Comparison of MRD monitoring by IG/TR rearrangement and BCR/ABL1 transcript in PH+ pediatric acute lymphoblastic leukemia patients treated with imatinib in the EsPhALL study

Background. In the European Study for Pediatric Ph+ ALL (EsPhALL), a total of 128 patients aged 1 to 18 years diagnosed with ALL and evidence of t(9;22)(q34;q11) received post-Induction Imatinib on top of BFM high risk chemotherapy backbone, either in Good Risk (GR) (after randomization, n=58) or in Poor Risk (PR) (n=70), as defined according to early response to the National induction treatment. Although not used for treatment decision, MRD was prospectively measured at relevant time-points (TPs) by real-time quantitative PCR analysis of both rearranged immunoglobulin (IG) and T-cell receptor (TR) genes and BCR/ABL1 (B/A) transcript. Previous studies reported a different levels of concordance of the two methods, which could underline biological differences and might have implication in the clinical practice Aim. To define the concordance of MRD measured at DNA (IG/TR) and RNA (B/A) levels in a large clinically controlled study by applying consensus methods Methods. MRD was analyzed in national reference laboratories by real-time quantitative PCR of rearranged IG/TR genes as well as B/A fusion transcript, on subsets of study patients with available material at different TPs during therapy: 50/128, 39% at TP1 (end of IA), 56/128, 44% at TP2 (end of IB), 40/128, 31% at TP3 (end of HR1 Block), 37/128, 29% at TP4 (end of HR2 Block) and 13/128, 10% at TP5 (end of HR3 Block). Overall, 99/128 (77%) patients had at least one MRD measurement within TP5, and IG/TR versus B/A comparison was performed in 196 samples. Common guidelines developed within the EuroMRD consortium were applied for performing both methods and interpreting MRD results. Results were considered discordant in case of at least 1 log difference if both positive; similarly, in case of negative result by one of the two methods, discordance was called if the difference between the positive result and the sensitivity of the other method was more than 1 log Results. Overall, the two methods were concordant in 70% of tests (n=137) (see figure), with the lowest concordance at TP3 (65%) and the highest at TP5 (77%). The vast majority of the 59 discordances reflected B/A results higher than IG/TR (47, 80%). IG/TR results tend to reduce earlier than B/A, as indicated by persistent positivity by B/A only at later TPs. Indeed, at TP1 (before Imatinib administration) 3 cases (6% of analyzed) were negative by IG/TR and positive by B/A, and 2 vice versa (4%). At later TPs, no more cases were found negative by B/A and positive by IG/TR, while 6/56 cases (11%) were B/A+ and IG/TR- at TP2, 7/40 (17%) at TP3, 6/37 (16%) at TP4, and 2/13 (15%) at TP5. Out of the 3 cases already IG/TR- and B/A+ at TP1, 2 were evaluated by 2 IG/TR markers and were persistently positive by B/A at subsequent TPs (7.3E-02 to 1.1E-01 at TP4); however, all 3 are alive in CCR, 2 after HSCT (1 GR and 1 PR) and 1 (GR) after Imatinib + chemo only. On the other hand, of the 6 cases that were IG/TR- but B/A+ at TP4, 3 were found negative by IG/TR already at TP1 and are described above. The remaining 3 progressively reduced tumor load by IG/TR and were all negative at TP3. One was GR and was alive in CCR after HSCT, while 2 were PR, one relapsed and one died in CCR, both after HSCT Relapses in PR cases discordant at TP2 (2/10, 20%) (after the first exposure to Imatinib) were less frequent than in concordant cases (8/23, 35%), but were higher for discordant at TP3 (6/9, 67% vs 2/12, 17%) and were not different at TP4 (2/6, 33% vs 4/15, 27%) (too few data at TP5) If MRD would be used as a criteria for eligibility to HSCT (MRD≥5x10-4 at TP2 and if MRD