KEYWORDS: lung transplantation; valganciclovir; cytomegalovirus; bronchiolitis obliterans syndrome

Transcription

1 Long-term efficacy and safety of 12 months of valganciclovir prophylaxis compared with 3 months after lung transplantation: A single-center, long-term follow-up analysis from a randomized, controlled cytomegalovirus prevention trial C. Ashley Finlen Copeland, MSW, a W. Austin Davis, BS, a Laurie D. Snyder, MD, a Missy Banks, BS, b Robin Avery, MD, c R. Duane Davis, MD, d and Scott M. Palmer, MD, MHS a,b From the a Department of Medicine, Duke University Medical Center, and b Duke Clinical Research Institute, Durham, North Carolina; the c Department of Medicine, Cleveland Clinic, Cleveland, Ohio; and d Department of Surgery, Duke University Medical Center, Durham, North Carolina. KEYWORDS: lung transplantation; valganciclovir; cytomegalovirus; bronchiolitis obliterans syndrome BACKGROUND: The optimal approach to cytomegalovirus (CMV) prevention after lung transplantation is controversial. We recently completed a prospective, randomized, placebo-controlled study of CMV prevention in lung transplantation that demonstrated the short-term efficacy and safety of extending valganciclovir prophylaxis to 12 months vs 3 months of therapy. In the current analysis, we monitored a single-center subset of patients enrolled in the CMV prevention trial to determine if extended prophylaxis conferred a sustained long-term benefit and to assess its hematologic safety. METHODS: The sub-analysis included 38 randomized patients from Duke University Medical Center. All patients underwent consistent serial serum CMV monitoring and surveillance bronchoscopies. CMV was defined by viremia ( 500 CMV DNA copies/ml) or pneumonitis. The safety assessment included a review of all complete blood counts obtained from transplant onward. RESULTS: During a mean follow-up of 3.9 years in each group, extended-course compared with short-course prophylaxis provided a sustained protective benefit with a lifetime CMV incidence of 12% vs 55%, respectively (hazard ratio, 0.13; 95% confidence interval, ; p 0.009), an effect that persisted after adjustment for clinical risk factors. Patients in each group underwent a comparable number of peripheral blood draws and bronchoscopies. Post-transplant white blood cell, neutrophil, and platelet counts were similar between each treatment group during the course of follow-up. CONCLUSION: Extending valganciclovir prophylaxis to 12 months provides a durable long-term CMV protective benefit compared with short-course therapy, without increasing adverse hematologic effects. J Heart Lung Transplant 2011;30: International Society for Heart and Lung Transplantation. All rights reserved. Reprint requests: Scott M. Palmer, MD, MHS, DUMC Box , Durham, NC Telephone: Fax: address: Palme002@mc.duke.edu /$ -see front matter 2011 International Society for Heart and Lung Transplantation. All rights reserved. doi: /j.healun Cytomegalovirus (CMV) is the most prevalent and important opportunistic infection to occur among lung transplant recipients. Primary manifestations include tissueinvasive CMV pneumonitis (CMV-P) or systemic viral syndrome. In addition to the short-term morbidity of CMV, we recently demonstrated that CMV-P, occurring at any point after lung transplantation and despite treatment, significantly increases the risk for long-term allograft dysfunction, manifest as bronchiolitis obliterans syndrome and is associated with worse overall survival after transplant. 1 Effectively preventing CMV therefore is critical to improving long-term transplant outcomes; however to date, there remains a lack of consensus among lung transplant centers regarding the safest and most effective prophylaxis durations. Results from a 2010 international survey indicated current practice variations ranging from no prophylaxis to a lifetime of prophylaxis, although most centers often used 3 to 6 months of antiviral prophylaxis. 2

2 Copeland et al. Extended Prophylaxis Prevents CMV 991 In an effort to determine the most effective strategy to prevent CMV after lung transplantation, we recently conducted a prospective, randomized, multicenter, doubleblind, placebo-controlled trial comparing 12 months of valganciclovir prophylaxis (extended-course) with 3 months of prophylaxis (short-course). An extended course of prophylaxis significantly reduced the incidence of CMV infection and disease over the first 18 months after lung transplant, without increasing ganciclovir resistance or adverse events. 3 Despite the importance of these findings, the study was designed to describe short-term safety and efficacy of valganciclovir; therefore, whether the observed benefits are sustained over prolonged post-transplant follow-up remains unknown. Concerns have been raised that extended prophylactic courses may merely delay, not prevent, disease while potentially increasing hematologic risks and costs to patients. 