Outcomes of Older Adults With AML Treated in Community Versus Academic Centers: An Analysis of Alliance Trials

PURPOSE:

Clinical trials are important for managing older patients with AML. We investigated differences in outcomes of older patients with AML on the basis of whether patients participated in intensive chemotherapy trials at community versus academic cancer centers.

METHODS:

We used data from the Alliance for Clinical Trials in Oncology phase III trials that enrolled patients age ≥ 60 years with newly diagnosed AML between 1998 and 2002 in the Cancer and Leukemia Group B (CALGB) 9720 trial and between 2004 and 2006 in the CALGB 10201 trial. Centers funded by the NCI Community Oncology Research Program were identified as community cancer centers; others were designated as academic cancer centers. Logistic regression models and Cox proportional hazards models were used to compare 1-month mortality and overall survival (OS) by center type.

RESULTS:

Seventeen percent of the 1,170 patients were enrolled in clinical trials in community cancer centers. The study results demonstrated comparable rates of grade ≥3 adverse events (97% v 93%), 1-month mortality (19.1% v 16.1%), and OS (43.9% v 35.7% at 1 year) between community versus academic cancer centers, respectively. After adjusting for covariates, 1-month mortality (odds ratio, 1.40; 95% CI, 0.92 to 2.12; P = .11) and OS (hazard ratio, 1.04; 95% CI, 0.88 to 1.22; P = .67) were not statistically different among patients treated in community versus academic cancer centers.

CONCLUSION:

An older patient population, who have complex health care needs, can be successfully treated on intensive chemotherapy trials in select community cancer centers with outcomes comparable with that achieved at academic cancer centers.

INTRODUCTION

In contrast to several years ago, today about half of patients in the United States with AML are diagnosed and treated in community cancer centers.¹ Travel to an academic cancer center can be burdensome for many patients. Hence, the proportion of patients with AML treated at community cancer centers may further increase with the increasing use of low-intensity chemotherapy regimens. This issue of where AML treatment is rendered is particularly important for older patients with this disease.² Approximately 60% of new cases of AML and more than 70% of deaths of patients who had been diagnosed with AML are observed in patients older than 60 years.³ Many older patients with AML have significant comorbidities and poor functional status,^4,5 both of which can increase the risk of chemotherapy-related toxicities and early mortality. Older adults with AML may greatly benefit from clinical trials. The National Comprehensive Cancer Network guidelines on AML strongly recommend to treat older adults with AML on a clinical trial.⁶ Enrollment of older adults in clinical trials may be facilitated greatly if patients can get enrolled in community centers close to home. Such issues raise the question of whether older patients with AML require specialized resources and expertise within presumably high-volume academic cancer centers as opposed to community cancer centers in a clinical trial setting.

To investigate differences in outcomes among older patients with AML on the basis of whether patients were enrolled in clinical trials at a community versus academic cancer center, we performed an analysis of two Alliance AML clinical trials. These trials provide data on the specific prescribed chemotherapy regimens and their accompanying toxicities, which are not captured in large-scale registries in the United States Because these data were also acquired prospectively, they provided a higher level of accuracy with respect to longitudinal outcomes. Hence, these data provide an invaluable opportunity for determining whether clinical outcomes differ on the basis of where cancer treatment is rendered.

METHODS

Overview

The current study is a patient-level analysis of completed, prospectively conducted clinical trial data from the Alliance for Clinical Trials in Oncology (Alliance). Cancer centers that had been funded by the NCI Community Oncology Research Program (NCORP) or formerly Community Clinical Oncology Program (CCOP) were identified as community cancer centers. Minority underserved CCOP or NCORP centers or research bases affiliated with academic centers were categorized as academic centers. These categorizations are consistent with the NCI's definitions and list of NCORP/CCOP and research bases.^7,8 The study team reviewed all participating sites and categorized any other non-NCORP/CCOP centers as community centers, if they were not affiliated with a NCI-designated cancer center or an academic university.

Data Sources

This study included data from two large Cancer and Leukemia Group B (CALGB) phase III trials, CALGB 9720^9,10 and 10201.¹¹ CALGB is now part of the Alliance. We looked at data for all AML trials in the past 25 years. Trials were selected on the basis of the following criteria: the trial was designed specifically for older adults with AML; primary analysis conclusions of the trial were published; and each trial included more than 10 patients who were treated at a community cancer center. We identified the following four trials that enrolled older patients with AML for the analysis: CALGB 9720, 10201, 11001, and 11002. Only trials CALGB 9720 and 10201 had more than 10 patients treated in community sites. From within these trials, the current study focused on patients older than 60 years when diagnosed with AML. All patients had provided written consent when enrolled on the original trial. The original trials were approved by local institutional review boards.

