Hello. I am Joseph Solomkin. I am professor of surgery at the University of Cincinnati College of Medicine and I am welcoming you to our program today

Transcription

1 Hello. I am Joseph Solomkin. I am professor of surgery at the University of Cincinnati College of Medicine and I am welcoming you to our program today on Growing Resistance Epidemic: A Call to Action. Today, we will be discussing collateral damage. A lot of this discussion with circle around Clostridium difficile. I will be joined in this program by Dr. Robert Owens. Dr. Owens is co-director of the Antimicrobial Stewardship Program and clinical specialist in infectious diseases for the Department of Pharmacy Services and Division of Infectious Diseases at Maine Medical Center in Portland, Maine. Additionally, he is a clinical assistant professor in the department of medicine at the University of Vermont College of Medicine and he is a faculty member in the infectious diseases fellowship program at Maine Medical Center. He received his PharmD at Mercer University School of Pharmacy in Atlanta and he completed his residency in clinical pharmacy practice at Dekalb Medical Center, also in Atlanta. He then did the fellowship in infectious disease at Hartford Hospital in Hartford, Connecticut, a premiere program of antimicrobial pharmacokinetics and dynamics. It is a great pleasure for me to introduce Dr. Owens. Dr. Owens: Thanks very much, Dr. Solomkin and everybody joining us today to talk about this very interesting topic related to antimicrobial harms. If we take a look at the first slide here, Antimicrobial Resistance Collateral Damage Associated with Antimicrobial Use, you will notice a cartoon and another picture of a book that really focuses on antimicrobial side effects. In fact, the major text of Bitter Pills was associated with one of the 4 quinolones, an entire book written upon the insight of the hazardous world of legal drugs. The very prescriptions that we take can cause a great amount of harm. Even back in 1944 when we really did not have very many antibiotics around, there was still an appropriate use message attached to this advertisement of the sulfa drugs. I wanted to put this slide in here because obviously with penicillin back on the beaches of Normandy in 1944, you can clearly see, the dawn of our antibiotic era, all the way through today. In everyday aspects of clinical practice, antimicrobials have very well-known benefits. Sometimes we get caught up in talking a lot about the benefits, but we do not talk about the other equally sharp side of the sword, which is the antimicrobial harms. In today s teleconference, we do not have time to talk about the variety of harms that are attached to antimicrobials, but I did want to expose people to this topic. We will focus on a couple of different aspects of this topic. On this particular slide, some of the traditional harms that we are looking at deal with many of the severe side effects that most people in clinical practice have seen at one time or another anaphylaxis due to beta-lactam antimicrobials, Stevens-Johnson syndrome or toxic epidermal necrolysis due to Bactrim or to other products as well. Torsades de pointes has become an issue, where it previously was not one. Interstitial nephritis with aminoglycosides is well known, as well as other types of nephrotoxicity. We see this particularly in people on concurrent nephrotoxic drugs, as well as liver failure many examples of different antibiotic classes, seizures, high-dose beta-lactams, imipenem if the patient does not receive dosing adjustment adequately, or if the patient has underlying CNS disorders. In fact, in one of the clinical trials for development of imipenem, there was a 30% seizure rate in the meningitis study. In that particular disease state where we have somebody who has a CNS infection, imipenem is probably not a good idea. Fluoroquinolones also displace gaba from the CNS receptor sites, which also predisposes them to seizures. There are a variety of ways to get around these particular side effects. The drugs cannot be prescribed but that is not something that we are allowed to actually do. We need many of these drugs. They are very lifesaving. On the next slide, we talk about some of the harms that are a little bit distanced from the patient receiving the actual drug and the clinician taking care of that particular patient. Resistance developing during therapy is one of those particular non-traditional harms that is done to patients through inappropriate antibiotic use and appropriate antibiotic use, I might add. We facilitate this by 1 of 11

2 inappropriate dosing of antimicrobials, giving antimicrobials for too long, etc. Also now our new foe and nemesis is C. difficile-associated disease, and this is starting to become one of the more realtime harms that we see when patients are receiving antibiotics. If we look at some of the potential solutions, which I am going to talk a little bit more in depth about today, what we say is the holy trinity of infectious diseases is judicious use of antibiotics, infection control measures, and environmental stewardship as methods that we can actually use in a combined fashion to help decrease the risk of C. difficile-associated disease in clinical practice. Practicing antibiotic heterogeneity, antibiotic stewardship programs will be a focus on the third teleconference in the series, and more or less methods of how we can actually use antibiotics more wisely and the use of more ecologically friendly antimicrobials. Believe it or not, there are some antibiotics out there that actually have less potential