High Post-Remission WT1 Expression Is a Predictive Marker for Subsequent Molecular Relapse and Poor Survival Outcome in Adult Patients with Acute Promyelocytic Leukemia (APL)

Background: Acute promyelocytic leukemia (APL) is classified into favorable-risk group and long-term overall survival (OS) is estimated at around 80%. Relapse rate of APL is relatively lower than another acute myeloid leukemia (AML) subtypes, but we still confront relapse in 10-20% and some relapsed cases are hardly cured even after hematopoietic cell transplantation (HCT). Therefore, it is important to find out patients with high-risk of relapse and early intervention should be considered. In APL, PML-RARa RQ PCR is used as a marker for residual disease, but the marker is not useful for pre-emptive management for relapse prevention because its positivity directly indicates relapse of disease. WT1 expression is a well-known marker in AML, and the expression is higher in APL than the other AML subtypes (Cilloni et al., leukemia, 2002). We monitored WT1 decrement along the treatment courses to identify its significant role as a marker for relapse prediction.

Methods: In this study, 117 APL patients with a median follow-up duration of 38.5 months (range, 9.7-81.3) from 2008 to 2014 were analyzed. APL was diagnosed by RT-PCR method for detection of PML-RARa and all patients were available with cytogenetic results. All except 5 with normal karyotype was identified with t(15;17)(q22;q21) and 33 (28.2%) showed combination of additional chromosomal abnormalities. Our treatment protocol was based on LPA99 trial using ATRA and idarubicin monotherapy (Sanz et al., Blood, 2004). In relapsed patients, we applied ATO and some high-risk patients were treated with HCT (n=3). PML-RARa and WT1 expression in the BM samples were quantified by RQ-PCR method, and we used WT1 ProfileQuant¢â kit (Ipsogen) for WT1 monitoring. We measured RQ-PCR levels at diagnosis, post-induction and each of the 3 post-consolidation chemotherapies, and 3 months interval after starting maintenance therapy. FLT3-ITD/TKD mutation was evaluated by multiplex allele-specific PCR and concomitantly analyzed. Significant cut-off level of PML-RARa and WT1 expression was calculated by ROC curve analysis. According to the level, we calculated OS, disease free survival (DFS), and cumulative incidence of relapse (CIR).

Results: Hematological complete remission (CR) was identified in 117 (100.0%) patients but complete molecular remission (CMR) was identified in 68 (42.7%) after induction. Among 49 patients who failed to achieve CMR, 44 patients achieved CMR after 1^st consolidation and 5 patients after 2^nd consolidation. Three-year OS and EFS was 92.5% and 82.0%, and CIR rate was 14.7% (n=13). Three patients showed clonal evolution to therapy-related AML and 1 patient died in CR due to lung cancer. FLT3-TKD and FLT3-ITD mutation was identified in 6 (5.1%) and 25 (21.4%) patients, and PML-RARa BCR3 and BCR1 subtype was identified in 32 (27.4%) and 85 (72.6%) patients, respectively. For relapse prediction, we analyzed WT1 expression at several time points in association with CMR after induction, FLT3-ITD/TKD mutation, BCR subtype, and hyperleukocytosis at diagnosis and during first chemotherapy. High WT1 expression (>120 copies/10⁴ABL1) in early period (3 months) after maintenance therapy significantly predicted subsequent relapse. All paired PML-RARa RQ-PCR was not detected except one sample with early relapse after 3^rd consolidation. Patients with high post-maintenance WT1 expression (n=40) showed significantly higher CIR rate (30.7% vs. 4.2%, p=0.0003) and inferior 3-year OS (86.1% vs. 97.9%, p=.0103) and DFS (62.8% vs. 94.1%, p<.0001). Multivariate analysis revealed high leukocyte counts [HR=9.2, 95%CI=2.2-38.5, p=.002], WT1 at 3 months post-maintenance [HR=8.7, 95%CI=1.9-38.9, p=.0051], and FLT3 mutation [HR=5.4, 95%CI=1.5-19.4, p=.0092] were significant factors for relapse prediction. However, even in the BCR3 or FLT3-positive subgroup (n=48), low WT1 at post-maintenance 3 months was associated with lower CIR rate (13.4% vs. 50.8%, p<.0001) and better DFS (95.2% vs. 49.2%, p<.0001).

Conclusion: High post-remission WT1 expression was a reliable marker for the prediction of subsequent relapse in APL, even when PML-RARa was not detected at 3 months post-maintenance. In this high-risk group, early intervention with ATRA±ATO, WT1 vaccination or WT1-specific cytotoxic cell therapy may be used for relapse prevention. The role of WT1 expression needs to be validated by prospective studies in a large cohort.