Search

Introduction The paradigm for preventing clinically significant cytomegalovirus (CMV) infection and CMV disease after hematopoietic cell transplantation (HCT) has shifted from preemptive treatment to antiviral prophylaxis with the approval of letermovir. Previously, our group validated the use of quantitative CMV DNA viral load as a surrogate endpoint for tissue-invasive CMV disease (ASBMT 2018) in the pre-emptive treatment setting. Objectives To address whether CMV viral load kinetics could serve as valid surrogate endpoints for CMV disease in the prophylactic setting, we performed CMV viral load testing of frozen serum samples collected during a placebo-controlled randomized trial of ganciclovir (GCV) given at engraftment (Goodrich et al. Ann Intern Med 1993). Because this study pre-dated the approval of ganciclovir, CMV disease occurred in an extremely high proportion of patients, allowing us to link viral load kinetics with CMV disease, which would not be possible in the modern treatment environment. Methods We calculated each patient's CMV viral load kinetics (mean, peak, maximum change, percentage of positive viral loads) from samples collected during the first five weeks after randomization and analyzed their value as surrogates for CMV disease by 24 weeks after randomization. First, we tested each marker for fulfillment of the Prentice criteria for valid surrogate endpoints using multivariable logistic regression, including quantifying the degree to which each marker captured the effect of ganciclovir on CMV disease. Secondly, to determine the ability of the viral load markers to predict CMV disease, we developed a ‘SuperLearner' machine learning model and quantified prediction accuracy by cross-validated area under the receiver-operator characteristic curves (cv-AUCs). Results Mean, peak, change, and percentage of positive viral loads were significantly higher in the placebo group than in the ganciclovir group. See viral loads by group in Fig 1. Percentage of positive viral load satisfied Prentice criteria as a valid surrogate endpoint for CMV disease by week 24 after randomization (Fig 2). All kinetics explained greater than 80% of ganciclovir's reduction in CMV disease (mean: 86%, peak: 83%, maximum change: 95%, percent positive: 94%). We found that a SuperLearner model that included all four viral load kinetics, CMV donor serostatus, acute graft-versus-host disease, and baseline viral load predicted CMV disease with 91% cv-AUC in the placebo group, 78% in the ganciclovir group, and 82% in the combined groups (Fig 3). Conclusion We have shown for the first time in a clinical endpoint-rich, placebo-controlled antiviral prophylaxis trial that CMV viral load kinetics fulfill the Prentice criteria as valid surrogate endpoints and have high predictive value for CMV disease after HCT.

The United States Food and Drug Administration recently issued guidance that CMV DNAemia (quantitative CMV viral load as measured by PCR) could be used in a composite endpoint along with tissue-invasive CMV disease as the primary endpoint in allogeneic HCT CMV prophylaxis trials. However, the specific time points after antiviral treatment initiation at which to measure viral load and the viral load thresholds that most accurately predict clinical CMV disease have not been identified. In order to estimate the treatment effect of an antiviral agent on CMV clinical outcomes, placebo-controlled, randomized trial data with accompanying viral load measurements are needed. Unfortunately, in the modern era of PCR and early preemptive treatment, ethically, placebo-controlled, randomized trials that allow for the development of CMV disease after HCT can no longer be performed. To overcome this limitation, we performed CMV DNA PCR testing on frozen plasma samples collected during the first 100 days post-HCT in the only placebo-controlled, double-blind RCT of ganciclovir for the early treatment of CMV infection after HCT (Goodrich et al. NEJM 1991). We used three analytic methods (Cox proportional hazards models, machine learning, and a novel non-parametric method) to determine which of 14 viral load markers, measured in the first four weeks of treatment, are the most useful surrogates of antiviral treatment for the prevention of CMV disease. All methods identified peak viral load and percentage of positive samples by week 4 as highly associated with or predictive of tissue-invasive CMV disease by week 8 post-randomization. Cox proportional hazards models (Figure 1) yielded significant associations for peak viral load measured by week 4 (HR 1.4, 95% CI 1.1-1.9) and percentage of positive samples by week 4 (HR 1.6, CI 1.1-2.3). The machine-learning ensemble algorithm SuperLearner implemented in R (Figure 2) demonstrated cross-validated area under the receiver-operator curves of at least 75% for peak viral load (75%, CI 62-87%) and percentage of positive plasma samples (77%, CI 63-90%) measured by week 4. Using a novel statistical technique that allows direct estimation of biomarker surrogacy (Parast et al. Stat Med 2017), we estimated that the percentage of ganciclovir treatment effect explained by peak viral load and percentage of positive samples measured by week 4 was greater than 85%—peak viral load 89%, CI 56-100%; percent positive samples 91%, CI 60-100% (Figure 3). Peak viral load and percentage of positive samples by week 4 are highly predictive of CMV disease and the treatment effect of ganciclovir. Our analysis suggests that these may be the best surrogate markers of ganciclovir treatment and introduces novel methods for identifying optimal candidate biomarkers as surrogates for clinical endpoints.