to cause resistance, and some of those drugs we will talk about. Vancomycin is one of those actual drugs that resistance has not developed to Staphylococcus aureus until 30 years after clinical use. Hand hygiene compliance is also going to be talked about, and the solutions of environmental cleaning and vigilance related to that is a solution that needs to be sort of interplayed with appropriate antibiotic use in an effort to get rid of C. difficile-associated disease. For now, I am going to talk briefly about antibiotic resistance developing during therapy, and this is a very complicated slide. Nonetheless it does actually report in one slide what has happened in New York City over the last 10 to 12 years. Subsequent to the late 1980s, there were not many resistance problems in New York City. They did have an outbreak of ESBL-producing organisms following the introduction of ceftazidime use. And so, heavy reliance on ceftazidime primarily because of the reason that is mentioned on this slide multi-drug resistance but ceftazidime-susceptible Acinetobacter species, sort of squeezed the balloon and allowed everybody to use ceftazidime as the agent of choice in ICUs. Heavy reliance on that particular drug drove resistance to a variety of other organisms which led to an increased use of imipenem. This use in turn resulted in a variety of imipenem-resistant organisms. Throughout the last decade, we have learned a lot about antibiotic use and what has been termed squeezing the balloon, if you will. This particular concept is very true. This study published in Clinical Infectious Diseases in 2002 demonstrated clearly that imipenem use was associated with imipenem resistance not a big surprise. Ceftazidime use was associated with ceftazidime resistance. Interestingly enough, quinolone use or in this case ciprofloxacin use was actually associated not only with resistance to itself as identified in the last column, but it was also associated with imipenem resistance. There is a biologic basis for this occurrence too detailed for us to go into in this teleconference. Nonetheless, you can actually see that antimicrobials have collateral effects in terms of developing resistance not only to themselves but oftentimes to other antibiotics as well, almost as a collateral damage aspect of the antimicrobial use. Unfortunately, we are oftentimes seeing the patient. The patient maybe is off service now at home recovering when they come back in with this infection that may now be due to a resistant organism because of that antibiotic use. On the next slide, it is an example of this clinically that we have seen at our hospital thanks to John Quinn, our beta-lactamase guru, and his associate Karen Rolands. I just wanted to give them some recognition because they helped us out on a really vexing case of a 48-year-old female, whom we will call Patient A, who developed recurrent urinary tract infections due to E. coli--something that we see postoperatively in the intensive care unit from time to time--requiring near continuous treatment. Unfortunately in this case, patient A had to be treated with antimicrobials for literally years, almost 5 years to be exact. It kind of went like this. The patient was prescribed initially amoxicillin, very appropriate. The patient developed amoxicillin resistance. The patient was then treated with cephalosporin A, developed cephalosporin A resistance, and then she was treated with cephalosporin D and developed resistance to that as well, on down the line. Bactrim was knocked out, etc. She has a twin sister, and Patient A cares for this sister, Patient B. Again, this Patient A is having near continuous urinary tract infections. Patient B is a tad more debilitated. It came to light not too long ago that Patient B was admitted to our hospital with urosepsis and had never received 2 of 11

3 antibiotics previously within the last at least 2 to 3 years. Interestingly, she had actually developed this severe infection and went straight to the ICU. She unfortunately died 4 days later. Patient B had actually acquired the same organism; it turned out to be E. coli. We had it pulse-fielded by John Quinn and his group in Chicago. What they realized was, in fact, the same infecting strain from Patient A infected Patient B, so this is a classic example of what we call the societal consequences of antibiotic use. When we give antibiotics, they not only affect the patient receiving them, but also the patients around them and in this case family members around them. Patient B being the sister of Patient A was infected with what we think is the first CTXM ESBL enzyme to be identified in the United States, with this much clinical data around it anyway, as a cause of disease. In fact, we did not have enough time to adequately treat this patient as she obviously died due to this infection directly after being on pressors for a few days. We can see the resistance having a clear effect collaterally here in this patient who had never even received antibiotics. The next slide is an example of a patient who had suffered from C. difficile-associated disease and was actually treated both in the emergency department and by our surgical team. Everyone followed the algorithm that we developed at Maine Medical Center back in 2002 as we were one of the first sites in the United States actually to be inundated with this particular pathogen. Unfortunately, we had a number of mortalities, as have many people who have been unlucky enough to have the BI/NAP1 strain, which we will talk about in just a moment. I just wanted to start off by introducing this topic by a