and systemic therapy. This same animal had an episode of bilateral aural erythema and pruritus 20 months postoperatively that resolved after 3 days of topical therapy. Case no. 9 (bilateral ablation) developed multiple episodes of bilateral bacterial otitis externa starting 1 month postoperatively that persisted for 6 months. Treatment consisted of topical otic cleaning solutions, oral antibiotics, and nonsteroidal anti-inflammatory drugs. Clinical signs resolved once the dog was started on a hypoallergenic diet and a twice-weekly topical antibiotic/antifungal/corticosteroid ointment. Case no. 14 (left-sided ablation) developed left-sided otitis externa from Malassezia spp. 3 months postoperatively, which resolved following 2 weeks of topical and systemic therapy. Case no. 18 (bilateral ablation) developed bilateral otitis externa from Malassezia spp. 2 months postoperatively, which required 1 month of topical and systemic therapy, as well as a hypoallergenic diet. Discussion Removal of the entire vertical ear canal can result in altered ear carriage in dogs and cats with erect ears. 2-8,12-14 Total ear canal ablations have been less commonly performed in cats and have been performed primarily to treat neoplasia rather than chronic otitis. 9,11,15 In a prior report, ear carriage was maintained in six cats using a ventrally-based advancement flap, but greater dissection was required than with the technique reported here. 11 Attempts to minimize alteration

54 September/October 2006, Vol. 42 Subtotal Ear Canal Ablation 379 in ear carriage in dogs have included 1) resection of the skin caudal to the vertical portion of the T-shaped skin incision, so that the remaining cartilage at the base of the pinna is pulled into a shallow cup following skin closure; 2) using an inverted L-shaped skin incision to achieve cupping of the pinna during closure; 3) placement of a temporary suture between the dorsal pinna and the skin overlying the zygomatic arch; and 4) retention of the distal vertical ear canal and creation of a stoma just ventral to, and communicating with, the external orifice. 2,3,8,10,13 Long-term results regarding ear carriage following these techniques have rarely been reported. 10 In the last report describing stoma creation, it was theorized that the stoma would help eliminate recurrence of otitis externa by improving aeration and drainage. 10 Several animals in the current report had minor (three dogs with pendulous ears and one with erect ears) to protracted (two dogs with pendulous ears and two with erect ears) dermatological problems of the pinna or auricular cavity following surgery. It was unclear if the creation of a stoma would have reduced these complications. The subtotal ear canal ablation technique reported here resulted in an ear that appeared normal and was similar to a previously reported modification of the lateral ear canal resection technique. 12 While many owners of animals with erect ears are not concerned about ear carriage, ear-droop following standard TECA and methods to prevent its occurrence should be discussed prior to surgery. The subtotal ear canal ablation procedure used in this report preserved erect ear carriage and eliminated the dissection through and around the medial aspect of the auricular cartilage that is required for a standard TECA. Although necrosis of the pinna is a rare complication following standard TECA, subtotal ear canal ablation may also reduce this risk. 4,11 The subjective impression of the surgeons performing this procedure is that subtotal ear canal ablation is easier to perform than a standard TECA, results in less hemorrhage, and may be less painful to the animal postoperatively. Prospective studies are needed to investigate these issues. Because these potential benefits were observed in several dogs with erect ears, the technique was subsequently performed on nine dogs with pendulous ears and minimal to no distal ear canal involvement. Animals that have masses or gross changes to the vertical ear canal secondary to otitis externa are not candidates for subtotal ear canal ablation. Such dogs include cocker spaniels, because they typically have significant involvement of the distal ear canal and the auricular cavity by the time a TECA is considered. The risk of recurrent disease in the remaining vertical canal is too great to warrant alternative, tissue-sparing surgical techniques. Additionally, dogs with pendulous ears that have otitis externa from a cause other than a horizontal ear canal mass, horizontal ear canal stenosis, or traumatic separation are not good candidates for subtotal ablations because of the potential for otitis recurring within the remaining vertical canal. All of the animals in this report had grossly normal distal vertical ear canals and normal external orifices. In addition, masses were confined to the horizontal ear canal. Infectious complications associated with this procedure included superficial and transient skin infections involving the remaining auricular cavity (seen in four animals; five of 24 ears, 21%). Another four animals (six of 24 ears, 25%) developed protracted auricular cavity skin infections that eventually resolved with medical management. These results were comparable to the recurrent dermatological problems of the pinna that have been reported in 26% of dogs after standard TECA surgery. 4 Although all superficial infections were corrected with medical management, owners should be warned of the possibility of these complications prior to performing subtotal ear canal ablation, even if there is no involvement of the distal ear canal at the time of surgery. Although the blind pouch (i.e., distal vertical canal) created by the surgery is shallow, it may predispose individual animals to a buildup of wax and/or provide an appropriate environment for the recurrence of superficial infections. Five (26%) animals without neurological signs at presentation developed transient facial nerve paresis or paralysis. During subtotal ear canal ablation, the soft tissues were retracted away from the horizontal ear canal with Gelpi retractors. b Care was taken to ensure the retractors were placed superficially to the facial nerve, as it courses cranially and ventrally to the horizontal ear canal. Although retraction may have contributed to the facial nerve trauma in these cases, the frequency of this complication was similar to that previously reported for the standard TECA technique. 2,4,5 The subjective impression of the surgeons performing this procedure is that facial nerve identification and retraction are no more difficult than during a standard TECA; however, in the cat, the skin incision is modified to improve exposure. As with the standard technique, owners should be made aware of the possibility of postoperative neurological complications. Conclusion Based on the results of this study, subtotal ear canal resection appears to be an option for dogs and cats with masses or other anatomical abnormalities of the horizontal ear canal. The procedure may also be an option for erect-eared animals with otitis externa that does not affect or minimally involves the distal vertical ear canal. The potential risk of continued or recurrent local infection should be discussed with the owner prior to surgery. Erect ear carriage was maintained with this modification. Further investigation is required before the procedure can be recommended in pendulous-eared dogs as a treatment for otitis externa not caused by masses or anatomical abnormalities of the horizontal ear canal. a PDS II; Ethicon, Inc., Somerville, NJ b Gelpi retractors; Sklar Instruments, West Chester, PA References 11. Schaller O, ed. Illustrated Veterinary Anatomical Nomenclature. Stuttgart: Ferdinand Enke Verlag, 1992:541.

55 380 JOURNAL of the American Animal Hospital Association September/October 2006, Vol Beckman SL, Henry Jr WB, Cechner P. Total ear canal ablation combining bulla osteotomy and curettage in dogs with chronic otitis externa and media. J Am Vet Med Assoc 1990;196: Harvey CE. Ear canal disease in the dog: medical and surgical management. J Am Vet Med Assoc 1980;177: Matthiesen DT, Scavelli T. Total ear canal ablation and lateral bulla osteotomy in 38 dogs. J Am Anim Hosp Assoc 1990;26: Smeak DD, DeHoff WD. Total ear canal ablation clinical results in the dog and cat. Vet Surg 1986;15: Lane JG, Little CJL. Surgery of the canine external auditory meatus: a review of failures. J Small Anim Pract 1986;27: White RAS, Pomeroy CJ. Total ear canal ablation and lateral bulla osteotomy in the dog. J Small Anim Pract 1990;31: Lane JG. ENT and oral surgery of the dog and cat. Boston: PSG Wright, 1982: Williams JM, White RAS. Total ear canal ablation combined with lateral bulla osteotomy in the cat. J Small Anim Pract 1992; 33: Okamoto Y, Miyatake K, Inoue T, et al. Total ear-canal ablation preserving the auricular annular cartilage. J Jpn Vet Med Assoc 2001;54: McNabb AH, Flanders JA. Cosmetic results of a ventrally based advancement flap for closure of total ear canal ablations in 6 cats: Vet Surg 2004;33: Bardens JW. Relocation of the auditory canal in chronic otitis. Small Anim Clin 1962;2: Fraser G, Withers AR, Spreull JSA. Otitis externa in the dog. J Small Anim Pract 1961;2: McCarthy RJ, Caywood DD. Vertical ear canal resection for endstage otitis externa in dogs. J Am Anim Hosp Assoc 1992;28: Marino D, MacDonald JM, Matthiesen DT, et al. Results of surgery in cats with ceruminous gland adenocarcinoma. J Am Anim Hosp Assoc 1994;30:54-58.

