Although kidney transplantation is recognized as the ideal treatment option for end-stage renal disease patients,1 the requirement for long-term maintenance immunosuppression to preserve the function of the graft renders kidney transplant recipients (KTR) vulnerable to infections. In particular, the use of immunosuppression and antibiotics, both of which are ubiquitous in the kidney transplant population, are established risk factors for 1 particular infectious agent, Clostridium difficile.2-4
Clostridium difficile is a Gram-positive bacteria that colonizes the intestinal tract.4C. difficile is transmitted via the fecal-oral route and is capable of forming extremely resistant spores; this promotes its spread in nosocomial settings.4 Colonization may be asymptomatic, and disease is caused by the proliferation of toxigenic strains.4 Usually constrained by the indigenous intestinal microbiota, C. difficile may proliferate and cause symptoms when the protection of the microbiota is disrupted by the use of broad-spectrum antibiotics.4 Signs and symptoms are caused by toxins produced by the bacteria and include diarrhea, abdominal pain, pseudomembranous colitis, which can progress to toxic megacolon, and even death.3-5
The reported incidence of C. difficile infections (CDI) in KTR varies greatly from less than 1% to 8%.5-7 The incidence of CDI in the general population has been shown to be on the rise, with an increase in the United States from 0.45% of all discharges in 2001 to 0.82% of all discharges in 2010,8 with similar increases in Canada.9 The increase in incidence has also been linked to an increase in morbidity and mortality from these infections.10 Immunosuppressive regimens required posttransplant to prevent rejection, antibiotics for both prophylaxis and treatment, and frequent exposure to healthcare settings all position KTR to be at increased risk for severe CDI. Despite this, the presentation of CDI, as well as transplant-related risk factors and graft-related outcomes associated with CDI, remains poorly understood in the KTR population. Thus, the aim of this observational cohort study was to comprehensively explore the incidence, risk factors, and outcomes for CDI in a large contemporary cohort of KTR.
MATERIALS AND METHODS
Study Design and Population
An observational cohort study was conducted to analyze the incidence and outcomes of CDI, whereas a case-control study nested within this cohort was used for risk factor analysis. The population included all adult (age, ≥ 18 years) KTR at the Toronto General Hospital, who received a kidney transplant between January 1, 2000, and December 31, 2013. The exclusion criteria included (i) transplants done at other institutions, (ii) patients who had primary graft nonfunction, and (iii) patients who had prior multiorgan transplants. Data were obtained from an in-center research database, the Comprehensive Renal Transplant Research Information System,11 with additional elements collected from the electronic patient record and the Organ Transplant Tracking Record systems. For the risk factor analysis, patients were followed up until they developed CDI (for cases), and controls without CDI were matched to cases at the time of CDI occurrence. For the outcomes analysis, patients were followed up from kidney transplant to graft loss, death, or until December 31, 2014.
Maintenance Immunosuppression and Treatment for Biopsy-proven Acute Rejection
Most patients received induction therapy before transplant, comprised of either interleukin-2 receptor blocker or rabbit antithymocyte globulin. Some patients in the earlier era (ie, before 2007) did not receive induction based on physician discretion. Maintenance immunosuppression consisted of a calcineurin inhibitor, mycophenolic acid, and prednisone. Calcineurin inhibitor therapy was either a cyclosporine microemulsion (before 2007) or tacrolimus (since 2007). Patients also received mycophenolate mofetil or mycophenolate sodium. Prednisone was rapidly tapered to 5 mg daily at 1 week and maintained for the life of the allograft. A slower taper over 3-months was undertaken for patients experiencing delayed graft function or early biopsy-proven acute rejection (BPAR). Biopsy-proven acute rejection was diagnosed by kidney transplant biopsy and classified using Banff criteria.12 Depending on the type of BPAR, treatment included intravenous corticosteroid pulses, rabbit antithymocyte globulin, intravenous immunoglobulin, plasmapheresis, and/or rituximab.
Clostridium difficile Infection
Cases of CDI were defined as a positive laboratory test either by enzyme immunoassay (before 2010) or real-time polymerase chain reaction (after 2010). Laboratory test results were obtained directly from the University Health Network microbiology laboratory. The diagnosis of CDI was based on a positive laboratory test result and evidence of clinical symptoms (ie, diarrhea).