particular case. In fact again, this patient was seen and never had a history of being in the hospital. He was about 51 years old and had never received antibiotics. Within the last 30 days, he had actually had a course of Augmentin or amoxicillin and clavulanic acid. In this particular case, the patient this was his only antibiotic exposure was treated very appropriately in the emergency department by our folks. The patient was given metronidazole according to our pathway. The next slide actually has the details of the case. The patient was treated in the emergency department and released that day. He came back in about 6 hours, was diaphoretic this time, toxicappearing after getting his prescription filled, and only had 1 dose of metronidazole. Very quickly, within a matter of about 12 hours, this patient had gone from suffering some nuisance-type diarrhea to being seen in the emergency department, to getting a prescription for metronidazole, and to being re-admitted to the hospital. The reason I saw the ED folks and the surgeons actually treated this patient very appropriately was the fact that they employed a test that Dr. Solomkin is going to go in more detail about. They performed a CT scan on this individual which noted a couple of severe signs and markers for severe C. diff. He had ascites, colon wall thickening, and unfortunately he had free air in his abdomen. They found a place in his colon where he had actually perforated. He was then treated, as on the previous slide, according to our algorithm with intravenous metronidazole, oral vancomycin started via the NG tube, and then sent to the OR for what Dr. Solomkin is going to talk more about as well later in this teleconference, and an emergency colectomy was performed. The patient was fully recovered. About 7 days later, he was actually out of the hospital and off all antibiotics. Interestingly enough, we sent a stool sample away, and it came back positive for being the BI/NAP1 strain, the epidemic strain that is causing all of this disease and virulence. We saw again a clear example of a guy who had never been admitted to a hospital and has had fairly remote antibiotic history but came down with this disease and almost died from consequences of this particular disease as well. If it is not impacting your neighborhood, it is coming to one near you soon. It has now been identified in many of the states in the U.S. and in certain countries around the world, so it is spreading very rapidly. Some basic concepts about C. diff., what we know about it anyway, is it is a 3-hit process to get the disease. Disruption of the intestinal microbiota in fact previously has been attributed primarily to antibiotic use. There is a subset of patients now who have no antibiotic use but rather have been exposed to proton pump inhibitors. In this category of patients, again if you are interviewing them and 3 of 11

4 they have had no antibiotic use in the past 6 months to a year but they have been on a proton pump inhibitor, that is now taking the place of antibiotics as a risk factor for C. diff. through potential minor disruption in the intestinal flora. The second hit is you need to be exposed to the organism, the toxigenic strain of the organism, in fact. The BI/NAP1 is becoming so prevalent that we are now seeing community-acquired cases of this particular organism. The third wildcard variable is that some people get antibiotics where they have disrupted intestinal microflora. They get exposed to the organism. They have been in a long-term care facility or been hospitalized a long period of time, and they have not gotten C. diff. The fourth wildcard hit that people need to have in order to get the disease is thought it is an immune-based problem where the patients are actually unable to mount a serum IgG or a mucosal IgA antitoxin A antibody. This is particularly prominent in patients over the age of 65 as well, so we are seeing the large predominance of C. diff. right now in the elderly population, whom we think has a waning immune response, particularly the IgG antibody response to antitoxin A. And so, putting this perfect storm together, this is the scenario for developing actual disease for C. diff. What makes this organism different than the other ones? I will just boil it down to a couple of different facts really quickly. The graph to the left is data from Michel Warney where he illustrated very nicely in blue that the toxigenic BI/NAP1 strain, in fact, produces 16 to 23 times more toxin than shown in red, which is the usual customary strain of C. diff. toxinotype 0 strains that most of us have been dealing with for the last 20 years. This is a new strain. It produces again 23 to 16 times more toxin than previous strains. It has also acquired resistance determinates not to the treatments themselves but rather to different antibiotic classes, which makes the administration of those particular antibiotics risky. The number one area where resistance has been acquired is in the area of fluoroquinolones, and so again receiving a fluoroquinolone or cephalosporin again are 2 major risk factors. I will go over those data in a moment as well. We also have recently learned in the study that we initiated with Mark Wilcox in the United Kingdom the fact that this is a hyper-spore-producing strain. Putting all of this together, what does this mean? We have a hyper-toxin-producing strain of C. diff. It stays in the environment longer because of the sporulation characteristics that we know and we have learned from it. In addition, more antibiotics are on the table now than we previously had from previous