56 Left Lateral and Left Middle Liver Lobe Torsion in a Saint Bernard Puppy A 5-month-old, male Saint Bernard was presented for acute collapse and abdominal discomfort. Significant findings were a cranial abdominal mass, hemorrhagic abdominal effusion, anemia, and disseminated intravascular coagulation. An exploratory surgery revealed torsion of both the left lateral and middle liver lobes, a condition that has not been previously described in the veterinary literature. Torsion of one or more hepatic lobes is a rare condition but should be considered as a differential diagnosis for acute abdomen syndrome in both young and mature dogs. Early diagnosis and prompt surgical intervention may be curative. J Am Anim Hosp Assoc 2006;42: Dirsko J.F. von Pfeil, Dr.med.vet. L. Ari Jutkowitz, VMD, Diplomate ACVECC Joe Hauptman, DVM, MS, Diplomate ACVS CR From the Department of Small Animal Clinical Sciences, College of Veterinary Medicine, Michigan State University, East Lansing, Michigan Introduction Hepatic lobe torsion is a rare condition that has been reported in dogs, cats, pigs, rabbits, horses, and humans Signs associated with liver lobe torsion are variable and may include abdominal pain, anemia, anorexia, abdominal effusion, and fever. 1-8 Different etiologies for this condition have been described or suggested; they include trauma, stretching or aplasia of the hepatic ligaments, distinct lobulation of the liver, entrapment of the liver by the hepatogastric ligament, torsion secondary to gastric dilatation volvulus or other gastric surgeries, and hepatic neoplasia. 1-7,18 In one review of liver lobe torsion in the dog, the left lateral lobe was most commonly affected (48%, 11 cases), followed by the caudate lobe (22%, five cases), the right lateral lobe (17%, four cases), the right medial lobe (4%, one case), and the entire liver (4%, one case). 7 Double liver lobe torsion is an extremely rare condition and was reported in only one case from In that report, torsion of the left lateral lobe and papillary process of the caudate liver lobe was described in a puppy, secondary to congenital aplasia of the left triangular ligament. 6 The purpose of this report is to describe torsion of both the left lateral and left medial lobes in a young dog. No underlying cause for these abnormalities could be determined. Case Report A 5-month-old, 36-kg, male Saint Bernard was referred for acute weakness, collapse, and abdominal discomfort. The dog had experienced lethargy, inappetence, and soft stools for 2 days prior to examination and had vomited twice in the 24 hours prior to referral. Initial diagnostic tests performed by the referring veterinarian included a fecal flotation and an in-house parvovirus test, both of which were negative, and abdominal radiographs that showed loss of serosal detail. A complete blood count (CBC) revealed marked anemia, leukocytosis, and thrombocytopenia [see Table]. A serum biochemical panel was unremarkable except for a mildly elevated blood urea nitrogen (BUN) [see Table]. JOURNAL of the American Animal Hospital Association 381

57 382 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Table Selected Laboratory Results for a Puppy With Torsion of the Left Lateral and Left Middle Liver Lobe Day Prior Laboratory Test (unit) Reference Range to Referral Day 1 Day 2 Day 3 Complete Blood Count ph Venous pco * 2 (mm Hg) Venous po 2 (mm Hg) HCO 3 (mmol/l) Lactate (mmol/l) Hematocrit (%) manual Hematocrit (%) auto-analyzer Red blood count (cells/µl) Hemoglobin (g/dl) Reticulocytes (10 3 cells/µl) Granulocytes (10 3 cells/µl) Leukocytes (10 3 cells/µl) Neutrophils (10 3 cells/µl) Lymphocytes (10 3 cells/µl) Monocytes (10 3 cells/µl) Platelets (10 3 cells/µl) Serum Biochemical Panel BUN (mg/dl) Total calcium (mg/dl) Phosphorus (mg/dl) Serum total solids (g/dl) Total protein (g/dl) Albumin (g/dl) ALP \ (U/L) Total bilirubin (mg/dl) Creatine kinase (U/L) Coagulation Panel FDP (µg/ml) < D-dimers (ng/ml) <250 > PT # (sec) APTT ** (sec) Fibrinogen (mg/dl) * pco2 =partial pressure of carbon dioxide po2 =partial pressure of oxygen HCO3 =bicarbonate BUN=blood urea nitrogen \ ALP=alkaline phosphatase FDP=fibrin degradation products # PT=prothrombin time ** APTT=activated partial thromboplastin time

58 September/October 2006, Vol. 42 Atypical Liver Lobe Torsion 383 Physical examination findings at the time of presentation included lethargy, pale mucous membranes, prolonged capillary refill time (3 seconds), moderate dehydration, tachypnea (96 breaths per minute), tachycardia (200 beats per minute), weak pulse quality, fever (39.6 C; F), and a distended, painful abdomen. A fluid wave was present on abdominal palpation, and a firm structure could be palpated in the cranial abdomen. Packed cell volume (PCV) and total solids on admission were 14% and 5.1 g/dl, respectively (normal 43% to 59% and 5.5 to 7.4 g/dl, respectively). A venous blood gas analysis revealed respiratory alkalosis [see Table]. Abdominocentesis yielded a nonclotting, sanguinous fluid with a PCV and total solids of 21% and 5 g/dl, respectively. Cytological examination showed large numbers of erythrocytes, occasional nondegenerate neutrophils, and occasional macrophages, consistent with hemorrhagic effusion. Initial treatment included the intravenous (IV) administration of lactated Ringer s solution a (90 ml/kg) and one unit (350 ml) of packed red blood cells (RBCs). A CBC revealed normocytic, normochromic, nonregenerative anemia; thrombocytopenia; leukocytosis; neutrophilia; and lymphopenia [see Table]. Elevations in BUN, alkaline phosphatase, total bilirubin, and creatine kinase, as well as hypoproteinemia and hypoalbuminemia, were detected on a biochemical panel. Mild elevations in calcium and phosphorus were considered normal for a puppy [see Table]. A urinalysis was unremarkable except for isosthenuria (specific gravity 1.009). A coagulation profile revealed prolongation of prothrombin time and activated partial thromboplastin time, decreased fibrinogen, and positive d-dimers consistent with disseminated intravascular coagulation (DIC) [see Table]. The dog was blood typed as dog erythrocyte antigen 1.1 positive. Abdominal ultrasonography revealed a moderate amount of hypoechogenic peritoneal effusion. A large, hypoechoic mass was identified on the left lateral aspect of the midabdomen. There were hyperechoic parallel lines within the mass [Figure 1]. No evidence of blood flow was seen within the mass on Doppler examination. The lateral aspect of the left lateral liver lobe was hypoechoic, with hyperechoic speckles scattered throughout the stroma [Figure 2], and it lacked blood flow on Doppler examination. These findings were interpreted as splenic torsion and possible thrombosis of a portion of the left lateral liver lobe. An exploratory celiotomy was performed for diagnostic and therapeutic purposes. Prior to surgery, the dog received one unit of fresh-frozen plasma (275 ml) over a 2-hour period. Anesthesia was induced using a 50:50 mixture of diazepam b (5 mg/ml) and ketamine c (100 mg/ml) at a dose of 0.1 ml/kg IV to effect. Anesthesia was maintained with isoflurane d (1.5% to 2%) in oxygen (0.4 L per minute). Lactated Ringer s solution a was also given at 20 ml/kg per hour IV, and oxymorphone e (0.1 mg/kg IV) was administered at the beginning of surgery. A standard ventral midline approach to the abdomen was made, and 2.5 L of serosanguinous effusion was suctioned from the abdominal cavity. Figure 1 Longitudinal view of a left lateral liver lobe torsion in a 5-month-old, male Saint Bernard puppy. Note the hyperechoic and hypoechoic lines, consistent with vasculature and twisted stroma (white and black asterisks). These findings were originally believed to represent a splenic torsion. Figure 2 Transverse ultrasonographic view of the left liver of the dog in Figure 1, showing an area of decreased echogenicity (circumscribed in black). Doppler examination of this area did not reveal blood flow. Both the left lateral and left middle liver lobes were torsed approximately 180 around their pedicles [Figure 3]. Both lobes were dark red to black in color [Figure 4]. They were easily movable, and there was no gross evidence of infection or neoplasia. Without untwisting the lobes, resection was performed using stapling equipment. f No bleeding occurred at the resection site after the lobes were removed. A thorough abdominal exploration revealed no other abnormalities. The abdominal cavity was flushed with 4 L of warm, sterile saline solution. The liver lobectomy sites were again checked for bleeding, and the abdomen was closed in a

