BCL2L10 Quantification Is a Predictive Factor of Response to Azacitidine in Myelodysplastic Syndromes (MDS) and Acute Myeloid Leukemia (AML)

Abstract

▪Background:Azacitidine (AZA) is the reference treatment for higher-risk MDS patients ineligible for intensive chemotherapy (IC) (Lancet Oncol 2009). It also improves overall survival (OS) in elderly AML patients with more than 30% marrow blasts ineligible for IC over conventional care (Dombret et al., EHA 2014). To date, no reliable biological marker predictive of AZA response has been reported. In a preliminary retrospective work in 32 patients, we found that quantification of BCL2L10 (an anti-apoptotic member of the Bcl-2 family of proteins) bone marrow mononuclear cells (BMMC) positive cells by flow cytometry (FCM) in HR-MDS patients represents a new potential biomarker for AZA response (Cluzeau et al. Oncotarget 2012). The aim of the present study was to validate those preliminary findings in a larger prospective multicenter cohort, analyzed blindly in 2 different laboratories.Methods:FCM was performed on fresh BMMC obtained at different times during AZA treatment: at treatment onset, after 3 or 6 cycles of AZA and at relapse, as previously described (Cluzeau et al., Oncotarget 2012) after several steps of fixation, permeabilization, and consecutive treatment with i) an anti-BCL2L10 antibody (Cell Signaling) and ii) a donkey anti-Rabbit FITC-antibody (Santa Cruz). All assays were performed in two different laboratories with two kinds of cytometers: Paris (Canto Becton Dickinson), Nice (Miltenyi Biotec). MDS and AML patients treated with AZA were prospectively included from 6 centers in this correlative study (clinicaltrial.gov: NCT 01210274). Response was assessed by IWG 2006 criteria for MDS or by Cheson et al (2003) for AML.Results:75 MDS or AML patients were included. Median age was 73 years (range 35-91) and M/F was 37/38. 20%, 19%, 36% and 25% patients had RA, RAEB-1, RAEB-2 and AML respectively. IPSS was low, int-1, int-2 and high in 2%, 26%, 35% and 37% respectively. IPSS-R was very low, low, int, high and very high in 3%, 2%, 16%, 20% and 59% respectively. Patients were treated by AZA (75mg/m²/day, 7 days every 4 weeks) for a median number of 6 cycles (range 1-50). Overall response rate (ORR) was 60%, including 28% CR, 17% marrow (m) CR, 7% PR and 8% stable disease (SD) with hematologic improvement (HI). In MDS, the ORR was 57% (33 % CR, 11% mCR, 7% PR and 6% SD with HI). In AML, the ORR was 62% (23% CR, 23% mCR, 8% PR and 8% and SD with HI, based on MDS criteria).The median % of BCL2L10 positive cells was 9.5% (range 0-95) and no correlation was observed between % of BCL2L10 positive cells and marrow blasts. The median % of BCL2L10 positive cells was 30% (range 0-95) in non-responders and 10% (range 0-56) in responders (p=0.01). The response rate was 7% and 64% in patients with ≥ 50% vs < 50% BCL2L10 positive cells, respectively (p<0.0001). Median OS after FCM analysis performed before or during AZA treatment was 5.8 months in the 11 patients with more than 50% versus 11.7 months in the 64 patients with less than 50% of BCL2L10 positive cells (p=0.03). In 8 patients studied sequentially before, during AZA treatment and at relapse, the % of BCL2L10 positive cells remained stable below 50% and increased above 50% few months before relapse. The best prognostic cut off value for BCL2L10 positive cells was 50%. Flow cytometry results were reproducible in the two laboratories, with two different cytometers.Conclusion:We confirmed in this larger prospective multicenter cohort that the percentage of BCL2L10 positive cells, analyzed in 2 different labs, is inversely correlated with response and survival after AZA treatment in both MDS and AML patients, the best prognostic cut-off value for BCL2L10 positive cells being 50%. Our flow cytometry assay is reproducible in different laboratories and can be performed routinely at diagnosis and during AZA treatment. A multivariate analysis including other prognostic factors of response and OS with AZA will be presented. DisclosuresNo relevant conflicts of interest to declare.