Inter and intra-fractional stability of rectal gas in cervical cancer patients

Abstract

Purpose or Objective Due to the Electron Return Effect during Magnetic Resonance guided Radiotherapy (MRgRT), rectal gas during pelvic treatments can potentially result in hotspots in the rectal wall. Determining the clinical impact of this effect on rectal toxicity requires estimation of inter and intrafractional stability of rectal gas. We investigate local inter and intra-fractional changes of rectal gas in cervical cancer patients. Material and Methods Six cervical cancer patients underwent four scanning sessions (~20 minutes each). Within each session, T2-w MRIs were taken at seven time-points (TP, mostly ~3 minutes apart). The rectum on each image was delineated by a clinical oncologist. Slices where rectal delineations were present across all TPs in a scan session were selected for analysis. The length of the rectum was normalised, going from 0 (most cranial) to 1 (most caudal). The area of gas within the rectum delineations was identified on each slice using thresholding and recorded (ATPX) {X = 1,…,7} (fig. 1). The (ATPX) for averaged over all TPs was used to calculate the local inter-fractional local group Mean (M), systematic (Σ) and random (σ) error. To determine the intra-fractional changes, the difference in gas area between TP2, TP3, … TP7 vs. TP1 was found on each slice (ΔATPX) and used to calculate the local M, Σ, σ and σf, but now for intra-fractional motion. Results Figure 2 shows the local statistics (M, Σ, σ (and σf)) for inter (panel A) and intra (panel B) fractional changes across all patients. Both inter and intra-fractional changes are larger in the cranial aspect of the rectum, and gas appears more stable at the caudal end. Further, interfractional changes (~2cm2/slice) are larger than intrafractional changes (~1cm2/slice). The intra-fractional changes (~1cm2/slice) are small compared with the absolute area of gas in each slice, which is up to 10cm2/slice in the patient in figure 1. This indicates that gas remains relatively stable during a 20 minute period. Conclusion We are the first to statistically evaluate stability of rectal gas in cervical patients. Inter-fractional gas changes are larger than intra-fractional gas changes. I.e. gas is appears more stable at the caudal end. Further, interfractional changes (~2cm2/slice) are larger than intrafractional changes (~1cm2/slice). The intra-fractional changes (~1cm2/slice) are small compared with the absolute area of gas in each slice, which is up to 10cm2/slice in the patient in figure 1. This indicates that gas remains relatively stable during a 20 minute period. Changes are more likely to occur at the cranial aspect of the rectum than the caudal end. This work will feed into assessment of the total dosimetrical effect of rectal gas during MRgRT. However, in practical MRgRT the relevant time period may exceed our measurement period.