ORIGINAL ARTICLE ANZJSurg.com
Clostridium colitis: challenges in diagnosis and treatment Kenneth N. Buxey, Chris Sia, Stephen Bell, Roger Wale, Daniel Wein and Satish K. Warrier Department of Colorectal Surgery, The Alfred Hospital, Melbourne, Victoria, Australia
Key words Clostridium difficile, colitis. Correspondence Mr Satish K. Warrier, Department of Colorectal Surgery, The Alfred Hospital, High Street, Melbourne, Vic. 3181, Australia. Email: [email protected] K. N. Buxey MBBS (Hons); C. Sia MBBS; S. Bell MBBS, FRACS; R. Wale MBBS, FRACS; D. Wein MBBS Candidate; S. K. Warrier MBBS, FRACS. Accepted for publication 3 August 2014. doi: 10.1111/ans.12840
Abstract Background: Clostridium difficile infection (CDI) has been reported to occur with increasing frequency and with more severe presentations being encountered. This article presents data from The Alfred Hospital highlighting the increased incidence, the increased severity and the broader clinical presentations observed. A case series highlights a variety of clinical scenarios that provided diagnostic and management challenges. We additionally describe a novel form of treatment for fulminant colitis. Methods: A retrospective review of C. difficile toxin (CDT)-positive and culturepositive cases was performed at The Alfred Hospital (2010–2012). Six cases are then presented as a case series to highlight the broad and atypical types of presentations one may encounter. Finally, a novel method for managing fulminant colitis operatively is presented. Results: A fourfold increase in cases of toxin-positive and culture-positive cases was noted over the initial 14 months of the period of analysis, the rate of cases detected then plateaued. This increase could not be explained by increased testing being undertaken. It is also not associated with increased usage of antibiotics nor with increased patient numbers being treated. Conclusion: CDI can present in various clinical forms. In our hospital, the number of cases of toxin-positive and culture-positive detection is increasing. A low threshold is required to identify and adequately treat patients with CDI. Fulminant colitis can be managed successfully with the creation of a diverting loop ileostomy, colonic washout and subsequent antegrade colonic vancomycin enemas.
Introduction Clostridium difficile infection (CDI) commonly results in nosocomial diarrhoea.1 In the hospital setting, it often arises following a period of antibiotic therapy. In the United States, there has been an increase in cases of severe colitis.2 We have also noted an increase in the absolute number of C. difficile toxin (CDT)-positive cases at our institution over the last three years. This increase is not accounted for by a proportional increase in Clostridium testing. A hypervirulent strain of CDT known as the NAP-1, or PCR ribotype 027 subtype, has been identified. This strain produces the two major toxins (A and B) as well as an additional potentially virulent toxin (binary toxin).3 The strain, while commonly found in both the United States and European countries, has more recently been isolated in Australia.4,5 Toxin A is an enterotoxin and produces damage to the colonic mucosa, whereas toxin B is a cytotoxin and has 10 times the potency. Not surprisingly, infection with this pathogen can result in acute severe colitis, a source of major morbidity and can even result in © 2014 Royal Australasian College of Surgeons
death. In one recent study, the mortality of severe colitis requiring subtotal colectomy secondary to CDI was 50%.6 With an increasing prevalence and virulence of Clostridium colitis, the aim of the current study is to report on our hospital experience with CDI, in particular highlighting the increased incidence, the increased severity, and the broader clinical presentations observed. The case series highlights a variety of clinical scenarios that provided diagnostic and management challenges at our institution. It also highlights a case of a novel form of treatment for fulminant colitis.
Methods A retrospective review of CDT-positive and culture-positive cases was performed at The Alfred Hospital (2010–2012). Results were attained from the microbiological laboratory at the institution. Ethics approval was granted for the study. Specific strains of C. difficile are not included in our data. ANZ J Surg •• (2014) ••–••
2
Buxey et al.
Fig. 1. The trend of Clostridium difficile toxin-positive cases per month. ( ) Toxin-positive cases.
Fig. 2. The trend of Clostridium difficile toxin- or culture-positive cases per month. ( ) Toxin- or culture-positive cases.
Jointpoint regression program version 4.0.1 was used to calculate the monthly percentage change. From these cases, six with varied clinical presentations of CDI were identified. These are presented as a case series to highlight the varied clinical manifestations of the disease. All patients in the case series required colorectal surgical input within the last 12-month period.
Results Clostridium toxin-positive and culture-positive cases (2010–2012) From June 2010 to May 2012, the monthly percentage change in toxinpositive cases was 8.0% (95% confidence interval (CI) = 5.4–10.7). This trend is shown in Figure 1. In absolute figures, this equated to a fourfold rise in toxin-positive cases (from 2010 – 7.7 toxin-positive cases/month to 2012 – 29.2 toxin-positive cases/month). This trend is mirrored by an equal increase in culture-positive or toxin-positive cases, in which an 8.8% monthly percentage change was observed (95% CI 5.9–11.8). This change is shown in Figure 2. There was no significant increase in antibiotic prescribing rates to explain this trend. The collective defined daily doses of all prescribed antibiotics over the study period were not increasing (monthly percentage change 0.3, 95% CI −0.3–0.9). There was no significant increase in the dispensing rates of fluoroquinolone antibiotics, cephalosporin antibiotics or broad-spectrum antibiotics. There was a
minor increase in total hospital admissions (monthly percentage change 0.3, 95% CI 0.0–0.6). The increase is not accounted for by a proportionate increase in the number of stool samples tested (2010 – 2100 (175/ month), 2011 – 2400 (200/month), 2012 – 1300 (217/month)). Sixtythree cases (12%) were culture-negative but toxin-positive. Genotypes of C. difficile colitis were collected in some cases but only when specifically sought by the primary referrer. With regard to the cohort of patients that tested positive for CDI, 70% had undergone treatment with beta lactam or fluoroquinolone-based antibiotics 4 weeks prior to testing. Forty-two percent were aged greater than 75 years and 9% were noted to be obese with a body mass index greater than 30. Twenty-six percent were immunocompromised, either due to an underlying illness, treatment with immunosuppressive medications or chemotherapy. These are all recognized as being risk factors for the development of CDI. There were no reported mortalities in our cohort.
