Tomosynthesis screening: about a quarter fewer interval cancers, but not significant

Design

Multicentre, open-label, randomised superiority trial

Population

99,689 women aged 50 to 70; about 49,000 screen-negative per arm

Primary outcome

Invasive interval cancer rate over 24 months

Effect

14.2 vs 18.6 per 10,000; adjusted OR 0.76 (95% CI 0.55 to 1.06)

TOSYMA randomised 99,689 women aged 50 to 70 at 17 German screening sites to digital breast tomosynthesis plus synthesised two-dimensional mammography or to digital mammography. The trial had already reported a higher invasive cancer detection rate with tomosynthesis. This second primary endpoint asked whether that translated into fewer invasive interval cancers in the 24 months after a negative screen.

Among about 49,000 screen-negative women in each arm, interval cancers occurred in 70 with tomosynthesis (14.2 per 10,000) and 92 with digital mammography (18.6 per 10,000), an adjusted OR of 0.76 (95% CI 0.55 to 1.06). That missed significance. In the first 12 months, the difference was significant (4.5 vs 6.3 per 10,000; OR 0.71, 0.52 to 0.97).

Fewer interval cancers is the outcome that matters most, because it suggests the extra detections are cancers that would otherwise have surfaced clinically, not only overdiagnosis. The direction here supports tomosynthesis, but the primary endpoint does not prove it.

For units deciding on tomosynthesis, this adds weight on the benefit side without closing the question. In India, where most screening is opportunistic and equipment varies, the practical value lies more in diagnostic workup than in population screening for now.

• Tomosynthesis plus synthesised mammography detects more cancers and may reduce interval cancers.
• The 24-month interval cancer reduction was not statistically significant.
• The 12-month reduction, a secondary outcome, was significant.
• Weigh reading time, cost and dose against the likely benefit when planning a switch.
• Report interval cancers as part of your unit's own audit.

Why it matters

It is the first large randomised signal that tomosynthesis may cut cancers that surface between screens.

Don't overread it

The primary endpoint was not significant; this supports but does not prove a reduction in interval cancers.

The statistics, in plain English

An odds ratio of 0.76 suggests about a quarter fewer interval cancers, but the interval (0.55 to 1.06) crosses 1, so no difference cannot be ruled out. The 12-month result was a secondary outcome, which is weaker evidence than the primary.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.