A 66-year-old male patient who is hepatitis C RNA–positive without cirrhosis has been found to have a mass on ultrasound examination of the liver. The physical examination is unremarkable. The hemoglobin, white cell count, and platelet counts are normal. The alkaline phosphatase level is mildly elevated to 212 U/L (normal level < 115 U/L), the aspartate aminotransferase level is 112 U/L, and the alanine aminotransferase level is 128 U/L. The carbohydrate antigen 19-9 level is 82 U/mL (normal level < 55 U/mL), and the alpha-fetoprotein level is 12 ng/mL. A computed tomography (CT) scan of the abdomen shows a vague mass in segment V of the liver without definite arterial enhancement. A magnetic resonance imaging (MRI) examination of the abdomen is carried out with contrast, but the mass is still indeterminate on T1 and T2 imaging.
What is the role of a positron emission tomography (PET) scan in the diagnosis and staging of the liver mass?
Incidentally identified hepatic masses are a frequent clinical finding because of the widespread use of cross-sectional imaging modalities (i.e., CT and MRI abdominal scan) and surveillance imaging studies in patients with chronic liver diseases. Up to 20% of the population has a benign hepatic lesion (most commonly cavernous hemangioma and focal nodular hyperplasia). Benign masses can be subdivided into lesions requiring serial monitoring or intervention and lesions not requiring any further evaluation. Benign lesions not requiring any follow-up or further evaluation include hepatic cysts, cavernous hemangiomas, focal nodular hyperplasia, focal fat in a normal liver, and a focus of normal tissue in a fatty liver. Hepatic adenomas are benign lesions but may require interval monitoring or intervention (ablation or resection) because of their potential for malignant transformation and hemorrhage.
The most common primary malignant hepatic masses include intrahepatic cholangiocarcinoma (CCA) and hepatocellular carcinoma (HCC). Patients with cirrhosis are predisposed to developing HCC and CCA, and the incidence of both these malignancies has increased during the last 3 decades. The characterization of hepatic lesions in cirrhosis frequently requires multiple imaging modalities. HCC is diagnosed on the basis of early arterial enhancement on contrast CT or MRI examinations, early washout, and the presence of a pseudocapsule. Venous washout, defined as a hypervascular mass that becomes hypointense with respect to the adjacent parenchyma on delayed postcontrast images, has been reported as an imaging finding that increases the specificity for HCC. CCAs are well-defined homogeneous masses that enhance somewhat later than HCC. These characteristics are typical, but size, calcification, necrosis, and hemorrhage may alter the appearance of both HCC and CCA. However, the diagnosis of HCC and CCA may not be clear despite a combination of imaging techniques. In such situations, functional imaging techniques might be considered to help further characterize the lesion and thereby guide management.
Positron emission tomography (PET) is a functional imaging modality that has been proven useful as a diagnostic tool for a number of tumor types. It is based on the tumor-specific high intracellular accumulation of the glucose analogue fluorodeoxyglucose (18F-FDG). Tumor cells have increased glucose uptake and glycolysis. Glucose uptake is mediated by members of the glucose transporter family, which are called GLUTs and can be regulated by hypoxia, oncogenes, and growth factors. After its uptake, glucose is phosphorylated, and this results in its intracellular trapping as glucose-6-phosphate, an intermediate in the glycolytic pathway. In hepatocytes, this reaction is mediated by the enzyme glucokinase, whereas in tumor cells, it is mediated by hexokinase (particularly hexokinase II). In contrast to glucokinase, hexokinase has a low Michaelis constant, and this results in high maximum velocities even at low glucose concentrations. Subsequently, glucose-6-phosphate is phosphorylated at its 2′-OH group by phosphofructokinase 1; this is the key enzymatic step in glycolysis. In the glucose analogue 18F-FDG, the hydroxyl group (-OH) at the 2′ position in the glucose molecule is substituted by the positron-emitting radioactive isotope fluorine-18, and further metabolism of 18F-FDG is thereby inhibited. Thus, fluorodeoxyglucose 6-phosphate (18F-FDG-6-P) cannot move out of the tumor cell or be metabolized before radioactive decay (Fig. 1). Certain radioisotopes decay by positron emission (a process called positron annihilation), and this results in the emission of two 511-KeV photons at a 180° angle. These photons are detected simultaneously by detectors encircling the patient; this process is called coincidence detection. PET images are reconstructed from large numbers of detected coincident events and represent the radiotracer distribution in the body. For quantification of the tracer uptake, a semiquantitative measurement known as the standardized uptake value (SUV) is most commonly used. It is calculated by the division of the activity per unit of volume by the injected activity per body weight. Tumor images based on the quantification maximum [maximum standardized uptake value (SUVmax)] rather than average SUV are used for analysis. The analysis of hepatic hot spots can be challenging because of signal attenuation by the liver size. Attenuation correction and signal quantification can help in the interpretation of hepatic lesions, and frequently, the tumor-to-benign tissue ratio is used for analysis.
