Abstract
Kidney transplantation (KT) from deceased donors with hepatitis C virus (HCV) into HCV-negative recipients has become more common. However, the risk of complications such as BK polyomavirus (BKPyV) remains unknown. We assembled a retrospective cohort at four centers. We matched recipients of HCV-viremic kidneys to highly-similar recipients of HCV-aviremic kidneys on established risk factors for BKPyV. To limit bias, matches were within the same center. The primary outcome was BKPyV viremia ≥1000 copies/mL or biopsy-proven BKPyV nephropathy; a secondary outcome was BKPyV viremia ≥10,000 copies/mL or nephropathy. Outcomes were analyzed using weighted and stratified Cox regression. The median days to peak BKPyV viremia level was 119 (IQR 87–182). HCV-viremic KT was not associated with increased risk of the primary BKPyV outcome (HR 1.26, p=0.22), but was significantly associated with the secondary outcome of BKPyV ≥10,000 copies/mL (HR 1.69, p=0.03). One-year eGFR was similar between the matched groups. Only one HCV-viremic kidney recipient had primary graft loss. In summary, HCV-viremic KT was not significantly associated with the primary outcome of BKPyV viremia, but the data suggested that donor HCV might elevate the risk of more severe BKPyV viremia ≥10,000 copies/mL. Nonetheless, one-year graft function for HCV-viremic recipients was reassuring.
Introduction
Until the year 2015, most organs from potential deceased donors with hepatitis C virus (HCV) infection were discarded or never procured because historic antiviral medications had major side effects, including allograft rejection. The advent of direct-acting antiviral agents (DAA) has markedly transformed the HCV treatment landscape with virologic cure rates approaching 99%,(1–3) such that transplanting HCV infected kidneys into HCV naïve recipients is now commonplace.(4–8) An increasing number of transplant programs are accepting these kidneys,(9) either in the setting of clinical trials (5–8) or in real-world settings.(4, 10) Moreover, a study of decision-making by HCV-naïve candidates suggested that most were willing to accept HCV-infected kidneys under at least some circumstances, such as a kidney from a young donor.(11, 12) A single-center trial and national cohort studies using OPTN registry data have demonstrated that 1-year allograft function after transplantation is similar between recipients of HCV-infected kidneys and recipients of kidneys from matched donors without HCV.(4, 5, 9, 13) However, important concerns have emerged suggesting that donor-derived HCV viremia might be associated with unexpected immunological complications such as BK polyomavirus (BKPyV) infection.(4)
Human BKPyV is frequently found in the urine of healthy individuals.(14, 15) BKPyV resides in the urothelium and approximately 10–12% of kidney transplant recipients develop BKPyV viremia, most often during the first year after transplant.(16) Severe viremia and BK nephropathy are associated with increased risk of kidney graft loss.(17–19)
Given the high prevalence of BKPyV infection in kidney recipients, guidelines recommend routine surveillance for BKPyV in the first post-transplant year. (20, 21) No therapy has been demonstrated to effectively treat BKPyV viremia, but centers typically perform routine surveillance of blood and/or urine and reduce immunosuppression when BKPyV is detected, which can often resolve the infection.(22) Guidelines differ regarding clinically-important levels of BKPyV viremia. The Banff Working Group on Polyomavirus Nephropathy suggested more than 10,000 copies in the plasma should be treated as presumptive nephropathy.(21) Likewise the AST Infectious Diseases Community of Practice characterizes BK nephropathy when viremia is >10,000 copies/mL, and recommends stepwise immunosuppression reduction when viremia is >1,000 copies/mL sustained for 3-weeks, or the viremia level is >10,000 copies/mL.(20) Unfortunately, allograft fibrosis can occur even when BKPyV is no longer detectable in the blood or urine.(23)
Identifying new risk factors for development of severe BKPyV is a key component of preventing and addressing infection early. A single-center study in HCV-uninfected recipients of kidneys from HCV-viremic donors, with delayed administration of DAA therapy post-transplant, showed that 34% of newly-infected recipients developed BKPyV viremia in the first few months.(4) Most of these recipients developed severe BKPyV viremia (median and interquartile [IQR] range was 37,646 [5,959–255,475] copies/mL).(4) In addition, >70% of these BKPyV viremia cases developed before HCV treatment initiation.(4) While these data are concerning, the study lacked matched comparators and therefore, it is unknown whether this high cumulative incidence of BKPyV viremia is attributable to HCV or other features of the center practice or population.
The aim of our study was to determine the association between deceased donor HCV-infection and risk of BKPyV viremia in HCV-naive kidney transplant recipients across multiple centers. We hypothesized that HCV-viremic kidney recipients would be at elevated risk of BKPyV viremia versus precisely-matched recipients of HCV-aviremic kidneys, with matches formed within the same center.
Materials and Methods
Consortium to study Outcomes After Transplanting HCV-viremic kidneys into HCV-negative Recipients (COAUTHOR)
We assembled a retrospective cohort of deceased donor kidney transplant recipients who underwent transplantation from 4/1/2015 to 3/31/2019 at 1) Massachusetts General Hospital (Massachusetts), 2) the University of Pennsylvania (Pennsylvania), 3) University of Tennessee Health Science Center, Methodist University Hospital (Tennessee) or 4) Vanderbilt University Medical Center (Tennessee). These centers had the advantages of geographic diversity as well as practice variation. Specifically, Massachusetts General Hospital and the University of Pennsylvania performed HCV-viremic KT within a clinical trial setting and started DAAs in the first post-transplant week, while the other two centers did the transplants outside a trial with delayed initiation of DAAs. At all centers, recipients of HCV-infected kidneys received similar induction and maintenance immunosuppression as recipients of HCV-uninfected kidneys. Immunosuppression was not modified if recipients developed HCV-antibody.
We included recipients who were ≥18 years at listing and received a deceased donor transplant. We excluded recipients if they were listed for a multi-organ transplant, were seropositive or Nucleic Amplification Testing (NAT) positive for HIV, or had pre-transplant HBV or HCV NAT or seropositivity.
