Precision Medicine Trial in Cancer Prevention Identifies Chemoprevention Strategy
A team of scientists has reported that the genetic biomarker loss of heterozygosity (LOH) is able to predict which patients with premalignant mouth lesions are at highest risk of developing oral cancer.
The findings, published in JAMA Oncology (doi:10.1001/JAMAONCOL.2015.4364), present a new tool that could be used to identify patients most likely to benefit from chemoprevention. They also may be applicable to preventing other types of cancer.
Oral cancer is diagnosed in more than 300 000 people worldwide each year, and 145 000 will die from this disease, which is the most common type of head and neck cancer.
"One of the greatest challenges in developing chemopreventive agents is to identify the population at highest cancer risk," said William N. William, Jr, MD, associate professor at MD Anderson in Houston, Texas, which led the enrollment and design of the study. "Not all patients with an oral premalignant lesion will develop oral cancer. By using a molecular test that can identify those at highest risk, we can focus preventive efforts on these specific individuals."
Erlotinib (Tarceva) is an inhibitor of the epidermal growth factor receptor (EGFR), a protein responsible for the development of many cancer types, including head and neck cancer. It is currently approved for the treatment of specific lung and pancreatic cancers and has shown promise in preventing oral cancer in animal models.
The Erlotinib Prevention of Oral Cancer (EPOC) trial involved 379 patients at 5 sites across the country. All had premalignant lesions in their mouths. Following study enrollment, participants were evaluated for LOH, a chromosomal abnormality characterized by the loss of chromosomal regions, which include tumor suppressor genes.
Patients who tested positive for LOH were considered to be at high risk for oral cancer and were randomized to receive either erlotinib or a placebo for 1 year. The study's primary end point was to determine if fewer patients treated with erlotinib would develop oral cancer, compared with those that received placebo.
"To the best of our knowledge, EPOC represents the first study of oral premalignant lesions to use cancer as the primary end point, and the first molecularly based precision medicine trial design in cancer prevention," said Scott Lippman, MD, director of Moores Cancer Center at University of California (UC) San Diego Health and senior author of the 6-year study. The study began in 2006 when Lippman was chair of the Department of Thoracic/Head and Neck Medical Oncology at MD Anderson.
The EPOC study demonstrated that LOH predicted a higher oral cancer risk. LOH-negative patients had a 3-year cancer-free survival rate of 87% compared with 74% for the LOH-positive group. However, patients who took erlotinib showed no statistical difference in terms of cancer-free survival rates after 3 years: 74% for participants given erlotinib compared to 70% for those taking placebo. Patients with both LOH and EGFR copy number gains had the highest incidence of cancer, but still did not benefit from erlotinib.
The lack of erlotinib benefit overall is disappointing, said Lippman, but still an important finding.
"It's vital to know what doesn't work as well as what does, and this research furthered progress in other ways," Lippman said. "For example, we demonstrated that erlotinib-treated patients who develop a skin rash, a common side effect of the drug, had the lowest incidence of oral cancer. Further investigations of this association could point to novel biomarkers, mechanisms, and trial designs related to the effects of EGFR-targeted agents in patients with oral premalignant lesions or head and neck cancers."