Inger Lilliehöök and Harold Tvedten Clinical Pathology Department of Clinical Sciences and University Animal Hospital (UDS) Swedish University of the Agricultural Sciences (SLU) Uppsala Sweden SIGNALMENT: A 10-year-old, female Flat-coated retriever living south of Stockholm, Sweden. History and Clinical Findings She has had poor appetite, polydipsia, sporadic vomiting and been depressed for two weeks. Ticks have been found on her earlier. At admittance May 24 to an Animal Clinic in Stockholm she had an elevated body temperature 39.6 o C and was slightly depressed. She was otherwise in good condition, had normal appearing mucous membranes, normal heart and lung sounds and showed no pain on palpation of her abdomen. Anaplasma phagocytophilium infection was suspected based on fever, thrombocytopenia and increased CRP (218mg/L, reference value <7mg/L), but no Anaplasma inclusions were seen on blood smear examination. She was treated with doxycycline despite not seeing Anaplasma inclusions. The clinical signs did not improve and she went back to the clinic for a second examination May 31. She was slightly more alert after one week of treatment, but still had fever (39.2 o C, 39.4 o C) and polydipsia. At admittance to the clinic she had slightly pale mucous membranes but no other specific clinical findings. CRP was 135 mg/l (reference value <7mg/L) Serum biochemistry May 24 and 31: ALP was slightly increased, while ALAT, creatinine, glucose, Na and K were within normal ranges. EDTA blood samples were sent to the SLU University Animal Hospital for evaluation because of a leukocytosis and abnormal blood picture both from May 31 and June 16. Both samples had a similar and unusual WBC dot plot pattern on the Advia 2120 graphics. HEMATOLOGY DATA: Advia Automated Differential WBC Counts Test May 31* June 16* Reference WBC Basophil (default) 24.6 x 10 9 /L 38.0 x 10 9 /L 5.8-16.0 WBC Peroxidase 18.2 x 10 9 /L 22.2 x 10 9 /L 5.8-16.0 Neutrophils 7.1 x 10 9 /L 8.3 x 10 9 /L 3.3-10.4 Lymphocytes 11.9 x 10 9 /L 19.3 x 10 9 /L 1.5-4.7 Monocytes 2.6 x 10 9 /L 4.9 x 10 9 /L 0.1-1.0 Eosinophils 0.6 x 10 9 /L 2.0 x 10 9 /L 0.2-1.6 LUC 2.3 x 10 9 /L 2.9 x 10 9 /L 0-0.2 LUC is large unstained cell category for the Advia 2120. *Instrument leukocyte alarms: WBC-CE, PX-NV She had a slight anemia (Hgb 99 g/l) on May 31 without an increase in reticulocytes (50 x 10 9 /L) and no polychromasia. A slight thrombocytopenia was noted by a manual estimate on blood smear. Many large PLT were seen.
Tasks 1. Interpret the instruments graphical reports and give suggestions for what could cause the changes in the cytograms and histogram. 2. Which total WBC count (Baso or Perox) should be more accurate? 3. Was the lymphocyte count 19.3 x 10 9 /L and if so what would that indicate? 4. Indicate what we should expect to see in the blood smear.
Figure 1 Advia WBC graphic May 31, 2017. Mailed-in sample one day old. Figure 2 Advia WBC graphic June 16, 2017. Mailed-in sample four days old Figure 3 Normal Advia WBC graphics dog
Boule Hematology Analyzer results May 31, 2017. Fresh sample Test May 31 Reference WBC 19.3 x 10 9 /L 6.0-17.0 Granulocytes 16.0 x 10 9 /L 3.5-12 Lymphocytes 2.5 x 10 9 /L 0.9-5.0 Monocytes 0.8 x 10 9 /L 0.3-1.5 Figure 4. Boule WBC histogram for patient May 31 Figure 5. A normal appearing canine Boule WBC histogram Lyserade LPK means lysed WBC.