studies that show there are greater risks from other antibiotics that we have out there. Putting all those together, this is sort of the perfect storm that people have been talking about. Now, this slide depicts all antibiotics have an odds ratio greater than 1, which indicates that they all are associated with C. diff. It is just the question of which antibiotics may be more risk than others. Cephalosporins by far and away have been the most risky antibiotics to give in which C. diff. is going to occur during therapy. I will show you more recent data, which is in a very busy graph. Here are 4 studies, including one done at our hospital during the time of the initial outbreak in 2002 and also from 3 other centers, which illustrate a common theme in that cephalosporins possess the greatest odds ratios, and fluoroquinolones were close behind. Most antibiotics, again as I said, are associated with C. diff., but these 2 classes seem to be most associated with the current strain that is roaming around right now. On the next slide, some people suggest that surgical prophylaxis is only a single dose in many cases. It may be upwards of 3 doses potentially, so this should be less of a risk for C. diff. This is not borne out to be true in the literature. In fact, if you look at a couple of different studies, perioperative prophylaxis actually was the number one risk factor for getting C. diff. in this particular paper published now about 10 years ago. Even back then it was well noted as a risk factor. 4 of 11

5 If we fast-forward to 2005 and data from Canada, they demonstrated that cefoxitin use with the antianaerobic cephalosporin for surgical prophylaxis was very strongly associated with C. difficileassociated disease development in Canada. Other studies also pointed to this fact. If you look at the various different agents studied over time for prophylaxis and we go way back here with a variety of different older antibiotics as well, cefoperazone and mezlocillin, that most people have never even prescribed before. If you look at the other ones that people are commonly prescribed cephalosporins in general they all have odds ratios that demonstrate risk for C. difficile-associated disease in this case, a percent who are colonized, which does not indicate disease but rather does shows that it does increase the risk for C. diff. Here is another paper. This is the pre-bi/nap1 era. Again, this is surgical prophylaxis in patients who actually ended up developing C. diff. This is in an era where a variety of different agents were used. Ciprofloxacin, cefoxitin, cefotan, meropenem were the most common antibiotics used in this particular study. Again, it did not include only patients who were undergoing operations, but it did include sort of all-comers, if you will. What they noted was that these patients had higher indirect mortality or crude mortality, but they did not have higher direct or actual mortality related to C. diff. Nonetheless, they still were put in the position to have a poor outcome if they did develop C. difficile during their length of stay. The average time from the completion of antibiotic therapy to the diagnosis of C. diff. in this case was 7 days, and 16% developed C. diff. after prophylactic antibiotics only. On the next slide, if you take a look at this paper, this is actually comparing a cephalosporin, cefotaxime, versus Pipracil and tazobactam. This is an interesting study design, an open ward-based crossover design, in a population enriched for people who were at risk for C. diff. and elderly patients, sort of elderly wards. C. difficile was defined as you see it on this slide. The duration of hospitalization in the Pip-tazo group was actually longer than that of prior to developing C. diff. than the cephalosporin antibiotic, sort of stacking the deck against patients receiving Pip-tazo. Those people would obviously be more likely to have acquired C. difficile or have contact with the organism prior to getting antibiotics. What they found in this particular study, if you pan over to the far right, was that cefotaxime actually had the highest by many fold risk for C. diff., and 53% of people receiving cefotaxime developed C. diff., whereas only 7% of people receiving Pip-tazo developed C. diff. This is obviously a statistically significant finding, but it is something that has been borne out of all of the most recent studies looking at C. difficile-associated disease. Pipracil and tazobactam again has demonstrated that it is more or less one of those ecologically friendly antimicrobials in the fact that it has not been pinpointed as a risk factor specifically in all of these studies that have been done previously and in current times with the BI/NAP1 strain. This is illustrated again in a very recent study published in 2006 by the CDC, Cliff McDonald and his group. On this slide here, cephalosporins had the highest odds ratios followed by the fluoroquinolones. Again as I had mentioned previously, proton pump inhibitors also popped out as a risk factor. One of the problems that we face with C. difficile-associated disease is the fact that antibiotic restrictions or antibiotic stewardship, if you will, is not going to be the only solution. In fact, C. difficile behaves very much like anthrax. In order to actually rid the environment in the post offices that were infected by anthrax, they had to shut them down and actually chlorine gas the place. This is the only known sporicidal agent except for a very, very high concentration of hydrogen peroxide. Even though we cannot actually chlorine gas hospitals, we can use chlorine bleach as a sporicidal disinfectant to rid bed rails and very high-contact areas of C. diff. The problem again is that it is a spore-forming organism, and this organism can last literally for years outside the body. Therefore, that is why environmental interventions are very key in addition to using antibiotics wisely in terms of getting rid of C. diff. If you look at this particular study done by John Boyce, it is actually a culmination of studies. You can see that these are surfaces that are positive for C. diff. spores in 5 of 11