59 384 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 using a Doppler flow probe j for 12 hours after surgery, and pressure remained between 100 and 120 mm Hg. A second coagulation profile, taken 1 day after surgery, showed mild improvement in the dog s coagulation status [see Table]. The dog improved clinically and was discharged 44 hours after surgery, with a PCV of 25% and a total serum protein of 5.4 g/dl. Telephone contact with both the owner and referring veterinarian 2 months after surgery revealed that the puppy had completely recovered and was healthy. Figure 3 Schematic drawing of the diaphragmatic aspect of the liver, demonstrating the torsion of both the left lateral (1) and left middle (2) liver lobes. The arrow depicts the direction of the torsion. Note: 3=quadrate lobe; 4=right middle lobe; 5=right lateral lobe; 6=caudate process; 7=papillary process; 8=coronary ligament; 9=right triangular ligament; 10=left triangular ligament; 11=falciforme ligament; 12=round ligament; 13=esophagus; 14=gallbladder; 15=caudal vena cava; 16=hepatic veins. Figure 4 Intraoperative view of the torsed left middle liver lobe (TL) following removal of the left lateral liver lobe. routine manner. Packed cell volume and total protein immediately after surgery were 18% and 5.0 g/dl, respectively. Postoperative treatment consisted of hydromorphone g (0.1 mg/kg IV q 4 hours), midazolam h (0.2 mg/kg IV q 4 hours as needed), a second unit of fresh-frozen plasma, a second unit of packed RBCs, and lactated Ringer s solution a supplemented with 20 meq/l potassium chloride. i Indirect blood pressure monitoring was performed every 4 hours Discussion Clinical signs associated with hepatic torsion are variable and may be consistent with shock. 1-5,7,8,18 The dog in this report presented with signs consistent with hypovolemic shock. Radiographs of the abdomen may show a masseffect in the cranial abdomen and loss of serosal detail, as seen in this case. Pneumoperitoneum or a gas-filled mass in the cranial abdomen may also be seen as a result of hepatic abscessation. 1,4 Ultrasonographic examination of liver lobe torsions may reveal a mass of mixed echogenicity as a result of ischemia and necrosis, and the findings in this case were similar. 18 As in this report, it has been found in previous cases that it can be difficult to diagnose hepatic lobe torsion with diagnostic imaging alone. 4,18 In the dog of this study, liver lobe torsion was initially not high on the list of differential diagnoses, as the condition is not commonly reported in young dogs. The preliminary diagnosis of splenic torsion was made from the sonographic appearance of the mass on the left side of the abdomen and the inability to identify a normal spleen. A final diagnosis of liver lobe torsion was reached only at the time of surgery, which was consistent with previous reports where diagnosis was made only at surgery or necropsy. 1-5,7,8,10-12,15,18 Computed tomography and magnetic resonance imaging are noninvasive options that could have been considered to better visualize the lesion. However, as emergency surgery was dictated by the dog s clinical status, these techniques were not explored. Rapid surgical intervention is generally recommended in the treatment of liver lobe torsion in order to prevent massive hemorrhage into the abdominal cavity, DIC, and possible bacterial hepatitis leading to peritonitis. 2-4 The dog in this case was diagnosed with DIC, and treatment for DIC principally includes elimination of the inciting cause, restoration of effective circulating volume, and possibly administration of fresh-frozen plasma and/or heparin. In the dog of this report, it was felt that the risk of bleeding outweighed the possible benefits of heparinization, so DIC was addressed through restoration of effective circulating volume, removal of the inciting cause, and plasma transfusions. A coagulation profile, taken 1 day after surgery, showed improvement in coagulation parameters. A second platelet count was not taken postoperatively because of the improvement in the clinical signs and the financial constraints of the owner. It is recommended that resected liver lobes be submitted for histopathological examination to rule out underlying

60 September/October 2006, Vol. 42 Atypical Liver Lobe Torsion 385 conditions. 7 In the dog of this report, the resected lobes were not submitted for histopathological examination because of financial constraints. Neoplasia was considered to be an unlikely explanation for the liver lobe torsion in this 5-month-old puppy, as there was no gross evidence of neoplasia in any liver lobe. The histopathological findings most commonly reported in dogs with liver lobe torsion include necrosis, hepatic atrophy, and intense accumulations of neutrophils in the affected tissue. 7 Conclusion Torsion of the left lateral and middle liver lobes was found in a 5-month-old, male Saint Bernard puppy. Clinical signs associated with the liver lobe torsion included lethargy, inappetence, abdominal pain, vomiting, and soft stools. Hypovolemic shock, hemoabdomen, an abdominal mass, and signs of DIC were diagnosed. Immediate stabilization, resection of the affected liver lobes, and adequate postoperative care provided a successful outcome. Although liver lobe torsion has been reported primarily in mature dogs, this report illustrates that regardless of age, torsion of one or more liver lobes should be considered as a differential diagnosis in young dogs presenting with abdominal pain, hemoabdomen, and signs of DIC. 19. Morin M, Sauvageau R, Phaneuf JB. Torsion of abdominal organs in sows: a report of 36 cases. Can Vet J 1983;25: Fitzgerald AL, Fitzgerald SA. Hepatic lobe torsion in a New Zealand white rabbit. Canine Pract 1992;19: Wilson RB, Holscher MA, Sly DL. Liver lobe torsion in a rabbit. Lab Anim Sci 1987;37: Turner TA, Brown CA, Wilson JH. Hepatic lobe torsion as a cause of colic in a horse. Vet Surg 1993;22: Hamir AN. Torsion of the liver in a sow. Vet Rec 1980;106: Feins NR, Borger J. Torsion of the right lobe of the liver with partial obstruction of the colon. J Pediatr Surg 1972;7: LeFaucher C, Dupuis E, Muller J. Torsion du lobe de Riedel. J Chir (Paris) 1978;115: Tate P. Hepatic torsion and dislocation with hypotension and colonic obstruction. Am Surg 1993;59: Betge L. Torquierter akzessorischer Leberlappen als seltene Ursache eines akuten Abdomens. Dtsch Med Wochenschr 1995;120: Sonnenfield JM, Armbrust LJ, Radlinsky MA, et al. Radiographic and ultrasonographic findings of liver lobe torsion in a dog. Vet Radiol Ultrasound 2001;42: a Lactated Ringer s Solution; Abbott Laboratories, Chicago, IL b Diazepam Injection; Abbott Laboratories, Chicago, IL c Ketamine HCl Injection; Phoenix Scientific, Inc., St. Joseph, MO d Isoflurane; Abbott Laboratories, Chicago, IL e Oxymorphone HCl Injection; Endo Pharmaceuticals, Chadds Ford, PA f TA premium ; United States Surgical, Norwalk, CT g Hydromorphone HCl Injection; Baxter Healthcare Corporation, Deerfield, IL h Midazolam HCl Injection; Baxter Healthcare Corporation, Deerfield, IL i Potassium Chloride; Abbott Laboratories, Chicago, IL j Model 811-B Ultrasonic Doppler Flow Detector; Parks Medical Electronics, Inc., Aloha, OR References 11. McConkey S, Briggs C, Solano M, et al. Liver torsion and associated bacterial peritonitis in a dog. Can Vet J 1997;38: Downs MO, Miller MA, Cross AR, et al. Liver lobe torsion and liver abscess in a dog. J Am Vet Med Assoc 1998;212: Tomlinson J, Black A. Liver lobe torsion in a dog. J Am Vet Med Assoc 1983;183: Sato AF, Solano M. Radiographic diagnosis: liver lobe entrapment and associated emphysematous hepatitis. Vet Radiol Ultrasound 1998;39: Glardon O, Fluckiger M, Keller M, et al. Leberlappentorsion: Ein Beitrag zur Differentialdiagnose des akuten Abdomens. Kleintierpraxis 1987;32: Woolfe DT, English B. Torsion of the left lateral and papillary lobes of the liver in a pup. A case report. J Am Vet Med Assoc 1959;134: Swann HM, Brown DC. Hepatic lobe torsion in 3 dogs and a cat. Vet Surg 2001;30: Perl S, Yakobson B, Nobel TA, et al. Torsion of a liver lobe in a dog. Refuah Veterinarith 1981;38:

61 Combined Use of Surgery and Radiation in the Treatment of an Intradural Myxoid Liposarcoma in a Dog An intradural-extramedullary myxoid liposarcoma of the high cervical spine was diagnosed in a 9-year-old, spayed female Cavalier King Charles spaniel that was presented for a 2-month history of cervical pain and tetraparesis. Radiation therapy applied after surgery resulted in complete remission of the neurological deficits. The tumor recurred 18 months after surgical excision. A second surgery and another course of radiotherapy again resulted in complete remission of the clinical signs. The dog was euthanized 11 months after the second surgery because of tumor recurrence. J Am Anim Hosp Assoc 2006;42: Sergio Rodenas, DVM Isabel Valin, DVM Patrick Devauchelle, DVM Françoise Delisle, DVM Michel Baron, DVM CR From the Clinique Vétérinaire de référence en chirurgie Baron-Valin (Rodenas, Valin, Baron), 5 rue fernet, 94700, Maisons Alfort, France; and the Centre Radiotherapie-scanner (Devauchelle, Delisle), 7 Avenue Charles de Gaulle, 94700, Maisons Alfort, France. Doctor Rodenas current address is the Department of Animal Medicine and Surgery, Veterinary School, Autonomous University of Barcelona, Bellaterra, Barcelona, Spain. Introduction Liposarcoma is a malignant tumor of soft tissues. 1 Although metastatic spinal liposarcoma has been reported in humans, primary spinal liposarcoma remains a very rare condition. 2 Only a few cases of primary liposarcoma of the spine have been reported in humans, and a single case exists in the dog. 2-4 In all published cases, the tumor was in an extradural location. 2-4 The purpose of this report is to describe an unusual case of spinal cord compression from an intradural myxoid liposarcoma in a dog, which has not been reported previously. The clinical and imaging findings and details on the use of surgery and radiation therapy in the dog are all provided. Case Report A 9-year-old, spayed female Cavalier King Charles spaniel was referred for evaluation of intermittent lameness in the right forelimb and cervical spinal hyperesthesia of 2 months duration. The dog had been treated with prednisone by the referring veterinarian, without significant improvement. The only abnormalities found on physical examination were the presence of an ulcerative stomatitis and resistance to manipulation of both thoracic limbs. Neurological examination revealed normal mental status, a mild ataxia in all four limbs, and weakness predominantly in thoracic limbs. Conscious proprioception was decreased in the right pelvic and thoracic limbs. Spinal reflexes were normal in all limbs. Marked hyperesthesia was elicited with manipulation of the head and neck. On the basis of the neurological examination, a focal spinal cord lesion between the first cervical (C 1 ) and the fifth cervical (C 5 ) vertebrae was considered likely. Differential diagnoses included intervertebral disk disease, neoplasia, meningomyelitis, syringomyelia or hydromyelia, and diskospondilitis. Results of hematology and biochemical analyses were normal. Survey radiographs of the spine, thorax, and abdomen did not reveal any abnormalities. A cerebellomedullary myelogram was performed by injecting iopamidol a (0.3 ml/kg intravenously [IV]) into the subarachnoid space. Myelography indicated a marked, focal compression of the spinal cord 386 JOURNAL of the American Animal Hospital Association

