Screening for Liver Cancer

• Serum alpha-fetoprotein (AFP) levels are elevated in many patients with hepatocellular carcinoma (HCC). However, AFP measurement is not an ideal test because it lacks both sensitivity and specificity.

• Real-time ultrasonography is the most cost-effective modality for evaluating patients with HCC.

• Modifications in computed tomography have increased its sensitivity in patients with HCC from 50% or 60% to more than 90%.

• In North America, screening for HCC in patients with HCV infection is widely accepted by hepatologists, even though clearer demonstration of effectiveness is not currently available. In general, patients are screened by AFP measurements and ultrasonography every 6 to 12 month

Hepatocellular carcinoma (HCC) is the fifth-leading cause of cancer in men worldwide and the ninth-leading cause in women. It is the most common primary tumor of the liver and is responsible for 1 million deaths per year.^182,183 Its incidence is rising in both the Western world and Japan, and increases in the incidence of HCV-associated HCC have been reported in the United States, Italy, and Japan.

El-Serag¹⁸⁴ and El-Serag and Mason¹⁸⁵ reported that the number of cases of HCV-induced cirrhosis that led to the development of HCC in the United States doubled between 1979 and 1998. Moreover, they may have underestimated the incidence because they studied only biopsy-proven cases of HCV. Similar increases have been seen in Western Europe and especially in Japan. The higher incidence in Japan appears to be associated with an increase in the number of blood transfusions and the use of contaminated needles during and shortly after World War II, which led to the development of hepatitis C-induced chronic liver disease many years later.^186,187

The presence of chronic HCV infection leading to cirrhosis is the necessary precursor in the natural history of HCC. Several series from the United States, Italy, Japan, and France reported annual rates that ranged from 1.2% to 6.9%, although cirrhosis appears to have been the common denominator in most of these cases.¹⁸⁸ There are rare reports of HCC presenting in hepatitis C patients in the absence of cirrhosis.¹⁸⁹

Screening for HCC in North America remains controversial. The U.S. Preventive Services Task Force, for example, does not endorse screening for this disorder. The cumulative costs of testing and evaluation of false positives, combined with the unproven value of detection of early HCC, explain this controversy. The availability of liver transplantation for patients with limited HCC is an additional factor to consider.

Improvements in survival and cost-effectiveness have not been demonstrated in randomized controlled trials of screening in patients with HCV-induced HCC. It is well known that screening low-incidence populations provides minimal benefit. Screening for HCC in patients with HCV infection is widely accepted by hepatologists, even though there is no evidence that it is effective. In general, patients are screened by alpha-fetoprotein (AFP) measurements and ultrasonography every 6 months. This interval was chosen on the basis of tumor-volume doubling time, which ranges from 1 to 20 months (median, 6).¹⁹⁰

Sherman outlined in great detail the components of a good surveillance program (Table 13).¹⁹¹ These may be applicable to patients with HCV infection, considering its potential to progress to HCC.
  

Table 13

Characteristics of a successful surveillance program191

The disease must be common and must cause substantial morbidity and mortality. The target population must be easily identifiable. The surveillance test must carry low morbidity and high sensitivity and specificity. Recall policies must be standardized. The surveillance test must be acceptable to the target population. There must be effective therapy. Surveillance should reduce mortality from the disease.

A recent search of the literature found only one randomized prospective cohort study of screening tests for HCC in patients with chronic HCV infection.¹⁹² In it, Gebo et al found that twice-yearly screening by AFP measurements and ultrasonography can detect tumors smaller than 3 cm. Small tumors, of course, facilitate resection or transplantation.

The recommendation of this advisory board, while consistent with much published literature, will undergo revision as better data become available.

Screening and surveillance for HCC and underlying cirrhosis begins with the recognition that the patient has advanced fibrosis or cirrhosis.¹⁹³ Predisposing factors for the progression of cirrhosis to HCC should be clearly identified by the clinician. Among these factors are age at acquisition of HCV infection, excessive alcohol consumption (>50 g/d), and coinfection with HBV¹⁹⁴ or HIV.¹⁹⁵

The diagnosis of HCC in the Western world usually occurs 3 to 4 decades after the initial acquisition of HCV. The diagnosis is usually based on clinical deterioration, laboratory test results, and radiographic and histologic findings. Clinical manifestations of HCC are sometimes characterized by anorexia, increasing fatigue, weakness, weight loss, abdominal pain, occasional hemoperitoneum, and variceal bleeding. Physical examination may demonstrate significant ascites and an abdominal mass with a prominent bruit.

