Reactivation of Cytomegalovirus Infection in Critically Ill Immunocompetent Patients

To the Editor: The cohort study by Dr Limaye and colleagues¹ evaluated reactivation of cytomegalovirus (CMV) infection in critically ill patients. The study used the current generation of extremely sensitive DNA amplification tests for the detection of CMV reactivation in patients who did not have traditional reasons to be immunosuppressed. Although the findings provided the rationale for further investigation, we believe there are aspects of this study that require further research to better understand the implications of the findings.

First, the authors combined many different patient populations into the category of “critical illness,” ranging from acute myocardial infarction to burn injury to sepsis. This difference in patient populations is reflected, for example, in substantial variability between the burn intensive care unit (ICU) and cardiac care ICU in the incidence of CMV reactivation (55% vs 15%, respectively) and duration of CMV reactivation (median, 20 days vs 4 days) and between the burn ICU and medical ICU for time to first detectable CMV viremia (median, 19 days vs 8 days). The authors did not address these patient differences or perform a stratified statistical analysis.

Second, the authors did not capture use of corticosteroids, a common practice in intensive care settings and potentially a precipitant of CMV reactivation. Also, it appears that they did not exclude recent chemotherapy or targeted immunomodulatory agents such as rituximab or alemtuzumab. All of these factors can greatly modulate the risk of CMV reactivation.²

While trauma, sepsis, myocardial infarction, and burn injury can all cause critical illness, the targeted effects of each process on the inflammatory and immune system are substantially different. Effects of burn injury in particular have been well studied, and profound changes in cell-mediated immunity and anergy develop following burn injury, predisposing these patients to a variety of opportunistic infections.^{3 - 5} Although a randomized controlled trial of antiviral prophylaxis may be warranted, as Limaye et al argued, the correct (and ideally uniform) patient populations must be identified to ensure that the potential benefits (clinical and cost) of antiviral agents outweigh the harms. Finally, more work is needed to determine the immune deficits underlying CMV reactivation in critically ill patients.