Importance of patient selection and testing strategies in the determination of the cost-effectiveness of hereditary cancers: The case of mismatch repair genes

LS Wilson PhD^a, CM Reyes PhD^a, BA Allen MS^b, JP Terdiman MD^b

^aDepartment of Pharmacy, ^bDepartment of Medicine; University of California San Francisco, San Francisco, CA, United States

AIM: Genetic testing for hereditary cancers is available for many types of cancer, however cost-effectiveness analyses of the use of these genetic tests are few. In addition to the disease prevalence and penetrance, major factors in the both the decision to use genetic testing and their cost-effectiveness are both the determination of the population to be tested and methods of testing. The individuals tested can range from population based screening to testing of all patients with cancer and their families to testing smaller defined high risk patients. In addition, the method of testing can affect both the costs and the sensitivity and specificity of testing. We look at the cost-effectiveness of various selection strategies for one type of hereditary cancer (HNPCC), including both different method of population selection and approaches to testing. In addition we review the ability of this model to apply to similar selection and testing strategies for other cancers associated with mismatch repair genes and other selected hereditary cancers for which clinical tests are available. Methods: We use a decision analytic method to evaluate the effectiveness and cost-effectiveness of 4 commonly used selection and testing strategies to detect HNPCC gene carriers. We determine which strategy identifies the most gene carriers as well as which is most cost-effective in identification of gene carriers. We then review other genetically defined hereditary cancers to determine which selection and testing strategies might fit into a model such as this, and how they will affect the cost-effectiveness of that strategy. Results: We found that for HNPCC, which has several well-defined patient selection strategies based on patient history, that the Amsterdam criteria identified the fewest gene carriers but cost the least, while screening all cancer patients cost the most and identified the largest number of gene carriers when using gene sequencing. However a mixed model which includes both identification of high risk groups using gene sequencing and then testing others meeting less stringent history requirements first with microsatelite instability testing (MSI) and then sequencing only those who are MSI-H positive was the most cost-effective. The mixed method detects 59.6 mutation carriers and has an incremental cost-effectiveness of $6,441. We think that similarly, patient selection and testing strategies will effect the cost-effectiveness of other hereditary cancers and that our decision model could be applied directly to other cancers affected by mismatch repair genes. Conclusion: It is important to consider both different patient selection strategies and different methods of testing to determine the most cost-effective approach to the genetic testing of hereditary cancers.

Paper presented at the International Symposium on Predictive Oncology and Intervention Strategies; Nice, France; February 7 - 10, 2004; in oral session 892 (Genetic predisposition - Part II).