6 commodes, on toilet floors, bed rails, bed pans, room floors, etc. and are all very high-contact areas, including call buttons and bed sheets as well. Just to realize that C. diff. is all around our practicing environment, we can certainly take it home to our families, much less pass it along to other patients. If you take a look at this study by Matt Seymour and colleagues, published again about 10 years ago, it noted an interesting concept. As environmental levels of contamination of C. diff. spores increased, so did that of the health care workers hands. Again, in the last row, greater than 50% of environmental sites were positive for C. diff. spores. You can actually see that the number of people with C. difficile on their hands increased as well, up to 36% in this particular study. If you are wondering is it just physicians or medical students? No, it is everybody. It is nurses, physical therapists, dialysis technicians, housekeepers. Everybody who comes into contact with the hospital or long-term care facility environment has the potential to carry this organism and the spores particularly in the oil layers on your hands, which is why hand-washing with good old-fashioned soap and water rather than alcohol hand gels is the treatment of choice actually to get C. difficile off your hands. That should be reiterated. Alcohol hand hygiene products are very, very good for getting rid of Pseudomonas, VRE, MRSA, and other problem pathogens off the hands of health care workers in a quick and convenient way, but what we do know about C. difficile anyway is that C. difficile loves alcohol. It certainly does nothing to remove this in a statistically significant way from your hands. That is why hand-washing with soap and water is encouraged very strongly. In fact, if you put all of these interventions together, in 2005 we had about 586 cases of C. diff. Simply by adding environmental maneuvers, such as the chlorine bleaching rooms, to our armamentarium of interventions that we can do to get rid of C. diff. at our hospital, we reduced the overall numbers of C. diff. by 100 cases, which we previously were not doing. That is illustrated in the blue bars there. With that, I am actually going to conclude. Just to summarize briefly on the C. diff. end of things is that again it is more of an allen wrench approach. We need a different-sized wrench for each of the different problems that we face. We need something to address hand hygiene practices. We need something to address environmental cleaning of surfaces, and we also need something to address the better, appropriate, and judicious use of antibiotics as well. With this, I am going to turn it over to Dr. Solomkin who is going to go into more detail regarding C. diff. and some of the surgical solutions. Dr. Solomkin: Rob, thanks. That was really an extraordinarily thorough and very interesting discussion. I like the work that Dr. Owens has done personally in this area, and he has really been one of the leaders in doing research and then trying to publicize the size and magnitude of this problem and therapeutic, preventive measures that can be taken. I am going to deal with a much narrower area and deal primarily with surgical issues surrounding C. difficile colitis. This has become a very important problem at many medical centers because of the increasing prevalence of C. difficile colitis in our practices. This particular slide taken from a very nice Medicare survey done in the northwest of the United States shows the gradual increase in an even more recent spike. Whether or not the spike was due to the more invasive or more virulent strain that Dr. Owens described is not clear. Clearly, this is very rapidly becoming one of our most lethal nosocomial or healthcare-associated infection. The next slide actually shows the fatality rate of patients and about how that is also increasing. This would seem to correlate with the appearance of these more virulent strains that are hyperencapsulated and also producing much higher levels of C. difficile toxin. I am just going to mention very quickly because I think Dr. Owens has already described these that basically the infection begins with ingestion of spores. These survive gastric acidity. Obviously, their survival is facilitated by the administration of both proton pump and other antacid inhibitors. They germinate in the small bowel and then colonize the lower intestinal tract. The spores can then elaborate to toxins A and B. A is an enterotoxin, which we will describe a bit more in the next slide, and B is the cytotoxin that is actually 6 of 11

