Critically ill patients with COVID-19 frequently exhibit reactivation of latent viruses. Herpes simplex virus type 1 (HSV-1) and cytomegalovirus (CMV) are commonly detected in intensive care settings, but their clinical significance in patients with severe SARS-CoV-2 infection has been debated. The study reported here aimed to assess the incidence, timing, clinical course, and association with outcomes of HSV-1 and CMV reactivation in a large single-center cohort of ICU patients with COVID-19.
This was a retrospective single-center analysis including 455 consecutive ICU patients with PCR-confirmed SARS-CoV-2 infection treated at the anesthesiology ICU of University Hospital Essen, Germany, between March 2020 and December 2021. Data extraction for research occurred between July 1, 2025 and February 28, 2026. Routine virological monitoring consisted of CMV PCR in whole blood (weekly) and HSV-1 PCR in bronchoalveolar lavage fluid (BALF) for invasively ventilated or tracheotomized patients or in selected non-invasive ventilation cases with clinical concern.
Reactivation was defined virologically as any positive PCR detection (HSV-1: any positive in BALF; CMV: any positive in whole blood) regardless of viral load or additional clinical criteria. Institutional treatment guidance recommended acyclovir for HSV-1 in BALF above a specified viral-load threshold combined with respiratory deterioration; ganciclovir was recommended for CMV reactivation with dose adjustments for renal function. Immunosuppressed patients were prioritized for prompt antiviral therapy, whereas treatment in other patients relied on clinical judgement. Since July 2020, dexamethasone 6 mg daily for 10 days was used for SARS-CoV-2 pneumonia per institutional practice.
The primary endpoint was independent risk factors for ICU mortality. Secondary endpoints included 30-day mortality, crude incidence of HSV-1 and CMV, organ failure, need for invasive ventilation, and need for vvECMO. Standard statistical methods were used, including chi-squared testing, t-tests or Mann–Whitney U tests, logistic regression, multivariable adjustment, Kaplan–Meier survival analysis, and sensitivity analysis with HSV-1 modeled as a time-dependent covariate in a Cox proportional-hazards framework.
The cohort’s mean age was 57 years (range 13–94), with 316 (69%) male patients. The median ICU length of stay and median follow-up were both 14 days. COVID-19 pneumonia accounted for 423 (93%) of ICU admissions; 260 (57%) were secondary referrals from external hospitals.
Viral reactivation (HSV-1, CMV, or both) occurred in 164 patients (36%). Isolated HSV-1 reactivation was observed in 54 patients (12%), isolated CMV in 53 patients (12%), and combined reactivation in 57 patients (13%). The median time from ICU admission to detection was 4 days for HSV-1 (IQR 0–10) and 4 days for CMV (IQR 2–12). Detection frequency increased with prolonged ICU stays.
Patients with any viral reactivation were more frequently invasively mechanically ventilated (96% vs 49%), had longer durations of invasive ventilation (median 509 vs 112 hours), and had lower initial PaO2/FiO2 ratios (115 vs 135 mmHg) compared with patients without reactivation. vvECMO was used more often in the reactivation group (57% vs 18%) with longer support durations; renal replacement therapy and sepsis were also more frequent among those with reactivation. ICU length of stay was longer (median 20 vs 6 days).
Isolated HSV-1 reactivation showed a particularly severe phenotype: almost all patients required invasive mechanical ventilation, vvECMO use and duration were substantially higher, renal replacement therapy was more common, and pulmonary embolism rates were increased compared with patients without herpesvirus detection.
Overall, viral reactivation was associated with higher 30- and 90-day mortality (49% vs 33%). Detection of HSV-1 was associated with markedly decreased ICU survival (32% versus 67%; p < 0.001). This relationship for HSV-1 remained significant after multivariable adjustment and was explored further using time-dependent modeling as a sensitivity analysis. In contrast, isolated CMV detection was not reported to have the same independent association with ICU mortality as HSV-1 in the findings summarized.
Among patients with isolated HSV-1, 53 of 54 (98%) received intravenous acyclovir per institutional practice; dosing recommendations were described (e.g., 5 mg/kg every 8 hours, adjusted for renal replacement therapy). For CMV, ganciclovir was recommended with renal dose adjustments, but only 9 of 53 patients with isolated CMV received ganciclovir. In combined reactivation, all patients received acyclovir and a subset received additional ganciclovir.
Patients with viral reactivation had higher peak inflammatory markers (CRP and leucocyte counts). Higher neutrophil-lymphocyte ratio (NLR) correlated with detection of HSV-1 and CMV and was independently associated with increased ICU mortality in multivariable analyses. Pulmonary embolism was more frequent among those with viral reactivation, and secondary referral status was associated with higher reactivation rates.
Univariable and multivariable logistic regression assessed ICU survival and confounding variables. HSV-1 status was also modeled as a time-dependent covariate in a Cox model to address the time-dependent nature of viral reactivation; this sensitivity analysis supported the association between HSV-1 detection and adverse outcomes. Statistical significance was defined as p < 0.05.
In this large retrospective single-center cohort of critically ill patients with COVID-19, herpesvirus reactivation was common. HSV-1 reactivation in BALF was associated with a more severe clinical course, greater needs for respiratory and renal support, prolonged ICU stay, and increased ICU mortality, with the association persisting after multivariable adjustment. CMV reactivation was frequently detected but did not show the same independent association with mortality in the presented summary. Findings were constrained by the retrospective single-center design and by the study definition of reactivation as any positive PCR without viral-load thresholds; detailed supplementary data and analyses were provided by the authors. These results highlight the prevalence of herpesvirus detection in critically ill COVID-19 patients and suggest that HSV-1 detection identifies a subgroup at higher risk of adverse ICU outcomes.