Serial Cerebrospinal Fluid CMV DNA Monitoring During Extended Induction Therapy for HIV-Associated Central Nervous System Cytomegalovirus Infection: A Retrospective Cohort Study
Year: 2026
Type: preprint
Study design: Observational Study
Abstract: Background Central nervous system (CNS) cytomegalovirus (CMV) infection remains a devastating complication among people with HIV (PWH), with historical mortality exceeding 80% despite standard 2–3-week antiviral induction. Although plasma CMV DNA predicts mortality, cerebrospinal fluid (CSF) viral load dynamics and their therapeutic implications remain poorly characterized. Methods We conducted a retrospective review of 33 PWH with confirmed CNS CMV infection (CD4 < 200 cells/µL) admitted to a tertiary hospital in Nanjing, China, between March 2019 and August 2024. All patients received induction therapy (94% ganciclovir + foscarnet) with serial CSF CMV DNA monitoring. Clinical features, virologic dynamics, treatment responses, and 182-day outcomes were examined. Delayed CSF clearance was defined as CSF CMV DNA detectable at the last available CSF observation (never cleared). Results Median baseline CD4 count was 11 cells/µL (IQR 4–34); 87.9% had CD4 < 50 cells/µL. Extended induction (median 6 weeks, IQR 4–8) achieved 72.7% 182-day survival. CSF CMV clearance was substantially slower than plasma clearance: 63.6% remained CSF PCR-positive (> 500 copies/mL) at week 2 vs. 42.4% with detectable plasma viremia. Non-survivors had higher baseline CSF CMV loads (4.9 vs. 3.9 log₁₀ copies/mL, P = 0.075) and pronounced CSF inflammation. Grade ≥ 3 hematologic toxicity occurred in 54.5%. Neurological symptom burden was prognostically relevant: patients with ≥ 2 of four neurological symptoms at presentation had significantly worse 182-day survival than those with ≤ 1 symptom (42.9% vs. 94.7%; log-rank P = 0.0010). Conclusions In this cohort, patients receiving CSF-monitored extended induction therapy had a 72.7% 182-day survival rate. Compartmentalized viral kinetics suggest that CSF-specific monitoring may identify patients who require prolonged therapy, though randomized trials are warranted. Symptom burden may help identify patients at highest risk.
Source: Research Square
Institution: Nanjing University of Chinese Medicine
Language: English
Subfield: Epidemiology
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7213473003