4 7 The relevance of CMV reactivation is illustrated in the article by Humar et al, 5 where CMV developed in 45% of the lung transplant recipients who received 3-month courses of prophylaxis after discontinuation of prophylaxis. 5 Although previous studies suggest that most CMV infections occur in the first year after transplantation, these studies are based on cohorts treated with short-courses of prophylaxis; thus, the natural history of CMV reactivation after 12 months of prophylaxis is unknown. Therefore, to determine if extended prophylaxis to 12 months safely confers a long-term durable benefit in the prevention of CMV, we performed additional detailed follow-up in a single-center subset of patients prospectively enrolled and randomized in the multicenter CMV prevention trial. Materials and methods The Duke University Institutional Review Board approved this study (Pro ), and written informed consent was obtained from patients to allow collection of long-term transplant outcomes. Patient population and study design Inclusion criteria for this follow-up analysis consisted of the 38 patients enrolled and randomized in the previous CMV prevention study from Duke University Medical Center. 3 Criteria for enrollment in the original study included adult, first lung recipients at risk for CMV, defined as donor (D) or recipient (R) serology positive (D /R or R ). Full trial inclusion and exclusion criteria, patient screening, enrollment, and randomization study flow are presented in Supplemental Table 1 and Figure 1 (see Appendix). All patients initially received 3 months of valganciclovir prophylaxis (900 mg daily, adjusted for renal function) and then were randomized in a double-blind manner to receive an additional 9 months of prophylaxis (extended-course) or placebo (shortcourse). During the randomized study period, CMV immunoglobulin preparations were not used for prophylaxis, and treatment with other anti-virals was not permitted. Clinical outcomes were extracted from our Duke Lung Transplant Research Database through June 1, Clinical management All patients received clinical care according to the Duke centerspecific lung transplant management protocols, which included triple immunosuppression with tacrolimus, azathioprine, and corticosteroids, as well as induction with basiliximab, as previously described. 8 All patients underwent surveillance bronchoscopies at 1, 3, 6, 9, and 12 months after transplantation, then yearly or at any time as clinically indicated. It is our center s practice that all patients remain under the care of the Duke Lung Transplant team for the lifetime of their follow-up care, as was the case for patients enrolled in this study. Acute rejection ( A1) was first treated with methylprednisone, followed by a prednisone taper. Recurrent or refractory acute rejection was treated primarily with antithymocyte globulin (ATG) or alemtuzumab. Valganciclovir prophylaxis was not used outside of the context of the original randomized study, with the exception of 2 extended-course patients with CMV D /R serology who were continued on valganciclovir after study conclusion ( 1 year after transplant). In the setting of augmented immunosuppression with ATG or alemtuzumab, patients received intravenous (IV) ganciclovir prophylaxis. Patients were seen at least quarterly for pulmonary function testing and routine laboratory screening and received continued follow-up at Duke from transplant onward. CMV monitoring All patients underwent prospective, serial serum and bronchoscopic monitoring for CMV. CMV hybrid capture (before 2006) or plasma polymerase chain reaction (PCR; 2006 onward) was performed weekly for the first 3 months after transplant, and then quarterly to coincide with regular clinic visits or more frequently if clinical indications developed. Every transbronchial biopsy specimen was prospectively stained for CMV by immunohistochemistry, as described previously. 1 Primary and secondary outcomes The primary study outcome was time to (1) CMV viremia, defined as 500 copies/ml of CMV DNA in serum measured by hybrid capture or PCR or (2) CMV pneumonitis, defined as 1 positive CMV-infected cells on immunohistochemistry staining of lung tissue. The date of the first CMV event was recorded for time-toevent analysis. The date of the last CMV screening test was used for a patient s CMV censor date; if CMV had not developed up to that point, patients were censored. Because our center-specific clinical protocol initiates treatment with IV ganciclovir for CMV viremia, defined as 500 copies/ml of CMV DNA by serum PCR, irrespective of clinical symptoms, we did not attempt to distinguish CMV syndrome from viremia. Secondary outcomes included absolute white blood cell (WBC), neutrophil, and platelet counts examined from every peripheral blood draw obtained from randomization onward to evaluate whether extended prophylaxis affected hematologic values long-term. Severe neutropenia and severe thrombocytopenia were defined as absolute counts /liter and /liter, respectively. Statistical analysis Descriptive statistics were used for recipient demographics and include median and interquartile range (IQR). Continuous variables