Case series Pseudo-obstruction Case 1 A 73-year-old man underwent elective closure of a loop ileostomy following previous laparoscopic anterior resection for rectal cancer. His post-operative course was complicated by a rapid exacerbation of known atrial fibrillation (AF) with accompanying fever. Helical computed tomography (CT) scan of the abdomen demonstrated a © 2014 Royal Australasian College of Surgeons
Challenges in Clostridium infection
colonic pseudo-obstruction. Resolution of the pseudo-obstruction was achieved with neostigmine. Diarrhoea ensued and therefore testing for clostridial toxin was performed. He was anti-coagulated for stroke prevention in the setting of AF. This resulted in a period of rectal bleeding and ongoing diarrhoea. The C. difficile culture and toxin proved positive, and an interval helical CT scan performed in the setting of bleeding demonstrated a pancolitis. The clinical symptoms resolved with oral vancomycin. Comment In this case, CDI was identified from stool cultures as the pseudoobstruction resolved. We have experienced one other case of intractable pseudo-obstruction with C. difficile-positive cultures that finally settled with medical treatment of the C. difficile and believe that CDI may in fact be a possible cause of some cases of colonic pseudo-obstruction. The case also shows that CDI may complicate a loop ileostomy closure. A single dose of ceftriaxone and metronidazole was used as prophylaxis intraoperatively. We have observed this in three other cases in our institution.
Crohn’s disease and CDI Case 2 A 19-year-old man was admitted following a laparoscopic appendicectomy with right iliac fossa (RIF) pain, fever and obstructive symptoms. Helical CT scan revealed a phlegmon in the RIF and stool cultures demonstrated CDI. Histopathology of the appendix showed chronic transmural inflammation without granulomata. Symptoms partially resolved with antibiotic therapy. Subsequent colonoscopy demonstrated aphthous ulcers throughout the colon and pus in the terminal ileum. Histopathology failed to demonstrate conclusive evidence of Crohn’s disease. The patient however failed to respond completely to antibiotic therapy, but had symptom resolution with intravenous (i.v.) steroids.
3
regarding an appendiceal stump leak in the former and an anastomotic leak in the latter case. The predominant feature was fever and not diarrhoea. A low threshold for C. difficile testing allowed both cases of CDI to be identified and treated.
Sepsis and recurrent CDI Case 4 A 77-year-old man was admitted with small bowel obstruction secondary to parastomal herniation. In the previous month, the patient had undergone a cystectomy for transitional cell carcinoma of the bladder. That admission had been complicated by C. difficile colitis, which was treated with 10 days of oral metronidazole. He presented with herniation of small bowel around his urostomy site. This resolved spontaneously. He then developed sepsis with fever and rigor preceding diarrhoea by 24 h and initially treated as urosepsis. Subsequently, C. difficile was demonstrated on stool cultures, and an interval helical CT scan confirmed pancolitis. A resolution of symptoms was achieved with administration of oral vancomycin. Comment The current case presents an unusual scenario where the patient had dual pathology – small bowel obstruction that resolved spontaneously and CDI. Again, fever pre-dated the symptoms of diarrhoea, but on first loose bowel motion a C. difficile toxin assay was sent, which confirmed the diagnosis. The issue with this gentleman is that he has a relapsing or recurrent course of C. difficile infection. Fidoxamicin (Dificid Optimer Pharmaceuticals, San Diego, CA, USA) is a new drug that has been introduced in the United States, and is now becoming available in Australia, and has been shown to be superior at preventing recurrences of CDI but is not presently available in Australia. More extreme measures to abate chronic courses of C. difficile colitis or prevent recurrent symptoms include monoclonal antibody therapy or faecal transplantation.
Fulminant colitis Case 3 A 43-year-old woman presented with nausea, vomiting, fever and bloating one day following discharge from hospital after undergoing a redo ileocolic resection for high-grade fibrostenotic ileal Crohn’s disease. The patient had undergone previous resections in 2000 and 2009 and prior to surgery had been medically managed with budesonide, adalimumab (anti-tumour necrosis factor therapy) and azathioprine. The adalimumab therapy was ceased preoperatively. Helical CT scan showed ileus and a collection remote from her anastomotic site, which was demonstrated to be a non-infected seroma by CT-guided aspiration. Stool cultures demonstrated C. difficile. The patient’s symptoms resolved with oral vancomycin and i.v. metronidazole therapy as well as supportive measures for her ileus. Comment The two aforementioned cases demonstrate that CDI can complicate patients with Crohn’s disease. In both cases, the presence of a toxin-positive result presented a diagnostic dilemma. In the first, the probable diagnosis of Crohn’s disease was not recognized at index surgery. Both presented with collections and initial concerns existed © 2014 Royal Australasian College of Surgeons
Case 5 A 46-year-old woman was referred to our service with fulminant C. difficile colitis having failed 48 h of medical therapy (i.v. metronidazole and oral vancomycin). The patient had previously undergone an elective cervical discectomy with anterior fixation for C6 nerve root impingement. In the resultant rehabilitation phase of her care, she had been placed on a short course of broad-spectrum antibiotics for abdominal pain symptoms in the absence of a clear diagnosis. When the patient arrived at our hospital, she was tachycardic, hypotensive with high fever and diarrhoea with stool cultures that confirmed CDI. Following resuscitative measures, operative management consisted of forming a laparoscopic loop ileostomy, intubating the efferent limb and irrigating the colon. The colon was irrigated with 8 L of polyethylene glycol (PEG). In the post-operative period, the patient received antegrade QID vancomycin colonic flushes (250 mg) through the ileostomy for 10 days. This produced a dramatic resolution of her colitis clinically within the first 48 h of therapy. Her treatment was complicated by line sepsis, but 2 weeks after surgery, she returned to her spinal rehabilitation facility.
