Article Text
Abstract
In 2016, a French independent review proposed that the Organized Breast Cancer Screening program should be
stopped
or
stopped as existing today and re–organized after a complete revision.
The first proposition was immediately dismissed by the French authorities. At first sight, the second recommendation seems to be fulfilled by the MyPeBS randomized clinical trial, to be launched in 7 European countries in 2019. Its promoters presented it as seeking to know whether screening tests based on ‘individual risks of cancer are as efficient or better than standard screening tests’. They already announced that its results will serve as a basis to organize a new screening in France (and possibly in other countries).
MyPeBS will compare two groups.
Women from the ‘standard group’ will be screened according to the usual screening program in their geographical area: mammography screening every 1 to 3 years, from the age of 45 or 50. Following a complex process, women from the ‘individual risk group’ will be divided into four risk subgroups. Low-risk will lead to no screening; moderate risk to mammography screening every two years; high-risk to annual mammography and very high-risk to an additional MRI scan every year. After four years, every woman will undergo mammography screening.
This design is quite complex. In the ‘individual risk group’, mammography screening will begin at the age of 40, instead of 45 or 50 in the standard screening group. Screening programs organized in different countries will be, inaccurately, considered as identical. Paradoxically, in the individual risk group, young women with a low or moderate risk factors should undergo more mammography screenings than in the standard screening group. Conversely, some women over 50 will undergo fewer mammography screenings, and others, more. MyPeBS poses several problems: absence of non-screened group; risk evaluation methods not validated; heterogeneous groups; insufficient information given to participants.
The primary outcome is the incidence of advanced cancers. This doesn’t seem a bad choice: advanced cancer incidence is linearly linked to breast cancer mortality. But the comparison methods are flawed. MyPeBS uses a non-inferiority design, with a 25% threshold. This means that if advanced cancer incidence (and breast cancer mortality) rises by less than 25%, the ‘individual risk screening’ will be considered ‘non-inferior’ to the standard screening. Now, for the Cochrane meta-analysis (and others), mammography screening seems to lower breast cancer mortality by 20%. So, even if the ‘individual risk screening’ does not perform better than no screening at all, it will be considered as non-inferior to the standard screening. Whatever the results, ‘individual risk screening’ will be considered a success.
When the results of a trial are known in advance, you cannot call it scientific trial: you have to call it marketing. And what does MyPeBS market? Screening from the age 40 (whatever the results of the AgeX trial) and extensive use of MRI scan for ‘high risk women’. In other words: overdiagnosis and overtreatment.
Beware of rigged trials!