Long-term results of the Israeli National Studies in childhood acute lymphoblastic leukemia: INS 84, 89 and 98

By the 1980s, improved results in childhood acute lymphoblastic leukemia (ALL) were achieved with risk-directed combination chemotherapy and central nervous system (CNS) prophylaxis, mostly cranial radiotherapy (CRT). However, concerns about long-term toxicity, neuropsychological and secondary brain tumors prompted our group and others to seek alternative strategies for eliminating preventive CRT (pCRT), in a stepwise increasing proportions of patients.1, 2, 3, 4, 5 Since 1984, we have conducted three Israeli National Studies (INS) of childhood ALL: the INS 84, INS 89 and INS 98.6, 7 In the first study, using a randomized design, we compared the effect of pCRT with extended triple intrathecal therapy (TIT) in the B-lineage ALL low-risk (LR) or standard-risk (SR) groups (25% of total ALL population). In the next two studies, based on the Berlin-Frankfurt-Munster (BFM) backbone, pCRT was omitted in all non-high-risk B- and T-lineage ALL patients (86% of total ALL population). From August 1984 to October 2003, a total of 786 newly diagnosed patients up to 18 years of age were enrolled in these three studies (all were eligible). In all, 181 patients (23%) had T-cell ALL (T-ALL). Four patients were lost to follow-up within the first 5 years from diagnosis, and 10 patients within 10 years. In December 2008, 28, 33 and 74% of event-free survivors in the studies INS 84, INS 89 and INS 98, respectively, had been observed within 12 months. In the INS 84 study, the patients were assigned to four risk groups (Supplementary Table 1): one T-cell group and three B-lineage risk groups. They are defined as follows. LR: age 3–8 years and white blood cells (WBCs) <10 × 109/l and liver <3 cm and common ALL antigen (CALLA) positive. SR: age 1–14 years and/or WBC 10<50 × 109/l or liver 3–7 cm or CALLA negative. High risk (HR): all the remainder.6 Randomization of CNS prophylaxis between pCRT (18 Gy) and TIT × 6 versus extended TIT × 18 doses was assessed in the B-lineage LR and SR groups, in the context of reduced and intensified systemic protocols, respectively. The B-lineage HR and T-ALL patients received pCRT 18 and 24 Gy, respectively (Supplementary Figure 1), and patients with CNS involvement were treated with 24 Gy. Systemic therapy for the very restricted LR group was limited to prednisone 60 mg/m2/day × 28 days, vincristine (VCR) weekly × 4, cyclophosphamide (CTX) 500 mg/m2 intravenous (i.v.) for 2 days and Escherichia Coli Asparaginase 20 000 IU/m2/day intramuscular (i.m.) for 10 days, and continuation with 16 monthly pulses of VCR and prednisone (40 mg/m2/day × 7days) and oral methotrexate (MTX) 20 mg/m2/week and 6-mercaptopurine (6-MP) 75 mg/m2/day for 4 years. For the other risk groups, therapy was intensified: in induction by daunorubicin 45 mg/m2 × 2 and cytarabine (ARA-C) 150 mg/m2 subcutaneous × 2; and in consolidation by four pulses of VCR 1.5 mg/m2, MTX 200 mg/m2 i.v. (escalated dose by 50 mg/m2 every pulse) followed by L-asparaginase 15 000 IU/m2 i.m. (Capizzi schedule). Continuation therapy consisted of 16 monthly pulses of VCR, CTX 300 mg/m2, ARA-C 150 mg/m2 subcutaneous, daunorubicin 45 mg/m2 (total cumulative dose of 350 mg/m2) and prednisone. Maintenance was intensified for T-ALL with the following schedules of oral medications repeated every 3 months: CTX 100 mg/m2/day × 30 days, MTX 30 mg/m2 × 2 per week × 4 weeks and 6-MP 90 mg/m2/day × 30 days, for 4 years (Supplementary Figure 1). For patients with non-T ALL and WBC 100 × 109/l or T-ALL and WBC 20 × 109/l, treatment was further intensified with high-dose (HD) ARA-C and L-asparaginase, HD MTX i.v., and low-dose (LD) ARA-C with i.v. VP-16. From August 1984 to October 1989, 134 patients were enrolled in the ALL-INS 84 protocol (Table 1); 26% had a T-cell phenotype (Table 2), pCRT was indicated in a total of 75% of patients (age >1-year old): all the T-cell patients (24 Gy), and 70% of the non-T-cell patients (18 Gy). At a median follow-up of 21.8 years (range 18–24.3 years) for the 84 patients remaining alive (71 in first complete remission (CCR)), the 15-year event-free-survival (EFS) was 57.1±4.3% and the overall survival (OS) was 66.1±4.1% (Figure 1a). There was only 1 relapse after 10 years: bone marrow+mediastinum+lung in T-ALL subtype at 148 months. The cumulative incidence (CI) of isolated CNS and any CNS relapse for 15 years or more was 8.3±2.4 and 9.0±2.5%, respectively (Table 1 and Figure 1a). Among B-lineage LR ALL patients (n=11), in the context of reduced systemic therapy, four patients were randomized to receive pCRT (18 Gy) and all of them remained alive in first CR; 7 patients were randomized to receive extended TIT and only two of them are alive in first CR; three patients presented with CNS relapse; overall, 5 patients are alive at 15 years. In the SR group (n=54) with intensified treatment, patients randomized to extended TIT (n=31) had similar results to patients receiving pCRT (n=23) in terms of systemic and CNS protection: 10-year EFS 73.8±8 versus 65.2±9.9% (P=0.298); OS 90.3±5.3 versus 65.2±9.9% (P=0.061); and CI of isolated or any CNS relapse 6.5±4.5 versus 13±7.2% (P=0.41), respectively. Secondary malignancies occurred in four patients, all irradiated, and CI for the whole study at 15 and 20 years was 1.5±1.1 and 3.3±1.6%, respectively; for the irradiated (n=80) patients it was 2.5±1.8 and 5.3±2.6%, respectively. The earliest malignancy, carcinoma of the bladder, occurred 6.5 years from diagnosis in a patient in whom fibroadenoma of the breast and papillary carcinoma of thyroid developed 14 and 16 years from diagnosis, respectively. Three patients developed astrocytoma, papillary carcinoma of thyroid and breast carcinoma (7.2, 18.8 and 19.8 years since diagnosis, respectively). In addition, meningiomas were detected in three patients (two irradiated) 17–20 years after diagnosis. In the ALL-INS 89 study, treatment was based on the ALL-BFM 86 and 90 studies in 35 and 65% of the patients, respectively.7, 8, 9, 10 HD MTX (5 g/m2 i.v. × 4) of the BFM protocols, TIT × 18, without pCRT were given as CNS prophylaxis to all non-high-risk patients, both non-T and T-ALL, comprising 86% of the patients. Criteria for the HR group were prednisone poor response (PPR) on day 8 with 1 × 109 blasts/l in peripheral blood (PB) and/or no BM remission after 1 month of induction (BM M2/3 on day 33 or 41) and/or Philadelphia chromosome. All remaining patients were considered non-HR. Systemic treatment of the non-HR group (B and T lineages) (Supplementary Table 2), included i.v. VP-16 150 mg/m2 1 week before every HD MTX pulse. In the HR group (14%), IT triple instead of IT MTX was administered; pCRT 18 or 12 Gy was indicated for all HR patients. A total of 337 patients (25% T cells) were enrolled between October 1989 and December 1997 (Tables 1 and 2). Median follow-up duration for 256 survivors (203 in first CCR) was 15 years (range 11–20.5 years). Despite the lower cumulative doses of anthracyclines and alkylating agents and much larger proportion of patients who were spared preventive cranial irradiation (only 23 patients: 13 T cell and 10 non-T HR, actually received pCRT (5 with total body irradiation (TBI)), markedly improved results were achieved in the INS 89 when compared with the INS 84 (Table 1 and Figure 1b). The estimated 15-year EFS was 70.2±2.5%, OS was 75.2±2.4% and CI of isolated and any CNS relapse was 2.4±0.8 and 4.5±1.1%, respectively (Table 1). Secondary hematological malignancies, that is, acute monoblastic leukemia (AMOL), malignant histiocytosis and acute myeloid leukemia (AML), one patient each (which occurred 17, 24 and 140 months after diagnosis, respectively), were attributable to the VP-16 added to the antimetabolites. In three other patients, the preceding chronic myeloproliferative disease (Philadelphia-positive chronic myeloid leukemia (CML) in two patients) recurred while in remission of the ALL. Other secondary malignancies included Hodgkin's lymphoma, primitive neuroectodermal tumor (PNET) of the hip, papillary carcinoma of the thyroid, embryonal carcinoma of the brain, malignant melanoma and colon adenocarcinoma (at 34, 45, 105, 110, 112 and 209 months from diagnosis, respectively). In the last patient, recurrent benign meningiomas had developed a few years before the carcinoma. CI of secondary malignancies in 15 years was 3.3±1.0%. The ALL-INS 98 study was based on the ALL-BFM 95 study,11 with the same risk group criteria. TIT was used in all patients. In view of the, Associazione Italiana di Ematologia ed Oncologia Pediatrica (AIEOP) 91 report showing inferior EFS in non-irradiated T-ALL Medium Risk (MR) patients with a high WBC,12 the indication for pCRT 12 Gy was reinstated