7 able to kill by itself colonic epithelial cells. Both toxins work very quickly by dysregulating the cytoskeleton of the epithelial cells. This means the cells then lose their shape and their ability to adhere to each other and to their connective tissue matrix. This results in fluid leakage commonly seen as bowel edema and diarrhea. Toxins result in formation of pseudomembranes as the epithelial cells gap formation or through actual death of the cells. The host s response to the exposure of the connective tissue of the bowel to fecal content is to form pseudomembranes. These pseudomembranes attach to the connective tissue and contain dead leukocytes and a variety of other products, including host defense protein in mucus. These are just some slides to illustrate what this process looks like. You can see in this first slide at the top with the arrows the pseudomembranes as they appear endoscopically, and then in the lower panel this is the surgically excised slide showing how these pseudomembranes look. Histologically, and I will show some other ones, this is what the pseudomembranes look like microscopically. I think you can see the loss of the mucosa and the formation of the pseudomembranes that are the body s attempts to protect itself from colonic content and colonic organisms. This is a higher-power view of what that looks like. I thank the Case Western Reserve University for providing these photos. This shows really almost a diffuse pseudomembrane formation. You can see the mucous layer on the surgical specimen. Again, here is a higher-power view of the pseudomembrane showing again the loss of epithelial surface and the host s reaction at the point of contact with the formation of these mucoid pseudomembranes. It is important to recall that colectomy is a very radical form of therapy, but it terminates the disease effectively. There are very rare reported cases of some forms of C. diff. occurring in the distal small bowel in severely effected patients. Generally a subtotal colectomy is a curative operation and will terminate the septic process that leads to the deaths of these patients. I would remind you of how high this death rate is. Current indications for subtotal colectomy are listed here, including toxic megacolon, perforation or peritonitis, and fulminant disease associated with shock and multiple organ failure. This disease is most likely caused by the very large loss of the membrane of the colon and absorption contact of the host with bowel content, particularly the standard bowel flora that we will discuss including anaerobes and facultative aerobic organisms. In the 3 prior cases of toxic megacolon, peritonitis, and then fulminating disease, it is very clear that operation is required even though this is a very late phase. A much more controversial area that I would like to spend some time on is an area where I think there is some room for us to improve on treatment outcomes of difficile colitis, and this is what is meant by a failure of medical therapy. There has been some work done to try to characterize patients that are going to fail with a colectomy or without a colectomy. This is data from a Canadian site that was one of the centers severely affected by the highly toxigenic strain that Dr. Owens described. In brief, the patients who died. If you will look to the no-colectomy column displaying the number who died as divided by the total number of patients in the group. Patients with age greater than 75 had a very high mortality rate. Shock requiring pressors was also associated with a very high mortality rate, as was a white count greater than 20,000, as was lactic acidosis. It is very apparent in this table that colectomy in these groups reduces the mortality in each group. However, if you look at these figures, we are still looking at experience with the 30% to 40% mortality rate overall. This is really very high, and one must certainly question whether this represents our best effort. So the question then becomes how do we identify patients and perhaps operate on them earlier as we await the development of a wonder drug. I would point out that because the disease process involves loss of the mucosa in the bowel, it is very likely that more aggressive C. diff. antimicrobial therapy will not reduce the mortality rate from this extreme form of the disease. The question then becomes is there 7 of 11

8 some basis for defining which patients would benefit from earlier operation. One way that is only now beginning to be explored is to use CT findings to both make the diagnosis and then to define stages of it. This is data from a recent study where they were comparing CT findings of patients scanned for intraabdominal processes where they were suspected of having colitis. The groups are divided into those with positive and negative toxin assays in their stool. I would point out that probably within the next 2 years there will be a much more sensitive genetic-based test producing a variant of what is called PCR technology. It appears to increase greatly the true positive rate. Had this test been applied to some of these patients with a negative screening test, this table might look different. In any case, colonic wall thickening seems to be very specific for the disease; nodularities and what is called an accordion sign, which is very uncommon; pericolic stranding is very common; and then ascites in these patients is also fairly common. I will go through some examples of these. In brief, the point being and the question I put on the table is whether more aggressive use of CT scanning as a staging tool would help us to define these patients. This is a slide taken from a very nice review article in the American Journal of Roentgenology that demonstrates many of these findings. The colonic wall thickening is illustrated by the white arrow with the long tail. The other arrows demonstrate edema in the mesentery very similar to what would be seen with very aggressive appendicitis, and then the suspicion of ascites or even perforation is raised also on this slide. The next slide is an effort to demonstrate the so-called