Why there is still no screening test for ovarian cancer

Ovarian cancer is the kind of disease screening was built to fight: it is usually found late, and it is far more survivable when caught early. Yet there is no recommended screening test for it, and the reason is one of the most instructive results in the whole field. When screening was finally put to a large randomised trial, it did exactly what it was supposed to do to the timing of diagnosis — and it still did not save lives.

The disease

Ovarian cancer causes approximately 14,000 deaths a year in the United States, making it the fifth most common cause of cancer death among women and the leading cause of death from a gynaecologic cancer [s1]. More than 95 percent of those deaths occur in women aged 45 and older [s1]. Because it tends to be diagnosed at an advanced stage, the appeal of a test that could find it earlier is obvious.

The trial that settled the question

The UK Collaborative Trial of Ovarian Cancer Screening, UKCTOCS, was built to answer it directly. More than 200,000 postmenopausal women were randomly assigned to one of three groups: annual multimodal screening using a blood test and ultrasound, annual transvaginal ultrasound alone, or no screening — 50,625, 50,623 and 101,314 women respectively [s2]. They were followed for a median of 16.3 years [s2].

Screening worked, in the narrow sense it was designed to. Compared with no screening, the multimodal group had a 47.2 percent higher incidence of stage I disease and a 24.5 percent lower incidence of stage IV disease; overall, stage I or II disease was 39.2 percent more common in the screened group [s2]. The test was shifting diagnoses earlier, which is the entire theory of screening.

Then came the outcome that matters. In total, 1,206 women died of ovarian or tubal cancer: 296 in the multimodal group, 291 in the ultrasound group and 619 in the no-screening group — a death rate of 0.6 percent in every group [s2]. There was no significant reduction in deaths from screening, in either the multimodal arm (p=0.58) or the ultrasound arm (p=0.36) [s2]. The authors' interpretation is unusually blunt: the reduction in late-stage disease was not sufficient to translate into lives saved, and given that screening did not significantly reduce deaths, general population screening cannot be recommended [s2].

Why catching it earlier was not enough

The gap between a stage shift and a mortality benefit is the crux. Finding more early-stage cancers only saves lives if those particular cancers would otherwise have killed the patient and are curable when found early. If a screen picks up slow-growing cancers that would never have proved fatal, or finds aggressive ones a little sooner without changing their course, the stage numbers improve while the death toll does not. UKCTOCS is the textbook demonstration of why cancer mortality, not stage or detection rate, has to be the primary outcome in a screening trial [s2].

The US Preventive Services Task Force reached the corresponding recommendation. It found adequate evidence that screening does not reduce ovarian cancer mortality, and adequate evidence that the harms — chiefly unnecessary surgery in women who turn out not to have cancer — are at least moderate and can be substantial [s1]. Concluding that the harms outweigh the benefit, it recommends against screening asymptomatic women, its grade D [s1].

The one important exception

That recommendation applies specifically to asymptomatic women who are not known to carry a high-risk hereditary cancer syndrome [s1]. Women with an inherited predisposition — the group that includes BRCA mutation carriers — are explicitly outside its scope, and their management is a specialist matter involving risk-reducing options that have nothing to do with population screening. For them, the evidence and the decisions are different.

What this means for a reader

There is no ovarian-cancer screening test to request, because the best evidence shows that screening the general population does not save lives while exposing healthy women to the risks of unnecessary surgery [s1][s2]. Ovarian cancer's symptoms are real but nonspecific, which is part of why early detection is so hard; a woman with persistent, unexplained abdominal or pelvic symptoms should be assessed by a clinician rather than reassured by the absence of a screening programme. And a woman with a strong family history of ovarian or breast cancer is in a separate category, for whom inherited-risk assessment — not screening — is the relevant conversation.

Sources

• [s1] Screening for Ovarian Cancer: USPSTF Recommendation Statement, JAMA, 2018-02-13
• [s2] Ovarian cancer population screening and mortality after long-term follow-up (UKCTOCS), The Lancet, 2021-05-12