Leukemia (2017) 31, 318–324 © 2017 Macmillan Publishers Limited, part of Springer Nature. All rights reserved 0887-6924/17 www.nature.com/leu
ORIGINAL ARTICLE
A clinical trial for patients with acute myeloid leukemia or myelodysplastic syndromes not eligible for standard clinical trials G Montalban-Bravo1, X Huang2, E Jabbour1, G Borthakur1, CD DiNardo1, N Pemmaraju1, J Cortes1, S Verstovsek1, T Kadia1, N Daver1, W Wierda1, Y Alvarado1, M Konopleva1, F Ravandi1, Z Estrov1, N Jain1, A Alfonso1, M Brandt1, T Sneed1, H-C Chen2, H Yang1, C Bueso-Ramos3, S Pierce1, E Estey1, Z Bohannan1, HM Kantarjian1 and G Garcia-Manero1 Most clinical trials exclude patients with poor performance or comorbidities. To study whether patients with these characteristics can be treated within a clinical trial, we conducted a study for patients with acute myeloid leukemia (AML) or myelodysplastic syndromes (MDS) with poor performance, organ dysfunction or comorbidities. Primary endpoint was 60-day survival. Study included stopping rules for survival and response. Treatment consisted on a combination of azacitidine and vorinostat. Thirty patients (16 with MDS, 14 with AML) were enrolled. Median follow-up was 7.4 months (0.3–29). Sixty-day survival was 83%. No stopping rules were met. Main adverse events (AEs) were grades 1 and 2 gastrointestinal toxicities. In view of these results, we expanded the study and treated 79 additional patients: 27 with azacitidine (AZA) and 52 with azacitidine and vorinostat (AZA+V). Median follow-up was 22.7 months (12.6–47.5). Sixty-day survival rate was 79% (AZA = 67%, AZA+V = 85%, P = 0.07). Median overall survival was 7.6 months (4.5–10.7). Median event-free survival was 4.5 months (3.5–5.6). Main AEs included grades 1 and 2 gastrointestinal toxicities. Our results suggest this subset of patients can be safely treated within clinical trials and derive clinical benefit. Relaxation of standard exclusion criteria may increase the pool of patients likely to benefit from therapy. Leukemia (2017) 31, 318–324; doi:10.1038/leu.2016.303
INTRODUCTION Participation in clinical trials is fundamental for the development of new therapeutic interventions. New drugs are first studied in the context of clinical trials not only to test specific clinical hypothesis, but also to protect patients from unanticipated side effects of such therapy.1 Importantly, for the patient and the physician, there is an expectation that the investigational therapy will have a beneficial clinical effect. This is particularly important for patients for whom there is no optimal standard of care.2 Despite this need, only 3–5% of patients with cancer treated in the United States currently enroll in clinical trials.3 Although enrollment rates may differ regionally, there are a number of reasons for such low rate of participation, with clinical trial eligibility criteria likely representing one of them. Most clinical studies in oncology exclude patients with comorbidities, active or recent malignancies, organ dysfunction or poor performance status. How these criteria protect patients is unclear. Although some of them are based on clinical reasoning it seems that these criteria are in place more to protect the drug or intervention being studied than the patient itself. In addition, such stringent criteria not only may limit the access of ‘unfit’ patients to potentially beneficial therapies, but also limit our ability to extrapolate the safety and efficacy outcomes of new drugs being tested in conventional clinical trials to this subset of patients with unfavorable clinical features.4 Of importance, patients with such characteristics and poor expected outcomes represent the population that
could benefit the most from investigational approaches considering their otherwise dismal prognosis.5–7 This is particularly important in patients with leukemia in whom tissue infiltration or disease burden may result in organ dysfunction or poor performance that could be reversible with appropriate therapy.8–10 In view of this, it seems necessary to determine whether excluding these patients from new therapeutic interventions within monitored clinical trials is ultimately beneficial, and if development of clinical trials for this subset of patients is safe and feasible. To test this concept we designed a clinical trial for patients with acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS) with organ dysfunction, poor performance, other malignancies or other comorbidities. Patients were treated with azacitidine,11–13 a hypomethylating agent, and vorinostat,14,15 a histone deacetylase inhibitor. This treatment modality was selected based on the acceptable toxicity profile of azacitidine,11 and vorinostat,16 and existing preclinical data suggesting a potential synergistic effect of combining hypomethylating agent with histone deacetylase inhibitors.17–19 The study was designed with stopping rules for survival and response. In view of the results of this initial study, suggesting it is both feasible and safe to treat this group of patients within a clinical trial, we expanded the study to increase the treated patient population and allow further experience in the development of clinical trials for this group of patients.
1 Department of Leukemia, University of Texas MD Anderson Cancer Center, Houston, TX, USA; 2Department of Biostatistics, University of Texas, MD Anderson Cancer Center, Houston, TX, USA and 3Department of Hematopathology, University of Texas, MD Anderson Cancer Center, Houston, TX, USA. Correspondence: Dr G Garcia-Manero, Department of Leukemia, University of Texas, MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston 77015, TX, USA. E-mail: [email protected] Received 4 July 2016; revised 21 August 2016; accepted 30 August 2016; accepted article preview online 31 October 2016; advance online publication, 18 November 2016
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
MATERIALS AND METHODS Study design and patient population We conducted a study for patients ineligible to conventional clinical trials, due to organ dysfunction, poor performance, other malignancies or comorbidities at the University of Texas, MD Anderson Cancer Center following institutional guidelines. The study was registered at clinicaltrial.gov as NCT00948064. We enrolled patients older than 17 years of age with previously untreated AML or MDS (intermediate-2 or higher risk by International Prognostic Scoring System20) with at least one of the following inclusion criteria: serum creatinine ⩾ 2 mg/dl, total bilirubin ⩾ 2 mg/dl, ECOG performance status (PS) equal to 3 or 4, or ineligibility to participate on a protocol of higher priority due to comorbidities, other active malignancies or malignancies with a remission period of o2 years. Diagnosis was confirmed at MD Anderson Cancer Center and classified following the revised 2016 WHO critera.21 Pregnant patients were excluded from study participation. Main exclusion criteria included presence of favorable cytogenetic abnormalities, such as inv(16), t(16;16), t(8;21) or t(15;17) due to available effective standard of care therapies,22,23 prior anti-leukemic therapy (except hydroxyurea for cytoreduction in case of proliferative disease); not being able to receive oral medications due to the oral route of administration of vorinostat; prior therapy with histone deacetylase inhibitors; active hepatitis A, B, or C infection due to risk of potential hepatic toxicity with both azacitidine and vorinostat; or known allergy or sensitivity to vorinostat or azacitidine. Study was designed with predefined stopping rules for survival and response that were generated by the use of a Bayesian sequential monitoring design for single-arm clinical trials.24 Multc Lean desktop v1.1 program (https://biostatistics.mdanderson.org/SoftwareDownload/Default. aspx) was used to generate this model. To define the minimum expected survival and response rates that would trigger these stopping rules, prior data of 181 patients treated at our institution who met the study′s eligibility criteria was used. These data were also used to determine the study sample size and patient cohort sizes. In this group, 50% of patients had a survival of at least 60 days, and the complete response (CR) rate was 28%. The maximum sample size was 30. For the survival stopping rule, the study would stop if survival by Day 60 was unlikely (chance o5%) to be at least 20% higher than in the historical group. This would be met if the number of patients alive at ⩾ 60 days was ⩽ 0 of 3 (0/3) patients treated, 2/6, 3/9, 5/12, 6/15, 8/18, 10/21, 11/24, or 13/27. For the response stopping rule, the study would stop if the CR was unlikely (chance o10%) to be higher than in the historical group. This would be met if the number of patients with CR is ⩽ 0/6, 1/12, 2/18, 3/24 or 4/30. After study conclusion, in view of the survival and toxicity results, and after approval from the institutional review board, we expanded the original trial to increase the treated patient population and further explore the feasibility of treating patients with unfavorable clinical characteristics within clinical trials. The study was independently monitored by the Department of Biostatistics at MD Anderson. The primary objective was survival by Day 60. The maximum sample size was 80. The initial 40 patients were randomized in a 1:1 ratio to either azacitidine or azacitidine in combination with vorinostat. As an additional safety measure, treatment assignment for subsequently enrolled patients was performed in an unbalanced manner in favor of the safest treatment modality determined by higher 60-day survival rate. Assignment of patients to a given therapy would stop if 60-day survival rate within a given group was unlikely to be 450%. All calculations were performed using Multc99 and Adaptive Randomization software. This study included a stopping rule for toxicity were by the trial would stop if at any given time average rate of grade 3 or greater non-hematologic toxicity was 420%. Additional details for both phases of the trial are shown in Supplementary Material.
Therapy, baseline evaluation, follow-up during study and response criteria All patients received therapy with 5-azacitidine at a dose of 75 mg/m2 daily on a 5 day schedule (days 1–5). In addition, all patients in the initial exploratory study and those receiving combination therapy in the expansion study also received oral vorinostat at a dose of 200 mg three times a day on the same days as 5-azacitidine. Courses could be repeated every 3–8 weeks. Therapy was planned to continue until disease progression, patient request, or intercurrent illness or until 12 courses of therapy. After 12 courses, patients could continue to receive single agent azacitidine. All patients were evaluated at baseline with a bone marrow aspirate with cytogenetic assessment, complete history and physical examination, © 2017 Macmillan Publishers Limited, part of Springer Nature.
