79 Sense and nonsense of yT-staging on MRI after chemoradiotherapy in rectal cancer 5 yT4 stage The study dataset included 4 cases with ypT4 disease. As shown in Table 1, most of these cases were understaged on MRI (mainly as ymrT3). One case with peritoneal invasion (ypT4a) was correctly classified as ymrT4 by only 12/22 (55%) readers; three cases with ypT4b invasion (vagina, bladder and external sphincter invasion) were correctly classified as ymrT4 on MRI by only 7 (32%), 11 (50%) and 4 (18%) of the 22 readers, respectively. An example is shown in Figure 3. Conversely, when readers assigned a ymrT4 stage, this was not confirmed at histology in on average 66% of the cases (80% for ymrT4a; 58% for ymrT4b), indicating significant overstaging. Figure 3 Example of a ypT4b case with histology confirmed invasion of the vagina after CRT that was understaged by 3/5 of the expert and 12/17 of the non-expert readers. The primary staging MRI (A) shows a solid tumour that invades the dorsal vaginal wall (white arrows) and left part of the puborectalis muscle (black arrowhead). On the restaging MRI (B), all readers recognized the presence of residual tumour (*) but there was substantial disagreement on whether the surrounding fibrosis involving the vagina and puborectalis was still suspicious for persistent tumour infiltration. Discussion This study shows that accuracy for yT-staging on restaging MRI after CRT is greatly reduced in patients with a more favourable tumour response and increasing levels of fibrosis. Accuracy for ymrT-staging is significantly higher in patients who show predominant tumour at histology (pTRG4-5). In this subgroup, the overall ymrT-/ypT-stage concordance was 55% for non-expert and 67% for expert readers, with overstaging rates of only ≤15%. In the good responders who showed predominant fibrosis after CRT (pTRG1-3), ymrT-stage accuracy was low with a ymrT-/ypT-stage concordance of only 40-43% and up to 44% overstaging.
RkJQdWJsaXNoZXIy MjY0ODMw