Though an important cause of morbidity and mortality in solid organ transplantation (SOT), the long-term outcomes of cytomegalovirus (CMV) disease treatment have not been well studied. In a randomized trial, 321 SOT recipients with CMV disease were followed 1 year after treatment with either twice daily intravenous ganciclovir or oral valganciclovir (for 21 days) followed by once daily valganciclovir until day 49 in all patients. Clinical and viral eradication of CMV disease was similar between groups. Clinical recurrence beyond day 49 was found in 15.1% and virological recurrence in 30.0%, no difference between groups (p > 0.77). In a multivariable logistic regression analysis, the only independent predictor for recurrence was failure to eradicate DNAemia by day 21 (clinical: OR 3.9 [1.3–11.3], p = 0.012; virological: OR 5.6 [2.5–12.6], p < 0.0001). Eight patients developed ganciclovir resistance, with no difference between groups (p = 0.62). Twenty patients (valganciclovir: 11, ganciclovir: 9, p = 0.82) died, 12 due to infections, two involving CMV disease. There were no differences in long-term outcomes between treatment arms, further supporting the use of oral valganciclovir for treatment of CMV disease. Persistent DNAemia at day 21, CMV IgG serostatus and development of resistance may be relevant factors for further individualization of treatment.
Cytomegalovirus (CMV) disease is a common clinical infection in solid organ transplant (SOT) recipients despite various prophylaxis strategies (1–4). CMV disease generally responds well to antiviral treatment and the short-term efficacy of intravenous ganciclovir and oral valganciclovir is well documented (5–10). However, the long-term outcomes and predictors of recurrence after an episode of treated CMV disease need further investigation (11–18).
The outcome of infection and the risk of recurrence, is dependent on a complex interplay between host and viral factors (19). For example, the nature of the primary infection (CMV IgG serostatus; D+/R–) and the therapy of acute rejection have been shown to predict recurrence (11,15,17). Important virologic factors associated with high recurrence rates include viral load at time of initiation of treatment, longer viral load half-lives and persistent viremia after end of treatment (11,14,17,18).
Antiviral therapy resistance may also be an important determinant of CMV recurrence. Resistance rates following ganciclovir treatment range from below 2% up to almost 10%, with a higher incidence in CMV D+/R– transplants and in lung transplant recipients (20–25).
We have previously shown that oral valganciclovir is noninferior to intravenous ganciclovir for eradication of CMV disease in SOT recipients (26). In this article, we report the follow-up results for 1 year.
Material and Methods
One-year outcomes were investigated following treatment of CMV disease in SOT recipients with either oral valganciclovir (900 mg twice daily) or intravenous ganciclovir (5 mg/kg twice daily) for 21 days, followed by maintenance treatment with valganciclovir (900 mg once daily) in all patients up to day 49, in a randomized (1:1), open-label, parallel-group, active drug-controlled, multicenter and noninferiority trial (26). Valganciclovir and ganciclovir doses were adjusted to renal function (Cockroft and Gault estimation). The study was conducted in accordance with the Declaration of Helsinki, good clinical practice guidelines and applicable local regulatory requirements (ClinicalTrials.gov Identifier: NCT00431353). Forty-two centers from around the world recruited patients in the study.
Any type of adult SOT recipient with both virological and clinical evidence of CMV disease, disregarding CMV serostatus was eligible for inclusion as previously presented (26). Most importantly, patients with life-threatening CMV disease (by the decision of the investigator) or creatinine clearance (Cockroft and Gault) below 10 mL/min were not eligible for inclusion. Figure 1 outlines the patient flow in the study.
Ganciclovir resistance mutations and CMV IgG serostatus were determined at days 0, 21 and 49. Resistance was also analyzed in samples drawn in case of clinical recurrence. Clinical examinations were performed at all visits, also at month 12.