4
Buxey et al.
Comment This case highlights an unusual presentation of fulminant segmental colitis requiring operative intervention that did not resolve with medical therapy. Again, the colitis occurred in the setting of C. difficile-positive testing, and at the time of operative intervention, it was thought that Crohn’s colitis was the more likely aetiology. This suspicion was not validated by the eventual histopathology. An anastomosis was performed as normal ileum could be joined to macroscopically normal colon. The risk of anastomotic failure was therefore deemed acceptable and the patient recovered well.
Discussion
Fig. 3. The colon lavage technique that was used for the fulminant colitis case.
Comment The case illustrates an acute severe colitic patient (due to CDI) that requires surgical intervention. The gold standard treatment for this disease is a subtotal colectomy with end ileostomy. In part due to delays of diagnosis and a reluctance to operate on CDI, the patients who undergo emergent colectomy are sick, and hence surgical outcomes are poor.7 A recent publication by Neal and colleague have shown a significant reduction in morbidity and mortality with creation of diverting ileostomy and antegrade lavage of the colon with subsequent antegrade vancomycin enemas of the efferent limb of the ileostomy. The case is the first Australian report utilizing this technique. The technique is illustrated in Figure 3.
Segmental colitis Case 6 A 60-year-old man presented with left iliac fossa pain and helical CT scan demonstrated mild diverticulitis. He was treated with i.v. metronidazole and ceftriaxone. Within two days, his original pain resolved, but he developed right-sided abdominal pain associated with a segmental colitis involving the right colon on CT. There was no diarrhoea or rectal bleeding at any stage. Subsequent colonoscopy showed a segmental colitis from the caecum to the hepatic flexure, with no evidence of colitis or diverticulitis elsewhere. He failed to respond to i.v. hydrocortisone, metronidazole and ceftriaxone. Stool cultures were only possible late in the disease with a slightly looser motion, and this proved positive for C. difficile. He developed toxic dilation and clinical signs of impending perforation while on treatment and thus required a right hemicolectomy. The pathology showed a non-specific colitis and C. difficile, with no evidence of Crohn’s disease.
The current study shows that the number of cases of toxin-positive and culture-positive C. difficile cases is increasing in an Australian tertiary institution. We have experienced an 8% monthly increase in cases and advocate methods to improve prevention of CDI transmission. Increasing usage of antibiotics or simply increasing the number of patients a hospital treats would both be possible confounding elements to consider when considering the reason for a rise in detection of CDI. This, however, does not appear to be the case as data from the National Antimicrobial Utilisation Surveillance Program demonstrate stable prescribing of antibiotics with no increase in the subgroup most likely to be associated with CDI. Institutional data also do not support the increasing numbers of CDI being simply a function of treating more patients. With regard to other possible confounding factors, we have looked at obesity and immunodeficiency within our cohort. To what end these factors may be associated with increasing CDI is difficult to say. We do not have data available to make ready comparison to. It is possible that changes in patient populations with respect to representation of these two risk factors for CDI may play some role in the increase in numbers but that is something we can only speculate on at this time and may be an area of future research interest. What is clear with an increase in CDI cases is a need for increased vigilance and awareness of this important condition. In addition, the case series challenges the dictum of CDI presenting as pseudomembranous colitis with symptoms of diarrhoea, abdominal pain, nausea and vomiting. The diverse clinical patterns presented include symptoms of a colonic pseudo-obstruction, appendicitis, inflammatory bowel disease, segmental colitis as well as possible anastomotic leak. Due to the diverse range of symptoms, the authors would advocate a high level of vigilance to diagnose CDI. This message is particularly important as more harm may be done if the disease is not diagnosed (e.g. by continuing broadspectrum antibiotics). The current study highlights challenges that clinicians may face in the diagnosis of CDI. CDI is diagnosed on the basis of stool tests. These tests can be divided into toxin and organism detection assays. The current series shows that patients with CDI may have an ileus or pseudo-obstruction, and in such circumstances, stool sampling may be problematic.8 Often placement of a rectal tube with liquidy bowel motions is enough to gain a sample for analysis. Infection results when organisms release toxin A or B (usually both but on occasion one or the other). Cytotoxicity assays are the most sensitive and specific (94–100% and 100%, respectively) of the toxin assays;9,10 however, they are costly and have a 48-h © 2014 Royal Australasian College of Surgeons
Challenges in Clostridium infection