target sign where you see these very bright rings circumferentially. This is felt to be almost pathonemonic for advanced C. difficile colitis. The next slide shows interesting observation of colitis confined to the right side of the colon. I am showing this mainly to point out that there has been some effort in the small numbers of these patients to do less than subtotal colectomy; in other words, to do either restricted right or left colectomy. These patients have had a relatively high recurrence rate in the retained colon. Some have required reoperation to remove this. Currently the recommendation is that patients operated upon for C. diff. colitis undergo a subtotal colectomy. Whether they undergo an immediate anastomosis would depend on their physiologic condition at the time of operation. This next slide, which I assume is almost impossible for you to read, tries to identify further some of the risk factors that in combination with CT scanning might be used to identify patients who preferentially should undergo early operation given non-responsiveness over a day or two to anticlostridial therapy. Increasing age is certainly an important risk factor. The figures given are per year of age. This slide also demonstrates at the very top data line that as the time went by as this study was performed, there was an increasing risk of death and of colectomy as the time or the year of appearance of the patient increased. More recently in the 2000s, the mortality rate and the colectomy rate have increased as I showed you in the previous study. As usual, women seem to do better with this acute disease than men do, having less mortality rate and about 15% less risk of requiring operation. Additionally, using Caucasians as the standard group, patients of other races, including African Americans, have a higher mortality rate from this. An interesting sidebar is that using privately insured patients as the standard seem to do better than patients in other insurance schemes and payment schemes with no explanation for this available at this time. In any case, these are the kinds of data that demonstrate the risk factors that determine mortality also to an extent determine those who would under current rules undergo a colectomy. 8 of 11

9 The other side of this that I want to get into has to do with management of these patients because aside from attempting to treat the underlying infectious etiology that Clostridium difficile, these patients again have lost, when they are acutely ill are really suffering from invasive infection from colonic flora. It is very important, and may in fact be the most important element, that a patient who is going to undergo an operation is to treat the manifestations of severe sepsis. These are definitions of sepsis that are now used and guidelines for the surviving sepsis campaign. They include a sustained systolic blood pressure less than 90, a mean arterial pressure less than 65 both typically indications for pressors reduced urine output, reduced central venous oxygen saturation and meaning peripheral hyper-utilization and a lactic acidosis, as I showed you on a prior slide as a major risk factor for death from this disease. One could indeed argue that a patient with lactic acidosis and other manifestations for shock should undergo surgery as soon as they can be hemodynamically stabilized. Certainly it is important to recognize the impact of hemodynamic instability on the outcome of this disease. Again, to come back to it, is that I would think of a point in time where we should be considering CT scanning as a staging modality and then base the decision on when to operate on the evidence of severity of sepsis and in particular sepsis that does not rapidly by that, I mean no more than 24 hours respond to aggressive hemodynamic resuscitation and antibiotics and should undergo rather rapid operation. Regarding antibiotics, one of the more interesting studies that has been done recently to reinforce the surviving sepsis guidelines is presented here. These authors looked at the timing from the onset of hypotension to the administration of effective antibiotics. The output from this was the odds ratio of death. What they demonstrated, as has been seen with severe community-acquired pneumonia, is that relatively small delays in instituting antibiotic therapy even as small as 4 to 5 hours results in a significant increase in the risk of death. Avoiding use of aminoglycosides in these patients who may be volume depleted and having renal insults both from the sepsis and from hypoperfusion may be very important in preventing toxicities. In this, we are very fortunate to have many of the drugs we have now. The next issue regarding managing these patients is what organisms need to be covered. What you see in this slide are organisms that are typically seen in community-acquired infections and are typically fairly easy to treat, including E. coli, strep species which are highly pathogenic, and Bacteroides fragilis and other reactive species. This would apply to a patient who has had only a brief disease and has not been extensively treated with antibiotics in the recent past. With this, I will introduce some thoughts based upon the SIS IDSA guidelines that were published in The first comment, which I make only because I think it is probably no longer correct, is I think we recognize it is important to initiate antibiotic therapy very early in the course, as I showed you from the Washington database, Washington-Vancouver data set, on impact on mortality of delayed antimicrobial therapy. The other point is that therapy should begin