Israa Imam: Rectal Cancer Treatment and Response Prediction

Abstract

Rectal cancer is one of the most common cancers in Sweden with 2000 new cases annually. Outcomes have improved over recent decades due to better staging, using magnetic resonance imaging (MRI), refined surgical techniques and optimised use of radiotherapy (RT) or chemoradiotherapy (CRT). More recently, total neoadjuvant therapy (TNT), RT/CRT and chemotherapy before surgery, has emerged as a superior treatment for locally advanced rectal cancers (LARC). Rectal cancers are stratified into risk groups for systemic and loco-regional recurrence (LRR), requiring different treatments. Response to neoadjuvant therapy varies considerably between tumours, and despite extensive research, reliable predictors of response remain to be identified. The LARCT-US study treated LARC patients with TNT; 5x5 Gy + 4 CAPOX, an abbreviated RAPIDO-schedule. The aim of this thesis was to explore factors of importance for treatment choice and outcome in rectal cancer patients.

Paper I assessed the neoadjuvant rectal (NAR) score, a short-term surrogate endpoint to compare different regimens in clinical trials. The NAR-score discriminated between different treatments having different cytotoxic effects and applies irrespective of therapy given.

Paper II describes treatment selection over time in an unselected patient cohort from two adjacent Swedish regions. MRI-based risk grouping most strongly influenced treatment choice, with fewer patients receiving RT over time. Variations in MRI-interpretation and a stronger desire to decrease RT may explain regional differences. Accuracy of MRI was poor and requires improvement.

Paper III investigated long-term outcome and recurrence predictors in LARCT-US. The TNT schedule achieved excellent systemic control (25%) and few LRR (6%). The low LRR risk compared with RAPIDO may reflect more adequate distal resection margins practiced at Swedish centres. Besides treatment response, multiple high risk-criteria, tumour deposits, low tumour level and a sub-optimal resection plane were associated with recurrence risk.

Paper IV evaluates quality of life (QoL) and late toxicity following LARCT-US. QoL was comparable to that in the RAPIDO TNT-arm. Major bowel dysfunction was less frequent (45% vs 59%). Grade 3+ late toxicity occurred in 12% at three years and 8% at five years. Overall, the abbreviated TNT schedule achieves favourable oncological outcomes with acceptable QoL and limited late toxicity.