In Adults with Ph-Negative Acute Lymphoblastic Leukemia (ALL), Age-Adapted Chemotherapy Intensity and MRD-Driven Transplant Indication Significantly Reduces Treatment-Related Mortality (TRM) and Improves Overall Survival – Results from the Graall-2014 Trial

Background

During the 2005-2014 period, the GRAALL conducted the GRAALL-2005 trial in patients (pts) with Philadelphia chromosome (Ph)-negative ALL aged 18-59y. In this trial, all pts received a pediatric-inspired chemotherapy, whatever their age. Post-hoc analysis revealed an unacceptable TRM in pts aged 45-59y (Huguet et al. J Clin Oncol 2018; 36:2514-23). Allogeneic hematopoietic stem cell transplantation (HSCT) was offered in first complete remission (CR) to most CR pts, defined at high-risk (HR) by at least one baseline clinical or biological HR factor (CNS involvement, complex karyotype [≥5 abns.], low hypodiploidy/near triploidy, poor early blast clearance, late CR, as well as WBC ≥ 30x10⁹/L, lack of CD10 expression, KMT2A rearrangement or TCF3::PBX1 fusion for B-cell precursor [BCP] ALL). Minimal residual disease (MRD) response was not considered for HSCT indication at that time.

In the next GRAALL-2014 trial conducted in a similar population, we introduced two major changes in the treatment strategy. First, chemotherapy intensity (steroids, anthracycline and L-asparaginase) was reduced in pts aged 45-59y to decrease excessive TRM. Secondly, the indication for HSCT was modified, relying on IG/TR MRD response only (post-induction MRD ≥10^-3 and/or post-consolidation MRD ≥10^-4) while not on any former baseline HR criterion. Here, we compare the outcomes of the two trials, focusing on the impact of these two major evolutions.

Patients & Methods

A total of 743 pts treated in the GRAALL-2014 trial between 2015 and 2020 were compared to the 787 pts from the historical GRAALL-2005 trial, in terms of CR and induction death rates, rate of HSCT in CR1, cumulative incidence of TRM (CITRM) and relapse (CIR), relapse-free (RFS) and overall (OS) survival. It should be noted that two nested Phase 2 trials were introduced by amendment during the GRAALL-2014 course (in 2017 and 2018, respectively), aiming to evaluate post-remission addition of nelarabine and blinatumomab in 87 T-ALL and 94 BCP-ALL selected pts, respectively.

Results

Main patient characteristics, including age, gender, BCP/T-ALL, WBC, and historical baseline HR features, did not significantly differ between the two trial cohorts, even if CNS disease at diagnosis was more frequent in the 2014 cohort (12 vs 7%, p= 0.001).

Outcome comparisons are shown in Table 1. Overall, the induction death rate was significantly reduced in the GRAALL-2014 (3 vs 6%, p= 0.005). As expected, this was only observed in pts aged 44-59y (3 vs 11%, p= 0.001), in whom dose-intensity reduction also translated into a higher need for second induction course (9 vs 5%, p= 0.05) eventually resulting in a higher CR rate (92 vs 86%, p= 0.05). Due to the newly introduced MRD-based stratification, the rate of pts with HSCT indication was markedly reduced (30 vs 65%, p<0.001), especially in pts aged 18-44y (21 vs 40%, p<0001). Consequently, the rate of pts transplanted in CR1 dropped down from 38 to 23% (p<0.001).

The GRAALL-2014 strategy also yielded a significant reduction in CITRM (5 vs 11% at 3 years; p<0.001; Figure 1B) after CR achievement. This reduction was more pronounced in pts aged 45-59y (7 vs 17%, p<0.001 compared to 4 vs 8%, p= 0.02 in 18-44y pts). This was associated with an increased CIR (35% vs 28% at 3 years; p= 0.01), with more late relapses as depicted in Figure 1A. Even if the resulting RFS was similar in both cohorts (59 vs 62% at 3 years; p= 0.77), OS was significantly longer in the GRAALL-2014 (71 vs 64%; p= 0.002) (Figure 1C) likely due to better post-relapse outcomes. When censoring those GRAALL-2014 pts who received nelarabine or blinatumomab in CR1, observations were basically unchanged, even if the difference in CIR was even more marked.

Finally, using a 3-month RFS landmark in the 211 GRAALL-2014 pts eligible for HSCT in CR1 based on their poor early MRD response, HSCT significantly prolonged RFS (HR= 0.46 [95% CI, 0.27-0.78]; p= 0.004).

Conclusions

In adults with Ph-negative ALL enrolled in the GRAALL-2014, age-adapted chemotherapy intensity and MRD-driven indication for HSCT significantly reduced induction and post-remission TRM. This translated into a prolonged OS, indicating that this strategy was safe, despite a higher incidence of late relapses. Newly available salvage options along with HSCT in CR2 might also have played a role.