Divarasib lung cancer trial extends survival in new study

Roche said on Sunday that its experimental drug divarasib cut the risk of death by 44% compared with existing treatments for people whose lung cancer had already been treated once and then returned or kept growing, as reported by Reuters and in a Roche announcement. The divarasib lung cancer trial compared the new drug directly with two drugs already used for the same kind of cancer, and the result gives doctors a head-to-head comparison rather than a result measured against a placebo or against older chemotherapy alone. According to Roche, the findings were presented at a major oncology congress and published in a leading medical journal on the same day, a pairing that usually signals that the company expects the data to be scrutinised closely. For patients who have run out of good options after an earlier targeted treatment, a direct comparison of this kind matters because the next choice is often made with thin evidence about which drug works better after the first one stops working.

What the survival results show

Patients who received divarasib lived for a median of nearly 18 months, compared with 12 months for patients who received one of the two comparison drugs. That gap in median survival is the headline result of the divarasib lung cancer trial that will draw the most attention from oncologists, because median survival describes the point at which half of the patients in a group were still alive, and a longer median suggests that the benefit was not limited to a small group of exceptional responders. The comparison matters for another reason as well. Both comparison drugs block the same cancer-driving signal as divarasib, so the trial was testing whether a newer version of the same approach could do the job more effectively, not whether blocking the signal helps at all. Roche described the survival advantage as statistically significant, which in plain terms means that the difference was large enough, given the size of the trial, that it is unlikely to be explained by chance alone.

The divarasib lung cancer trial also found that divarasib lowered the risk of the disease getting worse or of death by 42%, with median progression-free survival of 9.8 months against 5.7 months in the control group. Progression-free survival measures how long patients live before scans show that tumours have grown again, so it captures a different part of the story from overall survival. A longer stretch before progression can mean more months without a change of treatment, fewer scans showing bad news, and more time before side effects from a new regimen begin. Doctors often weigh that measure alongside overall survival because a drug that delays progression but does not extend life still changes daily experience for patients, while a drug that extends life without delaying progression can be harder to interpret. Here, both measures moved in the same direction, which strengthens the case that the newer drug was doing more than shifting the timing of a scan result.

How the trial was run and what comes next

The divarasib lung cancer trial enrolled 338 patients and compared divarasib with sotorasib or adagrasib, the two approved drugs in the same class. The objective response rate was 58% for divarasib, versus 32% for sotorasib or adagrasib, a measure that records how often tumours shrank by a defined amount on imaging. Tumour shrinkage does not guarantee longer life on its own, and oncologists are careful to separate a scan that looks better from a patient who actually lives longer. That caution is part of why the survival result in this trial carries weight. When shrinkage, delayed progression, and longer survival all point the same way, the picture becomes harder to dismiss as a quirk of one measurement. The patients in the trial all had previously treated disease, which means that they had already received at least one earlier course of treatment before entering the study, and that history makes the population a realistic test for a second-line option rather than an easier first-line group.

On safety in the divarasib lung cancer trial, Roche said that no new safety findings were observed, and the most common treatment-related side effects were mainly low grade and manageable. Some 4.8% of patients on divarasib stopped treatment because of treatment-related adverse events, versus 11.7% in the control group. A lower discontinuation rate matters in practice because a drug that patients can stay on is more likely to deliver its full benefit, while frequent stops and dose changes can erode an advantage that looks strong on paper. The company also said that it was working with health authorities to bring the treatment to patients, a step that usually involves submitting the full trial data for review rather than relying on a congress presentation alone. Regulators will look beyond the headline figures to the detailed safety tables, the consistency of benefit across patient subgroups, and the quality of the trial conduct before deciding whether the drug should be available more widely.

The cancer at the centre of the trial is a form of non-small cell lung cancer driven by a mutation in a gene that helps control cell growth. When that gene is altered, the signal telling cells to divide can stay switched on, and drugs in this class are designed to block the altered protein and interrupt that signal. Earlier drugs in the class proved that the approach could work, but responses did not always last, and patients often needed another option when the first drug stopped working. A newer drug that works better in that second setting would fill a clear gap, because lung cancer remains one of the most common and most deadly cancers worldwide, and options after an initial targeted treatment fails are limited. The divarasib lung cancer trial therefore speaks to a practical question that clinics face every week, which is what to offer when the first targeted drug has run its course.

There are limits to what the divarasib lung cancer trial can settle on its own, and the publication in a peer-reviewed journal will let independent researchers examine the methods, the patient mix, and the follow-up period in detail. Questions that remain include how long the benefit lasts for patients who respond, whether certain subgroups benefit more than others, and how the drug performs outside the controlled conditions of a trial, where patients often have other health problems and take other medicines. For now, the reported results place divarasib in a stronger position as a candidate for wider use, but they do not by themselves make it a standard treatment. That decision will rest with regulators and with the guidelines committees covered in health coverage and with the guidelines committees that weigh trial evidence, safety, and the experience of patients who have already lived through a first round of treatment and are looking for a second chance that actually lasts.