The primary outcome of the study (eradication of CMV plasma DNAemia at day 21) has been reported earlier (26). Secondary outcome measures that have not previously been presented include recurrence (both viral and investigator assessed clinical recurrence), ganciclovir resistance and patient survival over the 12-month follow-up.
Patients assessed in this article were those with positive DNAemia measurements (central analysis) at start of the study, the viral efficacy population. Safety information was, however, done on the intention to treat population (26). Viral recurrence was assessed in patients with proven eradication of CMV DNAemia (two consecutive negative viral loads) by day 49 and defined as positive CMV DNAemia (≥600 copies/mL) occurring after the end of treatment, either at the scheduled (months 3 and 6) visits or upon evidence of clinical recurrence. Patients with clinical disease defined as absent at day 49 were monitored for clinical recurrence up to month 12. Any report of clinically active disease according to the attending physician, regardless of DNAemia values, was considered clinical recurrence. Intensity of recurrent disease was assessed by the local investigator to be; mild, moderate, severe or life-threatening.
Blood samples were drawn in EDTA vacutainers at the following time points: days 0, 3, 7, 10, 14, 17, 21, 28, 35, 42 and 49 as well as months 3 and 6. Samples were shipped on dry-ice to the central laboratory and stored at −70°C.
CMV viral load measurements: Plasma viral loads were measured in all patients at a central laboratory facility, blinded with regards to treatment allocation, using the Amplicor CMV Monitor® Test (600 copies/mL plasma cut-off; Roche Diagnostics, IN).
Investigators were not encouraged to do local viral load measurements, but it was permitted according to the protocol.
CMV IgG serostatus: Donor and recipient CMV IgG serostatus at the time of transplantation was captured from local measurements. Measurement of CMV IgG levels at days 0, 21 and 49 was determined in plasma at a central laboratory using the Architect CMV-IgG kit (Abbott Diagnostic Division, Abbott Park, IL). The lower level of quantification was 6 U/mL and the upper limit was 250 U/mL.
Ganciclovir resistance mutations: Polymerase chain reaction amplification and sequencing analysis of mutations in the UL97 and UL54 loci associated with ganciclovir resistance was used for the determination of ganciclovir resistance as previously described (21,27–32).
Absolute recurrence risk was estimated as number of recurrences over number recovered by day 49 and reported as proportion with the relevant confidence interval according to Brown et al. (33). Missing data were not replaced. Possible differences in absolute risk by initial treatment were tested with common odds-ratio (Cochran–Mantel–Haenszel) and 95% confidence interval (CI) estimates. Time to virologic recurrence was estimated by Kaplan–Meier analysis, in which patients without confirmed virologic and, respectively, clinical recurrence were censored at the time of the last DNAemia or, respectively, clinical observation available.
Potential predictors of recurrence were tested in multivariable binomial logistic regression analysis, monitoring ORs and relevant 95% CI.
Demographic and baseline characteristics
There were no clinical or statistically relevant differences in baseline characteristics between the virological efficacy population consisting of 133 patients in the oral valganciclovir and 126 patients in the intravenous ganciclovir group (Figure 1, Table 1). In the oral valganciclovir group 27 patients did not reach the full 1-year follow-up due to death (N = 10), treatment failure (N = 5), protocol violation (N = 2, use of lamivudine and interferon, respectively), adverse events (N = 5, one with CMV recurrence) and clinical recurrence of CMV (N = 5). Corresponding reasons for the 16 patients in the intravenous ganciclovir group were: death (N = 6; 2 other died shortly after the 1-year observation), treatment failure (N = 4), protocol violation (N = 1, use of lamivudine), withdrawn consent (N = 1), adverse event (N = 1) and CMV disease recurrence (N = 3).
Table 1. Demographic and baseline characteristics of the virological efficacy population
Valganciclovir (N = 133)
Ganciclovir (N = 126)
1chi-square. 2t-test. 3Mann–Whitney test.
