Introduction: T-MN is a serious complication of prior chemotherapy and/or radiation therapy. In 2008, the World Health Organization created a new category for t-MN due to the distinct variance in etiology when compared to de novo myelodysplastic syndromes (MDS) or acute myelogenous leukemia (AML). Compared to de novo MDS/AML, t-MN characteristically has poor risk cytogenetics and a higher incidence of TP53 mutations which are both associated with inferior survival. Small, retrospective analyses have evaluated the efficacy of front-line HMA therapy for t-AML and t-MDS, with response rates of approximately 38% (Klimek et al. Leukemia Res, 2012 and Bally et al, Leukemia Res, 2013). However, the response duration (RD), progression-free survival (PFS), and overall survival (OS) outcomes are poorly characterized in t-MN treated with HMA. This retrospective study evaluates the efficacy and survival of t-MN patients (pts) treated with azacitidine (AZA) or decitabine (DAC) in the largest cohort analyzed to-date.
Methods: Our cohort included pts with t-MN at Memorial Sloan Kettering Cancer Center treated with either AZA or DAC from January 1, 2002 to September 30, 2016. The primary objective was to compare the best overall response rate (ORR) between AZA and DAC, including complete remission (CR), marrow CR (mCR), hematologic improvement (HI), and partial response (PR). The best ORR was then compared to stable disease (SD), treatment failure (TF), and progressive disease (PD). Secondary objectives included OS, PFS, and RD. Responses were assessed using the 2006 MDS International Working Group (IWG) guidelines and 2003 AML IWG guidelines. We also assessed the impact of TP53 mutations on response to HMA and survival outcomes.
Results: Of the 118 pts (56% male), 76 pts received DAC and 42 pts received AZA. Median age was 70 years (yrs) (range, 25-90). DAC recipients were younger (median 67 yrs, range, 25-85) compared to AZA (median 73 yrs, range, 42-90). More pts with AML were treated with DAC (n=27, 35.5%) and more pts with MDS were treated with AZA (n=36, 85.7%). Most MDS pts had Poor (19.3%) or Very Poor (42%) risk cytogenetics, resulting in a High (23.9%) or Very High (54.5%) overall IPSS-R risk category. Time to response was faster with DAC (9.4 weeks) versus (vs.) AZA (13.3 weeks) with more pts proceeding to hematopoietic cell transplant (HCT) in the DAC group (n=19, 25%) vs. AZA (n=3, 7.1%).
Overall, there were no statistical differences in response between AZA and DAC when we analyzed combined TF/PD (45.2% and 32.9%, respectively), SD (14.3% and 25%, respectively), and combined CR/HI/PR (40.5% and 42.1%, respectively) with a p-value of 0.28 (Table 1). However, more patients receiving DAC achieved CR and mCR (18.4% and 14.5%, respectively) vs. AZA (4.8% and 7.1%, respectively) and AZA had a higher HI rate (28.6%) vs. DAC (5.3%). DAC pts also appeared to have higher TF rates (25%) vs. AZA (19%) with more PD in the AZA group (26.2%) vs. DAC (7.9%).
Median OS was 10.9 months (mo) and PFS was 5.3 mo (Figure 1). There was no difference in median OS between AZA (11.2 mo) and DAC (10.9 mo, p=0.61) and no difference in the median PFS between AZA (5.1 mo) and DAC (5.3 mo, p=0.83). Lastly, there was no difference in median RD between AZA (4.5 mo) and DAC (3.4 mo, p=0.39).
Of the 58 pts with known TP53 mutation status, 29 (50%) had ≥1 TP53 mutation(s). There was no difference in OS and PFS between mutant and wildtype TP53 pts, and no difference in survival between AZA and DAC recipients in the mutant TP53 cohort (Figure 2).
Conclusions: This large, retrospective study showed no difference in the ORR between AZA and DAC. The faster response and higher rates of CR and mCR may favor the use of DAC as a bridge to HCT in eligible pts. Conversely, the higher rates of HI seen with AZA can improve the quality of life in MDS pts who are ineligible for HCT. Our results confirm prior reports of poor OS seen in t-MN. In our cohort, we showed no difference in survival outcomes between AZA and DAC. There was also no statistical difference in OS/PFS between DAC and AZA in mutant TP53 pts, however point estimates may indicate a pattern favoring DAC over AZA that needs to be explored in a larger sample size. Although we confirmed the efficacy of HMA in this difficult to treat population, response duration and survival outcomes remain poor. Development of more effective therapies that specifically target the biology of t-MN are needed to improve survival outcomes.
