Optimization of Valganciclovir for Treatment of Cytomegalovirus Infection in Pediatric Liver Transplant Recipients: A Prospective Pharmacokinetic Study

Abstract

Background.

Valganciclovir (VGCV) is the first-line therapeutic agent for cytomegalovirus (CMV) infection following pediatric liver transplantation (LT). The dosage for pediatric patients is either determined by body weight (BW dose) or body surface area (BSA) and renal function (7 × BSA [m²] × creatinine clearance (CrCl) [mL/min/1.73 m²]) (BSA dose). However, the optimal dosing strategy for achieving target drug exposure for CMV treatment remains unclear.

Methods.

We conducted a prospective pharmacokinetic study in pediatric LT patients. The area under the concentration-time curve (AUC_0–24) was estimated using previously reported pediatric population pharmacokinetic parameters. An AUC_0–24 of 80–120 µg·h/mL was defined as the therapeutic target. The primary endpoint was to compare the achievement rate of target AUC between BW and BSA doses. Although BW dose is the standard at our institution, a hypothetical BSA dose was estimated proportionally for comparison with BW dose.

Results.

Fourteen patients with a median age of 1.1 y were included. The median VGCV dose was 15.3 mg/kg, whereas the hypothetical median BSA dose was 24.5 mg/kg. The estimated AUC_0–24 in BW and BSA dose groups were as follows: <80 µg·h/mL, 85.7% and 35.7%; 80–120 µg·h/mL, 14.3% and 42.9%; >120 µg·h/mL, 0.0% and 21.4% (P = 0.027). There was a negative correlation between AUC_0–24 and the duration of CMV viremia, indicating that higher AUC values were associated with faster viral clearance (ρ = −0.72, P = 0.003).

Conclusions.

Based on our estimation, BSA dosing may be more likely to achieve the target AUC_0–24 compared with BW dosing in infant/toddler LT recipients. Higher AUC values correlated with faster viral clearance.

Valganciclovir (VGCV) is an L-valyl ester prodrug of ganciclovir (GCV), which is converted into its active form by intestinal and hepatic esterases after oral administration.¹ VGCV is widely used for both the prophylaxis and preemptive therapy of CMV infections in immunocompromised patients, including pediatric liver transplantation (LT) patients.²

The standard dosing regimen of VGCV varies depending on the indication. For prophylaxis, the drug is administered once daily, whereas for preemptive therapy, it is given twice daily.² The pediatric dosage is determined using either of 2 approaches: body weight based (BW dose: 15–18 mg/kg per dose) or body surface area (BSA) and renal function-based (BSA dose: 7 × BSA [m²] × creatinine clearance [CrCl, mL/min/1.73 m²] per dose).² It has been reported that BW dose often results in lower dose compared with BSA dose.³ Although once-daily BSA dose is the FDA-approved regimen for prophylaxis, a multicenter study reported that dosing strategies were not standardized owing to the variety of institutional protocols and individual patient’s condition.⁴

Vaudry et al⁵ reported that a BSA dose of VGCV could achieve optimal GCV exposure in pediatric solid organ transplantation (SOT) patients, comparable to established adult targets. In contrast, Villeneuve et al⁶ demonstrated that in pediatric SOT patients aged 6 mo to 3 y, the BW dose was more likely to achieve appropriate therapeutic drug exposure than the BSA dose. According to these findings, a consensus on the optimal dosing regimen remains unestablished, necessitating further pharmacokinetic (PK) and pharmacodynamic (PD) evaluations in this population.

The area under the concentration-time curve (AUC) has been established as a key PD parameter correlating with both antiviral efficacy and toxicity of GCV.⁷ However, in pediatric LT patients, the relationship between AUC and therapeutic efficacy or adverse effects remains not fully established. Because both underexposure and overexposure can potentially result in unfavorable clinical outcomes, optimizing VGCV dosing in this vulnerable population is critically important.

