Abstract
Background:
Despite approval of azacitidine in 2004 and decitabine in 2006 in the US for chronic myelomonocytic leukemia (CMML), overall survival (OS) benefit with hypomethylating agent (HMA) therapy is unclear.
Methods:
We selected older adults (age ≥66 years) diagnosed with CMML during 2001–2011 from SEER-Medicare and used propensity score to match patients diagnosed after HMA approval (2007–11) who received HMA treatment with patients diagnosed before HMA approval (2001–03). Cox proportional hazards models with the matched sample were used to assess the change in OS. A second matched cohort between those who did not receive HMA after approval and patients diagnosed before HMA approval were used to evaluate survival change attributable to other potential changes between the two time periods such as improved supportive care.
Results:
Among 1,378 older adults diagnosed with CMML, median OS was 13 months, and 18.8% received HMAs. In the primary matched analysis, with 225 HMA users diagnosed in 2007–2011 compared to 395 patients diagnosed in 2001–2003, OS was 17 vs. 11 months (hazard ratio [HR] =0·72, 95% confidence interval [CI]: 0·58–0·91; P=·01). In a secondary analysis, the risk of death was not different between 395 propensity score matched HMA non-users diagnosed in 2007–11 compared to 484 patients diagnosed in 2001–03 (HR=1·09, 95%CI, 0·91–1·32, P=·34).
Conclusions:
Despite limited evidence, HMAs are commonly used to treat older CMML patients. Use of HMAs was associated with a 28% reduction in risk of death in adjusted analyses. Improvement in supportive care does not solely account for temporal improvement in OS.
Keywords: Chronic myelomonocytic leukemia (CMML), survival, effectiveness, hypomethylating agents, azacitidine, decitabine, SEER, Medicare
PRECIS:
Despite limited evidence, HMAs are commonly used to treat older CMML patients. This study demonstrated that the use of HMAs was associated with a 28% reduction in risk of death in adjusted analyses. Improvement in supportive care does not solely account for temporal improvement in OS.
Background
Chronic myelomonocytic leukemia (CMML) is a rare hematologic malignancy with an estimated age-adjusted annual incidence in the United States (US) of 0·3–0·4 per 100,000 persons.1 CMML was long classified as a subtype of myelodysplastic syndromes (MDS) before it was placed in a new separate category (MDS/myeloproliferative neoplasms [MDS/MPN] overlap) in the World Health Organization (WHO) 2008 classification.2 Epidemiologic, biologic and therapeutic evaluations of patients with CMML were historically conducted within MDS studies and not as a separate group. Therefore, most of our knowledge about therapeutic interventions are derived from large randomized MDS trials that included small subsets of CMML patients or from small single-center trials.1,3
Older adults with CMML classified as lower-risk by various risk stratification tools are generally treated with observation or supportive measures. The management of higher-risk patients may include the use of hypomethylating agents (HMAs) azacitidine or decitabine with the goal of extending survival. Very few patients undergo allogeneic hematopoietic stem cell transplantation (alloSCT), the only potentially curative approach.
The use of HMAs in large randomized trials in patients with higher risk MDS led to hematologic responses, quality of life improvements, and survival improvement (azacitidine only), resulting in the Food and Drug Administration approval of azacitidine in 2004 and decitabine in 2006 in the US.4–7 Both drugs were also approved for use in CMML because the disease was still classified as a subtype of MDS. However, a survival advantage with HMA therapy has not been demonstrated specifically in CMML.
The Surveillance, Epidemiology, and End Results (SEER)-Medicare database is particularly well suited to study patterns of care, outcomes, and effectiveness of therapies of rare malignancies in whom the majority of patients are age 65 years or older at time of diagnosis, including CMML (median age at diagnosis, 76 years with 84% of patients ≥65 years in age at diagnosis).8–10 In a prior SEER-Medicare analysis of patients diagnosed with CMML in 2001–2005, we demonstrated that patients diagnosed with CMML have a significantly shorter median overall survival (OS; 13·3 vs. 23·3 months; P<·0001) and lower 3-year survival rate (19% vs. 36%; P<·0001) than patients with MDS.11 However, the majority of patients in this analysis was diagnosed before the approval of the first HMA, and only 5·8% of the 792 patients in this cohort received a HMA.11 In the current study, we evaluated patterns of HMA use in CMML in the US over a longer time span with significant coverage of the post-HMA approval period (patients diagnosed in 2001–2011), and to study the impact of HMA use on survival using a large population-based database.