This study was approved by the Institutional Review Boards at each center and at the University of Pennsylvania (IRB #833840), which was the data coordinating center.(24)
Data sources
We used the baseline data on recipients and deceased donors that are reported to the OPTN to match each recipient of an HCV-viremic kidney transplant to one or more recipients of HCV-aviremic transplants. Investigators then extracted additional data from recipients’ charts regarding ureteral stent placement, DAA therapy, serum BKPyV viremia levels, clinical complications, deaths, and allograft function for 1-year following transplantation.
Outcomes
The primary outcome was BKPyV viremia level ≥1,000 copies/mL or biopsy-proven BK nephropathy, as per guidelines from the Infectious Disease Community of Practice of the AST.(20) The secondary outcome was BKPyV viremia ≥10,000 copies/mL or biopsy-proven BK nephropathy. If the center BKPyV screening protocol included urine screening first and the recipient had negative urine results, we categorized that patient as having no BKPyV viremia. The BKPyV screening protocols and management of immunosuppression are shown in Supplementary Tables 1 and 2.
Main exposure
The main exposure variable was the HCV-viremia status of the deceased donor. The criterion of HCV-viremia positivity was determined by NAT performed during usual care by the organ procurement organization, regardless of HCV antibody status. HCV-aviremic matched donors had negative testing for HCV-NAT and HCV-antibody.
Matching Process
We used optimal matching to generate highly-similar pairs of recipients of HCV-viremic kidneys (the “focal patients”) and recipients of non-viremic kidneys (“comparators”) who were matched on characteristics that studies have reported as plausible risk factors for BKPyV infection, as well as selected additional characteristics, and are known at the time of transplantation.(25–27)
We matched recipients within the same center. This design minimized bias, because the matched focal patients and comparators were subject to the same center-based clinical protocols, including detection of BKPyV and treatment. We note that at Vanderbilt, only two nephrologists routinely performed BKPyV screening post-transplantation and these two nephrologists cared for all recipients of HCV-viremic kidneys. Therefore, prior to matching, we excluded from the pool of potential comparators all kidney transplant recipients assigned to other nephrologists; we only matched HCV-viremic kidney recipients at Vanderbilt to HCV-aviremic recipients cared for by the same two nephrologists.
Our matching strategy allowed each focal patient to be matched to a variable number of comparators (minimum 1 and maximum of 5 comparators). We matched exactly on transplant center, induction agent at transplant, and CMV serostatus of the donor and recipient (because a subsequent study will examine CMV outcomes.)
We then estimated the propensity score for treatment assignment (transplantation with an HCV-viremic donor) using all the covariates used in the matching algorithm (Supplementary Appendix 1). A Mahalanobis distance matrix was built using the propensity score, recipient age, calculated panel reactive antibody (cPRA) at allocation, and the allocation Kidney Donor Profile Index (KDPI). Penalties were subsequently applied on covariates to ensure that the matching algorithm first prioritized allograft cold ischemia time (CIT), PRA, recipient age, the KDPI, zero Human Leukocyte Antigen (HLA) mismatch transplants, and recipient race. The R package “designmatch” was used to construct the distance matrix and apply penalties.(28)
We used a standardized difference of <0.10 as the target for excellent covariate balance and <0.20 as the conventional cut-off for acceptable balance.(29–31)
Statistical analysis
We computed the rate of BKPyV viremia or biopsy-proven BK nephropathy as the number of events divided by the amount of person-time; rates were expressed as events per 100 patient-years (PY) of follow-up. Each patient was followed from the date of transplant (time 0) until the earliest of transplant failure, death or 365 days post-transplant. Since ureteral stent placement is a risk factor for BKPyV viremia, we adjusted for ureteral stent placement.(16, 20, 25)
Time to BKPyV viremia event was modeled using Cox regression, stratified by matched set and weighted as described above.
Secondary and exploratory analyses
We also conducted analyses using BKPyV viremia ≥10,000 copies/mL or biopsy-proven BK nephropathy as the outcome. Second, we examined the effect of time until DAA initiation; only recipients of HCV-viremic donor kidneys were included. Based on clinical judgement, we classified patients as receiving DAA therapy within or later than 3 weeks, because we estimated that it would typically take centers 3-weeks to detect recipient genotype and obtain insurance approval for DAA therapy. Follow-up stratified on the day DAA therapy was initiated, with follow-up time measured from that date forward. The covariate of interest was time-to-treatment, with the outcome again being BKPyV viremia incidence. Third, since delayed graft function (DGF, i.e. dialysis in the first week post-transplant) was previously reported to be a risk factor for BKPyV viremia, we performed an exploratory analysis with adjustment for DGF.(32) We did not match on DGF or adjust for DGF in primary analyses, because DGF (or treatment for DGF) could be in the causal pathway between donor HCV viremia and the outcomes. Some studies have identified male recipient sex as a possible risk factor for BK nephropathy; (27, 33) we performed a secondary analysis adjusting for male sex post-match.
We reported post-transplant eGFR using the Chronic Kidney Disease Epidemiology Collaboration Equation.(34, 35) All analyses were performed in either R statistical package (R Core Team 2019, Vienna, Austria) or SAS (v9.4; Cary, NC). We used the R package “cobalt” to visually assess covariate balance.(36)
Results
Baseline recipient, donor, and transplantation characteristics
Figure 1 shows cohort generation. In the overall cohort of 1,137 kidney recipients (prior to matching), the mean (SD) age was 52 (+/−13) years, 55% were male, and 49% were black. A higher percentage of recipients of HCV-viremic kidneys were male, black, and had diabetes mellitus or hypertension as cause of kidney failure versus recipients of HCV-aviremic kidneys. HCV-viremic kidney recipients also had lower maximum PRA versus recipients of HCV-aviremic kidneys. (Supplementary Table 3 and Supplementary Figure 1 show further data about comparators at Center 4 [Vanderbilt]).
Figure 1. Flow chart of selection of the patients*.
* We excluded recipients who were seropositive or NAT positive for HIV, HBV, and HCV.
At transplant Center 4, all recipients of HCV-viremic kidneys were followed by two nephrologists who screened recipients regularly for BK nephropathy. To limit bias, and only compare participants who underwent routine BK screening, we excluded 394 recipients of HCV-aviremic kidneys prior to any matching because these recipients were followed by nephrologists who did not perform routine BK screening.
