Loading
Dying of something that is not the cancer. In prostate cancer most men do, which makes it the outcome that decides how much any treatment can possibly help. It is also where the inequality that survives equal cancer care shows up, and almost no cancer service measures it.
Other-cause mortality is death from any cause other than the disease under study, counted as a competing risk: once a man has died of a heart attack he can no longer die of prostate cancer, so the two are not independent and cannot be estimated with an ordinary Kaplan-Meier curve. The correct handling is a cumulative incidence function with a Fine and Gray subdistribution hazard, which is what the prostate literature reports as a subdistribution hazard ratio. Prostate cancer has an unusually strong competing-risk structure: the disease is common, slow, and diagnosed at a median age at which other causes of death are already the majority, so other-cause mortality sets a ceiling on how much benefit any prostate cancer treatment can deliver.
It carries the equity finding, which is why it has an entry here. Dess and Spratt assembled individual patient data on men with clinical T1 to T4, N0 to N1, M0 prostate cancer from three cohorts with progressively tighter control of access to care: the Surveillance, Epidemiology, and End Results registry (296,273 men), five equal-access Veterans Affairs medical centres (3,972 men, all treated surgically) and four pooled National Cancer Institute randomised radiotherapy trials (5,854 men), with inverse probability weighting for demographic, cancer and treatment differences. Prostate cancer-specific mortality in Black men fell from an age-adjusted subdistribution hazard ratio of 1.30 in the registry to 1.09 after weighting, was not significantly different in the equal-access surgical cohort (0.85), and was significantly lower in the randomised trial cohort (0.81). Other-cause mortality stayed significantly higher in two of the three: 1.30 in the weighted registry cohort and 1.17 in the weighted trial cohort.
What that means operationally. Once treatment and access are equalised, the excess prostate cancer death largely disappears and the excess death from everything else does not. A prostate cancer service that measures only cancer-specific mortality has therefore made itself blind to the larger surviving gap, in a population it sees regularly for years and treats with androgen deprivation, a therapy that worsens metabolic and bone health. The caveats belong with the finding: these are United States cohorts with a United States access gradient, the analysis is retrospective despite the weighting, and it addresses mortality after diagnosis at a given stage rather than the higher incidence and younger age at presentation in Black men.
The evidence base for setting the age at which screening starts by risk rather than by birthday. It is also a warning: the same score performs differently across ancestries, so a risk model built and validated in European cohorts will under-serve the men at highest risk.
The disparity in prostate cancer death among Black men in the United States is, stage for stage and treatment for treatment, largely a disparity in getting standard care rather than in tumour biology. The disparity that survives equal access is in dying of everything else, which is the part a cancer service is least organised to fix and most able to measure.
The clearest evidence that radical treatment of localised prostate cancer saves lives when the cancer was found clinically rather than by a blood test, and the clearest single statement of what grade does: a Gleason score above 7 carried ten times the risk of death of a score of 6 or lower in the same trial.
The current shape of the screening question in the United States, and the best short statement of the trade-off in numbers a man can weigh. The three-to-one ratio between metastatic cases prevented and deaths prevented is also the argument for using metastatic presentation, not mortality, to judge a screening programme sooner.
The trial that made observation a defensible choice for low-risk prostate cancer found by a blood test, and that supplied the number a man needs when weighing surgery: the progression it prevents is mostly progression on a scan or a blood test, and the harms it causes are felt every day.
The reason intermittent androgen deprivation is offered as a choice in metastatic disease rather than recommended, and a case study in what an inconclusive non-inferiority trial should say. For a man weighing the side effects, the honest statement is that the quality-of-life gain is real but brief and the survival question is open.
Shares Treatment-induced bone loss, Androgen deprivation therapy (ADT), SWOG 9346: intermittent versus continuous androgen deprivation in metastatic prostate cancer, Report death from other causes as an outcome of the prostate cancer service, split by deprivation and ethnicity and the tags gu, prostate-glossary.
Shares Trans-ancestry genome-wide association meta-analysis of prostate cancer identifies new susceptibility loci and informs genetic risk prediction, Association of Black race with prostate cancer-specific and other-cause mortality, Localised prostate cancer, intermediate risk, Localised prostate cancer, very low and low risk and the tags gu, prostate-glossary.
Shares USPSTF 2018: screening for prostate cancer, recommendation statement (grade C at 55 to 69, grade D at 70 and over), Overtreatment, Localised prostate cancer, intermediate risk, Localised prostate cancer, very low and low risk and the tags gu, prostate-glossary.
Shares USPSTF 2018: screening for prostate cancer, recommendation statement (grade C at 55 to 69, grade D at 70 and over), Non-metastatic castration-resistant prostate cancer, Hazard ratio (HR), Metastatic hormone-sensitive prostate cancer and the tags gu, prostate-glossary.
Shares Androgen deprivation therapy (ADT), Quality of life, Androgen deprivation therapy (ADT), Prostate cancer and the tags gu, prostate-glossary.
Shares Overtreatment, Localised prostate cancer, intermediate risk, Localised prostate cancer, very low and low risk, Screening and the tags gu, prostate-glossary.
Shares Localised prostate cancer, intermediate risk, Localised prostate cancer, very low and low risk, Screening, Localised prostate cancer, high and very high risk and the tags gu, prostate-glossary.
Shares Localised prostate cancer, intermediate risk, Localised prostate cancer, very low and low risk, Localised prostate cancer, high and very high risk, Prostate cancer and the tags gu, prostate-glossary.