AIM: To evaluate the value of positron emission tomography (PET)/computerized tomography (CT) in surveillance of colorectal cancer (CRC) patients with different carcinoembryonic antigen (CEA) concentrations.

METHODS: One hundred and six postoperative CRC patients who had suspected recurrence or metastasis and received fluorodeoxyglucose (FDG) PET/CT within one week were included in this study. The final diagnosis was confirmed by histological examination or clinical follow-up over at least six months.

RESULTS: The sensitivity, specificity, and accuracy of FDG PET/CT were 95.2%, 82.6%, and 92.5%, and 94.8%, 81.4% and 92.8%, respectively, in the case- and lesion-based analyses. The sensitivity and accuracy of FDG PET/CT significantly differed from CT in both analyses (χ² = 8.186, P = 0.004; χ² =6.201, P = 0.013; χ² =13.445, P = 0.000; χ² =11.194, P = 0.001). In the lesion-based analysis, the sensitivity, specificity, and accuracy of FDG PET/CT in the abnormal CEA group were 97.8%, 82.6%, and 95.6%, compared with 81.3%, 80%, and 80.6% for patients with normal CEA levels. In case-based analysis, the sensitivity, specificity, and accuracy of FDG PET/CT were 97.2%, 77.8%, and 95% in abnormal CEA group. Only in lesion-based analysis, the sensitivity and accuracy of FDG PET/CT in the abnormal CEA group were significantly superior to those in the normal CEA group (χ² =6.432, P = 0.011; χ² =7.837, P = 0.005). FDG PET/CT changed the management in 45.8% of patients with positive scans.

CONCLUSION: FDG PET/CT showed superior diagnostic value and is an advisable option in surveillance of postoperative CRC patients with a vague diagnosis.

Colorectal cancer (CRC) is one of the most common cancer entities worldwide[1]. Despite the fact that 70% of the patients have a chance of radical operation, 30%-50% of them will develop metastasis or local recurrence within two years after operation[2,3]. For CRC patients, both local recurrence and metastasis can be addressed by reoperation. Only 10%-30% of recurrent patients can be treated by salvage surgery[4,5], but curative-intent surgeries are associated with a 5-year survival rate of 30%-40% in selected patient populations with single organ metastasis[6]. Thus, the surveillance of postoperative colorectal cancer should enhance the proportion of resectable cases by early detection of the recurrence and metastasis in order to improve the survival of CRC patients.

The surveillance is usually performed by a regular physical examination, determination of the serum carcinoembryonic antigen (CEA) level, colonoscopy, and conventional imaging techniques such as ultrasound of the liver, contrast-enhanced computerized tomography (ceCT) and magnetic resonance imaging (MRI). It remains unverified what strategy can provide significant survival benefits in routine follow-up of CRC patients[7,8]. Serum CEA is generally used as a tumor marker and CEA elevation predicts a high risk of recurrence and poor survival for CRC[9]. Elevated CEA values can not provide any accurate information regarding the sites of recurrence and complementary imaging techniques should be provided for the diagnosis of CRC recurrence. Conventional imaging techniques primarily offer morphologic data based on anatomical information. It is difficult to identify recurrent disease from nonmalignant changes, such as scars, inflammation lesions and radiation necrosis by this morphological imaging tool. Fluorine-18 (¹⁸F) fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT is an integrated imaging modality of anatomic and functional imaging and can show metabolic changes before morphological ones. In the literature, ¹⁸F-FDG PET/CT was considered to be superior to the conventional imaging in the early detection of recurrence and metastases in CRC patients with elevated CEA[2,10]. Recently, it has been reported that PET/CT showed a high-positive predictive value for metastases in postoperative colorectal cancer patients with normal CEA levels[6,11]. However, due to some limitations in previous studies, such as small number of patients and unsatisfactory statistic analysis, the clinical value and efficacy of FDG PET/CT in surveillance are not yet fully established.

In this retrospective study, we aimed to evaluate the diagnostic performance of ¹⁸F-FDG PET/CT in surveillance of postoperative CRC patients as compared with CT and to investigate the role of FDG PET/CT in patients with different CEA concentrations.

In surveillance of postoperative CRC, CT is the preferred imaging technique for local recurrence detection[6] and MRI is regarded as the most sensitive conventional imaging tool for liver metastases[12,13]. However, the early diagnosis of recurrence is still difficult through conventional surveillance strategies. It has been reported that PET/CT showed better sensitivity and specificity (87%-100% and 90%-98%, respectively) for detection of hepatic and extra-hepatic metastasis than CT[14-16]. In some studies, the sensitivity of PET/CT for hepatic lesions was 91%-100%, which was similar to that of MRI[17,18]. In our study, ¹⁸F-FDG PET/CT showed an excellent diagnostic performance in the detection of CRC recurrence and metastasis and its sensitivity and accuracy were significantly superior to those of ceCT. We found that many positive lesions in FDG PET/CT, which were once classified as negative lesions by CT (Table 5), were finally confirmed as true positives by gold standard. This showed that it would be possible that some CRC patients with recurrences and metastasis might obtain the chance for a second remission and improve the prognosis through PET/CT findings.