319 complete blood count and chemistries. During the first month, patients were assessed weekly and then monthly. Bone marrow aspiration with cytogenetics, if abnormal at baseline, were performed on Day 28 of Course 1 and afterwards with each course of therapy, to document response or to decide on therapy administration. Once a response was obtained, bone marrow sampling was repeated at the discretion of the treating physician to confirm remission status or document loss of response. Response assessment in patients with AML was performed following the International Working Group 2003 recommendations.25 For patients with MDS, response assessment was performed following the revised 2006 International Working Group criteria.26
Assesment of comorbidities Evaluation of patient comorbidities was performed using the Adult Comorbidity Evaluation-27 (ACE-27) index at baseline. Prior studies have confirmed the use of this index in comorbidity evaluation of patients with MDS.27,28 Outcome of patients in terms of overall survival (OS) was correlated with ACE-27 score.28
Safety evaluation Toxicity was assessed on the basis of adverse events (AEs). All AEs were recorded and coded according to the National Cancer Institute Common Terminology Criteria for Adverse Events v4.0. Toxicity monitoring determined criteria for stopping rule for toxicity.
Correlative studies Analysis of global DNA methylation29 was performed in patients included on the initial exploratory study. Please see Supplementary Methods, Supplementary Table S1 and Supplementary Figure S1 in Supplemental Appendix for details.
Statistical analysis Continuous and categorical variables were evaluated using the Wilcoxon rank sum and Fisher’s exact or χ2 tests, as appropriate. OS was defined as the time interval between treatment start date and death date, and was censored at the last follow-up date for patients who were alive. Event-free survival (EFS) was defined as the time interval between date of treatment start and date of resistance, relapse or death. Patients who were alive with CR and without relapse were censored at the date of last follow-up. The probabilities of OS and EFS were estimated using the Kaplan–Meier methods. Cox proportional hazards regression models were used to assess the association between patient characteristics and OS or EFS. Logistic regression models were conducted to assess the relationship between the patient characteristics and survival at 60 days. Statistical analyses were performed in SAS 9.3 (The SAS Institute, Cary, NC, USA) and Stata 13.1 (Stata Corp, College Station, TX, USA). All P-values were two sided, with significance being Po0.05.
RESULTS Initial exploratory study A total of 30 consecutive patients were enrolled between September 2009 and December 2010. Patient characteristics are shown in Table 1. A total of 14 (47%) patients had AML and 16 (53%) had MDS. Patient specific inclusion criteria and evolution are illustrated in Figure 1. Specific inclusion criteria, number of cycles, survival, response, subsequent therapies and cause of death for each individual patient is detailed in Supplementary Table S4 of Supplementary Material. The median follow-up was 7.4 months (range 0.3–29) with a median number of cycles of therapy of 3.5 (range 1–12). Sixty-day survival rate was 83% (24/30) and the stopping rule for survival was never met. No differences in OS (5.9 months vs 7.9 months; HR = 0.67, 95% CI 0.32–1.4, P = 0.308) and EFS (3.6 vs 4.4 months; HR = 0.74, 95% CI 0.35–1.57, P = 0.432) were observed between patients with AML or MDS (Figure 2). The overall response rate was 40% (12/30) with 8 (27%) patients achieving CR: 4 patients with MDS and 4 with AML. Therefore, the stopping rule for response was never met. Remarkably, two Leukemia (2017) 318 – 324
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
320 Table 1.
Patient characteristics
Characteristics
Age Sex Female Diagnosis AML MDS MDS-SLD MDS-RS MDS-MLD MDS-EB CMML Cytogeneticsa − 5/5q − − 5/5q − and − 7/7q − − 7/7q − +8 Diploid Other Complex Serum creatinine (mg/dl) Median ⩾2 Total bilirubin (mg/dl) Median ⩾2 ECOG performance status 0–1 2 ⩾3 Hemoglobin (g/dl) WBC (×109/l) Platelet count (×109/l) Bone marrow blasts (%) Inclusion criteriab Other malignancy Bilirubin ⩾ 2 mg/dl Creatinine ⩾ 2 mg/dl PS ⩾ 3 Comorbidities Ineligible to other trial
Initial exploratory study (N = 30) n (%)
Extension study (N = 79) Azacitidine (N = 27) n (%)
Azacitidine+vorinostat (N = 52) n (%)
73 (44–83)
71 (54–90)
70 (30–88)
9 (30)
8 (30)
16 (31)
14 16 1 0 2 11 2
(47) (53) (3) (0) (7) (37) (7)
14 (52) 13 (48) 2 (7) 0 (0) 3 (11) 5 (19) 3 (11)
18 (35) 34 (65) 5 (10) 2 (4) 6 (12) 17 (33) 4 (8)
3 (10) 1 (3) 1 (3) 1 (3) 7 (24) 3 (10) 14 (47)
6 (22) 3 (11) 3 (11) 2 (7) 4 (15) 1 (4) 14 (52)
8 (16) 8 (16) 4 (8) 2 (4) 15 (29) 3 (6) 22 (42)
1 (0.3–2.4) 3 (10)
1.2 (0.47–4.76) 5(19)
0.93 (0.38–5.59) 6 (12)
1 (0.3–4.1) 4 (13)
0.9 (0.3–3.3) 3 (11)
0.6 (0.3–3.2) 2 (4)
19 (63) 6 (20) 5 (17) 9.2 (7.2–12) 3 (0.6–112) 41 (12–310) 14 (1–78)
21 (78) 3 (11) 3 (11) 9 (7.7–12.2) 3.8 (0.2–44) 32 (8–354) 17 (1–87)
45 (88) 3 (6) 4 (8) 9.05 (5.8–14.8) 2.45 (0.2–102) 44 (2–714) 8 (1–89)
19 (63) 4 (13) 3 (10) 5 (17) 4 (12) 0 (0)
13(48) 3 (11) 5 (19) 5 (19) 5 (19) 0 (0)
28 (54) 3 (6) 7 (13) 4 (8) 9 (17) 3 (6)
Abbreviations: AML, acute myeloid leukemia; CMML, chronic myelomonocytic leukemia; MDS, myelodysplastic syndrome; MDS-EB, MDS with excess blasts; MDS-MLD, MDS with multilineage dysplasia; MDS-SLD, MDS with single lineage dysplasia; WBC, white blood cell count. aA total of five patients treated with azacitidine+vorinostat in the extension study did not have evaluable conventional cytogenetics at diagnosis due to absence of metaphase growth. bAbsolute numbers and percentages for eligibility criteria are represented in terms of total number of patients meeting each criterion. A total of six patients met more than one inclusion criteria.
patients, one enrolled due to other active malignancy and another due to PS ⩾ 3, underwent allogeneic stem-cell transplantation after achieving CR. In univariate analysis, creatinine or bilirubin ⩾ 2 mg/dl, PS ⩾ 3, ACE-27 ⩾ 2 and presence of other malignancies were not significant predictors of response or survival. Only presence of more than one eligibility criteria was predictive for lower likelihood of 60-day survival (P = 0.04). No significant differences in survival were observed between patients with ACE-27 scores 0–1 compared with 2–3 (6.3 vs 7.0 months; HR = 0.88, 95% CI 0.41–1.91, P = 0.755; Figure 3a). A total of 21 (70%) patients experienced at least one AE (Table 2). In general, therapy was well tolerated with most toxicities being grades 1 and 2. Grades 1 and 2 gastrointestinal toxicity and febrile neutropenia were the most common toxicities. Therapy had to be held or withdrawn due to toxicities in eight patients (27%). Mortality at 4 and 8 weeks was 10% and 20%, respectively. Leukemia (2017) 318 – 324
Expansion study A total of 79 patients were enrolled between September 2011 and March 2014: 27 (34%) received azacitidine (AZA) and 52 (66%) azacitidine in combination with vorinostat (AZA+V). Patient characteristics and inclusion criteria are shown in Table 1. Specific inclusion criteria, number of cycles, survival, response, subsequent therapies and cause of death are detailed in Supplementary Table S5 of Supplementary Material. Survival. The median duration of follow-up was 22.7 months (range 12.6–47.5). Median number of cycles administered was 3 (1–12). Consistent with the initial exploratory study results, 60-day survival was observed in 62 (79%) patients: 18/27 treated with AZA and 44/52 with AZA+V. Median OS was 7.6 months (4.5–10.7) with patients with MDS having significantly longer OS (10.6 vs 4.4 months, HR = 0.5, 95% CI 0.1–8.3, P = 0.07) and EFS (5.7 vs 2.9 months, HR = 0.1, 95% CI 0.3–0.8, P = 0.04) compared with © 2017 Macmillan Publishers Limited, part of Springer Nature.
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
321
Figure 1. Evolution of patients on initial exploratory study. Patients are presented by consecutive enrollment into trial. Patient inclusion criteria are represented by bar color as specified in the legend. Cri, complete response with insufficient hematological recovery (platelets or neutrophils); CRp, complete response with insufficient platelet recovery; HI, hematological improvement; PR, partial response. Absence of any of the previous indicators on a patient survival bar is indicative of no response to therapy.
