Cytomegalovirus Incidence Between Everolimus Versus Mycophenolate in De Novo Renal Transplants: Pooled Analysis of Three Clinical Trials

Authors


Daniel C. Brennan, brennan@wudosis.wustl.edu

Abstract

Everolimus (EVR) in heart and renal transplant (RTx) recipients may be associated with a decreased incidence of cytomegalovirus (CMV). A detailed analysis of the association between EVR versus mycophenolic acid (MPA) and CMV events has not been reported. CMV data from 2004 de novo RTx recipients from three-randomized, prospective, EVR studies A2309 (N = 833), B201 (N = 588) and B251 (N = 583) were retrospectively analyzed to identify differences between two EVR dosing groups and MPA. EVR groups received 1.5 mg/day, or 3 mg/day with either standard (SD-CsA) or reduced dose cyclosporine (RD-CsA). Controls received MPA with SD-CsA. CMV prophylaxis was as per center practice. CMV incidence (infection/syndrome, disease, viremia) was captured per local center evaluations. Kaplan–Meier analyses demonstrated that freedom from CMV viremia and infection/syndrome was significantly greater for EVR versus MPA for recipients without CMV prophylaxis. Among recipients who received prophylaxis, freedom from viremia was greater for EVR 3.0 mg; freedom from infection/syndrome was greater for EVR 3.0 and 1.5 mg. Although freedom from organ involvement was numerically greater for EVR, it was not statistically significant. This analysis documents significant reductions in the incidence of CMV infection/syndrome and viremia in EVR-treated de novo RTx recipients, especially those who did not receive CMV prophylaxis versus MPA.

Abbreviations: 
ANOVA

analysis of variance

ATG

antithymocyte globulin

CMV

cytomegalovirus

CNI

calcineurin inhibitor

CsA

cyclosporine

EC-MPS

enteric coated-mycophenolate sodium

EVR

everolimus

IF/TA

interstitial fibrosis and tubular atrophy

MMF

mycophenolate mofetil

mTOR

mammalian target of rapamycin

MPA

mycophenolic acid

RD-CsA

reduced dose cyclosporine

SD-CsA

standard dose cyclosporine

Introduction

Cytomegalovirus (CMV) is the most common viral pathogen occurring postrenal transplantation (1–3). The success of CMV prophylaxis with antivirals has resulted in a decrease in the incidence of CMV infection (4–7). However, CMV remains a significant pathogen, associated with allograft rejection and loss, mortality, interstitial fibrosis and tubular atrophy (IF/TA) in protocol biopsies, and increased post-transplant costs (8–12). In addition, the issue of CMV infection occurring after cessation of CMV prophylaxis (“late CMV infection and disease”) has arisen as an important clinical problem (13–15). An immunosuppressive drug effective in preventing rejection and that attenuated the incidence of CMV events would be beneficial for patients.

Everolimus (EVR) is a mammalian target of rapamycin (mTOR) inhibitor that has both immunosuppressive and antiproliferative effects (16). EVR has been shown to be efficacious in preventing acute allograft rejection (17). Use of EVR may allow for minimization of calcineurin inhibitor (CNI) exposure with preservation of efficacy and renal function (17,18). Clinical data also suggest that EVR is associated with a decrease in CMV compared to mycophenolic acid (MPA) (19,20). CMV replication is dependent upon 1 of 2 mTOR pathways and in vitro studies support an association between mTOR inhibitors and decreased CMV (21). Following cardiac transplantation, use of EVR was associated with a lower incidence of CMV events (viremia, syndrome, and infection) compared to MPA (19). Individual clinical trial data have also shown that use of EVR was associated with a lower incidence of CMV infection compared to MPA following renal transplantation (20). However, in renal transplantation, a complete understanding of the reduction of CMV events with the use of EVR remains unaddressed. Utilizing data from several large clinical trials (B201, B251, A2309), an analysis pooling data was conducted to determine more completely the association between EVR and the incidence of specific CMV events compared to MPA.