Diagnosis: Hepatozoon canis. H. canis gamonts were seen in about 40-54 % of the leukocytes causing unusual Advia WBC dot plots, see Advia Interpretation below. Blood smear examination indicated 40 % of leukocytes on May 31 (about 54% June 16) contained elongated oval H. canis gamonts. These protozoa had a distinct cell membrane and one round granular nucleus. The infected leukocytes often had a segmented shape of the nucleus but their chromatin pattern was more open (more euchromatin) than normal neutrophils. The large gamont often displaced the nucleus to one side of the leukocyte. PCR confirmed the parasite to be Hepatozoon canis. Figure 6. Hepatozoon canis gamont in a leukocyte Advia interpretation Both samples from May 31 and June 16 had a distinct, large population of large peroxidasenegative leukocytes seen in an unusual position of the Advia 2120 Perox dot plots (Fig 7). That new cell population was mainly found high in the Advia s lymphocyte counting area, but also in LUC and monocyte area. Canine cells commonly seen in this area are monocytes and basophils. Mast cells, immature lymphocytes and macrophages may also be present in this area. No basophils, mast cells or immature lymphocytes were seen in the blood smear. The blood sample from May 31 also had a distinct population of monocytes in the usual monocyte area found more to the right of the unusual cells (Fig 7). Neutrophil, lymphocyte and eosinophil cell clusters were present in normal areas of the Advia peroxidase scattergram. In fresh made blood smears from sample May 31 and June 16 Hepatozoon gamonts were seen in 40% and 54%, respectively, of the leukocytes. The cells in the distinct new cell population seen in Advia PEROX dot plots were interpreted to be the leukocytes containing Hepatozoon. The Advia s Baso cytograms were more difficult to interpret. There were two long, linear and parallel cell populations. The upper population appeared abnormal and resembled a mirror image population of the lower and more typical population. In the Advia Baso channel the leukocytes are pretreated with a reagent that removes the cytoplasm from the leukocytes. The size of the nucleus is shown on the y-axis and the optical density that reflects the complexity of the nucleus on the x-axis. The Baso cytogram is often compared to a worm with a round head to the left containing mononuclear cells (lymphocytes and monocytes) and the linear body to the right being mainly neutrophils. In the Baso scattergram from this dog the most dense cell cluster (representing the largest number of cells) of the leukocytes was seen in the mononuclear cell area more to the left of the lower population (at the head of the worm ) (Fig 7). The Baso x-axis histograms confirmed that the largest number of cells were in the mononuclear cell area. In addition, a diffuse, less dense elongated cell population was seen above the leukocyte worm and more to the right. The upper cluster was located higher on the Y-axis and represented larger structures remaining after action of the basophil reagent than normal leukocyte nuclei. There were likely abnormal cells/particles. It is possible that these abnormal
cells/particles were free parasites or parasites within cell debris remaining after the effect of the lysing reagent. The Baso WBC count was greater than WBC Perox perhaps reflecting the presence of free protozoa (abnormal cells/particles), so the Perox WBC count was judged to more likely be correct than the WBC Baso count. Figure 7. Advia Perox and Baso scattergrams and x/y-histograms. The Perox cytogram has a dense population of atypical cells (blue oval) that are large (higher on the Y-axis) and mainly peroxidase negative (located to the left along the X-axis). A smaller population of monocytes (light green dots) are in the more typical monocyte area to the left of neutrophils (red dots). The Baso cytogram has two parallel and linear populations of cells. The most dense cell population (yellow oval) was in the mononuclear area (lymphocytes and monocytes) of the lower linear cluster of cells. Boule is an impedance instrument. The Boule WBC histogram (Fig 8) of this dog extended more to the right indicating it had a population of larger than normal WBC (blue circle in figure). This was interpreted to represent leukocytes containing Hepatozoon. The Boule interprets mid-cells in the area (red rectangle) between lymphocytes (smaller cells on the left) and neutrophils on the right side of the histogram to be monocytes.