Supplementary Figure 2 shows the number of matched comparators for each focal patient. In the matched cohort, recipients of HCV-viremic and HCV-aviremic donor kidneys had similar baseline characteristics related to risk of BKPyV viremia (Table 1), as evidenced by the small standardized differences, as well as the similar distributions (Supplementary Figure 3). Table 1 shows that post-match, the mean (SD) age was 54 vs. 53 years, the percentage of Black patients was 61 vs. 67%, mean maximum PRA was 14.8 vs. 16.6% and induction therapy was identical among focal patients vs. comparators, respectively.
Table 1.
Recipient, donor, and allograft characteristics pre- and post-match*
| Variables | HCV NAT D + / R – (N = 148) | HCV NAT D - / R – (N = 989) | HCV NAT D + / R – (N = 146) | HCV NAT D - / R – (N = 146)* | Standardized difference** |
|---|---|---|---|---|---|
| Pre-Match | Pre-Match | Post-Match | Post-Match | ||
| Recipient Variables *** (Matched) | |||||
| Age in years (Mean, SD) | 54.2 (10.4) | 51.9 (13.3) | 54.1 (10.4) | 53 (10.3) | 0.0863 |
| Sex Female (%) | 51 (34.5%) | 465 (47%) | 49 (33.6%) | 64 (44%) | −0.1017 |
| Black Race (%) | 89 (60.1%) | 471 (47.6%) | 89 (60.9%) | 98 (67%) | −0.0606 |
| Maximum PRA (Mean, SD) | 15.3 (28) | 30.1 (41.1) | 14.8 (27.8) | 16.6 (31.6) | −0.0521 |
| Zero HLA Mismatch (%) | 2 (1.4%) | 54 (5.5%) | 2 (1.4%) | 7 (4.8%) | 0.0356 |
| Induction Agent * | |||||
| Rabbit anti-thymocyte globulin (%) | 128 (86.5%) | 820 (82.9%) | 128 (87.7%) | 128 (87.7%) | 0 |
| Alemtuzumab (%) | 16 (10.8%) | 92 (9.3%) | 16 (10.9%) | 16 (10.9%) | 0 |
| Basiliximab (%) | 4 (2.7%) | 47 (4.8%) | 2 (1.4%) | 2 (1.4%) | 0 |
| None (%) | 0 (0%) | 30 (3%) | 0 (0%) | 0 (0%) | 0 |
| Cause of ESKD | |||||
| Diabetes (%) | 63 (42.6%) | 257 (26%) | 63 (43.2%) | 50 (34.2%) | 0.0871 |
| Hypertension (%) | 46 (31.1%) | 261 (26.4%) | 46 (31.5%) | 48 (32.8%) | −0.0151 |
| Glomerular Disease (%) | 13 (8.8%) | 154 (15.5%) | 13 (8.9%) | 14 (9.6%) | −0.0059 |
| Cystic Disease (%) | 10 (6.7%) | 71 (7.2%) | 9 (6.2%) | 11 (7.5%) | −0.0103 |
| Other / Missing (%) | 16 (10.8%) | 246 (24.9%) | 15 (10.3%) | 23 (15.6%) | −0.0558 |
| History of Prior Transplant (%) | 9 (6.1%) | 156 (15.8%) | 9 (6.2%) | 14 (9.6%) | −0.0305 |
| Donor and allograft variables (Matched) | |||||
| Allocation KDPI % | 46 (16) | 44 (26) | 47 (16) | 46 (21) | 0.0104 |
| CIT in hours (Mean, SD) | 16.7 (5.9) | 14.9 (6.1) | 16.8 (5.9) | 16.2 (5.6) | 0.1053 |
| Donor-Recipient CMV Serostatus | |||||
| D-/R- (%) | 15 (10.1%) | 156 (15.8%) | 14 (9.6%) | 14 (9.6%) | 0 |
| D-/R+ (%) | 28 (18.9%) | 188 (19%) | 28 (19.2%) | 28 (19.2%) | 0 |
| D+/R+ (%) | 67 (45.3%) | 359 (36.3%) | 66 (45.2%) | 66 (45.2%) | 0 |
| D+/R- (%) | 38 (25.7%) | 286 (28.9%) | 38 (26%) | 38 (26%) | 0 |
Abbreviations: CMV: Cytomegalovirus; D: Donor; R: Recipient; HLA: Human leukocyte antigen; CIT: Cold ischemia time; HCV: Hepatitis C; NAT: Nucleic Acid Amplification Testing; ESKD: End Stage Kidney Disease; PRA: Panel reactive antibody; SD: Standard deviation; Std diff: Standardized difference
Because each recipient of an HCV-viremic kidney was matched to a varying ratio of 1:5 comparators, the results presented here are the weighted averages after accounting for variation in matched stratum size
A standardized (mean) difference is a measure of distance between two group means in terms of a variables and is a conventional metric used to evaluate the quality of a match. A standardized mean difference of < 0.2 is considered a sign of acceptable covariate balance, while <0.1 is considered excellent balance.
Exact matching on center, induction therapy, and donor/recipient serostatus; other matching methods used to create groups similar in terms of BK risk factors – recipient age, sex, race, cPRA, as well as KDPI, CIT, HLA mismatch.
Transplant center variation in testing for BKPyV viremia
Table 2 shows considerable variation between the four centers in the testing performed for BKPyV. For instance, at the University of Tennessee Health Science Center the primary method for BKPyV screening was monthly urine testing, and confirmatory testing was performed using serum testing if BKPyV was detected in urine. However, we detected no substantial differences between the rates of BKPyV testing within each center for recipients of HCV-viremic kidneys versus comparator recipients of HCV-aviremic kidneys.
Table 2.
