The impact of tumor heterogeneity on patient treatment decisions.

The existence of intratumor phenotypic heterogeneity was described a few decades ago, but the biological impact of this clonal diversity remains unknown. In 1976, Peter Nowell stated in Science that tumors arise from a single cell that has undergone several expansion rounds and whose descendants eventually acquire genetic aberrations. He also posited that additional mutational events would drive the behavior of the tumor and determine how the tumor will evolve (1 ). Nowell proposed the concept that tumor progression is linked to a genetic evolution that follows a Darwinian pattern, thus bringing the notion of tumor evolution to bear on this heterogeneity. Since that time, numerous studies have reported the presence of subpopulations within tumors, and it has been proposed that this clonal diversity could affect tumor evolution by (a) providing the diversity on which selection can work and (b) modulating progression and resistance through biological interactions between the clones (2 ). The New England Journal of Medicine recently published a report by Gerlinger et al. entitled “Intratumor Heterogeneity and Branched Evolution Revealed by Multiregion Sequencing” (3 ), which evoked the concept of intratumor heterogeneity playing a role in cancer progression and metastasis and in complicating molecularly driven treatment strategies. The authors investigated primary renal carcinomas and their associated metastases via multiregion whole-exome sequencing, chromosome aberration analysis, and DNA ploidy profiling of spatially separated parts of each sample. Immunohistochemical and gene expression analyses were also performed. Their data showed that 63% to 69% of all somatic mutations were not detectable across every tumor region and that a phylogenetic reconstruction of the tumors revealed an evolutionary tumor growth pattern. Additionally, the authors observed different gene expression signatures and divergent allelic-imbalance profiles within some of the tumors. The authors concluded that a single tumor biopsy sample might reveal only a fraction of the genetic aberrations present and therefore might not represent the diversity present in the entire tumor. Furthermore, they showed that reconstructing the tumor clonal architectures by localizing aberrations in the root of the phylogenetic tree might bring substantial insights that could contribute to improving therapeutic approaches, if these “root” events can be targeted. Although the magnitude of intratumor heterogeneity and the mutational status at different cancer stages are unclear, several studies have demonstrated that the clinical implications can be huge. One study has shown a strong correlation between the degree of clonal diversity and the probability of invasive progression of Barrett esophagus to esophageal lesions, thus correlating heterogeneity with the development of malignancy [reviewed in (2 )]. Other studies have shown that mutational status can predict resistance to treatment by preventing the binding of a drug to its target. One of the best examples is the BCR-ABL fusion protein in chronic myeloid leukemia. Its presence is a predictor of a patient’s response to imatinib mesylate, but some mutant forms of the BCR-ABL protein have been implicated in the relapse of chronic myeloid leukemia. This pattern has been found with other tumor types, including melanoma and gastrointestinal stromal sarcomas [also reviewed in (2 )]. With the accumulating evidence that intratumor heterogeneity seems to be closely related to tumor progression and the development of resistance to therapy, one of the most challenging aspects of studying this heterogeneity is to determine how best to account for heterogeneity when planning sample acquisition for molecular analysis and clinical decision-making. A decision about a patient’s treatment might require consideration of not only the dominant clone but also any minor clones that might present clinically relevant aberrations. As Gerlinger et al. (3 ) and others have indicated, a single biopsy may not reflect the composition of the entire tumor, owing to the clonal heterogeneity. One option is to collect samples from spatially separated parts of a tumor for analysis. Although this approach might be feasible for surgical samples, it is more 1 Sheikh Khalifa bin Zayed Al Nahyan Institute of Personalized Cancer Therapy (IPCT) and Departments of 2 Surgical Oncology, 3 Experimental Therapeutics, and 4 Systems Biology, University of Texas MD Anderson Cancer Center, Houston, TX. * Address correspondence to this author at: University of Texas MD Anderson Cancer Center, 7435 Fannin St., Houston, TX 77054. Fax 713-563-4235; e-mail [email protected]. Received August 14, 2012; accepted October 15, 2012. Previously published online at DOI: 10.1373/clinchem.2012.194712 Clinical Chemistry 59:1 000 – 000 (2013) Perspectives