Background.The incidence and clinical and virologic aspects of ganciclovir-resistant cytomegalovirus (CMV) disease have not been well-characterized in heart transplant recipients.
Methods.We retrospectively analyzed all patients who underwent their first heart transplantation during the period from 1 January 1995 through 30 June 2005 at a single health care center. Cox proportional hazard regression was used to assess the relationship between clinical variables and CMV disease. Portions of the UL97 gene were sequenced in patients with slow clinical and/or virologic response to ganciclovir therapy.
Results.Cytomegalovirus disease developed in 32 (11.7%) of 274 patients at a median of 4.2 months after transplantation (range, 1.8–11.6 months after transplantation) and was independently associated with donor-seropositive/recipient-seronegative (D+/R-) serostatus (adjusted hazard ratio, 6.93; P < .001). The incidence of ganciclovir-resistant CMV disease was 1.5% overall (4 of 274 patients), 5% among D+/R- serostatus recipients (4 of 80 patients), and 12.5% among patients who developed CMV disease (4 of 32 patients). Ganciclovir-resistant CMV disease was significantly associated with D+/R- serostatus (4 [5%] of 80 vs. 0 [0%] of 153 patients; P = .02), greater prior exposure to ganciclovir (median duration of exposure, 150 vs. 69 days; P = .003), and substantial morbidity, including prolonged CMV-associated hospitalization (median duration of hospitalization, 66 vs. 0 days; P < .01).
Conclusions.CMV disease, including ganciclovir-resistant disease, is an important clinical problem in D+/R- heart transplant recipients who receive antiviral prophylaxis. Strategies specifically designed to reduce the incidence and impact of CMV disease in this population are warranted.
Cytomegalovirus (CMV) is an important cause of direct and indirect morbidity in transplant patients , and various preventive strategies have been developed. The use of antiviral prophylaxis has been shown to be an effective strategy for prevention of CMV in solid-organ transplant (SOT) recipients [2,3–4]. However, late-onset CMV disease (i.e., CMV disease that occurs after the discontinuation of prophylaxis) has been recognized as a potential limitation of the antiviral prophylaxis strategy in several SOT populations [5,6–7]. However, the incidence of and risk factors for late-onset CMV disease have been less well-studied in heart transplant recipients who received antiviral prophylaxis.
The use of antiviral prophylaxis has also been associated with the emergence of ganciclovir resistance in SOT recipients [8,9–10], and several prior studies have characterized the incidence and clinical manifestations in specific SOT populations, including abdominal organ transplant recipients  and lung transplant recipients [11,12–13]. Among these SOT recipients, ganciclovir-resistant CMV was associated with significant morbidity and mortality [8, 11]. However, much less information is available about the incidence and clinical aspects of ganciclovir-resistant CMV disease in heart transplant recipients. Thus, given these limitations in the existing literature, we sought to characterize the incidence and clinical and virologic features of ganciclovir-resistant CMV disease in a cohort of heart transplant recipients who received antiviral prophylaxis, with the goal of identifying specific risk factors for the development of ganciclovir-resistant CMV disease in this clinical context.
Patient population.All patients who received their first heart transplant at the University of Washington Medical Center (Seattle) during the period from 1 January 1995 through 30 June 2005 were included in the study. All data were collected through review of clinical laboratory, pathology, and clinical transplantation databases, as well as electronic and paper medical records, using standardized data collection forms. All patients were observed for a minimum of 12 months or until death. The study was approved by the Institutional Review Board/Human Subjects Division of the University of Washington Medical Center.
CMV prophylaxis and treatment of CMV disease.All CMV-seropositive (R+) patients or patients who received an organ from a CMV-seropositive donor (D+) received prophylactic ganciclovir therapy. Before the US Food and Drug Administration approved oral ganciclovir in December 1996, patients received intravenous ganciclovir at 5 mg/kg every 12 h (adjusted for renal function) for postoperative days 1–14 and 6 mg/kg per day for postoperative days 15–28 , followed by oral acyclovir (800 mg orally 4 times daily) for 3 months (regimen A). After the US Food and Drug Administration approved oral ganciclovir, patients received intravenous ganciclovir at 5 mg/kg per day (adjusted for renal function) until they were able to tolerate oral medications (usually ⩽7 days after transplantation), followed by oral ganciclovir (1 g 3 times per day, adjusted for renal function) for 3 months (regimen B). CMV-seronegative transplant recipients with CMV-seronegative donors (the D-/R- group) received oral acyclovir (400 mg 3 times per day) for a period of 3 months after transplantation as standard herpes simplex prophylaxis.