Sex N (%)
Ethnicity N (%)
Mean ± SD
47.5 ± 13.7
44.4 ± 14.1
Mean ± SD
67.9 ± 15.5
68.0 ± 16.8
Transplanted organ N (%)
HLA-A mismatches N (%)
1 or 2 mismatches
HLA-B mismatches N (%)
1 or 2 mismatches
Mean ± SD
248 ± 920
245 ± 594
Baseline viral load (copies/mL plasma)
Mean ± SD
58984 ± 106590
53507 ± 77614
Clinical diagnosis N (%)
Tissue invasive CMV disease
Type of Tissue invasive CMV disease N (% of all patients)
Mean ± SD
No. of patients > 38°C
Previous anti-CMV strategy N (%)
Previous anti-CMV therapy N (%)
Donor/Recipient CMV IgG serostatus at time of transplantation N (%)
Evaluation of clinical and virological disease at follow-up
There were no significant differences between the valganciclovir and ganciclovir group in success rates as number of patients experiencing clinical CMV disease or plasma DNAemia were similar at all time-points during the 12-month follow-up period (Table 2).
Table 2. Prevalence of CMV disease and plasma DNAemia (>600 copies/mL plasma) during the study period for the two different treatment groups (virological efficacy population), oral valganciclovir (valganciclovir) and intravenous ganciclovir (ganciclovir)
Disease (both groups together)
(n = 247)
(n = 235)
(n = 229)
(n = 222)
(n = 216)
(n = 124)
(n = 118)
(n = 112)
(n = 107)
(n = 106)
(n = 123)
(n = 117)
(n = 117)
(n = 115)
(n = 110)
DNAemia (both groups together)
(n = 259)
(n = 259)
(n = 222)
(n = 199)
(n = 133)
(n = 133)
(n = 106)
(n = 97)
(n = 126)
(n = 126)
(n = 116)
(n = 102)
Central analysis of CMV IgG levels showed no difference between the two treatment arms at days 0, 21 or 49 (p = 0.31). The overall number of patients with positive CMV IgG serostatus increased during the treatment period; day 0 (214/246; 87.0%), day 21 (237/240; 98.8%) and day 49 (228/230; 99.1%; p < 0.001, Cochran's test).
During the treatment phase (up to day 49), five (3.6%) new-onset ganciclovir-resistance mutations appeared in the valganciclovir group and three (2.3%) appeared in the ganciclovir group (p = 0.51). The development of resistance during treatment was associated with failure to eradicate CMV DNAemia in plasma both at day 21 and at day 49. A detailed presentation of the resistance data are to be published elsewhere Boivin G et al. Cytomegalovirus (CMV) resistance insolid organ transplant recipients with CMV disease treated withintravenous ganciclovir or oral valganciclovir, submitted.
The absolute risk of recurrent CMV disease was 15.1% (95% CI: 11.1% to 20.5%) (n = 34/225) during the follow-up period. The risk of recurrent CMV disease was independent of initial treatment given: 14.8% (95% CI: 9.6% to 22.8%) among those who received oral valganciclovir and 15.5% (95% CI: 10.0% to 23.8%, p = 0.89) among those with intravenous ganciclovir. Two of the patients in the valganciclovir arm and one in the ganciclovir arm had not eradicated the virus by day 49. Recurrence was not different between patients primarily presenting a CMV syndrome or a tissue invasive disease (p = 0.35). Neither the intensity of recurrent CMV disease (p = 0.58) nor the type (CMV syndrome vs. tissue-invasive disease, p = 0.25) differed significantly between the treatment groups. Overall equal number of patients presented with CMV syndrome and tissue-invasive disease at recurrence.
As shown in Figure 2, patients initially treated with valganciclovir had recurrent CMV disease 104 ± 39 days after treatment and those randomized to ganciclovir after 143 ± 72 days (p = 0.22).