To address this gap, we conducted a prospective PK study to evaluate the drug exposure levels achieved with BW or BSA dose in pediatric LT patients and to explore the relationship between AUC and antiviral efficacy or toxicity. This study aims to provide insights into the optimal dosing regimen for VGCV treatment in this population.

MATERIALS AND METHODS

Study Design, Setting, and Participants

This is a prospective, PK study at the National Center for Child Health and Development (NCCHD), a tertiary children’s hospital with the largest pediatric transplant center in Japan. We enrolled pediatric patients (<18 y old) who had undergone LT and were receiving twice-daily VGCV. Patients ≥18 y of age or those without consent were excluded.

Management of CMV Infection After LT

In our institution, preemptive therapy is used for the management of CMV infection following LT.⁸ Prophylactic antiviral therapy is not administered. Instead, CMV antigenemia is monitored once weekly and antiviral therapy is initiated when the result exceeds 2 or more positive cells per 50 000 leukocytes. CMV treatment consists of either oral VGCV or intravenous GCV. When using VGCV, the dosage is 16 mg/kg/dose administered twice daily.

Data Collection

Clinical information was extracted from electronic medical records. We collected data on demographic information (age, height, weight, BSA estimated using the Mosteller formula BSA [m²] = $\surd \left\{\frac{(height\left[cm\right]*weight\left[kg\right])}{3600}\right\}$, sex, and underlying diseases); laboratory test results (white blood cell counts, serum creatinine, CrCl estimated using the Schwartz formula based on the VGCV interview form in Japan,⁹ and CMV antigenemia); VGCV administration-related information (formulation, dosage, timing of administration, and blood sampling times); and clinical information (concomitant medications, clinical course, and outcomes).

Blood Sampling, GCV Concentration Measurement, PK Analysis, and Definition of the Therapeutic Target of AUC

Blood sampling was performed as part of routine clinical care. The timing of blood sampling was determined by the clinician, and GCV concentrations were measured using residual samples from clinical specimens. Serum GCV concentrations were measured by liquid chromatography-mass spectrometry using a validated method.¹⁰ The details of the concentration measurement methods are provided in the Supplemental File, SDC, https://links.lww.com/TXD/A873.

A population PK model established in pediatric patients after liver and kidney transplantation was used to estimate the AUC_0–24 in this study.¹¹ The therapeutic target was set at AUC_0–24 80–120 µg·h/mL.² PK analyses were performed using Phoenix NLME (v. 8.3) (Certara, Inc., Tokyo, Japan).

Primary and Secondary Endpoints

The primary endpoint was to compare the achievement rate of target AUC between BW and BSA doses. Although BW dose is the standard at our institution, a hypothetical BSA dose was calculated by substituting the individual patient data into the formula: 7 × BSA (m²) × CrCl (mL/min/1.73 m², if the value was >150, it was treated as 150 for the calculation). AUC for BSA dose was estimated proportionally for comparison with BW dose. The secondary endpoint was the time to clearance of CMV viremia.

Evaluation of Adverse Events

Adverse events were evaluated based on clinical laboratory values, including bone marrow suppression (WBC counts, lymphopenia, neutropenia, anemia, and thrombocytopenia) and renal impairment (elevated serum creatinine levels). Common terminology criteria for adverse events version 5.0 was used to grading adverse events,¹² with the baseline being before VGCV administration, and the most abnormal value during the VGCV administration period was evaluated. Additionally, renal impairment was assessed using pediatric renal function assessment criteria and classified according to the CKD staging system (stage 1: estimated glomerular filtration rate [eGFR] ≥90 mL/min/1.73 m², stage 2: 60–89 mL/min/1.73 m², stage 3: 30–59 mL/min/1.73 m², stage 4: 15–29 mL/min/1.73 m², stage 5: <15 mL/min/1.73 m²).¹³ Renal impairment was defined as an increase of 1 or more stages.