Treatment of CMV disease consisted of intravenous ganciclovir (5 mg/kg every 12 h, adjusted for renal function) until signs and symptoms of CMV disease resolved and pp65 antigenemia test results became negative (usually within 4–6 weeks). Intravenous foscarnet treatment (90 mg/kg every 12 h, adjusted for renal function), given either alone or in combination with intravenous ganciclovir, was administered to patients if ganciclovir resistance was documented or clinically suspected.
Immunosuppression, rejection therapy, and prophylaxis.Standard induction and maintenance immunosuppressive therapy was administered per protocol and varied during the study period. Patients routinely received antithymocyte globulin and methylprednisolone as induction therapy. Patients who were intolerant to antithymocyte globulin received an IL-2 receptor antagonist (daclizumab). Routine maintenance immunosuppression consisted of a calcineurin inhibitor (tacrolimus or cyclosporine), an antimetabolite (mycophenolate mofetil or azathioprine), and prednisone. Endomyocardial biopsies were performed as surveillance for allograft rejection on a standardized schedule (weekly for 6 weeks, every 2–4 weeks for up to 6 months after transplantation, then 1–4 months for up to 3 years after transplantation) and when clinically indicated. Clinically significant rejection was defined as a rejection determined to be grade ⩾2 using the International Society for Heart and Lung Transplant grading system and accepted pathologic criteria . Episodes of rejection with International Society for Heart and Lung Transplant grade 2 were routinely treated with pulsed-dosed oral prednisone (40–100 mg orally per day for 3 days), whereas episodes of severe rejection (grade ⩾3A) were treated with methylprednisolone (125–1000 mg intravenously per day for 3 days) and/or antithymocyte globulin (1–1.5 mg/kg intravenously per day for 4 days).
CMV definition and diagnosis of disease.Standard definitions of CMV disease were used as described elsewhere [5, 15], with a previously described modification . We performed genotypic resistance testing by nucleotide sequencing, either directly from plasma and/or from blood , if ganciclovir resistance was suspected (i.e., if patients did not experience clinical improvement or if the viral load had not decreased after 14 days of full-dose intravenous ganciclovir therapy [8, 9]).
Statistical analysis.In these analyses, patients with primary allograft failure who required re-transplantation during the first year of follow-up were censored at the time of subsequent transplantation. Estimation of the cumulative incidence of CMV disease was calculated with death as a competing risk. The Wilcoxon rank sum test was used to compare the times to onset of CMV disease, and Fisher's exact test was used to compare proportions. Cox proportional-hazard regression analysis was performed to identify independent predictors of CMV disease and ganciclovir-resistant CMV disease, as well as mortality at 1 year. Variables with P values <.1 in the univariate model were entered into multivariable models. Log rank tests were used to assess the significance of covariates. P values <.05 were considered to be statistically significant. Allograft rejection was assessed as a time-dependent risk factor for CMV disease, and CMV disease was assessed as a time-dependent risk factor for mortality.
Patient population.The characteristics of the study population are shown in table 1.
The incidence and clinical manifestations of CMV disease.The cumulative incidence of CMV disease, stratified by donor and recipient CMV serostatus, is shown in figure 1. Because the CMV prophylaxis regimen varied during the study period, the incidence of CMV disease was also analyzed on the basis of prophylactic regimen (figure 1). The incidence of CMV disease at 1 year after transplantation was higher among patients who received the shorter antiviral prophylaxis regimen (i.e., regimen A, which consisted of 28 days of intravenous ganciclovir) than among those who received the longer antiviral prophylaxis regimen (i.e., regimen B, which consisted of 3 months oral ganciclovir) in both the D+/R- group (53.8% vs. 25.8%; P < .01) and the R+ group (18.5% vs. 2.4%; P < .01). The time to onset of CMV disease in the D+/R- group also occurred significantly earlier among patients who received regimen A, compared with those who received regimen B (median, 83 vs. 165 days; P < .01). The clinical manifestations of CMV disease, as stratified by serostatus, are shown in table 2.
Risk factors for CMV disease.Multivariate analysis of risk factors associated with the development of CMV disease is shown in table 3. Only D+/R- serostatus and the receipt of the shorter antiviral prophylaxis regimen (regimen A) were independent risk factors for the development of CMV disease (adjusted hazard ratio [HR], 6.93 [95% CI, 3.11–15.5; P < .001] and 4.17 [95% CI, 2.03–8.57; P < .001], respectively). After restricting the analysis to only patients with tissue-invasive CMV disease, D+/R- serostatus remained the only independent risk factor for the development of CMV disease (adjusted HR, 3.73; 95% CI, 1.35–10.27; P = .01). The year of transplantation was significantly associated with an increased risk of CMV disease in univariate analysis, but it was highly colinear with antiviral prophylaxis regimen and was thus excluded from the multivariable model.