Sixty-six out of 220 eradicated patients (30.0%[95% CI: 24.5% to 36.7%]) had virologic recurrence (Figure 3). The type of antiviral treatment did not affect the virologic recurrence; valganciclovir, 30.9% (95% CI: 23.4% to 40.8%) and ganciclovir, 29.1% (95% CI: 21.8% to 38.8%, p = 0.77). Time to virologic recurrence was not significantly different in patients initially treated with valganciclovir (97 ± 22 days) compared with those given ganciclovir treatment (107 ± 33 days, p = 0.20, t-test).
Predictive factors for recurrence
Univariate analyses of factors predictive of recurrence revealed that the type of organ transplant (p = 0.028), persistent CMV DNAemia in plasma at day 21 (p = 0.001) and negative CMV IgG serostatus at start of treatment (p = 0.039) were significant predictors for recurrent CMV disease. The rate of recurrent CMV disease for lung transplant recipients was 38.5% compared to 14.6% for kidney, 11.8% for heart and 0% for liver transplant recipients. The absolute yearly risk of recurrent CMV disease was 24.4% (95% CI: 16.9 to 35.2%) for patients with persistent CMV DNAemia in plasma at day 21 versus 8.8% (95% CI: 5.3% to 14.7%) for those eradicated at day 21. Eight of 29 patients (27.6%, 95% CI: 15.6% to 48.6%) with negative CMV IgG serostatus at start of treatment had recurrent CMV disease compared with 26/200 (13.0%, 95% CI: 9.1% to 18.5%) of patients with positive CMV IgG serostatus.
In a multivariable logistic regression analysis including eradication of CMV DNAemia in plasma at day 21 (yes/no), eradication of CMV DNAemia in plasma at day 49 (yes/no), baseline viral load (copies/mL plasma), baseline CMV IgG serostatus (positive/negative), transplanted organ and treatment arm (valganciclovir/ganciclovir) the only independent predictor for recurrent CMV disease was viral eradication at day 21 (p < 0.0001). Subjects classified as failing to eradicate CMV DNAemia in plasma at day 21 had an odds ratio of experiencing recurrent CMV disease of 3.9 (95% CI: 1.3 to 11.3) and virologic recurrence of 5.6 (95% CI: 2.5 to 12.6). Patients who failed eradicate DNAemia by day 21 also suffered recurrence significantly earlier as shown in Figure 4.
Treatment of recurrent CMV disease
Treatment of recurrent CMV disease was not given in the protocol, but the investigators chose valganciclovir (52.6%) or intravenous ganciclovir (47.4%) as treatment in all cases of clinical disease. Two patients also received additional foscarnet.
Safety and adverse events
Adverse events in general during the study treatment period have been reported previously (26).
Patient survival was analyzed in the intention-to-treat population. During or immediately after the 1-year follow-up, 20 of the 321 patients died, including two who had been excluded from the efficacy population (absolute risk: 6.2%, 95% CI: 4.1% to 9.5%). There was no significant difference between the two treatment arms: 11/164 (6.7%; 95% CI: 3.8% to 11.7%) in the valganciclovir group and 9/157 (5.7%, 95% CI: 3.1% to 10.7%) in the ganciclovir group (P = 0.72). Twelve of the deaths were due to infections two of them involving CMV disease, one in each arm. None of the deaths were considered to be related to the study drug. The median time to death was 78 days (range 4–396 days).
Univariate analyses of predictive factors for death in this population revealed that failure to achieve eradication of CMV DNAemia in plasma by day 49 was significant (OR: 2.9; 95% CI: 1.2 to 7.2; p = 0.024; absolute risk: 10.9% vs. 4.1%). Other significant predictors for death were transplanted organ (p = 0.003), geographic area (p = 0.043) and ethnicity (p = 0.001). Lung transplant recipients had the highest absolute risk of 1-year mortality (21.1%) as compared to liver (15.4%), heart (10.0%) and kidney (3.9%) transplant recipients. Asian-Pacific geographic area showed the highest absolute risk of 1-year mortality (12.4%) as compared the others who ranged from 2.8 to 5.1%. Blacks and Orientals also had a higher probability of death during the study than Caucasians and Hispanics.
