Response to High Dose Cytarabine (HIDAC) As First Salvage for Relapsed Acute Lymphocytic Leukemia in Patients Receiving HIDAC As Initial Therapy

Consecutive pts with ALL in first relapse treated with HIDAC-containing regimens either at the Cleveland Clinic (CC) between the years 1993–2010 or at any institution participating in SWOG trial S9030 (HIDAC 3000 mg/m² Days 1–5, mitoxantrone 80 mg/m² Day 1) (1992–1993) were included. HIDAC was defined as a cycle of at least 3000 mg/m² × 5 days. Remission was defined according to standard criteria. The outcome analysis [CR and overall survival (OS)] was adjusted for the following factors: age, WBC at diagnosis, cytogenetic (CG) risk, immunophenotype, transplant, and prior HIDAC exposure.

Sixty-six pts were included (39 treated at CC, and 27 as part of SWOG protocol S9030). All pts received a vincristine/prednisone/anthracycline/steroid-based induction regimen (S8417, CALGB 19802, CALGB 8811) except for 1 pt who was treated with hyperCVAD. Seventeen pts treated at CC had HIDAC incorporated into their initial treatment (1: hyperCVAD; 16: CALGB 19802), but none of the SWOG pts did. The median age was 35 yrs (range 17 to 73). The median WBC at the time of diagnosis for CC pts was 21.4 K/uL (range 0.5–260.0) and median WBC at the time of study registration for SWOG patients was 17.6 K/uL (range 0.4–198.4). Three pts (5%) had a mixed (B/T) lineage leukemia. Three patients had lymphoblastic lymphoma (1 B-cell; 2 T-cell) at the time of initial diagnosis, and had ALL at the time of relapse. For the 39 CC pts, the median time from diagnosis to relapse was 12 mos (range 1–55 mos). CG risk was ascribed by CALGB criteria. Of the 50 pts with evaluable pre-study CG, 20 pts (40%) had normal CG, 18 (36%) miscellaneous, and 12 (24%) poor risk CG. Twenty pts (30%) received HIDAC alone, and 46 (70%) received HIDAC in combination with other drugs for relapsed ALL. The CR rate for all relapsed pts was 32% (CC 36% and SWOG 26%) and was not affected by the addition of other drugs to HIDAC. Twenty-nine patients (44%) were able to proceed to HSCT; and the median OS was 5.4 mos (95% CI: 4.8–6.0 mos). After adjusting for all baseline and demographic factors, the CR rate and OS between pts receiving or not receiving HIDAC during initial treatment was not significantly different. Five of 17 (29%: 95% CI 10%-56%) pts with prior exposure to HIDAC achieved CR while 16 of 49 (33%: 95% CI 20%-48%) pts without prior HIDAC exposure achieved a CR (p=0.80). The 1 year OS (from salvage) for pts treated with HIDAC for relapse who also were treated with prior HIDAC was 12% (95% CI: 0%-27%) and for pts not treated with prior HIDAC was 33% (95% CI: 20%-46%)(p=0.17). Since additional information was available on the CC pts, additional analyses were performed on this subgroup of pts. With the exception of lymphoblastic lymphoma at the time of diagnosis, no other factors correlated with achievement of CR. Achievement of CR was the only factor associated with proceeding to HSCT (79% vs. 36%, p=0.01). Variables associated with improved OS included: lymphoblastic lymphoma at the time of diagnosis (p=0.04; 2 of the 3 pts are still being followed at 80+ and 96+ mos), achievement of CR (p=0.0001), longer remission (> 30 mos, p=0.005), and transplantation (p=0.0001).

The outcome of relapsed ALL with HIDAC salvage therapy is dismal, regardless of prior HIDAC exposure; and novel treatments are needed. There was a suggestion that the OS of pts with prior HIDAC exposure may be lower, but further study of 1 year OS with larger pt numbers will be needed to evaluate this. An interesting finding in this study was the favorable outcome of pts with lymphoblastic lymphoma at diagnosis, who subsequently relapsed in the leukemic phase and were treated with HIDAC. However, few pts carried this diagnosis and a larger number of pts are required before drawing firm conclusions.

No relevant conflicts of interest to declare.