Is Prevention the Best Treatment? CMV After Lung Transplantation


Scott M. Palmer,


Cytomegalovirus (CMV) is the most prevalent opportunistic infection that occurs in lung-transplant recipients. In addition to its direct morbidity, multiple studies have demonstrated that CMV, in particular CMV pneumonia, is associated with an increased risk for chronic graft dysfunction manifested as bronchiolitis obliterans syndrome (BOS) and worse posttransplant survival. Therefore, prevention of CMV remains an important goal to improve long-term lung-transplant outcomes. Although centers often employed 3 months of prophylaxis in at-risk patients after lung transplantation, a significant proportion of patients still developed infection or disease after the discontinuation of prophylaxis, highlighting the need for more effective approaches to CMV prevention. A number of early single-center reports suggested benefit to extending prophylaxis to longer durations, but concerns regarding cost, late-onset CMV disease, viral resistance and bone marrow toxicity limited enthusiasm for longer durations. However, several recent studies including a multicenter, prospective, randomized, double-blinded clinical trial have demonstrated significant benefits to extending CMV prophylaxis beyond 3 months. Although some areas of controversy remain, the clinical implications of these recent studies suggest that extending prophylaxis with valganciclovir up to 12 months is clearly beneficial for CMV prevention after lung transplantation.


bronchoalveolar lavage fluid


bronchiolitis obliterans syndrome




CMV pneumonia

D+/R–, donor positive

recipient negative


human leukocyte antigen


quality of life


Cytomegalovirus (CMV) is the most prevalent opportunistic infection that occurs after human lung transplantation (1). Common manifestations of CMV disease in lung-transplant recipients include pneumonitis or viral syndrome, and less frequently gastrointestinal disease (1). Modern diagnostic tests have also made it possible for the early identification of CMV infection in the form of asymptomatic viral replication in the peripheral blood. With the advent of ganciclovir treatment, direct CMV morbidity and mortality have been dramatically reduced. However, concerns remain regarding the indirect consequences of CMV, particularly the association with an increased risk for chronic allograft dysfunction manifested as bronchiolitis obliterans syndrome (BOS; Ref. 1).

In an effort to limit CMV and graft dysfunction, the use of antiviral prophylaxis with ganciclovir or valganciclovir has become standard at most lung-transplant centers. A recent international survey regarding prophylactic strategies reported that the majority of at-risk patients receive prophylaxis for 3–6 months after transplantation, whereas a small minority receive no treatment or durations longer than 6 months (2). Although a few single-center reports have described the initial experience with extending ganciclovir to 12 months or longer, there has been considerable hesitation to incorporate longer durations of prophylaxis into clinical practice (3–5). In particular, concerns regarding cost, delayed onset of CMV disease, viral resistance and bone marrow toxicity has limited enthusiasm for longer courses of antiviral prophylaxis (6,7).

In contrast, a major limitation of short-course prophylaxis is that almost half of at-risk patients develop CMV infection or disease shortly after the discontinuation of antiviral prophylaxis (8). Therefore, more effective approaches to CMV prevention are needed to improve overall transplant outcomes. In this minireview, we highlight several recent studies, including results from a prospective, multicenter, double-blinded study of CMV prevention comparing the effectiveness of different durations of valganciclovir prophylaxis, to formulate an evidence-based approach to CMV prevention after lung transplantation.

Why Is Prevention of CMV Important?

Long-term survival after lung transplantation is limited by BOS, a condition of chronic airflow obstruction associated with progressive airway fibrosis. By 5 years posttransplant, almost 50% of patients meet criteria for this condition (9). Once BOS ensues, quality of life (QOL) is diminished and median survival is less than 3 years (10). Although the precise biological mechanisms that lead to BOS are uncertain, it seems to be a manifestation of chronic allograft rejection. Consistent with that idea, the severity and frequency of prior acute rejection or lymphocytic bronchiolitis are the strongest predictors for the development of BOS. In addition to cellular rejection, a number of other potential risk factors for BOS have been identified that include CMV infection, respiratory viral infection, fungal infection, gastric reflux and the development of donor-specific anti-human leukocyte antigen (HLA) antibodies (11).