very quickly without waiting for any kind of supportive diagnostics. All of these patients would fall into the heading of very high-severity infections. The recommendations from those guidelines were to treat them either with Pip-tazo or with carbapenems. Alternatively for combination regimens, the extended-spectrum cephalosporins with metronidazole or quinolone with metronidazole were also recommended. The other point that I want to highlight very briefly with this slide is that one of the important determinants of the antibiotic therapy that is being given is the frequency and role of the various Bacteroides species. Most of these patients will be receiving metronidazole. It is hard to treat the underlying C. difficile infection. This agent has remained extremely effective against Bacteroides organisms. The other point that is important to remember and that we will come back to when we discuss issues in a subsequent conference on stewardship programs is duration of therapy. Even if 9 of 11

10 the patient is being treated and perhaps even more importantly because the patient is being treated for complication of antimicrobial therapy, considerable attention should be focused on limiting therapy for all of these patients all patients with intra-abdominal infections. One approach which was taken in our 2003 document was to wait until all signs of inflammation have resolved. In other words, the patient was afebrile and had a normal white count and return of GI function. The difficulty is that infection and inflammation are not the same thing. It is very likely that in many of these patients with a white count that comes down more slowly than others, the infection has been gone for some time after they have normalized the parameters we use to measure inflammation. This is leading to recommendations looking at or advocating shorter durations of treatment. In patients, for example, without perforation who undergo an emergency subtotal colectomy for C. difficile colitis likely require surprisingly brief therapy for no more than 4 to 5 days. The other issue is what happens if the patients do not appear to be responding following aggressive surgical therapy and broad-spectrum antibiotics. These patients typically, if there are organisms cultured and operation was successful, do not require new antibiotics. They do, however, require very aggressive diagnostic investigations that would typically be CT based. Of course, it is always important to make sure that the antibiotic therapy is appropriate and adequate for the organisms that were identified. This should occur as quickly as possible. The other issue is that most of the patients that we are talking about have been in the hospital and have received prior antibiotic therapy which led to this infection. Perhaps the best study that has been done to look at what organisms those patients are infected with is that displayed here by Montravers and his colleagues. They looked at asking the question that patients who have undergone prior antibiotic treatment following intra-abdominal surgery and who then develop peritonitis, how important is early, adequate, and effective empiric antibiotic therapy meaning given at the time of reoperation. What they discovered in their study is that the organism shifted greatly. There now was much less E. coli, much more Pseudomonas aeruginosa, and other species that are in community infections very uncommon. Along with this, they saw a significant increase in MRSA. MRSA is very rarely if ever a problem in community-acquired intra-abdominal infections. They also saw a disappearance of Bacteroides and an increase in Candida species. Most importantly, they discovered in this retrospective study that patients who received adequate empiric therapy meaning being given at the time of their second operation when the postop peritonitis was identified at that time did much better in terms of treatment failure and mortality as compared to patients who received therapy directed by cultures only begun days after the procedure. This is a cautionary tale that argues that in these types of patients who have been in the hospital, and even if not, have been exposed to antibiotic therapy that they be presumed to have a nosocomial flora and be treated accordingly. To summarize my part of this discussion, I believe that mortality of C. difficile colitis can be reduced from earlier surgical intervention and that this would depend, even in the absence of earlier diagnostic testing and even more active therapy, this can be I think achieved by a more aggressive stance on when these patients should go to operation. Reported mortality rates currently in my opinion represent delayed consequences of delayed intervention. The final point I would make in managing these patients perioperatively is that sepsis and nonresponsiveness to antisepsis therapy, which includes broad-spectrum antibiotics, is primarily based on fluid resuscitation. It is an urgent indicator for strong consideration of reoperation. I would like to invite the audience to attend the third of this series of teleconferences, and that will be entitled Antimicrobial Resistance: Clinical Countermeasures. We are really going to put together the solutions to many of the problems that we have talked about during the first and second teleconferences. I think the third one is going to really start putting things together in terms of how we 10 of 11

11 really deal with these clinical problems that we have brought up in the first 2 teleconferences. We look forward to your attendance during that teleconference. UK Wyeth Roundtable /ssb 11 of 11