Encouraged by a growing arsenal of targeted cancer drugs and the ability to molecularly profile tumours, oncologists are increasingly using targeted drugs off-label to treat patients. Up to 30% of cancer drug use is off-label. A first randomized trial of off-label targeted drugs versus chemotherapy has now shown that this personalized approach does not benefit patients (Lancet Oncol. 2 Sep 2015 [epub ahead of print]).

Christophe Le Tourneau, an oncologist at the Institut Curie in Paris, France, and the lead author of the study, randomized 195 patients with 'actionable' abnormalities in a range of cancers to one of two arms: treatment with a potentially relevant, targeted drug, or treatment with chemotherapy. Progression-free survival was 2.3 months in the experimental personalized-treatment arm, non-significantly different from the 2.0 months on chemotherapy. There was a non-significant trend towards increased toxicity with the targeted drugs.

“Our trial has drawbacks,” admits Le Tourneau. It was a small trial, in patients with advanced cancer who may have been unlikely to benefit. The investigators were also limited in terms of the drugs that they could use (only approved drugs), and applied a simplistic algorithm for matching molecular abnormalities to targeted drugs. But many of these limitations also apply for oncologists who use drugs off-label in the clinic, he adds.

The trial “offers robust evidence for the deficiencies in assigning therapy based on the various loose associations between biomarkers and inhibitors that are often provided in commercial clinical diagnostic reports. The results suggest that off-label use of molecularly targeted agents in this manner should be restricted,” writes Daniel Catenacci, an oncologist at the University of Chicago, Illinois, USA, in an editorial about the trial (Lancet Oncol. 2 Sep 2015 [epub ahead of print]).

Several other trials are also ongoing to test the broader use of targeted cancer drugs, including the NCI-MATCH trial, which will test 25 targeted drugs in 1,000 patients with different tumour types (Nat. Rev. Drug Discov. 14, 513–515; 2015).