Efficacy and Safety of Imatinib on Top of BFM-Like Chemotherapy in Pediatric Patients with Ph+/BCR-ABL+ Acute Lymphoblastic Leukemia (Ph+ALL). the EsPhALL Study
Abstract
Background. Philadelphia chromosome positive (Ph+) ALL accounts for 3–5% of pediatric ALL. An international survey on 640 children diagnosed between 1995 and 2005 and treated with chemotherapy and stem cell transplantation (SCT) without the use of tyrosin kinase inhibitors (TKI) recently reported an overall 7-year event free survival (EFS) and overall survival (OS) of 31.2% and 44.2%, respectively. In those years, only limited experience was accumulated on the use of Imatinib (IM) for children with Ph+ leukemia. The EsPhALL study was designed as an intergroup, open-label, randomized Phase II/III study, within the I-BFM-SG network, to assess the safety and efficacy of IM in association with chemotherapy. Ten national study groups participated in the study: AIEOP, BFM-G/CH, COALL, FRALLE, NOPHO, MRC, DCOG, CPH, PINDA and HONG KONG. Methods. Patients 1 to 18 years of age diagnosed with Ph+ ALL were eligible to the study. After the induction phase according to national treatment protocol, patients were classified as Good Risk (GR) or Poor Risk (PR) according to their response to treatment. GR patients were those who achieved both the early response (i.e. blast cell count <1000/ml in peripheral blood after 7 days of prednisone and a single dose of intrathecal methotrexate or M1/M2 bone marrow at day 15 or M1 at day 21) and the complete remission after the frontline Induction course (1st complete remission, CR1). They were randomized to receive IM in combination with chemotherapy (GR-IM) or chemotherapy alone (GR-noIM). PR patients (those who did not achieved early response or CR1 or both) received IM in combination with chemotherapy. Due to the availability of external evidence, the randomization in GR was stopped in 2009 and an amended trial started, with all patients receiving IM continuously. The chemotherapy regimen was modeled upon a BFM high risk backbone and IM was delivered at the dose of 300 mg/m2/day. SCT in CR1 was recommended for all PR patients (any donor) and for GR patients if a genotype-matched donor (9/10 or 10/10 alleles) was available. The primary analysis for the randomized question (ITT) was on disease-free survival (DFS) in GR and EFS in PR patients, not censoring for SCT in CR1, with comparison based on the log-rank test. Results. Between 01-Jan-2004 and 31-Dec-2009, 178 patients (age 1.5–17.9 years) were enrolled and stratified as GR (108; 61%) or PR (70; 39%). Ninety GR patients were randomized (18 excluded for parental refusal or clinical decision); of these, 77% underwent SCT in CR1. Out of 35 PR patients who were resistant to Induction, 80% achieved CR1 after consolidation (Phase IB). 84% of PR patients received SCT in CR1. Relapse was the first event in 23 (33%), 12 (27%) and 10 (22%) of PR, GR-noIM and GR-IM patients, respectively. The most common site was the bone marrow (74%, 92% and 60% in PR, GR-noIM and GR-IM patients, respectively). Deaths in CR1 were 8, 4 and 2 in PR, GR-noIM and GR-IM patients respectively, none being related to IM. The 4-year DFS was 73% (95% CI: 56% – 84%) in the GR-IM arm and 62% (95% CI: 45% – 75%) in the GR-noIM arm (p=0.24), with a 4-year OS of 85% (95% CI: 70% – 93%) and 73% (95% CI: 54% – 85%), respectively (p= 0.37). A secondary ‘as treated' analysis was performed accounting for 13 patients who switched from GR-noIM to GR-IM, with 4-year DFS of 56% (95% CI: 36% – 72%) and 75% (95% CI: 61% – 85%), respectively (p=0.06). The EFS in PR patients was 54% (95% CI: 40% – 65%) at 2-years and remained constant through 3 and 4 years, with a 4-year OS of 64% (95% CI: 50% – 74%).The most frequently reported adverse events (AEs) across the treatment arms were decreased leukocytes, platelets and granulocyte counts, decreased hemoglobin, and infections. There was no significant difference in the overall frequency of AEs across all 3 patients' groups. Severe adverse events rate was 28% in GR-IM group, 32% in GR-noIM and 34% in PR group. Conclusions. Results suggest that the addition of IM to intensive BFM-type chemotherapy regimens was associated with an approximate 10% advantage in long-term DFS in GR patients which, however, the study was not powered to detect. The PR group treated with IM had improved EFS and OS as compared to historical controls. IM was generally well tolerated on top of intensive chemotherapy with a reassuring safety profile.