3 992 The Journal of Heart and Lung Transplantation, Vol 30, No 9, September 2011 were analyzed using non-parametric t-tests, and dichotomous variables were analyzed using chi-square and Fisher s exact test, as appropriate. All randomly assigned patients were included and analyzed on an intent-to-treat basis (ie, patients were analyzed according to the group to which they were randomized). The primary outcome of CMV pneumonitis or viremia was analyzed first as unadjusted Kaplan-Meier rates and then with univariate and bivariate Cox proportional hazards models with prespecified covariates. Hazard ratios (HR) are presented with 95% confidence intervals (CI). The proportional hazard assumptions were verified. Time-to-event analysis began at study randomization and ended at the last known follow-up before June 1, All statistical analyses were performed using SAS 9.2 software (SAS Institute, Cary, NC). Results Recipient characteristics Table 1 Baseline Demographic and Clinical Characteristics Before Randomization of the Extended-Course and Short- Course Groups Baseline Extended-course Short-course characteristics a (n 19) (n 19) p-value Female Sex 11 (58) 7 (37) 0.19 Caucasian Race 16 (84) 18 (95) 0.6 Age at transplant, 60 (30 63) 53 (37 63) 0.62 years Native disease Obstructive 8 (42) 9 (47) 0.74 Cystic 5 (26) 3 (16) 0.43 Restrictive 5 (26) 7 (37) 0.49 Other 1 (5) 0 (0) 0.99 Bilateral transplant 18 (95) 19 (100) 0.99 CMV serology D /R 6 (32) 3 (16) 0.45 Antithymocyte 1 (5) 2 (11) 0.99 globulin Incidence of acute 13 (68) 11 (58) 0.5 rejection BOS incidence b 6 (38) 6 (43) c 0.76 Mortality incidence b 7 (41) 5 (35) 0.57 BOS, bronchiolitis obliterans syndrome; CMV, cytomegalovirus; D, donor; R, recipient. a Categoric data are given as number (%), and continuous data as median (interquartile range). b Kaplan-Meier survival estimates. c One patient was excluded from analysis due to insufficient number of pulmonary function tests to assess BOS. Baseline demographic and clinical features of the extendedand short-course groups were similar before randomization (Table 1). Most patients were white, and the most common indications for transplant were chronic obstructive pulmonary disease, idiopathic pulmonary fibrosis, or cystic fibrosis. Reflective of our center s practice, all but 1 patient received bilateral lung transplant. D /R patients comprised 32% of the extended-course group and 16% of the short-course group. The incidence of acute rejection was 68% in the extended-course group and 58% in the shortcourse group. ATG was given to 1 patient in the extendedcourse group and to 2 patients in the short-course group. No patients received alemtuzumab before randomization. Mean follow-up was 3.9 years in each group. At the conclusion of the study, bronchiolitis obliterans syndrome had developed in a comparable proportion of patients in the extended and short-course groups (38% and 43%) in a mean time of 1551 (SE, 147) days and 1239 (SE, 100) days after transplant, respectively. Overall Kaplan-Meier mortality was 41% in the extended-course group vs 35% in the shortcourse group, which was not significantly different. Effect of extended-course valganciclovir on CMV Extended-course prophylaxis was associated with a sustained long-term CMV protective benefit (HR, 0.13; 95% CI, ; p 0.009). Kaplan-Meier CMV cumulative incidence estimates in the extended- and short-course groups were 12% and 55%, respectively (p 0.002; Figure 1). Within the extended-course group, pneumonitis accounted for 2 late-onset CMV events, one of which was concurrent with viremia. The 2 CMV events occurred 52 and 266 days (median, 159 days) after discontinuation of prophylaxis and occurred in 1 D /R and in 1 D /R patient, respectively. Ten late-onset CMV events occurred in the short-course group: pneumonitis comprised 4 events, 3 of which were concurrent with viremia, and isolated viremia accounted for the other 6 events. These 10 CMV events occurred in a median time of 91 days (IQR, ) after discontinuation of prophylaxis and included 9 R patients and 1 D /R patient. Neither group experienced CMV while receiving prophylaxis. Because 2 extended-course patients received additional valganciclovir beyond the first year, we repeated the analysis excluding those 2 