As observational studies are subject to selection bias based on unobserved factors, we sought to take advantage of the market entry of HMAs in the US in 2004–2006 as a natural experiment to study the effect of HMAs on survival. In this study, our primary analysis compares older adults diagnosed with CMML and treated with HMAs after 2006 to those diagnosed prior to 2004, the year of first HMA approval, who had similar characteristics to those in the post-2006 period. An alternative hypothesis is that general improvements in supportive care improved survival over time. To test this alternative, we compared survival for older adults diagnosed after 2006 who did not receive HMAs to those diagnosed in the pre-HMA approval period.
Methods
Data Source and Study Sample
In this retrospective cohort study, we used the SEER-Medicare linked database. The SEER database captures information on a longitudinal cohort with data on cancer diagnosis and selected demographics derived from cancer registries in 17 regions in the US. SEER data are linked to Medicare enrollment and claims files,10–12 which contain detailed individual-level information regarding healthcare utilization and costs of care.13–14
The study cohort included Medicare beneficiaries diagnosed with CMML in 2001–2003 and 2007–2011, who had no additional MDS diagnosis and were observed until death or end of follow-up on 12/31/2011 (diagnosis in 2001–2003) or 12/31/2013 (diagnosis in 2007–2011). The diagnosis of CMML was ascertained using International Classification of Disease for Oncology, 3rd edition code of 9945 in the SEER registry.15 Date of diagnosis is assigned the first day of the month of diagnosis provided by SEER. Inclusion criteria included age 66 years or older at diagnosis, continuously enrolled in Medicare Parts A and B, and not enrolled in Medicare Advantage for the 12 months preceding CMML diagnosis until death or end of follow-up. Patients were excluded if their diagnosis was reported from death certificate or autopsy only, they had a missing diagnosis date or received a HMA before CMML diagnosis [Figure 1]. The Yale Human Investigation Committee determined that this study did not directly involve human subjects.
Figure 1. Study sample selection.
CMML, Chronic Myelomonocytic Leukemia; MDS, Myelodysplastic syndromes; AB, Medicare Parts A and Part B; HMO, Health Maintenance Organization.
Variables of Interest
Our key variable of interest was the receipt of any HMA. We identified therapies including the HMAs, erythropoietin stimulating agents (ESAs), and alloSCT using the Healthcare Common Procedural Coding System (HCPCS) codes on claims (Supplementary Table 1). HMA receipt was identified as any claims for azacitidine or decitabine administration from CMML diagnosis. We also counted the number of HMA cycles, with a cycle defined as at least three distinct days of HMA use, separated by a gap of two weeks or more. ESAs were assessed from 12 months before through the date of CMML diagnosis. As classical disease-specific measures of CMML severity such as cytogenetics and blood/bone marrow blast proportions were not available in the dataset, we used alternative measures of disease severity including transfusion needs and hospitalization for bleeding or infections. We used HCPCS codes to identify transfusions of red blood cells (RBC) and platelets from eight weeks prior to diagnosis to four weeks after diagnosis. We also observed for hospitalizations with a diagnosis of bleeding or infection from six months prior to diagnosis through one month after diagnosis month (eight months total).
For patient-level covariates, we used the census tract level (zip code level if not available) measures linked to SEER-Medicare including median household income and percentage of adults with high school education or less. In the 12-month period before the diagnosis date, we identified the number of comorbid conditions based on the approach developed by Elixhauser et al.16 We required the relevant diagnosis codes to be present in either a single inpatient claim or ≥ 2 physician or outpatient claims, with at least two that were ≥30 days apart for the comorbidity to be considered present. Disability status, a validated claims-based proxy for performance status, was assessed using claims in the year prior to diagnosis.17,18 Finally, we calculated OS from the date of diagnosis to death or end of study, whichever came first.
Statistical Analysis
We used chi-squared tests to evaluate categorical variables and Student’s t-test to study continuous variables in comparing the distributions of clinical and demographic covariates between different groups of CMML patients.
