A surge of genetic testing services has many people asking a difficult question: What is my risk of getting cancer? But for those just diagnosed, doctors must field a new question: How long do I have? For patients with non–small cell lung cancer, this question is particularly difficult to answer. Detected early, during stage I or II of the disease, patients have pretty good odds at five-year survival. But actual survival times vary widely and, for most patients, diagnosis does not come until the disease has reached stage III or IV.
New work from HSPH, however, has identified in the lung cancer tumor genome a set of five common genetic variations—called single nucleotide polymorphisms, or SNPs—that predict survival. Patients with stage I or II lung cancer and none of these risk-associated variations, called risk alleles, can expect to live five or more years. But for those with tumors that have three or more risk alleles, the odds drop precipitously to a 50–50 chance of surviving just 20 months.
The study’s authors, led by David Christiani, HSPH professor of occupational medicine and epidemiology and HMS professor of medicine, examined tumor tissue collected at Massachusetts General Hospital, sequenced the cells, and then validated the data against the Cancer Genome Atlas. In collaboration with colleagues from the HSPH Program in Quantitative Genomics, the researchers filtered out the genetic variations not statistically associated with survival. To further narrow the field, the team analyzed tumor tissue from Norwegian patients with Stage I or II lung cancer to identify the risk alleles present in the samples, assembled by the National Institute of Occupational Health in Oslo, Norway. This step trimmed the set to five risk-associated variants located within four known genes: STK39, PCDH7, A2BP1 and EYA2. The work is described in the June Journal of Clinical Oncology.
While these molecular markers have the potential to add accuracy to survival predictions made by the traditional, histological, cells-under-a-microscope stage classifications, it is unlikely that the staging system will incorporate such markers in the near term. Globally, genetic analyses remain out of reach for most clinics.
But for early-stage lung cancer, these markers could help clinicians stratify patients into different risk groups, said co-author Rebecca Heist, HMS associate professor of medicine at MGH. For instance, assuming these markers are validated in a larger study, a clinician might advise a patient with multiple risk alleles to endure the side effects of postsurgical chemotherapy, increasing the odds of long-term survival.
Students may contact David Christiani at dchris@hsph.harvard.edu for more information.
Conflict Disclosure: The authors declare no conflicts of interest.
Funding Sources: The National Institutes of Health (National Cancer Institute) and the Norwegian Cancer Society; the content of the work is the responsibility solely of the authors.
