G.J. den Heeten Nijmegen/Amsterdam NL

Discussions about subgroups in organized breast cancer screening populations carry the risk that the well-known discussions about benefits and harms of screening which are being calculated on the base of the whole population are extrapolated to this particular subgroup. This applies especially for age limits for invitation. Most programs in Europe stop inviting after the age of 70 except for the Netherlands. After the start of the population based screening program in 1990 the invited group was extended in 1998 with the cohort of 70-74, meaning more than 15 years of follow up. It is plausible that the life expectancy of this group will particularly affect one of the factors in the balance between harms and benefits, i.e.  overdiagnosis and overtreatment. In the Netherlands the group of women at the age of 70-75 has a mean life expectancy of 15 years (17 – 13 years). Modeling studies on the basis of invasive tumors show that only the low 5 percentile fraction of tumors take more than 6.3 years to reach the size of 15 mm (DCIS excluded). The often heard assumption that growth rates of tumors in screening programs within the target population, substantially decline with age has never been proved, and even contradicted. The decision to continue screening over 70 years of age should first be based on the question whether the disease is an important health problem.  It is a well-known fact that the higher the age (up to 84) the higher the incidence. During the yearly evaluations of the Dutch screening program we have seen the strongest decline in mortality in the older age group of 75-84. This is very relevant because in this particular age group one should expect the effects of screening at the age of 70-74. Computer simulations showed that the impact of breast cancer screening in this age group is twice that of adjuvant treatment. Another aspect brought up against breast cancer screening in the elderly, is comorbidity. But there is no sudden appearance of comorbidity in the older ages. Even opposite: at the age of 50 half of the population in Scotland had at least one morbidity, and at the age of 65 most are multimorbid. However in absolute terms most people with multi morbidity are under the age of 65. In the US 70% of the population aged 74 has no co morbidity that influences the life expectancy. There are also worries about intensive therapies like chemotherapy. According to the guidelines in the Netherlands women over 70 year will not receive chemotherapy and the treatment will consist of breast conserving treatment with most of the time radiotherapy and if relevant hormonal therapy. In analyses in literature there is a fair agreement that women with a life expectancy of less than 7-10 years do not benefit anymore from screening mammography. This is on an individual base a hard to use criterion but useful in population based decision making.

Conclusion: According to the results of the screening in the Netherlands with a relatively high and still increasing life expectancy of 87 years (of those who reach 70), screening the 70 – 74 cohort is one of the most successful interventions regarding prevented death. It is obvious that overdiagnosis will become more and more prominent at higher ages, but this is most relevant for DCIS. Because of the ageing of the European population, the contingent of healthy elderly women will further increase in the upcoming decades with a further increase of life expectancy. We expect that a substantial part of this group will actively choose to continue participating in screening. And if you are over 70 and healthy (as most of them are), it is the right decision.