Potential Risk Factors for CDI
A nested case-control design was used to examine risk factors for CDI. This approach facilitated the efficient collection of data elements not found in the in-center research database. Controls without CDI were selected at a 4:1 ratio and matched to cases by transplant year and follow-up time (ie, risk-set sampling). Potential risk factors explored included recipient age at transplant, sex, recipient race, body mass index (BMI), cause of end-stage renal disease, time on dialysis before transplant, peak panel-reactive antibody status, donor age, donor type (deceased vs living), total days spent in hospital posttransplant (days), type of induction therapy, type of calcineurin inhibitor, BPAR, prior antibiotic treatment, prior proton pump inhibitor or histamine H2-receptor antagonist use, and a history of gastrointestinal surgery. The types of antibiotic treatments used were classified. Of note, prophylactic antibiotics, such as trimethoprim/sulfamethoxazole and dapsone, were excluded from the analysis due to their ubiquitous usage in all KTR posttransplant.
The primary outcome of the nested case-control was the development of a positive C. difficile laboratory test on a stool sample. The main outcomes for the cohort study were total graft failure (defined as either return to chronic dialysis, preemptive retransplant, or death with graft function) and BPAR. Additional risk factors, such as cold ischemia time, HLA mismatch, delayed graft function, BPAR before CDI, and transplant era, were considered potential confounders in the cohort study.
Stata/MP, version 12 (StataCorp, College Station, TX) was used for all statistical analyses. Donor, recipient, and transplant characteristics were summarized using descriptive statistics and expressed as mean and standard deviation or median and interquartile range. For the nested case-control study, conditional logistic regression models were used to assess risk factors for the development of the first CDI event. Backward stepwise selection of risk factors (P < 0.10) was used for covariate selection. Descriptive statistics were used to describe the types of antibiotics used by patients. Cumulative failure probabilities (eg, cumulative probability of developing BPAR) were analyzed using Kaplan-Meier survival curves, and statistical differences between curves were assessed using the log-rank test. Cox proportional hazards models were fitted for the cohort study examining graft outcomes with CDI handled as a time-dependent exposure. All statistical tests were 2-tailed, and a P value less than 0.05 was considered statistically significant. The study was approved by the University Health Network Research Ethics Board.
After applying the inclusion and exclusion criteria, the final study cohort included 1816 patients (Figure 1). The median follow-up time was 5.34 years, with a total follow-up of 10 694 person-years. The incidence rate was 0.64 cases per 100 person-years with a cumulative probability of CDI of 4.8% at 10 years posttransplant (Figure 2). There were 68 cases of first CDI over follow-up. A total of 272 controls were selected using risk-set sampling, resulting in a total of 340 patients for the nested case-control study. The median time to first CDI was 3.52 years. The majority of CDI events developed close to the time of transplant, with 35 of the 68 cases occurring within 6 months posttransplant.
Table 1 shows the baseline characteristics at transplant of the overall study cohort. There were approximately equal numbers of men and women, with the average age being 50 years. Most patients were white and had glomerulonephritis as their cause of end-stage renal disease. Compared with controls, patients who developed CDI generally had a longer time on dialysis before transplant and a higher prevalence of deceased donors (Table S1, SDC,http://links.lww.com/TP/B554). Cases had increased days spent in the hospital and a larger proportion had BPAR, antibiotic treatment, proton pump inhibitor usage, and gastrointestinal surgeries during the year before the development of CDI compared with controls.
Several recipient, donor, transplant, and treatment variables were assessed as potential risk factors for CDI in both univariable (Table 2) and multivariable (Table 3) conditional logistic regression models. In the univariable analysis, having a higher BMI (odds ratio [OR], 1.08; 95% confidence interval [CI], 1.01-1.16; P = 0.02), longer time on dialysis before transplant (OR, 1.16; 95% CI, 1.03-1.30; P = 0.01), deceased donor (OR, 4.64; 95% CI, 2.08-10.36; P < 0.001), increased length of posttransplant hospitalizations (OR, 1.05; 95% CI, 1.01-1.09; P = 0.01), and BPAR (OR, 7.46; 95% CI, 1.85-30.07; P = 0.01), antibiotic therapy (OR, 4.07; 95% CI, 1.68-9.88; P = 0.002), or gastrointestinal surgery (OR, 4.03; 95% CI, 1.23-13.20; P = 0.02) during the year before case or control selection increased the risk of CDI. In the multivariable analysis, time on dialysis was no longer significant, however having a deceased donor (OR, 2.98; 95% CI, 1.47-6.05; P = 0.002), increased length of posttransplant hospitalizations (OR, 1.04; 95% CI, 1.02-1.06; P < 0.001), and BPAR (OR, 5.82; 95% CI, 2.22-15.26; P < 0.001) or antibiotic therapy (OR, 2.88; 95% CI, 1.35-6.15]; P = 0.01) during the year before case or control selection remained significant risk factors for CDI.