Materials and Methods

Data sources and endpoints

Data from three randomized clinical trials were pooled for the present analysis (Figure 1). Freedom from and incidence of CMV events, including viremia, infection or syndrome, or CMV with organ involvement, was compared between the EVR groups and the MPA group. We elected to examine CMV events categorized as viremia (no clinical symptoms), infection/syndrome (viremia and clinical symptoms present) or CMV with organ involvement as an attempt to capture all the events while grouping them in a logical manner. In addition to examining the specific categories above, we also included as an outcome, “any CMV event” which was a composite of the above categories. In study 2309, investigators specifically reported individual CMV events (laboratory-defined CMV [antigenemia-positive, PCR positive], CMV syndrome [fever that lasted 2 days, neutropenia, leucopenia, viral syndrome] and CMV-disease [organ involvement]) on a case report form designed to capture such events. In studies B201 and B251, CMV events were reported as part of the data collection on infections and adverse events or serious adverse events and classified by the investigator as CMV viremia (infection with no clinical symptoms), CMV infection or syndrome (viremia plus clinical symptoms), or CMV tissue invasive disease. Endpoints assessed were stratified by use of CMV prophylaxis; incidence of CMV infection or syndrome was additionally stratified by donor/recipient CMV serostatus.

Figure 1.

Description of the study designs of the three trials from which data were pooled for the present analysis.

Patient population and treatment

Patient selection and treatment have been described previously (22–24). In brief, studies B201 (n = 588) and B251 (n = 583) were both multicenter, randomized, double-blind, parallel-group equivalence trials of two oral fixed doses of EVR (1.5 mg/day, B201: n = 194; B251: n = 193; or 3 mg/day, B201: 198; B251: n = 194) versus mycophenolate mofetil (MMF) (B201: n = 196; B251: n = 196) in combination with standard dosing of cyclosporine (CsA) microemulsion (Neoral) and corticosteroids in adult, de novo RTx recipients. EVR levels were collected prospectively within these studies. Study A2309 (N = 833) was a multicenter, randomized, open-label, parallel-group noninferiority trial of two oral concentration controlled doses of EVR (1.5 mg/day, targeted to 3–8 ng/mL, n = 277; 3.0 mg/day, targeted to 6–12 ng/mL, n = 279) versus enteric coated mycophenolate sodium (EC-MPS) (n = 277). EVR was administered with reduced dose CsA while EC-MPS was with standard dose CsA. All three groups received basiliximab induction, and steroids were used according to local center practice.

Key exclusion criteria for all three studies included kidneys donated after cardiac death or with a cold ischemia time >40 h; donor age >65 years; or recipients of multiorgan-, ABO-incompatible-, positive T-cell cross-match- or HLA-identical living-related-donor transplants. In A2309 an additional exclusion criteria was most recent anti-HLA class I panel-reactive antibodies (PRA) >20% by a complement-dependent cytotoxicity (CDC) based assay or >50% by flow cytometry or ELISA.

All studies were approved by the Ethics Committee at all participating institutions and conducted according to the recommendations of Good Clinical Practice and the Declaration of Helsinki. All patients gave written informed consent to participate in the original studies.

CMV prophylaxis

In studies B201 and B251 CMV prophylaxis with ganciclovir, CMV hyperimmune globulin or acyclovir was mandatory for all CMV-negative recipients of a transplant from a CMV positive donor and was recommended after antibody treatment of an acute rejection episode. In study A2309 CMV prophylaxis (≥30 days; ganciclovir, CMV hyperimmune globulin, acyclovir or valacyclovir; according to local practice) was mandatory for all CMV-negative recipients who received a kidney from a CMV-positive donor. Other patients received CMV prophylaxis according to local practice.

Statistical analysis

To maximize the precision in the analyses, and given the similarity between the studies in terms of design (e.g. randomized), length of follow-up (at least 24 months) and treatment groups, data from all three studies were pooled. Cochran–Mantel–Haenszel generalized association test controlled for the study was used to compare categorical data between groups. Analysis of variance (ANOVA) with treatment and study as factors was used to compare continuous data. Kaplan–Meier survival analysis was used to determine the proportion of recipients in each group who were free from CMV events at 720 days post-transplantation; the log-rank test was used to compare survival curves between the EVR and MPA groups. Cox proportional hazard regression was used to estimate the risk for CMV events between the groups (MPA therapy vs. EVR), after adjustment for a potential study effect, and donor/recipient CMV serology. A p-value of < 0.05 was used to determine statistical significance.

