Search

Aims: Keratinocyte apoptosis is a major pathogenic mechanism in dermal complications, such as graft versus host disease (GVHD), after allogeneic bone marrow transplantation. However, the mechanisms by which recipient target cells undergo apoptosis in GVHD are still unclear, but may result from DNA damage caused by chemotherapeutic agents and/or by direct cytokine action. The basis of this investigation was to correlate keratinocyte apoptosis with (1) the severity of graft versus host reactions (GVHR) in vitro and (2) the clinical grade (0-III) of GVHD. Methods: Skin sections generated from an in vitro skin explant model for detecting experimental or clinically relevant GVHR were investigated for the detection of apoptotic nuclei using the terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) technique. This investigation also aimed to establish whether the TUNEL assay could be used as an additional, predictive method for the severity of GVHD before transplantation in potential patient/donor pairs given standard GVHD prophylaxis (cyclosporin A and methotrexate). Results: By comparing mean values of apoptosis for each GVHR grade in a cohort of 83 retrospective skin sections it was shown that as the severity of GVHR increased there was a parallel increase in the percentage of apoptotic cells (p < 0.0001). However, the correlation between clinical GVHD grade II-III and overall keratinocyte apoptosis (> 2.6%) did not reach this degree of significance (χ²: 4.2; degrees of freedom, 1; p = 0.04; Fisher's exact test: p = 0.06). Conclusions: The detection of apoptosis correlated with degree of GVHR using an in vitro assay and a higher degree of apoptosis tended to correlate with more severe GVHD. Further studies in a larger cohort of patients, using other methods to detect apoptosis in conjunction with the TUNEL assay, may give additional insight into the complex immunopathophysiology of GVHD.

Graft-versus-host disease (GVHD) is the most serious complication following bone marrow transplantation, with an incidence of 40-60%. The disease can be fatal in 50% of cases, even in patients receiving marrow from an HLA identical sibling. Several assays have been developed to try to predict the development of GVHD, including mixed lymphocyte culture reaction, cytotoxic T lymphocyte precursor, and helper T lymphocyte precursor frequency assays. This review describes an in vitro skin explant model which has been used since 1988 for both predicting acute GVHD in HLA identical sibling bone marrow transplantation and studying the pathophysiology of the disease. The model involves sensitising donor lymphocytes in vitro in a primary mixed lymphocyte reaction and then evaluating the secondary response on patient skin biopsies by grading the graft-versus-host reactivity (grades I-IV) histopathologically. From analysis of collective data the model is a clear predictor of GVHD and superior to the other assays widely used, with a correlation of 82% with clinical outcome. The skin explant model allows the investigator to study the pathogenesis of GVHD. The cytokines TNFα and IFNγ are shown to be important mediators of cellular damage in graft-versus-host reactions. Recent work has also involved using the model to study the alloreactivity of cord blood. The model is currently being assessed in several bone marrow transplantation centres in Europe on different patient groups including those who receive marrow from haploidentical and matched unrelated donors.