Overall Outcome of Relapsed Acute Lymphoblastic Leukemia Can Be Improved by Stem Cell Transplantation and Early Treatment in Stage of Molecular Relapse

Abstract 250

Survival of adult ALL has improved over the past decade to >50% with contemporary, risk adapted, targeted treatment strategies. However, published results of relapsed ALL are poor. In 4 retrospective studies with 1494 patients (pts) the CR rate after first salvage therapy was 40% and the long-term survival was only 6% including stem cell transplantation (SCT) (Thomas et al 1999, Tavernier et al 2007, Fielding et al 2007, Oriol et al, 2010).

1638 patients with newly diagnosed ALL were included in studies 06/99 and 07/03 of the German Multicenter Study Group for Adult ALL (GMALL). 547 patients with first hematologic relapse (HemRel) were evaluable for a retrospective analysis (378 chemotherapy, 169 post SCT). Median age was 33 (15–55) yrs. 432 pts had early (< 18 mo of 1st remission duration) and 115 late relapse. Salvage therapy was given according to physicians' choice and results are available in part of the pts. The aim was to achieve a CR and to perform SCT. Molecular relapse (MolRel) defined by standard criteria (Brüggemann et al, Leukemia, 2010) as reappearance of minimal residual disease (MRD) above 10-4 after prior molecular CR and measured by quantitative PCR of individual IgH/TCR rearrangements was detected in 43 pts.

Response to salvage was evaluable in 224 pts with Ph-neg ALL with bone marrow (BM) relapse (no CNS involvement) during/after chemotherapy. The CR rate after first salvage was 42% and significantly inferior in early vs late relapse (36% vs 58%; p=.003). In early relapse the most frequently used regimens were FLAG-IDA (N=38; 42% CR) or a combination of HDAC/HDMTX/VP16/VCR/DEXA (N=38; 29% CR) in B-lineage and CLAEG (N=16; 19% CR) in T-lineage. In late relapses most frequently induction was repeated (N=30; 90% CR). In pts with failure after 1^st salvage (N=82) the CR-rate after 2nd salvage was 33%. In relapse after SCT (N=48) the CR rate after 1^st salvage was 23%.

Median overall survival after relapse was 8.4 months and survival at 5 yrs 24%. Survival was significantly inferior in relapse post-SCT versus relapse on chemotherapy (15% vs 28%; p<.0001) and in BM±other vs CNS±other vs isolated extramedullary relapse (23% vs 27% vs 47%; p=.02).

Prognostic factors for survival were analysed in 291 pts with Ph-neg ALL and BM relapse (no CNS involvement) on chemotherapy. Survival was superior in late vs early relapse (43% vs 22%; p<.0001), in pts aged 15–25 yrs vs 26–45 yrs vs 46–55 yrs (38% vs 28% vs 12%; p<.0001) and in pts with CR compared to failure/PR after 1^st salvage (47% vs 13%; p<.0001). 75% of evaluable pts received SCT in any stage after relapse. Their survival at 5 yrs was 38% vs 0% without SCT (p<.0001). Survival was significantly better if SCT was performed in CR after 1st salvage vs later CR vs no CR (56% vs 39% vs 20%;p<.0001).

Of pts with MolRel (N=43) 26% remained untreated, 19% received specific salvage therapy and in 55% first-line treatment was continued without modification. 11% (N=5) remained in CCR, 30% (N=13) underwent SCT in 1st CR and 58% (N=25) developed HemRel. Median remission duration without SCT in CR1 was 92 d until HemRel and no pt was in continuous CR after 3.5 yrs. Survival after 5 yrs was significantly superior in pts with SCT in CR1 (N=13; 100%) compared to those without (N=30; 24%) (p=.0006). Overall survival after 5 yrs was significantly superior after MolRel (45%) compared to HemRel (22%) (p=.003).

Survival of pts with relapse during/after chemotherapy was 28% in the GMALL trials and superior compared to published data. Outcome after relapse was not uniformly poor but depended on prognostic factors such as age and time to relapse. The most important prognostic factor, however, was response to first salvage therapy, being very poor in early relapse. These pts suffer from chemorefractory disease and are candidates for experimental, targeted approaches. Improved overall outcome was mainly an effect of a high rate of SCT (75%) which was possible due to stringent donor search (related/unrelated) at diagnosis of relapse. Survival was significantly better if SCT was performed in CR. For the first time it could be demonstrated that outcome after molecular relapse is superior to outcome after hematologic relapse and this result underlines the relevance of MRD follow-up testing. Further improvement should therefore be achievable by early detection of molecular relapse, stratified relapse treatment, experimental approaches in early relapse and rapid transplant realisation with optimised procedures.