Previously published data have found that acyclovir exhibits inadequate anti-CMV activity (25,26). Therefore, we conducted the following sensitivity analyses: in one analysis, recipients who were administered acyclovir were classified as not receiving CMV prophylaxis; in another analysis recipients who were administered acyclovir were classified as receiving CMV prophylaxis. The results from the Kaplan–Meier analyses and the Cox proportional hazard regression for the two analyses were very similar. The results presented in this paper reflect those with acyclovir classified as CMV prophylaxis.

Results

Patients

A total of 2004 patients comprised the pooled data sample (EVRs 1.5 mg, n = 664; EVR 3.0 mg, n = 671; MPA, n = 669). The three groups were similar in age, race, underlying cause of end-stage renal disease, cold ischemia time, HLA mismatches and donor age and type. A significant (p = 0.005) difference was observed among the three groups for recipient gender (% males: MPA 69%; EVR 3.0 mg: 66%; EVR 1.5 mg: 60%). Donor and recipient CMV serostatus combinations were similar among the three groups with the most frequent combination being D+/R+; approximately 17% in each group were D+/R− (Table 1).

Table 1.  Baseline characteristics of pooled population, studies B201, B251, A2309
 Everolimus 1.5 mg/day N = 664Everolimus 3 mg/day N = 671MPA N = 669
  1. MPA = mycophenolic acid; SD = standard deviation; HLA = human leukocyte antigen; D = donor; R = recipient; ATG = antithymocyte globulin (rabbit); CMV = cytomegalovirus.

  2. 1For studies B201 and B251, reflects any recorded use of ATG through 24 months posttransplantation, for study A2309 reflects any recorded use of ATG through 12 months posttransplantation.

  3. 2Denominator used in calculating percentages represent total number of recipients who received CMV prophylaxis in each group: EVR 1.5 mg, n = 345; EVR 3.0 mg, n = 348; MPA, n = 351; recipients may have received >1 prophylaxis agent.

  4. *Overall group comparison, p = 0.005.

Recipient age in years, mean ± SD44.9 ± 12.344.5 ± 12.645.8 ± 12.5
Male, n (%)401 (60)*441 (66)*460 (69)*
Caucasian, n (%)510 (77)482 (72)494 (74)
Cause of end-stage renal disease, n (%)
Hypertension/nephrosclerosis96 (15)120 (18)104 (16)
Glomerular disease166 (25)170 (25)162 (24)
Diabetes mellitus73 (11)77 (12)94 (14)
Polycystic disease104 (16)73 (11)97 (15)
Unknown/other225 (34)231 (34)212 (32)
≥3 HLA mismatches, n (%)487 (73)471 (70)477 (71)
Donor age in years, mean ± SD39.8 ± 14.240.2 ± 13.540.5 ± 14.4
Deceased heart beating, n (%)384 (58)388 (58)367 (55)
D/R CMV serology, n (%)
D+/R−105 (17)105 (16)107 (17)
D+/R+286 (45)298 (47)276 (43)
D−/R+101 (16)119 (19)119 (19)
D−/R−143 (23)116 (18)139 (22)
Received ATG1, n (%)32 (5)28 (4)37 (6)
CMV prophylaxis, n (%)
Any CMV prophylaxis345 (52)348 (52)351 (53)
Specific CMV prophylactic agents2, n (%)
Acyclovir134 (39)139 (40)142 (40)
Ganciclovir133 (39)130 (37)133 (38)
Valganciclovir95 (28)95 (27)98 (28)
Valacyclovir21 (6)15 (4)20 (6)
CMV hyperimmune globulin9 (3)5 (1)9 (3)

Incidence of biopsy-proven acute rejection (BPAR), which was assessed over the duration of each of the studies, did not significantly differ (p = 0.1891) between the groups (BPAR rates: 21.4% for EVR 1.5 mg; 19.7% for EVR 3 mg; and 23.8% for MPA). The percentage of patients receiving CMV prophylaxis after a rejection episode was EVR 1.5 mg: 25.5%; EVR 3.0 mg: 32.4%; MPA: 37.5%.