Figure 2.
Kaplan–Meier estimates of (a) OS and (b) event-free survival of patients in the initial exploratory study.
patients with AML (Figure 4). A total of 31 patients (39%) died during study treatment with 48 patients (61%) receiving subsequent therapy including allogeneic stem-cell transplantation after CR in 8 (10%). Among the 31 patients not receiving subsequent therapy, causes of death included: infectious complication in 13 (42%) and cardiovascular event in 2 (6%) patients with unknown cause of death in 16 (52%). The cause of death was unknown in 38 patients (48%) due to transition to hospice, loss of follow-up or death at an outside institution. By univariate analyses © 2017 Macmillan Publishers Limited, part of Springer Nature.
neither creatinine nor bilirubin levels ⩾ 2 mg/dl, ACE-27 ⩾ 2, presence of other malignancy or PS ⩾ 3 were predictive of 60day survival, OS or EFS (Supplementary Tables S2 and S3 of Supplementary Material). Response. The overall response rate was 47% (37/79) including 13 responses in patients treated with AZA and 24 in patients treated with AZA+V. A total of 20 (25%) patients achieved CR, 13 with MDS and 7 with AML. Median number of cycles to response Leukemia (2017) 318 – 324
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
322
Figure 3. Kaplan–Meier estimates of OS based on ACE-27 score: (a) among patients in the initial exploratory study. (b) Among patients in the extension study. Table 2.
Adverse events
Adverse event
Extension study (N = 79) Initial exploratory study (N = 30)
Constipation Dermatologic (skin rash) Diarrhea Respiratory/cardiac Fatigue Febrile neutropenia Gastrointestinal-other Elevated bilirubin Elevated ALT Mucositis/stomatitis Muscle weakness Nausea Opportunistic infection Pain Urinary frequency Blurred vision Voice changes Vomiting
Azacitidine+vorinostat (N = 52)
Any grade %
Grade ⩾ 3%
Any grade %
Grade ⩾ 3%
Any grade %
12 7 20 3 13 17 0 7 3 3 3 20 3 0 0 0 0 3
0 0 7 3 0 17 0 0 3 0 3 3 3 0 0 0 0 0
22 0 0 0 44 7 0 0 0 0 0 7 11 7 4 0 0 0
4 0 0 0 0 7 0 0 0 0 0 0 11 0 0 0 0 0
31 2 27 25 65 17 2 0 0 2 19 37 2 4 0 6 2 6
was 2 (1–12) with median duration of response of 7.4 months (4.9– 9.9). A total of 44 patients were evaluable for cytogenetic response. Eleven (25%) patients achieved CCyR. Achievement of CCyR was associated with significantly improved OS (28.3 m vs 9.4 m; HR 0.27, 95% CI 0.11–0.67, P = 0.001; Supplementary Figure S2). In univariate analysis, creatinine or bilirubin ⩾ 2 mg/ dl, PS ⩾ 3, ACE-27 ⩾ 2 and presence of other malignancies were not significant predictors of response (Supplementary Tables S2 and S3 of Supplementary Material). Safety and toxicity. A total of 65 (82%) patients experienced at least one AE: 15/27 patients treated with AZA and 42/52 patients treated with AZA+V. Adverse events ranked by grade are shown in Table 2. Grade 3 or higher AEs were experienced by 28 (35%) Leukemia (2017) 318 – 324
Azacitidine (N = 27)
Grade ⩾ 3% 6 0 2 12 23 13 0 0 0 0 0 8 2 0 0 0 0 2
patients: 5/27 patients treated with AZA and 23/52 treated with AZA+V. Grade 3 fatigue, dyspnea and neutropenic fever were the most frequent grade 3 AEs. However, stopping rule for toxicity was not met. At the present time of follow-up, 67 (85%) patients have died. Early mortality at 4 and 8 weeks was 10% and 19%, respectively. Effect of comorbidities on therapy. The most common comorbidities by ACE-27 scoring system included hypertension in 50 (63%), other malignancy in 41 (52%), diabetes mellitus in 22 (28%) and coronary artery disease in 21 (27%). No significant differences in survival were observed between patients with ACE-27 scores 0–1 compared with 2–3 (12.2 vs 7.1 months, HR = 1.2, 95% CI 0.6–2.3, P = 0.591; Figure 3b). © 2017 Macmillan Publishers Limited, part of Springer Nature.
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
323
Figure 4.
Kaplan–Meier estimates of (a) OS and (b) event-free survival of patients in the extension study.
DISCUSSION Standard clinical trial eligibility criteria classically exclude patients with organ dysfunction, poor PS or other comorbidities. Here we show that treating this patient population within a clinical trial is feasible. In our initial exploratory study, a majority of patients were able to complete at least one cycle of therapy without major toxicity, and obtain clinical benefit with acceptable responses and survival despite their high comorbidity burden. The fact that we could compute the toxicity profile, response rates, assess treatment exposure for at least 4 weeks and perform pharmacodynamics assays, strongly support the concept that this group of patients can be included in clinical trials. The presence of renal or hepatic dysfunction did not compromise therapy administration or negatively impact survival and overall response rate. Similarly, patients with prior malignancies or PS ⩾ 3 did not have significantly shorter survival and, in two such patients, therapy allowed achievement of CR and subsequent allogeneic stem-cell transplantation. Furthermore, treatment was able to negate the predictive effect of ACE-27, a previously validated comorbidity index with prognostic significance in MDS,27 further suggesting the likely benefit of therapy irrespective of comorbidity burden. In view of these results, we were able to further explore the feasibility of treating this patient population within a clinical trial by conducting a larger confirmatory study. As in the initial exploratory trial, a majority of patients could complete at least one cycle of therapy with acceptable toxicity, although with an overall higher proportion of AEs and grade 3 AEs when combining vorinostat to azacitidine. Importantly, most patients met the study’s primary endpoint of survival at 60 days of therapy suggesting there is no excess treatment-related toxicity and early mortality in this patient population. In fact, as in the initial exploratory phase, therapy was able to negate the prognostic impact of ACE-27 further suggesting the likely benefit of therapy, irrespective of comorbidity burden, in this patient population. Remarkably, a total of 8 patients were able to proceed to allogeneic stem-cell transplantation. Organ dysfunction8–10,30 and poor PS31 in patients with leukemia may appear in the context of tissue infiltration representing a reversible condition with appropriate therapy, and should not always jeopardize treatment. The presence of other comorbidity, particularly when chronic, or if the patient is not anticipated to succumb to it for a short period, should not © 2017 Macmillan Publishers Limited, part of Springer Nature.
impact the clinical information that could be obtained when treating this group of patients on a clinical trial. We hypothesized, that this group of patients, with dismal prognosis if untreated,32–34 would have a more favorable risk/benefit ratio when treated with an investigational intervention than patients traditionally included in such studies. In fact, a previous retrospective study published by Mengis et al.4 evaluating the outcome of patients with AML treated at their institution, suggested patients with comorbidities who were not eligible to conventional trials had similar OS and disease-free survival when treated with off protocol intensive chemotherapy compared with patients treated within a clinical protocol with similar regimens. We recognize our study has several limitations including the heterogeneity in patient population in terms of diagnosis, comorbidities and clinical features which limits our ability to extrapolate the outcomes to specific patient populations. In addition, there are a significant proportion of patients in whom the cause of death was not able to be determined. However, most of such patients died after transition to hospice and, therefore, a potential association of the cause of death with study therapy is unlikely. Also, we recognize our study did not assess quality of life parameters, an important clinical outcome in a patient population with high comorbidities and short expected survival such as the one included in our study. In addition, we recognize it could be arguable whether some of the included patients with AML could have been treated with reduced doses of conventional chemotherapy. However, we believe the age and comorbidities of such patients, as is described in Supplementary Data, favor considering lower intensity forms of therapy and, therefore, allowed inclusion in the present study to be a reasonable therapeutic approach. Finally, we selectively treated patients with AML and MDS using low-intensity therapies and, therefore, we cannot extrapolate our results in terms of toxicity, response and survival to other malignancies or treatment modalities such as high-dose chemotherapy or other forms of therapy. In conclusion, our results support the feasibility of treating patients with MDS or AML not eligible to other clinical trials due to poor PS, comorbidities, other malignancies or organ dysfunction, with low-intensity therapies within a clinical trial. These findings suggest relaxation of such criteria may likely increase the pool of clinical trial patient candidates and allow access to potential beneficial therapies for patients with otherwise dismal prognosis. Leukemia (2017) 318 – 324
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
324 We believe this sets the basis to future larger clinical trials specifically treating this patient population. CONFLICT OF INTEREST GG-M work has been supported by Celgene, maker of azacitidine, and Merck, maker of vorinostat. Merck supported in part this study. GG-M has served as a consultant and speaker for both Celgene and Merck.
ACKNOWLEDGEMENTS Merck Sharp and Dohme Corporation supported this clinical trial (NCT00948064). Support for this project was also provided by a number of other sources, including the MD Anderson Cancer Center Support Grant P30 CA016672, the Dr Kenneth B McCredie Chair in Clinical Leukemia Research endowment, the Edward P Evans Foundation, the Fundacion Ramon Areces, the Cancer Prevention & Research Institute of Texas (CPRIT) award RP140500, and by generous philanthropic contributions to MD Anderson’s MDS/AML Moon Shot Program.
