Mortality in cancer patients previously diagnosed with herpes zoster in the hospital setting: a nationwide cohort study.

ABSTRACT

BACKGROUND: Herpes zoster (HZ) is associated with underlying immunodeficiency and may thereby predict mortality of subsequent cancer. METHODS: By using Danish nationwide medical databases, we identified all cancer patients with a prior hospital-based HZ diagnosis during 1982-2011 (n=2754) and a matched cancer cohort without prior HZ (n=26 243). We computed adjusted mortality rate ratios (aMRRs) associating prior HZ with mortality following cancer. RESULTS: Prior HZ was associated with decreased mortality within the year after cancer diagnosis (aMRR 0.87; 95% confidence interval (CI) 0.81-0.93), but not thereafter (aMRR 1.07; 95% CI 0.99-1.15). However, prior HZ predicted increased mortality throughout the entire follow-up among patients aged <60 years (aMRR 1.39; 95% CI 1.15-1.68) and those with disseminated HZ (aMRR 1.18; 95% CI 1.01-1.37). The increased mortality rates were observed primarily for haematological and immune-related cancers. CONCLUSIONS: Overall, HZ was not a predictor of increased mortality following subsequent cancer.