We classified the sample into three analytic subsamples based on the approval dates and HMA treatment [Figure 2]. The three groups were A) diagnosed after 2006 (2007–2011) and received HMA; B) diagnosed after 2006 but did not receive HMA; and C) diagnosed prior to HMA approval (2001–2003). There were 18 patients diagnosed in 2001–2003 survived long enough through HMA approval and received HMA. Given assumed survival benefits of HMA, inclusion of these 18 patients may bias our results towards the null, whereas exclusion of these 18 patients may cause bias away from the null. Therefore, we decided to adopt the more conservative approach and kept them in the pre-HMA approval group. We conducted a multivariable logistic regression to estimate HMA treatment propensity using the post-approval samples (A & B). The covariates in the propensity model included age at diagnosis, sex, race, census tract level education and income, metro residency, Medicaid dual eligibility, Elixhauser comorbidity index, disability status, hospitalizations for bleeding or infection, RBC and platelet transfusions. The model estimates were used to generate treatment propensity scores (PS) for the entire sample (including pre-approval patients). Our primary analysis focused on group A compared to PS-matched patients in group C. The secondary analysis compared group B with PS-matched patients from group C to evaluate other potential differences between the two time periods (2001–2003 vs. 2007–2011) such as supportive care on survival. We used Kaplan-Meier methods and log-rank test to compare unadjusted survival between HMA users and HMA non-users. We ran Cox proportional hazards models to assess the differences in survival using the PS matched sample. We also report Kaplan-Meier survival estimates and differences at from 12–60 months from the matched sample using a boot strapping approach to calculate confidence intervals, which are not subject to the proportional hazards assumption of a Cox model. Finally, we used Kaplan-Meier methods and log-rank test to compare unadjusted survival between azacitidine- and decitabine-treated patients. We used SAS version 9.4 (SAS Institute, Inc., Cary, NC) to conduct all analyses, with two-sided statistical tests and an alpha of 0·05.
Figure 2. Propensity scoring matching methodology.
HMA, Hypomethylating agents; PS, Propensity Score; PH, Proportional Hazards.
Results
A total of 1,378 patients with a CMML diagnosis in SEER-Medicare met the eligibility criteria [Figure 1]. The median age was 79 years [Interquartile range (IQR): 73–84], 89·7% were white, and 60·6% were male. A total of 1,242 patients (90·1%) died during follow-up. The median follow-up from diagnosis was 13 (IQR: 4–29) months. The median OS of the entire cohort from date of CMML diagnosis was 13 (95% CI: 12–14) months.
During the follow-up period, 259 patients (18·8%) received at least one dose of a HMA including 119 patients who received only azacitidine (8·6%), 94 who received decitabine (6·8%), and 46 patients (3·3%) who received both agents at some point during follow-up [Table 1]. Among CMML patients diagnosed in pre-HMA approval period (2001–2003), only a small number (18, 3·6%) subsequently received a HMA. For patients who diagnosed from 2007 to 2011, the proportion of HMA recipients ranged from 23·5% to 31·0% each year [Figure 3]. Among those who received a HMA, the median duration from diagnosis to first initiation of the HMA was 2 (IQR: 1–10) months. Only 14 patients underwent alloSCT during follow-up (1%). In unadjusted analyses, the OS of patients diagnosed after 2006 was significantly better compared to those diagnosed in the pre-approval period (median OS 14 [95%CI: 12–16] months vs. 10 [95% CI: 8–12] months, P=·01, Supplementary Figure 1]). Regardless of the year of diagnosis, patients who received HMAs had an OS of 18 (95% CI: 15–20) months compared to those who did not receive any HMAs (median OS 11 [95% CI: 10–12] months, P<·001)[Supplementary Table 2]. We also compared the OS among HMA users per the type of HMA received (excluding the patients who received both drugs) and found no difference in OS in unadjusted analysis (median OS of 18 months [95%CI, 14–22] for azacitidine-treated patients [n=119] vs. 20 months [95%CI, 16–25] for decitabine-treated patients [n=94], p=0.40, Figure 4).
Table 1.
