Lung cancer patients on Roche's immune system-boosting drug Tecentriq lived on average 4.2 months longer than those taking chemotherapy in a pivotal study, pressuring Bristol-Myers Squibb's dominant position in the field.

The trial involved second-line patients, who have already used chemotherapy, and it found the Roche drug worked even in people with low or no levels of a protein called PD-L1, which is often used to test suitability for immunotherapy.

The study results were presented at the annual European Society of Medical Oncology congress on Sunday.

Bristol's Opdivo is currently the only immunotherapy drug approved without a PD-L1 test for non-small cell lung cancer (NSCLC), the most common form of the disease. Merck & Co's Keytruda requires a test.

That has favored Opdivo and fueled its sales, but the arrival of a rival with the same broad effectiveness now promises increased competition.

U.S. regulators are due to decide whether to approve Tecentriq in lung cancer by Oct. 19. The Roche drug is already cleared for use in bladder cancer.

Tecentriq, Opdivo and Keytruda are the first of a highly promising class of immunotherapy drugs that are revolutionizing cancer treatment and are tipped to generate huge sales. Lung cancer, the biggest cancer killer globally, is the top commercial opportunity.

Analysts believe Tecentriq sales across all cancers will reach $4 billion in 2021, according to consensus figures compiled by Thomson Reuters.

In the 1,225-patient Roche study, the mean survival for patients on Tecentriq was 13.8 months, which was 27 percent better than the 9.6 months seen among those on docetaxel chemotherapy.

The benefit of the new drug was greatest in patients with high levels of PD-L1 but even in people with no PD-L1 expression there was still a significant improvement.

Martin Reck of Germany's Lung Clinic Grosshansdorf said the findings suggested doctors had a problem if they planned to use a negative PD-L1 reading as a reason to exclude second-line NSCLC patients from treatment.

Companies are also racing to prove that immunotherapy can help in first-line NSCLC, where it would replace chemotherapy as the initial drug given. Here, however, the idea of giving immunotherapy to all patients has proved problematic, as evidenced by the failure of a recent Bristol trial.

Roche has eight late-stage Phase III lung studies under way evaluating Tecentriq alone or in combination with other treatments in patients with early and advanced stages of lung cancer.

No silver lining in failed Bristol cancer drug trial

Bristol-Myers Squibb disappointed investors on Sunday as researchers detailed data for its Opdivo cancer immunotherapy, which was already known to have failed to do better than older chemotherapies in a closely watched clinical trial.

A trader passes by a screen displaying the tickers symbols for Bristol-Myers Squibb and Intelsat, Ltd. on the floor at the New York Stock Exchange, April 25, 2013.

Some analysts had speculated the study would at least show Opdivo worked in a subset of lung cancer patients with a biomarker that should make them more receptive to immunotherapy, but there was no sign of this.

Bristol first announced the trial failure in August, since when its shares have fallen by around a quarter, but experts have been waiting eagerly to learn exactly what went wrong and see if something might be salvaged from the setback.

In the event, data on Sunday showed patients actually did worse on Opdivo, surviving only 4.2 months before their disease worsened against 5.9 months for those on chemotherapy, although the difference was not statistically significant.

What is more, researchers told the European Society for Medical Oncology congress that they were unable to point to any group of patients who did better in the trial, possibly due to imbalances between patients in different subsets.

Bristol took a big gamble by accepting a wide range of previously untreated lung cancer patients into its trial, in contrast to Merck which succeeded in a similar study by only taking people with high levels of a protein called PD-L1.

Immunotherapies like Opdivo and Merck's Keytruda work by taking the brakes off the immune system and are known to be most effective when tumor cells express lots of PD-L1.

Some investors had hoped Bristol would be able to point to a clear benefit in patients whose tumors had at least 50 percent of cells producing PD-L1. However, even in this group the trial did not show a benefit.

Commenting on the results, Naiyer Rizvi of Columbia University Medical Center, who was not involved in the study, said the failure to see a benefit was unexpected and hard to resolve with other clinical trials.

Bristol's bad news day contrasted sharply with that of Merck, which was able to point to a double success in clinical trials using Keytruda.

Source: Reuters