62 September/October 2006, Vol. 42 Spinal Myxoid Liposarcoma 387 between C 1 and the second cervical (C 2 ) vertebrae. The contrast medium outlined a large filling defect within the dorsal aspect of the subarachnoid space, causing a golf tee sign [Figure 1]. The radiographic findings were consistent with an intradural-extramedullary mass. Computed tomography b (CT) of the neck was performed before and after administration of iodinated contrast medium c (2 ml/kg IV) in order to clarify the size, location, extent of the mass, and degree of spinal cord compression, and to assist with surgical planning. Results of plain CT showed no Figure 1 Lateral view of a cervical myelogram of a 9- year-old, spayed female Cavalier King Charles spaniel showing ventral deviation and loss of the dorsal contrast column (white arrows). An intradural filling defect ( golf tee sign, black arrow) occurred within the dorsal aspect of the subarachnoid space at the level of the first two cervical vertebrae. abnormalities. Images taken after injection of contrast medium revealed a large (6 9 mm), contrast-enhanced mass within the vertebral canal at the level of C 1-2. The mass was located dorsolaterally and was more prominent on the right side [Figure 2]. There was no evidence of bony changes, but marked spinal compression was evident. Treatment with prednisolone d (0.5 mg/kg per os [PO] q 24 hours) was initiated until an exploratory laminectomy could be performed 1 week later. Clinical signs improved with the treatment, but the dog still had proprioceptive deficits in both the right thoracic and pelvic limbs. The dog was anesthetized, and a dorsal, right-sided laminectomy was performed at the level of the atlas and axis. The laminectomy of the axis extended from the spinous process to the articular process, allowing removal of two-thirds of the length of the vertebra. The spinous process of the axis was not removed completely to avoid as much vertebral instability as possible. 5 One-half of the dorsal atlas was also removed. The interacuate ligament was partially eliminated to expose the spinal cord, and this resulted in significant hemorrhage that was controlled by gentle compression. A durotomy was performed to expose a firm, friable, and yellowish mass [Figure 3] located principally along the right dorsal aspect of the spinal cord. The mass did not appear to infiltrate the cord parenchyma. Abnormal tissue was removed as completely as possible by gentle dissection and submitted for histological evaluation. The laminectomy was covered with a fat graft and cervical musculature. Subcutaneous tissues and skin were apposed routinely. Postoperatively, the dog recovered rapidly and was discharged in 4 days. At that time, conscious proprioception was decreased in all four limbs, but the dog was able to walk unassisted and no neck pain was detected. Postoperative medications included prednisolone (0.5 mg/kg PO q 24 hours) for 10 days and amoxicillin/clavulanic acid e (15 mg/kg PO q 12 hours) for 1 week. Histopathology of the Figure 2 Postcontrast, transverse computed tomographic (CT) image at the first cervical (C 1 ) to second cervical (C 2 ) vertebral space of the dog in Figure 1. Note the contrastenhanced mass located dorsally (black arrows) and involving mainly the right side of the vertebral canal (white arrow). Widening of the vertebral canal and marked spinal cord compression are also present. R=right; L=left. Figure 3 Intraoperative photograph after durotomy, showing a firm mass in the subarachnoid space on the right side of the cervical spinal cord (arrows). The dog s head is to the right of the photograph.

63 388 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 mass revealed abundant and atypical myxoid stroma. Clusters of epithelioid cells with ovoid nuclei and nucleoli of variable sizes were regularly seen. Large, clear spaces were present in the extracellular compartment, and clear vacuoles were common within the tumor cells [Figure 4]. Material in the spaces and vacuoles stained poorly with periodic acid- Schiff stain, indicating the presence of glucosaminoglycans. Multinucleated giants cells were also seen. Histopathology suggested a poorly differentiated sarcoma (e.g., myxoid liposarcoma or myxoid chondrosarcoma) or a chordoma. Three weeks after surgery, radiation therapy of the site was performed because of the extensive nature of the mass, the poor cellular differentiation seen on histopathology, and the inability to achieve clear, wide margins at surgery. Radiation therapy was done with a linear accelerator f using bilaterally opposed portals and 4 4-cm lateral fields centered on the C 1-2 area. Treatment consisted of 33 gray (Gy) delivered in 11 fractions of 3 Gy each, administered on a Monday-Wednesday-Friday schedule. At a recheck examination 3 months following surgery, the dog was normal. A CT scan performed 6 months after surgery did not reveal any recurrence of the mass. Eighteen months after the surgery, the dog became reluctant to move its neck. A neurological examination at that time revealed only cervical spinal hyperesthesia. Recurrence of the tumor was suspected, and a third CT scan was performed. It revealed a 6 10-mm, contrast-enhanced mass located dorsolaterally in the vertebral canal in the area of C 1-2. The mass was most prominent on the left side and was compressing the spinal cord [Figure 5]. Following CT scan, a dorsal left-sided laminectomy was performed at the level of C 1-2, and a new mass was observed through the dura mater. After durotomy, a yellowish, well-circumscribed mass was exposed. The abnormal tissue was well demarcated cranially, caudally, and mediodorsally, but it seemed to infiltrate the spinal cord parenchyma laterally. The mass had the same appearance as the initial tumor. Complete excision of the mass was not performed to avoid excessive damage of the spinal cord. The dog was discharged 2 days following surgery, still exhibiting signs of mild neck hyperesthesia. Histopathological evaluation of the second mass revealed a myxomatous stroma. Rows and clusters of cells with abundant or eosinophilic, vacuolated cytoplasm were Figure 5 Postcontrast, transverse CT image at the level of C 1-2 vertebral space of the dog in Figure 1, 18 months after the initial surgery. A contrast-enhanced mass involving the left side of the spinal cord (arrows) indicated probable recurrence of the tumor. R=right; L=left. Figure 4 Photomicrograph of a poorly differentiated, intradural myxoid liposarcoma in the dog of Figure 1. Notice the rows and clusters of tumor cells (black arrows) surrounded by a myxoid matrix. Intracytoplasmic vacuoles are also seen (white arrows) (Hematoxylin and eosin stain, 100 ; bar=50 µm). Figure 6 Photomicrograph of the mass removed at the second surgery, showing a more differentiated myxoid liposarcoma. Some tumor cells had small, dark nuclei (black arrows) and an eosinophilic, vacuolated cytoplasm (white arrows) (Hematoxylin and eosin stain, 400 ; bar=200 µm).