Baseline characteristics and demographics (N=1,378). HMA, Hypomethylating agents; ESA, Erythropoiesis-stimulating agent; RBC, Red blood cell; alloSCT, allogeneic hematopoietic stem cell transplantation.
| N | % | |
|---|---|---|
| Total | 1378 | 100.0 |
| Year of Diagnosis | ||
| 2001–2003 | 497 | 26.6 |
| 2007–2009 | 536 | 28.7 |
| 2010–2011 | 345 | 18.5 |
| HMA | ||
| No | 1119 | 81.2 |
| Yes | 259 | 18.8 |
| HMA Type | ||
| Azacitidine only | 119 | 8.6 |
| Decitabine only | 94 | 6.8 |
| Both | 46 | 3.3 |
| Age Group | ||
| 66–69 | 147 | 10.7 |
| 70–74 | 255 | 18.5 |
| 75–79 | 310 | 22.5 |
| 80+ | 666 | 48.3 |
| Sex | ||
| Female | 543 | 39.4 |
| Male | 835 | 60.6 |
| Race | ||
| White | 1236 | 89.7 |
| Other | 142 | 10.3 |
| Hispanic | ||
| Non-Hispanic | 1314 | 95.4 |
| Hispanic | 64 | 4.6 |
| Marital Status | ||
| Married | 739 | 53.6 |
| Unmarried | 524 | 38.0 |
| Other | 115 | 8.3 |
| High School Education (% adults 25+ with <= HS education) | ||
| <33% | 406 | 29.5 |
| 33%−66% | 774 | 56.2 |
| ≥ 66% | 198 | 14.4 |
| Median Household Income | ||
| <$33,000 | 289 | 21.0 |
| $33,000–40,000 | 208 | 15.1 |
| $40,000–50,000 | 291 | 21.1 |
| ≥ $50,000 | 590 | 42.8 |
| Metro Residency | ||
| Metro | 1128 | 81.9 |
| Non-metro | 250 | 18.1 |
| Medicaid Dual Eligibility | ||
| No | 1219 | 88.5 |
| Yes | 159 | 11.5 |
| Elixhauser Comorbidity Index | ||
| None | 447 | 32.4 |
| 1 to 2 | 500 | 36.3 |
| More than 3 | 431 | 31.3 |
| Disability status | ||
| Not disabled | 1231 | 89.3 |
| Disabled | 147 | 10.7 |
| Hospitalization for infection or bleeding | ||
| No | 1058 | 76.8 |
| Yes | 320 | 23.2 |
| RBC transfusion | ||
| Naïve | 1304 | 94.6 |
| User | 49 | 3.6 |
| Dependent | 25 | 1.8 |
| Prior Platelet transfusion | ||
| No | 1342 | 97.4 |
| Yes | 36 | 2.6 |
| ESA Use | ||
| No | 1266 | 91.9 |
| Yes | 112 | 8.1 |
| Allogeneic hematopoietic stem cell transplantation | ||
| No | 1364 | 99.0 |
| Yes | 14 | 1.0 |
Abbreviations: HMA, Hypomethylating agents; ESA, Erythropoiesis-stimulating agent; RBC, Red blood cell
Figure 3. Proportion of patients with CMML who received a HMA by year of diagnosis.
CMML, Chronic Myelomonocytic Leukemia; HMA, Hypomethylating agents.
Figure 4. Kaplan Meier survival estimates for HMA Users (N=213) stratified by Azacitidine Only (N=119) and Decitabine Only (N=94).
HMA, Hypomethylating agents. Number at risk at 48 months not reported to meet data use requirements (Number decreased from 36 months to 48 months <11)
The median number of HMA cycles received was 5 cycles (IQR: 2–11), with 63% of HMA users receiving 4 or more cycles and 46% receiving 6 or more cycles, indicating adequate exposure to the drugs which typically require prolonged exposure to provide clinical benefit. The median OS of those patients who received 4 or more cycles of HMAs was 22 (95% CI: 19–25) months.
