Hematopoietic stem cell transplantations (HSCT) were performed between 1995 and 2012. Diseases were non malignant inherited diseases (63%) and hematological malignancies (37%). Median age was 4, 3 years. Donors were MRD (27.5%), MUD (32%), mismatch-MUD (13%), Haplo-identical (20.5%) and Cord Blood (CB) (7%). The graft source was BM (80%), PBSC (12%) and CB (8%). 37% patients had a myeloablative conditioning. The median time to aGVHD was 13 days after HSCT. The baseline grade of aGVHD was I in 19%, II 45.5%, III 25% and IV 10.5%. The median delay between aGVHD and Inolimomab was 16 days. The median dose was 0.4mg/kg and the median treatment duration of 27 days.
We observe 55% complete responses and 14% partial responses for a total response rate of 69% and with no side effect. Among the responders, 26% relapse. Incidence of infection during treatment was 19.5%.
A logistic model on complete response provide evidence of the high predictive negative effect of the baseline grade [Odds Ratio (OR): 5.24]. No particular target organ was a significant predictive factor for treatment response or survival. However, multi-organ involvement predicted worse survival with 34%, 45% and 52% of death for patients with one, two, and three or more organs involved, respectively (p=0.234).
The observed overall survival probability was 59%, with a median survival time of 463 days. The Overall unadjusted survivals were 87%, 77 %, and 61% at day 100, 200, and one year.
Survival was also significantly improved for responders (Relative Risk: 0.413; p=0.002).
As shown by multivariate Cox model, the survival was significantly lower for patients with non malignant inherited diseases [Hazard Ratio (HR) =2.88, p<.001] compared with hematological malignancies, and female patients were associated with higher mortality (HR=1.72 , p=.049). For both response and survival, the success rates were significantly increasing with year of transplantation, and non-significant differences were found among centers.
Chronic GVHD (cGVHD) occurred in 56% patients, extensive in 42%. cGVHD incidence was 78%, 72% , 42% for Non Responders, Partial Responders and Complete Responders (Anova, p=.017, difference only significant between Complete Responders and the others). Patients with hematological malignancies were observed with significantly less cGHVD (OR=.29, p=.012).
Fifty-one patients died, 17(33%) before day 100 and 34(67%) after day 100, mostly of infections (55%) and GVHD (51%).
In conclusion, Inolimomab is well tolerated and effective for steroid-refractory aGVHD in pediatrics. Further randomized studies are requires to define the optimum regimen.