Among these potential BOS risk factors, CMV has remained one of the most controversial, particularly with the development of more effective CMV prophylaxis and treatment. In a systematic review of BOS risk factors, Sharples et al. identified a number of studies that correlated CMV with BOS (12). Study results, however, were inconsistent likely because of variable CMV definitions, small sample sizes, varying CMV prophylaxis regimens and methodological issues such as the failure to adjust for the time-dependent nature of many predictor and outcome variables. More recently, treated CMV pneumonia (CMV-P) was not found to be predictive of the development of BOS or worse survival in a large cohort of Australian lung and heart–lung recipients (13). Despite many strengths of the Tamm study, the analysis suffered from a number of limitations. The study included patients transplanted over a 14-year period during which the diagnosis, prophylaxis and treatment of CMV changed (13). In addition, the study diagnosed CMV-P based upon viral cytopathic effect creating the potential for misclassification. Most importantly, the study failed to properly adjust for the time-dependent nature of CMV as a BOS risk factor, thus potentially biasing results and obscuring the true effect of CMV-P on BOS or survival (13).

In contrast, recent studies have highlighted the continued importance of CMV and its indirect consequences for lung-transplant recipients despite effective antiviral prophylaxis or treatment. Snyder et al. recently demonstrated that posttransplant CMV-P led to an increased risk for BOS (HR 2.19; 95% CI 1.36–3.51) and worse posttransplant survival (HR 1.89; 95% CI 1.11–3.23) in a modern cohort of 231 lung-transplant recipients (11). In this study, patients received a short course of intravenous (IV) ganciclovir prophylaxis after transplantation, CMV-P was diagnosed using immunostaining prospectively performed on every biopsy sample (>1800 samples reviewed in total), and results were analyzed using multiple statistical approaches that considered the time-dependent nature of CMV and BOS. Similarly, a translational study by Weigt et al. also confirmed the relationship between pulmonary CMV infection and BOS as well as provided new insights into the potential mechanisms by which CMV might lead to BOS (14). The group confirmed the findings that earlier BOS and diminished survival were significantly worse in patients with pulmonary CMV infection or disease as compared to those without CMV (14). In addition, CC chemokines in bronchoalveolar lavage fluid (BALF) of patients with CMV infection had significantly increased levels of CCL2/MCP-1 and CCL5/RANTES as compared to healthy transplant controls, with highest levels in the setting of CMV-P (14). These novel observations provide a potential link between pulmonary CMV and BOS as both CCL2 and CCL5 can recruit and activate immune cells and their elevation in the BALF has been previously reported as a biomarker for BOS (14). In addition, another recent study found elevations of serum CXCL-10 (IP10) correlated with CMV replication in the BALF or blood of lung-transplant recipients. Furthermore, higher elevations of CXCL-10 were correlated with a greater decline in lung function (15). Although the role of CXCL-10 in BOS is controversial, human and murine studies have also implicated this as an important chemokine in the development of BOS. Thus, several independent studies now provide strong evidence that CMV replication, particularly CMV-P, is a significant risk factor for BOS, despite modern prophylaxis and treatment, and that this effect could be mediated by local and systemic upregulation of specific proinflammatory chemokines.

What Is the Optimal Approach to Prevent CMV After Lung Transplantation?

Initially, lung transplant CMV prophylaxis consisted of IV or oral ganciclovir with or without CMV hyperimmune globulin (CMV IVIg) and the duration was typically 1–3 months posttransplant (1). However, with the availability of valganciclovir and its comparable bioavailability to IV ganciclovir, many centers switched to oral valganciclovir as standard prophylaxis. As confirmed by a recent international survey, the most commonly used approaches to CMV prevention after lung transplantation included IV ganciclovir followed by oral valganciclovir or solely valganciclovir for duration of 3–6 months. This practice is also consistent with the recently updated American Society of Transplant infectious disease guidelines that recommend 3–6 months of prophylaxis in at-risk lung-transplant recipients, favoring 6 months in high-risk donor positive, recipient negative (D+/R–) patients (16). These prophylactic strategies and recommendations, however, are primarily adapted from other organs because of the limited primary data in lung transplantation. This extension of results from other solid organs may not be appropriate to apply to lung recipients because of their intense immunosuppression, higher rates of CMV in at-risk patients and pneumonia as the primary manifestation of disease (11,17).

