Lung cancer · Nat Commun · 2025
The paper’s summary: half of 63 patients lived 11.1 months without cancer growing. Its data: 5.5.
Months until half the patients’ cancer had grown or they had died
The paper says
Lung-cancer pill trial: each cancer had extra DNA in one gene, at spot A or B. Does the spot matter? For 63 spot-A patients, the summary at the top of the paper says half lived 11.1 months without cancer growing; for 19 spot-B patients, 3.5.
Its own data show
The paper’s detailed results section gives 5.5 months for the same 63 patients, and redoing the calculation from its patient-by-patient file gives 5.5 too. That file labels 42 of the 63 “grew or died” and 21 “not yet”. Flip those two labels and the file gives the summary’s 11.1, the only change found that does.
“Not yet” means alive with no growth seen at the last check-up; such patients count as lasting at least that long. Flip the labels and 21 count as grown or died, not 42, so 5.5 becomes 11.1: near a rival drug’s 12.0 months, which the summary sets beside it. The 5.5 is under half that, yet beats spot B’s 3.5.