Dose and volume effects of gastrointestinal toxicity during neoadjuvant IMRT for rectal cancer

Purpose/Objective: Preoperative chemoradiotherapy for rectal cancer reduces the risk of local recurrence, but also induces gastrointestinal (GI) toxicity. Acute GI toxicity might, in addition to the discomfort caused to patients, threaten treatment compliance. Modern radiotherapy techniques, such as intensity modulated radiotherapy (IMRT), can reduce dose delivered to the bowel, potentially limiting acute GI toxicity. Previous studies have reported dose-volume relationships for acute GI toxicity for patients treated with 3D conformal techniques (3D-CRT), but few have examined patients treated with IMRT. Materials and Methods: We explored dose metrics correlating with acute diarrhea and chemotherapy compliance for a single-institution cohort of rectal cancer patients (n=115) treated with IMRT. Acute diarrhea during treatment was scored prospectively by trained RT nurses (CTCAE v3.0). The highest toxicity score was used for analysis. Treatment charts were retrospectively reviewed for chemotherapy compliance; defined as no treatment interruptions, discontinuations, or dose reductions. The entire intestinal cavity containing small bowel loops up to the upper level of L5 was delineated by experienced radiation oncologists. IMRT treatment planning used a 5- or 7-field technique with 6 MV photon beams. Most patients were given 60 Gy in 30 fractions to the tumour and 50 Gy in 30 fractions to lymph node targets, and many received an additional tumour boost (either external beam or brachytherapy). Ordinal and binary logistic modelling was used to correlate absolute volumes of intestinal cavity receiving x Gy or above (Vx) to graded toxicity scores and chemotherapy treatment compliance. Optimal value of the dose cut-off x was chosen by maximum likelihood technique accounting for scanning the cut-point. Clinical factors (see Table 1) were included one by one in both models to examine their association with toxicity. Significance levels were estimated using likelihood ratio tests. Results: V32Gy correlated with acute diarrhea (p=0.0001), see Fig 1a; females had an increased risk of toxicity (OR=2.13, 95% CI 1.03-4.42, p=0.04). V46Gy correlated with chemotherapy compliance (p=0.005, 100 patients receiving concurrent chemotherapy), see Fig 1b. Women had lower treatment compliance (OR=2.41, 1.09-5.34, p=0.03), as had patients with low BMI (OR=0.89, 0.80-0.99, for 1 point increase in BMI, p=0.03) and patients with diabetes (OR=7.29, 1.21-43.8, p=0.03). Age, brachytherapy boost, prior abdominal surgery, smoking history, or domestic status had no influence on any of the two endpoints, nor had concurrent chemotherapy on the risk of acute diarrhea. Conclusions: We found that dose to the intestinal cavity (V32Gy and V46Gy) was associated with acute diarrhea and chemotherapy treatment compliance in patients treated with IMRT for primary rectal cancer. The results are not in direct agreement with results from patient cohorts treated with 3DCRT, where V15Gy has consistently been reported as an optimal dose cut-off.