Several reports have suggested that extending prophylaxis in lung-transplant recipients beyond 3–6 months is both feasible and efficacious. In particular, Palmer et al. described successful indefinite treatment of high-risk D+/R– with oral ganciclovir and Zamora et al. described successful CMV prevention with IV ganciclovir and oral valganciclovir when administrated with CMV IVIg for durations that extended from 180 to 365 days (5,18). Interestingly, no difference was observed in CMV infection or disease with 6, 9 or 12 months of prophylaxis at 180 days after prophylaxis discontinuation in the Zamora study. More recently, Valentine et al. published a single-center experience that demonstrated that indefinite ganciclovir treatment almost completely eliminated CMV-P (4). Although all intriguing results, these single-center, nonrandomized reports using historical controls limited the strength and generalizability of the conclusions. At the same time, early experience with extended valganciclovir prophylaxis in solid organ transplantation also led to several reports raising concerns for an increased risk for late-onset CMV disease, increased ganciclovir resistance and significant bone marrow toxicity, particularly neutropenia (6,7).

In light of these controversies and to better define the optimal approach to CMV prevention after lung transplantation, Palmer et al. conducted a prospective, multicenter, randomized, placebo-controlled clinical trial in which serologically at-risk lung-transplant recipients received 3 months of prophylaxis with valganciclovir and then were randomly assigned in a double-blind manner to either an additional 9 months of prophylaxis (extended course) or placebo (short course; Ref. 3). The study was conducted across 11 US lung-transplant centers and randomized 136 patients with 66 receiving short-course and 70 receiving extended-course treatment (3). All patients were analyzed on intent to treat basis and freedom from CMV was estimated from the time of randomization onward. Overall, there was a significant reduction in CMV disease, the primary study endpoint, in patients who received extended-course as compared to short-course prophylaxis (4% vs. 32%, p < 0.001) at 13 months posttransplantation (3). The secondary analysis also demonstrated a significant reduction in CMV infection with extended prophylaxis as compared to short-course therapy (10% vs. 64%; p = 0.001) as well as a reduction in CMV disease severity (3200 viral copies/mL vs. 110 000 viral copies/mL; p = 0.009; Ref. 3). After adjusting for CMV mismatch status, a significant decrease in CMV disease persisted with extended treatment as compared to short-course treatment (HR 0.09; 95% CI 0.021–0.39; p = 0.001; Ref. 3). In a 6-month poststudy follow-up across all centers, the rates of additional CMV infection and disease remained low in both groups (3% with extended-course and 2% with short-course). Thus, within the context of the first 18 months after lung transplantation, it seems that extending prophylaxis to 12 months compared to 3 months effectively prevents CMV infection or disease (3).

Although this study represents the largest prospective multicenter effort at CMV prevention in lung transplantation, there are a number of limitations to consider. First, the study design does not address the efficacy of intermediate durations of prophylaxis such as 6 months. Second, several important questions remain regarding the impact of extended therapy upon ganciclovir resistance and patient toxicity. The study genotyped every positive CMV isolate by PCR for ganciclovir resistance mutations and found only one known ganciclovir resistance mutation in each group, but longer term resistance testing was not performed. As suggested by numerous studies, the development of resistance seems to be linked to improper dose reduction of oral valganciclovir; the strict dose reductions in this study protocol according to changes in renal function and not leucopenia could have contributed to this low rate of resistance in both groups (3). Also, from a safety point of view, there was a statistically significant reduction in the mean platelet count in patients treated with extended prophylaxis. In general, most other measures of safety including adverse events were similar among patients receiving extended-course as compared to short-course treatment suggesting that extending prophylaxis does not significantly increase patient toxicity (3). Finally, the study included follow-up only through the first 18 months, perhaps not long enough to fully ensure late-onset disease was prevented.

To further clarify the issue of late-onset disease (beyond 18 months) Finlen-Copeland et al. subsequently published a follow-up analysis that considered the lifetime risk for CMV disease or infection in a subset of 38 patients from the largest enrolling center (19 of whom received short-course and 19 of whom received extended-therapy). By confining the analysis to a single center, they ensured comparable serial bronchoscopy, clinical management and CMV monitoring over the course of the analysis. In this study, patients receiving extended-course prophylaxis had a significantly reduced lifetime CMV incidence as compared to those receiving short-course prophylaxis (12% vs. 55%, respectively; HR 0.13; 95% CI 0.03–0.61), an effect that persisted after bivariate adjustments for clinical risk factors and over nearly 4 years of posttransplant follow-up (19). Collectively, these studies demonstrate that 12 months of valganciclovir prophylaxis after lung transplantation is generally safe, well tolerated and significantly more effective at CMV prevention than 3 months.

