Abstract 432P
Background
Preoperative radiotherapy (RT) is frequently used in rectal cancer to improve results, but when it is best used alone or with chemotherapy (CT) is subject to discussions and guidelines differ. To better understand selection mechanisms in a changing landscape, we analysed chosen initial treatment in all patients diagnosed during an eleven-years period in two Swedish regions, one with and one without a RT-department.
Methods
All patients with rectal cancer in Uppsala (with a RT department) and Dalarna, between 2010-2020 were prospectively identified. Information on staging and initial treatment (direct surgery, preoperative short-course RT or combinations of RT and CT) in the Swedish Colorectal Cancer Registry were used. Information from the staging magnetic resonance imaging (MRI) permitted division into three risk groups, early/good, intermediate/bad and locally advanced/ugly according to the national guidelines for treatment recommendations during the period. Cohen’s kappa was used to test congruence between clinical and pathologic stage, and logistic regression to explore associations between baseline characteristics and treatment choice.
Results
1150 patients without synchronous metastases (M0) were analysed (Uppsala 542, Dalarna 608). Patients from Dalarna were older, had less advanced tumours and received presurgical treatment less often (63% vs 52%, p<0.001). All MRI-characteristics (T-/N-stage, MRF, EMVI, tumour level) were important for treatment choice but not age and sex. Age was, however, important for if CT was added. Independent associations were seen for all MRI-characteristics, but not for region. The treatment choices followed the recommendations in the good and ugly groups but varied in the bad group. The correlation between clinical and pathological T-stage was fair/moderate and poor for N-stage.
Conclusions
The risk grouping based on MRI-characteristics had the most influence on treatment choice. Since the risk grouping was modified, due to a wish to diminish the proportion pre-treated, fewer patients were irradiated with time. Regional differences in interpretation of the MRI characteristics probably explained that fewer patients were pre-treated in Dalarna. MRI-staging is far from optimal.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Cancerfonden (Swedish Cancer Society).
Disclosure
All authors have declared no conflicts of interest.