CEA is used as the early indicator for the recurrent disease and is elevated in approximately 60%-70% of patients with recurrence[10,19]. It has been observed that some potentially curable recurrent tumors were detected by routine imaging techniques, while CEA levels were still normal. Therefore, CEA measurements had only a marginal effect on survival[11]. Moreover, for postoperative CRC patients who have a suspicion of recurrence based on the rise in the CEA level, there is controversy on the most accurate imaging technique. Ozkan et al[2] reported that the sensitivity and specificity of ¹⁸F-FDG PET/CT in the detection of disease recurrence in postoperative CRC patients were 97% and 61%, while they were 51% and 60% for CT, respectively. Mittal et al[20] reported that PET/CT showed recurrences in 71% of CRC patients and the positive rate increased with the CEA level. Our data also showed coincidentally that PET/CT was superior to CT in terms of sensitivity, specificity, and accuracy in all groups of patients and had a tendency to increase with the CEA level (Table 4). Lee et al[21] evaluated a group of CRC patients with normal CEA levels and reported that the sensitivity, specificity, and accuracy for FDG PET/CT were 95%, 76.6%, and 88.8%, respectively. In our study, the sensitivity, specificity, and accuracy of PET/CT in patients with normal CEA was similar to the data. In addition, the sensitivity and accuracy of FDG PET/CT in patients with abnormal CEA levels significantly differed from those with normal CEA levels in the lesion-based analysis. But in the case-based analysis, the FDG PET/CT diagnostic performance did not show a significant difference between these two groups. It may be related to the small quantity of patients with normal CEA levels, which could influence the statistical analysis. However, the data still suggested that FDG PET/CT was effective similarly in patients with normal and abnormal CEA levels. For patients with a vague diagnosis of recurrence or metastasis based on a routine examination, FDG PET/CT might provide much benefit to patients by increasing the diagnostic accuracy. Based upon these results, FDG PET/CT is an advisable option for CRC patients with an indefinable diagnosis and should be recommended in surveillance of post-operative CRC patients even with normal CEA.

In this study, the specificity of FDG PET/CT did not significantly differ from that of ceCT in both case- and lesion-based analyses. FDG PET/CT scans showed false-positive results which mainly included the inflammatory lesion and single hypermetabolic lesion in the organs. Our data demonstrated that inflammatory processes can result in hypermetabolism and consequently false-positive results, for example, the inflammatory processes of colitis in the anastomotic site. In addition, for a single hypermetabolic lesion of the bone, metastasis should be differentiated from the injuries and restorations correlated to other imaging systems. For lung nodules, metastasis should be differentiated from inflammatory lesions and tuberculosis.

It has been reported that mucinous carcinoma usually showed little uptake of the FDG and the sensitivity of FDG-PET imaging for detection of mucinous carcinoma was significantly lower than that for nonmucinous carcinomas[3]. In our results, two quarters of FN cases and three sixths of FN lesions were mucinous adenocarcinoma and it demonstrated that mucinous carcinoma was the main factor responsible for FN scans. Therefore, it was suggested that for mucinous adenocarcinoma patients with negative FDG-PET imaging results, other imaging modalities should be recommended for further diagnosis. Otherwise, nonvisualization of FDG PET/CT detection should be attributed to the little FDG uptake for lymph nodes and peritoneal micrometastases, consistent with the literature report. Moreover, our FDG PET/CT scanning data were acquired in a 2D mode with 4.25 mm spatial resolution and a 256 x 256 matrix. Given better spatial and temporal resolution (3D mode with 2 mm spatial resolution, 400 x 400 matrix and continuous table movement), more tiny lymph nodes and metastases can be revealed precisely. With the decrease of false diagnosis, the further accuracy elevation can be expected.

In terms of the superiority of FDG PET/CT over CT in detection of recurrence and metastasis, FDG PET/CT might provide chances to select suitable patients for surgical resection or other local treatments (radiotherapy, embolization, and radio-frequency ablation). Meanwhile, some unnecessary operations might be avoided. Many studies demonstrated some disease management change as a result of PET/CT usage in 30%-56% of patients with a suspected or confirmed recurrence of CRC[22-24]. It is possible to benefit from this strategy in terms of patient survival with early detection and treatment of tumor recurrence by FDG PET/CT[6,25]. In our studies, 45.8% of FDG positive patients had a changed disease management and received suitable treatments. Second operation was performed only in 12% of retreatment candidate patients, because most patients of the groups had already developed multi-organ metastases when they received the FDG PET/CT scan. It suggested that the early detection due to the use of PET/CT in surveillance of CRC could correct the disease management strategy and improve the treatment efficacy by reoperation for potential curative patients.

Our study had two potential limitations. Due to the retrospective nature of the study, we were unable to obtain unified clinical date for the patients and there was an inter-observer variation for imaging interpretation that might have had some influence on the sequences. The second was that histopathological confirmation was only performed in some of the patients in this study.

In conclusion, FDG PET/CT showed a superior diagnostic performance in surveillance of postoperative CRC patients. For patients with suspicious recurrence or metastasis based on a routine examination, our data suggest that PET/CT is an excellent option to replace CT in the follow-up of CRC patients even when CEA is normal.

The authors thank Dr. Hui-Qing Li and Yuan-Yuan Liu for their assistance with this paper.