Low Rates of Toxicity, GVHD, and Relapse Using Sirolimus (SRL)-Based GVHD Prophylaxis in Pediatric Related and Unrelated Transplant Recipients with High-Risk ALL.

Relapse after allogeneic SCT remains a significant risk for children with ALL. In-Vitro and NOD-SCID xenogeneic models have shown significant activity of sirolimus (SRL) against human ALL blasts, resulting in apoptosis and cell death at serum levels commonly used for immune suppression. We hypothesized that SRL would decrease relapse after allogeneic transplantation for ALL, as well as contribute to prevention of GVHD. We conducted a pilot trial at four pediatric centers designed to assess the feasibility and toxicity of SRL-based GVHD prophylaxis using multiple stem cell sources. After a preparative regimen of TBI (1200cGy in 6 fractions over 3d), Thiotepa (5mg/kg/d x 2), and cyclophosphamide (60mg/kg/d x 2), recipients received continuous infusion tacrolimus starting on day -2 (target level 5–10ng/mL), PO sirolimus starting on day 0 (2.5mg/m², max 4mg daily, target level 3–12ng/mL), and IV methotrexate (5mg/m², d+1, +3, and +6 for all recipients, d+11 for recipients of UD BM/PBSC). Tacrolimus was given orally when tolerated, and tapered between d+42–96 for MSD and d+100–180 for mismatched related, UD or UCB. Sirolimus was tapered at 6 months after transplant over 1 month. Thirty patients were enrolled, with median age of 9 yrs (1–18). Immunophenotypes included pre-B 23, T-cell 6, and MLL 1. Five patients were in CR1 (Ph+ 2, failed remission 2, 2° ALL 1), 16 were in CR2 (9 CR1 <18 months), and 8 were in CR3. Stem cell sources included: UCB 15, MSD-BM 13, 5/6 mismatched RD-BM 1, 9/10 mismatched UD-PBSC 1. SRL was successfully administered to all pts and therapeutic drug levels achieved. PK studies were obtained and are being analyzed. Three patients had SRL held briefly for high levels or toxicity and three patients stopped the medication due to toxicity (VOD/TAM). Engraftment occurred in 28 pts at median of 20d (13–62); one CB recipient engrafted successfully after a second infusion and a second patient relapsed prior to engraftment. Overall grade II–III aGVHD occurred in 11/30 pts (37%) including 3/14 MRD (21%) and 8/16 CB/UD (50%, 38% grade III). No grade IV aGVHD occurred and the grade III aGVHD (all diarrhea) was rapidly controlled with prednisilone in all patients. 3/18 (17%) of evaluable patients developed chronic GVHD. Relapses occurred in 5 pts at a median of 94 days (34–362); 3 were T cell, 1 was CR3. TRM occurred in 3 patients (10%: MSOF/sepsis/VOD in 2 pts d+34 and d+39, cGVHD/infection in 1pt d+789). Other significant toxicities included one case of IPS with engraftment requiring intubation (resolved), 1 case of nonfatal VOD, and two cases of reversible TAM. With a median f/u of 7m (1–36m) the 1 year EFS and OS are 68.5% (SE 11.5) and 77.1% (SE 8.6), respectively. In this very high risk cohort of children with ALL, we have demonstrated that GVHD prophylaxis with SRL, tacrolimus, and methotrexate is feasible using multiple stem cell sources, resulting in low rates of TRM, acute and chronic GVHD, and relapse. A randomized phase III trial comparing this approach to tacrolimus/methotrexate in children with CR2 ALL is currently being planned through the Children’s Oncology Group.