Blood Viral Load (VL) Not Clinically Meaningful in Symptomatic Congenital Cytomegalovirus (cCMV) Infection

Background Sensorineural hearing loss (SNHL) and neurodevelopmental (ND) outcomes are favorably impacted by antiviral therapy in infants with symptomatic cCMV disease. We correlated blood VL before and during therapy with clinical findings at presentation and follow-up in this population. Methods Post-hoc analysis of two clinical trials conducted by the CASG from 2002 to 2013 evaluating valganciclovir therapy. 120 subjects (73 treated × 6 weeks, 47 treated × 6 months) were included. Whole blood VL was determined by real-time PCR at a central laboratory before therapy (baseline, BL) and periodically for 6 months. Results In subjects treated for 6 months, increases in BL VL correlated with decreased probability of better hearing outcomes at 12 months (Figure 1), but clinically meaningful VL thresholds that predict SNHL were not identified (Table 1). Subjects treated for 6 weeks had no correlation between BL VL and SNHL. No correlation was found between BL VL and Bayley ND testing at 12 and 24 months for subjects receiving either treatment duration. Subjects treated for 6 months who achieved and sustained VL suppression (<2.5 log) between treatment day 14 and month 4 had better hearing outcomes at 6, 12, and 24 months (89% vs. 56%, P = 0.01; 100% vs. 63%, P = 0.0007; 94% vs. 68%, P = 0.04), but 56%–68% of subjects not achieving suppression still had improved hearing. Higher BL VL correlated with BL CNS involvement, thrombocytopenia, and transaminase elevation for subjects receiving either treatment duration, but with substantial overlap in quantity of virus detected (Figure 2). Subjects with >3 symptoms of congenital CMV at presentation had higher BL VL than subjects with ≤3 symptoms (3.75 log, range 1.00–5.65, vs. 3.38 log, range 1.00–5.36; P = 0.005). Conclusion Blood VL at BL and during therapy has little clinically meaningful predictive value for long-term outcomes in symptomatic congenital CMV. Table 1 Hearing outcome BL VL (log genome equivalent/ml) Improved/protected (no.) Others (no.) P-value Negative predictive value (CI) Positive predictive value (CI) 12 months >3 43 20 0.10 93 (79–100) 32 (20–43) ≤3 13 1 >4.5 8 9 0.01 80 (70–90) 53 (29–77) ≤4.5 48 12 24 months >3 42 14 0.72 83 (62–100) 25 (14–36) ≤3 10 2 >4.5 10 5 0.32 79 (68–90) 33 (9–57) ≤4.5 42 11Figure 1Figure 2 Disclosures J. Englund, Gilead: Consultant and Investigator, Research support; Chimerix: Investigator, Research support; Alios: Investigator, Research support; Novavax: Investigator, Research support; MedImmune: Investigator, Research support; GlaxoSmithKline: Investigator, Research support