Superior Survival of Combined Gastrointestinal and Skin GvHD Compared to Isolated Gastrointestinal GvHD in Children Post Allogeneic Stem Cell Transplantation

To compare the prognosis of isolated gastrointestinal (GI) graft versus host disease (GvHD) vs combined (GI and skin) GvHD in children post allogeneic stem cell transplantation (SCT).

Allogeneic SCT recipients from January 2000-Dec 2009 were included and patients who were diagnosed with skin and GI GVHD were identified. Demographics and outcome data were collected and analyzed.

Four hundred and fifty children (0–18 years) underwent allogeneic SCT in the study period. Seventy-nine (17.5%) patient underwent endoscopy and GI biopsy for GI GvHD. Of the seventy-nine patients, 49 had isolated GI GvHD and 30 had combined GvHD. Fifty-one patients received SCT for a malignant indication and 28 for a non-malignant cause. Donor sources were; isolated GI group, related 10, unrelated 20: combined group, related 20, unrelated 29). All patients received myeloablative conditioning regimens and the majority received cyclosporine A and methotrexate for GvHD prophylaxis. Sixty-seven percent of patients with combined GvHD had grade III-IV while only 31% had grade III-IV in patients with isolated GI GvHD. To ensure there was no bias in a quicker initiation of anti GvHD therapy in children with skin involvement, we included only children who had a GI scope and GI biopsy looking for GI GvHD and also we compared if GvHD therapy initiation was different in patients with or without skin involvement [In the combined group 16 patients (53%), therapy was started prior to GI biopsy versus 16 patients (33%) in the isolated GIT group P=0.2]. Interestingly, 40% of children with combined GvHD had also liver GvHD while only 6% of patients with isolated GI GvHD had liver involvement (p≤0.0001). Overall response to GvHD therapy was similar in both groups (83% vs 87%). However, despite having less patients in the isolated GI group with severe (grade III-IV GvHD), there was a significant increase in non-relapse mortality (NRM) in the isolated GI group compared to the combined group (37% vs 16%) resulting in a statistically significant superior overall survival for those children with combined GvHD (79% vs 49% P= 0.02). Disease relapse was similar in both groups (8% and 6% for combined and isolated GIT group, respectively).

Children with GI GvHD has better prognosis when there is skin involvement. Further research is warranted to examine if there is certain biological features that make GI GvHD extends to skin and other organs and why isolated GI GVHD is associated with worse prognosis?

No relevant conflicts of interest to declare.