64 September/October 2006, Vol. 42 Spinal Myxoid Liposarcoma 389 detected [Figure 6]. The tumor appeared similar to but more differentiated than the initial mass. Histopathological findings were consistent with a diagnosis of myxoid liposarcoma. A fourth CT scan was performed 1 month after the second surgery. A contrast-enhanced lesion measuring 2 7 mm was detected along the right side of the lateral vertebral canal. Additionally, another contrast-enhanced lesion measuring 3 9 mm was visualized in the dorsal vertebral canal. Spinal cord compression was not observed. These findings were thought to represent postoperative scarring and/or residual tumor. A second course of radiotherapy was performed. The dog was treated with 21 Gy delivered in 3 Gy fractions. A fifth CT scan performed 3 months after the radiotherapy did not show any abnormalities, and the dog was clinically normal at that time. Eleven months after the second surgery and 29 months after the initial presentation, the dog became depressed and reluctant to move its neck. On recheck examination, the dog was mildly tetraparetic and had marked cervical hyperesthesia. A sixth CT scan was performed and revealed two contrast-enhanced lesions located dorsally and laterally along the right side of the vertebral canal [Figures 7A-7C], suggesting recurrence of the original tumor. The dog was given prednisolone d (1 mg/kg PO q 24 hours), and another course of radiotherapy was instituted. This third treatment consisted of 15 Gy delivered in five fractions of 3 Gy each. After radiotherapy, no improvement was noted, so a third surgery was offered to evaluate the degree of spinal cord infiltration and the possibility of removing more tumor as a palliative measure. At surgery, a mass was found dorsally and laterally along the right side that was infiltrating the spinal cord almost completely. Only a small amount of tumor could be removed. Because of the poor prognosis, the owners elected euthanasia of the animal. Necropsy was not permitted. Histopathological evaluation of tissue removed at the third surgery was similar to the prior biopsies. Discussion Primary spinal cord tumors are relatively uncommon in dogs. 6,7 Based on anatomical location, spinal cord tumors can be classified as extradural, intradural-extramedullary, or intramedullary. 6,8,9 The intradural-extramedullary tumors are located in the subarachnoid space, within the dura mater but outside the spinal cord. Some studies have estimated that intradural-extramedullary tumors comprise approximately 35% of all spinal cord tumors of the dog, while extradural and intramedullary tumors in the dog represent 50% and 15%, respectively. 5,6 However, several other studies have reported approximately an equal distribution between extradural and intradural tumors. 8,10-12 The most common intradural-extramedullary neoplasms of dogs are the meningioma and peripheral nerve sheath tumor. 6,9 Other tumors with an intradural-extramedullary location that have been reported in dogs include nephroblastoma, lipoma, myxoma/myxosarcoma, glioma, mixed germ cell tumor, lymphoma, and metastatic tumors. 7,13-17 Liposarcoma is an uncommon neoplasm in dogs. 18,19 The spinal cord is a rare site for a liposarcoma. 2 While metastatic spinal liposarcomas have been reported in people, primary spinal liposarcomas have been poorly documented. 2 All liposarcomas in humans have had an extradural location. 2,3 A myxoid liposarcoma located extradurally at the level of the first two lumbar vertebrae has been reported in an 8-year-old, female Doberman pinscher. 4 On the basis of histological findings, the three main features of a myxoid liposarcoma are a well-developed vascular pattern, the presence of mononuclear cells, and a copious myxoid matrix containing abundant, nonsulfated glucosaminoglycans. 21 Myxoid liposarcomas usually contain cells with an abundant, eosinophilic, or vacuolated cytoplasm; a large nucleus; and a small nucleolus. 18,22 In the case reported here, the tumor was classified as a myxoid liposarcoma based primarily on histological findings of myxoid areas associated with cells having prominent cytoplasmic vacuolization, and a vascular pattern typical of liposarcoma. The goals of the surgery for spinal cord tumors in dogs are to improve spinal cord function by decompression, to remove as much tumor as possible, and to collect tissue for histopathological evaluation. 23,24 In the case reported here, the extensive involvement of the tumor did not allow wide, clear margins to be achieved; therefore, radiation therapy was also recommended. The primary purpose of radiotherapy for spinal cord tumors is to decrease the potential for recurrence following incomplete excision of the mass. 6,23 Although the efficacy of postoperative irradiation of spinal cord tumors in dogs is not well documented, previous reports have suggested it may be beneficial Postoperative radiotherapy for the treatment of spinal sarcomas is limited by the presence of the spinal cord. 28 Data in humans suggest that the threshold for radiation injury of the spinal cord is between 45 and 50 Gy using conventional fractionation. 4,29,30 Radiation doses that are considered effective for eradication of microscopic residual tumors are in the range of 60 to 70 Gy. 4,28 Despite suboptimal doses, however, postoperative radiation seems to delay local recurrence and is routinely recommended in humans after resection of high-grade tumors. 3 In dogs, liposarcoma has been classified as a tumor with poor responsiveness to radiation therapy. 20 In contrast, combined surgery and radiotherapy has achieved local control of liposarcomas in most affected humans and is the therapeutic preference for this tumor. 31 The long-term prognosis for either humans or animals with primary spinal myxoid liposarcoma has not been well established. 28 In the previously mentioned report of spinal liposarcoma in a dog, surgical resection alone provided 7 months of clinical remission before recurrence of the tumor. 4 In the dog reported here, surgical therapy was considered inadequate, because clean margins could not be achieved. Another stimulus for considering radiation therapy was the poor differentiation noted on histopathological examination of the tumor after the first surgery. 32 The recommended dosage for the treatment of spinal cord tumors in people is a total dose of 45 to 50 Gy given in

65 390 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Figures 7A, 7B, 7C Postcontrast, transverse (7A, 7B) CT images taken of the dog in Figure 1, 11 months after a second course of radiotherapy and 29 months after initial presentation. Note the presence of two contrast-enhanced lesions (arrows) located dorsally (7A) and laterally (7B) along the right side of the vertebral canal, indicating probable recurrence of the tumor. The sagittal reconstructed CT image (7C) allowed better visualization of the extent of the mass (arrows). R=right; L=left. 1.8 to 2.0 Gy daily fractions. 30 Data regarding radiation dosages for spinal cord tumors in dogs are limited. In one study, the thoracic spinal cords of 42 dogs were irradiated. 33 The doses delivered were 0, 44, 52, 60, or 68 Gy in 4 Gy fractions, and dogs were evaluated 1 and 2 years after radiation. 33 Dosages of 60 Gy given in 4 Gy fractions were associated with clinical signs. In dogs with spinal cord tumors, it is desirable to administer as large a total dose as possible without inducing radiation myelopathy. 25 In the case reported here, a total dose of 69 Gy was delivered in 3 Gy daily fractions. The doses were delivered over a 30- month period, and the three treatments were separated by 18 and 11 months. No clinical signs attributable to radiation myelopathy were observed during the 30 months of followup. Unfortunately, necropsy was not permitted in this dog, so the presence of secondary radiation injury could not be assessed. This is the second report of spinal liposarcoma in a dog and the first report of an intradural location. It is possible the tumor originated from epidural fat. Another explanation for the location is that the inner layer of the dura mater and the arachnoidea are of mesodermal origin, and this may give rise to neoplastic adipose tissue. 14 Although a necropsy was not performed, it was possible that the spinal tumor was a metastatic lesion but was considered primary, since no other lesions were discovered during the long treatment period.

66 September/October 2006, Vol. 42 Spinal Myxoid Liposarcoma 391 Conclusion An intradural-extramedullary myxoid liposarcoma was diagnosed in the region of the C 1-2 vertebrae of a 9-year-old dog. Treatment included three dorsal laminectomies and three courses of radiation therapy, and the dog survived for 30 months. Further investigations are indicated to evaluate the efficacy of postoperative radiation therapy for dogs with spinal cord tumors that cannot be completely removed at surgery. a Iopamidol 30`62g; Laboratoire Shering A.G. Allemagne, France b HISpeed CT/E Dual; GE Medical Systems, Milwaukee, WI c Sodium ioxitalamate; Guerbet, Villepinte, France d Megasolone; Merial, Lyon, France e Synulox; Pfizer, Paris, France f Orion, linear accelerator (5 mv, 760 S); GE Medical Systems, Milwaukee, WI Acknowledgments The authors thank Dr. Mercedes Estrada and the histopathology laboratory (Laboratoire d histo-cytopathologie vétérinaire, 95 rue raspail, Maisons Alfort, France) for their assistance with the photomicrographs and the preparation of this manuscript. References 11. Estourgie SH, Nielsen GP, Ott MJ. Metastatic patterns of extremity myxoid liposarcoma and their outcome. J Surg Oncol 2002;80: Turanli S, Ozer H, Ozyürekoglu T, et al. Liposarcoma in the epidural space. Spine 2000;25: Bilsky MH, Boland PJ, Panageas KS, et al. Intralesional resection of primary and metastatic sarcoma involving the spine. Neurosurgery 2001;49: Lewis DD, Kim DY, Paulsen DB, et al. Extradural spinal liposarcoma in a dog. J Am Vet Med Assoc 1991;199: Fingeroth JM, Smeak DD. Laminotomy of the axis for surgical access to the cervical spinal cord. Vet Surg 1989;18: Lecouteur RA. Tumors of the nervous system. In: Withrow S, MacEwen GE, eds. Small Animal Clinical Oncology. Philadelphia: WB Saunders, 2001: Levy MS, Kapatkin AS, Patnaik AK, et al. Spinal tumors in 37 dogs: clinical outcome and long term survival ( ). J Am Anim Hosp Assoc 1997;33: Wright JA. The pathological features associated with spinal tumors in 29 dogs. J Compar Pathol 1985;95: Gilmore DR. Neoplasia of the cervical spinal cord and vertebrae in the dog. J Am Anim Hosp Assoc 1983;19: Luttgen PJ, Braund KG, Brawner WR, et al. A retrospective study of twenty-nine spinal tumors in the dog and cat. J Small Anim Pract 1980;21: Kippenes H, Gavin PR, Bagley RS, et al. Magnetic resonance imaging features of tumours of the spine and spinal cord in dogs. Vet Radiol Ultrasound 1999;40: Drost WT, Love NE, Berry CR. Comparison of radiography, myelography, and computed tomography for the evaluation of canine vertebral and spinal cord tumors in 16 dogs. Vet Radiol Ultrasound 1996;37: Summers BA, delahunta A, McEntee M, et al. A novel intraduralextramedullary spinal cord tumor in young dogs. Acta Neuropathol (Berl) 1988;75: Umphlet RC, Vinici DS, Godshalk CP. Intradural-extramedullary lipoma in a dog. Compend Contin Educ Pract Vet 1989;11: Hara Y, Nezu Y, Harada Y, et al. Secondary chronic respiratory acidosis in a dog following the cervical cord compression by an intradural glioma. J Vet Med Sci 2002;64: Ferreira AJ, Peleteiro MC, Carvalho T, et al. Mixed germ cell tumor of the spinal cord in a young dog. J Small Anim Pract 2003;44: Jeffery ND. Ancillary aids. In: Slatter D, ed. Textbook of Small Animal Surgery. London: WB Saunders, 1995: Straffus AC. Liposarcoma in dogs. J Am Anim Hosp Assoc 1973;9: Baez JL, Hendrick MJ, Shofer FS, et al. Liposarcomas in dogs: 56 cases ( ). J Am Vet Med Assoc 2004;224: MacEwen GE, Powers BE, Macy D, et al. Soft tissue sarcomas. In: Withrow S, MacEwen GE, eds. Small Animal Clinical Oncology. Philadelphia: WB Saunders, 2001: Enzinger FM, Weiss SW. Liposarcoma. In: Enzinger FM, Weiss SW, eds. Soft Tissue Tumors. 3rd ed. St Louis: Mosby, 1995: Wakely PE. Myxomatous soft tissue tumours: correlation of cytopathology and histopathology. Ann Diagn Pathol 1999;3: Wheeler SJ, Sharp NJ. Neoplasia. In: Enzinger FM, Weiss SW, eds. Soft Tissue Tumors. London: Mosby-Wolfe, 1997: Luttgen PJ. Spinal neoplasia: diagnosis and treatment. Seminar Vet Med Surg 1990;5: Siegel S, Kornegay JN, Thrall DE. Postoperative irradiation of spinal cord tumors in 9 dogs. Vet Radiol Ultrasound 1996;37: Bell FW, Feeney DA, O Brien TJ, et al. External beam radiation therapy for recurrent intraspinal meningioma in a dog. J Am Anim Hosp Assoc 1992;28: Dickinson PJ, Mc Entee MC, Lipsitz D, et al. Radiation-induced vertebral osteosarcoma following treatment of an intraduralextramedullary spinal cord tumor in a dog. Vet Radiol Ultrasound 2001;42: Merimsky O, Kollender Y, Bokstein F, et al. Radiotherapy for spinal cord compression in patients with soft-tissue sarcoma. Int J Radiat Oncol Biol Phys 2004;58: Garcia DM. Primary spinal cord tumors treated with surgery and postoperative irradiation. Int J Radiat Oncol Biol Phys 1985;11: Lindstadt DE, Wara WM, Leibel SA, et al. Postoperative radiotherapy of primary spinal cord tumors. Int J Radiat Oncol Biol Phys 1989;16: Zagars GK, Goswitz MS, Pollack A. Liposarcoma: outcome and prognostic factors following conservation surgery and radiation therapy. Int J Radiat Oncol Biol Phys 1996;36: Gillete EL, Gillete SM. Principles of radiation therapy. Seminar Vet Med Surg 1995;10: Powers BE, Beck ER, Guillette EL, et al. Pathology of radiation injury to the canine spinal cord. Int J Radiat Oncol Biol Phys 1992;23:

67 Unilateral Leiomyoma in the Mesosalpinx of a Dog Routine ultrasonographic evaluation of the genital organs of a 3-year-old terrier bitch revealed a mass at the level of the left ovary. The mass was located next to the caudal pole of the left kidney and ventrocaudal to the left ovary. Ultrasonographically, the uterus was not enlarged and had no luminal contents. Exploratory laparotomy revealed a mass attached to the left ovarian bursa with a small and thin pedicle. The mass had smooth margins, was whitish in color, and was lobulated on cut surface. The histopathological diagnosis of the mass was leiomyoma associated with the mesosalpinx. J Am Anim Hosp Assoc 2006;42: Kemal Eker, DVM Mehmet Rifat Salmanoglu, DVM, PhD Sevil Atalay Vural, DVM, PhD CR From the Department of Obstetrics and Gynecology, Faculty of Veterinary Medicine, Ankara University, 06110, Diskapi, Ankara, Turkey. Address all correspondence to Dr. Salmanoglu. Introduction Leiomyomas are reported to be noninvasive, nonmetastatic, and slowgrowing structures. 1 They rarely cause clinical signs. 1 In some cases, leiomyomas become extremely enlarged and may exert pressure on the internal organs and cause pain. 2 Usually they are coincidentally detected during necroscopy or ovariohysterectomy. 2 Leiomyomas and leiomyosarcomas of mesenchymal origin may be located within the digestive system, heart, ovaries, uterus, and vulva of dogs and cats. 2 In one report of canine uterus tumors of mesenchymal origin, leiomyomas accounted for 85% to 90%, and leiomyosarcomas accounted for 10% of the masses. 2 In a retrospective study involving 49 dogs with ovarian tumors, only one (2.04%) bitch had a leiomyoma of ovarian origin. 3 The purpose of the case reported here is to present the ultrasonographic evaluation of a rare case of leiomyoma located at the mesosalpinx of a dog. Case Report A 3-year-old, intact, female terrier was presented to the Obstetrics and Gynecology Department of the Veterinary Faculty of Ankara University for routine gynecological examination. On physical examination, there was no vulvar edema or discharge. On vaginoscopy, the cervix was closed, and there was no mucosal folding or edema of the vaginal mucosa. The general condition and appetite of the dog were good. The uterus and ovaries were ultrasonographically evaluated by the use of a linear array, 8-MHz a probe. On ultrasonography, the uterus was normal. A mass was visualized ventrocaudal to the left kidney (i.e., at the level of the ventral edge of the kidney), and it had lower echogenicity than that of the renal cortex [Figure 1]. The mass was easily distinguished from the neighboring tissues by its size ( cm), oval shape, and irregular borders on the side nearest the kidney. The mass was not attached to the kidney. The mass contained two oval, cyst-like structures, with anechoic centers that were 0.51 cm and 0.55 cm in size. The mass appeared to mimic an ovary with follicular structures. Acoustic enhancement was noted within the distal part of the mass. Further examination 392 JOURNAL of the American Animal Hospital Association

68 September/October 2006, Vol. 42 Mesosalpinx Leiomyoma 393 revealed that the mass was just ventral to the left ovary [Figure 1] and was thought to be adherent to the ovary. No stalk was observed between the mass and left ovary. The size of the left ovary was cm. The echogenicity of the ovary was significantly higher than that of the mass. No follicular structures were detected on the left ovary. Figure 1 The ultrasonographic appearance of a mass near the left ovary in a 3-year-old, female terrier. The mass (white arrow and between cursors) is located at the ventrocaudal edge of the ipsilateral kidney (K). It is hypoechoic compared to the renal cortex and can be differentiated from the surrounding tissues by its hyperechogenic borders. The left ovary is located dorsal to the mass (between the arrowheads). (The bottom of the picture is the ventral abdomen.) The right ovary was easily imaged. It was smooth, ovalshaped, located near the caudal pole of the kidney, and had similar echogenicity to the renal cortex. Structures with lower echogenicity than the ovarian parenchyma were detected and were compatible with regressing corpora lutea. Considering the probability that the mass was a tumor, the rest of the abdomen was thoroughly evaluated. The intestinal loops observed around the structure appeared normal. There was no corrugation of the intestines, and no ascites was observed. 4 The ipsilateral kidney was normal. Vaginal samples were stained via the Papanicolaou method, revealing abundant parabasal and small intermediary cells and a few neutrophils. No erythrocytes were detected. A blood sample was submitted for hormonal assays. Serum progesterone and 17 ß estradiol values were determined by radioimmunoassay (RIA) b and were normal for late diestrus (2.76 ng/ml, reference range 2 to 10 ng/ml; 3 pg/ml, reference range 2 to 7 pg/ml, respectively). Based on vaginoscopic, vaginal cytological, and hormonal test results, the mass was considered to be nonfunctional tumor adherent to the ovary. Exploratory laparotomy revealed no fluid accumulation in the abdomen. The mass was connected to the ovarian bursa of the left ovary by a thin, small stalk. No adhesions to any other organs or adjacent structures were found, and there were no apparent metastatic lesions on the intestines, kidneys, or surrounding tissues. Both ovaries and the uterus were surgically removed along with the mass. The mass was fixed in 10% buffered formalin for routine histopathological evaluation by the Department of Pathology, Faculty of Veterinary Medicine, Ankara University. Grossly, the mass was cm in size, with a weight of 5 g [Figures 2A, 2B]. The mass was smooth and well defined. On microscopic examination, the mass was composed of oval or round cells that swirled in various directions [Figure 3A]. Based on the results of Masson s trichrome staining [Figure 3B], the cells were identified as smooth muscle cells, and the diagnosis was established as leiomyoma. Discussion Leiomyomas and leiomyosarcomas of the canine genital system rarely cause clinical signs and are usually detected coincidentally during necroscopy or ovariohysterectomy. 5 Similarly, the leiomyoma in the case presented here was coincidentally detected during ultrasonography conducted during routine gynecological examination. The dog was clinically asymptomatic, and vaginoscopic, vaginal cytological, and hormonal evaluations indicated no sexual activity. Leiomyomas in bitches have been reported to produce estrogen and progesterone. 1 The neoplastic transformation of myometrium into a leiomyoma in humans may involve somatic mutation of normal myometrium and complex interactions of local growth factors and ovarian steroids. 6 The estrogen and progesterone levels in the dog reported here were within reference ranges, which suggested that the mass was not secreting significant levels of estrogen or progesterone. Because the mass was located in the mesosalpinx and had no direct relationship to the ovary, it would likely have been hormonally inactive. Ultrasonography is a reliable tool for the detection of abdominal masses. 7,8 The ultrasonographic appearance of leiomyomas of the genital organs of dogs has not been previously described. Ovarian and uterine tumors have been reported to be heterogeneous on ultrasonography, regardless of whether they are malignant or benign. 8,9 In the present case, the hypoechoic appearance of the leiomyoma allowed both its definition and detection. Conclusion A leiomyoma of the mesosalpinx was detected on abdominal ultrasonography in a 3-year-old, female terrier. The leiomyoma was not associated with the production of steroid hormones or any clinical signs. Definitive diagnosis was achieved by histopathological evaluation, although ultrasonography was a reliable diagnostic tool for identification and localization of the mass. a 100 Falco Vet; Pie Medical Equipment B.V., Philipsweg 1, 6227 AJ, The Netherlands b Immunotech; A Beckman Coulter, BP Marseille Cedex 9, France

