Predictors of Enrollment in Lung Cancer Clinical Trials. BACKGROUND. Clinical trials may offer patients innovative therapeutic options with

Transcription

1 420 Predictors of Enrollment in Lung Cancer Clinical Trials Wei Du, Ph.D. 1 Shirish M. Gadgeel, M.D. 2 Michael S. Simon, M.D., M.P.H. 2 1 Department of Pediatrics, Wayne State University, Detroit, Michigan. 2 Karmanos Cancer Institute at Wayne State University School of Medicine, Division of Hematology and Oncology, Detroit, Michigan. BACKGROUND. Clinical trials may offer patients innovative therapeutic options with potentially better outcomes, which are particularly relevant for patients afflicted with lung carcinoma, because current therapies provide only modest survival benefits. Only approximately 5% of patients with newly diagnosed cancer participate in clinical trials nationwide, and African-American (AA) patients are particularly under-represented. METHODS. To determine predictors of clinical trials enrollment, the authors reviewed the medical records of 427 patients with lung carcinoma (175 AA patients and 252 non-aa patients) who were eligible for clinical trials between 1994 and 1998 at the Karmanos Cancer Institute in Detroit, Michigan. Logistic regression analysis was used to assess the association of patient demographic characteristics and clinical trial enrollment. RESULTS. Ninety-one patients (21%) were enrolled onto a lung cancer clinical trial during the period of the current study. Enrollment was associated significantly with race (P 0.001), gender (P 0.048), age (P 0.005), and insurance type (P 0.024). After multivariable adjustment, only race and gender remained significant predictors of enrollment. AA patients were less likely to enroll than non-aa patients (odds ratio [OR], 0.485; 95% confidence interval [95% CI], ), and men were more likely than women to enroll (OR, 1.812; 95% CI, ). CONCLUSIONS. The current results suggest disparities by race and gender in the enrollment of patients onto lung cancer clinical trials and support the need to improve educational and outreach endeavors that would make clinical trials available to a wider range of eligible patients. Cancer 2006;106: American Cancer Society. Supported by American Cancer Institute Grant TURPG PBP. Address for reprints: Wei Du, Ph.D., Clinical Pharmacology and Toxicology, Department of Pediatrics, Wayne State University, Children s Hospital of Michigan, 3901 Beaubien, Room 3N47, Detroit, MI 48201; Fax: (313) ; duw@med. wayne.edu Received June 9, 2005; revision received August 12, 2005; accepted October 18, KEYWORDS: clinical trials, lung carcinoma, enrollment, predictors, disparity. Most of the major clinical advancements in cancer treatment have come from clinical trials, which are essential for improving patient outcomes. However, despite the large number of available studies, only approximately 5% of adult patients with newly diagnosed cancer participate in clinical trials nationwide, 1,2 whereas enrollment rates are considerably higher in the pediatric population. 3,4 In particular, African American (AA) patients in the United States are underrepresented in clinical trials compared with their proportion in the general population. 4 6 Lung carcinoma is the leading cause of cancer death in the United States for both men and women, with a 5-year survival rate of approximately 14%. 7 These poor survival statistics highlight the limitations of current therapies as well as the need for significant improvement in available treatments. Most major advances in cancer treatment have come from appropriately conceived and conducted clinical trials Every public and private organization that has ex American Cancer Society DOI /cncr Published online 12 December 2005 in Wiley InterScience (

2 Predictors of Lung Cancer Trial Enrollment/Du et al. 421 amined this issue has determined that participation in high-quality clinical trials is the best way to find better treatment options for patients with cancer. 11,12 Barriers to clinical trials participation have been attributed to several aspects of the enrollment process, including physician and patient factors, trialspecific factors, and cost issues. 13,14 A few studies have demonstrated that lack of adequate information is a significant patient-level barrier to clinical trials participation It also has been shown that fears based on the awareness of a history of abuses of the rights of human participants in biomedical research, like the Tuskegee Syphilis Study, 18,19 also are significant barriers to participation, particularly in the AA population. To assess enrollment rates and factors that are predictive of enrollment onto lung cancer clinical trials, we evaluated the clinical trials enrollment experience at a large academic medical center in Metropolitan Detroit. MATERIALS AND METHODS Study Population Data for this analysis came from clinical and demographic information collected on men and women with newly diagnosed nonsmall cell lung carcinoma (NSCLC) and small cell