in the INS 98 for T-ALL MR patients with WBC 100 × 109/l; all other MR patients continued to receive extended TIT × 18; MR group received six pulses of VCR and dexamethasone during maintenance (Supplementary Table 2). SR patients received TIT × 11 doses. HR patients (16%) were treated with the ALL-BFM 95 HR arm or the augmented BFM (16/51 patients),13, 14 and were all candidated to pCRT −12 Gy (actually only 19 patients received pCRT (of them 8 TBI): 15 T cells, 4 non-T cells). Between January 1998 and October 2003, 315 patients (19% T cells) were enrolled. Of the 256 remaining alive (235 in first CCR), the median follow-up duration was 8.1 years (range 5.2–11 years). At 8 years the EFS was 76.5±2.4%, OS was 82.3±2.2% and CI of isolated or any CNS relapses was 1.9±0.8% (Table 1 and Figure 1c). These results did not differ significantly from the previous INS 89 trial (Table 1), even though the anthracycline and TIT doses were reduced in the SR group (30% of patients) and VP-16 was omitted with the addition of six pulses of VCR/Dexa to the MR group (54%). In the T-cell group, there was a significant decrease in relapse rate in the INS 98 compared with INS 89 (9.8±3.9 vs 29.8±5%, respectively, P=0.004) that resulted in a trend of improved (8 years) EFS (73.8±5.6 vs 61.9±5.3%, respectively, P=0.170) and OS (75.4± 5.5 vs 65.5±5.2%, respectively, P=0.279). Secondary malignancies, within the relatively short follow-up period, developed in five patients: two with early lymphoproliferative diseases after cord blood SCT and TBI and antithymocyte globulin (8 and 11 months from diagnosis), and one each with Hodgkin's lymphoma, abdominal small round-cell tumor and AML (50, 117 and 118 months from diagnosis, respectively). The CI of secondary malignancies at 8 years was 1.0±0.5% (Table 1). On univariate analysis of EFS and OS (Table 2), the most consistent adverse prognostic factors across the three trials were: within the B-lineage patients, the National Cancer Institute (NCI) HR group and high WBC count, for both B- and T-lineages in the INS 89/98, the poor early response to prednisone treatment. T-cell phenotype fared worse than the B-lineage in the first two studies, whereas in the last study the EFS and OS differences from B-lineage were almost abolished. Infants did worse, but the number of patients was small. The outcome of young adolescent patients (10 years) was poorer only for B-lineage. The influence of the main cytogenetic groups could be analyzed only in the last study, in which it was tested in a relatively high number of patients. Philadelphia chromosome was associated with a poor prognosis, whereas TEL/AML1 and high hyperdiploidy (50–65 chromosomes) had a favorable effect on OS (Table 2). In summary, in our first trial, the INS 84, we showed the efficacy of extended TIT compared with pCRT in the SR group of B-lineage ALL, in the context of intensive systemic therapy. In the subsequent studies, INS 89/98, based on the ALL BFM 86/90/95 protocols, pCRT was eliminated from >85% of the patients: all non-HR T- and B-lineage ALL. Significant improvement (in terms of EFS, OS and CNS relapse rate) was achieved when compared with our previous study, and the results were comparable with the reported studies using pCRT in a larger proportion of patients. In the last study, INS 98 (based on ALL-BFM 95), despite some reduction in therapy, the outcome in the B-lineage did not change from before, with a trend of improvement in the T-lineage, achieving similar outcome as the B-lineage. The INS studies in the 1980s and 1990s were among the first to replace cranial irradiation with CNS-directed chemotherapy. However, much remains to be studied on the potential late morbidity and adverse neurocognitive effects of HD MTX, dexamethasone and TIT,15 especially in light of the currently high cure rate of children with ALL. The authors declare no conflict of interest. This work was supported by the Israel Cancer Association. This paper is dedicated to the memory of Professor Rina-Zaizov, the founder of the Israel National Studies for childhood ALL. We thank Professor David Steinberg for his helping with the statistical analysis in earlier years, Vardit Shai and Dina Kugel for data processing and Bella Vakulenko-Lagun and Geffen Kleinstern for the statistical analyses. Supplementary Information accompanies the paper on the Leukemia website (http://www.nature.com/leu)