AUTHOR CONTRIBUTIONS G Garcia-Manero, E Estey, C DiNardo, N Pemmaraju, N Daver, Z Estrov, J Cortes, Y Alvarado, S Verstovsek, N Jain, W Wierda, M Konopleva, E Jabbour, G Borthakur, T Kadia, F Ravandi and H Kantarjian designed the study, analyzed data, contributed patients and participated in writing the manuscript. X Huang and C Hsiang-Chun performed the statistical analysis and participated in statistical design. G Montalban-Bravo participated in data collection, analyzing the data and writing the manuscript. M Brant and S Pierce collected and analyzed data. C Bueso-Ramos performed histopathological analysis. H Yang performed correlative studies. C Foudray and T Sneed participated as the research nurses of the study.
REFERENCES 1 Yao B, Zhu L, Jiang Q, Xia HA. Safety monitoring in clinical trials. Pharmaceutics 2013; 5: 94–106. 2 Yanada M, Okamoto A, Inaguma Y, Tokuda M, Morishima S, Kanie T et al. The fate of patients with acute myeloid leukemia not undergoing induction chemotherapy. Int J Hematol 2015; 102: 35–40. 3 Bell JA, Balneaves LG. Cancer patient decision making related to clinical trial participation: an integrative review with implications for patients’ relational autonomy. Support Care Cancer 2015; 23: 1169–1196. 4 Mengis C, Aebi S, Tobler A, Dahler W, Fey MF. Assessment of differences in patient populations selected for excluded from participation in clinical phase III acute myelogenous leukemia trials. J Clin Oncol 2003; 21: 3933–3939. 5 Kantarjian H, O’Brien S, Cortes J, Giles F, Faderl S, Jabbour E et al. Results of intensive chemotherapy in 998 patients age 65 years or older with acute myeloid leukemia or high-risk myelodysplastic syndrome: predictive prognostic models for outcome. Cancer 2006; 106: 1090–1098. 6 Kantarjian H, Ravandi F, O’Brien S, Cortes J, Faderl S, Garcia-Manero G et al. Intensive chemotherapy does not benefit most older patients (age 70 years or older) with acute myeloid leukemia. Blood 2010; 116: 4422–4429. 7 Menzin J, Lang K, Earle CC, Kerney D, Mallick R. The outcomes and costs of acute myeloid leukemia among the elderly. Arch Intern Med 2002; 162: 1597–1603. 8 Sevinc A, Buyukberber S, Camci C, Koruk M, Savas MC, Turk HM et al. Granulocytic sarcoma of the colon and leukemic infiltration of the liver in a patient presenting with hematochezia and jaundice. Digestion 2004; 69: 262–265. 9 Wandroo FA, Murray J, Mutimer D, Hubscher S. Acute myeloid leukaemia presenting as cholestatic hepatitis. J Clin Pathol 2004; 57: 544–545. 10 Ruger W, Kruip MJ, Betjes MG. Reversible renal failure due to bilateral renal sarcoma in a patient with acute myeloid leukemia. Ren Fail 2009; 31: 606–609. 11 Fenaux P, Mufti GJ, Hellstrom-Lindberg E, Santini V, Finelli C, Giagounidis A et al. Efficacy of azacitidine compared with that of conventional care regimens in the treatment of higher-risk myelodysplastic syndromes: a randomised, open-label, phase III study. Lancet Oncol 2009; 10: 223–232. 12 Lee YG, Kim I, Yoon SS, Park S, Cheong JW, Min YH et al. Comparative analysis between azacitidine and decitabine for the treatment of myelodysplastic syndromes. Br J Haematol 2013; 161: 339–347.
13 Thepot S, Itzykson R, Seegers V, Recher C, Raffoux E, Quesnel B et al. Azacitidine in untreated acute myeloid leukemia: a report on 149 patients. Am J Hematol 2014; 89: 410–416. 14 Silva G, Cardoso BA, Belo H, Almeida AM. Vorinostat induces apoptosis and differentiation in myeloid malignancies: genetic and molecular mechanisms. PLoS One 2013; 8: e53766. 15 Schaefer EW, Loaiza-Bonilla A, Juckett M, DiPersio JF, Roy V, Slack J et al. A phase 2 study of vorinostat in acute myeloid leukemia. Haematologica 2009; 94: 1375–1382. 16 Verma A, Odchimar-Reissig R, Feldman EJ, Navada SC, Demakos EP, Baer MR et al. A phase II trial of epigenetic modulators vorinostat in combination with azacitidine (azaC) in patients with the Myelodysplastic Syndrome (MDS): initial results of study 6898 of the New York Cancer Consortium. Blood 2013; 122: 386. 17 Cameron EE, Bachman KE, Myohanen S, Herman JG, Baylin SB. Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer. Nat Genet 1999; 21: 103–107. 18 Esteller M. CpG island methylation and histone modifications: biology and clinical significance. Ernst Schering Res Found Workshop 2006; (57: 115–126. 19 Yang H, Hoshino K, Sanchez-Gonzalez B, Kantarjian H, Garcia-Manero G. Antileukemia activity of the combination of 5-aza-2′-deoxycytidine with valproic acid. Leuk Res 2005; 29: 739–748. 20 Greenberg P, Cox C, LeBeau MM, Fenaux P, Morel P, Sanz G et al. International scoring system for evaluating prognosis in myelodysplastic syndromes. Blood 1997; 89: 2079–2088. 21 Arber DA, Orazi A, Hasserjian R, Thiele J, Borowitz MJ, Le Beau MM et al. The 2016 revision to the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia. Blood 2016; 127: 2391–2405. 22 Lo-Coco F, Avvisati G, Vignetti M, Thiede C, Orlando SM, Iacobelli S et al. Retinoic acid and arsenic trioxide for acute promyelocytic leukemia. N Engl J Med 2013; 369: 111–121. 23 Prebet T, Boissel N, Reutenauer S, Thomas X, Delaunay J, Cahn JY et al. Acute myeloid leukemia with translocation (8;21) or inversion (16) in elderly patients treated with conventional chemotherapy: a collaborative study of the French CBFAML intergroup. J Clin Oncol 2009; 27: 4747–4753. 24 Thall PF, Simon RM, Estey EH. Bayesian sequential monitoring designs for singlearm clinical trials with multiple outcomes. Stat Med 1995; 14: 357–379. 25 Cheson BD, Bennett JM, Kopecky KJ, Buchner T, Willman CL, Estey EH et al. Revised recommendations of the International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. J Clin Oncol 2003; 21: 4642–4649. 26 Cheson BD, Bennett JM, Kantarjian H, Pinto A, Schiffer CA, Nimer SD et al. Report of an international working group to standardize response criteria for myelodysplastic syndromes. Blood 2000; 96: 3671–3674. 27 Daver N, Naqvi K, Jabbour E, Kadia T, DiNardo C, Cardenas-Turanzas M et al. Impact of comorbidities by ACE-27 in the revised-IPSS for patients with myelodysplastic syndromes. Am J Hematol 2014; 89: 509–516. 28 Naqvi K, Garcia-Manero G, Sardesai S, Oh J, Vigil CE, Pierce S et al. Association of comorbidities with overall survival in myelodysplastic syndrome: development of a prognostic model. J Clin Oncol 2011; 29: 2240–2246. 29 Soriano AO, Yang H, Faderl S, Estrov Z, Giles F, Ravandi F et al. Safety and clinical activity of the combination of 5-azacytidine, valproic acid, and all-trans retinoic acid in acute myeloid leukemia and myelodysplastic syndrome. Blood 2007; 110: 2302–2308. 30 Tapper EB, Luptakova K, Joyce RM, Tzachanis D. A 78-year-old man with acute myeloid leukemia (AML) and acute renal failure. Am J Case Rep 2014; 15: 364–367. 31 Bhatnagar B, Duong VH, Gourdin TS, Tidwell ML, Chen C, Ning Y et al. Ten-day decitabine as initial therapy for newly diagnosed patients with acute myeloid leukemia unfit for intensive chemotherapy. Leuk Lymphoma 2014; 55: 1533–1537. 32 Medeiros BC, Satram-Hoang S, Hurst D, Hoang KQ, Momin F, Reyes C. Big data analysis of treatment patterns and outcomes among elderly acute myeloid leukemia patients in the United States. Ann Hematol 2015; 94: 1127–1138. 33 Oran B, Weisdorf DJ. Survival for older patients with acute myeloid leukemia: a population-based study. Haematologica 2012; 97: 1916–1924. 34 Ostgard LS, Norgaard JM, Sengelov H, Severinsen M, Friis LS, Marcher CW et al. Comorbidity and performance status in acute myeloid leukemia patients: a nation-wide population-based cohort study. Leukemia 2015; 29: 548–555.
Supplementary Information accompanies this paper on the Leukemia website (http://www.nature.com/leu)
Leukemia (2017) 318 – 324
© 2017 Macmillan Publishers Limited, part of Springer Nature.
