August 12, 2020

expert reaction to study looking at breast screening women in their 40s and cancer deaths

A study, published in the Lancet Oncology, looked at breast screening women in their 40s and cancer deaths.

Dr Imogen Locke, Consultant Clinical Oncologist at The Royal Marsden NHS Foundation Trust said:

“This important study suggests that offering mammographic screening to women from a younger age results in breast cancers being detected at an earlier and more curable stage.  Diagnosing breast cancers before they cause symptoms results in less women dying from breast cancer and may also mean that women require less aggressive cancer treatment.  It is critical that we detect cancer as early as possible to give us the best chance of successfully treating patients and saving their lives.  We are continuing to see the importance of early diagnosis and screening across breast and other types of cancer too including lung cancer.

“The strengths of this study include the fact it’s a randomised trial conducted in a significant proportion of breast screening units and involving a large number of patients with long (more than 20 yr) median follow up.  One weakness is that since the patients were recruited, there have been significant advances in the sensitivity of imaging technology to diagnose cancer and also improvements in the treatment of cancer, so the impact of screening on reducing breast cancer mortality could possibly have been under or over estimated.”

Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“This research report covers long-term follow-up of a large, well-conducted trial of annual mammography screening for breast cancer, involving over 160,000 women.  The statistical analysis is sound, and it’s good to see the follow-up continuing for such a long time (23 years) after the start of the original trial, so that any very long-term effects could be picked up.

“There’s a marked contrast between the explicit top line of the press release, “Breast screening women in their forties saves lives”, and the much more cautious conclusion in the actual research report, “Reducing the lower age limit for screening from 50 to 40 years could potentially reduce breast cancer mortality.”  So are lives saved by breast screening women in their forties, or is this just something with potential?  If lives are saved, are there any reasons why the lower age limit should not be changed?  The trouble is that the answers to both these questions are not entirely clear.  I’d support the researchers’ recommendation that we need further evaluation of breast screening in women younger than 50, using more modern screening and treatment approaches than were available when this trial began more than two decades ago.

“I think it’s fair to say that, like earlier reports on this trial, there’s evidence that lives were saved by the sort of screening that was done, in the circumstances that existed when the women involved were in their forties, which is now quite a long time ago.  But, as the researchers are careful to point out in their research paper, things have changed since then.  Screening methods have improved, as have the treatments available to women diagnosed with breast cancer.  The researchers state that, because of changes to treatments, “there might be less scope for screening to reduce mortality in our current era.”  Against this, they also make clear that there is still evidence (from other studies) of an advantage in survival from early diagnosis and treatment, but it is not certain how these two possibilities would balance out.

“It’s also the case that the number of deaths saved by the earlier breast screening in this trial was not very large, and subject to quite a lot of statistical uncertainty.  This follow-up indicates that the gain in survival was concentrated in the first ten years after the women began to be screened.  In those first ten years, out of every 10,000 women invited for screening, on average about 16 died of breast cancer, while in every 10,000 women in the control group who did not get the screening, on average 21 died.  These numbers indicate that lives were saved.  But they also indicate that death from breast cancer was pretty rare in women of that age.  Because breast cancer deaths in younger women are not common, the estimates of breast cancer death rates are not very precise, despite the fact that the trial involved 160,000 women.  Over the whole follow-up period so far, the difference in numbers of deaths between those who were screened in their forties and those who weren’t, is 6 deaths for every 10,000 women, but because of the statistical uncertainty, this figure could plausibly be larger, at 13 per 10,000, or in fact the data are also consistent with a very slightly higher death rate (1 death per 10,000 women) in those who had the screening.  But none of those numbers is very large, out of 10,000 women.  Allowing for the fact that not every woman invited for screening will actually attend the screening, the researchers estimate that 1,150 women would have to be screened in their forties to prevent one breast cancer death.  They also report that 11.5 years of life would be saved per thousand women invited for screening in their forties.  That’s not 11.5 years of life saved per woman screened – it’s 11.5 years in total across a thousand women, or about 4 days of life (on average) per woman.  (The 11.5 years figure again is subject to considerable statistical uncertainty – it could plausibly be as high as 22 years or as low as 1 year per thousand women invited, so between 8 days and a few hours of life saved per woman, on average.)

