The Thirteen-Month Dividend

Thirteen months can hold a birthday, a conversation postponed, a season someone thought they would not see. That is why a survival headline matters. It is also why its arithmetic deserves care. The title of this essay names a trial median. It cannot name the time a particular person will receive, and it does not mean thirteen additional months beyond the alternative treatment.

On 26 August 2026, the FDA approved daraxonrasib, sold as RASONQUE, for adults with metastatic pancreatic adenocarcinoma who have received at least one prior systemic therapy or are not candidates for multiagent systemic therapy. That indication establishes a particular clinical setting. It is not an approval for every pancreatic cancer diagnosis. My question begins after that distinction: what would count as evidence that a treatment benefit has become reachable?

The number and the person

The FDA label’s section 14 describes RASolute 302: 500 patients randomized to daraxonrasib or physician-selected chemotherapy. They had progression after one prior treatment line and ECOG performance status 0 or 1. In the overall population, median overall survival was 13.2 months, with a 95% confidence interval from 10.0 months to an upper bound not estimable, versus 6.7 months, with an interval of 5.8–8.0. The hazard ratio was 0.40, with a 95% interval of 0.30–0.53. The RAS G12 subgroup’s chemotherapy median was 6.6 months.

Subtracting the overall-group medians gives 6.5 months. This is a comparison of two group summaries, not the extra time assigned to each person. A survival median is the point at which the estimated survival curve reaches one-half; it is neither an expiry date nor an average number of months gained. The hazard ratio is another summary of the comparison. It cannot be read as a sixty-percent extension of every life.

The randomized comparison supplies strong evidence of benefit in the studied setting. The eligibility criteria still matter when that evidence travels into a consulting room. The trial population and the approved population are not identical descriptions. Nor does a survival endpoint settle how someone values treatment, symptoms, independence or additional appointments. Those questions need their own evidence and a conversation in which the patient has agency.

The label also warns about skin and soft-tissue toxicity, mouth disorders, diarrhea, gastrointestinal perforation and lung inflammation. A tablet changes the mode of delivery; it does not remove the need for clinical support. I would count appropriate supportive care as part of access, rather than treat dispensing the first bottle as the completion of the system’s work.

Which price is being counted?

Revolution Medicines’ 26 August SEC filing disclosed a wholesale acquisition cost of $39,800 for a 30-day supply at the recommended dose. Twelve such supplies total $477,600 and cover 360 days. That multiplication makes the scale visible. It does not establish a full calendar-year treatment duration, a payer’s net expenditure, a patient’s bill or a cost per life-year gained. Each requires a different denominator and additional information.

A large list price raises a legitimate question about financing. It does not answer the access question by itself. The money charged upstream, the amount paid by an insurance system and the cash someone must find before starting treatment are connected quantities, but they are not interchangeable. Collapsing them into one dramatic figure can obscure both the protections that exist and the gaps those protections leave.

CMS’s instructions set a $2,100 annual out-of-pocket threshold for covered Medicare Part D drugs in 2026, after which covered-drug cost sharing is zero for the rest of the year. The word “covered” is essential. This program rule does not verify RASONQUE coverage in any particular plan, erase premiums and other medical expenses, or establish that navigating the process is easy. It does show why the launch list price cannot be presented as every Medicare patient’s personal liability.

The manufacturer’s (ON)Path terms describe a commercial copay program that excludes beneficiaries of government healthcare programs. A separate patient-assistance program may be available to uninsured or underinsured people who meet medical and financial criteria. These are conditional routes, not universal guarantees. Their existence belongs in an honest account of access; so do their limits. A program page supplies terms, not the number of people helped or turned away.

The people outside the median

Consider a constructed audit, not patient data. In one cohort, 80 of 100 clinically eligible people begin treatment within the observation window, and the median wait among those starters is seven days. In another, 60 of 100 begin, with a five-day median among starters. Judged only by that median, the second service looks faster. Judged by the fraction who reach treatment, it performs worse. The missing forty people do not become visible merely because the chart is precise.

A useful access record would therefore begin at a declared clinical decision point, count eligible people who do and do not start, and state the observation window. It would distinguish coverage problems, financial obstacles, clinical changes, missing information and a patient’s informed decision to decline. Non-starting is not automatically a system failure. Treating every non-starter as a denial would substitute another crude story for the first.

I would also ask who can correct that record. A person whose application is marked incomplete needs a practical route to find the missing step. A patient choosing a different treatment should not disappear into a category labelled “failed access.” The audit should publish aggregate results with the relevant privacy protections, and explain its categories before ranking services. These are proposed requirements, not observations about the launch.

A benefit and an obligation

The clinical result deserves recognition on its own terms. The access argument gains force when it identifies what remains unknown, rather than filling those gaps with a predicted decade of exclusion or a numerical dignity score. Approval, financing, appropriate care and patient choice are distinct achievements. None becomes complete because another has occurred.

The dividend I want to watch is the opportunity to use a beneficial treatment when it is appropriate, with its burdens understood and its practical route open. A rigorous account would preserve the trial’s result while following the people whom a launch announcement cannot see. The question is not whether a headline can carry hope. It is whether the institutions behind it can make that hope answerable to a person.