From January 2005 to December 2010, 470 related and unrelated HCT were realised at Maisonneuve-Rosemont Hospital. Of this group, 103 patients (21.9%) developed grade III to IV aGVHD. This study examines 91 of these 103 patients (88%). 12 patients with grade III to IV aGVHD (11.6%) were diagnosed with PC by tomodensitometry. This study compares the clinical characteristics, treatment and evolution of these 12 patients with PC with the control group of 91 patients with severe GI GVHD without PC.
All patients (100%) who developed PC received oral ciprofloxacin at the time of HCT compared with 42 patients (53.2%) in the control group. Donor and recipent CMV status were negative in 75% of patients with PC compare to 35.4% in patients with severe GVHD without PC. The median time of onset for acute GVHD was 64 days in the PC cohort (84 days control group). The median time for PC development was 199 days from HCT. Stem cell donors were female in 72.7% of patients with PC (47.3% in the control cohort). Sexe mismatch between donor and recipient was present in 50%, ( 26% in the control group). 92% of patients received peripheral blood as their source of stem cells (82.2% in the no PC cohort). HLA disparity was present in 33.2% of patients with PC (15.2% in the no PC cohort). 41.7% of patients developed GI tract infections before the onset of PC (29.1% control group).
All patients were on calcineurin inhibitors at time of acute GVHD onset. Treatment of aGVHD consisted of corticosteroids MP 2 mg/kg in 91.7% of PC cohort (65.8% in the control group). The succcessive lines of treatment were: MMF 66.7% (78.5% in the rest of the cohort), daclizumab 75% (34.2% in the control group) and pentostatin 41.7.4% (5% in the no PC cohort) with complete resolution of GVHD symptoms in 16.7% (32.9% in the rest of the cohort).
PC is a severe complication of early onset GI GVHD. Risk factors included HLA disparity between donor and recipient, female donor, high dose and prolongued corticosteroids treatment for GVHD. GI tract infection may contribute to its pathophysiology and frequently preceeded PC onset. Antibioprophylaxis with ciprofloxacin was a statistically significant risk factor for PC development. Oral cipro for GI tract decontamination is suboptimal and should be revised. Treatment of GI GVHD with daclizumab or pentostation increase PC risk. CMV seronegatvity was not protective. GVHD related PC is associated to a high mortality rate (75% in the PC cohort and 59% no PC group) not statistically different from the dismal outcome of steroid refractory aGVHD.