patients and obtained nearly identical results, demonstrating a beneficial effect of extended prophylaxis on lifetime risk for CMV, with Kaplan-Meier overall CMV incidence estimates of 13% with extended-course vs 55% with short-course therapy (HR, 0.15; 95% CI, ; p 0.01). CMV screening frequency To ensure that variation in CMV screening did not bias our results, we examined the median number of peripheral blood draws and transbronchial biopsy specimens obtained over the entire follow-up period for patients in each group. A comparable median number of peripheral blood draws analyzed for CMV infection occurred in the extendedcourse and short-course groups (14 vs 16, respectively, p 0.27). A comparable number of transbronchial biopsy specimens were analyzed for CMV pneumonitis in the extendedand short-course groups (10 vs 11, respectively; p 0.50). Further, when sampling was considered exclusively within the first year of randomization, or exclusively after the first year, when clinic visits occur less frequently, there were no significant differences in sampling rates between the ex-

4 Copeland et al. Extended Prophylaxis Prevents CMV 993 Figure 1 Extended-course prophylaxis conferred a durable long-term cytomegalovirus protective benefit. Overall Kaplan-Meier incidence estimates in the extended-course and short-course groups were 12% and 55%, respectively (p 0.002). tended-course and short-course groups for peripheral blood draws or transbronchial biopsies (data not shown). Bivariate analyses To determine if the durable benefit of extended-course prophylaxis persisted independently of other CMV risk factors, we conducted a series of bivariate analyses adjusting for CMV risk status (D /R vs D /R ), time to first acute rejection episode ( A1) after randomization, or time to first administration of ATG after randomization. Acute rejection and ATG were treated as time-dependent covariates, and only events occurring before CMV (or censor date) were considered. The durable benefit of extended-course prophylaxis remained present after bivariate adjustment for CMV risk status, acute rejection, or ATG (Table 2). Equal numbers of patients in both groups used alemtuzumab after randomization (4 patients in each group). Because CMV did not develop in any of these 8 patients, bivariate adjustment did not change the original model results (data not shown). Table 2 Bivariate Models Show the Protective Effect of Extended-Course Prophylaxis Bivariate models HR (95% CI) p-value Model 1 Extended-course 0.14 ( ) 0.01 CMV serology D /R 0.76 ( ) 0.72 Model 2 Extended-course 0.13 ( ) Acute rejection a 1.25 ( ) 0.72 Model 3 Extended-course 0.12 ( ) Antithymocyte gobulin a 3.34 ( ) 0.06 CI, confidence interval; CMV, cytomegalovirus; D, donor; HR, hazard ratio; R, recipient. a Time to first event occurring after randomization. Long-term hematologic safety of extended-course treatment Finally, we considered the effect of extended-course prophylaxis on absolute WBC, neutrophil, and platelet counts. When all measured values were considered within the first year of randomization (time when the extended-course group received the study drug) or beyond the first year, there were no significant differences in the absolute median WBC, neutrophil, or platelet counts between the groups (Table 3). Furthermore, the incidence of severe neutropenia and thrombocytopenia was infrequent and similar among both groups within and beyond the first year of randomization (Table 3). The frequency with which complete blood counts were obtained during the follow-up period was comparable between both the extended-course and short-course groups (44 vs 52, respectively; p 0.27). Discussion Effective strategies to prevent CMV are essential to minimizing its serious direct and indirect consequences and improving long-term transplant outcomes. Current approaches vary widely and primarily include 3- to 6-month post-transplant courses of IV or oral ganciclovir, valganciclovir, or CMV hyperimmune globulin. 2,9 11 A recent international survey, however, found that extending prophylaxis to 12 months is rare in clinical practice. 2 We recently completed a prospective, multicenter, randomized, placebocontrolled trial of CMV prevention in lung transplantation and demonstrated that extending prophylaxis to 1 year significantly reduced the incidence of CMV infection and disease during the first 18 months after transplant, without increasing ganciclovir resistance or adverse events compared with a shorter course of treatment. 3 In the current analysis, we leverage the previous randomized, prospective, double-blind clinical trial design and pa-