Statistical Analysis

Categorical variables were expressed as absolute number and percentage, and continuous variables as median with interquartile ranges (IQR). The proportions of patients achieving the target AUC in the BW and BSA dose groups were compared using the Fisher–Freeman–Halton exact test. Continuous variables were compared by Mann–Whitney U test. A 2-sided P value of <0.05 was considered statistically significant. The correlation between AUC and the duration of CMV-positive days to assess the viral clearance was evaluated using the Spearman’s rank correlation test. All statistical analyses were performed by R Version 4.3.2. (Vienna, Austria: R Foundation for Statistical Computing; 2023. Available from: https://www.R-project.org/).

Ethical Consideration

Written informed consent was obtained from the parents or legal guardians of all study participants. This study was approved by the Ethics Committee of the NCCHD (NCCHD-1328, 2024-215).

This study was conducted in an effort to comply with the STROBE guidelines.

RESULTS

Patient Backgrounds

This study included 14 VGCV-treated cases from 12 pediatric LT patients <18 y old. All cases received VGCV as preemptive therapy, and no case developed tissue-invasive CMV disease. Eight of the 14 cases (57.1%) were male. The median age at onset of VGCV administration was 1.1 y (IQR: 0.5–3.1 y), the median weight was 6.5 kg (IQR: 4.7–10.1 kg), and the median BSA was 0.33 m² (IQR: 0.27–0.47 m²). The median CrCl was 165.8 mL/min/1.73 m² (IQR: 129.7–187.7 mL/min/1.73 m²). Indications for LT included cholestasis in 6 cases (42.9%), fulminant hepatic failure in 4 cases (28.6%), metabolic disorders in 2 cases (14.3%), and other conditions in 2 cases (14.3%). The donor source was living donor in 12 cases and deceased donor in 2 cases. Regarding the use of immunosuppressants at the time of detection of CMV antigenemia, prednisolone was used in all cases. Other medications included tacrolimus (TAC) in 9 cases (64.3%), antithymocyte globulin (ATG) in 6 cases (42.9%), mycophenolate mofetil (MMF) in 5 cases (35.7%), and cyclosporine A (CyA) in 5 cases (35.7%) (Table 1).

TABLE 1.

Patient characteristics

Variables	Number (%) or Median (IQR)
Case number	14*
Age (y)	1.1 (0.5–3.1)
Male sex	8 (57.1)
Body weight (kg)	6.5 (4.7–10.1)
Body surface area (m2)	0.33 (0.27–0.47)
Creatinine clearance (mL/min/1.73 m2)	165.8 (129.7–187.7)
Underlying diseases
Cholestasis	6 (42.9)
Fulminant hepatitis failure	4 (28.6)
Metabolic diseases	2 (14.3)
Others	2 (14.3)
Donor source
Living	12 (85.7)
Deceased	2 (14.3)
Days postliver transplantation	53 (30–79)
Immunosuppressive agents
Prednisolone	14 (100.0)
Tacrolimus	9 (64.3)
Antithymocyte globulin	6 (42.9)
Cyclosporin A	5 (35.7)
Mycophenolate mofetil	5 (35.7)
Cytomegalovirus serostatus
Donor+ / Recipient+	2 (14.3)
Donor+ / Recipient −	7 (50.0)
Donor − / Recipient+	0 (0)
Donor − / Recipient–	1 (7.1)
Donor unknown / Recipient–	2 (14.3)

^*Two patients had 2 episodes of VGCV treatment, yielding a total of 14 cases from 12 patients.

IQR, interquartile range; VGCV, valganciclovir.

GCV Dose and AUC

The median (IQR) actual VGCV dosage was 15.3 (14.7–16.1) mg/kg/dose, administered orally twice daily. The dosage form was oral solution in 13 (92.9%) of 14 cases and tablet suspension in 1 (7.1%) case. The median (IQR) serum sample collection time after VGCV administration was 5.8 (2.3–11.2) h, and the median (IQR) detected GCV concentration was 1.3 (0.5–4.8) μg/mL. The median (IQR) AUC_0–24 estimated from the detected GCV serum concentrations was 71.8 (49.0–110.1) μg·h/mL (Table 2). The calculated median (IQR) VGCV dose per body weight for BSA dose was 24.5 (22.1–27.1) mg/kg/dose. The median AUC_0–24 when administered based on BSA using individual GCV PK parameters was 92.2 (66.8–108.3) μg·h/mL.

