Electronic Thesis and Dissertation Repository

Degree

Doctor of Philosophy

Program

Epidemiology and Biostatistics

Supervisor

Dr. Gregory S. Zaric

Abstract

Histopathological analysis of tumor is currently the main tool used to guide cancer management. Gene expression profiling may provide additional valuable information for both classification and prognostication of individual tumors. A number of gene expression profiling assays have been developed recently to inform therapy decisions in women with early stage breast cancer and help identify the primary tumor site in patients with metastatic cancer of unknown primary. The impact of these assays on health and economic outcomes, if introduced into general practice, has not been determined. I aimed to conduct an economic evaluation of regulatory-approved gene expression profiling assays for breast cancer and cancer of unknown primary for the purpose of determining whether these technologies represent value for money from the perspective of the Canadian health care system. I developed decision-analytic models to project the lifetime clinical and economic consequences of early stage breast cancer and metastatic cancer of unknown primary. I used Manitoba Cancer Registry and Manitoba administrative health databases to model current “real-world” Canadian clinical practices. I applied available data about gene expression profiling assays from secondary sources on these models to predict the impact of these assays on current clinical and economic outcomes. In the base case, gene expression profiling assays in early stage breast cancer and cancer of unknown primary resulted in incremental cost effectiveness ratios of less than $100,000 per quality-adjusted life-year gained. These results were most sensitive to the uncertainty associated with the accuracy of the assay, patient-physician response to gene expression profiling information and patient survival. The potential application of these gene expression profiling assays in clinical oncology appears to be cost-effective in the Canadian healthcare system. Field evaluation of these assays to establish their impact on cancer management and patient survival may have a large societal impact and should be initiated in Canada to ensure their clinical utility and cost-effectiveness. The use of Canadian provincial administrative population data in decision modeling is useful to quantify uncertainty about gene expression profiling assays and guide the use of novel funding models such as conditional funding alongside a field evaluation.


Share

COinS