The Multi-Institutional Medullary Thyroid Cancer Collaborative Registry: Can a Rare Tumor Registry Accurately Represent the Real-World Patient Population?

Background: Large population-based registries, such as the Surveillance, Epidemiology and End Results (SEER) Registry, help in the study of rare tumors, including medullary thyroid cancer (MTC), but lack data to understand the natural history of the disease. The Medullary Thyroid Cancer Collaborative Registry (MTCCoRe) is an exhaustive multi-institutional collection of demographic, clinical, and pathological data. To determine the extent to which MTCCoRe represents the real-world MTC population, we compared the characteristics of patients enrolled in MTCCoRe with patients enrolled in population-based cancer registries. Methods: Comparison of demographic and clinical characteristics of MTC patients who were enrolled in MTCCoRe, Texas Cancer Registry (TCR), California Cancer Registry (CCR), and SEER between 1995 and 2018. Results: A total of 1416 patients were identified in MTCCoRe, 329 in TCR, 2105 in CCR, and 3820 in SEER. Percentages of patients 20-54 years in MTCCoRe were 58.0%, 50.2% in TCR, 47.2% in CCR, and 44.8% in SEER (p < 0.0001). About half of the patients were female (55.9% in MTCCoRe, 61.4% in TCR, 59% in CCR, and 57.5% in SEER (p = 0.3). Percentages of Hispanic and Black patients differed among cohorts (10.1% and 3.8% for MTCCoRe, 23.7% and 8.2% for TCR, 24.8% and 4.9% in CCR, and 15.9% and 8.2% for SEER, respectively; p < 0.001). MTCCoRe patients presented with more advanced T and N classifications than patients in the other registries (MTCCoRe, 28.6% T3-4 and 49.4% N1; TCR, 12.7% and 32.2%; CCR, 18.6% and 32.4%; and SEER, 24% and 37.8%; p < 0.0001). Prevalence of M1 disease was 10% in MTCCoRe, 11.9% in TCR, 14.1% in CCR, and 9.5% in SEER (p < 0.0001). In the MTCCoRe, 11.4% underwent systemic therapy (compared with 0.3% in TCR and 5.6% in CCR). Conclusions: The clinicodemographic profile of patients with MTC enrolled in a multi-institutional registry differs from those enrolled in population-based databases, with lower proportions of Hispanic and Black patients but additive data on treatment modalities. Moving forward, MTCCoRe and other registry and clinical trial enrollment efforts should intentionally include underrepresented groups via community engagement techniques, patient stakeholder involvement, and inclusion of languages other than English in study materials to yield more generalizable results and conclusions.