Milea Timbergen

176 The selection of patients suitable for the AS approach remains challenging. The results of this systematic review suggest that AS is mainly described as a treatment for tumours localised in the extremity/girdles and in the trunk. This might be due to the predilection sites of DTF tumours to these locations 43 , or due to a selection upfront because of the higher risk of recurrence after surgery for these groups 12 . Based on the current systematic review, drawing a single conclusion with regard to tumour sites and the success of AS remains challenging. This is mainly due to the inclusion of studies with homogeneous cohorts in terms of tumour site (e.g., mesenteric, or breast), or a preselection of patients upfront (e.g., inoperable tumours due to localisation adjacent to vital structures [e.g., nerves, blood vessels]). Furthermore, the exact tumour site was not specified in a large number of patients. About one-third of the patients needed a shift to an ‘active’ form of treatment. Although no uniform results could be drawn from the current studies, several studies reported that larger tumours were more likely to shift 10, 36 , whilst age, sex and pregnancy before the development of DTF were not associated with this shift 10, 31, 36 . Colombo et al. reported that the sex, tumour site and tumour size did not predict progression and/or shift to change in treatment; the non-surgical group (n = 106) also contained patients receiving medical treatments (n = 4) 9 . Few studies described β-catenin mutation of the included cohort, and none of these studies analysed the influence of these mutations on the success or failure of the AS approach 6, 21 . The same applies for FAP-related DTF tumours. The variable results from these retrospective studies highlight the need for the identification of predictive factors for progression and changes in treatment strategies. In the current study, progression was often reported within two years after diagnosis 14 , however the length of follow-up of the included studies varied highly. Few studies reported the median follow-up duration of the AS subgroup, and time to intervention was often lacking. The minimal available information about the type and frequency of follow-up during AS underlines the need for standardisation of the AS approach. This includes defining a follow-up schedule with the use of MRI or CT, depending on the tumour site. As few studies reported progression after stabilisation, a maximum AS term should be discussed with the patient. The major limitation of the current study is the inclusion of retrospective, small sample- sized studies, which often evaluate several treatment regimens, with various follow-up schedules and limited information about disease outcomes, or reasons for shifting to ‘active’ treatment. Only part of the studies used and reported disease response based on RECIST 18 . Some included studies selected patients for the AS approach based on the fact that 6