Propensity score matching and survival
Predictors of HMAs receipt for patients diagnosed after 2006 are shown in supplementary Table 3. Older patients and patients who were hospitalized for bleeding or infection were less likely to receive HMAs, while those who underwent RBC or platelet transfusions were more likely to receive HMAs. Standardized differences between the 225 PS-matched HMA users who were diagnosed in 2007–2011 (group A) and the 395 PS-matched patients who were diagnosed in 2001–2003 (Group C) suggested that the distribution of covariates was balanced after matching [Table 2]. The median OS was 17 months compared to 11 months, respectively, which translated into a significantly improved OS with a hazard ratio (HR) of 0·72 (95% CI: 0·58–0·91, P=·01) [Figure 5A]. The probability of survival was higher among HMA users at one year after diagnosis (0·64 [95% CI: 0·57–0·70] vs. 0·47 [95% CI: 0·42–0·52]) but difference disappeared by the 2-year mark and thereafter [Figure 4A]. Small sample sizes precluded a comparison between azacitidine and decitabine for effects on survival.
Table 2.
Patient characteristics after PS matching (HMA users post approval, N=225; and all CMML patients pre approval, N=395). HMA, Hypomethylating agents; ESA, Erythropoiesis-stimulating agent; RBC, Red blood cell.
| All before 2004 | HMA users after 2006 | P-value | Standardized Difference | |
|---|---|---|---|---|
| Total N | 395 | 225 | 8.79 | |
| Age Group | 0.31 | |||
| 66–69 | 12.9% | 17.8% | 13.53 | |
| 70–74 | 23.3% | 21.8% | −3.62 | |
| 75–79 | 27.6% | 23.1% | −10.32 | |
| 80+ | 36.2% | 37.3% | 2.35 | |
| Sex | .75 | |||
| Female | 35.9% | 34.7% | −2.68 | |
| Male | 64.1% | 65.3% | 2.68 | |
| Race | .93 | |||
| White | 90.9% | 90.7% | −0.76 | |
| Other | 9.1% | 9.3% | 0.76 | |
| Hispanic | .36 | |||
| Non-Hispanic | 97.0% | >95.1% | 9.77 | |
| Hispanic | 3.0% | <4.9% | −9.77 | |
| Marital Status | .19 | |||
| Married | 60.3% | 57.8% | −5.03 | |
| Unmarried | 32.2% | 30.2% | −4.17 | |
| Other | 7.6% | 12.0% | 14.86 | |
| Percent adults 25+ with <= High school education | .79 | |||
| <33% | 28.6% | 26.2% | −5.35 | |
| 33%−66% | 57.0% | 59.6% | 5.26 | |
| ≥ 66% | 14.4% | 14.2% | −0.59 | |
| Median Household Income | .58 | |||
| <$33,000 | 20.0% | 18.2% | −4.52 | |
| $33,000–40,000 | 17.5% | 15.1% | −6.39 | |
| $40,000–50,000 | 18.5% | 22.7% | 10.37 | |
| ≥ $50,000 | 44.1% | 44.0% | −0.10 | |
| Metro Residency | .34 | |||
| Metro | 79.5% | 82.7% | 8.11 | |
| Non-metro | 20.5% | 17.3% | −8.11 | |
| Medicaid Dual Eligibility | .15 | |||
| No | 94.2% | 91.1% | −11.77 | |
| Yes | 5.8% | 8.9% | 11.77 | |
| Elixhauser Comorbidity Index | .092 | |||
| None | 41.3% | 35.6% | −11.76 | |
| 1 to 2 | 39.0% | 37.3% | −3.41 | |
| more than 3 | 19.7% | 27.1% | 17.45 | |
| Disability | .46 | |||
| Not disabled | 95.9% | 94.7% | −6.07 | |
| Disabled | 4.1% | 5.3% | 6.07 | |
| Hospitalization for infection or bleeding | .46 | |||
| No | 83.5% | 85.8% | 6.20 | |
| Yes | 16.5% | 14.2% | −6.20 | |
| RBC transfusion | .78 | |||
| Naïve | 94.2% | 92.9% | 5.29 | |
| User | >3.0% | <4.9% | −9.77 | |
| Dependent | <2.8% | <4.9% | −10.93 | |
| Prior Platelet transfusion | .61 | |||
| No | >97.2% | >95.1% | 10.93 | |
| Yes | <2.8% | <4.9% | −10.93 |
Abbreviations: HMA, Hypomethylating agents; ESA, Erythropoiesis-stimulating agent; RBC, Red blood cell. Values with < or > are suppressed according to our data use agreement not reveal cell sizes <11. Standardized differences reflect values shown in table if a value is suppressed.