Briefly, CALGB 9720 evaluated valspodar (PSC-833) during induction and consolidation therapy with cytarabine, daunorubicin, and etoposide (ADE ± PSC-833). Patients were enrolled on the study before starting induction therapy. Patients who did not achieve a complete remission were not randomly assigned (non–randomly assigned cohort), but those who did achieve a complete remission were randomly assigned to no further therapy or a subcutaneous interleukin-2 regimen as maintenance therapy. This trial did not demonstrate a survival benefit with interleukin-2 maintenance. CALGB 10201 randomly assigned patients to cytarabine and daunorubicin followed by cytarabine consolidation with or without the Bcl-2 antisense oligodeoxynucleotide oblimersen (G3139). This second trial did not demonstrate any differences in toxicity or survival between the two arms.

Study End Points and Variables Analyzed

The current study compared clinical outcomes including rates of allogeneic stem-cell transplantation, grade ≥3 toxicities, 1-month mortality, and overall survival (OS) between patients with AML treated at community versus academic cancer centers. The Common Terminology Criteria for Adverse Events version 2.0 in CALGB 9720 and version 3.0 in CALGB 10201 were used to assess adverse events. One-month mortality was defined by means of vital status 1 month after registration to the trial. OS time was defined as the time from trial registration until death or last follow-up. Survival of patients alive at last follow-up was censored at that time.

Demographics (age, sex, race, ethnicity, and insurance type), Eastern Cooperative Oncology Group performance status, AML risk categories (favorable, intermediate, adverse, unclassified, or missing), cancer therapy rendered, and treatment location (community cancer center or academic cancer center) were all incorporated into the analyses.

Statistical Analysis

Demographics and baseline characteristics were summarized by center type (community v academic) using median (range) for continuous variables and frequency (percentage) for categorical variables and were compared between center types using the Wilcoxon rank sum test and chi-square or Fisher's exact test, as appropriate, respectively. The proportion of patients with grade 3 or higher adverse events and 1-month mortality was estimated by center type and was compared between center types using a chi-squared test or Fisher's exact test. Logistic regression models were used to compare 1-month mortality by center with center type (community v academic) as the main effect and adjusting for age group (60-64, 65-69, ≥ 70 years). Other demographic and baseline characteristics included in the logistic models were associated with the outcomes using a P value of .05 as the threshold. The interaction between center type and age was also evaluated in the logistic regression model.

For OS, the distribution of survival probability was estimated using the Kaplan-Meier estimator. OS probabilities at 1, 3, and 5 years along with 95% CIs were estimated by center type and age group. Cox proportional hazards models^12,13 were used to compare OS using center type as the main effect, adjusted for age group. A similar selection procedure as described above was used to select the final models. The interaction between center type and age was also evaluated in the Cox proportional hazards models. The proportional hazards assumption was checked, and there was indication that the effect of performance status on OS was not proportional. Hence, the final Cox model was stratified by performance status. All tests were performed at a two-sided .05 significance level. Analyses were conducted by the Alliance Statistics and Data Management Center using SAS version 9.4 (SAS Institute Inc, Cary, NC).

RESULTS

Overview

Two of four CALGB trials in older adults with AML had more than 10 patients treated in community centers. In total, 1,175 patients were registered to CALGB 9720 and CALGB 10201. Five patients were excluded from this study for the following reasons: lack of receipt of the study drugs (n = 2), death before or on the day of registration (n = 2), and age younger than 60 years (n = 1). Thus, 1,170 patients (665 from CALGB 9720 and 505 from CALGB 10201) were included in the current study. Community cancer centers enrolled 12% (78 of 665) of all the trial participants in the CALGB 9720 trial and 25% (126 of 505) of the participants in the CALGB 10201 trial, which opened in more recent years than the CALGB 9720.

