Desley van Zoggel

Lateral lymph node recurrence and induction chemotherapy 59 CHAPTER 4 Neoadjuvant treatment and intraoperative radiotherapy Basic neoadjuvant chemoradiotherapy (nCRT) and intraoperative radiotherapy (IORT) treatment regimens in the Catharina Hospital for LRRC have been previously described.12 From 2007, induction chemotherapy (ICT), consisting of combination chemotherapy with oxaliplatin and fluorouracil (or capecitabine), was administered to a growing number of patients prior to the start of nCRT. This was mainly because of the poor results of the basic treatment in these patients and the remarkable results that were observed in a fewpatients with irresectable tumours, that became resectable after ICT. After three cycles of CAPOX or four cycles of FOLFOX, tumour response was evaluated by MRI and positron emission tomography–computed tomography (PET–CT) scan. There was no fixed protocol yet regarding the further administration of chemotherapy, but the refer- ring hospitals were advised that if therewas a response, a few more courses were to be administered followed by the nCRT, or chemotherapy could be administered during thewaiting period after nCRT. If no responsewas notified onMRI and/or PET–CT, nCRTwas started immediately, with no further administration of chemotherapy, except for capecitabine. The waiting period between the last radiotherapy day and surgery was between 8 and 10 weeks. Statistical analysis Statistical analysiswasperformedusingSPSSversion23 forWindows (IBMCorporation, Armonk, NY, USA). T tests and Chi square tests were used to compare individual variables. Local re-recurrence (LRR) rate, distant metastases (DM) rate, cancer-specific survival (CSS) and overall survival (OS) were estimated using the Kaplan–Meier method. CSSwas defined as the time between rectal cancer surgery anddeath causedby rectal cancer, and OS was defined as the time between surgery and death by any cause. Differences were assessed using the log-rank test. P-values were two-sided and were considered statistically significant at ≤0.05. For determinationof risk factors, univariate analyses were performed by analysing the effect of the covariates in a univariate Cox regression. Covariates with trend-significant effects (P < 0.10) were then selected for multivariate analysis, using stepwise Cox proportional hazards regression modelling; a two-sided P-value of < 0.05 was considered significant.