Figure 5A. Kaplan Meier survival estimates for Primary Propensity Score matched cohort (N=620) stratified by HMA User (N=225) and Pre-HMA (N=395).
HMA, Hypomethylating agents.
We matched 395 patients who were diagnosed in 2001–2003 (Group C) with 484 non-HMA users diagnosed in 2007–2011 (Group B). The risk of death from diagnosis was not significantly different between non-HMA users diagnosed in 2007–2011 (Group B) compared to patients diagnosed in 2001–2003 (Group C) (HR=1·09, 95% CI: 0·91–1·32, P=·34) [Figure 5B].
Figure 5B. Kaplan Meier survival estimates for Secondary Propensity Score matched cohort (N=879) stratified by Non-HMA User after 2006 (N=395) and Pre-HMA (N=484).
HMA, Hypomethylating agents.
Discussion
The results of our study, based on one of the largest cohorts of older adults with CMML, indicate a significant survival benefit associated with receipt of HMA. Specifically, we found that HMA use was associated with a 28% of reduction in risk of death for older CMML patients who were diagnosed after 2006 and received HMAs compared to PS-matched patients who were diagnosed in 2001–2003. It does not seem that this significant improvement in survival can be simply explained by improvement in supportive care or other potential changes in patient characteristics or treatments over the study period as there was no change in OS between the untreated in the post-HMA period compared to those diagnosed pre-HMA approval. Importantly, we noticed that the OS difference is significant in the first year after diagnosis but the benefit in OS associated with HMA use disappears by two years from diagnosis. This observation of survival benefit being limited to the first 24 months after initiating therapy with HMAs is in line with what was observed in patients with MDS who typically experience failure of HMAs by that time point.5 While considered the only potentially curative therapy, only 1% of older patients with CMML underwent alloSCT.
Our results confirm that HMAs are commonly used in the US for the treatment of CMML, which may be related to the lack of other disease-modifying therapies.1,9 Only 1% of older CMML patients in our study underwent the alloSCT, and the vast majority of patients were treated with palliative approaches. We observed that OS of older CMML patients as a group has improved after approval of HMAs (median, 14 vs. 10 months, P=·013). The factors underlying this improvement could be related to better supportive measures and/or wider use of active therapies. Our results suggest that the use of HMAs is associated with a reduced risk of death in CMML patients diagnosed 2007–2011 who received HMAs compared to PS-matched patients diagnosed 2001–2003. It does not seem that this significant effect can be explained by improvement in supportive care over the period of study.
Clinical trial data associated with HMA use in CMML are sparse, but our study results are generally consistent with reported outcomes regarding survival impact of those agents.4–7 In each of the two large randomized phase 3 trials of azacitidine in MDS, less than 10% of patients had CMML.4,5 While azacitidine use in the AZA-001 trial was associated with a median OS benefit of 9·5 months compared to conventional care (24·5 vs 15 months, HR=0·58, P<·0001), such small numbers precluded meaningful OS comparison for the CMML patients.5 Similarly, the two large randomized phase 3 studies of decitabine in MDS only included few CMML patients.6,7 Neither study showed an improvement in OS in the overall cohorts compared to supportive care, and it is not clear if this is related to biologic differences between azacitidine and decitabine or due to differences in study designs. Small retrospective and prospective single-arm studies using decitabine and azacitidine in CMML were also reported but improvements in OS could not be ascertained.20–23 Finally, while we observed no difference in OS between azacitidine- and decitabine-treated patients, it should be noted this was an unadjusted comparison which did not account for potential differences in the patient- or disease-related characteristics which might have impacted the OS. Small numbers limited the conduction of an adjusted comparison by the type of HMA used.
Like any retrospective analysis, our study has limitations. We cannot observe for some disease-specific characteristics such as blast percentage, blood counts and cytogenetic/molecular abnormalities in this dataset, all of which are factors that likely affect both the survival of patients and the decision by the physician to use HMAs. Nonetheless, we were able to assess the impact of other important prognostic predictors such as age, RBC and platelet transfusion needs (which are a proxy for severe anemia and thrombocytopenia), and the disability status index17,18 and Elixhauser comorbidity index16 which are proxies for the performance status and burden of comorbidities, respectively. In addition, our study design takes advantage of a natural experiment, HMA market entry in 2004 and 2006, which tends to limit bias associated with unobserved confounders. Furthermore, even if the disease-specific characteristics we could not observe for were not randomly distributed in the matched samples, it is conceivable that patients who received HMAs had higher risk disease (i.e. higher blast percentage, worse cytogenetics) which would underestimate, rather than inflate, the impact of HMAs use on survival.