69 394 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 A A B B Figures 3A, 3B (A) Histopathological appearance of the mass in Figure 2, showing smooth muscle cells coursing in various directions (Hematoxylin and eosin stain, bar=50 µm). (B) The appearance of smooth muscle tumor cells (Masson s trichrome stain, bar=50 µm). Figures 2A, 2B (A) The gross appearance of the ovaries (white arrows), uterus (black arrows), and leiomyoma (at the scalpel blade point; blade length is 3.9 cm). (B) The leiomyoma is connected to the mesosalpinx with a thin stalk (arrow). (Gray arrow=bursa ovarica) References 11. Klein MK. Tumors of the female reproductive system. In: Kersey R, ed. Small Animal Clinical Oncology. Philadelphia: WB Saunders, 2001: Withrow SJ, Susaneck SJ. Tumors of the canine female reproductive tract. In: Morrow DA, ed. Current Therapy in Theriogenology. 2nd ed. Philadelphia: WB Saunders, Sforna M, Brachelente C, Lepri E, et al. Canine ovarian tumours: a retrospective study of 49 cases. Vet Res Comm 2003;Suppl 27: Moon ML, Biller DS, Armbrust LJ. Ultrasonographic appearance and etiology of corrugated small intestines. Vet Radiol Ultrasound 2003;44: Herron MA. Tumors of the canine genital system. J Am Anim Hosp Assoc 1983;19: Rein MS, Barbieri RL, Friedman AJ. Progesterone: a critical role in the pathogenesis of uterine myomas. Am J Obstet Gynecol 1995;172 (1Pt 1): Nyland TG, Matton JS, Wisner ER. Physical principles, instrumentation and safety of diagnostic ultrasound. In: Nyland TG, Motton JS, eds. Veterinary Diagnostic Ultrasound. Philadelphia: WB Saunders, 1995: Diez-Bru N, Garcia-Real I, Martinez EM, et al. Ultrasonographic appearance of ovarian tumors in 10 dogs. Vet Radiol Ultrasound 1998;39: Poffenbarger EM, Feeney DA. Use of gray scale ultrasonography in the diagnosis of reproductive disease in the bitch: 18 cases ( ). J Am Vet Med Assoc 1986;189:90-95.

70 Ectopic Ureterocele in a Male Dog: A Case Report and Review of Surgical Management A 16-week-old, male border terrier was presented for urinary incontinence. Intravenous urography demonstrated a right-sided, extravesical ectopic ureterocele. Neoureterocystostomy and ureterocele omentalization were performed. Urinary incontinence persisted after surgery. Retrograde urethrography revealed communication between the ureterocele and urethra. Urinary incontinence resolved following partial ureterocelectomy and reconstruction of the proximal urethra. J Am Anim Hosp Assoc 2006;42: James A. Tattersall, BVSc, BSc, MRCVS Elizabeth Welsh, BVMS, PhD, MRCVS CR From the Department of Small Animal Surgery, Royal (Dick) School of Veterinary Studies, Division of Veterinary Clinical Studies, The University of Edinburgh, Roslin, Midlothian, EH25 9RG Scotland, United Kingdom. Introduction A ureterocele is defined as a cystic dilatation of the intravesicular submucosal segment of the distal ureter. 1 Ureteroceles have been commonly reported in humans. 2 In children, the incidence of ureterocele is as high as one in 500, although the incidence of clinically significant ureterocele in children is reported to be one in 5000 to 12,000. 2,3 Reports of ureteroceles in small animals are few. Twelve case reports in dogs and a single case report in a cat have been published between 1971 and ,4-15 A further five cases have been documented in the veterinary literature. 16 Although diagnosed infrequently, clinical signs associated with ureteroceles in dogs may include dysuria, stranguria, pollakiuria, and urinary incontinence. 17 The presence of ureteroceles in male dogs has been described in three cases. Two dogs were Siberian huskies with intravesical ureteroceles, and the third was a Labrador retriever that developed an ectopic ureterocele secondary to surgical correction of a right ectopic ureter. 6,8,12 To the authors knowledge, there are no reported cases of congenital ectopic ureteroceles in male dogs. Ureteronephrectomy, transurethral endoscopic incision, neoureterocystostomy, and ureterocelectomy are all previously reported surgical techniques used to manage ectopic ureteroceles in dogs. 4,5,7-9,13,15,16 Omentalization is the surgical application of omental grafts, which utilize the immunogenic, angiogenic, adhesive, and lymphatic properties of the omentum to fill dead space, induce vascularization, enhance physiological drainage, provide support for grafted tissues, and adhere to and reinforce other organs during healing. 18,24,25 Although the use of omentum to treat ectopic ureteroceles in dogs has not previously been reported, its application for abdominal, urogenital, thoracic, and reconstructive surgery has been described This report describes the use of neoureterocystostomy and ureterocele omentalization for the surgical treatment of a young, male dog with urinary incontinence associated with a congenital ectopic ureterocele. Neoureterocystostomy and ureterocele omentalization failed to correct the urinary incontinence, and a second surgical pro- JOURNAL of the American Animal Hospital Association 395

71 396 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 cedure to perform partial ureterocelectomy was necessary before urinary incontinence resolved. Case Report A 16-week-old, male border terrier was presented for investigation and treatment of urinary incontinence that had been present from 6 weeks of age. The dog dribbled urine continuously, which was most noticeable when the dog was asleep. Voluntary urination was considered normal, and there was no stranguria, dysuria, or hematuria. On physical examination, penile glandular hypospadia was present, and persistent dribbling of urine was evident. A serum biochemical profile demonstrated normal blood urea nitrogen (14.6 mg/dl, reference range 4.8 to 21 mg/dl), creatinine (0.98 mg/dl, reference range 0.45 to 1.5 mg/dl), total protein (5.9 g/dl, reference range 5.8 to 7.3 g/dl), potassium (4.2 meq/l, reference range 3.6 to 5.6 meq/l), sodium (146 meq/l, reference range 139 to 154 meq/l), and phosphorus (3.0 mg/dl, reference range 2.7 to 6.2 mg/dl). Urine analysis was normal. Urine specific gravity was No bacteria were cultured from a cystocentesis sample of urine. Intravenous (IV) urography was performed under general anesthesia. Diuresis was maintained by administration of IV lactated Ringer s solution a at 10 ml/kg per hour. Initially a pneumocystogram was performed by inflating the bladder with 50 ml of carbon dioxide through a 6-French urinary catheter. b This was followed by an IV injection (600 mg iodine/kg) of iothalamic acid. c Ventrodorsal abdominal radiographs were taken at 2 and 5 minutes postinjection and were followed by right lateral and oblique abdominal radiographs at 10 and 15 minutes. On pneumocystogram, the bladder appearance and location were normal. A radiolucent, spherical structure was evident immediately caudal to the bladder neck at the pelvic inlet. Intravenous urography demonstrated mild dilatation of the distal third portion of the right ureter. The right kidney, renal pelvis, and proximal ureter were radiographically normal. A well-circumscribed, spherical structure (approximately mm in diameter) that filled with contrast material was present caudal to the bladder neck [Figure 1]. Fluoroscopy confirmed a direct communication of the spherical structure with both the distal part of the right ureter and proximal urethra. The left kidney, ureter, and ureterovesical junction were normal. These radiographic features were highly suggestive of a right extraluminal ectopic ureter and ectopic ureterocele. After routine surgical preparation, a ventral midline celiotomy was performed. Cranial traction was applied to the bladder using stay sutures to improve exposure of the bladder neck and prostate. A combined ventral cystotomy and proximal urethrotomy was performed, extending caudally into the prostatic urethra. Multiple openings in the dorsal wall of the prostatic urethra were found to communicate directly with the ureterocele [Figure 2]. The right ectopic ureter was ligated at the level of the bladder neck with transfixing and circumferential 1.5 metric polypropylene d sutures. A neoureterocystostomy was created midway Figure 1 Intravenous urogram in a 16-week-old, male border terrier, showing a mildly dilated, distal right ureter (thick arrow) communicating with an ectopic ureterocele (thin arrows) of the proximal urethra. Figure 2 Intraoperative photograph of the same dog as in Figure 1, showing the bladder (thick arrow) and prostate (open arrows) following cystostomy. Multiple openings in the wall of the prostatic urethra can be seen (thin arrows), which communicate with the ectopic ureterocele. between the trigone and apex of the bladder on the right side. The spatulated end of the ureter was sutured to the bladder mucosa using 1 metric poliglecaprone e in a simple continuous suture pattern. The openings between the ureterocele and prostatic urethra were excised, and the edges of the urethral mucosa were apposed with 1 metric poliglecaprone in a simple continuous suture pattern. A stab incision through the wall of the ureterocele was made from the serosal surface and omentum packed into the cavity, similar to that described for prostatic omentalization. 25 The urethra and cystotomy were repaired with 1 metric poliglecaprone in a simple continuous suture pattern, and the abdomen was

