Session Item

The various rectal endoluminal brachytherapy techniques all have steep, but different, depth dose gradients, resulting in very inhomogeneous dose delivery. In rectal Contact X-ray Therapy (CXT) doses are typically prescribed and reported to the applicator surface and not to the GTV, CTV and OARs, the latter of which are crucial to understand tumor response and toxicity rates. To quantify this problem, we performed a dose modeling study using a fixed prescription dose to the surface of the applicator and varying tumor response scenarios (fig 1).

We created 3D-volume models of tumors and regression patterns based on endorectal ultrasound images of 5 patients with (y)cT1N0 rectal cancers. A “thick” and a “thin” tumor model were developed. At the start of treatment, the “thick” tumor model had a thickness of 1 cm, whereas the “thin” model had a thickness of 0.5 cm. The initial circular surface diameter for both tumor models was 2 cm. Treatment was modeled for our Maastricht HDR CXT applicator (Bellezzo M. et al. 2020), with a fixed-dose prescription to the applicator surface (equivalent to 3 fractions of 30 Gy CXT). Multiple different response scenarios were simulated: no or good response between fractions, concentric and non-concentric shrinkage, with or without recovery of rectal wall layers (Fig 1). Simulations were performed using MATLAB v2018b by creating a mask of the GTV, CTV, and different wall layers. Tissue composition was considered to be equivalent to water. The center of the surface of the tumor was positioned in the center of the high-dose region (Fig 2, tumor in gray, 1 of 5 source channels in red, dotted lines are isodose lines) for all simulations. We used a GTV to CTV margin of 5.5 mm (Verrijssen AS et al, 2020)

A fixed prescribed dose to the surface of the applicator resulted in a broad range of cumulative doses delivered to the GTV, CTV and healthy intestinal wall. Depending on the tumor response scenario the equivalent dose in fractions of 2 Gy (EQD2) (α/β 10) received by 90% of the GTV varied between 63 (thick tumor, no response) and 231 Gy (thin tumor, good response). The highest D2cc of the normal bowel wall was seen in the thin tumors showing good response and concentric shrinkage (EQD2 D2cc 165 Gy), and the lowest was seen in thick tumors without shrinkage (EQD2 D2cc 96.5 Gy). Differences in GTV/CTV and bowel wall dose were predominantly determined by initial tumor thickness and magnitude of tumor thickness regression during treatment.

We have demonstrated that doses prescribed at the surface are not representative for the dose received by the tumor and the bowel wall over a multi-fraction contact therapy treatment course. The results stress the relevance of dose reporting and prescription to GTV/CTV volumes and OARs over repeated imaging in order to gain insight between delivered dose, local control and toxicity to optimize treatment protocols.