TABLE 2.

Valganciclovir dosage and concentration

Variables	Number (%) or median (IQR)
Dosage form
Syrup	13 (92.9)
Tablet suspension	1 (7.1)
Dose (actual), mg/kg/dose, twice daily	15.3 (14.7–16.1)
Dose (hypothetical BSA), mg/kg/dose, twice daily	24.5 (22.1–27.1)
Number of measurement points per case	7 (3–9)
Time after administration of measurement points (h)	5.8 (2.3–11.2)a
Detected serum GCV (µg/mL)	1.3 (0.5–4.8)a
Estimated GCV AUC0–24 after VGCV dose (µg·h/mL)	71.8 (49.0–110.1)

^aMedian value from all measurement points.

AUC, area under the concentration-time curve; BSA, body surface area; GCV, ganciclovir; IQR, interquartile range; VGCV, valganciclovir.

Target AUC attainment was evaluated using a predefined target AUC_0–24 of 80–120 μg·h/mL. For the VGCV BW-based dose and BSA-based dose, the numbers of patients with AUC <80 μg·h/mL (underexposure) were 12 (85.7%) and 5 (35.7%), those with AUC 80–120 μg·h/mL (target exposure) were 2 (14.3%) and 6 (42.9%), and those with AUC >120 μg·h/mL (overexposure) were 0 (0.0%) and 3 (21.4%), respectively (P = 0.027; Table 3).

TABLE 3.

Valganciclovir AUC target attainment rate (n = 14).

AUC ranges μg·h/mL	<80 (underexposure)	80–120 (target exposure)	>120 (overexposure)
Body weight–based dose	12 (85.7%)	2 (14.3%)	0 (0.0%)
BSA-based dose	5 (35.7%)	6 (42.9%)	3 (21.4%)

P = 0.027 (Fisher–Freeman–Halton exact test).

AUC, area under the concentration-time curve; BSA, body surface area.

CMV Clearance

The presence of 1 or more CMV antigen-positive cells in peripheral blood polymorphonuclear leukocytes by CMV antigenemia test was defined as positive, and the duration until the antigen-positive cells disappear (positive days) was investigated. The median (IQR) positive days was 14 (13–24) d. A higher AUC tended to be associated with more rapid viral clearance (the correlation coefficient [ρ] with AUC_0–24 was –0.72, P = 0.003; Figure 1).

FIGURE 1.

Relationship between AUC and CMV viremia duration. The black circles represent the AUC estimated from the actual administered dose and the number of days until viral clearance. The dotted line represents the nonparametric regression line. AUC, area under the concentration curve; CMV, cytomegalovirus.

Adverse Events and AUC

White blood cell counts, lymphocyte counts, neutrophil counts, hematocrit values, and platelet counts were investigated during VGCV administration, and the lowest values of each test during VGCV administration were assessed. According to the common terminology criteria for adverse events adverse event classification, grade 1 or higher increases were observed in 10 cases of leucopenia, 4 cases of lymphopenia, 8 cases of neutropenia, 2 cases of anemia, and 3 cases of thrombocytopenia. The estimated median AUC_0–24 (IQR) was compared between patients without and with leukopenia, lymphopenia, neutropenia, anemia, and thrombocytopenia. The corresponding values were 102.1 (83.6–118.5) versus 61.2 (45.1–101.3) for leukopenia, 89.2 (49.0–111.6) versus 64.6 (46.3–79.8) for lymphopenia, 110.1 (83.8–124.4) versus 50.2 (44.0–74.8) for neutropenia, 81.2 (51.3–110.6) versus 59.7 (51.9–67.4) for anemia, and 68.4 (46.0–111.1) versus 94.0 (74.0–110.0) for thrombocytopenia, respectively. Among these, the estimated AUC_0–24 was significantly lower in patients with neutropenia (P = 0.029; Figure 2). No patients met the criteria for renal impairment.