Recipient BKPyV Viremia by donor HCV NAT status
| Donor HCV NAT+ | Donor HCV NAT - | |
|---|---|---|
| Number of patients in the matched cohort | 146 | 453 |
| Center 1 | 7 (4.8%) | 25 (5.5%) |
| Center 2 | 48 (32.9%) | 212 (46.8%) |
| Center 3 | 74 (50.7%) | 180 (39.7%) |
| Center 4 | 17 (11.6%) | 36 (8%) |
| Median number of times BKPyV PCR was checked in the 1 st year | 4 [0 – 8] | 4 [2 – 7] |
| Center 1 | 7.5 [5 – 9] | 7 [6 – 9] |
| Center 2 | 4.5 [4 – 6] | 5 [4 – 7] |
| Center 3 | .5 [0 – 10] | 1 [0 – 6] |
| Center 4 | 8 [7 – 9] | 5 [5 – 7] |
| Median number of times BKPyV PCR was checked in urine | 1.5 [0 – 12] | 0 [0 −8] |
| Center 1 | 0 | 0 |
| Center 2 | 0 | 0 |
| Center 3 | 12 [7 – 16] | 11 [7 – 13] |
| Center 4 | 0 | 0 |
| Number of patients with max > 1,000 copies/mL | 33 (22.6%) | 70 (15.4%) |
| Center 1 | 1 (0.7%) | 2 (0.4%) |
| Center 2 | 5 (3.4%) | 25 (5.5%) |
| Center 3 | 22 (15.1%) | 37 (8.2%) |
| Center 4 | 5 (3.4%) | 6 (1.3%) |
| Number of patients with max > 10,000 copies/mL | 25 (17%) | 41(9.1%) |
| Center 1 | 1 (0.7%) | 2 (0.4%) |
| Center 2 | 3 (2%) | 15 (3.3%) |
| Center 3 | 18 (12.3%) | 22 (4.9%) |
| Center 4 | 3 (2%) | 2 (0.4%) |
| Max BKPyV Viremia level if copies are > 1,000 copies/mL | 46,510 [11,147 – 102,345] | 15,088 [3,799 – 104,325] |
| Center 1 | 46,510* | 50,750* |
| Center 2 | 12,025 [6,825 – 23,725] | 13,000 [7,800 – 80,925] |
| Center 3 | 59,343 [11,542 – 303,823] | 20,980 [3,200 – 110,000] |
| Center 4 | 20,653 [7,330 – 70,691] | 6,146 [3,075 – 397,066] |
Abbreviations: HCV: Hepatitis C; NAT: Nucleic Acid Amplification Testing; PCR: Polymerase chain reaction
Since only one recipient of a HCV NAT+ kidney and two recipients of HCV NAT- kidneys received developed the primary and secondary outcomes, we did not provide an inter-quartile range.
Risk of BKPyV Viremia
Rates of BKPyV viremia and the results of Cox regression are presented in Table 3. Overall, there were 103 BKPyV viremia events (i.e., instances wherein the measurements revealed ≥1,000 copies/mL of BKPyV viremia or BK nephropathy) in the matched cohort; total follow-up was approximately 502 patient-years. Supplementary Table 4 shows that 3 recipients of HCV-viremic kidneys had biopsy-proven BK nephropathy and one recipient of a HCV-aviremic donor kidney had biopsy-proven BK nephropathy. In all instances, recipients of both HCV-viremic and HCV-aviremic kidneys had a BKPyV viremia level ≥10,000 copies/mL before biopsy showed BK nephropathy. After weighting, the BKPyV viremia rate for HCV-viremic kidney recipients (27.1 per 100 PY) was not markedly different from the comparator recipients of HCV-aviremic kidneys rate (18.6 per 100 PY). From the Cox model, the hazard ratio (HR) for HCV-viremic donor (versus comparator) was 1.26 (95% Confidence Interval [CI]: 0.87–1.84; p=0.22). This HR can be interpreted as the ratio of BKPyV viremia incidence rates (i.e., rate for HCV-viremic organ divided by comparator rate) adjusted for ureteral stent placement.
Table 3.
Incidence of BKPyV Viremia
| Group | BKPyV Viremia events | Patient-years of follow-up | Rate per 100 PY | Weighted rate Per 100 PY | Hazard Ratio (95% CI) |
|---|---|---|---|---|---|
| BKPyV >1,000 copies/mL (Primary outcome) | |||||
| Donor HCV NAT- | 70 | 380.3 | 18.4 | 18.6 | 1 (reference) |
| Donor HCV NAT+ | 33 | 121.6 | 27.1 | 27.1 | 1.26 (0.87, 1.84) p=0.22 |
| Total | 103 | 501.9 | 20.5 | -- | N/A |
| BKPyV >10,000 copies/mL | |||||
| Group | BKPyV Viremia events | Patient-years of follow-up | Rate per 100 PY | Weighted rate Per 100 PY | Hazard Ratio (95% CI) |
| Donor HCV NAT- | 41 | 400.5 | 10.2 | 10.4 | 1 (reference) |
| Donor HCV NAT+ | 25 | 126.5 | 19.8 | 19.8 | 1.69 (1.04, 2.74) p=0.03 |
| Total | 66 | 527.1 | 12.5 | -- | N/A |
Abbreviations: HCV: Hepatitis C; NAT: Nucleic Acid Amplification Testing; PCR: Polymerase chain reaction; PY: patient-years
The median peak BKPyV viremia level (among patients ever measuring >1,000 copies) was 20,980 (IQR: 6,350–104,325). Median time to peak BKPyV viremia load was 119 (IQR: 87–182) days. A plot of the crude BKPyV Viremia rate function is included in Supplementary Figure 4. Based on this plot, it appears that BKPyV viremia incidence increased sharply up to approximately day 90 post-transplant, then decreases steadily thereafter.
Effect of timing of DAA initiation on BKPyV Viremia
Supplementary Table 5 shows DAA regimens at each center. Median time to DAA initiation was 46 (IQR: 3–76) days. Site-specific medians were 71 (IQR: 62–86) days for the University of Tennessee Health Science Center, 0 (IQR: 0–0) days for Massachusetts General Hospital, 3 (IQR: 3–4) days for the University of Pennsylvania, and 28 (IQR: 22–37) days for Vanderbilt University. Time to treatment did not show a significant association with subsequent BKPyV viremia incidence, with HR=1.15 for >3 weeks relative to ≤3 weeks (p=0.90).
Sensitivity and secondary analyses
First, as a sensitivity analysis, we fitted a Cox model which also adjusted for DGF; results were almost identical to those presented in Table 3, with HR=1.25 (CI: 0.86–1.81) for HCV-viremic donor (p=0.25). Results were also very similar when we changed the outcome to any detectable BKPyV viremia; in this case, HR=1.27 (CI: 0.92, 1.77) with p=0.15.