In summary, to our knowledge, this is the largest reported cohort of patients with CMML treated with HMAs. Despite lack of randomized clinical trial evidence for their use, these drugs are often used in the US. In this observational cohort study that used market entry of HMAs as a natural experiment to assess impact of HMAs on survival, our findings support the use of HMAs in CMML. In the absence of randomized prospective data, our findings provide the best evidence to date for a survival advantage with the use of HMA in CMML.
Supplementary Material
Acknowledgements:
The results of this study were presented in part in an oral presentation at the American Society of Hematology (ASH) Annual Meeting in San Diego, CA, December 2016. This research was partly funded by the Dennis Cooper Hematology Young Investigator Award (AZ) and a P30 CA016359 from the National Cancer Institute (XM). The collection of the California cancer incidence data used in this study was supported by the California Department of Public Health as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute (NCI)’s Surveillance, Epidemiology and End Results (SEER) Program under contract N01-PC-35136 awarded to the Northern California Cancer Center, contract N01-PC-35139 awarded to the University of Southern California, and contract N02-PC-15105 awarded to the Public Health Institute; and the Centers for Disease Control and Prevention’s National Program of Cancer Registries, under agreement #U55/CCR921930–02 awarded to the Public Health Institute. The ideas and opinions expressed herein are those of the author(s) and endorsement by the State of California, Department of Public Health, the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors is not intended nor should be inferred. The authors acknowledge the efforts of the Applied Research Program, National Cancer Institute; the Office of Research, Development and Information, Centers for Medicare and Medicaid Services; Information Management Services, Inc.; and the SEER Program tumor registries in the creation of the SEER-Medicare database. The interpretation and reporting of the SEER-Medicare data are the sole responsibility of the authors.
Disclosures: Drs. Zeidan, Ma, Huntington, and Gore served as consultants for Celgene. Drs. Gore and Davidoff received research funding from Celgene. These sources of support were not used for any portion of the current study. Dr. Podoltsev consulted for Incyte. None of the other coauthors have conflicts to report.
REFERENCES
- 1.Nazha A, Prebet T, Gore S, Zeidan AM. Chronic myelomoncytic leukemia: Are we finally solving the identity crisis? Blood Rev. 2016;30(5):381–8. [DOI] [PubMed] [Google Scholar]
- 2.Vardiman JW, Thiele J, Arber DA, Brunning RD, Borowitz MJ, Porwit A, et al. The 2008 revision of the World Health Organization (WHO) classification of myeloid neoplasms and acute leukemia: rationale and important changes. Blood. 2009;114(5):937–51. [DOI] [PubMed] [Google Scholar]
- 3.Padron E, Steensma DP. Cutting the cord from myelodysplastic syndromes: chronic myelomonocytic leukemia-specific biology and management strategies. Curr Opin Hematol. 2015;22(2):163–70. [DOI] [PubMed] [Google Scholar]
- 4.Silverman LR, Demakos EP, Peterson BL, Kornblith AB, Holland JC, Odchimar-Reissig R, et al. Randomized controlled trial of azacitidine in patients with the myelodysplastic syndrome: a study of the cancer and leukemia group B. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2002;20(10):2429–40. [DOI] [PubMed] [Google Scholar]
- 5.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(3):223–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kantarjian H, Oki Y, Garcia-Manero G, Huang X, O’Brien S, Cortes J, et al. Results of a randomized study of 3 schedules of low-dose decitabine in higher-risk myelodysplastic syndrome and chronic myelomonocytic leukemia. Blood. 2007;109(1):52–7. [DOI] [PubMed] [Google Scholar]
