Does using a drug sooner buy time over the course of treatment?

In the SAFFRON lung cancer trial, adding a MET inhibitor benefited patients whose tumors progressed after osimertinib and showed high MET expression or amplification. That raises a separate question. If we could detect the same change before progression showed up on a scan, would adding the drug then be better than waiting? Original SAFFRON study ↗

There is a rationale for acting early. If other mechanisms of resistance develop while we wait, suppressing the pathway first detected might control the disease for less time. On the other hand, even when a molecular change is detectable, patients can continue to benefit from their current treatment while it controls the disease and defer the side effects of a new drug. We need to follow what happens after a treatment change to find out which benefit is greater.

SERENA-6, a breast cancer trial we previously covered, gives us a way to think about this difference. In both groups, the clock started when an ESR1 mutation was detected. Median progression-free survival was 16.8 months among patients who switched to camizestrant early and 9.2 months among those who continued their existing drug. Measured from that same point, the time to progression after the next treatment or to death was 25.7 months versus 19.1 months. The difference persisted into the next line of treatment. Subsequent therapy in the continuation group was left to clinical judgment at progression, however, so these results cannot isolate the value of using the same drug sooner compared with using it at progression. Our previous breast cancer article ↗ · SERENA-6 follow-up analysis ↗ · FDA advisory committee transcript ↗

The value of early detection and early treatment has diverged in practice. In a randomized trial of recurrent ovarian cancer, starting chemotherapy when the blood marker CA125 rose yielded no demonstrated improvement in overall survival compared with waiting for clinical recurrence. The disease and treatment differ from those in the breast cancer trial. The example illustrates why a test that detects a change early leaves a separate question about the benefit of acting on it. Randomized trial in ovarian cancer ↗

When the FDA granted accelerated approval to camizestrant, it required a confirmatory trial that compares using the drug when the mutation is detected with waiting until progression appears on imaging and then using the same drug. The trial will follow time to failure of each treatment strategy and overall survival. If early use delays the first progression alone, the benefit may reflect bringing forward the next drug’s effect. A longer period of control across the treatment strategy would point to a benefit from intervening earlier itself. FDA accelerated approval letter, confirmatory trial requirements ↗

In CheckMate 067, a study in advanced melanoma, overall survival at ten years was 43% in the group assigned to nivolumab plus ipilimumab. In a separate analysis, 39% of the combination group were alive at five years, having stopped their assigned treatment and without having started another systemic treatment. The researchers separately measured the time spent with side effects that persisted after treatment ended. CheckMate 067 ten-year follow-up ↗ · Analysis of time without subsequent systemic treatment ↗

You can hope for a cure and still find yourself asking how long the next treatment will work. In advanced cancer where a cure is difficult, that is sometimes what progress looks like. The gap between those two hopes is hard to sit with.

Even so, the same few months can feel very different. Were symptoms under control enough for someone to keep working or get on with life at home? Could they go without starting the next drug? Or did an earlier switch bring clinic visits and side effects forward as well? When trials track the time to progression, we also want to know what those days were like for the people living them.

If we detect resistance early, does using the drug sooner give patients more time to live as they wish than waiting to use the same drug? Answering that question would tell us what an earlier switch means from the patient’s point of view.