ORIGINAL ARTICLE
A clinical trial for patients with acute myeloid leukemia or myelodysplastic syndromes not eligible for standard clinical trials G Montalban-Bravo1, X Huang2, E Jabbour1, G Borthakur1, CD DiNardo1, N Pemmaraju1, J Cortes1, S Verstovsek1, T Kadia1, N Daver1, W Wierda1, Y Alvarado1, M Konopleva1, F Ravandi1, Z Estrov1, N Jain1, A Alfonso1, M Brandt1, T Sneed1, H-C Chen2, H Yang1, C Bueso-Ramos3, S Pierce1, E Estey1, Z Bohannan1, HM Kantarjian1 and G Garcia-Manero1 Most clinical trials exclude patients with poor performance or comorbidities. To study whether patients with these characteristics can be treated within a clinical trial, we conducted a study for patients with acute myeloid leukemia (AML) or myelodysplastic syndromes (MDS) with poor performance, organ dysfunction or comorbidities. Primary endpoint was 60-day survival. Study included stopping rules for survival and response. Treatment consisted on a combination of azacitidine and vorinostat. Thirty patients (16 with MDS, 14 with AML) were enrolled. Median follow-up was 7.4 months (0.3–29). Sixty-day survival was 83%. No stopping rules were met. Main adverse events (AEs) were grades 1 and 2 gastrointestinal toxicities. In view of these results, we expanded the study and treated 79 additional patients: 27 with azacitidine (AZA) and 52 with azacitidine and vorinostat (AZA+V). Median follow-up was 22.7 months (12.6–47.5). Sixty-day survival rate was 79% (AZA = 67%, AZA+V = 85%, P = 0.07). Median overall survival was 7.6 months (4.5–10.7). Median event-free survival was 4.5 months (3.5–5.6). Main AEs included grades 1 and 2 gastrointestinal toxicities. Our results suggest this subset of patients can be safely treated within clinical trials and derive clinical benefit. Relaxation of standard exclusion criteria may increase the pool of patients likely to benefit from therapy. Leukemia (2017) 31, 318–324; doi:10.1038/leu.2016.303
INTRODUCTION Participation in clinical trials is fundamental for the development of new therapeutic interventions. New drugs are first studied in the context of clinical trials not only to test specific clinical hypothesis, but also to protect patients from unanticipated side effects of such therapy.1 Importantly, for the patient and the physician, there is an expectation that the investigational therapy will have a beneficial clinical effect. This is particularly important for patients for whom there is no optimal standard of care.2 Despite this need, only 3–5% of patients with cancer treated in the United States currently enroll in clinical trials.3 Although enrollment rates may differ regionally, there are a number of reasons for such low rate of participation, with clinical trial eligibility criteria likely representing one of them. Most clinical studies in oncology exclude patients with comorbidities, active or recent malignancies, organ dysfunction or poor performance status. How these criteria protect patients is unclear. Although some of them are based on clinical reasoning it seems that these criteria are in place more to protect the drug or intervention being studied than the patient itself. In addition, such stringent criteria not only may limit the access of ‘unfit’ patients to potentially beneficial therapies, but also limit our ability to extrapolate the safety and efficacy outcomes of new drugs being tested in conventional clinical trials to this subset of patients with unfavorable clinical features.4 Of importance, patients with such characteristics and poor expected outcomes represent the population that
could benefit the most from investigational approaches considering their otherwise dismal prognosis.5–7 This is particularly important in patients with leukemia in whom tissue infiltration or disease burden may result in organ dysfunction or poor performance that could be reversible with appropriate therapy.8–10 In view of this, it seems necessary to determine whether excluding these patients from new therapeutic interventions within monitored clinical trials is ultimately beneficial, and if development of clinical trials for this subset of patients is safe and feasible. To test this concept we designed a clinical trial for patients with acute myeloid leukemia (AML) and myelodysplastic syndromes (MDS) with organ dysfunction, poor performance, other malignancies or other comorbidities. Patients were treated with azacitidine,11–13 a hypomethylating agent, and vorinostat,14,15 a histone deacetylase inhibitor. This treatment modality was selected based on the acceptable toxicity profile of azacitidine,11 and vorinostat,16 and existing preclinical data suggesting a potential synergistic effect of combining hypomethylating agent with histone deacetylase inhibitors.17–19 The study was designed with stopping rules for survival and response. In view of the results of this initial study, suggesting it is both feasible and safe to treat this group of patients within a clinical trial, we expanded the study to increase the treated patient population and allow further experience in the development of clinical trials for this group of patients.
1 Department of Leukemia, University of Texas MD Anderson Cancer Center, Houston, TX, USA; 2Department of Biostatistics, University of Texas, MD Anderson Cancer Center, Houston, TX, USA and 3Department of Hematopathology, University of Texas, MD Anderson Cancer Center, Houston, TX, USA. Correspondence: Dr G Garcia-Manero, Department of Leukemia, University of Texas, MD Anderson Cancer Center, 1515 Holcombe Blvd, Houston 77015, TX, USA. E-mail: [email protected] Received 4 July 2016; revised 21 August 2016; accepted 30 August 2016; accepted article preview online 31 October 2016; advance online publication, 18 November 2016
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
MATERIALS AND METHODS Study design and patient population We conducted a study for patients ineligible to conventional clinical trials, due to organ dysfunction, poor performance, other malignancies or comorbidities at the University of Texas, MD Anderson Cancer Center following institutional guidelines. The study was registered at clinicaltrial.gov as NCT00948064. We enrolled patients older than 17 years of age with previously untreated AML or MDS (intermediate-2 or higher risk by International Prognostic Scoring System20) with at least one of the following inclusion criteria: serum creatinine ⩾ 2 mg/dl, total bilirubin ⩾ 2 mg/dl, ECOG performance status (PS) equal to 3 or 4, or ineligibility to participate on a protocol of higher priority due to comorbidities, other active malignancies or malignancies with a remission period of o2 years. Diagnosis was confirmed at MD Anderson Cancer Center and classified following the revised 2016 WHO critera.21 Pregnant patients were excluded from study participation. Main exclusion criteria included presence of favorable cytogenetic abnormalities, such as inv(16), t(16;16), t(8;21) or t(15;17) due to available effective standard of care therapies,22,23 prior anti-leukemic therapy (except hydroxyurea for cytoreduction in case of proliferative disease); not being able to receive oral medications due to the oral route of administration of vorinostat; prior therapy with histone deacetylase inhibitors; active hepatitis A, B, or C infection due to risk of potential hepatic toxicity with both azacitidine and vorinostat; or known allergy or sensitivity to vorinostat or azacitidine. Study was designed with predefined stopping rules for survival and response that were generated by the use of a Bayesian sequential monitoring design for single-arm clinical trials.24 Multc Lean desktop v1.1 program (https://biostatistics.mdanderson.org/SoftwareDownload/Default. aspx) was used to generate this model. To define the minimum expected survival and response rates that would trigger these stopping rules, prior data of 181 patients treated at our institution who met the study′s eligibility criteria was used. These data were also used to determine the study sample size and patient cohort sizes. In this group, 50% of patients had a survival of at least 60 days, and the complete response (CR) rate was 28%. The maximum sample size was 30. For the survival stopping rule, the study would stop if survival by Day 60 was unlikely (chance o5%) to be at least 20% higher than in the historical group. This would be met if the number of patients alive at ⩾ 60 days was ⩽ 0 of 3 (0/3) patients treated, 2/6, 3/9, 5/12, 6/15, 8/18, 10/21, 11/24, or 13/27. For the response stopping rule, the study would stop if the CR was unlikely (chance o10%) to be higher than in the historical group. This would be met if the number of patients with CR is ⩽ 0/6, 1/12, 2/18, 3/24 or 4/30. After study conclusion, in view of the survival and toxicity results, and after approval from the institutional review board, we expanded the original trial to increase the treated patient population and further explore the feasibility of treating patients with unfavorable clinical characteristics within clinical trials. The study was independently monitored by the Department of Biostatistics at MD Anderson. The primary objective was survival by Day 60. The maximum sample size was 80. The initial 40 patients were randomized in a 1:1 ratio to either azacitidine or azacitidine in combination with vorinostat. As an additional safety measure, treatment assignment for subsequently enrolled patients was performed in an unbalanced manner in favor of the safest treatment modality determined by higher 60-day survival rate. Assignment of patients to a given therapy would stop if 60-day survival rate within a given group was unlikely to be 450%. All calculations were performed using Multc99 and Adaptive Randomization software. This study included a stopping rule for toxicity were by the trial would stop if at any given time average rate of grade 3 or greater non-hematologic toxicity was 420%. Additional details for both phases of the trial are shown in Supplementary Material.
Therapy, baseline evaluation, follow-up during study and response criteria All patients received therapy with 5-azacitidine at a dose of 75 mg/m2 daily on a 5 day schedule (days 1–5). In addition, all patients in the initial exploratory study and those receiving combination therapy in the expansion study also received oral vorinostat at a dose of 200 mg three times a day on the same days as 5-azacitidine. Courses could be repeated every 3–8 weeks. Therapy was planned to continue until disease progression, patient request, or intercurrent illness or until 12 courses of therapy. After 12 courses, patients could continue to receive single agent azacitidine. All patients were evaluated at baseline with a bone marrow aspirate with cytogenetic assessment, complete history and physical examination, © 2017 Macmillan Publishers Limited, part of Springer Nature.
