PET-CT imaging in patients with locally advanced rectal cancer before and after the neoadjuvant chemo radiotherapy: staging, evaluation of treatment response with high resolution pelvic MRI, and correlation with Ki-67 and GLUT1

Aim: We investigated the correlation between the SUV parameters, pelvic MRI and histopathology in patients with locally advanced rectal cancer (LARC) with PET-CT before and after the neoadjuvant chemo radiotherapy (CRT). We also looked for the correlation between PET-CT parameters, Ki-67 and GLUT1.Methods: 29 patients diagnosed as LARC with biopsy between November 2009 and February 2012 were included in our study. Patients were referred to Department of Nuclear Medicine for staging and evaluation of treatment response. SUV parameters obtained from PET-BT for early and delayed images before (BT) and after (AT) treatment (early and delayed SUVmax-average, LBM SUVmax-average, tumor/liver (T/L) SUV and retention ratios, ?SUVmax) were calculated. Pathological staging, pathological complete response, tumor regression grade (TRG) and MRI (BT and AT) results were compared in subgroups. Immunohistochemical staining with Ki-67 and GLUT1 were performed on biopsy and surgery specimens. Correlation between staining ratios and SUV were also investigated.Results: AT SUV parameters were generally decreased comparing with BT SUV parameters (p<0.001). All SUV parameters in delayed scans were higher than early scans both AT and BT images (p<0.001).Eight different groups were generated according to the pathological complete response, TRG, the presence of macroscopic tumor, the presence of significant fibrosis, T staging regression compared with BT MRI, pathological N staging regression compared with BT MRI, T staging regression with BT-AT MRI, pathological N staging regression with BT-AT MRI. When we classified the patients according to the TRG; there were 12 (%41) patients with TRG1, 10 (%35) patients with TRG2, 7 (%24) patients with TRG3. There was significant difference in AT delayed T SUVaverage / L SUVaverage (p: 0.047) and LBM AT delayed T SUVaverage (p: 0.036) between three groups.There were 5 (17%) patients with complete response and 24 (83%) patients with non-complete response. There was significant difference in AT delayed SUVmax-average, LBM AT delayed SUVmax-average and T/L ratios between early and delayed scans. There were 17 (%59) patients with macroscopic tumor and 12 (%41) patients with non-macroscopic tumor. ATdelayed T SUVaverage, LBM ATdelayed T SUVmax and LBM ATdelayed T SUVaverage were significantly lower in the patients with non-macroscopic tumor. There was significant difference in T/L SUV values between AT and BT images in both early and delayed scans.There were 22 (%76) patients with significant fibrosis and 7 (%24) patients without significant fibrosis in the pathologic surgical specimen. ATdelay T SUVmax (p: 0.047) and LBM ATdelay T SUVaverage (p: 0.047) were calculated as significantly lower in the patients with significant fibrosis. There was no significant difference in SUV parameters in the other groups.AT T SUVmax, AT T SUVaverage, AT T SUVaverage/L SUVaverage, AT T SUVmax/L SUVaverage, LBM T SUVmax and LBM T SUVaverage in the AT early scan and AT Ki-67 values were moderately correlated. AT T SUVaverage/L SUVaverage, AT T SUVmax/L SUVort values and AT GLUT1 were also moderately correlated.Conclusion: PET-CT is an effective method for assessing the response to neoadjuvant CRT in the patients with LARC. The most significant parameter for assessing treatment response between SUV parameters is tumor/liver ratios. LBM SUV values may have a role to evaluate effectiveness of therapy.Since a significant correlation between AT Ki-67 and SUV values was found, it showed that SUV value is a good metabolic indicator. Because AT GLUT1 and AT T/L ratio were correlated, we thought that T/L ratio might reflected the rate of metabolic activity more accurately than the other SUV parameters in the assessment of treatment responseKey words: F-18 FDG PET-CT, SUV values, LARC, Ki-67, GLUT1.