Immunosuppression and CMV prophylaxis

Mean EVR trough levels ranged from 2.2 ng/mL to 5.9 ng/ mL for the 1.5 mg EVR group and 4.5 ng/mL to 8.5 ng/mL for the 3 mg EVR group. MPA doses were similar across the studies. Mean CsA trough levels were consistently higher for the MPA group compared to the EVR groups. Use of thymoglobulin®[Antithymocyte Globulin (Rabbit) [ATG]] as an induction agent was infrequent (Table 1).

Slightly more than half of the patients in each of the groups received CMV prophylaxis (EVR 1.5 mg: 52%; EVR 3.0 mg: 52%; MPA: 53%). As expected, use of CMV prophylaxis was greatest among recipients who were CMV− and received a kidney from a CMV+ donor (EVR 1.5 mg: 85%; EVR 3.0 mg: 79%; MPA: 86%). The specific CMV prophylaxis agent used did not differ significantly between the three groups (Table 1). Among the D+/R− group, the most common agent used was ganciclovir (EVR 1.5 mg: 55%; EVR 3.0 mg: 59%; MPA: 53%), followed by valganciclovir (EVR 1.5 mg: 30%; EVR 3.0 mg: 29%; MPA: 37%), acyclovir (EVR 1.5 mg: 19%; EVR 3.0 mg 11%; MPA, 15%), CMV immunoglobulin (EVR 1.5 mg: 10%; EVR 3.0 mg: 5%; MPA: 8%) and valacyclovir (EVR 1.5 mg: 4%; EVR 3.0 mg: 5%; MPA: 7%).

CMV events

Often, it is not appropriate to interpret Kaplan–Meier results as “incidence” since 1− the Kaplan–Meier estimator may result in an over estimation of the cumulative probability of the event in question. Therefore, we report both the Kaplan–Meier freedom from events data below, followed by comparisons of the cumulative incidence of CMV events.

Among recipients who did not receive CMV prophylaxis, a significantly greater proportion of recipients in the EVR 1.5 mg (93.5%) and 3.0 mg (95.5%) groups were free from CMV events (viremia, infection/syndrome, or disease) compared to the MPA group (85.2%) (Figure 2A). The adjusted hazard ratio (HR) from the Cox model demonstrated that MPA was associated with a significantly increased risk for CMV events (HR: 2.81, 95% CI: 1.77–4.50, p < 0.0001).

Figure 2.

Kaplan–Meier plot for any CMV event, comparison between everolimus 1.5 mg/day, everolimus 3.0 mg/day and MPA for adult renal transplant recipients (A) who did not receive CMV prophylaxis and (B) who received CMV prophylaxis.

Among recipients who received CMV prophylaxis, a significantly greater proportion of recipients in the EVR 1.5 mg (93.9%) and 3.0 mg (93.2%) groups were free from CMV events (viremia, infection/syndrome, or disease) compared to the MPA group (88.0%) (Figure 2B). The adjusted hazard ratio from the Cox model for MPA versus EVR demonstrated that MPA was associated with a significantly increased risk for CMV events (HR: 1.82, 95% CI: 1.18–2.80, p = 0.0063).

CMV viremia

Among recipients who did not receive CMV prophylaxis, a significantly greater proportion of recipients in the EVR 1.5 mg (96.7%) and 3.0 mg (97.4%) groups were free from CMV viremia compared to the MPA group (90.7%) (Figure 3A). The adjusted hazard ratio from the Cox model for MPA versus EVR demonstrated that MPA was associated with a significantly increased risk for CMV viremia (HR: 3.34, 95% CI: 1.79–6.42, p = 0.0002).

Figure 3.

Kaplan–Meier plot for CMV viremia, comparison between everolimus 1.5 mg/day, everolimus 3.0 mg/day and MPA for adult renal transplant recipients (A) who did not receive CMV prophylaxis and (B) who received CMV prophylaxis.