319 complete blood count and chemistries. During the first month, patients were assessed weekly and then monthly. Bone marrow aspiration with cytogenetics, if abnormal at baseline, were performed on Day 28 of Course 1 and afterwards with each course of therapy, to document response or to decide on therapy administration. Once a response was obtained, bone marrow sampling was repeated at the discretion of the treating physician to confirm remission status or document loss of response. Response assessment in patients with AML was performed following the International Working Group 2003 recommendations.25 For patients with MDS, response assessment was performed following the revised 2006 International Working Group criteria.26
Assesment of comorbidities Evaluation of patient comorbidities was performed using the Adult Comorbidity Evaluation-27 (ACE-27) index at baseline. Prior studies have confirmed the use of this index in comorbidity evaluation of patients with MDS.27,28 Outcome of patients in terms of overall survival (OS) was correlated with ACE-27 score.28
Safety evaluation Toxicity was assessed on the basis of adverse events (AEs). All AEs were recorded and coded according to the National Cancer Institute Common Terminology Criteria for Adverse Events v4.0. Toxicity monitoring determined criteria for stopping rule for toxicity.
Correlative studies Analysis of global DNA methylation29 was performed in patients included on the initial exploratory study. Please see Supplementary Methods, Supplementary Table S1 and Supplementary Figure S1 in Supplemental Appendix for details.
Statistical analysis Continuous and categorical variables were evaluated using the Wilcoxon rank sum and Fisher’s exact or χ2 tests, as appropriate. OS was defined as the time interval between treatment start date and death date, and was censored at the last follow-up date for patients who were alive. Event-free survival (EFS) was defined as the time interval between date of treatment start and date of resistance, relapse or death. Patients who were alive with CR and without relapse were censored at the date of last follow-up. The probabilities of OS and EFS were estimated using the Kaplan–Meier methods. Cox proportional hazards regression models were used to assess the association between patient characteristics and OS or EFS. Logistic regression models were conducted to assess the relationship between the patient characteristics and survival at 60 days. Statistical analyses were performed in SAS 9.3 (The SAS Institute, Cary, NC, USA) and Stata 13.1 (Stata Corp, College Station, TX, USA). All P-values were two sided, with significance being Po0.05.
RESULTS Initial exploratory study A total of 30 consecutive patients were enrolled between September 2009 and December 2010. Patient characteristics are shown in Table 1. A total of 14 (47%) patients had AML and 16 (53%) had MDS. Patient specific inclusion criteria and evolution are illustrated in Figure 1. Specific inclusion criteria, number of cycles, survival, response, subsequent therapies and cause of death for each individual patient is detailed in Supplementary Table S4 of Supplementary Material. The median follow-up was 7.4 months (range 0.3–29) with a median number of cycles of therapy of 3.5 (range 1–12). Sixty-day survival rate was 83% (24/30) and the stopping rule for survival was never met. No differences in OS (5.9 months vs 7.9 months; HR = 0.67, 95% CI 0.32–1.4, P = 0.308) and EFS (3.6 vs 4.4 months; HR = 0.74, 95% CI 0.35–1.57, P = 0.432) were observed between patients with AML or MDS (Figure 2). The overall response rate was 40% (12/30) with 8 (27%) patients achieving CR: 4 patients with MDS and 4 with AML. Therefore, the stopping rule for response was never met. Remarkably, two Leukemia (2017) 318 – 324
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
320 Table 1.
Patient characteristics
Characteristics
Age Sex Female Diagnosis AML MDS MDS-SLD MDS-RS MDS-MLD MDS-EB CMML Cytogeneticsa − 5/5q − − 5/5q − and − 7/7q − − 7/7q − +8 Diploid Other Complex Serum creatinine (mg/dl) Median ⩾2 Total bilirubin (mg/dl) Median ⩾2 ECOG performance status 0–1 2 ⩾3 Hemoglobin (g/dl) WBC (×109/l) Platelet count (×109/l) Bone marrow blasts (%) Inclusion criteriab Other malignancy Bilirubin ⩾ 2 mg/dl Creatinine ⩾ 2 mg/dl PS ⩾ 3 Comorbidities Ineligible to other trial
Initial exploratory study (N = 30) n (%)
Extension study (N = 79) Azacitidine (N = 27) n (%)
Azacitidine+vorinostat (N = 52) n (%)
73 (44–83)
71 (54–90)
70 (30–88)
9 (30)
8 (30)
16 (31)
14 16 1 0 2 11 2
(47) (53) (3) (0) (7) (37) (7)
14 (52) 13 (48) 2 (7) 0 (0) 3 (11) 5 (19) 3 (11)
18 (35) 34 (65) 5 (10) 2 (4) 6 (12) 17 (33) 4 (8)
3 (10) 1 (3) 1 (3) 1 (3) 7 (24) 3 (10) 14 (47)
6 (22) 3 (11) 3 (11) 2 (7) 4 (15) 1 (4) 14 (52)
8 (16) 8 (16) 4 (8) 2 (4) 15 (29) 3 (6) 22 (42)
1 (0.3–2.4) 3 (10)
1.2 (0.47–4.76) 5(19)
0.93 (0.38–5.59) 6 (12)
1 (0.3–4.1) 4 (13)
0.9 (0.3–3.3) 3 (11)
0.6 (0.3–3.2) 2 (4)
19 (63) 6 (20) 5 (17) 9.2 (7.2–12) 3 (0.6–112) 41 (12–310) 14 (1–78)
21 (78) 3 (11) 3 (11) 9 (7.7–12.2) 3.8 (0.2–44) 32 (8–354) 17 (1–87)
45 (88) 3 (6) 4 (8) 9.05 (5.8–14.8) 2.45 (0.2–102) 44 (2–714) 8 (1–89)
19 (63) 4 (13) 3 (10) 5 (17) 4 (12) 0 (0)
13(48) 3 (11) 5 (19) 5 (19) 5 (19) 0 (0)
28 (54) 3 (6) 7 (13) 4 (8) 9 (17) 3 (6)
Abbreviations: AML, acute myeloid leukemia; CMML, chronic myelomonocytic leukemia; MDS, myelodysplastic syndrome; MDS-EB, MDS with excess blasts; MDS-MLD, MDS with multilineage dysplasia; MDS-SLD, MDS with single lineage dysplasia; WBC, white blood cell count. aA total of five patients treated with azacitidine+vorinostat in the extension study did not have evaluable conventional cytogenetics at diagnosis due to absence of metaphase growth. bAbsolute numbers and percentages for eligibility criteria are represented in terms of total number of patients meeting each criterion. A total of six patients met more than one inclusion criteria.
patients, one enrolled due to other active malignancy and another due to PS ⩾ 3, underwent allogeneic stem-cell transplantation after achieving CR. In univariate analysis, creatinine or bilirubin ⩾ 2 mg/dl, PS ⩾ 3, ACE-27 ⩾ 2 and presence of other malignancies were not significant predictors of response or survival. Only presence of more than one eligibility criteria was predictive for lower likelihood of 60-day survival (P = 0.04). No significant differences in survival were observed between patients with ACE-27 scores 0–1 compared with 2–3 (6.3 vs 7.0 months; HR = 0.88, 95% CI 0.41–1.91, P = 0.755; Figure 3a). A total of 21 (70%) patients experienced at least one AE (Table 2). In general, therapy was well tolerated with most toxicities being grades 1 and 2. Grades 1 and 2 gastrointestinal toxicity and febrile neutropenia were the most common toxicities. Therapy had to be held or withdrawn due to toxicities in eight patients (27%). Mortality at 4 and 8 weeks was 10% and 20%, respectively. Leukemia (2017) 318 – 324
Expansion study A total of 79 patients were enrolled between September 2011 and March 2014: 27 (34%) received azacitidine (AZA) and 52 (66%) azacitidine in combination with vorinostat (AZA+V). Patient characteristics and inclusion criteria are shown in Table 1. Specific inclusion criteria, number of cycles, survival, response, subsequent therapies and cause of death are detailed in Supplementary Table S5 of Supplementary Material. Survival. The median duration of follow-up was 22.7 months (range 12.6–47.5). Median number of cycles administered was 3 (1–12). Consistent with the initial exploratory study results, 60-day survival was observed in 62 (79%) patients: 18/27 treated with AZA and 44/52 with AZA+V. Median OS was 7.6 months (4.5–10.7) with patients with MDS having significantly longer OS (10.6 vs 4.4 months, HR = 0.5, 95% CI 0.1–8.3, P = 0.07) and EFS (5.7 vs 2.9 months, HR = 0.1, 95% CI 0.3–0.8, P = 0.04) compared with © 2017 Macmillan Publishers Limited, part of Springer Nature.
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
321
Figure 1. Evolution of patients on initial exploratory study. Patients are presented by consecutive enrollment into trial. Patient inclusion criteria are represented by bar color as specified in the legend. Cri, complete response with insufficient hematological recovery (platelets or neutrophils); CRp, complete response with insufficient platelet recovery; HI, hematological improvement; PR, partial response. Absence of any of the previous indicators on a patient survival bar is indicative of no response to therapy.
Figure 2.
