Risk Prediction Models for Rare Melanomas
Abstract
The proposed research project, developing risk prediction tools for rare melanomas, including acral, mucosal, and uveal melanomas, aligns with the Fiscal Year 2020 (FY20) Melanoma Research Program (MRP) Focus Areas of (i) prevention of melanoma initiation by defining initiation factors for rare melanomas, and (ii) utilizing bioengineering (computational) approaches to address diagnostics and high-risk markers. As noted as key gaps in the FY20 MRP Challenge Statement, in order to redefine prevention of rare melanomas, our proposal aims to first identify the exposures, initiation events, and genetic risk factors, which are largely unknown for rare melanoma subtypes. Scientific Objective and Rationale: The objective of the proposed project is to identify clinical, environmental, and genetic risk factors for rare melanomas. While personal and family history of melanoma and ultraviolet (UV) radiation are known initiation and risk factors for non-acral cutaneous melanoma, genetic, and environmental factors associated with rare melanoma subtypes are largely unknown. The rationale to start with a population of Veterans is to first develop rare melanoma risk prediction models in large populations of patients diagnosed with melanoma with centralized medical records (The US Department of Veterans Affairs [VA] Central Data Warehouse [CDW]), that also have environmental and lifestyle history, and germline genetic data (Million Veteran Program [MVP]). In preliminary data, we estimate 3,000 patients diagnosed with melanoma annually included in the VA system (CDW). In addition to clinical factors including age, sex, race-ethnicity, hair color, family history, and prior medical history, we will be able to explore environmental factors such as area of work (e.g., indoors/outdoors, flight crew), exposures (e.g., agent orange), and deployment history (base location and duration, including latitude and longitude coordinates). To explore how the newly developed rare melanoma risk models work outside of the military, we will validate and adjust (calibrate) the models in two civilian cohorts. First, we will test the models in the Harvard (Mass General Brigham) Research Patient Data Registry. Next, we will test the models in patient-centered registries for rare melanomas being developed by the Melanoma Research Alliance and collaborators in coordination with a dozen patients and care givers of patients with mucosal, acral, and uveal melanoma. Applicability of the Proposed Research: Our proposed work will benefit patients – and prevention of future patients – with acral, mucosal, and uveal melanomas. Our research in these focus areas enables clinically meaningful progress that can significantly impact understanding of risk for rare melanomas. The rare melanoma risk prediction models developed in this proposal represent the first ever comprehensive attempt to identify and integrate clinical and genetic risk factors for these rare melanoma subtypes, and apply those factors to predictive models. The overarching goal is to be able to identify individuals who are at increased risk for the rare tumor subtypes, so that risk factors can be mitigated and they can be more frequently screened for earlier detection and improved outcomes.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Dec 05, 2021
- Source ID
- W81XWH2110819
Entities
People
- Marc Hurlbert
Organizations
- Melanoma Research Alliance
- United States Army