Desley van Zoggel

PET/CT response after induction chemotherapy 127 CHAPTER 7 differences between patients with an anastomosis or abdominoperineal surgery for their primary tumour (patients with local treatment or watch andwait for their primary were excluded in this paper) could be demonstrated. Figure 3 depicts the correlation between the PET/CT and histopathological responses. In 24 patients, PET/CT showed a CMR, which accurately predicted the major histopathological response in 15 patients (63 percent). However, 7/24 patients (29 percent) had a partial histopathological response and 2/24 patients (8 percent) had a poorhistopathological response. In54patients, PET/CTshowedaPMR,whichaccurately predicted a partial histopathological response in 23/54 patients (43 percent). However, 16/54 patients (30 percent) had a major histopathological response and 15/54 patients (28 percent) had a poor histopathological response. In 28 patients, PET/CT showed NR, which accurately predicted a poor histopathological response in 10 patients (36 percent). However, 5/28 patients (18 percent) had a major histopathological response and 13/28 patients (46 percent) had a partial histopathological response. Therewas a significant correlationbetween thePET/CT response andhistopathological outcomes (P = 0.010). The sensitivity to detect major histopathological response was 42 percent, and the positive predictive value was 63 percent. The sensitivity to detect a partial histopathological responsewas 53 percent, and the positive predictive valuewas 43 percent. The sensitivity to predict a poor histopathological responsewas 37 percent, and the positive predictive value was 36 percent. Overall, PET/CT predicted the histopathological response correctly in 48 patients (45 percent). PET/CT underestimated the actual histopathological response in 34 patients (32 percent), and PET/CT overestimated the actual histopathological response in 24 patients (23 percent). Table 1 shows potentially confounding clinical parameters that may affect the reliability of the PET/CT response category in predicting histopathological outcomes. The table shows that, for both male and female patients, a significant correlation persists. However, this significance was lost in several subsets. The presence of diabetes had a negative impact, as well as the presence of inflammation. The size of the largest post- treatment tumour diameter did not impact the predictability.

RkJQdWJsaXNoZXIy MTk4NDMw