Kaplan–Meier estimates of (a) OS and (b) event-free survival of patients in the initial exploratory study.
patients with AML (Figure 4). A total of 31 patients (39%) died during study treatment with 48 patients (61%) receiving subsequent therapy including allogeneic stem-cell transplantation after CR in 8 (10%). Among the 31 patients not receiving subsequent therapy, causes of death included: infectious complication in 13 (42%) and cardiovascular event in 2 (6%) patients with unknown cause of death in 16 (52%). The cause of death was unknown in 38 patients (48%) due to transition to hospice, loss of follow-up or death at an outside institution. By univariate analyses © 2017 Macmillan Publishers Limited, part of Springer Nature.
neither creatinine nor bilirubin levels ⩾ 2 mg/dl, ACE-27 ⩾ 2, presence of other malignancy or PS ⩾ 3 were predictive of 60day survival, OS or EFS (Supplementary Tables S2 and S3 of Supplementary Material). Response. The overall response rate was 47% (37/79) including 13 responses in patients treated with AZA and 24 in patients treated with AZA+V. A total of 20 (25%) patients achieved CR, 13 with MDS and 7 with AML. Median number of cycles to response Leukemia (2017) 318 – 324
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
322
Figure 3. Kaplan–Meier estimates of OS based on ACE-27 score: (a) among patients in the initial exploratory study. (b) Among patients in the extension study. Table 2.
Adverse events
Adverse event
Extension study (N = 79) Initial exploratory study (N = 30)
Constipation Dermatologic (skin rash) Diarrhea Respiratory/cardiac Fatigue Febrile neutropenia Gastrointestinal-other Elevated bilirubin Elevated ALT Mucositis/stomatitis Muscle weakness Nausea Opportunistic infection Pain Urinary frequency Blurred vision Voice changes Vomiting
Azacitidine+vorinostat (N = 52)
Any grade %
Grade ⩾ 3%
Any grade %
Grade ⩾ 3%
Any grade %
12 7 20 3 13 17 0 7 3 3 3 20 3 0 0 0 0 3
0 0 7 3 0 17 0 0 3 0 3 3 3 0 0 0 0 0
22 0 0 0 44 7 0 0 0 0 0 7 11 7 4 0 0 0
4 0 0 0 0 7 0 0 0 0 0 0 11 0 0 0 0 0
31 2 27 25 65 17 2 0 0 2 19 37 2 4 0 6 2 6
was 2 (1–12) with median duration of response of 7.4 months (4.9– 9.9). A total of 44 patients were evaluable for cytogenetic response. Eleven (25%) patients achieved CCyR. Achievement of CCyR was associated with significantly improved OS (28.3 m vs 9.4 m; HR 0.27, 95% CI 0.11–0.67, P = 0.001; Supplementary Figure S2). In univariate analysis, creatinine or bilirubin ⩾ 2 mg/ dl, PS ⩾ 3, ACE-27 ⩾ 2 and presence of other malignancies were not significant predictors of response (Supplementary Tables S2 and S3 of Supplementary Material). Safety and toxicity. A total of 65 (82%) patients experienced at least one AE: 15/27 patients treated with AZA and 42/52 patients treated with AZA+V. Adverse events ranked by grade are shown in Table 2. Grade 3 or higher AEs were experienced by 28 (35%) Leukemia (2017) 318 – 324
Azacitidine (N = 27)
Grade ⩾ 3% 6 0 2 12 23 13 0 0 0 0 0 8 2 0 0 0 0 2
patients: 5/27 patients treated with AZA and 23/52 treated with AZA+V. Grade 3 fatigue, dyspnea and neutropenic fever were the most frequent grade 3 AEs. However, stopping rule for toxicity was not met. At the present time of follow-up, 67 (85%) patients have died. Early mortality at 4 and 8 weeks was 10% and 19%, respectively. Effect of comorbidities on therapy. The most common comorbidities by ACE-27 scoring system included hypertension in 50 (63%), other malignancy in 41 (52%), diabetes mellitus in 22 (28%) and coronary artery disease in 21 (27%). No significant differences in survival were observed between patients with ACE-27 scores 0–1 compared with 2–3 (12.2 vs 7.1 months, HR = 1.2, 95% CI 0.6–2.3, P = 0.591; Figure 3b). © 2017 Macmillan Publishers Limited, part of Springer Nature.
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
323
Figure 4.
Kaplan–Meier estimates of (a) OS and (b) event-free survival of patients in the extension study.
DISCUSSION Standard clinical trial eligibility criteria classically exclude patients with organ dysfunction, poor PS or other comorbidities. Here we show that treating this patient population within a clinical trial is feasible. In our initial exploratory study, a majority of patients were able to complete at least one cycle of therapy without major toxicity, and obtain clinical benefit with acceptable responses and survival despite their high comorbidity burden. The fact that we could compute the toxicity profile, response rates, assess treatment exposure for at least 4 weeks and perform pharmacodynamics assays, strongly support the concept that this group of patients can be included in clinical trials. The presence of renal or hepatic dysfunction did not compromise therapy administration or negatively impact survival and overall response rate. Similarly, patients with prior malignancies or PS ⩾ 3 did not have significantly shorter survival and, in two such patients, therapy allowed achievement of CR and subsequent allogeneic stem-cell transplantation. Furthermore, treatment was able to negate the predictive effect of ACE-27, a previously validated comorbidity index with prognostic significance in MDS,27 further suggesting the likely benefit of therapy irrespective of comorbidity burden. In view of these results, we were able to further explore the feasibility of treating this patient population within a clinical trial by conducting a larger confirmatory study. As in the initial exploratory trial, a majority of patients could complete at least one cycle of therapy with acceptable toxicity, although with an overall higher proportion of AEs and grade 3 AEs when combining vorinostat to azacitidine. Importantly, most patients met the study’s primary endpoint of survival at 60 days of therapy suggesting there is no excess treatment-related toxicity and early mortality in this patient population. In fact, as in the initial exploratory phase, therapy was able to negate the prognostic impact of ACE-27 further suggesting the likely benefit of therapy, irrespective of comorbidity burden, in this patient population. Remarkably, a total of 8 patients were able to proceed to allogeneic stem-cell transplantation. Organ dysfunction8–10,30 and poor PS31 in patients with leukemia may appear in the context of tissue infiltration representing a reversible condition with appropriate therapy, and should not always jeopardize treatment. The presence of other comorbidity, particularly when chronic, or if the patient is not anticipated to succumb to it for a short period, should not © 2017 Macmillan Publishers Limited, part of Springer Nature.
impact the clinical information that could be obtained when treating this group of patients on a clinical trial. We hypothesized, that this group of patients, with dismal prognosis if untreated,32–34 would have a more favorable risk/benefit ratio when treated with an investigational intervention than patients traditionally included in such studies. In fact, a previous retrospective study published by Mengis et al.4 evaluating the outcome of patients with AML treated at their institution, suggested patients with comorbidities who were not eligible to conventional trials had similar OS and disease-free survival when treated with off protocol intensive chemotherapy compared with patients treated within a clinical protocol with similar regimens. We recognize our study has several limitations including the heterogeneity in patient population in terms of diagnosis, comorbidities and clinical features which limits our ability to extrapolate the outcomes to specific patient populations. In addition, there are a significant proportion of patients in whom the cause of death was not able to be determined. However, most of such patients died after transition to hospice and, therefore, a potential association of the cause of death with study therapy is unlikely. Also, we recognize our study did not assess quality of life parameters, an important clinical outcome in a patient population with high comorbidities and short expected survival such as the one included in our study. In addition, we recognize it could be arguable whether some of the included patients with AML could have been treated with reduced doses of conventional chemotherapy. However, we believe the age and comorbidities of such patients, as is described in Supplementary Data, favor considering lower intensity forms of therapy and, therefore, allowed inclusion in the present study to be a reasonable therapeutic approach. Finally, we selectively treated patients with AML and MDS using low-intensity therapies and, therefore, we cannot extrapolate our results in terms of toxicity, response and survival to other malignancies or treatment modalities such as high-dose chemotherapy or other forms of therapy. In conclusion, our results support the feasibility of treating patients with MDS or AML not eligible to other clinical trials due to poor PS, comorbidities, other malignancies or organ dysfunction, with low-intensity therapies within a clinical trial. These findings suggest relaxation of such criteria may likely increase the pool of clinical trial patient candidates and allow access to potential beneficial therapies for patients with otherwise dismal prognosis. Leukemia (2017) 318 – 324
A study for patients with MDS and AML ineligible to clinical trials G Montalban-Bravo et al
324 We believe this sets the basis to future larger clinical trials specifically treating this patient population. CONFLICT OF INTEREST GG-M work has been supported by Celgene, maker of azacitidine, and Merck, maker of vorinostat. Merck supported in part this study. GG-M has served as a consultant and speaker for both Celgene and Merck.
ACKNOWLEDGEMENTS Merck Sharp and Dohme Corporation supported this clinical trial (NCT00948064). Support for this project was also provided by a number of other sources, including the MD Anderson Cancer Center Support Grant P30 CA016672, the Dr Kenneth B McCredie Chair in Clinical Leukemia Research endowment, the Edward P Evans Foundation, the Fundacion Ramon Areces, the Cancer Prevention & Research Institute of Texas (CPRIT) award RP140500, and by generous philanthropic contributions to MD Anderson’s MDS/AML Moon Shot Program.