- 7.Lubbert M, Suciu S, Baila L, Ruter BH, Platzbecker U, Giagounidis A, et al. Low-dose decitabine versus best supportive care in elderly patients with intermediate- or high-risk myelodysplastic syndrome (MDS) ineligible for intensive chemotherapy: final results of the randomized phase III study of the European Organisation for Research and Treatment of Cancer Leukemia Group and the German MDS Study Group. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2011;29(15):1987–96. [DOI] [PubMed] [Google Scholar]
- 8.Fletcher SA, Cronin AM, Zeidan AM, Odejide OO, Gore SD, Davidoff AJ, et al. Intensity of end-of-life care for patients with myelodysplastic syndromes: Findings from a large national database. Cancer. 2016;122(8):1209–15. [DOI] [PubMed] [Google Scholar]
- 9.Zeidan AM, Wang R, Davidoff AJ, Ma S, Zhao Y, Gore SD, et al. Disease-related costs of care and survival among Medicare-enrolled patients with myelodysplastic syndromes. Cancer. 2016;122(10):1598–607. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Wang R, Zeidan AM, Yu JB, Soulos PR, Davidoff AJ, Gore SD, et al. Myelodysplastic Syndromes and Acute Myeloid Leukemia After Radiotherapy for Prostate Cancer: A Population-Based Study. Prostate. 2017;77(5):437–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Zandberg DP, Huang TY, Ke X, Baer MR, Gore SD, Smith SW, et al. Treatment and outcomes for chronic myelomonocytic leukemia compared to myelodysplastic syndromes in older adults. Haematologica. 2013;98(4):584–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zeidan AM, Davidoff AJ, Long JB, Hu X, Wang R, Ma X, et al. Comparative clinical effectiveness of azacitidine versus decitabine in older patients with myelodysplastic syndromes. Br J Haematol. 2016;175(5):829–40. [DOI] [PubMed] [Google Scholar]
- 13.Warren JL, Brown ML, Fay MP, Schussler N, Potosky AL, Riley GF. Costs of treatment for elderly women with early-stage breast cancer in fee-for-service settings. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2002;20(1):307–16. [DOI] [PubMed] [Google Scholar]
- 14.Warren JL, Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Med Care. 2002;40(8 Suppl):IV-3–18. [DOI] [PubMed] [Google Scholar]
- 15.April F, Constance P, Andrew J, Kanagaratnam S, Leslie S, D Max P, et al. International classification of diseases for oncology. 3rd ed. Malta: World Health Organization; 2000. [Google Scholar]
- 16.Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8–27. [DOI] [PubMed] [Google Scholar]
- 17.Davidoff AJ, Zuckerman IH, Pandya N, Hendrick F, Ke X, Hurria A, et al. A novel approach to improve health status measurement in observational claims-based studies of cancer treatment and outcomes. J Geriatr Oncol. 2013;4(2):157–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Davidoff AJ, Gardner LD, Zuckerman IH, Hendrick F, Ke X, Edelman MJ. Validation of disability status, a claims-based measure of functional status for cancer treatment and outcomes studies. Med Care. 2014;52(6):500–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Padron E, Garcia-Manero G, Patnaik MM, Itzykson R, Lasho T, Nazha A, et al. An international data set for CMML validates prognostic scoring systems and demonstrates a need for novel prognostication strategies. Blood Cancer J. 2015;5:e333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Aribi A, Borthakur G, Ravandi F, Shan J, Davisson J, Cortes J, et al. Activity of decitabine, a hypomethylating agent, in chronic myelomonocytic leukemia. Cancer. 2007;109(4):713–7. [DOI] [PubMed] [Google Scholar]
- 21.Wijermans PW, Ruter B, Baer MR, Slack JL, Saba HI, Lubbert M. Efficacy of decitabine in the treatment of patients with chronic myelomonocytic leukemia (CMML). Leuk Res. 2008;32(4):587–91. [DOI] [PubMed] [Google Scholar]
- 22.Braun T, Itzykson R, Renneville A, de Renzis B, Dreyfus F, Laribi K, et al. Molecular predictors of response to decitabine in advanced chronic myelomonocytic leukemia: a phase 2 trial. Blood. 2011;118(14):3824–31. [DOI] [PubMed] [Google Scholar]
- 23.Costa R, Abdulhaq H, Haq B, Shadduck RK, Latsko J, Zenati M, et al. Activity of azacitidine in chronic myelomonocytic leukemia. Cancer. 2011;117(12):2690–6. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.