Ig/TCR Monitoring of High Level MRD within the GRAALL-2003 Adult ALL Trial Has More Impact in B Than T Lineage Cases.

Abstract

The increasing number of prognostic markers available at diagnosis or during the early response to remission induction in ALL makes evaluation of their relative pertinence necessary, although it is largely protocol-dependent. Quantification of clonal Ig/TCR rearrangements has recognised prognostic value in pediatric and adult ALL with excellent informativity but variable sensitivity (10-3 by competitive genescan (GS), 10-4/10-5 by Q-PCR with CDR3 specific primers (CDR3 Q-PCR)) and is rather complex to perform. The proportion of patients with MRD greater than 10-2/10-3 at the end of induction is usually higher in adult than pediatric ALL and in T vs. B cell precursor (BCP) ALL. Modification of the French LALA94 and GOELAMS adult ALL trials towards the pediatric-inspired GRAALL03 protocol led to a significantly improved outcome, particularly in T- and younger BCP-ALLs (Huguet et al. this meeting). GRAALL03 for BCR-ABL neg. patients included reinforced induction and/or allo-SCT for patients with baseline poor risk markers (PRM) defined as WBC>30G/L in B-ALL, CNS+, MLL-AF4, E2A-PBX1, haploidy/near triploidy and/or poor response to induction, through corticoresistance (CoR) and chemoresistance (ChR) defined morphologically on BM. Comparable Ig/TCR quantification by GS and CDR3 Q-PCR was demonstrated for 179/189 FRALLE and LALA ALLs, and post-induction MRD was assessed in GRAALL03 by either technique for 121 adults (81 B, 40 T). Cases were classified as MRD+ and intensified as above if >10-2, as MRD+/− at 10-2/10-3 and MRD- if <10-3. Overall, 12/81 (15%) BCP-ALL were MRD+, compared to 5/31 (16%) in the preceding BCR-ABL negative LALA94 cohort. For T-ALL, 4/40 (10%) GRAALL patients were MRD+, compared to 7/17 (41%; p=0,006) LALA94 patients, consistent with the preferential improved clinical response in T-ALL. In GRAALL03 BCP-ALL, only 2/25 cases (8%) with no PRM were MRD+ and the 3 relapses observed in this group of 25 patients (12%) were MRD+/− or MRD-. None of the 16 baseline PRM+ but CoS/ChS patients were MRD+ and again all 7/16 (44%) relapses were MRD+/− or MRD-. Of the 40/81 CoR and/or ChR cases, 10 (40%) were MRD+, of which 7/10 (70%) relapsed and 5 died, compared to only 5/25 (20%) relapses and 5 deaths (20%) amongst MRD- CoR/ChR patients (p=0,0048). Of the 7 intermediate MRD+/− B lineage ALL, 5 were CoR/ChR and 5 relapsed, notably including both CoS/ChS cases. In GRAALL03 T-ALL, only 1/21 (5%) PRM negative cases was MRD+ and was identified as having TLX3/NUP214-ABL. Amongst PRM+ cases, 3/19 (16%) were MRD+ and 3 MRD+/−. Overall, 10/40 T-ALLs relapsed but the incidence correlated with neither the MRD status (3/7 amongst MRD+ or +/− T-ALL, compared to 7/33 MRD-, p=0,23) nor CoR/Chr (6/16 relapses; 38%) vs. CoS/ChS (4/21; 19%, p=0,16). We conclude that, in BCP-ALL, detection of high level Ig/TCR MRD (10-2) identifies a particularly poor prognostic group within the (frequent) CoR/ChR category but does not correlate with the baseline PRM used here, nor detect high risk cases in the CoS/ChS group. MRD high levels should be defined at >10-3 within the GRAALL but re-evaluated with techniques also detecting MRD <10-4 cases. The pertinence of Ig/TCR MRD evaluation is less clear in GRAALL T-ALLs. Given the large number of novel immunological and genetic markers in T-ALL, the place of MRD monitoring should be re-evaluated independently from B lineage ALL.