turnaround time, which may limit their clinical utility. Enzyme immunoassays are used by most laboratories and have specificities of 99%. Unfortunately, their sensitivity is reported to be as low as 60%.11 Polymerase chain reaction testing has been shown to have sensitivity and specificity of 93 and 97% with rapid turnaround times.12 At present, its availability in Australia is still limited. Our data show a 4.9% increase in CDI. The medium used for detection was altered during the study period; however, there was no point of transition identified in our statistical analysis to suggest that this was the reason for the change. Organism detection assays consist of either common antigen testing, which involves testing for an antigen borne by all C. difficile organisms including non-pathogenic strains. Stool culture is very sensitive but lacks both turnaround time (72 h) and the ability to distinguish non-toxin-elaborating resident bowel flora from pathogenic toxin elaborating organisms.13 In the current study, a 13.3% monthly percentage increase in toxin-positive or culture-positive cases was observed. Current guidelines suggest treating patients on clinical merits if toxin is negative but stool culture is positive. In cases where diagnosis is suspected but either the laboratory assays are negative or the patient cannot provide a stool sample (ileus or pseudo-obstruction), colonoscopy can also aid in diagnosis. Patients with mild C. difficile colitis can be treated with oral metronidazole. Where disease severity is worse, then dual therapy is recommended with i.v. metronidazole and oral vancomycin administered concomitantly. An alternative agent, fidaxomicin, is available in the United States, but is not yet available in Australia. This agent is a macrolide antibiotic with a narrow spectrum of activity, leading to less disruption of the colonic flora. It is associated with reduced recurrence rates when compared to conventional therapy but is not more active against the NAP-1 strain. Monoclonal antibodies against CDT are thought to reduce recurrence rates of infection10 but are not in routine clinical use, and are the subject of ongoing clinical trials that our hospital is involved. In the majority of the cases presented in the case series, dual therapy was used to help achieve resolution of symptoms. Patients who fail medical therapy require surgical management. Traditionally, patients who required surgery underwent emergent subtotal colectomy and end ileostomy formation. Primary anastomosis in the setting of fulminant pancolitis is not considered safe due to the risk of anastomotic failure. In this study, we present an alternative to this approach which is the first reported case in Australia utilizing this technique. A diverting loop ileostomy is created in the sick patient, therefore reducing the insult and timing of the procedure. In our case, it was performed laparoscopically. The colon is then lavaged with PEG solution and post-operatively the patient receives QID antegrade vancomycin flushes through the ileostomy for 10 days. The approach has been championed by Neale and colleagues from the University of Pittsburgh Medical Center (UPMC), PN. The authors compared 42 patients with severe colitis who received this treatment with historical controls who had undergone colectomy. In this study, the mortality difference was 19% versus 50% and achieved statistical significance.3 Both patient cohorts were matched in terms of disease severity and patient instability at the time of procedures. Further studies and longitudinal data are required to validate this approach. In theory, it makes sense as the Clostridium spores can be washed away © 2014 Royal Australasian College of Surgeons
5
and then antibiotics are delivered topically to the area of concern. Our patients’ clinical symptoms resolved quickly and support the contention that this is a valid means to treat such patients.
Conclusion CDI can present in various clinical forms. In our hospital, the number of cases of toxin-positive and culture-positive detection is increasing. This is not related to increasing antibiotics usage or increasing treatment of patients, but it is possibly due to other confounding elements – this is something we cannot be certain. A low threshold is required to identify and adequately treat patients with CDI. Fulminant colitis can be managed successfully with the creation of a diverting loop ileostomy, colonic washout and subsequent antegrade colonic vancomycin enemas.
Acknowledgement The authors would like to thank Dr Raf Ratinam for providing Figure 3.
References 1. Viscidi R, Wiley S, Bartlett JG. Isolation rates and toxigenic potential of Clostridium difficile isolates from various patient populations. Gastroenterology 1981; 81: 5–9. 2. Redelings MD, Sorvill F, Mascola L. Increase in Clostridium difficilerelated mortality rates, United States, 1999–2004. Emerg. Infect. Dis. 2007; 13: 1417–9. 3. McDonald LC, Kilgrove GE, Thompson A et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N. Engl. J. Med. 2005; 353: 2433–41. 4. Riley TV, Thean S, Hool G, Golledge CL. First Australian isolation of epidemic Clostridium difficile PCR ribotype 027. Med. J. Aust. 2009; 190: 706–8. 5. Cheng AC, Ferguson JK, Richards MJ. Australiasian Society for Infectious Diseases guidelines for the diagnosis and treatment of Clostridium difficile infection. Med. J. Aust. 2011; 194: 353–8. 6. Neal MD, Alverdy JC, Hall DE, Simmons RL, Zuckerbraun BS. Diverting loop ileostomy and colonic lavage: an alternative to total abdominal colectomy or the treatment of severe, complicated Clostridium difficile associated disease. Ann. Surg. 2011; 254: 423–9. 7. Connolly S, Ezekowitz M, Yusuf S et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med 2009; 361: 1139–51. 8. Peterson L, Robicsek A. Does my patient have Clostridium difficile infection? Ann. Intern. Med. 2009; 151: 176–9. 9. Laughon BE, Viscidi RP, Gdovin SL, Yolken RH, Bartlett JG. Enzyme immunoassays for detection of Clostridium difficile toxins A and B in fecal specimens. J. Infect. Dis. 1984; 149: 781–8. 10. Shanholtzer CJ, Willard KE, Holter JJ, Olson MM, Gerding DN, Peterson LR. Comparison of the VIDAS Clostridium difficile toxin A immunoassay with C. difficile culture and cytotoxin and latex tests. J. Clin. Microbiol. 1992; 30: 1837–40. 11. Blossom D, McDonald L. The challenges posed by reemerging Clostridium difficile infection. Clin. Infect. Dis. 2007; 45: 222–7. 12. Huang H, Weintraub A, Fang H, Nord C. Comparison of a commercial multiplex real-time PCR to the cell cytotoxicity neutralisation assay for diagnosis of Clostridium difficile infections. J. Clin. Microbiol. 2009; 47: 3729–31. 13. Lowy I, Molrine D, Leave B et al. Treatment with monoclonal antibodies against Clostridium difficile toxins. N Engl J Med 2010; 362: 197–205.
Clostridium colitis: challenges in diagnosis and treatment Kenneth N. Buxey, Chris Sia, Stephen Bell, Roger Wale, Daniel Wein and Satish K. Warrier Department of Colorectal Surgery, The Alfred Hospital, Melbourne, Victoria, Australia
Key words Clostridium difficile, colitis. Correspondence Mr Satish K. Warrier, Department of Colorectal Surgery, The Alfred Hospital, High Street, Melbourne, Vic. 3181, Australia. Email: [email protected] K. N. Buxey MBBS (Hons); C. Sia MBBS; S. Bell MBBS, FRACS; R. Wale MBBS, FRACS; D. Wein MBBS Candidate; S. K. Warrier MBBS, FRACS. Accepted for publication 3 August 2014. doi: 10.1111/ans.12840
Abstract Background: Clostridium difficile infection (CDI) has been reported to occur with increasing frequency and with more severe presentations being encountered. This article presents data from The Alfred Hospital highlighting the increased incidence, the increased severity and the broader clinical presentations observed. A case series highlights a variety of clinical scenarios that provided diagnostic and management challenges. We additionally describe a novel form of treatment for fulminant colitis. Methods: A retrospective review of C. difficile toxin (CDT)-positive and culturepositive cases was performed at The Alfred Hospital (2010–2012). Six cases are then presented as a case series to highlight the broad and atypical types of presentations one may encounter. Finally, a novel method for managing fulminant colitis operatively is presented. Results: A fourfold increase in cases of toxin-positive and culture-positive cases was noted over the initial 14 months of the period of analysis, the rate of cases detected then plateaued. This increase could not be explained by increased testing being undertaken. It is also not associated with increased usage of antibiotics nor with increased patient numbers being treated. Conclusion: CDI can present in various clinical forms. In our hospital, the number of cases of toxin-positive and culture-positive detection is increasing. A low threshold is required to identify and adequately treat patients with CDI. Fulminant colitis can be managed successfully with the creation of a diverting loop ileostomy, colonic washout and subsequent antegrade colonic vancomycin enemas.
Introduction Clostridium difficile infection (CDI) commonly results in nosocomial diarrhoea.1 In the hospital setting, it often arises following a period of antibiotic therapy. In the United States, there has been an increase in cases of severe colitis.2 We have also noted an increase in the absolute number of C. difficile toxin (CDT)-positive cases at our institution over the last three years. This increase is not accounted for by a proportional increase in Clostridium testing. A hypervirulent strain of CDT known as the NAP-1, or PCR ribotype 027 subtype, has been identified. This strain produces the two major toxins (A and B) as well as an additional potentially virulent toxin (binary toxin).3 The strain, while commonly found in both the United States and European countries, has more recently been isolated in Australia.4,5 Toxin A is an enterotoxin and produces damage to the colonic mucosa, whereas toxin B is a cytotoxin and has 10 times the potency. Not surprisingly, infection with this pathogen can result in acute severe colitis, a source of major morbidity and can even result in © 2014 Royal Australasian College of Surgeons
death. In one recent study, the mortality of severe colitis requiring subtotal colectomy secondary to CDI was 50%.6 With an increasing prevalence and virulence of Clostridium colitis, the aim of the current study is to report on our hospital experience with CDI, in particular highlighting the increased incidence, the increased severity, and the broader clinical presentations observed. The case series highlights a variety of clinical scenarios that provided diagnostic and management challenges at our institution. It also highlights a case of a novel form of treatment for fulminant colitis.
Methods A retrospective review of CDT-positive and culture-positive cases was performed at The Alfred Hospital (2010–2012). Results were attained from the microbiological laboratory at the institution. Ethics approval was granted for the study. Specific strains of C. difficile are not included in our data. ANZ J Surg •• (2014) ••–••
2
Buxey et al.
Fig. 1. The trend of Clostridium difficile toxin-positive cases per month. ( ) Toxin-positive cases.
Fig. 2. The trend of Clostridium difficile toxin- or culture-positive cases per month. ( ) Toxin- or culture-positive cases.
Jointpoint regression program version 4.0.1 was used to calculate the monthly percentage change. From these cases, six with varied clinical presentations of CDI were identified. These are presented as a case series to highlight the varied clinical manifestations of the disease. All patients in the case series required colorectal surgical input within the last 12-month period.
Results Clostridium toxin-positive and culture-positive cases (2010–2012) From June 2010 to May 2012, the monthly percentage change in toxinpositive cases was 8.0% (95% confidence interval (CI) = 5.4–10.7). This trend is shown in Figure 1. In absolute figures, this equated to a fourfold rise in toxin-positive cases (from 2010 – 7.7 toxin-positive cases/month to 2012 – 29.2 toxin-positive cases/month). This trend is mirrored by an equal increase in culture-positive or toxin-positive cases, in which an 8.8% monthly percentage change was observed (95% CI 5.9–11.8). This change is shown in Figure 2. There was no significant increase in antibiotic prescribing rates to explain this trend. The collective defined daily doses of all prescribed antibiotics over the study period were not increasing (monthly percentage change 0.3, 95% CI −0.3–0.9). There was no significant increase in the dispensing rates of fluoroquinolone antibiotics, cephalosporin antibiotics or broad-spectrum antibiotics. There was a
minor increase in total hospital admissions (monthly percentage change 0.3, 95% CI 0.0–0.6). The increase is not accounted for by a proportionate increase in the number of stool samples tested (2010 – 2100 (175/ month), 2011 – 2400 (200/month), 2012 – 1300 (217/month)). Sixtythree cases (12%) were culture-negative but toxin-positive. Genotypes of C. difficile colitis were collected in some cases but only when specifically sought by the primary referrer. With regard to the cohort of patients that tested positive for CDI, 70% had undergone treatment with beta lactam or fluoroquinolone-based antibiotics 4 weeks prior to testing. Forty-two percent were aged greater than 75 years and 9% were noted to be obese with a body mass index greater than 30. Twenty-six percent were immunocompromised, either due to an underlying illness, treatment with immunosuppressive medications or chemotherapy. These are all recognized as being risk factors for the development of CDI. There were no reported mortalities in our cohort.
Case series Pseudo-obstruction Case 1 A 73-year-old man underwent elective closure of a loop ileostomy following previous laparoscopic anterior resection for rectal cancer. His post-operative course was complicated by a rapid exacerbation of known atrial fibrillation (AF) with accompanying fever. Helical computed tomography (CT) scan of the abdomen demonstrated a © 2014 Royal Australasian College of Surgeons
Challenges in Clostridium infection
colonic pseudo-obstruction. Resolution of the pseudo-obstruction was achieved with neostigmine. Diarrhoea ensued and therefore testing for clostridial toxin was performed. He was anti-coagulated for stroke prevention in the setting of AF. This resulted in a period of rectal bleeding and ongoing diarrhoea. The C. difficile culture and toxin proved positive, and an interval helical CT scan performed in the setting of bleeding demonstrated a pancolitis. The clinical symptoms resolved with oral vancomycin. Comment In this case, CDI was identified from stool cultures as the pseudoobstruction resolved. We have experienced one other case of intractable pseudo-obstruction with C. difficile-positive cultures that finally settled with medical treatment of the C. difficile and believe that CDI may in fact be a possible cause of some cases of colonic pseudo-obstruction. The case also shows that CDI may complicate a loop ileostomy closure. A single dose of ceftriaxone and metronidazole was used as prophylaxis intraoperatively. We have observed this in three other cases in our institution.
Crohn’s disease and CDI Case 2 A 19-year-old man was admitted following a laparoscopic appendicectomy with right iliac fossa (RIF) pain, fever and obstructive symptoms. Helical CT scan revealed a phlegmon in the RIF and stool cultures demonstrated CDI. Histopathology of the appendix showed chronic transmural inflammation without granulomata. Symptoms partially resolved with antibiotic therapy. Subsequent colonoscopy demonstrated aphthous ulcers throughout the colon and pus in the terminal ileum. Histopathology failed to demonstrate conclusive evidence of Crohn’s disease. The patient however failed to respond completely to antibiotic therapy, but had symptom resolution with intravenous (i.v.) steroids.
3
regarding an appendiceal stump leak in the former and an anastomotic leak in the latter case. The predominant feature was fever and not diarrhoea. A low threshold for C. difficile testing allowed both cases of CDI to be identified and treated.
Sepsis and recurrent CDI Case 4 A 77-year-old man was admitted with small bowel obstruction secondary to parastomal herniation. In the previous month, the patient had undergone a cystectomy for transitional cell carcinoma of the bladder. That admission had been complicated by C. difficile colitis, which was treated with 10 days of oral metronidazole. He presented with herniation of small bowel around his urostomy site. This resolved spontaneously. He then developed sepsis with fever and rigor preceding diarrhoea by 24 h and initially treated as urosepsis. Subsequently, C. difficile was demonstrated on stool cultures, and an interval helical CT scan confirmed pancolitis. A resolution of symptoms was achieved with administration of oral vancomycin. Comment The current case presents an unusual scenario where the patient had dual pathology – small bowel obstruction that resolved spontaneously and CDI. Again, fever pre-dated the symptoms of diarrhoea, but on first loose bowel motion a C. difficile toxin assay was sent, which confirmed the diagnosis. The issue with this gentleman is that he has a relapsing or recurrent course of C. difficile infection. Fidoxamicin (Dificid Optimer Pharmaceuticals, San Diego, CA, USA) is a new drug that has been introduced in the United States, and is now becoming available in Australia, and has been shown to be superior at preventing recurrences of CDI but is not presently available in Australia. More extreme measures to abate chronic courses of C. difficile colitis or prevent recurrent symptoms include monoclonal antibody therapy or faecal transplantation.
Fulminant colitis Case 3 A 43-year-old woman presented with nausea, vomiting, fever and bloating one day following discharge from hospital after undergoing a redo ileocolic resection for high-grade fibrostenotic ileal Crohn’s disease. The patient had undergone previous resections in 2000 and 2009 and prior to surgery had been medically managed with budesonide, adalimumab (anti-tumour necrosis factor therapy) and azathioprine. The adalimumab therapy was ceased preoperatively. Helical CT scan showed ileus and a collection remote from her anastomotic site, which was demonstrated to be a non-infected seroma by CT-guided aspiration. Stool cultures demonstrated C. difficile. The patient’s symptoms resolved with oral vancomycin and i.v. metronidazole therapy as well as supportive measures for her ileus. Comment The two aforementioned cases demonstrate that CDI can complicate patients with Crohn’s disease. In both cases, the presence of a toxin-positive result presented a diagnostic dilemma. In the first, the probable diagnosis of Crohn’s disease was not recognized at index surgery. Both presented with collections and initial concerns existed © 2014 Royal Australasian College of Surgeons
Case 5 A 46-year-old woman was referred to our service with fulminant C. difficile colitis having failed 48 h of medical therapy (i.v. metronidazole and oral vancomycin). The patient had previously undergone an elective cervical discectomy with anterior fixation for C6 nerve root impingement. In the resultant rehabilitation phase of her care, she had been placed on a short course of broad-spectrum antibiotics for abdominal pain symptoms in the absence of a clear diagnosis. When the patient arrived at our hospital, she was tachycardic, hypotensive with high fever and diarrhoea with stool cultures that confirmed CDI. Following resuscitative measures, operative management consisted of forming a laparoscopic loop ileostomy, intubating the efferent limb and irrigating the colon. The colon was irrigated with 8 L of polyethylene glycol (PEG). In the post-operative period, the patient received antegrade QID vancomycin colonic flushes (250 mg) through the ileostomy for 10 days. This produced a dramatic resolution of her colitis clinically within the first 48 h of therapy. Her treatment was complicated by line sepsis, but 2 weeks after surgery, she returned to her spinal rehabilitation facility.
4
Buxey et al.
Comment This case highlights an unusual presentation of fulminant segmental colitis requiring operative intervention that did not resolve with medical therapy. Again, the colitis occurred in the setting of C. difficile-positive testing, and at the time of operative intervention, it was thought that Crohn’s colitis was the more likely aetiology. This suspicion was not validated by the eventual histopathology. An anastomosis was performed as normal ileum could be joined to macroscopically normal colon. The risk of anastomotic failure was therefore deemed acceptable and the patient recovered well.
Discussion
Fig. 3. The colon lavage technique that was used for the fulminant colitis case.
Comment The case illustrates an acute severe colitic patient (due to CDI) that requires surgical intervention. The gold standard treatment for this disease is a subtotal colectomy with end ileostomy. In part due to delays of diagnosis and a reluctance to operate on CDI, the patients who undergo emergent colectomy are sick, and hence surgical outcomes are poor.7 A recent publication by Neal and colleague have shown a significant reduction in morbidity and mortality with creation of diverting ileostomy and antegrade lavage of the colon with subsequent antegrade vancomycin enemas of the efferent limb of the ileostomy. The case is the first Australian report utilizing this technique. The technique is illustrated in Figure 3.
Segmental colitis Case 6 A 60-year-old man presented with left iliac fossa pain and helical CT scan demonstrated mild diverticulitis. He was treated with i.v. metronidazole and ceftriaxone. Within two days, his original pain resolved, but he developed right-sided abdominal pain associated with a segmental colitis involving the right colon on CT. There was no diarrhoea or rectal bleeding at any stage. Subsequent colonoscopy showed a segmental colitis from the caecum to the hepatic flexure, with no evidence of colitis or diverticulitis elsewhere. He failed to respond to i.v. hydrocortisone, metronidazole and ceftriaxone. Stool cultures were only possible late in the disease with a slightly looser motion, and this proved positive for C. difficile. He developed toxic dilation and clinical signs of impending perforation while on treatment and thus required a right hemicolectomy. The pathology showed a non-specific colitis and C. difficile, with no evidence of Crohn’s disease.
The current study shows that the number of cases of toxin-positive and culture-positive C. difficile cases is increasing in an Australian tertiary institution. We have experienced an 8% monthly increase in cases and advocate methods to improve prevention of CDI transmission. Increasing usage of antibiotics or simply increasing the number of patients a hospital treats would both be possible confounding elements to consider when considering the reason for a rise in detection of CDI. This, however, does not appear to be the case as data from the National Antimicrobial Utilisation Surveillance Program demonstrate stable prescribing of antibiotics with no increase in the subgroup most likely to be associated with CDI. Institutional data also do not support the increasing numbers of CDI being simply a function of treating more patients. With regard to other possible confounding factors, we have looked at obesity and immunodeficiency within our cohort. To what end these factors may be associated with increasing CDI is difficult to say. We do not have data available to make ready comparison to. It is possible that changes in patient populations with respect to representation of these two risk factors for CDI may play some role in the increase in numbers but that is something we can only speculate on at this time and may be an area of future research interest. What is clear with an increase in CDI cases is a need for increased vigilance and awareness of this important condition. In addition, the case series challenges the dictum of CDI presenting as pseudomembranous colitis with symptoms of diarrhoea, abdominal pain, nausea and vomiting. The diverse clinical patterns presented include symptoms of a colonic pseudo-obstruction, appendicitis, inflammatory bowel disease, segmental colitis as well as possible anastomotic leak. Due to the diverse range of symptoms, the authors would advocate a high level of vigilance to diagnose CDI. This message is particularly important as more harm may be done if the disease is not diagnosed (e.g. by continuing broadspectrum antibiotics). The current study highlights challenges that clinicians may face in the diagnosis of CDI. CDI is diagnosed on the basis of stool tests. These tests can be divided into toxin and organism detection assays. The current series shows that patients with CDI may have an ileus or pseudo-obstruction, and in such circumstances, stool sampling may be problematic.8 Often placement of a rectal tube with liquidy bowel motions is enough to gain a sample for analysis. Infection results when organisms release toxin A or B (usually both but on occasion one or the other). Cytotoxicity assays are the most sensitive and specific (94–100% and 100%, respectively) of the toxin assays;9,10 however, they are costly and have a 48-h © 2014 Royal Australasian College of Surgeons
Challenges in Clostridium infection
