CHICAGO — A personalized medicine initiative in heavily pretreated cancer patients found that they lived about 4 months longer when the choice of treatment matched their individual molecular abnormalities, investigators reported at the annual meeting of the American Society of Clinical Oncology.
Patients whose tumors displayed a single molecular abnormality and were treated in a phase I clinical trial of a matched therapy survived a median of 13.4 months, compared with 9.0 months for patients in trials that did not match a therapy to an abnormality (P = .017), said Dr. Apostolia-Maria Tsimberidou of the University of Texas MD Anderson Cancer Center in Houston.
Time-to-treatment failure was 5.2 months for matched therapy, vs. 2.2 months for unmatched (P less than .0001). In addition, 27% of patients on targeted therapies had a complete or partial response, compared with only 5% of controls (P less than .0001), and 23% of matched patients had stable disease for more than 6 months, vs. 10% of those without a match.
The findings point to the growing importance of personalized medicine based on the unique molecular profiles of individual patients, and to the role molecular profiling can play in bringing targeted therapies to market faster, commented Dr. Richard Schilsky, a professor of medicine in the section of hematology-oncology at the University of Chicago Pritzker School of Medicine and a past-president of ASCO.
"I think what this study illustrates is, if you have a way of identifying which therapy is likely to work in a patient, you can substantially increase the likelihood of benefit even in a patient with advanced cancer and even in a phase I trial," he said.
The M.D. Anderson investigators looked for molecular aberrations in tissue samples from 1,144 patients. Using various genetic sequencing or immunohistochemistry techniques, they identified abnormalities in a total of 460 patients (379 with 1 aberration, 73 with 2, and 8 with 3 or 4 aberrations).
The driver mutations or aberrations screened for included PIK3CA, KRAS, NRAS, BRAF, EGFR, PTEN (loss), p53, c-KIT, G-NAQ, c-MET, and ALK. All tissue samples were assessed for at least 2 mutations.
Tumors with frequent abnormalities included melanoma (in 73% of cases) and thyroid (56%), colorectal (51%), endometrial (43%), lung (41%), pancreatic (41%), and breast cancers (32%).
The median patient age was 56 years, and the median number of prior therapies was 4. A total of 291 patients went on to additional therapy, 175 with treatment matched to one of the aberrations, and 116 with unmatched treatment.
In multivariate analysis, investigators found that significant predictors for a therapeutic response were matched therapy vs. non-matched therapy (odds ratio 6.33, P less than .001) and normal lactate dehydrogenase (LDH; OR 2.16, P = .045). Predictors for time-to-treatment failure included matched therapy (OR 0.42), normal vs. low albumin levels (HR 0.45, P less than .0001), normal vs. high platelet levels (OR 0.34, P = .001), and 2 or fewer metastatic sites (OR 0.68, P = .003).
For the short term, the investigators plan to screen for additional genetic and/or molecular aberrations to identify new targets for new drugs, and to study combinations of drugs against more than one target. They hope to eventually test all new patients for abnormalities and conduct clinical trials demonstrating the clinical efficacy of testing, so that it can be reimbursed, Dr. Tsimberidou said.
Paula M. Fracasso, M.D., Ph.D., the Lawrence W. Penniston, M.D., Family Professor of Women’s Oncology Research at the University of Virginia Cancer Center in Charlottesville and the invited discussant, commented that the investigators failed to determine whether there may have been statistically significant differences between matched and unmatched treatment groups in the number and types of mutations.
Nonetheless, she applauded the attempt as a good early start. "Over the next few years, the future of personalized medicine has to happen, and should include molecularly profiling the entire tumor, characterizing and selecting the best targeted agents, getting the best of class agents, and performing prospective, randomized trials of the best agents," she said.
Dr. Tsimberidou disclosed a consultant/advisory role with Cephalon and Sanofi. Co-author David Hong disclosed research funding from Eisai. Dr. Schilsky disclosed consulting and receiving honoraria from Foundation Medicine. Dr. Fracasso had no disclosures.