Among recipients who received CMV prophylaxis, a significantly (p = 0.012) greater proportion of recipients in the EVR 3.0 mg (97.3%) group were free from CMV viremia compared to the MPA group (93.3%) (Figure 3B). The proportion of recipients free from CMV viremia in the EVR 1.5 mg group (95.1%) was numerically greater than the MPA group but the difference was not statistically significant (p = 0.505). The adjusted hazard ratio from the Cox model for MPA versus EVR for the prophylaxis group was 1.66 (95% CI: 0.93–2.93, p = 0.0821).

CMV infection or syndrome

Among recipients who did not receive CMV prophylaxis, a significantly (p ≤ 0.0001) greater proportion of recipients in the EVR 1.5 mg (94.8%) and 3.0 mg (96.1%) groups were free from CMV infection/syndrome compared to the MPA group (85.9%). The adjusted hazard ratio from the Cox model for MPA versus EVR for the no prophylaxis group demonstrated that MPA was associated with a significantly increased risk for CMV infection or syndrome (HR: 3.33, 95% CI: 2.03–5.54, p < 0.0001).

Among recipients who received CMV prophylaxis, a significantly greater proportion of recipients in the EVR 1.5 mg (94.2%) and 3.0 mg (93.8%) groups were free from CMV infection/syndrome compared to the MPA group (89.2%). The adjusted hazard ratio from the Cox model for MPA versus EVR for the prophylaxis group demonstrated that MPA was associated with a significantly increased risk for CMV infection or syndrome (HR: 1.71, 95% CI: 1.08–2.67, p = 0.0198).

CMV with organ involvement

The incidence of CMV with organ involvement was low regardless of prophylaxis use. Among recipients who did not receive CMV prophylaxis there was no significant difference in the proportion of recipients in the EVR 1.5 mg (97.4%), EVR 3.0 mg (99.0%), or MPA (98.7%) group free from CMV with organ involvement (Figure 4A). The adjusted hazard ratio from the Cox model for MPA versus EVR for the no prophylaxis group was 0.73 (95% CI: 0.20–2.14, p = 0.5911).

Figure 4.

Kaplan–Meier plot for CMV with organ involvement, comparison between everolimus 1.5 mg/day, everolimus 3.0 mg/day and MPA for adult renal transplant recipients (A) who did not receive CMV prophylaxis and (B) who received CMV prophylaxis.

Among recipients who received CMV prophylaxis, a significantly (p = 0.023) greater proportion of recipients in the EVR 1.5 mg (99.1%) group were free from CMV with organ involvement compared to the MPA group (96.5%) (Figure 4B). The proportion of recipients free from CMV with organ involvement in the EVR 3.0 mg group (97.9%) was not significantly (p = 0.259) different from the MPA group. The adjusted hazard ratio from the Cox model for MPA versus EVR for the prophylaxis group was 2.05 (95% CI: 0.86–4.94, p = 0.1032).

Incidence and timing of CMV events

No prophylaxis: Similar to the results from the Kaplan–Meier analysis, among the recipients who did not receive CMV prophylaxis, the incidence of CMV infection or syndrome was significantly lower (p < 0.0001) for the 1.5 mg (5%) and 3.0 mg (4%) EVR groups compared to the MPA group (14%). The incidence of viremia was also significantly (p ≤ 0.0016) lower for both the 1.5 mg and 3.0 mg EVR groups (3% for both) compared to the MPA group (9%) (Table 2).

Table 2.  Incidence of CMV events by 1 year post-transplantation by CMV prophylaxis
 Everolimus 1.5 mg/day N = 319Everolimus 3.0 mg/day N = 323MPA N = 318
No prophylaxis
 CMV infection/syndrome, n (%)16 (5.0)**14 (4.3)**44 (13.8)
 CMV viremia, n (%)10 (3.1)#10 (3.1)#29 (9.1)
 CMV with organ involvement, n (%)8 (2.5)3 (0.9)4 (1.3)
 Everolimus 1.5 mg/day N = 345Everolimus 3.0 mg/day N = 348MPA N = 351
  1. **p < 0.0001 versus MPA; #p ≤ 0.0016 versus MPA; *p < 0.04 versus MPA.