Incidence and clinical characteristics of ganciclovir-resistant CMV disease.The incidence of ganciclovir-resistant CMV disease, as stratified by donor and recipient serostatus and disease status, is shown in Figure 2. Although the overall incidence in the cohort was low (1.5%), nearly 13% of cases of symptomatic CMV disease were associated with ganciclovir-resistant strains of CMV. The incidence of ganciclovir-resistant CMV was significantly higher in the D+/R- group than in the R+ group (4 [5%] of 80 vs. 0 [0%] of 153 patients; P = .02). The median time of onset of ganciclovir-resistant CMV disease was 171 days after transplantation (range, 122–232 days). Patients with ganciclovir-susceptible CMV disease had received a median of 69 days of ganciclovir treatment before disease onset (range, 12–130 days), compared with 150 days (range, 124–154) for patients who had ganciclovir-resistant CMV disease. Three of the 4 patients with ganciclovir-resistant CMV disease developed drug resistance during the first episode of CMV disease, and 1 patient developed it during the second episode of CMV disease. Among the 4 patients who developed ganciclovir-resistant CMV disease, 1 died, 3 had documented tissue-invasive disease, and all underwent substantial CMV-related hospitalization after the diagnosis of ganciclovir-resistant CMV disease, compared with patients who received a diagnosis of ganciclovir-susceptible CMV disease (median duration of hospitalization, 66 vs. 0 days; P < .01) (table 4). The patient who died experienced substantial adverse effects associated with ganciclovir and foscarnet therapy, including renal failure that required dialysis, leukopenia that required the administration of granulocyte colony-stimulating factor, and peripheral neuropathy. The specific mutations that conferred ganciclovir resistance have all been previously well-characterized and were definitively shown to confer resistance to ganciclovir .
Risk factors for mortality during the first year after transplantation.The 1-year survival rate for the cohort was 93%, which is higher than the national average in the United States for all heart transplant recipients (88%) [19, 20]. Multivariate analysis of risk factors associated with mortality at 1 year after transplantation is shown in table 5. Among the variables included in the model, only ischemic heart disease as the indication for heart transplantation and development of CMV disease were independently associated with increased mortality at 1 year.
The major finding of the present study is that CMV disease (including ganciclovir-resistant CMV disease) remains an important clinical problem in heart transplant recipients with D+/R- status, despite the administration of antiviral prophylaxis. The overall incidence of ganciclovir-resistant CMV disease in this cohort of heart transplant recipients (1.5%) is similar to the figure of 2% reported for cohorts of abdominal organ transplants , but it is somewhat lower than the 5%–9% range reported for lung transplant recipients [11,12–13]. With the significant reductions in CMV disease associated with antiviral prophylaxis, the proportion of heart transplant recipients who developed symptomatic CMV infection due to ganciclovir-resistant strains was significant (4 [12.5%] of 32 patients), and clinicians must consider this possibility if they encounter transplant recipients with D+/R- status who do not improve virologically or clinically after receiving full-dose ganciclovir therapy. As found in prior studies, resistance occurred only among patients in the D+/R- group—and, more specifically, only among those in the D+/R- group who had received oral ganciclovir (compared with intravenous ganciclovir) prophylaxis . Recent studies have suggested that a newer formulation of ganciclovir (valganciclovir) has substantially improved bioavailability, reaches higher blood levels, and appears to be associated with a lower incidence of genotypic resistance, compared with oral ganciclovir [5, 21]. On the basis of the findings of our study (that ganciclovir-resistant CMV disease developed only in patients in the D+/R- group who had prolonged oral ganciclovir exposure, high viral loads, and intensive immunosuppression), we speculate that ganciclovir-resistant mutants were selected from high viral loads and expanded during prolonged and incompletely suppressive antiviral drug exposure in patients who had inadequate CMV-specific immunity. Future studies will be necessary to determine the relative contribution of each of these factors to the development of resistance.
