Outcome of Risk Adapted Therapy for Relapsed/Refractory Acute Lymphoblastic Leukemia in Children: Results From a Single Institution

Abstract 4237

Although the outcome of first line therapy for childhood ALL has improved significantly, the results of second line therapy remains suboptimal. The most important predictive factors identified for second line therapy are timing and site of relapse; early relapse and bone marrow (BM) involvement invoking poorer outcomes.

At our institution we have utilized a risk stratified therapeutic strategy for relapsed/refractory ALL which includes these factors. Re-induction for patients with first remission (CR1) duration of <18 months or those with remission failure was with high dose cytarabine (HDAraC) containing regimens, most often with idarubicin (Ida). For patients with CR1 >18 months, we utilized a standard 4-drug (prednisone, vincristine, daunomycin, asparaginase [PVDA]) induction regimen. Patients who relapsed in the BM while on chemotherapy or within 1 year of being off therapy were eligible for hematopoietic stem cell transplantation (SCT) after achieving CR, as were patients with isolated extra-medullary relapse with CR1 <18 months. Patients with late relapse were treated with post consolidation maintenance chemotherapy alone. Data collected prospectively in our database were reviewed. ALL patients who relapsed after receiving first line therapy at our institution and those who were referred to us at first relapse were included in this analysis. Only overall survival (OS) is presented as all patients who failed to achieve a CR or relapsed after achieving CR eventually died regardless of further therapeutic interventions.

Fifty-nine patients with ALL failed their first line therapy between January 1, 2005 and May 31, 2011. There were four induction failures and 55 relapses. These 59 patients included 39 males; 53 of the patients had B-cell ALL and 6 had T-cell ALL. The age at original diagnosis ranged from 0.55 to 13.8 years (mean 6; median 4.8; SEM 0.50) whereas the age at first failure ranged from 0.68 to 16.4 years (mean 8.0; median 7.8; SEM 0.50). 36/55 (65.5%) relapses occurred on first line chemotherapy, 17 (30.9%) after completion of therapy and 2 (3.6%) following SCT. For the relapsed patients, the duration of CR1 ranged from 0.09 to 9 years (mean 2.0; median 1.7; SEM 0.22); 24 (44.4%) patients relapsed less than 18 months from CR1. 36 (61%) patients had isolated BM relapse, 13 (22.1%) had isolated extra-medullary (7 CNS; 6 testicular) and the remaining 10 had BM with other sites.

Twenty four patients (40.7%) were re-induced with HDAraC/Ida (HIda), 28 (47.5%) with PVDA, 4 (6.8%) Fludarabine/AraC (FA) and 3 did not receive any second line therapy. For those who were treated 37 (62.7%) achieved a CR; 13 of the 19 patients who did not achieve CR with second-line therapy received third-line induction therapy (9 FA; 1 HIda; 1 VM26/AraC; 1 PVDA; 1 VP/CTX) and 7 (53.8%) achieved CR. The overall survival for all 59 patients at 2 years is 45.3%; for those who achieved CR the overall survival at 2 years is 64.7%.

When we compared those patients who had CR1 durations >18 months with those who were induction failures or had CR1 duration <18 months, we found no significant difference in CR2 induction rates (76.7% v. 80%; p=1.0), nor in OS (56.9% v. 34.5% at 2 years; p=0.14). CR induction rates did not change when we further categorized CR1 duration into 18 months, 18–36 months and >36 months (80% v. 72.2% v. 80%; p=0.74). However, the OS was significantly higher for the >36 month group (83.3%; p=0.023) when compared to the other 2 groups (38.7% and 38.4%, respectively). In patients with <36 months of CR1, HSCT resulted in a significantly improved OS as compared to chemotherapy alone (51.4% v. 30.9 at 2 years; p=0.05).

For those patients who achieved a CR2 there was no difference in the OS between those who underwent HSCT v. those who continued on chemotherapy (68% v. 58.5% at 2 years; p=0.33). Patients who achieved a CR after 2^nd line therapy fared much better than those who achieved a CR after 3^rd line therapy (25/36 [69.4%] alive v. 2/7 [28.6%] alive; OS at 2 years 68.3% v. 29.2%; p=0.06).

In conclusion, we believe that this risk stratified approach to the treatment of relapsed/refractory ALL in children is effective. While the majority of the patients can be re-induced into remission, maintenance of the remission is dependent on the duration of CR1; patients with late relapse do as well as treatment naïve patients. Increased intensity induction therapy for early relapses and SCT for patients with <36 months of CR1 may be beneficial.