Second, we evaluated the incidence of BKPyV viremia when the peak BKPyV viremia load was greater than 10,000 copies/mL (Table 3). Twenty-five recipients of HCV-viremic kidneys had a BKPyV level ≥10,000 copies/mL at any time during the year. Weighted rates were 19.8 and 10.4 per 100 PY for patients with HCV-viremic and HCV–aviremic donor kidneys, respectively. Cox regression yielded HR=1.69 (CI: 1.04–2.74) for HCV-viremic donors (p=0.03). Among those 25 recipients of HCV-viremic kidneys with BKPyV ever ≥10,000 copies/ml, by the end of the 12-month follow-up period, 10 recipients had a BKPyV level ≥1000 copies/mL, among whom 3 recipients still had a BKPyV level ≥10,000 copies/mL.
Third, we compared patients with HCV-viremic versus HCV–aviremic donor kidneys with respect to graft function and other one-year outcomes. As shown in Supplementary Table 6, 12.3% of recipients of an HCV-viremic organ experienced DGF, compared with 28.4% for HCV-aviremic kidney recipients. This contrast was significant (p=0.001) when we fitted a conditional logistic regression model for DGF. As shown in Figure 2, the 1-year eGFR was slightly higher for recipients of HCV-viremic kidneys compared to HCV-aviremic kidneys (65.8 vs. 60.2 mL/min/1.73 m2, p=0.005), a difference that is not clinically meaningful. The proportions of patients who experienced graft failure were 7.1% and <1% for HCV-aviremic and HCV–viremic kidney recipients, respectively; only one recipient of an HCV-viremic kidney had graft failure. There were no deaths among recipients of HCV-viremic organs, while two percent of HCV-aviremic kidney recipients died (Supplementary Table 6).
Figure 2.
Comparison of 1-year eGFR for recipients of HCV-viremic and HCV-aviremic deceased donor kidney transplants
Finally, ureteral stent placement was not associated a statistically significant increase in the risk of BKPyV viremia (HR=1.10; p=0.69) (Supplementary Table 7).
Discussion
In the current era of DAA therapy, many transplant centers have taken interest in accepting kidneys from donors with HCV-viremia. Early studies reported excellent short-term allograft function among recipients of HCV-viremic kidney transplants, but case series from some centers that noted a high prevalence of BKPyV viremia raised concerns that donor-derived HCV might increase the risk of opportunistic viral infections. In this matched cohort, we did not detect a statistically significant difference between rates of BKPyV viremia >1000 copies/mL among recipients of HCV-viremic and HCV-aviremic kidneys. We did find a higher risk of BKPyV >10,000 copies/mL associated with HCV-aviremic kidneys. These findings suggest the possibility that once established, BK viral infection is more difficult to control in the setting of donor-derived HCV infection. However, these findings must be taken in context. Very similar allograft function between groups and reassuring graft survival rates for HCV-viremic kidneys suggest that the overall risk profile related to donor-derived HCV is reasonable.
BKPyV and HCV infection might involve shared immunological pathways. Unlike viruses such as CMV and Epstein-Barr virus (EBV), the BKPyV serological status of donor and recipient is not used to stratify the risk profile of recipients for BKPyV viremia after transplant. Previous studies have shown that BKPyV IgG seropositive status did not protect against viremia and antibody response was not associated with clearance of the virus.(37) Cellular adaptive immunity is essential for BKPyV control and resolution of BK nephropathy.(38, 39) CD4+ and CD8+ T cells both play a role in BKPyV viremia clearance. In transplant recipients, a strong CD8 response is associated with lower BKPyV titers in blood and urine, whereas weak responses correlated with high BKPyV titers and viral persistence.(40, 41) Development of BKPyV specific cellular response also correlated with resolution of BK nephropathy.(42) HCV replication may create the milieu for secondary viral infections by suppressing host immune responses.(43, 44) HCV core protein has been associated with impairment of effector functions and survival potential of CD8+ T cells in-vitro.(45) HCV core protein reduces the activity and targets lysis-associated functions of CD8+ HCV-specific T-cells, but also non-HCV specific T-cells.(45) This may contribute to observed reductions in immunity to other viral coinfections such as BKPyV and CMV in HCV-infected individuals.(45) Moreover, it is not known whether this dysfunction caused by HCV core protein or its effect could persist after treatment and achievement of sustained virologic response.(45) Based on these potential pathophysiological effects of HCV on BKPyV viremia and a previous single-center study,(4) we hypothesized that HCV-viremic kidney transplant recipients would be at an elevated risk of BKPyV viremia versus recipients of HCV-aviremic kidneys.
After precisely-matching for established BKPyV viremia risk factors, we did not detect a statistically significant difference between the rates of BKPyV viremia >1000 copies/mL among recipients of HCV-viremic and HCV-aviremic kidneys in our adequately powered study. However, weighted rates of severe BKPyV viremia (>10,000 copies/mL) were higher among recipients of HCV-viremic (19.8 per 100 patient-years) versus recipients received HCV-aviremic kidneys (10.4 per 100 patient-years), with HR=1.69 (p=0.03). Therefore, it is possible that HCV infection does not lead to a higher risk of BKPyV viremia but when this infection manifests, the recipient’s immune system is less able to control replication.
The timing of DAA initiation may play an important role in the development of potential complications such as BKPyV viremia recipients of HCV-viremic kidneys. Sise et al. published their center’s experience where DAAs were initiated before transplantation and reported a lower cumulative incidence of BKPyV viremia than in another center with late DAA initiation (13% vs. 34%).(8) In the present study, we did not detect a difference in the incidence of BKPyV viremia within the first year between patients who were treated with DAA early (<3 weeks after transplant) versus after 3-weeks of transplant. This study did not address whether there is any difference in late BKPyV viremia (after 1 year of transplantation) in recipients of HCV-viremic and HCV-aviremic kidneys; however; based on prior data, we anticipate that majority of BKPyV viremia will occur within 1-year of KT.(46)
This matched cohort study advances the field by providing detailed data from four centers and reveals outcomes for recipients of HCV-viremic kidneys and matched comparators, which are lacking in almost all prior studies involving HCV-viremic transplantation. We matched HCV-viremic and HCV-aviremic transplant recipients within each center on risk factors for BKPyV viremia and adjusted for stent placement. By matching focal recipients to comparators within each center, we were also able to account for variation in clinical practice surrounding immunosuppression and BKPyV screening practices. All recipients in the matched cohort experienced regular BKPyV screening, so it is very unlikely that any clinically important BKPyV viremia went undetected.