72 September/October 2006, Vol. 42 Ectopic Ureterocele 397 lavaged and closed routinely. A 6-French Foley f urinary catheter was advanced through the urethra into the bladder at the end of surgery and was connected to a closed-collection system. Postoperative analgesia was provided by buprenorphine g (15 µg/kg IV q 8 hours) and carprofen h (2 mg/kg per os [PO] q 12 hours). The urinary catheter was removed 36 hours after surgery. At the time of discharge 5 days after surgery, the dog was urinating voluntarily and had no signs of incontinence or stranguria. The owner reported no urinary incontinence for 3 weeks, but then the incontinence recurred. A retrograde positivecontrast urethrogram was performed under sedation with medetomidine i (10 µg/kg IV) 6 weeks after the initial surgery. Iothalamic acid (10 ml) injected through a 6-French urinary catheter showed persistence of the ureterocele [Figure 3]. Ultrasonography of the bladder identified the reimplanted right ureter, and color-flow Doppler ultrasonography confirmed normal ejection of urine. On ultrasonography, the ureterocele was seen as a hypoechoic structure extending caudal to the bladder neck, with a narrow communication to the prostatic urethra. A second ventral midline celiotomy was performed. The ectopic ureterocele was identified in the same position as before. Omentum was adhered to the outside surface of the ureterocele. A combined cystotomy and urethrotomy was performed as described previously. The previous surgical repair of the dorsal wall of the urethra had dehisced completely, and the ureterocele was in direct communication with the prostatic urethra. A small amount of omentum was still located within the lateral aspect of the ureterocele. The edges of the deficit in the dorsal wall of the urethra were excised. The dorsolateral wall of the ureterocele was brought through the opening and partially excised (ureterocelectomy). The deficit in the dorsal wall of the prostatic Figure 3 Six weeks after the first surgery, a positive-contrast, retrograde urethrogram demonstrated persistence of the ureterocele (thick arrows) in the dog from Figures 1 and 2. The distal portion of the ligated, right ectopic ureter can also be seen (thin arrow) communicating with the ureterocele (open arrow). urethra was reconstructed with a flap created from the lateral ureterocele wall sutured to the axial border of the urethral opening with a full-thickness, 1 metric poliglecaprone simple continuous suture [Figure 4]. Complete excision of the dorsal portion of the ureterocele was not performed, in order to minimize iatrogenic damage to the neural and vascular structures dorsal to the prostate, including the pelvic and prostatic plexus, prostatic and urethral veins, and branches of the prostatic artery (caudal vesicular artery and artery of the ductus deferens). 27 The ureterocele remnant was omentalized using a technique similar to that described for partial resection and omentalization of prostatic retention cysts in dogs. 24 Urethral patency was checked by antegrade and retrograde urethral catheterization. Routine closure of the cystotomy, urethrotomy, and celiotomy was performed as described previously. Postoperatively, the dog had signs of stranguria. To ensure patency of the proximal lumen of the urethra, an indwelling, silicone-coated latex, 6-French Foley urinary catheter attached to a closed-collection system was inserted into the bladder and maintained for 1 week. Daily cytology of urine sediment was performed to assess for iatrogenic urinary tract infection. The urinary catheter was removed 1 week after surgery, and by the 10th postoperative day the dog was continent and capable of passing continuous streams of urine. One month after the second surgery, the dog had no signs of urinary incontinence, stranguria, or dysuria. Follow-up ultrasonography showed resolution of the ureterocele. The penile glandular hypospadia was surgically corrected by preputial reconstruction when the dog was 7 months old, at which stage the dog remained clinically normal. 28 Discussion Ureteroceles are defined as cystic dilatations of the intravesicular, submucosal segment of the distal ureter. 1 Dysembryogenesis of the ureteral bud is most likely involved in the etiology, but no single hypothesis has yet been formulated that completely explains all aspects of ureterocele formation The classification of ureteroceles in both human and veterinary literature is unclear. Ureteroceles have been categorized in a variety of ways, based on location, morphology, and the presence of other associated anatomical abnormalities. 1,14,32 In humans, ureteroceles are most commonly associated with a duplex kidney (i.e., a kidney that can be divided into upper and lower poles, each drained by a separate renal pelvis and associated ureter). 33 None of the reported cases of ureteroceles in dogs described the presence of a duplex kidney, and there was no gross evidence of renal duplication in the current case. 1,4-16 Single-system ureteroceles (i.e., ureteroceles associated with a kidney having only one renal pelvis and ureter) appear to predominate in animals; in humans, singlesystem ureteroceles are rare. 34 The Committee on Terminology, Nomenclature, and Classification of the Section of Urology of the American Academy of Pediatrics introduced standardized terms to

73 398 JOURNAL of the American Animal Hospital Association September/October 2006, Vol. 42 Figure 4 Schematic illustrations showing the second surgery performed in the dog from Figures 1-3, 6 weeks after neoureterocystostomy and ureterocele omentalization. Diagram A shows the bladder (B), prostate (P), reimplanted right ureter (RU), left ureter (LU), vas deferens (VD), and ligated distal portion of the right ureter (LRU) following cystostomy and urethrostomy. The ureterocele (U) communicated with the urethra (Ur) via multiple openings (UO) in the region of the prostate. Section a1: a transverse section through the prostate and ureterocele (a-a). Section a2: following partial resection of the ureterocele, a flap (F) created from the lateral ureterocele wall was used to close the communication between the urethra and ureterocele. Section a3: omentalization of the ureterocele remnant by suturing of omentum (Om) to the outer layer of the urethral repair and remainder of the ureterocele wall. classify human ureteroceles as either intravesical or ectopic. 35 Intravesical ureteroceles are defined as cystic dilatations of the ureter in the submucosa of the bladder wall, which are contained entirely within the bladder, and have an orifice opening into the bladder in the region of the trigone. 35 If any portion of the ureterocele is situated permanently at the bladder neck or urethra, regardless of the position of the orifice, the ureterocele is termed ectopic. 35 This classification system was used to define the ureterocele in the case reported here, and it has also been used previously to classify ureteroceles in dogs. 17 Other classification systems adapted from humans have been described in the veterinary literature and include abnormalities that may be seen concurrently with ureteroceles, such as hydroureter, hydronephrosis, and chronic renal disease. 14,32 Application of this classification system to previously reported cases of canine ureteroceles suggests that it might also be predictive of response to surgical intervention, with outcome being related to the severity of concurrent abnormalities. 14 The ureterocele presented in this report was associated with an ectopic, right ureter that opened into the prostatic urethra, and therefore was classified as an ectopic ureterocele. This report describes the management of a congenital ectopic ureterocele in a male dog. Of the previously reported cases in dogs, ureteroceles appear to occur most frequently in females (ratio 3.5:1). 1,4-16 Thirteen were classified as ectopic, and six were intravesical. 1,4-10,12-16 Ectopic ureteroceles associated with duplex kidneys have not been reported in dogs, and as such, the upper-pole heminephrectomy commonly used to manage human ectopic ureteroceles is not indicated. 36 Previously reported surgical techniques used in dogs to manage ectopic ureteroceles include transurethral endoscopic incision, ureteronephrectomy, ureterocelectomy, and neoureterocystostomy. 4,5,7-9,13,15,16 Following neoureterocystostomy in the dog reported here, the orifice of the ureterocele was identified, and primary closure of the urethral mucosa was performed. The ureterocele was then omentalized through a stab incision made from the serosal surface of the ureterocele, similar to that described for prostatic omentalization. 25 The rationale for ureterocele omentalization was based on its ability to obliterate dead space within cavitated lesions, provide physiological drainage, enhance vascularization, and promote adhesion formation that would provide a seal and help contribute to the strength of the mucosal repair. 18,24,25 Although the use of omentum has not been described for the treatment of ureteroceles in dogs, it has successfully been used in the reinforcement of gastrointestinal anastomosis; reconstruction of chronic, nonhealing wounds; treatment of pancreatic, prostatic, uterine stump, and lymph node cysts and abscess; and as a physiological drain for the treatment of idiopathic chylothorax Ureterocelectomy was not performed during the initial surgery, as complete resection would have required extensive dissection dorsal to the prostate, potentially resulting in iatrogenic damage to the neural and vascular structures in this region. 27 The prostatic plexus forms the middle portion of the pelvic plexus, derived from the pelvic and hypogastric nerves, and penetrates the dorsolateral capsule of the gland. 27 Sympathetic innervation is supplied to the detrusor muscle and smooth muscle of the urethra. 37 Stimulation of alpha-adrenergic fibers results in contraction of the smooth muscle in the trigone and proximal urethra and the formation