FIGURE 2.

AUC and hematologic adverse events. A, White blood cells, (B) Lymphocytes, (C) Neutrophils, (D) Hemoglobin, (E) Platelet count. Blue bars indicate the group without CTCAE grade increase (no adverse events), and orange bars indicate the group with CTCAE grade increase (with adverse events). *P < 0.05. AUC, area under the concentration curve; CTCAE, common terminology criteria for adverse events.

DISCUSSION

In this study, estimated BSA dose resulted in a higher VGCV dosing than BW dose and was suggested to achieve higher AUC, improving the likelihood of attaining the target therapeutic exposure. Furthermore, higher AUC values were found to be associated with faster clearance of CMV viremia.

In our study, BSA dose could achieve the target AUC more frequently than BW dose. To date, no clear consensus has been reached regarding whether dosing based on BSA or BW is preferable as a therapeutic VGCV use for twice-daily dosing in pediatrics.² In both the Japanese and United States package inserts for VGCV, the recommended dosing for prophylaxis is provided as a once-daily regimen based on BSA dose.^9,14 However, for treatment with a twice-daily regimen, no specific dosing recommendation is given. Several studies have reported on the VGCV dose for CMV prophylaxis. A prospective PK study demonstrated that weight-based VGCV dosing at 17 mg/kg once daily often resulted in lower-than-target GCV exposures (AUC_0–24), particularly in younger children. Despite the subtherapeutic exposures, no cases of CMV viremia or disease occurred during prophylaxis, suggesting sufficient clinical efficacy; therefore, the authors cautioned that manufacturer-recommended BSA dosing may result in substantially higher drug exposure, particularly in infants.¹⁵ In a PK study of 23 pediatric SOT patients treated with weight-based valganciclovir dosing (14–16 mg/kg), simulated manufacturer-recommended BSA-based dosing produced higher GCV AUCs, but was frequently supratherapeutic, resulting in lower attainment of the target AUC (AUC_0–12 for treatment and AUC_0–24 for prophylaxis: 40–60 μg·h/mL) compared with the BW-based approach.⁶ However, the majority of patients in these studies received once-daily dosing for prophylactic purposes, and the optimal dosing regimen for therapeutic use remains unclear. The international guidelines for CMV infection following SOT, published in 2025, state that in pediatric patients, BSA dose tends to result in higher doses—and consequently higher AUCs—than BW dose. BSA dose increases the likelihood of achieving the target AUC but also raises the risk of adverse events. Accordingly, the guidelines recommend that either BSA dose or BW dose may be used.² Our results support the use of BSA dose in terms of achieving the target AUC. However, to determine which dosing strategy is truly preferable in clinical practice, it is necessary to examine whether a higher AUC is associated with improved therapeutic efficacy and whether such higher exposures lead to unacceptable adverse effects.

In this study, a higher GCV AUC was associated with a more rapid improvement in CMV viremia. Data on the association between GCV AUC and therapeutic efficacy against CMV infection are limited. A target AUC of 80–120 μg·h/mL has been proposed for the treatment of CMV infection.² However, the value is not supported by prospective clinical trial data and is known to be based primarily on expert opinion.⁷ In fact, several studies have failed to identify a definitive association between AUC and therapeutic efficacy. In a prospective study of 7 adult D+/R– SOT patients treated with VGCV for CMV disease or asymptomatic viremia, it found that despite achieving apparently adequate GCV exposures, viral clearance rates varied widely, with a median time to undetectable CMV DNA of 34 d (range 28–82), suggesting that GCV exposure alone may not reliably predict virologic response.¹⁶ The other prospective study evaluated the relationship between VGCV exposure and the prevention of CMV infection in pediatric SOT patients. Although the majority of patients had an AUC below 40 μg·h/mL, no cases of CMV viremia or disease were observed, and no clear association between AUC and prophylactic efficacy was demonstrated.¹⁵ The reason for the relatively clear association observed in this study between higher AUC and faster viral clearance remains unclear; however, it is possible that this finding was affected by the inclusion of a relatively homogeneous population of pediatric LT patients, who generally experience lower levels of immunosuppression compared with other organ transplantation patients. Larger-scale studies are warranted for further investigation regarding the relationship between AUC and the therapeutic efficacy against CMV infection.

