Some studies clearly demonstrate that [¹⁸F]-FDG-PET/CT shows higher diagnostic performance than conventional imaging modalities (US, MDCT, magnetic resonance imaging [MRI]) in detecting or ruling out CRC recurrence, namely, in loco regional disease and evaluation of local lymph nodes [5–9, 14, 20, 34–36]. In fact, conventional imaging modalities, including MDCT, MRI, and US, often have limited sensitivity and specificity for detecting CRC recurrence, showing significant limitations in detecting tumor in the distorted pelvic anatomy following surgery and radiotherapy, as well as in the differential diagnosis between scar and viable tumor tissue in previously treated abdominal regions or in diagnosing lymph node metastasis, being unable to distinguish whether changes in size are a result of metastatic infiltration or other causes [37–39].

Maas and colleagues reported results of a meta-analysis in which the diagnostic performance of PET/CT, CT, and MRI as whole-body imaging modalities were compared for detecting local and distant recurrent disease in patients with a high suspicion of CRC relapse determined by clinical findings or CEA elevation [40]. In a patient-based analysis of the 14 observational studies included, PET/CT vs. PET and CT showed greater accuracy in depicting recurrent disease in patients with elevated suspicion of recurrence. The lower diagnostic performance of CT is caused by poor accuracy in detecting extrahepatic lesions, which makes this modality less effective for performing whole-body surveillance. The introduction of MDCT did not affect the sensitivity for detecting local recurrence, which remains low (38–82%), but significantly increased specificity (97–100%) [41, 42]. Metser and colleagues compared contrast-enhanced (CE) 64-row MDCT of the body with PET/CT in a series of 50 patients with a history of CRC and elevated serum CEA levels. The sensitivity of CE 64-row MDCT was 70.3% vs. 97.3% for PET/CT [43]. One possible explanation for such high discrepancy is the retrospective nature of the study. The anatomical location of CRC recurrences may also influence CT performance: in this case, MDCT attenuation differences are subtle among benign vs. malignant or normal vs. abnormal tissues. Identifying recurrent CRC within the presacral space in patients who previously underwent abdominoperineal resection may be missed by CT, as may recurrences in the border/periphery of liver nodules treated by radiofrequency ablation [43]. Such evidence suggests that PET/CT might be performed as the first-line imaging modality, instead of CT, for patients suspected of having recurrent CRC based on clinical findings or increasing CEA levels. In fact, in these patients, a negative CT finding does not rule out the presence of recurrence, and a PET/CT is mandatory. In addition, PET/CT is used in clinical practice when CT results are equivocal, as it provides better definition of suspected lesions. It is also used for patients with evident recurrences on CT to exclude distant metastases if curative surgery is considered.

The use of integrated FDG-PET/CE-CT is probably the most appropriate imaging modality for staging and restaging CRC. To date, a few groups have used intravenous CE administration for the CT component of the PET/CT scan, showing that PET/CE-CT provides more information than CE-CT or standard PET/CT performed separately [44–46]. The major advantages of FDG-PET/CE-CT are higher sensitivity in detecting metastatic lymph nodes vs. CE-CT alone and its ability to identify local recurrence [44–46].

In a series of 54 patients undergoing restaging for CRC, Soyka et al. [44] demonstrated that PET/CT provided accurate, additional information to CE-CT findings in 50% of patients and that PET/CE-CT was even more accurate, offering appropriate restaging in 72% of cases, with a therapeutic relevance in 23 patients compared with PET/CT. In patients suspected for recurrence, Kitajima et al. [45] demonstrated that PET/CE-CT findings influenced patient management; 12 were diagnosed by CE-CT and four by PET. In a patient-based analysis, Jiménez Londoño et al. found that PET, CE-CT, and PET/CE-CT had similar values in detecting unsuspected recurrence of CRC, with a slightly higher specificity for combined assessment of PET/CE-CT compared with the isolated evaluation of each technique [46].

In the standard follow-up of patients with CRC, these publications address the use of combined assessment using PET and CE-CT obtained in an integrated acquisition protocol. Results clearly emphasize the role of FDG-PET/CE-CT in detecting regional and distant relapse in asymptomatic patients with a high risk of recurrence. From these reported experiences, PET/CT would appear to be an adequate tool for staging and follow-up of patients with CRC, showing diagnostic performance superior to other conventional imaging modalities. However, the clinical position of PET/CT in routine management of CRC varies. For instance, in the National Comprehensive Cancer Network (NCCN) Guidelines, PET/CT is not routinely indicated either in staging or in the follow-up of CRC. It is, however, suggested/indicated in patients with serially increased CEA and resectable metastases in order to identify or exclude other possible lesions and avoid unnecessary surgery [47]. Other groups proposed the use of PET/CT for specific clinical scenarios, such as initial staging of primary rectal cancer, because it may modify staging in 40% of cases and affect therapeutic strategy in 18–24% of patients by showing unsuspected metastases or clarifying the nature of indeterminate lesions [48, 49]. To date, there are no firm criteria to define the appropriate use of PET/CT for surveilling CRC patients. Some reports clearly show that it should be used in addition to conventional follow-up imaging in patients with a higher risk of recurrence [50–52].