“Now these are averages, and in terms of years of life saved, they average out between the great majority of women whose lives are not saved by the early screening, and a small minority whose lives are saved at a relatively young age – and a life saved is a life saved.  So why isn’t mammography screening just extended to women in their forties as soon as possible?  That’s because there are other considerations.  I’ve already mentioned changes in screening and treatment methods, but there’s more.  The current leaflet sent to women who are invited for breast screening by the NHS points out that there are risks of several kinds from mammography.  The most important is from what’s called overdiagnosis – the screening will detect cancers that would otherwise never have been found and would never have become life-threatening.  But it’s not always possible to tell which cancers are in this category, so, having been found, they will often be treated, even though (if we had perfect knowledge) the treatment is unnecessary – and all treatments have risks and downsides.  The current NHS leaflet reports that, for every 1 woman whose life is saved by the current screening programme (starting at age 50), about 3 women are diagnosed with a cancer that would never threaten their lives.  These numbers might well not apply in the same way to screening in the 40s age group.  The researchers in this new report feel that the rates of overdiagnosis in their study were not high, and that most of the overdiagnosed cancers would be found anyway in screening after the age of 50.  But an earlier overdiagnosis does mean that a woman might have to live with the consequences of unnecessary treatment for longer.  It’s also the case, on the current screening programme, that most women whose mammogram shows something suspicious, so that they need further investigation, actually turn out not to have a cancer – currently that’s true for three out of every four women whose mammogram looks abnormal.  But these issues certainly aren’t reasons why screening should definitely never be offered to women in their forties – I’d say that they just illustrate the importance of giving good, clear information about such issues to women when they are invited for screening, as happens on the current programme, so that they can make their own decision whether or not to be screened.  And I continue to be grateful that, as a man, I don’t have to grapple with that decision for myself.

“One final issue, also mentioned by the researchers, is that they have not considered the cost-effectiveness of lowering the screening age.  Assuming that lowering the age, with current screening, diagnosis and treatment approaches, would still save lives – and that is not entirely certain yet – why should the cost matter?  Well, an extra pound spent on breast screening is very likely to be a pound less spent on something else in the NHS, and that something else might save more lives than lowering the breast screening age, particularly since the number of lives saved by earlier screening, according to this study, is not immense.  But this research does not throw any extra light on this aspect.”

Prof Shirley Hodgson, Professor of Cancer Genetics, St George’s, University of London, said:

“This is a large study of 160000 women aged 40-49 who were randomised to NHS Mammography screening 3 yearly from 50y, or annual screening mammograms from 40y.  The women in the screened group had a 25% reduction in mortality from breast cancer in the first 10y of follow-up, although the mortality from breast cancer after 10y was similar in both groups.  There was little evidence for over-diagnosis, so the clear implication from this study is that screening from 40y does appear to save lives from breast cancer, particularly early stage (1 and 2) breast cancer.  The findings indicate that 11.5 years of life were saved per 1000 women screened.  The fact that early screening involves annual screens which at the time of the study were less sensitive than those done nowadays, indicates that the sensitivity of screening may be greater now, but possibly this could result in more false positive diagnoses.  There is clear evidence from this study that mammography screening from 40y age has the potential to save lives from breast cancer.”

‘Effect of mammographic screening from age 40 years on breast cancer mortality (UK Age trial): final results of a randomised, controlled trial’ by Stephen W Duffy et al. was published in the Lancet Oncology at 23:30 UK time on Wednesday 12 August 2020.

DOI: 10.1016/S1470-2045(20)30398-3

Declared interests

Prof Kevin McConway: “I am a member of the SMC Advisory Committee, but my quote above is in my capacity as a professional statistician.”

Prof Shirley Hodgson: “I have no conflicts of interest.”

None others received