lung carcinoma (SCLC) from January 1, 1994 through December 31, 1998 who had received their entire initial course of cancer therapy at the Karmanos Cancer Institute (KCI) and who had at least 1 year of follow-up at the KCI. The KCI is one of the 39 National Cancer Institute (NCI)-designated Comprehensive Cancer Centers in the United States. It is affiliated with Wayne State University and the Detroit Medical Center and is located in Detroit, Michigan. The study population was identified through a search of the Metropolitan Detroit Cancer Surveillance System (MDCSS), which is 1 of 11 sites of the NCI s Surveillance, Epidemiology, and End Results Program. By using the MDCSS data base, we identified 649 individuals who had lung carcinoma diagnosed at the KCI, including 98 patients (15%) who were enrolled on a clinical trial and 551 patients (85%) who were not enrolled on a trial. Informed consent was waived by the Wayne State University Human Investigation Committee because of the finding that our study was a retrospective chart review study and that most of the patients had died. To evaluate the factors associated with clinical trials enrollment among a set of patients who would be eligible for participation in a clinical trial, 222 patients were excluded who were unlikely to be eligible. These included patients who had 1) Stage I disease (there were no available Stage I trials during the study period); 2) 2 comorbid conditions (diabetes, heart disease, or chronic obstructive pulmonary disease [COPD]); 3) liver failure; 4) renal failure; 5) a performance status (PS) 2; 6) age older than 80 years at diagnosis; 7) received no carcinoma-directed treatment, such as chemotherapy, radiation, or surgery; or 8) died within 2 weeks of their diagnosis. After these exclusions, there were 427 patients available, including 91 enrollees (21%) and 336 nonenrollees. There were 27 active lung cancer trials open at KCI during the study period, including 2 Phase I studies, 16 Phase II studies, and 9 Phase III studies. Sixteen of the studies were for patients with NSCLC, and 10 of the studies were for patients with SCLC. Study Measures Patient demographic, clinical, and enrollment data were collected by a medical records review, and additional clinical information was obtained through the MDCSS registry. The study outcome was whether a patient was enrolled onto a lung cancer clinical trial. Patient demographics included gender, age at diagnosis, race (classified as AA or non-aa), insurance coverage, and a Census-derived socioeconomic rank (Census rank). Insurance coverage was classified as commercial (conventional, health maintenance organization, or private provider organization), Medicaid, Medicare only, or Medicare plus any kind of supplemental insurance (e.g., Medigap or Medicaid). Using 1990 Census tract data, Census tracts were ranked as low, medium, or high based on the proportion of high school graduates (age 20 yrs and older) in that tract and the median household income within the Census tract in which the patient resided. 20 Clinical data included American Joint Committee on Cancer stage, histologic type (NSCLC or SCLC), South West Oncology Group PS (range, from 0 to 4), and the presence of five common comorbid conditions seen in this study population (heart disease, diabetes, COPD, renal failure, and liver failure). Statistical Methods This study was a retrospective cohort study. Chisquare and Student t tests were used to compare AA patients and non-aa patients as well as clinical trials enrollees and nonenrollees by demographic and clinical characteristics. Multiple logistic regression analyses were used to predict the simultaneous effect of demographic and clinical characteristics on enrollment in lung cancer clinical trials. RESULTS Patient Demographic and Clinical Characteristics Table 1 summarizes the demographic and clinical characteristics of 427 patients, including 173 AA pa-

3 422 CANCER January 15, 2006 / Volume 106 / Number 2 TABLE 1 Distribution of Demographic and Clinical Characteristics by Race among Lung Cancer Patients Eligible for Clinical Trials Characteristic All patients (n 427) AA (n 173) Non-AA (n 254) P value (2-sided) Male, % Age in yrs, mean/median 62/64 62/65 61/ Census rank (SES), % Low Medium High Insurance, % Commercial Medicaid Medicare Medicare/supplemental Stage/histology, % NSCLC Stage II Stage IIIA Stage IIIB Stage IV SCLC Limited Extensive Performance status 0, % Heart disease, % Diabetes, % COPD, % % with comorbidities AA: African American; NSCLC: nonsmall cell lung carcinoma; SCLC: small cell lung carcinoma; COPD: chronic obstructive pulmonary disease; SES: socioeconomic status. TABLE 2 Univariate Comparison of Enrollees versus Nonenrollees Characteristic All patients (n 427) Enrollees (n 91) Nonenrollees (n 336) P value (2-sided) African American, % Male, % Age in yrs, mean/median 62/64 60/60 62/ Age 70 yrs, % Census rank (SES), % Low Medium High Insurance, % Commercial Medicaid Medicare Medicare/supplemental