turnaround time, which may limit their clinical utility. Enzyme immunoassays are used by most laboratories and have specificities of 99%. Unfortunately, their sensitivity is reported to be as low as 60%.11 Polymerase chain reaction testing has been shown to have sensitivity and specificity of 93 and 97% with rapid turnaround times.12 At present, its availability in Australia is still limited. Our data show a 4.9% increase in CDI. The medium used for detection was altered during the study period; however, there was no point of transition identified in our statistical analysis to suggest that this was the reason for the change. Organism detection assays consist of either common antigen testing, which involves testing for an antigen borne by all C. difficile organisms including non-pathogenic strains. Stool culture is very sensitive but lacks both turnaround time (72 h) and the ability to distinguish non-toxin-elaborating resident bowel flora from pathogenic toxin elaborating organisms.13 In the current study, a 13.3% monthly percentage increase in toxin-positive or culture-positive cases was observed. Current guidelines suggest treating patients on clinical merits if toxin is negative but stool culture is positive. In cases where diagnosis is suspected but either the laboratory assays are negative or the patient cannot provide a stool sample (ileus or pseudo-obstruction), colonoscopy can also aid in diagnosis. Patients with mild C. difficile colitis can be treated with oral metronidazole. Where disease severity is worse, then dual therapy is recommended with i.v. metronidazole and oral vancomycin administered concomitantly. An alternative agent, fidaxomicin, is available in the United States, but is not yet available in Australia. This agent is a macrolide antibiotic with a narrow spectrum of activity, leading to less disruption of the colonic flora. It is associated with reduced recurrence rates when compared to conventional therapy but is not more active against the NAP-1 strain. Monoclonal antibodies against CDT are thought to reduce recurrence rates of infection10 but are not in routine clinical use, and are the subject of ongoing clinical trials that our hospital is involved. In the majority of the cases presented in the case series, dual therapy was used to help achieve resolution of symptoms. Patients who fail medical therapy require surgical management. Traditionally, patients who required surgery underwent emergent subtotal colectomy and end ileostomy formation. Primary anastomosis in the setting of fulminant pancolitis is not considered safe due to the risk of anastomotic failure. In this study, we present an alternative to this approach which is the first reported case in Australia utilizing this technique. A diverting loop ileostomy is created in the sick patient, therefore reducing the insult and timing of the procedure. In our case, it was performed laparoscopically. The colon is then lavaged with PEG solution and post-operatively the patient receives QID antegrade vancomycin flushes through the ileostomy for 10 days. The approach has been championed by Neale and colleagues from the University of Pittsburgh Medical Center (UPMC), PN. The authors compared 42 patients with severe colitis who received this treatment with historical controls who had undergone colectomy. In this study, the mortality difference was 19% versus 50% and achieved statistical significance.3 Both patient cohorts were matched in terms of disease severity and patient instability at the time of procedures. Further studies and longitudinal data are required to validate this approach. In theory, it makes sense as the Clostridium spores can be washed away © 2014 Royal Australasian College of Surgeons
5
and then antibiotics are delivered topically to the area of concern. Our patients’ clinical symptoms resolved quickly and support the contention that this is a valid means to treat such patients.
Conclusion CDI can present in various clinical forms. In our hospital, the number of cases of toxin-positive and culture-positive detection is increasing. This is not related to increasing antibiotics usage or increasing treatment of patients, but it is possibly due to other confounding elements – this is something we cannot be certain. A low threshold is required to identify and adequately treat patients with CDI. Fulminant colitis can be managed successfully with the creation of a diverting loop ileostomy, colonic washout and subsequent antegrade colonic vancomycin enemas.
Acknowledgement The authors would like to thank Dr Raf Ratinam for providing Figure 3.
References 1. Viscidi R, Wiley S, Bartlett JG. Isolation rates and toxigenic potential of Clostridium difficile isolates from various patient populations. Gastroenterology 1981; 81: 5–9. 2. Redelings MD, Sorvill F, Mascola L. Increase in Clostridium difficilerelated mortality rates, United States, 1999–2004. Emerg. Infect. Dis. 2007; 13: 1417–9. 3. McDonald LC, Kilgrove GE, Thompson A et al. An epidemic, toxin gene-variant strain of Clostridium difficile. N. Engl. J. Med. 2005; 353: 2433–41. 4. Riley TV, Thean S, Hool G, Golledge CL. First Australian isolation of epidemic Clostridium difficile PCR ribotype 027. Med. J. Aust. 2009; 190: 706–8. 5. Cheng AC, Ferguson JK, Richards MJ. Australiasian Society for Infectious Diseases guidelines for the diagnosis and treatment of Clostridium difficile infection. Med. J. Aust. 2011; 194: 353–8. 6. Neal MD, Alverdy JC, Hall DE, Simmons RL, Zuckerbraun BS. Diverting loop ileostomy and colonic lavage: an alternative to total abdominal colectomy or the treatment of severe, complicated Clostridium difficile associated disease. Ann. Surg. 2011; 254: 423–9. 7. Connolly S, Ezekowitz M, Yusuf S et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med 2009; 361: 1139–51. 8. Peterson L, Robicsek A. Does my patient have Clostridium difficile infection? Ann. Intern. Med. 2009; 151: 176–9. 9. Laughon BE, Viscidi RP, Gdovin SL, Yolken RH, Bartlett JG. Enzyme immunoassays for detection of Clostridium difficile toxins A and B in fecal specimens. J. Infect. Dis. 1984; 149: 781–8. 10. Shanholtzer CJ, Willard KE, Holter JJ, Olson MM, Gerding DN, Peterson LR. Comparison of the VIDAS Clostridium difficile toxin A immunoassay with C. difficile culture and cytotoxin and latex tests. J. Clin. Microbiol. 1992; 30: 1837–40. 11. Blossom D, McDonald L. The challenges posed by reemerging Clostridium difficile infection. Clin. Infect. Dis. 2007; 45: 222–7. 12. Huang H, Weintraub A, Fang H, Nord C. Comparison of a commercial multiplex real-time PCR to the cell cytotoxicity neutralisation assay for diagnosis of Clostridium difficile infections. J. Clin. Microbiol. 2009; 47: 3729–31. 13. Lowy I, Molrine D, Leave B et al. Treatment with monoclonal antibodies against Clostridium difficile toxins. N Engl J Med 2010; 362: 197–205.