In addition to defining the incidence of ganciclovir-resistant CMV disease among heart transplant recipients who received oral ganciclovir prophylaxis, we documented the significant morbidity of this complication by comparing the duration of CMV-related hospitalization among patients with ganciclovir-resistant CMV disease versus those with ganciclovir-susceptible CMV disease. Although only 1 of the 4 patients with ganciclovir-resistant CMV disease died, patients with ganciclovir-resistant CMV disease had a median duration of CMV-related hospitalization of 66 days, compared with a median duration of 0 days for those with disease due to ganciclovir-susceptible CMV strains. In addition, all of the patients who received foscarnet developed significant electrolyte abnormalities, mucositis, and transient renal dysfunction. Also, patient 2 received leflunomide, which did not have any apparent virologic or clinical effect. Given the limited antiviral options for treatment of ganciclovir-resistant CMV disease and their toxicities, strategies to reduce the incidence of this uncommon but serious complication are justified.
Overall, CMV disease developed in 11.7% of the cohort of heart transplant recipients and was significantly more common in D+/R- group than in the R+ group. Among patients in the R+ group who received antiviral prophylaxis with 3 months of oral ganciclovir, the incidence of CMV disease was <3%, and no cases of ganciclovir-resistant CMV disease occurred, suggesting that this prophylaxis strategy is highly effective for patients with R+ status. In contrast, the incidence of CMV disease in the D+/R- group among patients who received 3 months of oral ganciclovir was nearly 26%, and ganciclovir-resistant CMV disease was documented in 4 (6%) of 66 patients. Thus, there is significant room for improvement in the D+/R- group. Potential strategies might include the use of agents with greater bioavailability (i.e., valganciclovir rather than oral ganciclovir), a longer duration of prophylaxis, vaccination of CMV-seronegative recipients, or administration of preemptive therapy. Recent studies compared oral ganciclovir with valganciclovir and have suggested that the 2 agents are comparable . At our institution, we have recently changed to valganciclovir as the prophylactic agent for patients with D+/R- and R+ status. The finding that a 3-month course of oral ganciclovir was associated with lower CMV disease rates than the 28-day intravenous ganciclovir regimen suggests that, perhaps, longer durations of prophylaxis may be even more effective. However, because the route of administration for ganciclovir differed in the 2 regimens (i.e., the oral route for the 3-month regimen and the intravenous route for the 28-day regimen), it is not possible to definitively exclude that this factor might also have been contributory. However, given the cost, potential toxicity, risk for resistance, and potential for an even greater delay in the onset of CMV disease, the strategy of administration of prophylaxis for a longer duration should be specifically studied before it can be recommended. Several CMV vaccine candidates are currently under study and should specifically be assessed in CMV-seronegative organ transplant recipients as a strategy to decrease the incidence of and morbidity associated with CMV disease. In the meantime, heart transplant recipients with D+/R- status should be closely monitored for the possible development of CMV disease after they discontinue prophylaxis so that prompt antiviral therapy can be administered to decrease the risk of CMV-associated complications.
An interesting and unexpected finding was that CMV disease was independently associated with an increased risk of mortality at 1 year after transplantation. This association was significant even when we controlled for a number of variables, and the magnitude of the association was strong (adjusted HR, 6.97; 95% CI, 1.25–38.86). The association between CMV disease and increased mortality was previously reported for liver [16, 22] and kidney  transplant recipients, but to our knowledge, it has not been previously reported for heart transplant recipients . One possible mechanism may be through the indirect effect of CMV infection on the development of transplant cardiac allograft vasculopathy , which has been associated with late mortality in heart transplant recipients [24, 26]. Other mechanisms are also possible, but we were not able to define them further in this study. Given the retrospective nature of this study, the limited number of clinical events, and the inability to include other variables that may have influenced mortality for these patients, we consider the association of CMV disease with mortality in our study to be an interesting and important finding that needs to be confirmed by future studies.
The strengths of the study include the relatively large size of the cohort, the use of well-validated definitions of CMV disease, and the comprehensive analyses used. Potential limitations include the retrospective study design, the inclusion of a single health care center, the lack of use of valganciclovir prophylaxis, and the relatively small number of cases of ganciclovir-resistant CMV disease.
In summary, we have demonstrated that CMV disease, including disease caused by ganciclovir-resistant strains, remains an important clinical problem in heart transplant patients with D+/R- status and is associated with significant morbidity. Strategies specifically designed to reduce the incidence and impact of CMV disease in the D+/R- group of heart transplant recipients are needed.
Financial support.University of Washington Phillip A. Fialkow Award (to A.P.L.) and the National Institutes of Health (CA 18029 and CA 15704).
Potential conflicts of interest.A.P.L. has received research support and speaking fees from Roche and Fujisawa. All other authors: no conflicts.