Disparities in patient access and disease targets in upfront clinical trials for gynecologic cancers

Abstract

<h3>Objectives:</h3> Clinical trials are critical for guiding evidence-based cancer treatment that maximizes survival in balance with toxicities of therapy, yet only 2-4% of patients participate, more often in the setting of recurrence rather than upfront therapy. Access to trials depends on patient, disease, and treating facility characteristics and is not equitable. Our objective was to assess for disparities in upfront clinical trial participation and associations with overall survival. <h3>Methods:</h3> Retrospective study of patients with cancers of the uterus, cervix, ovary/fallopian tube, vulva, and vagina from the National Cancer Database (NCDB). NCDB samples approximately 70% of new diagnoses in the US and includes information on upfront treatment, including participation in a clinical trial. Patients with unknown trial participation status, or receiving unknown or no treatment were excluded. Disparities in upfront trial participation were assessed through comparison of characteristics using t-tests, chi2 tests, and multivariable logistic regression. Impact on overall survival was assessed with cox proportional hazards models. <h3>Results:</h3> From the NCDB, we identified 877,534 patients diagnosed 2004-16, of which 461 (0.05%) were enrolled in a clinical trial of upfront therapy. Participants in upfront clinical trials were on average younger, had fewer comorbidities, and were less likely to be Hispanic. Participation was focused on ovarian (70.1%) and advanced stage disease (77.9%), with over 60% of those in trials having stage III-IV ovarian cancer. Trial participants had similar rate of surgery (85.5% vs. 86.9%), but more frequently received chemotherapy (87.8% vs 39.4%; p<0.001) and immunotherapy (3.3% vs 0.5%; p<0.001), as opposed to radiation (15.0% vs 25.7%; p<0.001). Trial participation in NCDB institutions has fluctuated year-to-year, but tended towards higher participation rates in more recent years (0.06% 2012-16 vs 0.04% 2004-2011; <i>p</i><0.001). In a multivariable model for upfront trial participation, we find non-significant trends towards lower enrollment among non-White (OR=0.76, <i>p</i>=0.07) and Hispanic (OR=0.61, <i>p</i>=0.07), and significantly reduced odds of participation for patients of age ≥70 (OR=0.49; 95% CI 0.34-0.71; <i>p</i><0.001), from low-education zip codes (OR=0.52; 95% CI 0.36-76; <i>p</i>=0.001), and with public insurance (OR=0.67; 95% CI 0.53-0.85; <i>p</i>=0.001). Patients with ovarian (OR=4.90) and cervical (OR=1.78) cancers were more likely to be in trials than those with uterine cancer, and those with stage III-IV disease had significantly increased odds of participation (OR=6.79; 95% CI 5.0-9.1; <i>p</i><0.001). Participation rates were highest at Northeast facilities vs those in the Southern (OR=0.47; <i>p</i><0.001), Central (OR=0.70; <i>p</i>=0.004), and Western (OR=0.74; <i>p</i>=0.045) regions, and higher in academic facilities (OR=2.81; 95% CI 2.3-3.5; <i>p</i><0.001). After adjustment for patient and disease characteristics, we find no impact of trial participation on overall survival (HR=0.94; 95% CI 0.80-1.11; <i>p</i>=0.47). <h3>Conclusions:</h3> Participation in upfront clinical trials is rare across gynecologic cancers. Upfront trials disproportionately study treatment of advanced stage ovarian cancer with chemo- or immunotherapies. Access is not equitable, with greater participation among younger, better-educated, privately insured patients. Future clinical trial development should consider these disparities in disease targets, and strategies for outreach and recruitment.