Take every new chemotherapy drug approved over a dozen years and put the results in one pile. The average overall survival benefit comes to 2.1 months. That is the figure NutritionFacts highlights.

The measure matters. Overall survival is the plainest endpoint in oncology: the time from enrolment in a trial until death from any cause. It cannot be flattered. A drug that shrinks a tumour on a scan, or delays the date the tumour visibly grows, can look excellent on those measures without changing when the patient dies. Those softer endpoints arrive sooner and need fewer patients, which is why they are attractive to trial sponsors. Overall survival takes longer and is harder to game.

Two caveats belong here. An average hides spread: some drugs in the pile will have added far more than 2.1 months, others less, and some patients respond dramatically while most do not. And months are not the only thing a patient counts. Symptom control and quality of life are real benefits that a survival number does not capture. The snippet I have does not break down which drugs or which cancers, so I will not pretend to.

What it does establish is a floor for expectations. A dozen years of approvals, and the pooled gain in life is measured in weeks.

Set that beside a different market. CNBC reports that the wealthy and ultra-wealthy are spending more on longevity care than ever before, with the headline figure at $250,000 a year. Oncology drugs go through trials with a hard endpoint and still return 2.1 months on average. Longevity clinics sell a much larger promise, extra healthy years, and the buyer usually receives a set of biomarker readings rather than a survival curve.

My read: the chemotherapy number is the honest one, because it was measured against death. The question to put to any clinic charging six figures is what its own version of that number is, and whether anyone has ever measured it.

The science establishes that new cancer drugs, taken as a class, have added modest time on the endpoint that cannot be argued with. It does not establish which individual drugs are worth their cost to which patients. That answer sits in the per-drug trial data, drug by drug, not in the average.