Stage/histology, % NSCLC Stage II Stage IIIA Stage IIIB Stage IV SCLC Limited Extensive Performance status 0, % Heart disease, % Diabetes, % COPD, % % comorbidities NSCLC: nonsmall cell lung carcinoma; SCLC: small cell lung carcinoma; COPD: chronic obstructive pulmonary disease; SES: socioeconomic status. tients (40.5%) and 254 non-aa patients (59.5%), who received treatment for their newly diagnosed lung carcinoma at the KCI between 1994 and AA patients were more likely than non-aa patients to reside in a Census area ranked as low (83% of AA patients vs. 23% of non-aa patients; P 0.001), were less likely to have commercial insurance coverage (27% of AA patients vs. 49% of non-aa patients; P 0.001), were less likely to have a PS of 0 (27% of AA patients vs. 35% of non-aa patients; P 0.054), were less likely to have a diagnosis of heart disease (13% of AA patients vs. 21% of non-aa patients; P 0.030), and were more likely to have a diagnosis of diabetes (18% of AA patients vs. 9% of non-aa patients; P 0.004). AA patients also were more likely to have more advanced stage disease at diagnosis (32% of AA patients had Stage IV NSCLC vs. 28% of non-aa patients) and to have more comorbid conditions (38% of AA patients had comorbid conditions vs. 27% of non-aa patients), although these differences were not statistically significant. There were no significant differences by race among patients according to gender (59% AA males vs. 60% non-aa males), mean age at diagnosis (62 yrs vs. 61 yrs, respectively), or past history of COPD (16% vs. 16%, respectively). Table 2 summarizes the demographic and clinical characteristics of 91 clinical trial enrollees (21%) compared with 336 nonenrollees (79%). Nonenrollees were more likely than enrollees to be AA (45% of nonenrollees were AA vs. 25% of enrollees; P 0.001), female (43% of nonenrollees were female vs. 32% of enrollees; P 0.058), older at the time of diagnosis (24% of nonenrollees were age 70 yrs or older at diagnosis vs. 10% of enrollees; P 0.003), and to live in an area categorized as low Census rank (52% of nonenrollees vs. 37% of enrollees; P 0.028). Nonenrollees were less likely than enrollees to have commercial insurance coverage (55% of enrollees vs. 37% of nonenrollees; P 0.002) and were more likely to have Medicare, Medicaid, and Medicare plus supplemental insurance. There were no significant differences between nonenrollees and enrollees in the variables disease stage/ histology, PS, or the type or number of comorbid conditions. Table 3 shows the results of a multivariable logis-

4 Predictors of Lung Cancer Trial Enrollment/Du et al. 423 TABLE 3 Multivariable Logistic Regression Model of Predictors of Enrollment in lung Cancer Trials Predictor tic regression analysis using clinical trials enrollment (yes vs. no) as the dependent variable and race, gender, age, Census rank, insurance type, stage/histology, PS, and comorbidities as the predictors. The only significant predictors of enrollment were race and gender. The odds of enrollment for AA patients versus non-aa patients were (P 0.039; 95% confidence interval [95% CI], ), and the odds of enrollment for male patients versus female patients were (P 0.038; 95% CI, ). Age, socioeconomic status (SES), and insurance coverage were no longer significant predictors of enrollment onto a lung cancer clinical trial. DISCUSSION Cancer clinical trials bring the results of research on new therapeutic agents from the laboratory bench to the clinical setting and have the potential to help clinicians formulate better ways in which to both treat and prevent cancer. Clinical trials offer cancer patients access to novel and innovative therapeutic options that potentially may be associated with better treatment outcomes. 9,21,22 This is particularly relevant for patients with lung carcinoma, because current therapies provide them only modest benefits in terms of disease-free or overall survival. 8,10 The results of this analysis provide important information regarding the demographics of patients who participate in clinical trials. Although the results of univariate analyses suggested that, in our patient population, race, gender, age, SES (as measured by Census rank), and insurance type all had an impact on clinical trials enrollment, once these factors were evaluated simultaneously in multivariate analysis, only OR P value (2-sided) 95% CI African American vs. White Male vs. female Commercial insurance Census rank: Low vs. medium/high Age 70 yrs Stage IIIA vs. Stage Stage IIIB vs. Stage II Stage IV vs. Stage II Limited vs. Stage II Extensive vs. Stage II Performance status: 1 vs Any comorbidities vs. none OR: odds ratio; 95% CI: 95% confidence interval. race and gender were significant predictors. This is not surprising, because there is a high degree of collinearity (correlation between covariates) between race, SES, and insurance type. Other studies also reported differences in clinical trials accrual based on race and gender. In a prior analysis of breast cancer clinical trial accrual physicians at our institution, physicians were less likely to offer AA women than non-aa women participation in breast cancer clinical trials; however, once an offer was made, there were no racial differences in enrollment. 