GBC shares the same diagnostic challenges as CHC, which is an insidious course resulting in a late detection often at an advanced stage of the disease (Fig. 43.4) as well as the occurrence of nontumoral conditions that may simulate cancer on imaging studies such as inflammatory processes, especially xanthogranulomatous cholecystitis. Although few studies have investigated the role of PET for diagnosis, prognosis, staging, and tumor recurrence, this modality is seen as a potential useful technique that may eventually find its proper role in the GBC work-up. In a prospective study that evaluated FDG metabolism with PET, overall diagnostic accuracy was 83.3 % for the diagnosis of the primary lesion, 88.9 % for evaluation of lymph node involvement, and 85.1 % for the evaluation of metastatic disease. Mean SUV_max of malignant lesions was 6.14 and ROC curve showed a cutoff value of 3.56 g/ml in the SUV_max for malignancy. PET changes management in 14.8 % of patients as a result of identification of metastatic disease although proved to be limited in the setting of peritoneal carcinomatosis [25].

Another study that specifically addressed the role of PET in risk stratification of 1–2 cm gallbladder regarding potential surgical management found that various criteria derived from 18F-FDG-PET including visual assessment, SUV_max of the polyp, and, most notably, the ratio of SUV_max of the polyp to SUV_max of the liver were useful in determining the need for surgical intervention and data derived from PET was more effective than other clinical risk factors, except for the age of the patient [26]. Metabolic evaluation with PET/CT has also been found to be useful in detecting recurrent disease as well as metastatic distal involvement with sensitivity of 97.6 % and specificity of 90 %, whereas the positive predictive value, negative predictive value, and accuracy reported were 95.3 , 94.7, and 95.1 %, respectively. Therefore, PET has the potential to detect early recurrent locoregional as well as distant spread in the setting of GBC resulting in ability to allow for an impact on management [27].

There has been a recent interest to screen for hepatopancreaticobiliary malignancies by means of obtaining spectral analysis of metabolite profiles in human bile juice. Magnetic resonance spectroscopy (MRS) provides a noninvasive assessment of different metabolites whose normal and abnormal profiles correlate with the expected chemical alterations expressing tumor cell pathways [28].

Albiin et al. investigated the ability of 1H MRS to discriminate CHC from benign biliary conditions by obtaining spectra of sample bile from patients with and without PSC. The reported sensitivity, specificity, and accuracy were 88.9, 87.1, and 87.8 %, respectively [29]. Attempts have been tried to analyze differences in phospholipid metabolism by means of in vitro 1H and phosphorus 31P MRS with a high-field system (11.7 T). Khan et al. evaluated patients with nonmalignant pathologies and with CHC among other malignant hepatopancreaticobiliary malignancies and demonstrated that phosphatidylcholine levels were significantly reduced in the bile from cancer patients compared with those with no malignancy. They concluded that 1H and 31P MRS may be used to detect differences in phospholipid metabolism between these two populations of patients [30].

Other potential molecular target to be investigated by MRS is lactate level whose measurement in bile has been shown to correlate with hepatobiliary malignancies [31].

Wen et al. proposed a NMR-based metabolomic approach to aid in differentiating malignant from nonmalignant biliary tract diseases from bile samples obtained either percutaneously or endoscopically. They produced sensitivity and specificity of 88 and 81 %, respectively, performing better than conventional markers such as CEA, CA19-9, and bile cytology [32].

Although the aforementioned experiences are still far from current practice, they signal significant early attempts to employ novel imaging biomarkers that overcome present limitations of currently used noninvasive screening methods.

Although, at present, in vivo metabolic imaging through MRS is a very difficult task to perform in patients with anatomically demonstrated hepatobiliary masses, continued hardware and software innovations may result in a future useful approach to accomplish noninvasive molecular detection and characterization of tumors.