Several additional recent studies in lung and renal transplant also provide supporting evidence for the idea that extending prophylaxis actually prevents rather than simply delays the onset of CMV disease. In a retrospective study by Jaksch et al. that compared the impact of 3 months versus 12 months of valganciclovir in high-risk (D+/R–) lung-transplant recipients, the incidence of symptomatic CMV disease/syndrome was 44% with short-course and 13% with extended-course therapy, whereas CMV viremia was detected in 75% of short-course patients and 33% at 6 months after valganciclovir cessation (20). Although the incidence of disease and infection are higher than that reported in the CMV prevention study by Palmer, this likely reflects inclusion of only high-risk mismatch patients in the Jaksch study (20). Similarly, in a prospective randomized study of high-risk renal-transplant recipients, the IMPACT investigators found that longer durations of prophylaxis (200 days as compared to 100 days) significantly reduced CMV disease at 1 year posttransplantation, an observation confirmed in a 2-year follow-up analysis (8,21). Although certain aspects of the IMPACT study design, analysis and outcomes have recently been criticized, the study authors have provided a very detailed response that suggests the overall study conclusion is valid (22). Because of the significant differences in immunosuppression, clinical management, CMV rates and disease manifestations, direct extrapolation of the IMPACT results to lung transplantation is not particularly useful. However, it is interesting to note that a similar pattern has now emerged across multiple solid organ transplant populations treated with short as compared to extended prophylaxis durations. In particular, in patients treated with short-course prophylaxis, CMV events occur predominately within 6 months of prophylaxis discontinuation. In addition, few CMV events seem to occur beyond the first year (6,7). These results suggest that the first transplant year is the period of highest CMV risk likely related to the intensity of the immunosuppression or impairment of recipient antiviral immunity (6,7). As such, it seems that prophylaxis extended to longer durations can be safely discontinued after the first year with a low risk for subsequent CMV events.

Conclusions and Future Directions

What can we conclude about CMV prevention in lung transplantation? First, we now have the highest level of evidence from a prospective, multicenter, randomized control trial that 12 months is more effective than 3 months of prophylaxis. Furthermore, we have extended follow-up data from this trial that demonstrates, at least in the subset of patients, the benefits of extended prophylaxis persist over extended follow-up (3,19). Thus, based on the available data, 12 months should be considered as an acceptable approach to CMV prevention in lung transplantation. Similarly, based on these results, 3 months of prophylaxis should be considered inadequate. A more challenging question is the value of intermediate durations of prophylaxis. If intermediate durations were effective, this approach could reduce costs and potential toxicity as compared to 12 months of therapy. Unfortunately, it is not entirely clear whether intermediate durations of prophylaxis will behave more like 3 or 12 months. A single retrospective study by Zamora using CMV IVIg in combination with valganciclovir suggests that 6 or 9 months behave similar to 12 months of prophylaxis. Although we take a conservative approach to CMV prevention that involves 12 months for all at-risk patients, we recognize that durations of 6–9 months could also provide benefit as compared to 3 months. In particular, we note that center-specific variation in immunosuppression, clinical management protocols and patient risk profile will impact upon the observed incidence of CMV, making it difficult to define a single optimal duration of prophylaxis across all centers.

Second, the optimal approach to CMV prevention by serological risk status remains uncertain. For example, shorter durations of prophylaxis than 12 months might be most effective in medium-risk R+ patients whereas durations of 12 months or greater might be most appropriate in high-risk D+/R– patients. Clearly, high-risk patients lack any preexisting immunity to CMV, carry a high risk for disease over the entire posttransplant lifetime as compared to medium-risk patients, where CMV-specific immunity might be transiently suppressed but recover over time. From the Palmer et al. study and that of Jaksch et al., it seems that 1 year of prophylaxis is certainly a reasonable approach in high-risk patients but that a subset will still go on to develop CMV disease after prophylaxis is discontinued (3,20). Given the increased risk for BOS and death associated with CMV in these patients, one could make an argument to extend prophylaxis beyond 1 year in the mismatch patients (3,4). This subset represents an important population for further study and ideal for targeting with CMV vaccines or other approaches to boost CMV-specific immunity.