23 In a recent study of participation in nonsurgical cancer trials sponsored by the NCI (including breast, colorectal, lung, and prostate cancer clinical trials), AA women and Hispanic women had lower enrollment rates than Caucasian women, and men were more likely to be enrolled in both colorectal and lung cancer trials. 24 In another study of accrual to American College of Surgeons Oncology Group cancer protocols, AA patients were under-represented in thoracic clinical trials, although this was partially offset by the higher proportion of AA patients who were ineligible because of more advanced disease stage at diagnosis. 5 In an analysis that examined the impact of demographics on accrual in NCI-sponsored cancer treatment trials, it was noted that AA males accrued to clinical trials at lower rates than white patients of the same age, and there was a marked discrepancy in accrual with higher accrual rates recorded for geographic areas with more affluent communities. 4 In an analysis of barriers to clinical trials accrual at Duke University, AA patients were less willing than white patients to participate in cancer clinical trials, with education and income the major factors that affected these differences. Brown et al. 15 found that women across various racial and ethnic groups had little or no knowledge about clinical research and had lower participation rates than men. In addition, minority women were less knowledgeable about clinical research than their Caucasian counterparts. In our univariate data analysis, patients with commercial insurance coverage were 2-fold more likely (P 0.002) to participate in a clinical trial than patients who were insured by Medicare or Medicaid. Although the difference in enrollment cannot be attributed entirely to insurance, this finding is at least somewhat consistent with the finding that clinical trials were not considered a benefit covered by Medicare for the years included in the current study. Because of recent policy changes, Medicare and many commercial insurers now provide coverage for certain clinical trials (Michigan Cancer Consortium Update, February 2002). Subsequent analysis may provide further insight into the impact of this type of policy change. The strengths of the current study include the

5 424 CANCER January 15, 2006 / Volume 106 / Number 2 large proportion of AA patients treated at our cancer center, which enabled us to look effectively at clinical trial enrollment rates across racial groups. Through our review of the medical records, we also had information available on several possible confounding variables. However, our results should be interpreted with some degree of caution, because we studied enrollment only at a single, large, urban academic medical center; therefore, our findings may not generalize to other regions of the country or to smaller community settings. In addition, we did not collect patientdirected factors that may have provided other explanations for willingness or unwillingness to participate in clinical trials. Prominent among barriers to minority participation in clinical research is the experience of the Tuskegee Study, 18 with the results from a recent survey of potentially eligible patients from the Detroit area suggesting that lack of trust in the medical establishment continues to play a predominant role as a barrier to cancer clinical trial participation among AA patients. 19 Other often cited reasons for lack of participation among minority patients include lack of knowledge about clinical trials 15 and the effects of religious belief or spirituality on the willingness to participate. 6 Other studies have shown that higher rates of other comorbid conditions in AA patients versus white patients 23,25 have a large impact on accrual. In 1993, the National Institutes of Health Revitalization Act directed the healthcare community to ensure that women and members of minorities are included in clinical trials. 26 A recent survey by Harris Interactive suggested that the public is inclined toward participation in trials if they are allowed the opportunity to participate. It is important for the medical establishment to improve access to clinical trials from both a scientific standpoint, allowing us to examine the effect of treatments among different racial groups, 27 and from the standpoint of economic justice for all individuals, regardless of ethnic background. 28 This can be done only through the recognition of historic barriers, in-depth development of recruitment strategies before opening a trial, and the commitment of cooperative groups to design studies that will accommodate a broad spectrum of eligible patients. 