Table 4 shows the types of antibiotics used during the year before case or control selection. Sixteen percent of controls were prescribed fluoroquinolones, 24% β-lactams, and 24% other antibiotics. Cases were significantly more likely to be treated with antibiotics in each category versus controls—36.8% versus 15.8% were prescribed fluoroquinolones, 51.5% versus 24.3% were prescribed β-lactams, and 55.9% versus 24.3% other antibiotics, respectively.
The association of CDI and posttransplant outcomes were assessed using multivariable Cox proportional hazards models (Table 5). Although having a CDI was not found to be significant risk factor for total graft failure, graft loss, or death with a functioning graft, CDI was associated with a higher risk for developing subsequent BPAR (hazard ratio, 2.18; 95% CI, 1.34-3.55; P = 0.002). Notably, patients who had CDI experienced a 27% cumulative probability of BPAR over 10 years of follow-up compared with just 18% for patients who did not have CDI (log rank P = 0.01). An analysis of any or net dose reductions in calcineurin inhibitors, mycophenolic acid, or corticosteroids over the 4 weeks after a CDI diagnosis showed no statistically significant differences in patients who eventually did or did not develop BPAR (Table S2, SDC,http://links.lww.com/TP/B554).
This study confirms that CDIs are an important complication in KTR, especially within the first 6 months posttransplant. The 4.8% cumulative incidence of CDI in our cohort was within the 1% to 8% previously reported in the literature.5-7 The early presentation of CDI posttransplant may be due to asymptomatic colonization with C. difficile before and during the transplant. Previous studies have shown that approximately 15% of patients in similar settings, such as long-term care facilities, can be asymptomatic carriers of C. difficile.13 This hypothesis is partially supported by our risk factor analysis, where a longer total length of hospitalization posttransplant was associated with an increased risk of CDI. Longer time on dialysis pretransplant, representing a longer time for potential colonization, was also a significant risk factor in our univariable analysis.
As expected, antibiotic therapy was also significantly associated with CDI. Although we did not include antibiotic type in our multivariable analysis due to limited statistical power, descriptive statistics show that case patients were generally prescribed more fluoroquinolones, β-lactams, as well as other antibiotics. These results align with the previous literature, where β-lactams, such as penicillins and cephalosporins, have traditionally been shown to be the highest-risk antibiotics,7,14 with fluoroquinolones more recently being implicated as high-risk antibiotics as well.14,15
We also found that having a deceased donor and an BPAR episode during the year before a CDI were significantly associated with a higher risk for developing CDI. These relationships may be due to lower levels of graft function in deceased donor kidney recipients and the need for more intensive immunosuppressive therapy to manage BPAR episodes. This further compromises the immune system, increasing the risk of CDI either directly by reducing patients' abilities to respond to C. difficile colonization, or indirectly by reducing their ability to respond to other opportunistic infections and morbidities, leading to an increased need for antibiotics.
Additionally, we found in our univariable analysis that patients with a higher mean BMI had a higher risk of developing a CDI. Bishara et al16 showed an association between obesity and CDI (OR, 1.20; 95% CI, 1.12-1.27; P < 0.01), possibly due to differences in the intestinal microbiota of patients with a higher BMI. Our univariable analysis also showed that gastrointestinal surgeries during the year before the end of follow-up is a significant risk factor for CDI, consistent with prior studies.17,18 Surprisingly, proton pump inhibitor usage was not found to be a significant risk factor in our analysis as it previously has been in the literature; however, this may be due to the low event rate in our population and/or our center's tendency to reduce the use of PPI in recent years.
Our study also explored CDI as a potential risk factor for various graft outcomes. Although having a CDI was not associated with a higher risk of graft failure or death, it was found to be associated with a higher risk of BPAR. This result is of particular significance because BPAR was also found to be a predictor of CDI, making it both a potential risk factor and outcome of CDI. If the time sequence of BPAR and CDI is not properly accounted for in the statistical model, the association seen in the cohort study may have been a result of reverse causality. However, both BPAR and CDI were handled as time-dependent covariates in the Cox proportional hazards model thus significantly reducing this possibility. The mechanism for CDI as risk factor for BPAR is unclear. Dose reductions in immunosuppressive therapy that may occur in response to CDI did not differ significantly among patient show did or did not eventually develop BPAR. Moreover, it has been shown that CDI may not significantly impact tacrolimus levels.19
Our study is one of the largest cohorts of KTR in which CDI has been studied. We used a combination of differing study design approaches to evaluate novel risk factors and the prognosis of patients with CDI over follow-up. The size of the study cohort and the assessment of outcomes after CDI distinguish this study from other similar reports.7,20 Limitations of our study include the relatively low numbers of CDI despite the large sample size (68 cases of 1816 patients), which limited the number of variables that could be explored in multivariable analyses. Despite this, our study had the highest number of CDI cases in the recent literature. Although our cohort was a representative sample of KTR at large kidney transplant center, it was a single-center study, which potentially limits the generalizability of the results. Finally, the diagnostic test for C. difficile was changed during the study period (from an enzyme-linked immunosorbent assay and a polymerase chain reaction test), which may have increased case ascertainment in the latter period of the cohort.
