DANIEL KARP, MD
Results of a long-term intergroup study on the effect of selenium in early non-small-cell lung cancer highlight the differences between smokers and nonsmokers, and support the thesis that "good" supplements may be harmful in the presence of carcinogens.
The double-blind study (E5597) included 1,522 patients with stage IA and IB NSCLC who were randomized 2:1 to receive selenium yeast (200 μg daily) or placebo for four years. All patients had undergone surgical resection of their primary tumor, were cancer-free for at least six months after surgery, were not taking excessive vitamin supplements, and had normal liver function and normal chest x-ray. Prior to randomization, patients were stratified by smoking status (active, former, never), gender, and disease stage/prior treatment (1A, 1B no chemotherapy, 1B prior chemotherapy).
MARK G. KRIS, MD
Supplements aside, patients need to quit smoking
The study results highlight the importance of rigorously testing supplements in the same way medicines are tested, and the need for a strategy to improve prevention of secondary tumors, said Dr. Kris, chief of the thoracic oncology service at Memorial Sloan-Kettering Cancer Center in New York.
"We do have a strategy for these patients and that is to get them to stop smoking," Dr. Kris said. "It has been shown that if you can get patients to stop smoking, all of the outcomes will be better. That is something that every oncologists and physician can do today."
Conducted between October 2000 and November 2009, the nearly 10-year study was halted early at the four-year median follow-up mark because the patients in the selenium group showed a poorer progression-free survival (PFS) than did those in the placebo group.
Differences in PFS began to appear at about 28 months, and at 5 years, PFS was 78% in the control group vs 72% in the selenium group (P = .15), reported lead investigator Daniel Karp, MD, a professor of thoracic/head and neck medical oncology at Houston's M.D. Anderson Cancer Center.
"Paracelsus, the ancient alchemist and physician, stated that 'Poison is in everything, and no thing is without poison. The dosage makes it either a poison or a remedy,'" commented Donald I. Abrams, MD, president of the Society for Integrative Oncology and chief of hematology/oncology at San Francisco General Hospital. "Selenium has a very narrow range of safety. Perhaps instead of taking a selenium supplement, we should be asking people to eat a Brazil nut daily because it provides the recommended dietary allowance! I am a big supporter of obtaining essential nutrients from whole foods rather than capsules."
Overall, 216 second primary tumors of any type developed in 190 patients, occurring in 4.1% of patients treated with selenium and 3.7% in the placebo group. The selenium group also had a higher incidence of secondary lung cancer tumors. Overall, 84 of the 216 second primaries (38.9%) were lung cancers, occurring in 1.9% of patients treated with selenium and 1.4% in the placebo group. Overall survival was about 5% lower at three and five years in the selenium group, compared with placebo (ASCO 2010 abstract CRA7004).
Although the poorer outcomes in the selenium group were not statistically significant, Dr. Karp emphasized that they certainly weren't better, and a futility analysis showed that, over time, these results would not revert to a positive outcome for selenium.
Asked why the patients treated with selenium did worse, Dr. Karp drew on prior data that has shown a poorer outcome in patients with lung cancer who smoke and take large amounts of beta-carotene.
"Roughly 80% of lung cancer patients will be current or former smokers, and we saw in the beta-carotene study that patients with lung cancer taking beta-carotene in large amounts did worse," he said. "This was confirmed with a variety of data showing that these antioxidants may be harmful in the presence of carcinogens."
The poorer outcome with selenium supports these data, Dr. Karp said, emphasizing that the association was observational and not based on any "ironclad" evidence. But the take-home message is that lung cancer is a different disease in people who smoke and that lung cancer patients need to be stratified by smoking status, he said.