Third, it should be recognized that whereas longer durations of valganciclovir are very effective at CMV prevention, extended durations also create the potential for increased toxicity. Most commonly, this involves bone marrow suppression. Given the numerous concurrent medications employed in lung transplantation that also can cause leucopenia, frequent monitoring of a patient's blood counts for toxicity is mandatory. When leucopenia occurs, clinical judgment is needed to determine if dose reductions in other medications, addition of filgrastim or continued monitoring is the most appropriate course of action.

Fourth, although there is no clear consensus on posttransplant CMV monitoring, we would suggest that all at-risk patients should undergo frequent CMV monitoring particularly after the discontinuation of prophylaxis. In the study by Palmer et al., the majority of CMV events in the short-course prophylaxis group occurred over the next 6 months, consistent with most prior studies employing shorter durations of prophylaxis (3,17). We, therefore, favor monitoring with serum PCR for CMV at least monthly over the first 3–6 months after discontinuation of prophylaxis. As patients move farther out from transplant, however, CMV is much less frequent and testing based on clinical indications rather than routine surveillance seems appropriate.

Finally, it seems clear the lung-transplant community needs a better understanding of the factors that influence CMV reactivation beyond serological donor and recipient status to develop more effective and individualized prophylactic strategies. This point is highlighted by the fact that some patients who receive only 3 months of prophylaxis never develop CMV; whereas other patients who receive 12 months of prophylaxis still go on to develop CMV. At least one study suggests that the development of CMV could be linked to genetic variation in the interferon gene (23). It is likely that this and other underdetermined genetic factors influence host susceptibility to CMV infection.

The future of CMV prevention rests in our ability to extend the measurement of CMV-specific immunity to a level that is highly predictive of clinical outcomes after lung transplantation. Building upon early work by Bunde et al., the measurement of CMV-specific immunity using HLA-restricted tetramers or IE-1 and pp65 peptides offer the potential to directly measure the number and quality of CMV-specific CD4 or CD8 T cells (24). Although the preliminary work by Bunde et al. suggested a CD8 IE-1 response threshold predicted development of CMV disease, the study included only four lung-transplant patients and used older viral prophylaxis and detection methodology (24). More recent studies with larger numbers of lung-transplant recipients, have extended the feasibility of this approach in lung transplantation (25–27). Based on these studies, it seems there is considerable variation in time and magnitude of the CMV-specific immune response post lung transplant (25–27). Of note, CMV immunity does seem to develop normally among patients on valganciclovir as compared to those off prophylaxis in normal controls, a theoretical concern raised with regards to extended valganciclovir prophylaxis (26). Ultimately, larger prospective studies are needed to determine specific immune cell characteristics, including absolute numbers and functional cytokine production in response to CMV antigenic stimulation, to construct a clinically useful diagnostic test that could be applied to individual patients and assist in the decision to discontinue or extend prophylaxis. A future trial that compared fixed prophylaxis duration to variable durations, individualized based upon the results of serial CMV-specific immune monitoring assays would be an ideal mechanism to move translational research in CMV prevention forward.

Until such assays are developed, current studies clearly demonstrate the benefits of fixed duration of CMV prophylaxis extended up to 1 year posttransplantation. Although this approach does not achieve the ultimate goal of complete CMV prevention in all patients, it represents a significant advance with regard to the previously employed, shorter, 3-month duration of prophylaxis. Therefore, extending prophylaxis beyond this duration should now be standard practice after lung transplantation.


The authors acknowledge funding from the American Society of Transplantation Clinical Faculty Development Award (LS), National Heart Lung and Blood Institute P50 HL107180 (SMP), R34 HL105422–01 (SMP) and K24–091140-01 (SMP). In addition, S.M.P. receives funding from the Biomarker Factory, Roche Organ Transplant Foundation and the Lung Transplant Foundation.


The authors of this manuscript have conflicts of interest to disclose as described by the American Journal of Transplantation. Dr. Palmer's institution previously received funding from Roche to conduct an investigator initiated multicenter study of CMV prevention coordinated by the Duke Clinical Research Institute (3). Roche had no input in the conduct, design, analysis or publication of that study. Roche had no input into the design, content, or publication of this review.