Early Increase of Phospholipase Cbeta1 (PI-PLCbeta1) Gene Expression Predicts Azacitidine Responsiveness in MDS Patients

Introduction.: Azacitidine (AZA) is a DNA methyltransferase inhibitor currently approved for the treatment of high-risk MDS patients, which has been demonstrated to be feasible and effective also in lowrisk MDS (Fenaux P et al, Lancet Oncol 2009; Musto P et al, Cancer 2010). However, at least 4 or 6 cycles of therapy are required for assessing the hematologic response, and predictive markers of responsiveness are still lacking. PI-PLCbeta1 plays a role in the MDS progression to AML and is a specific target for AZA therapy (Follo MY et al, PNAS 2009). Indeed, PIPLCbeta1 has been demonstrated to be a dynamic marker for responsiveness to demethylating therapy, in that PI-PLCbeta1 mRNA increase or decrease could be associated with favourable response or failure, respectively. Stemming from these data, in this study we further investigated the role of PI-PLCbeta1 in MDS patients during AZA therapy. Methods.: The study included 60 patients, 22 low-risk MDS (WHO: RA, RARS, RCMD, RAEB-1, and IPSS risk Low or Int-1), and 38 high-risk MDS (WHO: RCMD, RAEB-1, RAEB-2, and IPSS risk Int-1 or High). All the patients received a minimum of 6 cycles, in the absence of disease progression or unacceptable toxicity. Hematologic response was defined according to the revised IWG criteria (Cheson et al, Blood 2006). Positive clinical responses were defined as: Complete Remission (CR), Partial Remission (PR) or Hematologic Improvement (HI). At a molecular level, for each patient we quantified the amount of PI-PLCbeta1 mRNA at baseline and before each cycle of AZA therapy. PI-PLCbeta1 ratio was calculated as the mean expression of PI-PLCbeta1 at cycles 1 to 3, as compared with the baseline level within the same subject. In case the mean value of PI-PLCbeta1 gene expression during the cycles 1 to 3 was above the baseline level, we defined it as a "PI-PLCbeta1 early increase". On the contrary, a "stable PI-PLCbeta1" expression was observed when subjects did not show any increase during the first three cycles of therapy, as compared with baseline. Results.: Patients' median age was 69 years (range 37–85) and the median follow-up was 23 months (range 1–103). The median number of AZA cycles was 11 (range 3–59) for high-risk MDS, and 8 (range 1–8) for low-risk MDS. Positive clinical responses were observed in 37/60 (62%) of the MDS patients (7 CR, 1 PR, 29 HI). In particular, 13/22 (59%) of our low-risk MDS and 24/38 (63%) of our high-risk MDS patients showed a positive clinical response to AZA, with 4 CR, 1 PR, and 19 HI in high-risk MDS, and 3 CR and 10 HI in low-risk MDS. Overall survival (OS), Progression- Free Survival (PFS), and Overall Response Rate (ORR) were analyzed using a Kaplan-Meier method, considering p-values<0.05 as statistically significant. No differences in OS nor in PFS were noted between patients with early increased or stable PI-PLCbeta1 (OS: 36 vs. 30 months, p=0.45; PFS: 28 vs. 24 months, p=0.06). However, PI-PLCbeta1 early increase was significantly associated with ORR (increase: 25/38 (65%) vs. stable: 4/22 (18%); p<0. 05). The predictive value of PI-PLCbeta1 was also analyzed: PI-PLCbeta1 early increase was significantly associated with duration of AZA response (increase vs. stable: 26 vs. 12 months; p<0.05), showing that an early increase of PI-PLCbeta1 was associated not only with a positive clinical response, but also with a higher probability of a longer response. Conclusions.: Taken together, our data confirm the role of PI-PLCbeta1 as a dynamic marker of response to AZA and show that the detection of an increase in PI-PLCbeta1 gene expression within the first three cycles of AZA therapy is associated with a better clinical outcome and a longer hematological response. Further analyses are needed to confirm in a larger group of patients the predictive role of PI-PLCbeta1 mRNA detection during AZA therapy.