Figure 8. Boule WBC histogram for patient May 31 What is the cell type containing H. canis? Findings in blood smear: The blood smear examination indicated 40 % of leukocytes on May 31 (about 54% June 16) contained oval H. canis gamonts. The infected leukocytes were larger and their cytoplasm was bluer than the normal looking neutrophils shown in Figures 9B and C. The infected leukocytes often had a segmented shape of the nucleus but their chromatin pattern was more open (more euchromatin) than the neutrophils. The large gamont often displaced the nucleus to one side of the leukocyte. Figure 9A and B. Hepatozoon canis gamonts were in leukocytes that were larger and had bluer cytoplasm than normal neutrophils. These leukocytes had nuclei often with a segmented shape but a more open chromatin density than the neutrophils in the same sample. They looked more like monocytes than neutrophils.
Figure 9C. About 40 % of leukocytes in the monolayer area contained H. canis on May 31 and 50 % of leukocytes on June 16. The infected leukocytes were large and the density (%) of infected cells at the feathered edge was greater. The infected leukocytes look more like monocytes than neutrophils. Neutrophils had clear cytoplasm and dense nuclear chromatin. Figure 10. Blood smear stained with Hemacolor. Note the different and more empty appearance of gamonts with Hemacolor than with Giemsa staining shown in Figure 9. Diff-Quick staining has also been described to reveal fewer morphological features of H. canis compared with Giemsa-stain (Mercer et al. 1988). The difference between neutrophils and monocytes is less apparent with this stain.
In most studies H. canis gamonts are described to be found in neutrophils (Elias et al. 1988, Baneth et al. 2001, PAŞA et al 2009, Paiz et al. 2016). H. canis has been described to infect both neutrophils and monocytes (Baneth et al. 2007, Vincent-Johnson 2013), Some authors avoid deciding what the cell type is and only describe the infected cells as leukocytes (Mercer et al. 1988, Otranto et al 2011). Canine monocytes were described to be coinfected with H. canis and E. canis in a study by Baneth (2015). In another study, the infected leukocytes were described as myeloperoxidase negative neutrophils (Ibrahim et al. 2008). The possibility that the leukocytes instead could have been monocytes was not discussed. There are some few studies that have tried to identify the origin of the infected leukocytes. Makimura el at. (1991) stained H. canis infected canine blood leukocytes for α-napthyl acetate esterase (ANAE) and chloroacetate esterase (CAE). The infected leukocytes were ANAE positive and CAE negative, which are monocyte characteristics (Makimura). Murata et al. 1993 studied ultrastructure and cytochemical characteristics of leukocytes infected with H. canis in 58 canine samples. The infected leukocytes were ANAE, CAE, myeloperoxidase (PER) and leukocyte alkaline phosphatase (LAP) negative which did not completely agree with either neutrophil (CAE, PER and LAP pos and ANAE neg) or monocyte characteristics (CAE, PER and LAP neg and ANAE pos) in that study. Hepatozoon infected peripheral blood leukocytes in flat-headed wild cats were negative for Sudan black B that stain lipids in secondary neutrophilic granules (Salakij et al. 2008). Myeloperoxidase is present in the primary granules from the promyelocyte stage to the segmented canine neutrophil and CAE activity is present in all maturation stages of canine neutrophils. SBB stain feline neutrophils in all stages of maturation (Raskin 2010). It seems unexpected that the infected leukocytes are neutrophils that have lost myeloperoxidase, and chloroacetate esterase activity. H. americanum were described to infect macrophages in the cysts and granulomas in skeletal muscle, as well as in peripheral blood based on morphology and immunohistochemical reaction, (Cummings et al. 2005). Several species of Hepatozoon are found fairly commonly in different types of birds. The parasite is seen usually in monocytes of birds but even lymphocytes and rarely thrombocytes (Shurulinkov 2005, Clark et al. 2009). Based on the Advia Perox and Baso scattergrams most of the infected cells were relatively large, peroxidase-negative cells found in the mononuclear cell (monocyte and lymphocyte) areas. The Boule histogram indicates the presence of cells larger than neutrophils. Most reports on Hepatozoon indicate they are found in neutrophils. We are unconvinced the infected cells in our case were neutrophils and suggest Hepatozoon were found in monocytes. The appearance on blood smears was more similar to monocytes than neutrophils. Compared with normal neutrophils the infected cells were slightly larger, the nucleus segments were wider with less dense heterochromatin density. The cytoplasm was darker blue. Neutrophils with normal appearance were present. The manual neutrophil count (33%) agreed with the neutrophils counted by Advia 2120 (29%). The leukocytes with gamonts often had very segmented nuclear shapes which can suggest they were neutrophils to many observers. However, canine monocytes may have very segmented nucleus as shown in figure 11 in blood from another dog.
