Incidence and Outcomes Associated With Clostridioides difficile Infection in Solid Organ Transplant Recipients

Key Points Question What are the incidence and outcomes associated with posttransplant Clostridioides difficile infection (CDI) in people who have undergone solid organ transplant? Findings In this cohort study including 10 724 patients, posttransplant CDI was associated with a 90-day mortality incidence of 17%. Late-onset CDI was associated with a greater risk of short- and long-term mortality compared with early-onset CDI. Meaning The findings of this large cohort of organ transplant recipients suggest a trend toward increasing risk of CDI over time.


Introduction
Patients who have undergone solid organ transplant (SOT) are at risk of Clostridioides difficile infection (CDI), which is associated with significant morbidity and mortality. [1][2][3] A wide range of the cumulative CDI incidence has been reported in single-center studies 4,5 ; however, SOT-related CDI incidence has been rarely estimated at the population level. 6 Most studies have investigated early-onset CDI occurring within 90 days following transplant. [4][5][6][7] However, the outcome of late-onset CDI has been rarely investigated. 4,5 Although early-onset infectious complications may affect allograft function, late-onset infectious diseases may be associated with allograft loss and death. 8,9 To our knowledge, post-CDI mortality has not been compared between patients who underwent SOT developing early-vs late-onset CDI.
SOT recipients are at risk of CDI complications. 10 The frequency of colectomy in patients with SOT who develop CDI is more than 3 times higher than the rate of post-CDI colectomy in the general population. 11 Complications of CDI are not limited to fulminant colitis. A recent study showed CDI considerably increases the risk of acute kidney injury (AKI) in patients without SOT. 12 SOT recipients are at risk of AKI following infectious complications due to continuous exposure to nephrotoxic drugs, allograft rejection, and comorbidities. 13 However, to our knowledge, the incidence and outcomes of AKI have not been investigated in patients with posttransplant CDI.
In this 15-year, population-based cohort study, we investigated the incidence and outcomes associated with posttransplant CDI among recipients of different organ allografts. Our findings may have implications for preventive strategies targeting SOT recipients who are at risk of CDI-associated complications and mortality.

Study Design
We undertook a population-based cohort study using Ontario administrative health care data held at ICES, a not-for-profit research institute in Ontario, Canada (eAppendix, eTable 1, and eTable 2 in the Supplement). 14 Use of the data without informed consent was authorized under section 45 of Ontario's Personal Health Information Protection Act and did not require research ethics board approval. Ontario has a universal, publicly funded, single-payer health insurance system that is administered by the provincial government. Data are reported herein in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline. 15 We conducted this population-based study in Ontario, Canada's most populous province (14.7 million in 2020). Approximately half of Canadian organ transplant procedures are performed in Ontario. 16 inflammatory bowel diseases, chronic obstructive pulmonary disease, chronic kidney disease, and the Deyo-Charlson Comorbidity Index (eTable 3 in the Supplement).
The exposure comprised SOT categorized according to allograft type (ie, kidney, liver, kidneypancreas, heart, lung, and multiorgan transplant). Owing to the small number of individuals in some subgroups, we created 3 transplant categories in some analyses: abdominal (ie, kidney, liver, and kidney-pancreas allografts), thoracic (ie, lung and heart allografts), and multiorgan transplant.
Multiorgan transplant recipients included patients who received at least 2 organ allografts during the same hospitalization other than those who received a kidney-pancreas transplant.

Study Outcomes
The primary outcome was hospital admission with a diagnosis of C difficile infection (International Classification of Diseases, 10th Revision, Canadian code A04.7), which has been reported to have high sensitivity (82.1%) and specificity (99.4%). 17 We defined early-onset CDI as C difficile infections occurring during the index hospitalization or within 90 days of discharge. Episodes of CDI that occurred beyond this point were considered as late-onset CDI. Patients were censored on the date of death, loss of insurance coverage, or the end of the study period (March 31, 2020), whichever occurred first.
The secondary outcomes included all-cause death, intensive care unit (ICU) admission, AKI requiring dialysis, and fulminant CDI comprising any of the following: toxic megacolon, ileus, perforation, and colectomy. We determined the risk of short-vs long-term death. We defined shortterm mortality as all-cause death occurring from the CDI hospitalization to 90 days postdischarge.
Death occurring after this point was considered long-term mortality. For the outcome of AKI requiring dialysis, patients receiving long-term dialysis following transplant were excluded.

Statistical Analysis
We determined baseline characteristics at the time of transplant as frequencies (percentages) for categorical variables and medians (IQRs) for continuous variables. We express incidence per 1000 person-years of observation (ie, patients with SOT and then recipients of different allografts). 18 To evaluate trends in CDI incidence, we estimated the incidence for each annual SOT cohort using 1-and 3-year follow-up periods. We used the Cochran-Armitage test to compare trends in CDI incidence.
We estimated the cumulative probability of CDI (1-Kaplan-Meier) and 95% CI) overall and for each SOT type, using the Kaplan-Meier estimator, and compared the probability among SOT types, using the log rank test. We used Cox proportional hazards regression to identify patient characteristics independently associated with posttransplant CDI.
Next, we restricted the cohort to SOT recipients who developed CDI and determined the frequency of fulminant illness, AKI requiring acute dialysis, ICU admission, and mortality. Subgroup comparisons were conducted with χ 2 tests.
We used logistic regression to quantify the association between short-term mortality and age, sex, Deyo-Charlson Comorbidity Index, SOT type (ie, thoracic vs abdominal transplant), early-vs lateonset CDI, fulminant CDI, ICU admission, and AKI requiring acute dialysis. We then used a Cox proportional hazards regression model to quantify the association between the same variables and long-term mortality, with patients censored on transplant of any organ type after the index hospitalization and the end of the study period (March 31, 2020). Statistical tests were 2-sided with the level of significance set at α = .05. All analyses were performed at ICES using SAS, version 9.4 (SAS Institute Inc).