The most common laboratory test for HCC is measurement of the serum AFP level, which is elevated in many patients with HCC. However, the AFP measurement is not an ideal test because it lacks both sensitivity and specificity. There appears to be no correlation between the AFP level and the size or extent of the intrahepatic disease or the presence of metastases. Patients with well-differentiated HCC and anaplastic lesions tend to have normal AFP levels and small lesions (<2 cm).¹⁹⁶ The presence of a mass lesion in the liver in a patient with cirrhosis and an AFP level greater than 200 ng/mL is considered diagnostic of HCC.¹⁹⁷ Fluctuating AFP levels may be present in patients who have underlying chronic active hepatitis and active cirrhosis. A rise in AFP levels is sometimes problematic for patients with negative imaging studies. These patients should be followed very closely with periodic imaging studies, such as magnetic resonance imaging (MRI) and triple-phase computed tomography (CT).

In addition, there are investigational tests such as measurements of AFP L-3, serum des-gamma-carboxyprothrombin, and prothrombin induced by vitamin K absence II (PIVKA); these levels are elevated in patients with HCC and are more specific. However, recent reports from the United States and Japan suggest increased sensitivity and specificity in patients with liver tumors smaller than 3 cm on PIVKA testing. Cirrhotic patients who present with hypoglycemia, hypercalcemia, hypercholesterolemia, and erythrocytosis should be thoroughly evaluated for the presence of HCC.

The following imaging studies are commonly used in establishing the diagnosis of HCC.

Ultrasonography. Real-time ultrasonography is the most cost-effective modality for evaluating patients with HCC. In the Far East, this modality is commonly used for detecting tumors smaller than 3 cm; in such cases, its sensitivity ranges from 50% to 90%. Doppler ultrasound can be used as an adjunct to real-time ultrasonography, and it can demonstrate the vascularity of the tumor and the patency of the portal and hepatic veins. Power Doppler, a further improvement in technology, can demonstrate the vascular nature of the tumor to a higher degree than can routine Doppler. The drawbacks of ultrasonography are that isodense lesions cannot be visualized and hyperechoic lesions cannot be differentiated from HCC. In addition, tumors smaller than 1 cm cannot be adequately differentiated from benign tumors and hepatic nodules.^198-200

CT. Improvements in technology have led to modifications in CT scanning, which have increased its sensitivity in patients with HCC from 50% or 60% to more than 90%. Helical CT allows the radiologist to scan patients during their arterial and portal venous phases. Because most HCCs derive their blood supply from the hepatic artery, helical CT can demonstrate HCC in more than 90% of cases when all three aspects of the CT scanning procedure (ie, nonenhanced, arterial phase, and portal venous phase) are utilized.^200,201

MRI. MRIs in patients with HCC demonstrate a high-intensity pattern on T₂-weighted imaging. Dynamic MRI, following an intravenous bolus injection of gadolinium diethylenetriamine penta-acetic acid (Gd-DTPA), reveals a high intensity in the early arterial dominant phase and better contrast resolution.

Magnetic resonance angiography. Magnetic resonance angiography demonstrates intrahepatic vascularity as well as vascularity in the inferior vena cava. This technology is advantageous during the staging of HCC. Hepatic angiography, once the gold standard for the radiologic diagnosis of HCC, is now used primarily for the delivery of chemotherapy and chemoembolization. However, in Western Europe and occasionally in some centers in the United States, lipiodol, an iodized oil, is injected into the hepatic artery, and 2 weeks later a CT scan of the liver will demonstrate the accumulation of lipiodol within the HCC.²⁰² This modality can also be used in patients with a rising AFP level and no demonstrable lesion on CT or as a staging procedure prior to surgery or liver transplantation.

Patients with HCV infection and cirrhosis may be screened for HCC by imaging and measurement of the AFP level, although HCC can develop in patients with a persistently normal AFP. The screening process is accelerated in patients with tumors smaller than 1 cm and in patients who experience a progressive increase in AFP levels in the absence of a mass lesion.