29 Our group recently completed a study that examined the effectiveness of an educational intervention that was developed to increase clinical trial enrollment in patients with lung cancer through 19-minute NCIdeveloped video. Recruitment strategies like this are inexpensive and easy to administer; however, information on the effectiveness of this intervention is pending. New recruitment strategies that target specific patient subgroups may be helpful in ensuring equal representation of women and minority groups in cancer clinical trials. REFERENCES 1. Harris Interactive. Misconceptions and lack of awareness greatly reduce recruitment for cancer clinical trials. Harris Interactive Healthcare News, issue 3 [serial online]. Available at URL: healthcare.as0. January 2001 [accessed November 2005]. 2. Spiro SG, Gower NH, Evans MT, Facchini FM, Rudd RM. Recruitment of patients with lung cancer into a randomised clinical trial: experience at two centres. On behalf of the Big Lung Trial Steering Committee. Thorax. 2000;55: Tejeda HA, Green SB, Trimble EL, et al. Representation of African-Americans, Hispanics, and whites in National Cancer Institute cancer treatment trials. J Natl Cancer Inst. 1996;88: Sateren WB, Trimble EL, Abrams J, et al. How sociodemographics, presence of oncology specialists, and hospital cancer programs affect accrual to cancer treatment trials. J Clin Oncol. 2002;20: Newman LA, Hurd T, Leitch M, et al. A report on accrual rates for elderly and minority-ethnicity cancer patients to clinical trials of the American College of Surgeons Oncology Group. J Am Coll Surg. 2004;199: Advani AS, Atkeson B, Brown CL, et al. Barriers to the participation of African-American patients with cancer in clinical trials: a pilot study. Cancer. 2003;97: Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, CA Cancer J Clin. 2004;54: [No authors listed.] Chemotherapy in non-small cell lung cancer: a meta-analysis using updated data on individual patients from 52 randomised clinical trials. Non-small Cell Lung Cancer Collaborative Group. BMJ. 1995;311: Green MR. Future randomized trials for non small-cell lung cancer. Clin Lung Cancer. 2000;2(Suppl 1):S29 S Depierre A, Lagrange JL, Theobald S, et al. Summary report of the Standards, Options and Recommendations for the management of patients with non-small-cell lung carcinoma (2000). Br J Cancer. 2003;89(Suppl 1):S35 S Lukens JN. Progress resulting from clinical trials. Solid tumors in childhood cancer. Cancer. 1994;74: Warnecke RB, Johnson TP, Kaluzny AD, Ford LG. The Community Clinical Oncology Program: its effect on clinical practice. Joint Comm J Qual Improv. 1995;21: Gotay CC. Accrual to cancer clinical trials: directions from the research literature. Soc Sci Med. 1991;33: Winn RJ. Obstacles to the accrual of patients to clinical trials in the community setting. Semin Oncol. 1994;21: Brown DR, Fouad MN, Basen-Engquist K, Tortolero-Luna G. Recruitment and retention of minority women in cancer screening, prevention, and treatment trials. Ann Epidemiol. 2000;10:S13 S Pinto HA, McCaskill-Stevens W, Wolfe P, Marcus AC. Physician perspectives on increasing minorities in cancer clinical trials: an Eastern Cooperative Oncology Group (ECOG) initiative. Ann Epidemiol. 2000;10:S78 S Fouad MN, Partridge E, Wynn T, Green BL, Kohler C, Nagy S. Statewide Tuskegee Alliance for Clinical Trials. A community coalition to enhance minority participation in medical research. Cancer. 2001;91: Brawley OW. The study of untreated syphilis in the Negro male. Int J Radiat Oncol Biol Phys. 1998;40:5 8.

6 Predictors of Lung Cancer Trial Enrollment/Du et al Shavers VL, Lynch CF, Burmeister LF. Knowledge of the Tuskegee study and its impact on the willingness to participate in medical research studies. J Natl Med Assoc. 2005;92: U.S. Bureau of the Census. Census of population and housing (1990) summary tape file 3 technical documentation. Washington, DC: U.S. Census Bureau, Stiller CA. Centralised treatment, entry to trials and survival. Br J Cancer. 1994;70: Watanabe H, Saijo N. [Evidence based chemotherapy for lung cancer]. Gan To Kagaku Ryoho. 2000;27: Simon MS, Du W, Flaherty L, et al. Factors associated with breast cancer clinical trials participation and enrollment at a large academic medical center. J Clin Oncol. 2004;22: Murthy VH, Krumholz HM, Gross CP. Participation in cancer clinical trials: race-, sex-, and age-based disparities. JAMA. 2004;291: Adams-Campbell LL, Ahaghotu C, Gaskins M, et al. Enrollment of African Americans onto clinical treatment trials: study design barriers. J Clin Oncol. 2004;22: Bennett JC. Inclusion of women in clinical trials policies for population subgroups. N Engl J Med. 1993;329: Miskimen T, Marin H, Escobar J. Psychopharmacological research ethics: special issues affecting US ethnic minorities. Psychopharmacology (Berlin). 2003;171: Brawley OW. The study of accrual to clinical trials: can we learn from studying who enters our studies? J Clin Oncol. 2004;22: Comis RL, Miller JD, Aldige CR, Krebs L, Stoval E. Public attitudes toward participation in cancer clinical trials. J Clin Oncol. 2003;21: