At the onset of acute GvHD after allogeneic hematopoietic cell transplantation (HCT), symptoms and signs are not severe in most patients. The subsequent development of severe acute GvHD despite treatment with high-dose corticosteroids, however, is associated with an increased risk of non-relapse mortality. 1 Skin biopsies are frequently obtained to verify or establish a diagnosis of acute GvHD. The degree of apoptosis or epidermal damage based on evaluation of skin sections by routine hematoxylin and eosin staining or immunostaining for lymphoid markers has not provided useful predictive information on eventual outcomes in patients with acute GvHD. In contrast, Nishiwaki et al. 2 recently demonstrated a higher content of dermal interface macrophages in sections from the skin biopsies taken soon after the onset of acute GvHD was significantly correlated with steroid unresponsiveness and increased mortality. Acceptance of a biomarker that predicts outcome requires confirmation by a second independent study obtained from a separate cohort of patients. Therefore, we sought to determine whether the findings of Nishiwaki et al. 2 could be replicated in a nested case-control study of patients who had skin biopsies shortly after the onset of acute GvHD at the Fred Hutchinson Cancer Research Center.

In a case-control study, we selected patients who had (1) HCT after myeloablative conditioning between 2000 and 2009,(2) acute GvHD involving the skin,(3) a skin biopsy taken o7 days after the diagnosis of acute GvHD and o7 days before or after the onset of treatment for acute GvHD and (4) signed written informed consent allowing the use of medical records and biospecimens for research. Cases were defined as patients who had non-relapse mortality within 1 year, whereas controls were defined as patients surviving without recurrent malignancy for at least 1 year after the onset of treatment for acute GvHD. The Institutional Review Board approved the study. Deparaffinized 4 μm sections from 3-4-mm punch biopsies were immunostained with a CD163-specific monoclonal antibody (Novocastra# NCL CD163 clone 10D6, Novocastra-Leica Biosystems, Buffalo Grove, IL, USA) with the use of a Ventana benchmark XT instrument and