Prevalence and risk factors of cytomegalovirus reactivation in critically Ill patients with COVID-19 pneumonia

Backgrounds In critically ill patients with COVID-19, secondary infections are potentially life-threatening complications. This study aimed to determine the prevalence, clinical characteristics, and risk factors of CMV reactivation among critically ill immunocompetent patients with COVID-19 pneumonia. Methods A retrospective cohort study was conducted among adult patients who were admitted to ICU and screened for quantitative real-time PCR for CMV viral load in a tertiary-care hospital during the third wave of the COVID-19 outbreak in Thailand. Cox regression models were used to identify significant risk factors for developing CMV reactivation. Results A total of 185 patients were studied; 133 patients (71.9%) in the non-CMV group and 52 patients (28.1%) in the CMV group. Of all, the mean age was 64.7 +/- 13.3 years and 101 patients (54.6%) were males. The CMV group had received a significantly higher median cumulative dose of corticosteroids than the non-CMV group (301 vs 177 mg of dexamethasone, p<0.001). Other modalities of treatments for COVID-19 including anti-viral drugs, anti-cytokine drugs and hemoperfusion were not different between the two groups (p>0.05). The 90-day mortality rate for all patients was 29.1%, with a significant difference between the CMV group and the non-CMV group (42.3% vs. 24.1%, p = 0.014). Median length of stay was longer in the CMV group than non-CMV group (43 vs 24 days, p<0.001). The CMV group has detectable CMV DNA load with a median [IQR] of 4,977 [1,365-14,742] IU/mL and 24,570 [3,703-106,642] in plasma and bronchoalveolar fluid, respectively. In multivariate analysis, only a cumulative corticosteroids dose of dexamethasone >= 250 mg (HR = 2.042; 95%CI, 1.130-3.688; p = 0.018) was associated with developing CMV reactivation. Conclusion In critically ill COVID-19 patients, CMV reactivation is frequent and a high cumulative corticosteroids dose is a significant risk factor for CMV reactivation, which is associated with poor outcomes. Further prospective studies are warranted to determine optimal management.