Background: Regulatory trials assessing CMV drugs have utilized symptomatic CMV disease as the primary endpoint. However CMV disease occurs infrequently, CMV syndrome is nonspecific, and early therapy of low-level viremia confounds the endpoint. Use of a viral surrogate may offer advantages in CMV drug development. We performed a literature meta-analysis and systematic review to define whether CMV viral load(VL) is an appropriate surrogate endpoint for use in clinical trials.

Methods: A systematic review of English language articles was conducted using available on-line data sources with appropriate search terms. Study selection: Randomized, controlled trials and cohort studies of CMV VL in solid organ transplant (SOT) recipients. Independent extraction of articles was by two authors using predefined data fields, and all pooled analyses were done with random or fixed effects model based on study heterogeneity.

Results: Of 2441 studies, 66 were included in the analysis. We first assessed if VL predicted disease. Based on 29 studies, the difference in VL in disease vs. asymptomatic patients was 16.3-fold, 7.0-fold, and 14.1-fold higher with all included studies, or Cobas Amplicor assay studies, or just natural history studies, respectively (p<0.001). We next assessed VL kinetics and disease prediction. Five studies were included and all suggested that a more rapid VL increase correlated with the development of CMV disease. Next, studies that assessed VL and disease during vs. after antiviral prophylaxis were evaluated. Based on 16 studies, the pooled incidence of CMV viremia and disease during prophylaxis was 2.2% and 1.0% respectively compared to an incidence of viremia and disease during the whole follow-up period of 33% and 12.6% respectively (p<0.001). We then assessed if therapy of asymptomatic viremia prevented disease (pre-emptive therapy; based on 22 studies). The pooled disease incidence was significantly lower than the pooled viremia incidence (p<0.001). Finally, we assessed studies that correlated symptom response with VL decline; achieving VL negativity correlated strongly with symptom response.

Conclusions: Based on this systematic review, we conclude that VL is likely to predict clinical endpoints in CMV clinical trials and may have some logistic and practical benefits over CMV disease endpoints.

CITATION INFORMATION: Natori Y, Alghamdi A, Kumar D, Miller V, Ljungman P, Husain S, Humar A. Can Viral Load Be Used as a Surrogate Marker in Clinical Studies of Cytomegalovirus in Solid Organ Transplantation: A Systematic Review and Metaanalysis. Am J Transplant. 2017;17 (suppl 3).

« Back to 2017 American Transplant Congress