Baseline Patient Characteristics

Patients who received treatment at community versus academic cancer centers differed on the basis of insurance types (P = .04) with greater proportion of governmental insurance (75% v 60%) and lower proportion of private insurance (23% v 31%) among those treated at community cancer centers. Patient characteristics between community versus academic cancer centers did not differ on the basis of age, sex, race, ethnicity, baseline performance status, or AML subtype. Genetic data were missing for many patients treated in community and academic cancer centers but even more so for those treated in the community cancer centers (Appendix Table A1, online only). Fifty-percent (586 of 1,170) of all trial participants were 70 years or older at the time of enrollment on the trials. Age distribution limited to adults ≥ 70 years also did not differ between community (median, 74 years; range, 70-85 years) versus academic (median, 75 years; range, 70-90 years) cancer centers.

Adverse Events

Grade ≥3 hematologic and nonhematologic adverse events were noted in 97% and 93% of patients treated at community versus academic cancer centers, respectively. No significant differences were noted in the rates of grade ≥3 adverse events among age-based groups between community versus academic cancer centers (Table 1). The rates of grade ≥3 adverse events were consistently high across all subgroups (Appendix Table A2, online only) and developed early after induction (data not shown).

TABLE 1.

Toxicity Rates (grade ≥3) by Cancer Center Type Within Age Group

Survival

The 1-month mortality rate was 19.1% and 16.1% for patients treated at community versus academic cancer centers, respectively. The 1-month mortality rate of community-treated patients was 10.2% (adults 60-64), 13.3% (adults 65-69), and 27.4% (adults ≥ 70 years), whereas in academic center-treated patients, these rates were 11.5% (adults 60-64), 13.7% (adults 65-69), and 19.6% (adults ≥ 70 years) (Appendix Table A3, online only). After adjusting for age, performance status, and treatment regimen, 1-month mortality did not differ between patients treated in community versus academic cancer centers (odds ratio, 1.40; 95% CI, 0.92 to 2.12; P = .11; Table 2). The interaction between age and center type did not reach statistical significance.

TABLE 2.

Multivariable Logistic Regression Model of 1-Month Mortality (OR > 1 means higher odds of mortality)

One-year OS was 43.9% (95% CI, 37.6 to 51.3) and 35.7% (95% CI, 32.8 to 38.9) for patients treated at community versus academic cancer centers (Fig 1). One-year OS rates among patients treated in a community cancer center were 46.9% (adults 60-64), 49.4% (adults 65-69), and 38.9% (adults ≥ 70 years), whereas these rates were 42.7% (adults 60-64), 35.1% (adults 65-69), and 32.7% (adults ≥ 70 years) for those treated in an academic cancer center (Appendix Table A4, online only; Appendix Figs A1 and A2, online only). OS did not differ between community versus academic cancer centers. The results from the multivariable Cox proportional hazard model stratified by performance status and adjusted for age, AML type, and treatment regimen did not demonstrate any difference in OS between community versus academic cancer centers (hazard ratio, 1.04; 95% CI, 0.88 to 1.22; P = .67; Table 3; Appendix Fig A3, online only). The interaction between age and center type was not statistically significant.

FIG 1.

Kaplan-Meier OS curves of older adults with AML by cancer center type. HR, hazard ratio; KM Est, Kaplan Meier estimates; OS, overall survival.

TABLE 3.

Multivariable Cox Model for Overall Mortality (stratified by performance status)

Transplant Rates

The rates of transplant appeared to decline with increasing age (15%-18% for adults 60-64 years to 2% for those ≥ 70 years) but appeared comparable between community versus academic cancer centers (Appendix Table A5, online only).

DISCUSSION

This study analyzed data from more than 1,000 patients who participated in the two CALGB/Alliance clinical trials designed specifically for older adults. Most patients were treated at academic cancer centers. Community centers enrolled more than 10 patients in two of the four CALGB trials. This is the first study to demonstrate that the rates of grade ≥3 adverse events, 1-month mortality, or OS can be comparable regardless of whether patients enrolled in clinical trials of intensive chemotherapy at select community cancer centers versus academic cancer centers. The study results demonstrate that an older patient population, who have complex health care needs, can be successfully treated on clinical trials in the select NCORP (formerly CCOP) community cancer centers with outcomes comparable with that achieved at academic cancer centers. Thus, the study results confirm that the NCORP program is a valuable part of clinical care system in the United States and that high-quality intensive treatment trials in older adults that span both academic and community cancer centers can be successfully performed.