Prophylaxis
 CMV infection/syndrome, n (%)21 (6.1)*21 (6.0)*37 (10.5)
 CMV viremia, n (%)18 (5.2)9 (2.6)*23 (6.6)
 CMV with organ involvement, n (%)3 (0.9)*7 (2.0)12 (3.4)

Prophylaxis: Among recipients who received CMV prophylaxis, the incidence of CMV infection or syndrome was significantly (p < 0.04) lower for both the 1.5 mg and 3.0 mg EVR groups (6% for both) compared to the MPA group (11%). The incidence of CMV viremia was significantly lower (p < 0.04) for the 3.0 mg (3%) EVR groups compared to the MPA group (7%), while the incidence of CMV with organ involvement was significantly (p < 0.04) lower for the EVR 1.5 mg group (0.9%) compared to the MPA group (3%) (Table 2).

By D/R serostatus: The incidence of CMV infection or syndrome was numerically lower for both the 1.5 mg and 3.0 mg EVR groups compared to the MPA group for all D/R CMV serostatus combinations. Statistically significantly lower incidences were found within the D+/R+ group for both the 1.5 mg (5%, p = 0.003) and 3.0 mg (3%, p < 0.001) EVR compared to MPA (12%), and within the D−/R+ group for 3.0 mg EVR (0.8%) compared to MPA (7%), p = 0.036. For the D+/R− group, the incidence of CMV infection or syndrome was EVR 1.5 mg: 16%, EVR 3.0 mg: 18%, and MPA: 25%, (p = 0.1278 for 1.5 mg vs. MPA and p = 0.2445 for 3.0 mg vs. MPA) (Table 3).

Table 3.  Incidence of CMV infection or syndrome by D/R CMV serostatus
Donor/recipient CMV serostatusEverolimus 1.5 mg/dayEverolimus 3 mg/dayMPA
  1. p = 0.0034 versus MPA; *p < 0.001 versus MPA; **p = 0.036 versus MPA.

N105105107
D+/R−, n (%)17 (16.2%)19 (18.1%)27 (25.2%)
N286298276
D+/R+, n (%)14 (4.9%)10 (3.4%)*33 (12.0)
N101119119
D−/R+, n (%)3 (3.0%)1 (0.8%)**8 (6.7%)
N143116139
D−/R−, n (%)1 (0.7%)2 (1.7%)6 (4.3%)

Among recipients who experienced a CMV event, both EVR 1.5 mg and 3.0 mg were associated with a significantly (p < 0.001) longer mean time to first CMV event compared to MPA (mean number of days to first CMV event, EVR 1.5 mg: 194; EVR 3.0 mg: 190; MPA: 124).

Discussion

Results from this pooled analysis of over 2000 de novo RTx recipients demonstrated that EVR was associated with a decrease in and delay in the time to onset of CMV events compared to MPA. The decrease in CMV infection or syndrome was found regardless of whether CMV prophylaxis had been used. While EVR was associated with numerically lower instances of CMV with organ involvement, the low frequency of events may have prevented detecting statistically significant differences between all group comparisons. When considering events by D/R CMV serostatus, the incidence of CMV infection or syndrome was lower for EVR versus MPA in each D/R category. However, statistical significance was only found within the D+/R+ and D−/R+ groups. Notably, the delay in the onset of the incidence of a CMV event occurred at a mean of 60–70 days later post-transplantation for EVR versus MPA.

CMV infection is associated with many deleterious indirect effects including rejection (8), IF/TA (12), mortality (8–10), and diabetes (8,27). Thus, preventing CMV infection is beneficial beyond the prevention of CMV disease-associated morbidity. In addition to the potential for undesirable clinical outcomes associated with CMV, there is also a negative economic aspect. Kidney transplant recipients who developed CMV disease have been found to use significantly more inpatient and outpatient resources than patients without CMV disease (28). CMV disease was also found to be an independent risk factor for increased resource utilization following liver transplantation (29). Perhaps most costly, is the economic burden of CMV-associated graft failure (30). Unfortunately, universal prophylaxis is associated with high treatment costs and the potential for drug-related toxicity. It can be speculated that use of EVR may offer additional economic benefits in terms of the decreased utilization associated with prevention of CMV disease, and decreased use of costly prophylaxis. However, whether a cost-savings, based on lower resource utilization, graft loss and/or decreased use of prophylaxis, is associated with EVR cannot be determined by the present study.

