Bone Marrow Levels of PARP1 mRNA Predict Response to Treatment with 5-Azacytidine in Patients with Myelodysplastic Syndrome

Introduction

Poly [ADP-ribose] polymerase 1 (PARP1) is the major member of a family of nuclear enzymes that participate in DNA repair and gene transcription, by recruiting various proteins involved in DNA repair to sites of DNA damage. PARP1 overexpression has been correlated to poor treatment response in children with acute lymphoblastic leukemia, while PARP1-driven apoptosis is important in patients with chronic lymphocytic leukemia. Moreover, PARP1 inhibitors have been tested in vitro in lymphoid malignancies, acute myeloid leukemia, myelodysplastic syndrome (MDS) and acute promyelocytic leukemia.

We have recently shown that bone marrow PARP1 mRNA levels correlate to the prognosis of patients with MDS (Diamantopoulos et al. Blood Cancer 2017). Moreover, PARP1 has been related to hypermethylation through DNA methyltransferase-1 (DNMT1) gene upregulation and protein activation (Zampieri et al. PLoS One. 2009). This correlation with DNA methylation is very important, since hypermethylation is a major event in the pathogenesis of high risk MDS, and hypomethylating agents, such as 5-azacytidine, are the most effective available treatment option for patients with higher risk MDS.

Aim

Based on the relation of PARP1 with hypermethylation, we investigated the correlation of bone marrow PARP1 mRNA levels of patients with MDS with treatment response to 5-azacytidine.

Methods

Bone marrow samples were collected from patients with MDS at diagnosis. Only samples from patients that were eventually treated with 5-azacytidine were selected for analysis. RNA extraction and reverse transcription were performed using standard protocols. Primers for PARP1 and beta-actin were designed using the primer3 software (University of Massachusetts, USA) and a SYBR green based real-time PCR was performed on a CFX96 RT-PCR system (Bio-Rad Laboratories, Hercules, CA, USA). PARP1 mRNA levels were expressed as a PARP1/beta-actin ratio. IBM SPSS statistics, version 23.0 (IBM Corporation, North Castle, NY, USA) was used for the analysis of the results.

Results

Sixty-seven patients diagnosed with MDS per the WHO 2008 classification and treated with 5-azacytdine were included in the study. The basic characteristics of the patients are shown in Table 1. The median follow-up of the patients was 25.8 (2.6-117.0) months. Response to treatment with 5-azacytidine per the International Working Group (IWG) criteria is shown in Table 1. The median PARP1 mRNA level was 0.0793 (range (0.0000424 - 3.43). Response to treatment differed between patients with high and low PARP1 mRNA levels (Independent samples Kruskal-Wallis test, 2-sided p=0.009). More precisely, the median level of PARP1 mRNA was 4 times higher in responders (partial response, complete response, hematologic improvement) than in non-responders (failure and stable disease), Independent samples Mann-Whitney U test, 2-sided p=0.006. We found no correlation of any other of the studied factors (MDS type, IPSS, IPSS-R, WPSS, karyotype risk, number of cytopenias, marrow blast percentage, hemoglobin level, neutrophil and platelet count, dose reduction of 5-azacytidine) to response to treatment in this cohort of patients. Although there was no correlation of PARP1 mRNA levels with overall survival, survival of patients after 5-azacytidine treatment tended to be longer in patients with higher PARP1 mRNA levels (38.7 versus 19.8 months, log rank test, p=0.136).

Discussion

We found that patients with MDS and high PARP1 mRNA levels have a better response to treatment with 5-azacytidine. This correlation has a theoretical basis, since upregulation of PARP1 is correlated to DNA hypermethylation, and hypomethylating agents may be more effective in cells with higher methylation levels. Moreover, in our cohort of patients, PARP1 level was the only factor affecting treatment response, although this response was not translated into a survival benefit.

PAPR1 is the only biological factor and one of very few identified factors to predict response to treatment with hypomethylating agents - low hemoglobin and platelet count were found to be an independent factor of poor response to hypomethylating agents (Jung et al. Oncotarget 2016), and absence of aberrant myeloid progenitor cells were found to predict response to treatment (Alhan et al. Cytometry B Clin Cytom 2014). A possible synergy between PARP1 inhibitors and hypomethylating agents in MDS should be further investigated.

View largeDownload slide

View largeDownload slide

Close modal