I read with interest and concern the Perspective piece, “Cancer Genes, Promiscuity, and the National Debt,” which appeared on page 1 of the February 1st issue of The ASCO Post. The underlying premise of this wide-ranging and provocative article would appear to be that in a rush to implement “genomics” (presumably precision therapy and molecular pathology) in oncology, old mistakes are being repeated in the absence of appropriate standardization and regulation, compounded by a detrimental “need for speed” in bringing this technology to the clinic.
It would of course be difficult to argue this premise, were it in fact the case. Space limitation does not permit a point-by-point rebuttal of this rather lengthy editorial. Thus, my major concern is that there exists a major dearth of knowledge, among even our most highly trained academic oncologists, related to fundamental aspects of genomic medicine, especially as it pertains to next-generation sequencing technology and precision medicine.
Gene Expression vs Mutation
Common, for example, is the conflation of “gene expression” and “gene mutation,” the latter of which is a driver of tumorigenesis for which the related aberrant protein product represents the target, in several cases, of an FDA-approved drug based on level 1 evidence (eg, vemurafenib [Zelboraf] for a BRAF mutation in melanoma or erlotinib [Tarceva] for several EGFR mutations in non–small cell lung cancer). Gene expression, on the other hand, refers to one or more RNA products transcribed by a gene (or the genome), and is not the target of precision medicine or the output of next-generation sequencing in this context.
The author writes, “Can someone explain to me the evidence-based standard algorithm for routine mathematical evaluation of gene-expression studies, including a way of dealing with the range of expressed mutations? Once that is done, perhaps we can start to consider a standard approach to the clinical interpretation of next-generation sequencing.”
The answer is, “no,” because the question does not distinguish gene expression from gene mutation. Mutation is a dichotomous classification; the BRAF gene is mutant or wild-type, with no algorithm or mathematical evaluation required. Such biostatistical algorithms are required, however, for making sense of quantitative alterations in the expression of tens to hundreds of genes that may relate to tumor classification, prognosis, or response to therapy.
Standardization and Cost Issues
With respect to standardization and regulation, the responsible production of next-generation sequencing data from a patient’s tumor (or germ line), if this information is to be used to guide clinical intervention, is performed in a Clinical Laboratory Improvement Amendments (CLIA)-certified, College of American Pathologists (CAP)-accredited laboratory context.
Regarding “profligate expenditure in health care, which is becoming a huge component of the national debt,” readers may be interested to learn that it is rapidly becoming cheaper to sequence large panels of genes by next-generation sequencing than to sequence a handful of genes using traditional Sanger-based technology. Furthermore, next-generation sequencing is substantially more sensitive for mutation detection, a critical issue when dealing with tissue specimens of low tumor cellularity. The cost of precision drugs is and will be driving the cost of health care in this context, not the sequencing test.
Foundation of Knowledge
Which brings me to my last point. You’re not going to find it if you don’t look for it. The sequencing-based interrogation of the exomes of thousands of tumors has made feasible the creation of a comprehensive catalogue of genetic mutations as they exist in most tumor types. By using sequencing-based gene panels in the clinic, and performing whole “omics”-based studies in the laboratory, we will soon know that mutations in theoretically druggable gene product X occur at a predictable frequency in tumor Y.
This state of knowledge allows for the design of rational, precision therapy–based clinical trials, generating the gold standard level 1 evidence that we all seek, in a most orderly and refined fashion. Thus, “trial designs that are extrapolated from fundamental Bayesian theory” will not be necessary. I wholeheartedly agree that there exists a “need for rational thinking” in this space, but that is possible only upon a foundation of knowledge of the fundamentals of genomic medicine. ■
—Jeff Boyd, PhD
Fox Chase Cancer Center
Disclosure: Dr. Boyd reported no potential conflicts of interest.
I read with interest the note from Jeff Boyd, PhD, Senior Vice-President for Molecular Medicine at Fox Chase Cancer Center, calling into question my recent commentary about the high costs of partly validated testing in the domain of molecular medicine. One of the oldest tricks in the book is to...