Renal function in kidney transplanted patients
In the kidney transplant subpopulation there were no significant difference in renal function between the two treatment arms at baseline, day 21, day 49 or at 3 and 6 months. Estimated GFR were; 50.8 ± 23.2 and 48.3 ± 20.2 (p = 0.37), 52.7 ± 22.1 and 54.9 ± 20.3 (p = 0.42), 52.9 ± 21.6 and 52.3 ± 19.3 (p = 0.84), 49.9 ± 19.7 and 52.7 ± 21.0 (p = 0.32) and 53.1 ± 21.5 and 52.5 ± 20.1 (p = 0.84) at the respective time point in the oral valganciclovir and intravenous ganciclovir group, respectively.
Eleven (6.7%) patients in the valganciclovir group and 13 (8.3%) patients in the intravenous ganciclovir group reported an opportunistic infection during the study (relative risk 0.81 (95% CI: 0.37–1.75), p = 0.59). The most frequent pathogens were candida and herpes simplex.
This is the first large randomized, controlled trial comparing oral valganciclovir and intravenous ganciclovir for the treatment of CMV disease in SOT recipients. As previously presented, initial treatment for 21 days with oral valganciclovir (900 mg twice daily) was noninferior to intravenous ganciclovir (5 mg/kg twice daily) in achieving eradication of CMV disease (26). The long-term follow-up presented here reveals that initial treatment for 21 days with either drug followed by 28 days of once daily oral valganciclovir (900 mg) maintenance, results in comparable treatment success rates, CMV recurrence rates and a low incidence of ganciclovir resistance. These long-term results further substantiate the initial report that oral valganciclovir is appropriate first-line treatment for SOT recipients with nonlife-threatening CMV disease. Applying valganciclovir to pediatric patients should be done with care due to the lack of data. A strategy for monitoring both renal function and hematologic parameters throughout treatment is probably advisable to provide patients with acceptable standard of care.
We found a virological recurrence rate of 30% that is in line with earlier reports ranging from 25% to 30% (11,13–15,17,18). The absolute risk of virological recurrence was twice that of clinical recurrence of CMV disease in the present study. DNAemia in plasma was only assessed cross-sectionally, so it is likely that the rate of virological recurrence would have been even higher if there had been more frequent assessment. The study design did not allow assessment of the effect of asymptomatic CMV DNAemia in plasma after the end of treatment. Interestingly, few patients with plasma DNAemia at the predefined time-points developed recurrent CMV disease or were retreated (samples were analyzed for plasma DNAemia centrally, after end of study). This leads one to question the need for monitoring after eradication. The present finding that patients failing to eradicate CMV DNAemia in plasma by day 21 have an almost fourfold risk of recurrent CMV disease and a higher risk for virological recurrence supports previous findings (14,18). It is probably good clinical practice to monitor viral loads weekly and to continue treatment until eradication is achieved. Development of resistance was also associated with reduced eradication of plasma DNAemia. In univariate analyses lung transplant recipients and patients with primary CMV disease were significantly more likely to develop recurrence. Turgeon and colleagues have also shown that primary infection is an independent risk factor for recurrence (11). Factors such as multiorgan transplantation, longer half-lives of viral decay, CMV pneumonia, treatment for acute rejection episodes and deceased donor transplantation that have been shown previously to be risk factors for recurrence, were not confirmed in this study (13–15).
The frequency of new-onset ganciclovir resistance during the treatment was low (<3%) in this large study with no difference between treatment groups. These results were, however, analyzed centrally at a later time and hence not available to the investigators at the time of treatment and did therefore not influence actual patient treatment.
The absolute 1-year risk of death was near 6% in this study. Patients still plasma DNAemia positive at day 49 had an almost threefold higher risk of death within 1 year. A higher risk of mortality was in addition seen for lung transplant recipients, patients treated in the Asian-Pacific area and in Black/Orientals. It is not clear from the present data why these factors should represent higher risk. Lung transplants in general have a worse prognosis, but many factors, including socioeconomic factors and health services might be involved in the higher risk between different regions and ethnicities. It should be emphasized that these results should be interpreted with care due to the relatively low death rate.
Although this study included a large number of patients, about 70% were kidney transplant recipients and the number of other organ transplants was lower. This may limit the validity of the overall results with regards to liver, heart and lung transplants. However, the absolute numbers of nonrenal organ transplant recipients in this study are higher than in most previous studies with a proper design (11,13,14,17,18). It should also be kept in mind that patients who were considered by the investigators to have life-threatening CMV disease were excluded from the study and one should not extrapolate the results to such patients.
In conclusion, there were no differences in long-term outcomes in transplant patients treated with oral valganciclovir or intravenous ganciclovir for CMV disease. This further supports the use of oral valganciclovir for treatment of CMV disease in SOT recipients. Clinicians should, however, remain alert to the risk of recurrent CMV disease in treated patients. Recurrence rates in patients with persistent DNAemia in plasma at day 21 are high and this subgroup may require prolonged initial treatment therapy. Monitoring of weekly viral loads until viral eradication is recommended. Negative CMV IgG serostatus at start of treatment and development of resistance may also be relevant factors to consider when individualizing treatment.
The authors would like to acknowledge the patients, contributing investigators and other study personnel for their work that made the study possible. Professor Guy Boivins’ lab is acknowledged for the determination of ganciclovir resistant mutations.
F. Hoffmann-La Roche Ltd. funded the study and local affiliates of F. Hoffmann-La Roche Ltd. were responsible for monitoring the study in most centers while local CROs were hired for monitoring of some centers. Hyperphar was subcontracted by F. Hoffmann-La Roche Ltd. to be the central CRO and study leader, also taking care of the electronic CRF program, data management and statistical analysis.
Conflict of Interest Statement
Dr. Åsberg is a former employee of F. Hoffmann-La Roche Ltd. (between 2000 and 2003) and served as a consultant to F. Hoffmann-La Roche Ltd. during the study dealing with study specific issues. Drs. Humar, Pescovitz, Jardine and Hartmann have performed consultancy work for F. Hoffmann—La Roche Ltd. Drs. Rollag, Töz and Montejo have no conflicts of interest to report. Dr. Mouas was until recently (April 2008) an employee of F. Hoffmann-La Roche Ltd. Dr. Bignamini was employed by Hyperphar during the study, which was a subcontractor for F. Hoffmann-La Roche Ltd., hired to perform data management and statistical analyses. Dr. Dittmer has received educational grants from F. Hoffmann-La Roche Ltd.
Appendix: Members of the VICTOR Study Group
North America: Sandra Cockfield. Latin America: Anabela Armino, Luis F.A. Carnargo, Carmen Garcia, Carlos Henriquez, Marilda Mazzali, Elias David Neto, Irene L. Noronha, Jose Osmar Medina Pestana, Rfael Reyes. Asia, Pacific Islands and Australia: Allan Glanville, George T. John, Vijay Kher, R.K. Sharma, C.M. Thiagarjan. Europe and Middle East: Jose Maria Aguado, Emel Akoglu, Hofman Blazenka, Felix Burkhalter, Magdalena Durlik, Antonio Franco Esteve, Marciej Glyda, Abdul Hammad, Rajko Hrvacevic, Madis Ilmoja, Marian Klinger, Dirk Kuypers, Phil Mason, Mai Ots, Patrik Peters, Rafails Rozentals, Boleslaw Rutkowski, Sabine Schmaldienst, Juerg Steiger, M. Tuncer, Zbigniew Wlodarczyk, Michael Zakliczynski.