5 994 The Journal of Heart and Lung Transplantation, Vol 30, No 9, September 2011 Table 3 Median Complete Blood Counts and the Incidence of Laboratory Abnormalities Between the Extended-Course and Short- Course Groups Variable Year 1 Beyond Year 1 Extended-course Short-course p-value Extended-course Short-course p-value Count a White blood cells 5.2 ( ) 5.1 ( ) ( ) 5.2 ( ) 0.69 Absolute neutrophils 3.9 ( ) 4.2 ( ) ( ) 3.8 ( ) 0.73 Platelets 208 ( ) 267 ( ) ( ) 241 ( ) 0.17 Severe neutropenia b 1/18 (6) 1/19 (5) /17 (6) 2/18 (11) 1.00 Severe thrombocytopenia c 1/19 (5) 2/19 (11) /18 (17) 6/18 (33) 0.44 a Median (interquartile range) 10 9 /liter. b Number (%) incidence /liter. c Number (%) icidence /liter. tient population to demonstrate that amongst a single-center cohort, extended prophylaxis confers a sustained, long-term protective benefit for CMV prevention. During a mean follow-up of 3.9 post-transplant years, the rate of CMV events after discontinuation of prophylaxis (also known as late-onset disease) among the extended course group was 12% compared with the 55% rate observed in patients who received shortcourse therapy. Furthermore, we demonstrate that extended prophylaxis does not increase the risk for long-term hematologic abnormalities compared with short-course therapy. This analysis has a number of notable features: First, the subset of patients for analysis was drawn from a clinical trial in which patients were randomly assigned in a prospective, double-blind manner to 3 months of valganciclovir, followed by 9 months of placebo, or to 12 months of valganciclovir. The randomized, double-blind study design served to minimize potential bias, and contrasts with previous retrospective studies of CMV prevention in lung transplantation that included sequential prophylaxis strategies among historical cohorts. 12,13 In fact, the effectiveness of randomization is illustrated in the baseline similarity of the extended vs short-course patients even within this single-center subset. Second, by focusing on a single-center subset, similar care was provided to all patients, which was particularly evident in the comparable frequency of bronchoscopic and serum monitoring during the extended follow-up, making it unlikely that sampling differences influenced results. Finally, the analysis included a mean follow-up of 3.9 years after transplant in all patients, making it very unlikely that any patients with late-onset disease were missed, even if extended prophylaxis were to delay the natural history of CMV reactivation compared with the shorter course of treatment. Although our current study suggests extending therapy to 12 months is beneficial in lung transplantation, there are also several limitations to our analysis: First, this analysis included only a subset of patients from the original randomized study. Because of concern regarding heterogeneity among clinical practices after study completion at each of the enrolling sites, we determined the single-center approach was the most scientifically meaningful and feasible for a long-term follow-up analysis. However, this approach limits the generalizability of study results, and it is possible that center-specific practices influenced these results. Second, precisely because of this single-center approach, an additional limitation of our study is the small sample size. Although the sample size is appropriate to address questions regarding the development of CMV, only much larger, appropriately powered studies could address the effect of extended prophylaxis on other long-term clinical outcomes such as bronchiolitis obliterans syndrome or long-term survival. Our small sample size also limits our power to detect differences in adverse events amongst the treatment groups. Third, death as a competing risk could bias the development of CMV in either group. To address this possibility, we considered time to death in both treatment groups. Among the 26 patients who were censored, 8 died within a median of 2.8 years (IQR, years). Two of these patients died early after transplant, 1 in the extended-course group at 269 days and the other in the short-course group at 147 days. All other patients survived a median of 3.6 years (IQR, ), making it unlikely that any additional CMV would have developed after these patients had died. Finally, another limitation relates to the original study design and duration of prophylaxis in the control group. Although 3 months was standard practice at the time the valganciclovir study was initiated, clinical practice has evolved to include longer durations of prophylaxis, which at many centers is up to 6 months. 2 Despite this tendency toward increased prophylaxis to 6 months, no prospective randomized studies support this practice. One study suggested that valganciclovir prophylaxis courses extended through 180, 270, or 365 days postoperatively were all associated with low rates of CMV events during the next 180 days, however, this study was limited by the non-randomized study design, comparison with historical controls, and results were confounded by initial treatment with 30 to 90 days of IV ganciclovir and use of CMV hyperimmune globulin. 