Abstract # 2957 – Landmark Response and Survival Analyses from 102 MDS and CMML Patients Treated with Guadecitabine in a Phase 2 Study Showing That Maximum Response and Survival Is Best Achieved with Adequate Treatment Duration
Authors: Michael R. Savona, MD1, Hagop M. Kantarjian, MD2, Gail J. Roboz, MD3, Casey L. O’Connell, MD4, Katherine J. Walsh, MD5, Raoul Tibes, MD, PhD6*, Karen W.L. Yee, MD7, Wendy Stock, MD8, Elizabeth A. Griffiths, MD9, Elias Jabbour, MD2, Scott D. Lunin, MD10*, Todd L. Rosenblat, MD, MS11*, Nikolai A. Podoltsev, MD, PhD12, Jean-Pierre Issa, MD13*, Xiang Yao Su, PhD14*, Mohammad Azab, MD14 and Guillermo Garcia-Manero, MD2
1Vanderbilt-Ingram Cancer Center, Vanderbilt University School of Medicine, Nashville, TN; 2The University of Texas MD Anderson Cancer Center, Houston, TX; 3Weill Cornell/NY Presbyterian Medical Center, New York, NY; 4USC Keck School of Medicine, University of Southern California, Los Angeles, CA; 5The Ohio State University, Columbus, OH; 6Mayo Clinic Arizona, Scottsdale, AZ; 7Princess Margaret Cancer Centre, Toronto, ON, CAN; 8University of Chicago Medical Center, Chicago, IL; 9Roswell Park Cancer Institute, Buffalo, NY; 10Sarah Cannon Research Institute, Florida Cancer Specialists, Venice, FL; 11Columbia University Irving Medical Center, New York, NY; 12Yale School of Medicine, New Haven, CT; 13Fels Institute, Temple University, Philadelphia, PA; 14Astex Pharmaceuticals, Inc., Pleasanton, CA
Background: Guadecitabine is a next generation subcutaneous (SC) hypomethylating agent (HMA) resistant to degradation by cytidine deaminase which results in prolonged in vivo exposure to the active metabolite decitabine. We conducted a phase 2 study of guadecitabine in 102 Myelodysplastic Syndromes (MDS), and Chronic Myelomonocytic leukemia (CMML) patients. International guidelines recommend a minimum of 4 to 6 cycles of HMA treatment to gain maximum benefit, but some suggest that treatment may not be beneficial if no response was observed after 4 cycles. No prospective studies have confirmed the correlation between an HMA number of cycles with response and survival using landmark methodology. We present here the results of landmark response and survival analyses based on number of cycles and whether patients had an objective response or not.
Methods: Landmark response based on 2006 IWG criteria, and overall survival (OS) analyses for patients alive at or beyond month 3 and month 5 (time of planned start of cycle 4 and cycle 6 respectively) were conducted. Objective response (OR) was defined as patients who had Complete Response (CR), Partial Response (PR), marrow (m)CR, or Hematological Improvement (HI). Landmark OS was compared between patients who received at least 4 or 6 cycles and those who did not. The landmark methodology avoids the bias of early deaths before cycles 4 and 6 attributing a survival benefit in those who did not die early and were able to get more cycles. We also compared the result in responding and non-responding patients to see if survival benefit was restricted to responding patients only.
Results: The study completed enrolment with 102 patients: 53 patients after HMA failure (relapsed/refractory or r/r), and 49 HMA-naive patients (Treatment Naïve or TN) with a median follow up for the entire study of 3.2 years (IQR 2.9-3.5 years). Median age was 71 and 72 years for TN MDS/CMML and r/r MDS/CMML patients respectively. Median OS was 23.4 months (m) for TN MDS/CMML patients and 11.7 m for r/r MDS/CMML patients. Of the 102 patients treated, 37 patients (36.3%) and 58 (56.9%) received less than 4 and 6 cycles respectively. The landmark analysis population was 91 patients for the 4-cycle analysis and 87 patients for the 6-cycle analysis. In those patients, the primary reasons for treatment discontinuation before cycle 4 or 6 respectively were patient decision (9.8% and 11.8%), and investigator decision (5.9% and 9.8%) while early progression accounted for 3.9% and 10.8% of those patients. There were no major baseline characteristics difference between patients who received at least 4 and 6 cycles and those who did not in the patients included in the landmark analyses. In the landmark analysis, patients who received at least 4 cycles (65 patients) had an OR rate of 68% compared to 15% in 26 patients who received <4 cycles (p <0.0001) and median OS of 20.4 m compared to 15.2 m respectively (HR 0.78, 95% CI 0.45-1.3, p 0.36). Those who received at least 6 cycles (44 patients) had an OR rate of 82% compared to 26% in 43 patients who received < 6 cycles (p<0.0001), and median OS of 23.8 m vs 13.6 m respectively (HR 0.51, 95% CI 0.3-0.85, p 0.009). Results were consistent when r/r MDS/CMML and HMA-naïve MDS/CMML were analyzed separately. Landmark OS analysis also favored those who received guadecitabine for at least 4 or 6 cycles compared to those who received <4 and < 6 cycles even in the absence of objective response (OS HR of 0.82 and 0.42 respectively) but the sample size was small to show statistical significance (p 0.58 and 0.10 respectively)
Summary/Conclusions: In a prospective phase 2 study of 102 MDS/CMML patients treated with the HMA guadecitabine, patients who were alive at the planned start of cycle 4 and cycle 6 did not continue treatment primarily because of patient or investigator decision in addition to early progression. Those who were alive and continued treatment for at least 4 or 6 cycles achieved highly significant objective response benefit compared to those who did not. Survival benefit was highly significant for those who received at least 6 cycles and was not restricted to patients who had an objective response. It is important to weigh reasons for treatment discontinuation carefully before discontinuing guadecitabine HMA treatment in MDS/CMML patients before 6 cycles to maximize response and survival benefit.