Liver Masses for Which PET Is Helpful: Lymphoma and Metastases
Primary hepatic lymphoma (PHL) is extremely rare, although it has been described in patients with hepatitis C virus (HCV)–positive liver disease. Most commonly, it presents as a solitary hepatic mass, but multiple masses as well as diffuse patterns have been described. On ultrasound, PHL is usually homogeneously hypoechoic; on CT, PHL typically presents as a hypoattenuating lesion. A central area of low intensity indicating necrosis may be present. Enhancement patterns on dynamic imaging are quite variable; 50% of PHL lesions do not enhance at all, 33% show patchy enhancement, and 16% show ring enhancement. On MRI, PHL presents as hypointense or isointense on T1-weighted images and as hyperintense on T2-weighted images. However, most of these radiological findings are nonspecific, and such lesions are often misdiagnosed as HCC or metastases. PHL accumulates 18F-FDG, and the feasibility of PET for the evaluation of PHL has been well described. However, because of its rarity, there are no large studies evaluating the accuracy, sensitivity, or specificity of PET for this disease entity. Secondary extranodal hepatic lymphoma is more common, and PET is used for the assessment of the treatment response in patients with lymphoma undergoing chemotherapy. Lesions >1.5 cm and 18F-FDG accumulation exceeding hepatic and splenic 18F-FDG are considered positive for lymphoma.1
Colorectal carcinoma (CRC) is the most common cause of hepatic metastases. Therefore, a majority of studies have evaluated the value of 18F-FDG-PET in the detection of hepatic metastases from CRC. The sensitivity and specificity of PET/CT for the detection of hepatic metastases from CRC are 88% to 96% and 75% to 96%, respectively.2, 3 A subgroup analysis showed a slightly higher sensitivity for the detection of non-CRC metastases (94% versus 98%) but an equal specificity of 75%.2 PET/CT outperforms contrast-enhanced CT when it is used for the evaluation of intrahepatic recurrence after hepatic resection or residual disease after local ablation. After neoadjuvant chemotherapy, the sensitivity of PET in detecting metastases decreases to 49%. In contrast, conventional CT has a sensitivity of 65%.4 PET/CT was found to cause a change in management in 25% to 32% of patients.2, 3 In summary, PET imaging is helpful in the evaluation of hepatic metastases from CRC, especially after hepatectomy and local ablative treatments, when it is superior to CT. However, its main advantage is its ability to detect extrahepatic disease that may lead to a change in management.
