Beyond genetics in personalized cancer treatment: assessing dynamics and heterogeneity of tumor responses.

Abstract

Cancer has been considered a genetic disease for over a century – ever since von Hansemann first described abnormal chromosomes in cancer cells [1] and Rous demonstrated the ability of transmissible agents (viruses) to cause cancer in chickens [2]. By linking genetic alterations with disrupted biochemical processes, Tatum and Beadle paved the way for others to identify individual genes responsible for oncogenic transformation [101]. Major breakthroughs came in the 1970s when Martin described that a viral gene (v-src) is required to maintain the transformed state of virally infected cells [3] and Varmus, Bishop and colleagues discovered that viral oncogenes have mammalian cell orthologs [4–6]. Subsequent studies determined that many cellular proto-oncogenes are kinases [7–9]. This realization paved the way for the development of targeted cancer therapies that specifically inhibit these oncogenic kinases. With the advent of high-throughput sequencing and the ability to screen thousands of genes from individual tumor samples, many tumors that were once evaluated solely by their histopathological characteristics are now being genetically characterized in astounding detail. With sufficient numbers of molecularly characterized tumors, statistical correlations have suggested links between specific alterations of oncogenic kinases within distinct tumor types or subtypes [10,11]. Some of these molecular alterations are the basis of successful personalized cancer therapy, including HER2 amplification in breast cancer [12], BCR–ABL translocations in chronic myelogenous leukemia [13], mutations within the kinase domain of the EGF receptor (EGRF) in adenocarcinoma of the lung [14,15] and activating BRAF mutations in melanoma [16]. In each of these cases, specifically targeting the genetically altered kinase has provided significant clinical benefit to patients by prolonging overall survival and decreasing treatment-associated morbidity when compared with conventional chemotherapy. It is thought that the success of these drugs is due to the tumor cells being ‘addicted’ to the oncogenic signaling such that when the oncogenic signal is abrogated, the tumor cells stop dividing or die [17,18]. However, the targeted treatments are not curative and the tumors ultimately progress even in the continued presence of the targeted agents, clearly indicating that not all tumor cells die in response to treatment and setting the stage for resistance. Progression of initially responsive tumors is a major challenge to personalized cancer therapy since the clinical benefit of treatment is generally reported in months rather than years, and many questions remain about how and when tumors recur. Answers to these questions would provide enormous clinical benefit, providing physicians with the information necessary to select optimal patient treatment options.