Patient Characteristics Associated With CDI
We used a series of Cox proportional hazards regression models to investigate the association between potential baseline contributing factors and posttransplant CDI (eAppendix; eTable 6 in the Supplement). We found an association between age at the time of transplant and posttransplant CDI (hazard ratio [HR], 1.14 for each 10-year increase in age; 95% CI, 1.07-1.22; P < .001). Patients who received an SOT other than kidney were at a significantly higher risk of CDI compared with those who received a kidney transplant. However, in kidney allograft recipients, delayed graft function was associated with a greater risk of CDI (HR, 1.6; 95% CI, 1.3-2.0). Comorbidities, such as diabetes (HR, 1.6; 95% CI, 1.4-1.9), inflammatory bowel disease (HR, 1.7; 95% CI, 1.1-2.5), chronic obstructive pulmonary disease (HR, 1.5; 95% CI, 1.2-1.8), and cancer (HR, 1.4; 95% CI, 1.2-1.6), were significantly associated with a higher risk of CDI.

JAMA Network Open | Infectious Diseases
Incidence     associated with an increased risk of longer-term mortality following CDI ( Table 3). As with short-term mortality, late-onset CDI was associated with a relatively greater risk of longer-term death following CDI (aHR, 2.49; 95% CI, 1.78-3.49).

Discussion
In the absence of population-based studies, estimation of posttransplant CDI incidence has been limited to single-center studies providing wide variation in cumulative CDI incidence. 2,19 Our study, which is, to our knowledge, the largest cohort of SOT recipients including 10 724 patients and 61 987 person-years of follow-up, estimated a CDI incidence of 11.7 (95% CI, 10.9-12.6) per 1000 personyears. We observed a trend toward increasing risk of posttransplant CDI over time. Recent surveillance data showed an increasing CDI incidence trend in the US general population in 2011 followed by a relative plateau in 2012 and thereafter. 20 The increasing CDI incidence in 2011-2012 could have been due to the fact that 52% of microbiology laboratories used nucleic acid amplification  factors that occur before AKI, such as decreased organ perfusion, 28 protein-losing enteropathy leading to reduced oncotic pressure, 29 and immunoglobulin A nephropathy. 30 We noted that dialysis-requiring AKI is an independent risk factor for short-term mortality. The association between AKI and long-term mortality remained significant in Cox proportional hazards regression analysis.
Experimental data noted an association between AKI and tissue injury in different organs, such as inflammatory and functional changes in the brain, pulmonary vascular integrity loss, cardiac tissue apoptosis, and severe fibrotic changes in the kidneys. 31,32 Acute kidney injury has been reported to increase the long-term risk of cardiovascular events. 33 Thus, the longstanding association between AKI and mortality appears to be biologically plausible.
Recipients of SOT with CDI are more likely to be admitted to the ICU compared with patients who have not received an SOT. 34 Patients who were admitted to the ICU were at greater risk of shortterm mortality. However, this risk did not remain significant in patients who survived 90 days following the CDI hospitalization.
Timing of the onset of CDI was an important factor in CDI outcome. Most studies only investigated early-onset CDI and their findings were likely affected by incomplete follow-up. 6 Prospective studies are required to estimate postcolectomy mortality.

JAMA Network Open | Infectious Diseases
Incidence

Limitations
This study has limitations. We did not have access to information on immunosuppressive regimens and antibiotic therapies at an individual level. The present study was conducted with a primary focus on CDI outcomes rather than risk factors. The outcomes of interest did not include mild CDI episodes that were managed in outpatient settings. This cohort was restricted to transplant centers across Ontario, which may affect the generalizability of our findings. However, within the accrual window of this cohort, 40% to 50% of SOT operations in Canada have been done in Ontario. 16

Conclusions
In this study, we observed increasing CDI trends in annual cohorts of SOT recipients. Posttransplant CDI was associated with considerable mortality. Although CDI was an early-onset disease in non-kidney allograft recipients, patients who underwent kidney transplant typically experienced late-onset CDI. Late-vs early-onset CDI was associated with a greater risk of death. Acute kidney injury was also associated with an increased risk of short-and long-term mortality following CDI.
Acute kidney injury preventive measures, such as adequate fluid repletion, avoidance of hypotension in critically ill SOT recipients, readjustment of nephrotoxic medications based on drug levels, and close kidney function monitoring, should be considered in the management of CDI. Clostridioides difficile infection is potentially preventable through risk reduction strategies. Further studies are required to reduce CDI incidence and related complications in SOT recipients.