AUTHOR CONTRIBUTIONS G Garcia-Manero, E Estey, C DiNardo, N Pemmaraju, N Daver, Z Estrov, J Cortes, Y Alvarado, S Verstovsek, N Jain, W Wierda, M Konopleva, E Jabbour, G Borthakur, T Kadia, F Ravandi and H Kantarjian designed the study, analyzed data, contributed patients and participated in writing the manuscript. X Huang and C Hsiang-Chun performed the statistical analysis and participated in statistical design. G Montalban-Bravo participated in data collection, analyzing the data and writing the manuscript. M Brant and S Pierce collected and analyzed data. C Bueso-Ramos performed histopathological analysis. H Yang performed correlative studies. C Foudray and T Sneed participated as the research nurses of the study.
REFERENCES 1 Yao B, Zhu L, Jiang Q, Xia HA. Safety monitoring in clinical trials. Pharmaceutics 2013; 5: 94–106. 2 Yanada M, Okamoto A, Inaguma Y, Tokuda M, Morishima S, Kanie T et al. The fate of patients with acute myeloid leukemia not undergoing induction chemotherapy. Int J Hematol 2015; 102: 35–40. 3 Bell JA, Balneaves LG. Cancer patient decision making related to clinical trial participation: an integrative review with implications for patients’ relational autonomy. Support Care Cancer 2015; 23: 1169–1196. 4 Mengis C, Aebi S, Tobler A, Dahler W, Fey MF. Assessment of differences in patient populations selected for excluded from participation in clinical phase III acute myelogenous leukemia trials. J Clin Oncol 2003; 21: 3933–3939. 5 Kantarjian H, O’Brien S, Cortes J, Giles F, Faderl S, Jabbour E et al. Results of intensive chemotherapy in 998 patients age 65 years or older with acute myeloid leukemia or high-risk myelodysplastic syndrome: predictive prognostic models for outcome. Cancer 2006; 106: 1090–1098. 6 Kantarjian H, Ravandi F, O’Brien S, Cortes J, Faderl S, Garcia-Manero G et al. Intensive chemotherapy does not benefit most older patients (age 70 years or older) with acute myeloid leukemia. Blood 2010; 116: 4422–4429. 7 Menzin J, Lang K, Earle CC, Kerney D, Mallick R. The outcomes and costs of acute myeloid leukemia among the elderly. Arch Intern Med 2002; 162: 1597–1603. 8 Sevinc A, Buyukberber S, Camci C, Koruk M, Savas MC, Turk HM et al. Granulocytic sarcoma of the colon and leukemic infiltration of the liver in a patient presenting with hematochezia and jaundice. Digestion 2004; 69: 262–265. 9 Wandroo FA, Murray J, Mutimer D, Hubscher S. Acute myeloid leukaemia presenting as cholestatic hepatitis. J Clin Pathol 2004; 57: 544–545. 10 Ruger W, Kruip MJ, Betjes MG. Reversible renal failure due to bilateral renal sarcoma in a patient with acute myeloid leukemia. Ren Fail 2009; 31: 606–609. 11 Fenaux P, Mufti GJ, Hellstrom-Lindberg E, Santini V, Finelli C, Giagounidis A et al. Efficacy of azacitidine compared with that of conventional care regimens in the treatment of higher-risk myelodysplastic syndromes: a randomised, open-label, phase III study. Lancet Oncol 2009; 10: 223–232. 12 Lee YG, Kim I, Yoon SS, Park S, Cheong JW, Min YH et al. Comparative analysis between azacitidine and decitabine for the treatment of myelodysplastic syndromes. Br J Haematol 2013; 161: 339–347.
13 Thepot S, Itzykson R, Seegers V, Recher C, Raffoux E, Quesnel B et al. Azacitidine in untreated acute myeloid leukemia: a report on 149 patients. Am J Hematol 2014; 89: 410–416. 14 Silva G, Cardoso BA, Belo H, Almeida AM. Vorinostat induces apoptosis and differentiation in myeloid malignancies: genetic and molecular mechanisms. PLoS One 2013; 8: e53766. 15 Schaefer EW, Loaiza-Bonilla A, Juckett M, DiPersio JF, Roy V, Slack J et al. A phase 2 study of vorinostat in acute myeloid leukemia. Haematologica 2009; 94: 1375–1382. 16 Verma A, Odchimar-Reissig R, Feldman EJ, Navada SC, Demakos EP, Baer MR et al. A phase II trial of epigenetic modulators vorinostat in combination with azacitidine (azaC) in patients with the Myelodysplastic Syndrome (MDS): initial results of study 6898 of the New York Cancer Consortium. Blood 2013; 122: 386. 17 Cameron EE, Bachman KE, Myohanen S, Herman JG, Baylin SB. Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer. Nat Genet 1999; 21: 103–107. 18 Esteller M. CpG island methylation and histone modifications: biology and clinical significance. Ernst Schering Res Found Workshop 2006; (57: 115–126. 19 Yang H, Hoshino K, Sanchez-Gonzalez B, Kantarjian H, Garcia-Manero G. Antileukemia activity of the combination of 5-aza-2′-deoxycytidine with valproic acid. Leuk Res 2005; 29: 739–748. 20 Greenberg P, Cox C, LeBeau MM, Fenaux P, Morel P, Sanz G et al. International scoring system for evaluating prognosis in myelodysplastic syndromes. Blood 1997; 89: 2079–2088. 21 Arber DA, Orazi A, Hasserjian R, Thiele J, Borowitz MJ, Le Beau MM et al. The 2016 revision to the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia. Blood 2016; 127: 2391–2405. 22 Lo-Coco F, Avvisati G, Vignetti M, Thiede C, Orlando SM, Iacobelli S et al. Retinoic acid and arsenic trioxide for acute promyelocytic leukemia. N Engl J Med 2013; 369: 111–121. 23 Prebet T, Boissel N, Reutenauer S, Thomas X, Delaunay J, Cahn JY et al. Acute myeloid leukemia with translocation (8;21) or inversion (16) in elderly patients treated with conventional chemotherapy: a collaborative study of the French CBFAML intergroup. J Clin Oncol 2009; 27: 4747–4753. 24 Thall PF, Simon RM, Estey EH. Bayesian sequential monitoring designs for singlearm clinical trials with multiple outcomes. Stat Med 1995; 14: 357–379. 25 Cheson BD, Bennett JM, Kopecky KJ, Buchner T, Willman CL, Estey EH et al. Revised recommendations of the International Working Group for Diagnosis, Standardization of Response Criteria, Treatment Outcomes, and Reporting Standards for Therapeutic Trials in Acute Myeloid Leukemia. J Clin Oncol 2003; 21: 4642–4649. 26 Cheson BD, Bennett JM, Kantarjian H, Pinto A, Schiffer CA, Nimer SD et al. Report of an international working group to standardize response criteria for myelodysplastic syndromes. Blood 2000; 96: 3671–3674. 27 Daver N, Naqvi K, Jabbour E, Kadia T, DiNardo C, Cardenas-Turanzas M et al. Impact of comorbidities by ACE-27 in the revised-IPSS for patients with myelodysplastic syndromes. Am J Hematol 2014; 89: 509–516. 28 Naqvi K, Garcia-Manero G, Sardesai S, Oh J, Vigil CE, Pierce S et al. Association of comorbidities with overall survival in myelodysplastic syndrome: development of a prognostic model. J Clin Oncol 2011; 29: 2240–2246. 29 Soriano AO, Yang H, Faderl S, Estrov Z, Giles F, Ravandi F et al. Safety and clinical activity of the combination of 5-azacytidine, valproic acid, and all-trans retinoic acid in acute myeloid leukemia and myelodysplastic syndrome. Blood 2007; 110: 2302–2308. 30 Tapper EB, Luptakova K, Joyce RM, Tzachanis D. A 78-year-old man with acute myeloid leukemia (AML) and acute renal failure. Am J Case Rep 2014; 15: 364–367. 31 Bhatnagar B, Duong VH, Gourdin TS, Tidwell ML, Chen C, Ning Y et al. Ten-day decitabine as initial therapy for newly diagnosed patients with acute myeloid leukemia unfit for intensive chemotherapy. Leuk Lymphoma 2014; 55: 1533–1537. 32 Medeiros BC, Satram-Hoang S, Hurst D, Hoang KQ, Momin F, Reyes C. Big data analysis of treatment patterns and outcomes among elderly acute myeloid leukemia patients in the United States. Ann Hematol 2015; 94: 1127–1138. 33 Oran B, Weisdorf DJ. Survival for older patients with acute myeloid leukemia: a population-based study. Haematologica 2012; 97: 1916–1924. 34 Ostgard LS, Norgaard JM, Sengelov H, Severinsen M, Friis LS, Marcher CW et al. Comorbidity and performance status in acute myeloid leukemia patients: a nation-wide population-based cohort study. Leukemia 2015; 29: 548–555.
Supplementary Information accompanies this paper on the Leukemia website (http://www.nature.com/leu)
Leukemia (2017) 318 – 324
© 2017 Macmillan Publishers Limited, part of Springer Nature.