PET for the Diagnosis of CCA: Sometimes Helpful
CCA frequently poses a diagnostic challenge, with conflicting results produced by different diagnostic modalities. Several centers have demonstrated the feasibility of PET in the diagnosis of CCA (Fig. 2). These reports are limited by their small sample numbers and the lack of distinction between intrahepatic, perihilar, and distal extrahepatic CCAs. The majority of studies have found SUVmax values of 2.5 to 3.5 to be cutoff points for the diagnosis of CCA. The overall sensitivity of PET for CCA was 61% to 92%, the specificity was 75% to 93%, the positive predictive value was 95%, and the negative predictive value was 25% to 64% (Table 1).5, 6 A direct comparison of CT and PET/CT for the diagnosis of the primary tumor revealed no significant differences in the sensitivity, specificity, accuracy, negative predictive value, or positive predictive value between the two imaging modalities.7 Factors influencing the sensitivity of PET for CCA include image acquisition, tumor location, morphology, and size. In extrahepatic CCA, a detection rate of 85% with PET was described for tumors 1.0 to 6.7 cm in size; this decreased to 36% in cases for which CT described only bile duct thickening.8 The sensitivity for the diagnosis of CCA in the absence of a mass on other imaging modalities was only 18%, whereas it was 85% for CCA forming a mass of more than 1 cm, and the positive predictive value was 94%.5 In one study, there was a higher sensitivity for intrahepatic CCA: the sensitivity was 95% for intrahepatic CCA versus 69% for extrahepatic CCA.6 With fusion PET/CT, there was a statistically significant difference in the sensitivity of PET/CT for the detection of intrahepatic CCA versus extrahepatic CCA: 93% for intrahepatic CCA and 55% for extrahepatic CCA.7 The overall sensitivity of PET/CT for the detection of perihilar CCA was 58.8%.9
Table 1. Sensitivity and Specificity of PET
Primary sclerosing cholangitis is a well-known risk factor for CCA. Studies evaluating 18F-FDG-PET for CCA in patients with primary sclerosing cholangitis have described significant intratumoral 18F-FDG uptake in these patients, and sensitivities and specificities of 75% to 100% and 80% to 95%, respectively, have been reported.10, 11 The positive predictive value has been reported to be 75%, and the negative predictive value has been reported to be 95% versus 83% with CT and ultrasound.11 The detection rate of PET for locoregional and distant lymph node metastases is suboptimal and has been reported to be between 16% and 50%.12
However, the authors of several studies have come to the conclusion that PET is suboptimal for the diagnostic evaluation of the primary CCA tumor because of false-positive results in the setting of inflammation and false-negative results due to the high desmoplastic reaction in these tumors.
PET for HCC: It May Not Be Helpful
18F-FDG uptake in HCC ranges from 38% to 70% with an overall sensitivity of only approximately 60%.13, 14 The sensitivity may be higher with delayed analysis at 2 and 3 hours because of a decrease in the SUV of normal tissue, which increases the tumor/normal (T/N) ratio.13 The sensitivity is higher with larger and poorly differentiated tumors. The reported sensitivities are 27%, 48%, and 93% for tumors of 1 to 2 cm, of 2 to 5 cm, and more than 5 cm, respectively.14 Poorly differentiated HCC accumulates more 18F-FDG, and this results in higher detection rates.
In two studies, the use of PET for the identification of HCC in patients with hepatitis C was evaluated. In the first study, 10 patients with chronic HCV and focal liver lesions were evaluated by conventional ultrasound, CT, and PET scans with subsequent liver biopsy.15 Only two of five HCCs were detectable by PET, so the authors concluded that there was no benefit from PET in this scenario. The second study evaluated eight patients with HCV-associated cirrhosis and elevated serum alpha-fetoprotein concentrations who had negative findings on conventional imaging modalities.16 Two of the patients were later found to have HCC, and neither was identified by a PET scan; this confirmed the lack of utility of PET for the detection of primary HCC.
