Breast-conserving therapy versus mastectomy

Abstract

There is an ongoing debate regarding the use of randomised controlled trials (RCTs) versus observational studies when investigating treatment effects in clinical practice [1]. This holds especially true for the comparison of breast-conserving therapy (BCT) and mastectomy (MAST), which gained much attention since the publication of our observational study in Lancet Oncology [2]. RCTs are highly appreciated as they are close to generate perfectly unbiased treatment comparison estimates. Treatment groups in a RCT are expected to be exchangeable; even when switching the treatment between the compared groups, results will be similar and are solely the effect of the treatment under study. Clinical decisions are largely based on this type of evidence. But is this always the best evidence? Is it always feasible or ethical? In the current era of personalised medicine and 'big data', clinical interpretation of an abundance of data (clinical reasoning) is becoming more and more crucial. It integrates all available and relevant information that may contribute to the best clinical decision-making for individual patients. This generally starts with existing guidelines, completed by evidence extracted from observational studies and clinicians' experiences [3]. Importantly, the patient's preference plays an important role in (shared) decision-making. In general, it is difficult to translate the overall results of a RCT in the response of an individual patient to the investigated treatment. Even for patients with identical characteristics to those in the trial population, the overall treatment effect observed in RCTs would only apply if the probability of treatment benefit and detriment was equally distributed in every individual participant [3]. Often, evidence forming the basis of treatment guidelines are based on RCTs conducted a long time ago, while observational studies include a more recently diagnosed population. For BCT and MAST, the trials were all conducted in the eighties. Another important discrepancy between the RCT populations and the real-world population is the increasing share of elderly breast cancer patients in the latter. This is not only due to the ageing population, but also to early detection of breast cancer in the national screening program (which upper age limit is 75 years in the Netherlands), leading to a higher incidence in the elderly. Furthermore, diagnostic and surgical procedures as well as local and systemic therapies improved considerably. Moreover, increasing knowledge about the biological features of breast tumours led to the introduction of more advanced tumour-directed therapies. The combination of these improvements are very likely to …