The results from the current study are at odds with earlier registry studies. These prior studies suggest that adults with AML who receive initial care at larger volume centers or academic cancer centers are more likely to receive chemotherapy¹⁴ and allogeneic stem-cell transplant,¹⁵ have lower early mortality rates, and experience longer OS.^1,16,17 The reasons for differences between prior studies and this study have several possible explanations. The NCORP is a competitive funding mechanism that selects for community-based cancer centers with high patient volumes, ability to enroll on an intensive chemotherapy trial of AML, and those with an investigational drug pharmacy for study drug and access to a hospital where patients enrolled on clinical trials can be monitored. The community-based medical centers that took part in CALGB 9720 and CALGB 10201 are likely not representative of all other community-based cancer centers.

Older patients who had enrolled in the trials examined here were all presumably fit enough to receive intensive chemotherapy. Prospectively gathered data on performance status, which are not captured in large-scale registries in the United States, enabled us to adjust for this variable in the OS analyses. Furthermore, registry studies provide information on less homogeneous patients, whereas patients treated within the context of a clinical trial must meet specific eligibility criteria to receive protocol-specified therapy. The receipt of protocol-specified therapy can standardize cancer treatment, with the possibility of reducing early mortality as demonstrated in acute promyelocytic leukemia.¹⁸ Overall, however, the results reported here provide reassurance that older adults who are treated in clinical trials can derive similar outcomes regardless of whether the treatment is rendered at community or academic cancer centers.

The current study has both strengths and limitations. The limitations include lack of complete data on genetic and molecular features of AML, which can have prognostic significance. This lack of data was observed in both settings but notably among those treated in the community cancer centers; this lack of genetic and molecular data is perhaps a major limitation, as such data are increasingly relied on to determine optimal cancer therapy. Information on comorbidity burden was not available for older CALGB trials. Most patients were treated at academic cancer centers. Details regarding patient volumes, clinicians' experience in treating AML, and expertise of personnel and resources available could not be ascertained for each site; however, participating community cancer centers needed to have resources to conduct clinical trials and many were part of the NCORP program. The results may not apply to patients receiving standard-of-care chemotherapy outside of clinical trials at community cancer centers because many community cancer centers may not have resources available at NCORP community cancer centers, clinical trials often provide robust guidance on managing toxicities and treatment modifications, and patients who are treated outside of clinical trials may have lower performance status and higher comorbidity burden. Such patients may have more complex needs and may benefit from multidisciplinary leukemia team including infectious disease specialists at academic centers. Because of the focus on older patients, these findings cannot be generalized to younger patients. The results from these trials that used intensive chemotherapy may not be generalizable to older adults with AML treated in the outpatient settings using a combination of hypomethylating agent and newer lower-intensity therapies since resources needed to manage outpatient treatments may differ from resources necessary for inpatient treatments, and outpatient clinics may not have as many resources and staff as in hospital settings. Future studies should examine the outcomes of older adults participating in lower-intensity outpatient treatments. The major strengths include a large-scale analysis of prospectively acquired clinical trial data from multiple centers across the United States that evaluated the outcomes of older adults with AML treated using a standardized research protocol. The current study demonstrates that older patients with AML who choose to enroll in clinical trials in a community cancer center as opposed to an academic cancer center are likely to receive a high standard of medical care with comparable clinical outcomes. These results open up venues to use select NCORP community cancer centers in other intensive chemotherapy trials, thus bringing clinical trials closer to home for older adults with AML.

APPENDIX

TABLE A1.

Patient Characteristics by Cancer Center Type

TABLE A2.

Demographics by Grade ≥3 Toxicities

TABLE A3.

One-Month Mortalities by Cancer Center Type Within Each Age Group

TABLE A4.

Overall Survival Estimates With 95% CIs by Patient Age and Center Type

TABLE A5.

Allogeneic Stem-Cell Transplantation Incidence by Patient Age and Center Type for CALGB Trial 10201

FIG A1.

Kaplan-Meier OS curves by age group. HR, hazard ratio; KM Est, Kaplan Meier estimates; OS, overall survival.

FIG A2.

Kaplan-Meier OS curves by age group and center type. HR, hazard ratio; OS, overall survival.

FIG A3.

Kaplan-Meier OS curves of performance status. HR, hazard ratio; KM Est, Kaplan Meier estimates; OS, overall survival.

AUTHOR CONTRIBUTIONS

Conception and design: Vijaya Raj Bhatt, Angela M. Ulrich, Jennifer Le-Rademacher, Aminah Jatoi

Collection and assembly of data: Vijaya Raj Bhatt, Angela M. Ulrich, Jennifer Le-Rademacher, Aminah Jatoi

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Outcomes of Older Adults With AML Treated in Community Versus Academic Centers: An Analysis of Alliance Trials

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/op/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).