Figure 11 A, B + C. A) Two monocytes and 1 neutrophil from another dog without Hepatozoon infection. These monocytes also had often very segmented shapes to their nuclei. B) Neutrophil C) Monocyte We did not do any special staining, except Advia myeloperoxidase staining. The infected cells were myeloperoxidase negative unlike normal neutrophils. It has been shown that endotoxin and severe infections can cause neutrophils in horses and dogs to be less myeloperoxidase positive (Klenner et al 2010, Lilliehöök et al 2016). Thus, we cannot totally exclude the possibility that the infected cells were very altered neutrophils.
In conclusion, this dog had a severe infection with H. canis with gamonts in more than 40% of the leukocytes in the first two blood samples that were analyzed with our Advia. The Advia and Boule cytograms indicated the presence of many cells found in atypical areas of cytograms. The infected leukocytes were leukocytes with a segmented nucleus. The cells were not proven to be typical neutrophils nor monocytes. Based both on morphology, Advia characteristics and earlier studies (Makimura et al. 1991, Murata et al. 1993, Cummings et al. 2005) the infected cells were more similar to monocytes in our opinion. Hepatozoon canis is a protozoan from the phylum Apicomplexa. H canis belongs to the family Hepatozoidae of suborder Adeleorina. This family consists of more than 300 species, described in reptiles, birds and mammals (Baneth et al. 2007). Dogs and other canids are the vertebrate intermediate host of H canis. Dogs become infected with the ingestion of a Rhipicephalus sanguineus tick, which is the definitive host. The ingested H. canis sporozoites are then released in the dog s intestinal tract, penetrate the gut wall, and are carried by blood or lymph to various tissues where they develop to merozoites. Some merozoites enter neutrophils or monocytes and develop into gametocytes (Ivanov et al. 2008). H. canis infection can be asymptomatic in dogs with low levels of parasitemia, while high parasitemia can cause severe life-threatening disease. Common clinical signs are anaemia, emaciation and intermittent fever. Other described signs are cachexia, depression, muscle hyperaesthesia, purulent conjunctivitis, rhinitis and diarrhea. (Ivanov et al. 2008). The dog in our case was treated with imidocarb dipropionate four times every two weeks and the number of infected cells declined with time. At sampling on June 30 only 5% of the leukocytes were infected. On July 16 no protozoa were seen and the hematological values were normal. Also, the clinical signs had disappeared so the treatment appeared to be effective. This Swedish dog had not been abroad and it is unclear how she was infected. In Sweden most ticks are Ixodes ricinus, H. canis sporogony is described not to occur in Ixodes ricinus (Giannelli et al. 2013). Though unlikely it is possible the dog may have eaten an infected Rhipicephalus sanguineus tick, which had been imported on another dog or a person. H. canis can infect transplacental, but this dog was 10 years old and her severe infection suggests an acute infection. References Baneth G, Samish M, Alekseev E, Aroch I, Shkap V Transmission of Hepatozoon canis to dogs by naturally-fed or percutaneously-injected Phipicephalus sanguineus ticks. J Parasitol 87;3:606-611, 2001. Baneth G, Samish M, Shkap V. Life cycle of Hepatozoon canis (Apicomplexa: Adeleorina: Hepatozoidae) in the tick Rhipicephalus sanguineus and dometic dog (Canis familiaris). Vet Parasitol 93:283-299, 2007. Baneth G, Harrus S, Gal A, Aroch I. Canine vector-borne co-infections: Ehrlichia canis and Hepatozoon canis in the same host monocytes. Vet Parasit 208:30-34, 2015. Clark P, Boardman W, Raidal S. Hemoparasites of birds. In Atlas of Clinical Avian Hematology, Wiley and Blackwell. p 127-136, 2009. Cummings CA, Panciera RJ, Kocan KM, Mathew JS, Ewing SA. Characterization of stages of Hepatozoon americanum and of parasitized canine host cells. Vet Pathol 42:788-796, 2005.