The lower frequency of CMV events associated with EVR use in the present pooled analysis is consistent with that of previously published results comparing EVR to MPA following kidney and heart transplantation (20,31–33). This is the first analysis to examine individual CMV events, however. Similarly, use of the mTOR inhibitor sirolimus has also been found to be associated with decreased risk and/or frequency of CMV events (32,34–36). The exact mechanism behind the association between mTOR inhibitors and decreased CMV events is not clear. However, the role of mTOR in CMV replication is beginning to be defined and compelling data suggest a role of mTOR inhibitors in CMV modulation (21,37). CMV requires mTOR for activation, EVR may affect viral amplification by blocking cellular proliferation and impairing pathways critical for CMV infection, signaling, and replication (21,37).

The limitations of the data should be taken into account when interpreting the results from the present analysis. The definition of CMV events varied between the individual studies from which the data were pooled, and all events were self-reported by the individual centers. As a result, the reported frequencies of the CMV events may be biased. However, a reporting bias would not be expected to differ among the EVR and MPA groups within each of the studies. Different agents were used as prophylaxis and this may have impacted the results as not all anti-CMV drugs are equally efficacious. For instance, acyclovir is not as effective as ganciclovir or valganciclovir in preventing CMV disease (38). The type and duration of prophylaxis may have been altered due to intolerance or other reasons. Importantly, the duration of prophylaxis was similar among the groups (median of 94 days for EVR 1.5 and 3.0 groups and 92 days for the MPA group). The use of antivirals and immunoglobulin did not differ between the EVR and MPA groups. In the analysis we only considered prophylaxis use that was recorded up to 10 days post-transplantation. Some recipients could have received antivirals after this time frame (e.g. preemptive therapy) and would have been misclassified as having received no CMV prevention therapy. It would not be expected for the misclassification to systematically differ between the drug groups, however. Furthermore, the result of decreased CMV infection or syndrome observed for the EVR group was regardless of prophylactic therapy. Finally, the association of less CMV with use of EVR might be explained by relatively less potent immunosuppression. However, the acute rejection rates were similar between those who received EVR compared to MPA which suggests that the “net state of immunosuppression” for any group was similar.

The results from this study add to the body of evidence suggesting EVR is associated with a decrease in CMV events compared to MPA. The next step in evaluating the association between EVR and CMV events is a prospective study designed to specifically evaluate CMV events in a well-defined systematic manner. As such, the currently on-going CRAD001AUS92 study (ClinicalTrials.gov identifier: NCT01025817) in kidney transplant recipients, which is specifically designed to evaluate the incidence of CMV events between EVR and MPA in a prospective manner by protocol specified testing, will contribute additional important information to corroborate the results presented here. Results from US92 are expected to be available January 2013.

Acknowledgments

The authors wish to thank all the A2309, B201, B251 investigators who contributed to the studies. [Text added after online publication August 3, 2011]

The authors thank Kristin Kistler of United BioSource Corp. for medical-writing support.

Portions of these data were presented at American Society of Nephrology, Renal Week, Nov. 16–20, 2010, Denver, CO.

Funding source: Funding for this study was provided by Novartis Pharmaceuticals Corporation.

Commercial Organizations: United BioSource Corporation provided assistance in the preparation of this manuscript. Novartis Pharmaceuticals Corporation provided financial support for the writing of this manuscript.

Disclosure

The authors of this manuscript have conflicts of interest to disclose as described by the American Journal of Transplantation. D.C.B. is a Consultant for Novartis, Genentech, Genzyme and Pfizer, Speaker honoraria for Novartis, Genentech and Genzyme. He received grant/research support from Pfizer to Washington University. C. L. received Lecture fees from Novartis, Roche, Pfizer and BMS. D.P. is an employee of Novartis Pharmaceuticals Corporation. K.M. is an employee of Novartis Pharmaceuticals Corporation. A.W. is an employee of Novartis Pharmaceuticals Corporation. K.M. is an employee of Novartis Pharmaceuticals Corporation. S.F.S. is a speaker and received grant support from Novartis.

Ancillary