Although it is not useful in diagnosis, several groups have suggested a role for 18F-FDG-PET as a prognostic factor for HCC. For example, overall survival was correlated with SUV intensities in HCC.17 Similarly, in HCC patients treated with surgical resection with curative intent, a high T/N ratio greater than 2 was found to be an independent predictor of recurrence and poor survival.18 In patients who had undergone liver transplantation, PET positivity, in contrast to PET negativity, was found to be an independent predictor of microvascular invasion, recurrence-free 3-year survival (35% versus 93%), and overall recurrence rates (50% versus 4%). A T/N ratio cutoff of 1.15 was found to be an independent predictor of 1-year progression-free survival on multivariate analysis (94% versus 54%).19
18F-FDG PET detection rates for extrahepatic metastases of HCC as high as 100% have been reported, but with a resulting change in management, the rates have been only 5% to 28%.20 The detection rates are again size-dependent: 83% for metastases larger than 1 cm and 13% for lesions with diameters equal to or less than 1 cm.21 In HCC patients treated with radiofrequency ablation or resection, 18F-FDG PET detected recurrences earlier and at higher rates in comparison with conventional CT.22 However, these observations have not been validated; therefore, 18F-FDG PET has not become a standard of care for either the diagnosis or staging of HCC.
Areas of Uncertainty
PET for Non-HCC Vascular Tumors.
Vascular tumors of the liver include cavernous hemangioma, epithelioid hemangioendothelioma (EHE), hepatic angiosarcoma, and rare cases of hepatic Kaposi sarcoma and spongiotic pericytoma. In particular, the latter two are extremely rare, and no reports describe the use of PET for these tumors. There are no reports on 18F-FDG PET for the evaluation of hepatic cavernous hemangioma. However, because of the high sensitivity and specificity of standard imaging techniques such as CT and MRI and the lack of change in management based on functional imaging, PET does not have a role in the evaluation of these lesions. Only a few studies have reported the use of 18F-FDG PET for the diagnosis of EHE.23 According to these reports, increased 18F-FDG uptake was observed in these tumors, and this allowed detection and evaluation for treatment response. However, there are no large studies evaluating the sensitivity and specificity of PET in these tumors. Thus, 18F-FDG PET cannot be considered standard of care in the diagnosis of EHE. Similarly, only two case reports describe increased 18F-FDG uptake in hepatic angiosarcoma, so the value of PET for this malignancy remains to be determined.24, 25 In summary, several vascular tumors of the liver tend to accumulate 18F-FDG, which allows their detection by PET, but at this time, PET cannot be recommended for the evaluation of these lesions.
PET for Neuroendocrine Tumors.
The current standard for functional imaging of neuroendocrine tumors is somatostatin receptor scintigraphy (SRS).26 Several studies have evaluated PET versus SRS. The overall sensitivity of PET was 58% versus 89% for SRS. The sensitivities for the detection of the primary tumors varied with the location. The sensitivity for the detection of ileal tumors was 91% with SRS but only 36% with PET, whereas for pancreaticoduodenal tumors, the sensitivities were 90% and 79%, respectively. The sensitivity for the detection of hepatic metastases with an unknown primary neuroendocrine tumor was 100% with SRS versus 86% with PET. In a direct comparison with PET, SRS showed significantly higher detection rates for hepatic and osseous metastases of neuroendocrine tumors but not for lymph node metastases. Only in neuroendocrine tumors with a proliferation index greater than 15% did PET outperform SRS with a sensitivity of 92% versus 69%.26 On multivariate analysis, 18F-FDG PET was found to be a predictor for progression-free survival.27 The predictive value of PET for progression-free survival was confirmed in another study including 96 patients with neuroendocrine tumors, of which 66% had hepatic metastases; a SUVmax value greater than 3 was found to be an independent predictor of progression-free survival on multivariate analysis.28 Thus, PET has only a complementary role in the detection of pancreaticoduodenal neuroendocrine tumors and poorly differentiated tumors (SRS can occasionally fail to identify the primary tumor) and in the prediction of progression-free survival.
In the current clinical scenario of noncirrhotic HCV-positive liver disease, the differential diagnosis includes HCC, CCA, lymphoma, and benign lesions. The sensitivity of PET for the identification of primary hepatic tumors is not high enough to rule out malignancy. Because the approaches to the management of HCC, CCA, lymphoma, and benign lesions are all radically different, the need for a specific diagnosis is absolute. Therefore, we would not perform a diagnostic 18F-FDG PET scan in this situation. Even if this lesion is HCC on biopsy, the role of a PET scan in determining the prognosis is unclear, and it is not recommended.