We draw attention to the substantial variation in the cumulative prevalence of BKPyV viremia between centers, which was not anticipated but could be explained in several ways. First, variation in maintenance immunosuppression and antibody induction therapy most likely played a major role in driving differences in the rates of BKPyV viremia. At some centers in our study, it is part of routine practice to either lower or discontinue the anti-metabolite when any BKPyV viremia is detected, however, at others, the anti-metabolite was discontinued only when the BKPyV viremia levels increased consistently above a certain level. Second, the assay used for detection of BKPyV varied between centers, and variation in screening protocols might have also played a small role leading to the difference in prevalence of BKPyV viremia. For instance, at Center 3, BKPyV viremia was only assessed when there was detectable viruria. Finally, geographic variation in the population prevalence of BKPyV might also have impacted the prevalence of BKPyV viremia between centers.
Our study has certain limitations. Even though we included four centers and 148 patients who received HCV-viremic kidney transplants (of whom 146 were matched to highly similar comparators), it is still possible that there might be a true underlying difference between BKPyV viremia > 1000 copies/mL between groups. However, we detected no graft losses due to BKPyV and graft function was excellent among HCV-viremic recipients. Our study only followed recipients for the first year; longer-term outcomes remain to be determined. We acknowledge clinical practice variation across centers in screening for BKPyV and responding to infection, although matching within center should limit any bias. Due to variation in clinical practice surrounding BKPyV virus screening at Vanderbilt, we excluded many HCV-aviremic patients as comparators because they were followed by nephrologists who did not routinely screen for BKPyV. However, we performed matching only after excluding those comparators who were not routinely screened for BKPyV, so that all matched pairs were subject to the same BKPyV screening practices by the same two nephrologists. There may be other unmeasured confounders, such as baseline immune function.
In conclusion, recipients of kidneys from deceased HCV-viremic donors did not have a higher risk of BKPyV viremia within 1-year after transplant, but our study did reveal a significantly higher rate of severe BKPyV viremia. These results suggest that HCV-viremic kidney recipients with detectable BKPyV should be closely monitored, given potential risk that BKPyV infection will be challenging to manage. Nonetheless, our study also demonstrated that recipients of HCV-viremic kidneys had excellent 1-year graft function, a general trend of resolution of BKPyV viremia over time, and allograft failure at 1-year was <1%. Given that opting-in for HCV-viremic kidney offers may increase the probability of receiving a kidney transplant sooner, or a transplant from a better quality donor, we conclude that there would need to be a very substantial risk of BKPyV viremia to offset the benefits of greater access to transplant.(5, 9) Our study detected no such substantial risk associated with HCV-viremic kidneys. This study provides additional evidence that HCV-viremic deceased donor kidneys offer a valuable resource for transplantation, while also signaling that immunological complications associated with donor HCV warrant further investigation.
Supplementary Material
Acknowledgments:
The results of this paper have not been published previously in whole or part.
Funding Support:
This study was supported by NIH grant K24-AI146137 (to Peter Reese). The THINKER trial of transplanting HCV-infected kidneys into uninfected recipients at the University of Pennsylvania was supported by an investigator-initiated grant from Merck. Vishnu S. Potluri was supported by a Satellite Dialysis Clinical Investigator Grant of the National Kidney Foundation. Douglas Schaubel was supported in part by R01-DK070869 from the National Institutes of Health.
The authors of this manuscript have conflicts of interest to declare as described by the American Journal of Transplantation. Dr. Molnar served as advisor for Merck, AbbVie, CareDx and Natera and received research support from CareDx and Viracor. Dr. Sise received grant support from Gilead Sciences, AbbVie and Merck & Co., and served as an advisor for AbbVie, Gilead Sciences, and as a consultant for Bioporto. Dr. Chung received grant support from AbbVie, Gilead Sciences, Merck & Co., Bristol Myers Squibb, Janssen, Boehringer, and Roche. Dr. Blumberg received research support from Merck and Takeda, and served as an advisor for Merck and Takeda. Dr. Sawinski served as an advisor for Veloxis, Natera, and CareDx. Dr. Goldberg received research grants from Merck & Co. and AbbVie to his institution to study the transplantation of HCV-infected organs into uninfected recipients followed by antiviral treatment. Dr. Reese received research grants from Merck and AbbVie to his institution to support research on transplantation of HCV-infected organs into uninfected recipients, followed by antiviral treatment, and from CVS Caremark to his institution to support research on medication adherence (focus: statins); and served as a consultant for VALHealth (focus: recognition of chronic kidney disease). Dr. Reese is an Associate Editor for the American Journal of Kidney Disease.
List of Abbreviations:
- ALT
Alanine Aminotransferase
- AST
Aspartate Aminotransferase
- BKPyV
BK Polyomavirus
- BMI
Body Mass Index
- CI
Confidence interval
- CMV
Cytomegalovirus
- cPRA
Calculated Panel Reactive Antibodies
- DAA
Direct Antiviral Agent
- DCD
Donation After Circulatory Death
- DSA
Donor Specific Antibodies
- EXPANDER
Exploring Renal Transplants Using Hepatitis C Infected Donors for HCV-Negative Recipients
- eGFR
Estimated Glomerular Filtration Rate
- FCH
Fibrosing Cholestatic Hepatitis
- HR
Hazard ratio
- HCV
Hepatitis C virus
- IRO
Increased Risk Organ
- IQR
Interquartile Range
- KDPI
Kidney Donor Profile Index
- NAT
Nucleic Acid Test
- PHS
Public Health Service
- PCR
Polymerase Chain Reaction
- REDCap
Research Electronic Data Capture
- RNA
Ribonucleic Acid
- SD
Standard Deviation
- SVR
Sustained Virologic Response
- THINKER
Transplanting Hepatitis C Kidneys into Negative KidnEy Recipients
- UNOS
United Network for Organ Sharing
Footnotes
Conflicts of Interest:
The rest of the authors declare no conflicts of interest.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request. Data requests will be reviewed by representatives of the Coauthor consortium.