Elias E, Homans PA. Hepatozoon canis infection in dogs : clinical and haematological findings ; treatment. J. small Anim. Pract 29:55-62, 1988. Giannelli A, Ramos RAN, Dantas-Torres F, Mencke N, Baneth G, Otranto D. Experimental evidence against transmission of Hepatozoon canis by Ixodes ricinus. Ticks and tick-borne diseases 4;5:391-339, 2013. Ibrahim NDG, Rahamathulla PM, Njoku CO. Neutrophil myeloperoxidase deficiency associated with canine hepatozoonosis. Int J Parasitol 19:8;915-918, 1989. Ivanov A, Tsachev I. Hepatozoon canis and hepatozoonosis in the dog. Trakia J Sci 6:2;27-35, 2008. Review Klenner S, Richartz J, Bauer N, Moritz A. Myeloperoxidase deficiency in dogs observed with the Advia 120. Tierärztliche Praxis Kleintiere 3:139-146, 2010. Lilliehöök I, Tvedten H, Bröjer J, Edner A, Nostell K. Time related changes in equine neutrophils after experimental endotoxinenia: myeloperoxidase staning (MPXI), size and numbers. Vet Clin Pathol 2016 Makimura S. Kinjo H. Cytochemical identification of canine circulating leucocytes parasitized by the gametocyte of Hepatozoon canis. J Vet Med Sci 53:5;963-965, 1991. Mercer SH, Craig TM. Comparison of various staining procedures in the identification of Hepatozoon canis gamonts. Vet Clin Pathol 17;3:63-65, 1988. Murata T, Inoue M, Kanoe M, Kono Y, Ishida Y, Horio M, Shimada M, Yokoyama M, Taura Y, Nakama S. Ultrastructure and cytochemical characteristics of leukocyte infected with Hepatozoon canis. J Vet Med Sci 55:6;1043-1045, 1993. Otranto D, Dantas-Torres F, Weigl S, Latrofa MS, Stanneck D, Decaprariis D, Capelli G and Baneth G. Diagnosis of Hepatozoon canis in young dogs by cytology and PCR. Parasites & Vectors 4:55;1-6, 2011. Paiz LM, Silva RC, Satake F, Fraga TL. Hematological disorders detected in dogs infected by Hepatozoon canis in a municipality in Mato Grosso do Sul State, Brazil. Arq Bras Med Vet Zootec 68:5;1187-1194, 2016. Pasa S, Kiral F, Karagenc F, Atasoy A, Seyrek K. Description of dogs naturally infected with Hepatozoon canis in the Aegean region of Turkey. Turk. J. Vet. Anim. Sci. 33:4;289-295, 2009. Raskin R. Cytochemical staining. In Schalm s Veterinary Hematology. Eds: Weiss DJ, Wardrop KJ. John Wiley & sons, 2011. Salakij C, Salakij J, Narkkong N, Sirinarumitr T, Pattanarangsan R. Hematologic, cytochemical, ultrastructural, and molecular findings of Hepatozoon-infected flat-headed cats (Prionailurus planiceps). Vet Clin Path 37:1;31 34, 2008. Shurulinkov, Peter. (2005). Occurrence of Haematozoan Parasites of Genus Hepatozoon (Apicomplexa: Hepatozoidae) in Wild Birds in Bulgaria. Acta Zoologica Bulgarica. Acta zool. Vincent-Johnson N. Canine and Feline Hepatozoonsis. In Canine and Feline Infectious Disease ED: Sykes JE, p 747-759, 2013.