Prognostic Value of Minimal Residual Disease by Real-Time Quantitative PCR in AML with CBFB-MYH11 Rearrangement: The French Experience.

Background. Despite the favorable prognosis of AML patients with inv(16)/t(16;16) leading to CBFB-MYH11 rearrangement, relapses still occur in about 40% of the cases. With the possible exception of receptor tyrosine kinase mutations, no pretreatment-factor can strongly predict risk of relapse.

Methods. To explore minimal residual disease (MRD) prognostic impact, we prospectively monitored CBFB-MYH11 rearrangement by real-time quantitative PCR (RQ-PCR) in patients with inv(16)/t(16;16) treated in French cooperative trials; most patients were treated in Acute Leukemia French Association (ALFA) trials and Leucémies Aiguës Myéloblastiques de l’Enfant (LAME) Cooperative Groups trials. RQ-PCR was performed in three molecular French labs (CHU of Lille, Necker Hospital and St-Louis Hospital) according to the EAC procedure and results are expressed as CBFB-MYH11/100 ABL copies(%).

Results. 61 AML patients with inv(16)/t(16;16) who had reached CR with one of the anthracyclin-aracytin regimens were analyzed. Median age was 37.5 years (range 2.1–76.5) and median baseline WBC 46 G/L (range 1.8–246). With a median follow-up of 23.3 months, median overall DFS was 23 months and median OS not reached (93.5% of patients alive). No initial clinical or hematological characteristic including age, gender and initial WBC was predictive of relapse in adults, but children had poorer DFS than adults (17.6 months vs not reached, p=0.03). Regarding chromosomal abnormalities in addition to inv(16)/t(16;16), trisomy 22 was the most frequent but no significant prognostic impact was observed. Of the 61 patients, MRD monitoring could be performed sequentially in 45. Pretreatment CBFB-MYH11 transcript level had no impact on DFS. However, after induction therapy, transcript levels >0.5% were significant predictors of poorer DFS (median 16.4 months vs not reached if <0.5%, p=0.002). Likewise, after first consolidation therapy course, transcript level >0.001% was also predictive of poorer DFS (median 22 months when >0.001% vs not reached when <0.001%; p=0.03). 34 paired analyses of bone marrow (BM) and peripheral blood (PB) samples revealed only moderate correlation (R²=0.86); especially for low MRD levels, CBFB-MYH11 transcript expression being generally higher (up to 78-fold) in BM samples than in PB samples.

Conclusion. MRD monitoring by RQ-PCR appears to be a major prognostic factor of DFS in AML with CBFB-MYH11 rearrangement and could be a powerful tool to early identify good and bad responders which could have an impact on therapeutic decisions for consolidation therapy.

View largeDownload slide

DFS according to MRD level after first consolidation treatment

View largeDownload slide

DFS according to MRD level after first consolidation treatment

Close modal