Diagnostic Performance of F-18 FDG PET/CT in the Detection of Recurrent Colorectal Cancer: Correlation with Biochemical Markers and Conventional Imaging Modalities

Abstract

Background/Objectives: Although the role of PET/CT imaging is well established in oncology, its diagnostic value in routine monitoring for recurrent colorectal cancer (CRC) is still controversial. The aim was to evaluate the diagnostic value of F-18 FDG PET/CT in detecting recurrent CRC in correlation with CEA, CA 19-9 levels, and conventional imaging modalities (CIM). Methods: Between 2009 and 2023, a retrospective study was performed including 134 CRC patients referred for PET/CT imaging on the suspicion of recurrence, based on elevated CEA and/or CA 19-9 and/or equivocal CIM findings. According to our institution's Tumor Board CRC protocol, after the initial treatment, which was dependent on the TNM stage (neoadjuvant therapy, primary resection, or adjuvant treatment), patients underwent a standard 5-year surveillance including CEA and CA 19-9 measurements, CIM, and colonoscopy, every six months. The statistics, including univariate and multivariate analyses were conducted using the IBM SPSS 20.0 statistical software. p-values < 0.05 were considered statistically significant. Results: Recurrent CRC was confirmed in 54/134 (40.3%) patients with elevated tumor markers. PET/CT showed high diagnostic performance in detecting recurrent CRC with sensitivity, specificity, PPV, NPV, and accuracy of 94.4%, 82.5%, 78.5%, 95.7%, and 87.3%, respectively. The CEA showed a high sensitivity of 98.1% but both low specificity and accuracy of 15% and 48.5%, respectively. The sensitivity, specificity, and accuracy for CA 19-9 and CIM for diagnosis of CRC recurrence were 44.4%, 67.5%, 58.2%, and 51.9%, 98.8%, 79.9%, respectively. The AUC for PET/CT, elevated CEA levels, CIM, and elevated CA 19-9 levels was 0.885 (95% CI: 0.824-0.946; p < 0.001), 0.844 (95% CI: 0.772-0.916; p < 0.001), 0.753 (95% CI: 0.612-0.844; p < 0.001), and 0.547 (95% CI: 0.442-0.652; p = 0.358), respectively. Univariate analysis showed that both PET/CT and CIM positive results were highly associated with CRC recurrence (p < 0.001 and p < 0.001, respectively). At the same time, gender, mucinous tumor type, presence of initial lymph node metastasis (N+), and presence of initial distant metastasis (M+) had no significance (p = 0.211, p = 0.158, p = 0.583, and p = 0.201, respectively). Our multivariate analysis showed that independent predictors for CRC recurrence are positive PET/CT scans (p < 0.001), positive CIM results (p = 0.001), and elevated CA 19-9 levels (p = 0.023). Although CA 19-9 was not detected as a statistically significant predictor in the univariate analysis (p = 0.358), in a multivariate analysis it was recognized as a significant predicting factor in detecting the CRC recurrence (p = 0.023). Conclusions: F-18 FDG PET/CT showed high diagnostic efficacy in CRC recurrence detection, in correlation with CEA levels, CA 19-9 levels, and CIM. This imaging modality should be routinely integrated into the post-operative follow-op in patients with elevated tumor markers.

Keywords: CA 19-9; CEA; F-18 FDG PET/CT; colorectal cancer; conventional imaging; detection; recurrence.

Figure 1

(a) PET/CT axial image detects a focal area of increased FDG uptake in the right pulmonary lobe, S4, SUVmax 7.50 (white arrow), corresponding to lung metastasis, (b) diagnostic CT scan in axial plane confirms a lesion in the right pulmonary lobe, S4 (white arrow), consistent with a metastatic lesion in the lung, (c) PET/CT axial image shows two hypermetabolic lesions in the liver in S8, SUVmax 8.9 (black arrow) and in S4a, SUVmax 8.0 (arrowhead), corresponding with liver metastases, and (d) diagnostic CT in axial plane shows two heterodense dominantly hypodense lesions in S8 (black arrow) and S4a (arrowhead), consistent with distant metastases in the liver.

Figure 2

(a) Axial T1W MR image shows focal T1W hypointense lesion in the right iliac bone (white arrow) corresponding with metastasis, and (b) PET/CT axial image does not detect FDG-avid lesion in the region of pelvic bone.

Figure 3

(a) Diagnostic CT in axial plane shows no lesion in the left pelvic region, and (b) PET/CT axial image detects an FDG-avid ovoid lesion in the left obturator area, SUVmax 7.20 (white arrow), corresponding with lymph node involvement.

Figure 4

ROC analysis.