References:
- 1.Afdhal N, Zeuzem S, Kwo P, Chojkier M, Gitlin N, Puoti M et al. Ledipasvir and sofosbuvir for untreated HCV genotype 1 infection. N Engl J Med 2014;370(20):1889–1898. [DOI] [PubMed] [Google Scholar]
- 2.Afdhal N, Reddy KR, Nelson DR, Lawitz E, Gordon SC, Schiff E et al. Ledipasvir and sofosbuvir for previously treated HCV genotype 1 infection. N Engl J Med 2014;370(16):1483–1493. [DOI] [PubMed] [Google Scholar]
- 3.Curry MP, O’Leary JG, Bzowej N, Muir AJ, Korenblat KM, Fenkel JM et al. Sofosbuvir and Velpatasvir for HCV in Patients with Decompensated Cirrhosis. N Engl J Med 2015;373(27):2618–2628. [DOI] [PubMed] [Google Scholar]
- 4.Molnar MZ, Nair S, Cseprekal O, Yazawa M, Talwar M, Balaraman V et al. Transplantation of kidneys from hepatitis C-infected donors to hepatitis C-negative recipients: Single center experience. Am J Transplant 2019;19(11):3046–3057. [DOI] [PubMed] [Google Scholar]
- 5.Reese PP, Abt PL, Blumberg EA, Van Deerlin VM, Bloom RD, Potluri VS et al. Twelve-Month Outcomes After Transplant of Hepatitis C-Infected Kidneys Into Uninfected Recipients: A Single-Group Trial. Ann Intern Med 2018;169(5):273–281. [DOI] [PubMed] [Google Scholar]
- 6.Durand CM, Bowring MG, Brown DM, Chattergoon MA, Massaccesi G, Bair N et al. Direct-Acting Antiviral Prophylaxis in Kidney Transplantation From Hepatitis C Virus-Infected Donors to Noninfected Recipients: An Open-Label Nonrandomized Trial. Ann Intern Med 2018;168(8):533–540. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gupta G, Yakubu I, Bhati CS, Zhang Y, Kang L, Patterson JA et al. Ultra-short duration direct acting antiviral prophylaxis to prevent virus transmission from hepatitis C viremic donors to hepatitis C negative kidney transplant recipients. Am J Transplant 2020;20(3):739–751. [DOI] [PubMed] [Google Scholar]
- 8.Sise ME, Strohbehn IA, Chute DF, Gustafson J, Van Deerlin VM, Smith JR et al. Preemptive Treatment With Elbasvir and Grazoprevir for Hepatitis C-Viremic Donor to Uninfected Recipient Kidney Transplantation. Kidney international reports 2020;5(4):459–467. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Potluri VS, Goldberg DS, Mohan S, Bloom RD, Sawinski D, Abt PL et al. National Trends in Utilization and 1-Year Outcomes with Transplantation of HCV-Viremic Kidneys. J Am Soc Nephrol 2019;30(10):1939–1951. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kapila N, Menon KVN, Al-Khalloufi K, Vanatta JM, Murgas C, Reino D et al. Hepatitis C Virus NAT-Positive Solid Organ Allografts Transplanted Into Hepatitis C Virus-Negative Recipients: A Real-World Experience. Hepatology 2020;72(1):32–41. [DOI] [PubMed] [Google Scholar]
- 11.Edwards GC, Shipe ME, Smith L, Gamble C, Shaffer D, Concepcion BP et al. Exploring patient willingness to accept hepatitis C-infected kidneys for transplantation. BMC Nephrol 2020;21(1):473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Couri T, Cotter TG, Chen D, Hammes M, Reddy B, Josephson M et al. Use of Hepatitis C Positive Organs: Patient Attitudes in Urban Chicago. Am J Nephrol 2019;49(1):32–40. [DOI] [PubMed] [Google Scholar]
- 13.Molnar MZ, Azhar A, Tsujita M, Talwar M, Balaraman V, Bhalla A et al. Transplantation of Kidneys From Hepatitis C Virus-Infected Donors to Hepatitis C Virus-Negative Recipients: One-Year Kidney Allograft Outcomes. Am J Kidney Dis 2021;77(5):739–747.e731. [DOI] [PubMed] [Google Scholar]
- 14.Antonsson A, Green AC, Mallitt KA, O’Rourke PK, Pawlita M, Waterboer T et al. Prevalence and stability of antibodies to the BK and JC polyomaviruses: a long-term longitudinal study of Australians. J Gen Virol 2010;91(Pt 7):1849–1853. [DOI] [PubMed] [Google Scholar]
- 15.Egli A, Infanti L, Dumoulin A, Buser A, Samaridis J, Stebler C et al. Prevalence of polyomavirus BK and JC infection and replication in 400 healthy blood donors. J Infect Dis 2009;199(6):837–846. [DOI] [PubMed] [Google Scholar]
- 16.Brennan DC, Agha I, Bohl DL, Schnitzler MA, Hardinger KL, Lockwood M et al. Incidence of BK with tacrolimus versus cyclosporine and impact of preemptive immunosuppression reduction. Am J Transplant 2005;5(3):582–594. [DOI] [PubMed] [Google Scholar]
- 17.Ramos E, Drachenberg CB, Papadimitriou JC, Hamze O, Fink JC, Klassen DK et al. Clinical course of polyoma virus nephropathy in 67 renal transplant patients. J Am Soc Nephrol 2002;13(8):2145–2151. [DOI] [PubMed] [Google Scholar]
- 18.Hirsch HH. Polyomavirus BK nephropathy: a (re-)emerging complication in renal transplantation. Am J Transplant 2002;2(1):25–30. [DOI] [PubMed] [Google Scholar]
- 19.Naesens M, Kuypers DR, De Vusser K, Evenepoel P, Claes K, Bammens B et al. The histology of kidney transplant failure: a long-term follow-up study. Transplantation 2014;98(4):427–435. [DOI] [PubMed] [Google Scholar]
- 20.Hirsch HH, Randhawa PS, Practice ASTIDCo. BK polyomavirus in solid organ transplantation-Guidelines from the American Society of Transplantation Infectious Diseases Community of Practice. Clin Transplant 2019;33(9):e13528. [DOI] [PubMed] [Google Scholar]
- 21.Nickeleit V, Singh HK, Randhawa P, Drachenberg CB, Bhatnagar R, Bracamonte E et al. The Banff Working Group Classification of Definitive Polyomavirus Nephropathy: Morphologic Definitions and Clinical Correlations. J Am Soc Nephrol 2018;29(2):680–693. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Babel N, Volk HD, Reinke P. BK polyomavirus infection and nephropathy: the virus-immune system interplay. Nat Rev Nephrol 2011;7(7):399–406. [DOI] [PubMed] [Google Scholar]
