Hematopoietic Cell Transplantation as Therapy for Pediatric Patients with Relapsed Acute Lymphoblastic Leukemia with and without CNS Involvement.

Allogeneic hematopoietic cell transplantation (HCT) is the standard of care for pediatric patients with acute lymphoblastic leukemia (ALL) with early bone marrow relapse. However, the role of HCT is less clear in patients with isolated central nervous system (CNS) relapse where intensive chemotherapy followed by craniospinal irradiation is often offered. To determine the potential role of HCT, we evaluated the transplant outcomes of 116 patients with relapsed ALL with and without CNS involvement, 1–18 years of age, treated at the University of Minnesota between 1991 and 2006. Patients not in remission at transplant, and those with isolated extramedullary disease not involving the CNS were excluded. Relapse site prior to HCT was CNS in 14 patients, bone marrow (BM) in 85 patients and both marrow and CNS (BM+CNS) in 17 patients. Forty-eight underwent HCT from 1991–1995, 39 from 1996–2000 and 29 from 2001–2006. There were no significant differences among groups in median age at diagnosis (CNS: 3.7 years, range 1.2–7.7; BM: 4.5, 1–16; BM+CNS 3.7, 1.4–17), median age at transplant (CNS: 8 years, range 3.2–17.3; BM: 8.3, 3.5–17.9; BM+CNS 7.8, 3–17.9), or length of CR1 (CNS: 22.8 months, range 8.6–58; BM: 26.9, 0.8–74; BM+CNS: 34.2, 3.3–74). Remission status was similar in all groups (CNS: 7 CR2, 14 CR3+; BM: 73 CR2, 92 CR3+; BM+CNS: 14 CR2, 17 CR 3+; p=.06). Graft source was also similar between groups with 44 patients (38%) having a related donor and 72 patients having an unrelated donor (36% marrow and 25% umbilical cord blood). The majority of patients received cyclophosphamide 120 mg/kg and total body irradiation 1320–1375 cGy alone (n=69), or with etoposide (n=38) or fludarabine (n=8). Preparative regimen had no effect on outcomes. The incidence of grade II–IV GVHD (CNS 51%, 95% CI 24–68; BM 36%, 26–46; BM+CNS 18%, 1–35; p=.27) and grade III–IV GVHD (CNS 21%, 95% CI 0–42%; BM 16%, 8–24%; BM+CNS 0%; p=.21) was similar between groups. Hazard ratios using Cox multiple regression analysis demonstrated reduced survival in recipients with T cell leukemia or BCR-ABL gene rearrangement, history of marrow relapse, or receipt of HLA mismatched marrow from an unrelated donor. Patients with isolated CNS relapse had the least transplant related mortality at 2 years (CNS: 0%; BM: 35%, 95% CI 11–59%, BM+CNS: 34, 24–44; p=.05) and the lowest incidence of relapse (CNS: 0%; BM: 12%, 95% CI 0–27%, BM+CNS: 30%, 20–40; p=.01). The probability of leukemia-free survival at 5 years was greatest for patients with isolated CNS relapse (CNS: 91%, 95% CI 51–99; BM: 35, 25–45; BM+CNS: 46, 22–.68; p<.05). Similarly, the probability of overall survival was also greatest for the CNS group (CNS 86%, 33–98%; BM: 38, 27–49; BM+CNS: 53, 28–73; p<.01). These data strongly support allogeneic hematopoietic cell transplantation as a viable therapeutic option for pediatric patients with isolated CNS relapse with a very high probability of survival. Furthermore, the data indicate that CNS disease in combination with marrow relapse does not adversely effect transplant outcome compared to results expected in those with marrow relapse alone.