The role of whole-body MRI in recurrent CRC is still under investigation. MRI, which is highly valuable in the staging of advanced primary and recurrent intestinal tumors — being accurate in depicting invasion of anatomical structures adjacent to the rectum — is suboptimal in evaluating involvement of pelvic sidewalls or female reproductive organs. The reported levels of accuracy for MRI are inferior to or in the same magnitude as those obtained by PET/CT [53–55].

There is growing interest in applying PET in settings that go beyond the sole purpose of identifying tumor lesions. The functional nature of this imaging method allows identifying changes in tumor metabolism that may occur very rapidly in response to effective treatment. A number of studies in the past 10 years have correlated changes in FDG uptake during and following treatment with pathologic response assessed on tumor specimens. Most data have been obtained in patients with locally advanced rectal cancer undergoing neoadjuvant treatment prior to surgical resection. Studies have assessed various combinations of chemotherapy and radiation treatment in this setting. Despite the wide heterogeneity in patient selection and the timing of PET scans during and after treatment and clinical end-points, there is a large consensus that FDG-PET may be a powerful tool for assessing treatment response and predicting outcomes [56].

Published data indicate that [¹⁸F]-FDG-PET has a high predictive value in the therapeutic management of CRC patients. This technique could be valuable for improving patient care by reducing costs, morbidity, and effort spent in pursuing ineffective treatment in patients who are identified as nonresponders. Available studies on chemotherapy response monitoring of patients with advanced CRC and studies on preoperative radiotherapy and multimodality treatment response evaluation in primary rectal cancer indicate that [¹⁸F]-FDG-PET is a reliable predictor of therapy outcome in both situations [56]. When [¹⁸F]-FDG-PET is able to predict the final outcome, it may be used to guide adjuvant chemotherapy for rectal cancer after optimal neoadjuvant and local treatments.

Data from our group indicate that patients with locally advanced rectal cancer who showed a complete or subtotal tumor regression could be accurately identified by evaluation of early average SUV change (ΔSUVmean: variation between baseline and 12 days from start of neoadjuvant radiochemotherapy) in the primary tumor. Early ΔSUVmean ≥52% [57] identified pathologic responders with 100% sensitivity and specificity. Long-term follow-up demonstrated a statistically significantly higher 5-year relapse-free survival compared with the 20 nonresponders (86 vs. 55%, p=0.014), who showed poor pathologic responses. A multivariate analysis demonstrated that early ΔSUVmean was the only pre surgical parameter correlated to the likelihood of recurrence (p=0.05) [58]. This study was the first prospective long-term evaluation demonstrating that FDG-PET is not only an early predictor of pathologic response but is also a valuable prognostic tool and therefore should be used in optimizing multidisciplinary management of patients with locally advanced rectal cancer.

Assessment of response by [¹⁸F]-FDG-PET may also be effective in other settings. In local ablative treatment of liver metastases, PET may allow early detection of incomplete tumor ablation that is not detectable by CT. Another potential application may be as a surrogate marker for treatment response in early clinical trials assessing new antineoplastic agents, where use of a rapid functional endpoint may shorten the observation times necessary to determine drug efficacy. Therefore, therapy response assessment with [¹⁸F]-FDG-PET is a worthwhile research topic and may provide a useful clinical tool in the near future.

[¹⁸F]-FDG-PET/CT imaging has limitations, which must always be taken into account when evaluating clinical studies. False positive findings may be observed in benign conditions such as inflammation, and may show varying levels of increased tracer uptake, which may be confused with disease. There are also various situations in which false negative findings are observed. The limited intrinsic spatial resolution of currently available scanners makes it so that lesions <1 cm in diameter may go undetected or may not yield accurate quantitation of radiotracer uptake. Similarly, certain histotypes, in which extracellular components such as mucin are more abundant than the cellular component — which is responsible for FDG concentration — may show reduced “biological” sensitivity and have low or no FDG uptake on imaging studies. The presence of high blood glucose levels also affects FDG uptake by the cancer cells by direct competition.

The survey of current literature supports that [¹⁸F]-FDG-PET/CT, when performed to address selected clinical issues in patients with suspected or known CRC, provides valuable information that has a significant and considerable impact on disease management, primarily related to improved staging accuracy and by sparing unnecessary surgical procedures. In addition, [¹⁸F]-FDG-PET/ CT imaging detects recurrent disease in the early stages in comparison with conventional imaging procedures, which means the opportunity to offer curative surgery to CRC patients. Ongoing work will clarify the possible role of PET to monitor and adjust systemic treatment in patients showing tumor recurrence.