12 These factors make it difficult to conclude that 6 months of oral valganciclovir is as effective as 12 months. Collectively, our results are consistent with the hypothesis that CMV develops primarily in the early post-transplant period, perhaps due to impaired recipient CMV-specific immunity, and that over time, post-transplant memory cells or other immune mechanisms emerge that are capable of keeping latent CMV infection in check. In support of this idea, a recent study among lung transplant recipients measured post-transplant

6 Copeland et al. Extended Prophylaxis Prevents CMV 995 CD8 T-cell CMV-specific immunity using human leukocyte antigen restricted peptides derived from CMV phosphoprotein (pp) 65 and immediate early 1 (IE-1) antigen at serial points after transplantation. 14 Although there was considerable variation in the timing and magnitude of CMV-specific immunity, CMV immunity stabilized or improved for most patients with time after transplant. Furthermore, we recently demonstrated the presence of CMV-specific immunity among lung transplant recipients who received valganciclovir prophylaxis. 15 More research is needed to determine if the duration of prophylaxis could be linked to functional assays of CMV specific immunity. The benefits of extended valganciclovir prophylaxis in preventing CMV must be balanced against any adverse consequences. Importantly, in the original clinical trial, we did not observe significant differences in adverse events, serious adverse events, or presumed treatment-related adverse events by treatment group, although the incidence of such events was high in both groups. In this follow-up analysis, we similarly demonstrate no significant hematologic abnormalities over time. The rate of severe neutropenia was not significantly different between groups, similar to results from several previous studies. 3,16 Collectively, our findings suggest that extending valganciclovir to 12 months does not add significant long-term hematologic risk compared with 3 months. In summary, our results demonstrate that a 12-month regimen of oral valganciclovir is a highly effective and safe approach for long-term prevention of CMV in at-risk lung transplant recipients compared with 3 months of therapy. Although alternative durations of prophylaxis might also be beneficial, there are currently only limited data to support such approaches. In the future, an approach in which duration of CMV prophylaxis is tailored to individual CMV-specific immune profiles might be possible. Until such an immune profile is defined and validated in clinical practice, however, our current results, in conjunction with other recent clinical studies, 3,13 provide evidence for 12 months of valganciclovir as the most effective approach to long-term CMV prevention after lung transplantation. Disclosure statement The authors acknowledge the VALGAN Study Committees and members for their contribution and dedication to the study: Steering Committee: S.M. Palmer (Principal Investigator and Study Chair); M. Banks, Project Leader. Independent Data and Safety Monitoring Committee: J. Govert (Chair), A. Smith, M. Vallee. Core Laboratory: Fred Hutchison Cancer Research Center for Molecular Virology, Seattle, WA A.P. Limaye (Director), L. Cook, J. Castor. Study Sites, Primary Investigators (PI), and Study Coordinators (SC): Duke University Medical Center, Durham, NC R.D. Davis (PI), J. Hawkins (SC); Cleveland Clinic Foundation, Cleveland, OH R. Avery and J. Chapman (PIs), S. Lubell and R. Rice (SCs); University of Minnesota, Minneapolis, MN J. Dunitz (PI), L. Schoelkoph (SC); Loyola University, Chicago, IL E. Garrity (PI), K. Kalnicky (SC); University of North Carolina, Chapel Hill, NC R. Aris (PI), J. Mascarella (SC); University of California San Diego, San Diego, CA G.L. Yung (PI), L. Koenig (SC); Vanderbilt University, Nashville, TN A. Milstone (PI); Emory University, Atlanta, GA E.C. Lawrence (PI), J. Gillespi (SC); Ochsner Clinic, New Orleans, LA V. Valentine (PI), C. Thompson (SC); University of Michigan, Ann Arbor, MI K.M. Chan (PI), J. Schmitt (SC); Indiana University, Indianapolis, IN J. Reynolds (PI), T. Isaacs, T. Oakes (SCs). The original VALGAN randomized controlled trial was funded by Roche Pharmaceuticals as an investigator-initiated clinical trial, coordinated by the Duke Clinical Research Institute. All money was paid directly to the Duke Clinical Research Institute for clinical trial operating expenses. Roche did not assist with preparation, writing, or review of original trial manuscript. Roche provided no additional funding for