Reducing dose to rectum by placement of a rectum-emptying tube in cervical cancer patients treated with brachytherapy

PURPOSE Only scarce data are available on the possibility to reduce rectal dose by controlling rectum filling before HDR (high dose rate) IGABT (image-guided adaptive brachytherapy) in LACC (locally advanced cervical carcinoma) patients. We compared dosimetric outcomes before and after the evacuation of gasses using a rectum emptying tube. METHODS AND MATERIAL Sixty CT (computed tomography) scans from 30 consecutive patients with cancer of the cervix undergoing HDR IGABT after EBRT were reviewed. Patients who underwent at least one gas evacuation were included in the analysis. The three-dimensional dosimetric data of the dosimetric plan performed before and after gas evacuation were compared. Primary endpoint was the difference between D2cc of the rectum before and after the procedure. Expected probability of grade 2–4 overall rectum morbidity was assessed using a probit model from the prospective multicenter EMBRACE study. RESULTS Thirty five (58.3%) CT scans from 23 patients (76.7%) requiring gas evacuation were analysed. The mean rectum volume, before and after gas evacuation, was 123.1 cc (sd, ± 60.4) and 66.4 cc (sd, ± 34.8), respectively. For each patient, the volume of the rectum after gas evacuation was lower than before. No major complication occurred during and after the procedure. After gas evacuation, a significant reduction in rectal dose per fraction was observed, on average -4.3 Gy (-38.4%, p < 0.001) for D0.1cc and -1.9 Gy (-30.6%, p < 0.001) for D2cc. Estimated mean probability to develop a grade 2–4 rectum morbidity was significantly lower after gas evacuation, 6.9% (sd,± 1.94) versus 9.5% (sd,± 3.17), p < 0.001. CONCLUSION Gas evacuation using a rectal emptying tube in selected LACC patients treated with HDR BT after chemoradiotherapy, allowed a substantial reduction in the dose to the rectum. Such procedure could be of particular interest when a dose escalation strategy is being considered.