Unexpectedly high seroprevalance of Kaposi's sarcoma-associated herpesvirus (HHV-8) in patients with stage IV melanoma

We thank Professor Hauschild for his recent commentary on our approach to predicting hepatitis after combined anti-PD-1 and anti-CTLA-4 therapy in patients with metastatic melanoma [1]. Our original publication reported that a subset of patients at high risk of developing hepatitis could be identified prior to treatment by cytomegalovirus (CMV)-seropositivity and expansion of effector memory CD4+ T cells (TEM) in peripheral blood [2]. This strong predictive model led us to hypothesize that patients with metastatic disease are susceptible to chronic or recurrent CMV reactivation that drives pre-treatment expansion of virus-specific CD4+ TEM cells that cause liver injury when checkpoint inhibitors are given.