In conclusion, our study reports a 4.8% incidence of CDI in a large Canadian population of KTR. Independent predictors of CDI included having a deceased donor, lengthier total days of hospitalization posttransplant, BPAR episodes, and treatment with antibiotics. Our study also showed that CDI was an independent predictor of BPAR, something which has not been previously reported in the literature. These results call for careful management of antibiotics and immunosuppression in patients who are at risk for developing a CDI. Future research is required to confirm the findings of this study, reliably identify high-risk patients for whom monitoring or screening for asymptomatic C. difficile colonization may be beneficial, and to elucidate the clinical management strategies for the prevention and treatment of CDI in KTR.
The authors would like to thank the students of the Multi-Organ Transplant Student Research Training Program for collecting, entering, and auditing data for the Comprehensive Renal Transplant Research Information System (CoReTRIS) at the Toronto General Hospital, University Health Network.
1. Tonelli M, Wiebe N, Knoll G, et al. Systematic review: kidney transplantation compared with dialysis in clinically relevant outcomes. Am J Transplant
2. Binion DG. Strategies for management of Clostridium difficile
infection in immunosuppressed patients. Gastroenterol Hepatol (N Y)
3. Freeman J, Bauer MP, Baines SD, et al. The changing epidemiology of Clostridium difficile
infections. Clin Microbiol Rev
4. Burke KE, Lamont JT. Clostridium difficile
infection: a worldwide disease. Gut Liver
5. Altiparmak MR, Trablus S, Pamuk ON, et al. Diarrhoea following renal transplantation. Clin Transplant
6. West M, Pirenne J, Chavers B, et al. Clostridium difficile
colitis after kidney and kidney-pancreas transplantation. Clin Transplant
7. Neofytos D, Kobayashi K, Alonso CD, et al. Epidemiology, risk factors, and outcomes of Clostridium difficile
infection in kidney transplant recipients. Transpl Infect Dis
8. Reveles KR, Lee GC, Boyd NK, et al. The rise in Clostridium difficile
infection incidence among hospitalized adults in the United States: 2001–2010. Am J Infect Control
10. Pépin J, Valiquette L, Alary ME, et al. Clostridium difficile
-associated diarrhea in a region of Quebec from 1991 to 2003: a changing pattern of disease severity. CMAJ
11. Famure O, Phan NA-T, Kim SJ. Health information management for research and quality assurance: the Comprehensive Renal Transplant Research Information System. Healthc Manage Forum
12. Haas M, Sis B, Racusen LC, et al. Banff 2013 meeting report: inclusion of C4d-negative antibody-mediated rejection and antibody-associated arterial lesions. Am J Transplant
13. Ziakas PD, Zacharioudakis IM, Zervou FN, et al. Asymptomatic carriers of toxigenic C. difficile
in long-term care facilities: a meta-analysis of prevalence and risk factors. PLoS One
14. Owens RC Jr., Donskey CJ, Gaynes RP, et al. Antimicrobial‐associated risk factors for Clostridium difficile
infection. Clin Infect Dis
. 2008;46(Suppl 1):S19–S31.
15. McCusker ME, Harris AD, Perencevich E, et al. Fluoroquinolone use and Clostridium difficile
-associated diarrhea. Emerg Infect Dis
16. Bishara J, Farah R, Mograbi J, et al. Obesity as a risk factor for Clostridium difficile
infection. Clin Infect Dis
17. Shah SA, Tsapepas DS, Kubin CJ, et al. Risk factors associated with Clostridium difficile
infection after kidney and pancreas transplantation. Transpl Infect Dis
18. Thibault A, Miller MA, Gaese C. Risk factors for the development of Clostridium difficile
-associated diarrhea during a hospital outbreak. Infect Control Hosp Epidemiol
19. Bonatti HJ, Sadik KW, Krebs ED, et al. Clostridium difficile
-associated colitis post-transplant is not associated with elevation of tacrolimus concentrations. Surg Infect (Larchmt)
20. Kennedy C, Waldron C, Skally M, et al. The epidemiology of Clostridium difficile
infection in a national kidney transplant center. Clin Transplant
Supplemental Digital Content
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.