this single center follow up analysis. Roche did not assist with the preparation, writing, or review of this manuscript. Scott M. Palmer is funded by National Heart Lung and Blood Institute SCCOR 1P50-HL and K Laurie D. Snyder is funded by National Institutes of Health KL2RR and American Society of Transplantation Clinical Faculty Development Award. None of the authors has a financial relationship with a commercial entity that has an interest in the subject of the presented manuscript or other conflicts of interest to disclose. Appendix Supplemental Table 1 Eligibility Criteria for Randomization in the VALGAN Trial Inclusion Criteria: Lung transplant recipient 18 years of age At risk for CMV (donor or recipient serology positive for CMV) Negative baseline serum PCR and bronchoscopy results for CMV Prophylaxis with IV ganciclovir 2 weeks immediately postoperatively Adequate hematological, liver, and renal function Able to tolerate oral medications Negative serial post-transplant serum PCR measurements for CMV Negative bronchoscopy for CMV through Day 75 measurement Provide informed consent Exclusion Criteria: Undergoing retransplantation On mechanical ventilation at study entry Previous or current intravenous or oral ganciclovir outside of the study protocol Invasive fungal infection Participating in another investigational study Using disallowed anti-cmv therapy or other prohibited medications Severe diarrhea or malabsorption, liver failure, renal failure, or pregnancy History of severe reaction to ganciclovir Experienced serious adverse event prior to randomization Withdrew consent to be randomized

7 996 The Journal of Heart and Lung Transplantation, Vol 30, No 9, September 2011 Supplemental Figure 1 Patients screened, enrolled, and randomized in the VALGAN trial. References 1. Snyder LD, Finlen Copeland CA, Turbyfill WJ, Howell D, Willner DA, and Palmer SM. Cytomegalovirus pneumonitis is a risk for bronchiolitis obliterans syndrome in lung transplantation. Am J Respir Crit Care Med 2010;181: Zuk DM, Humar A, Weinkauf JG, Lien DC, Nador RG, Kumar D. An international survey of cytomegalovirus management practices in lung transplantation. Transplantation 2010;90: Palmer SM, Limaye AP, Banks M, et al. Extended valganciclovir prophylaxis to prevent cytomegalovirus after lung transplantation: a randomized, controlled trial. Ann Intern Med 2010;152: Limaye AP, Bakthavatsalam R, Kim HW, et al. Late onset cytomegalovirus disease in liver transplant recipients despite antiviral prophylaxis. Transplantation 2004;78: Humar AD, Kumar D, Preiksaitis J, et al. A trial of valganciclovir prophylaxis for cytomegalovirus prevention in lung transplant recipients. Am J Transplant 2005;5: Limaye AP, Bakthavatsalam R, Kim HW, et al. Impact of cytomegalovirus in organ transplant recipients in the era of antiviral prophylaxis. Transplantation 2006;81: Kalil AC, Freifeld AG, Lyden ER, Stoner JA. Valganciclovir for cytomegalovirus prevention in solid organ transplant patients: an evidence-based reassessment of safety and efficacy. PLoS One 2009;4:e Snyder LD, Finlen Copeland CA, Lin SS, Davis RD, Palmer SM. Lung transplantation at Duke University Medical Center. Clinical Transpl 2007: Chmiel C, Speich R, Hofer M, et al. Ganciclovir/valganciclovir prophylaxis decreases cytomegalovirus-related events and bronchiolitis obliterans syndrome after lung transplantation. Clin Infect Dis 2008; 46: Palmer SM, Grinnan DC, Reams BD, Steele MP, Messier RH, Davis RD. Delay of CMV infection in high-risk CMV mismatch lung transplant recipients due to prophylaxis with oral ganciclovir. Clin Transplant 2004;18: Zamora MR. Use of cytomegalovirus immune globulin and ganciclovir for the prevention of cytomegalovirus disease in lung transplantation. Transpl Infect Dis 2001;3 (suppl 2): Zamora MR, Nicolls MR, Hodges TN, et al. Following universal prophylaxis with intravenous ganciclovir and cytomegalovirus immune globulin, valganciclovir is safe and effective for prevention of CMV infection following lung transplantation. Am J Transplant 2004;4: Jaksch P, Zweytick B, Kerschner H, et al. Cytomegalovirus prevention in high-risk lung transplant recipients: comparison of 3- vs 12-month valganciclovir therapy. J Heart Lung Transplant 2009;28: Westall GP, Mifsud NA, Kotsimbos T. Linking CMV serostatus to episodes of CMV reactivation following lung transplantation by measuring CMV-specific CD8 T-cell immunity. Am J Transplant 2008; 8: Snyder LD, Medinas R, Chan C, et al. Polyfunctional cytomegalovirus-specific immunity in lung transplant recipients receiving valganciclovir prophylaxis. Am J Transplant 2011;11: Zafrani L, Truffaut L, Kreis H, et al. Incidence, risk factors and clinical consequences of neutropenia following kidney transplantation: a retrospective study. Am J Transplant 2009;9: