Acute graft-versus-host disease (aGVHD) is a frequent complication of allogeneic hematopoietic stem cell transplantation (HSCT). Although the presence of a skin eruption is a cornerstone in the diagnosis of aGVHD according to the Glucksberg criteria and the 2005 NIH Consensus Conference, specific cutaneous features such as morphology and anatomic distribution have not been studied in a systematic manner. Subsequently, the relative incidences of specific skin lesions remain unclear and contribute to the ongoing challenge of delineating aGVHD from other commonly seen skin eruptions early after HSCT.
Methods:
A retrospective review of all patients receiving an allogeneic HSCT from 2010 to 2011 at Northwestern Memorial Hospital identified those individuals with cutaneous aGVHD. Each case of aGVHD was confirmed by both skin biopsy and evaluation by the dermatology consult service. Cutaneous features with respect to lesion morphology, location, and time of onset following transplantation were collected for each patient and further stratified by aGVHD clinical grade.
Results:
For all cutaneous aGVHD patients (n=37), the onset of rash was on average, day +45 (range +4 to +153). The most common skin lesion morphology was morbilliform (55%) followed by patchy erythema (38%). Follicular accentuation was seen in 29% of rashes. The cutaneous eruptions most commonly occurred on the trunk (69%), arms/legs (67%), face (62%), ears (38%) and palms (38%).
Comparing grade I and II skin aGVHD patients, grade I patients (n=11) had more patchy erythema (73% vs 23%, p=0.008), purpuric/violaceous lesions (55% vs 15%, p=0.038) or a reticular pattern (36% vs 0%, p=0.005) than grade II patients (n=26). Grade II aGVHD patients were found to have cutaneous eruptions located more often on the trunk (85% vs 36%, p=0.006) and arms/legs (77% vs 36%, p=0.028) when compared to grade I aGVHD patients. Comparison to grade III (n=3) and IV (n=2) aGVHD patients was not performed given the paucity of cases.
Conclusion:
We profiled the salient cutaneous features associated with aGVHD and their relative incidences based on clinical grade, skin lesion morphology and anatomic site of involvement. Application of this analysis will allow for improvements in the diagnosis of aGVHD and differentiation from other clinical mimickers. Differences in lesional morphology and location may differentiate grade I and II cutaneous aGVHD and serve to guide appropriate treatment.