Identification of genomic alterations susceptible to drug therapy in some solid tumors has led to increased interest in next-generation sequencing (NGS) testing to identify patients for biomarker-selected clinical trials. In an article published in JAMA Oncology earlier this year, Dr Erin F. Cobain and colleagues identified a high rate of pathogenic germline variants and a subset of patients who derived substantial benefit from sequencing information. The authors recommend directed germline testing for inherited cancer predisposition in all patients with advanced cancer.
In response to the publication by Cobain et al, Dr Catherine Dunn and colleagues suggested this may not be a universally beneficial approach for all patients. The authors highlighted the cost impact of NGS testing and suggested that lower-cost alternatives (eg, using immunohistochemistry to identify deficient mismatch repair as a marker for immune checkpoint inhibitor response) be considered. Tumor subtype is another factor for consideration, as it may alter the yield of molecular testing and may also be of little to no benefit for patients too unwell to receive any targeted therapy. Lastly, even when targeted therapies are available, they are often prohibitively expensive.
Dr Cobain and Dr A.M. Chimaiyan responded to the letter by agreeing that broad NGS testing may not be beneficial for all patients . . . yet. They support the use of lower-cost, limited gene panels or immunohistochemical assays for some patients. However, they also pointed out that the cost of NGS testing is far less than the cumulative cost of other diagnostic assays, such as computed tomography scans. Given the significant clinical implications of identifying pathogenic germline variants, they continue to advocate for germline testing in all patients with advanced cancer, but agree that more-limited gene panels than those used in their study may be more appropriate for many patients.
Dr Cobain and colleagues suggest that the value of NGS testing will increase over time. Researchers must continue to seek and identify effective biomarkers to guide therapeutic management of patients with advanced cancer. As suggested by Dr Dunn and colleagues, an “intent-to-test” analysis would be an appropriate objective of studies of biomarker-driven therapy, similar to an intent-to-treat analysis in pharmaceutical trials.
With regard to a personalized oncology approach, Dr Dunn and colleagues recommend that routine clinical practice be guided by prospective, randomized clinical trials that demonstrate clear patient and health system value. In their letter, they give the example of one trial that found the use of molecularly targeted agents outside their approved indications failed to improve progression-free survival compared with physician’s choice of treatment. Clinical decisions must continue to be individualized, as universal approaches to testing and treatment have yet to be identified.