In the present study, no clear association was observed between higher AUC and the frequency of hematologic adverse effects, and paradoxically, the group without neutropenia had significantly higher AUC. Various results have been reported to date regarding the relationship between the AUC of GCV and hematologic adverse effects. A prospective randomized trial investigating the relationship between GCV or VGCV exposure demonstrated that higher AUC values of GCV were associated with an increased risk of neutropenia and anemia.¹⁷ In contrast, in a large retrospective cohort study, multivariable analysis indicated that VGCV dosage, serving as a surrogate for AUC, did not affect the risk of developing neutropenia.¹⁸ These differences in results are likely attributable to variations in patient backgrounds, dosing regimens of GCV or VGCV, and AUC levels. The reason why patients without neutropenia had higher AUCs in our study is unclear; however, 1 possible explanation is that CMV infection itself exerts a myelosuppressive effect, and that more rapid viral clearance achieved with higher AUCs may have contributed to this finding. At the same time, unmeasured or unknown confounding factors related to cytopenia cannot be excluded, including the potential impact of differences in the type or dosage of immunosuppressive agents with bone marrow–suppressive effects. Nevertheless, within the context of the present cohort, in the setting of preemptive VGCV therapy following pediatric LT, hematologic adverse effects may not pose a major concern when the AUC remains within the appropriate range.

This study has several limitations. First, it was conducted at a single center with a relatively small number of patients, which limits the generalizability of the findings. Further studies involving larger patient cohorts are warranted. Second, it remains uncertain whether a target AUC of 80–120 μg·h/mL is appropriate for VGCV preemptive therapy in CMV infection. Although this range is described in recently published CMV treatment guidelines,² it was not established based on robust evidence. The optimal AUC may vary depending on factors such as the type of transplanted organ, the intensity of immunosuppression, and the severity of CMV infection. Further studies are needed to define a clinically relevant AUC target. Third, in this study, CrCl was estimated using the Schwartz equation in accordance with the VGCV interview form; however, this method may overestimate CrCl in neonates and young infants, even CrCl was treated as 150 mL/min/1.73 m², if the value was >150 mL/min/1.73 m². Future studies should consider alternative approaches, such as estimating CrCl using cystatin C, particularly in this age group. Fourth, viral load was assessed using the antigenemia assay. To more accurately evaluate viral clearance, future studies should use more quantitative methods, such as PCR-based assays. Finally, the efficacy of VGCV in this study was evaluated only by the duration of the days of CMV clearance. As all patients received VGCV for preemptive therapy while asymptomatic viremia was present, it remains uncertain as to whether these findings apply to patients who develop CMV disease, such as CMV pneumonia, colitis, or retinitis.

In conclusion, for preemptive therapy against CMV following pediatric LT, BSA dose of VGCV was estimated to achieve a higher proportion of target AUC attainment compared with BW dose. Further studies are warranted to determine whether the target AUC applied in this study is clinically appropriate.

ACKNOWLEDGMENTS

The authors initially used artificial intelligence (ChatGPT-5 and OpenEvidence) to support literature retrieval, grammatical corrections, and improvements in English expression during the preparation of this article. We thank the senior medical editor at the Center for Postgraduate Education and Training at NCCHD for authoritative, professional language editing. The authors thoroughly reviewed the content and take full responsibility for the final article. This study was partly presented at the ASM Microbe 2025, Los Angeles, USA.