- 23.Meehan SM, Kadambi PV, Manaligod JR, Williams JW, Josephson MA, Javaid B. Polyoma virus infection of renal allografts: relationships of the distribution of viral infection, tubulointerstitial inflammation, and fibrosis suggesting viral interstitial nephritis in untreated disease. Hum Pathol 2005;36(12):1256–1264. [DOI] [PubMed] [Google Scholar]
- 24.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42(2):377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Sawinski D, Goral S. BK virus infection: an update on diagnosis and treatment. Nephrology, dialysis, transplantation : official publication of the European Dialysis and Transplant Association - European Renal Association 2015;30(2):209–217. [DOI] [PubMed] [Google Scholar]
- 26.Limaye A, Brennan D. Kidney transplantation in adults: BK polyomavirus-associated nephropathy. In. Uptodate. Uptodate, 2020. [Google Scholar]
- 27.Thangaraju S, Gill J, Wright A, Dong J, Rose C. Risk Factors for BK Polyoma Virus Treatment and Association of Treatment With Kidney Transplant Failure: Insights From a Paired Kidney Analysis. Transplantation 2016;100(4):854–861. [DOI] [PubMed] [Google Scholar]
- 28.Zubizarreta J, Kilcioglu C, Vielma J. Matched Samples that are Balanced and Representative by Design. Package ‘Designmatch’. In. CRAN: R Package, 2018. [Google Scholar]
- 29.Rosenbaum PR, Rubin DB. Constructing a Control Group Using Multivariate Matched Sampling Methods That Incorporate the Propensity Score. The American Statistician 1985;39(1):33–38. [Google Scholar]
- 30.Silber JH, Rosenbaum PR, Trudeau ME, Even-Shoshan O, Chen W, Zhang X et al. Multivariate matching and bias reduction in the surgical outcomes study. Med Care 2001;39(10):1048–1064. [DOI] [PubMed] [Google Scholar]
- 31.Rosenbaum PR. Design of observational studies. Springer, 2010. [Google Scholar]
- 32.Balba GP, Javaid B, Timpone JG. BK polyomavirus infection in the renal transplant recipient. Infect Dis Clin North Am 2013;27(2):271–283. [DOI] [PubMed] [Google Scholar]
- 33.Malik O, Saleh S, Suleiman B, Ashqar B, Maibam A, Yaseen M et al. Prevalence, Risk Factors, Treatment, and Overall Impact of BK Viremia on Kidney Transplantation. Transplant Proc 2019;51(6):1801–1809. [DOI] [PubMed] [Google Scholar]
- 34.Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, Feldman HI et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150(9):604–612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Shaffi K, Uhlig K, Perrone RD, Ruthazer R, Rule A, Lieske JC et al. Performance of creatinine-based GFR estimating equations in solid-organ transplant recipients. Am J Kidney Dis 2014;63(6):1007–1018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Greifer N. Covariate Balance Tables and Plots: A Guide to the cobalt Package. In.: R Package version 3.6.1, 2018. [Google Scholar]
- 37.Bohl DL, Brennan DC, Ryschkewitsch C, Gaudreault-Keener M, Major EO, Storch GA. BK virus antibody titers and intensity of infections after renal transplantation. J Clin Virol 2008;43(2):184–189. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Schachtner T, Stein M, Babel N, Reinke P. The Loss of BKV-specific Immunity From Pretransplantation to Posttransplantation Identifies Kidney Transplant Recipients at Increased Risk of BKV Replication. Am J Transplant 2015;15(8):2159–2169. [DOI] [PubMed] [Google Scholar]
- 39.Weist BJ, Schmueck M, Fuehrer H, Sattler A, Reinke P, Babel N. The role of CD4(+) T cells in BKV-specific T cell immunity. Med Microbiol Immunol 2014;203(6):395–408. [DOI] [PubMed] [Google Scholar]
- 40.Schachtner T, Muller K, Stein M, Diezemann C, Sefrin A, Babel N et al. BK virus-specific immunity kinetics: a predictor of recovery from polyomavirus BK-associated nephropathy. Am J Transplant 2011;11(11):2443–2452. [DOI] [PubMed] [Google Scholar]
- 41.Trydzenskaya H, Sattler A, Muller K, Schachtner T, Dang-Heine C, Friedrich P et al. Novel approach for improved assessment of phenotypic and functional characteristics of BKV-specific T-cell immunity. Transplantation 2011;92(11):1269–1277. [DOI] [PubMed] [Google Scholar]
- 42.Prosser SE, Orentas RJ, Jurgens L, Cohen EP, Hariharan S. Recovery of BK virus large T-antigen-specific cellular immune response correlates with resolution of bk virus nephritis. Transplantation 2008;85(2):185–192. [DOI] [PubMed] [Google Scholar]
- 43.Bhattarai N, McLinden JH, Xiang J, Mathahs MM, Schmidt WN, Kaufman TM et al. Hepatitis C virus infection inhibits a Src-kinase regulatory phosphatase and reduces T cell activation in vivo. PLoS Pathog 2017;13(2):e1006232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Christiaansen A, Varga SM, Spencer JV. Viral manipulation of the host immune response. Curr Opin Immunol 2015;36:54–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Khan ST, Karges W, Cooper CL, Crawley AM. Hepatitis C virus core protein reduces CD8(+) T-cell proliferation, perforin production and degranulation but increases STAT5 activation. Immunology 2018;154(1):156–165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Imlay H, Whitaker K, Fisher CE, Limaye AP. Clinical characteristics and outcomes of late-onset BK virus nephropathy in kidney and kidney-pancreas transplant recipients. Transpl Infect Dis 2018;20(4):e12928. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.