Pegaspargase-modified risk-oriented program for adult acute lymphoblastic leukemia: results of the GIMEMA LAL1913 trial

Abstract

Pediatric-inspired chemotherapy is the standard of care for younger adults with Philadelphia chromosome-negative acute lymphoblastic leukemia/lymphoma (Ph- ALL/LL). In LAL1913 trial, the Gruppo Italiano Malattie EMatologiche dell'Adulto added pegaspargase 2000 IU/m2 to courses 1, 2, 5, and 6 of an 8-block protocol for patients aged from 18 to 65 years, with dose reductions in patients aged >55 years. Responders were risk stratified for allogeneic hematopoietic cell transplantation (HCT) or maintenance per clinical characteristics and minimal residual disease (MRD). Of 203 study patients (median age, 39.8 years), 91% achieved a complete remission. The 3-year overall survival, event-free, and disease-free survival (DFS) rates were 66.7%, 57.7%, and 63.3%, respectively, fulfilling the primary study end point of a 2-year DFS >55%. Although based on the intention-to-treat, the DFS being 74% and 50% in the chemotherapy (n = 94) and HCT (n = 91) assignment cohorts, respectively, a time-dependent analysis proved the value of HCT in patients who were eligible (DFS HCT 70% vs no HCT 26%; P <.0001). In multivariate analysis, age and MRD were independent factors predicting DFS rates of 86% (age ≤ 40 and MRD-negative), 64%-65% (MRD-positive or age > 40) and 25% (age > 40 and MRD-positive); P < .0001. Grade ≥2 pegaspargase toxicity was mainly observed at course 1, contributing to induction death in 2 patients but was rare thereafter. This program improved outcomes of patients with Ph- ALL/LL aged up to 65 years in a multicenter national setting. This trial was registered at www.clinicaltrials.gov as #NCT02067143.

An updated risk-oriented strategy resulting in 3-year OS and DFS rates >60% in adult patients 18-65 years with Ph-negative ALL

Figure 1.

Study strategy and treatment elements. Allo-SCT, allogeneic stem cell transplantation; FAR, folinic acid rescue; HD, high dose; IM, intramuscular; IT, intrathecal; PO, oral; SC, subcutaneous; w, week; u/k, unknown.

Figure 2.

Study flowchart. Patient disposition and outcome are shown overall and per risk-oriented treatment, by ITT, and as treated. Patients displaying refractory ALL after 2 treatment cycles were off study. Application of allogeneic HCT and chemotherapy based on patient age in the respective risk-oriented therapy cohorts is detailed in Supplemental Table 4; overall, patients aged from 18 to 40 years were more likely to complete the assigned chemotherapy steps, whereas HCT rates did not differ significantly across age groups. CCR, continuous first CR.

Figure 3.

Main outcome results. (A) OS: median was not reached; 3-year rate, 66.7% (95% CI, 60-74); (B) EFS: median was not reached; 3-year rate, 58% (95% CI, 51-65); (C) DFS, representing the primary study objective compared with prior GIMEMA study LAL 0904: median was not reached; 2-year rate, 70% (95% CI, 63-77) vs 45% (95% CI, 39-51); and 3-year rate, 63% (95% CI, 56-71) vs 38% (95% CI, 38-44), P < .0001; (D) 3-year DFS per ITT risk-oriented therapy: chemotherapy, 74% (95% CI. 65-83), allogeneic HCT, 50% (95% CI, 39-63), P = .0022; (E) 3-year DFS in the ITT allogeneic HCT group per time-dependent HCT realization: HCT, 75% (95% CI, 55-89) vs no HCT, 26% (95% CI, 15-45), P < .0001; (F) Cumulative incidence of TRM during induction (ID) and CR, and of resistance/relapse (Res/Rel) based on B- or T-ALL/LL diagnosis: 3-year incidence ID/CR death B-ALL/LL, 17.8% (95% CI, 11.3-24.3) vs T-ALL/LL, 4.8% (95% CI, 0-10.1), P = .0127; Res/Rel B-ALL/LL, 28.7% (95% CI, 20.9-36.6) vs T-ALL/LL, 28.1% (95% CI, 16-40.4), P = .773.

Figure 4.

Prognostic analysis 1. Forest plots from univariate and multivariate prognostic analysis for OS (A-B) and DFS (C-D), including major risk factors, MRD results, and risk-oriented therapy. The multivariate model analysis was performed on data from 149 patients with no missing values. All covariates were evaluated in univariate models and all relevant variables with univariate association within P < .15 were considered in the multivariate models. To compare the prognostic ability of multivariate models with the contribution of each variable, the Akaike information criterion was used to compare the models’ goodness of fit with the data. The final model includes the actual therapy received and not ITT therapy. The collinearity between treatment received and risk classification was evaluated using an interaction term into the multivariate model (resulting nonsignificant). Cyto, cytogenetics/genetics; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio.

Figure 5.

Prognostic analysis 2. (A) 3-year DFS per patient age (years): ≤40, 78% (95% CI, 70-87); 41 to 55, 49% (95% CI, 37-66); > 55, 44% (95% CI, 30-66); P = .00052. (B) 3-year DFS per MRD risk model (n = 151 evaluable): MRD_neg, 77% (95% CI, 70-86), MRD_pos, 41% (95% CI, 26-64), P < .0001). (C) 3-year DFS per patient age (years) group and MRD risk model interactions: 18 to 40/MRD_neg, 86% (95% CI, 78-95) vs >40/MRD_pos, 65% (95% CI, 52-82) vs 18 to 40/MRD_pos, 64% (95% CI, 43-95) vs >40/MRD_pos, 25% (95% CI, 11-59); P < .0001. (D) 3-year EFS per Ph-like ALL gene signature in 88 patients with B-ALL who were evaluable: Ph-like, 23% (95% CI, 10-49) vs non–Ph-like, 68% (95% CI, 57-81). P = .00049.

Figure 6.

Pegaspargase-related outcome and toxicity analysis. (A) 3-year DFS per the cumulative pegaspargase dose (IU/m²) administered at cycles (C) 1 and 2: ≤6000 IU/m², 60% (95% CI, 50-72) vs >6000 IU/m², 64% (95% CI, 55-76), that is, below or above the total median dose received at C1 and C2; the higher dose usually meant the full administration of the first 2 planned protocol doses. (B) 3-year DFS per BMI: BMI of ≤30, 63% (95% CI, 55-71) vs >30, 73% (95% CI, 53-100), P = .31. (C) Correlative analyses between BMI, cumulative pegaspargase dosing at cycles C1 and C2, and occurrence of grade (G) ≥2 severe adverse events (SAEs) in individual patients (nonsignificant P values). (D) Correlative analyses between BMI, cumulative pegaspargase dosing at C1 and C2, and SAE grading (nonsignificant P values). (E) Correlative analyses between BMI, cumulative pegaspargase dosing at C1 and 2, occurrence of SAE G ≥2, and age groups (nonsignificant P values). Scatter plots in panels C-E with linear model function to depict the correlation index.