Controversy over optimal surveillance methods for hepatocellular carcinoma continues to drive ongoing investigation into improved biomarkers and imaging techniques.
Controversy over optimal surveillance methods for hepatocellular carcinoma (HCC) continues to drive ongoing investigation into improved biomarkers and imaging techniques. Recent screening studies have focused on the utilization of alpha-fetoprotein (AFP) with ultrasound and abbreviated MRI approaches.
An article published in the American Journal of Gastroenterology demonstrated that AFP may be a helpful complement to traditional ultrasound-based surveillance programs, one that is particularly useful among patients with viral-induced cirrhosis.1 Additionally, MRI and CT are occasionally employed as a means of increasing HCC detection sensitivity among high-risk patients in place of conventional ultrasound. According to one recent study, per-patient sensitivities for HCC detection were 64%, 76%, and 85%, for ultrasound, CT, and MRI, respectively.2
Although alternative imaging techniques offer the potential for improved sensitivity, limiting factors often prevent their widespread use as an HCC surveillance tool. For instance, “despite higher diagnostic accuracy than ultrasound and CT,” explained Robert M. Marks, MD, from the Naval Medical Center San Diego, “the greater expense, long examination times, and relatively limited availability remain a challenge to using MRI in HCC surveillance.”3 Due to the factors limiting CT and MRI approaches, “their roles as surveillance tools for detecting early-stage HCC in patients with cirrhosis remain uncertain,” noted Te-Sheng Chang, MD, PhD, and colleagues in their recent report on the addition of AFP.1 The use of MRI, for instance, may be better suited to confirming positive surveillance results, rather than screening for HCC initially.3
In answer to these limitations, a recent study by Marks and colleagues in the American Journal of Roentgenology attempted to model an abbreviated MRI examination protocol and assess its potential as an HCC surveillance tool.3 With a gadoxetic acid-enhanced hepatobiliary phase and T2-weighted single-shot fast spinning echo (SSFSE), the abbreviated protocol was able to maintain the high per-patient sensitivity and negative predictive value of the MRI technique, while minimizing logistic limitations, such as cost and procedure time.3
“We found that per-patient sensitivity for the detection of HCC with an abbreviated MRI examination was similar to the reported sensitivity for the complete study with gadoxetic acid,” wrote the authors.3 According to their findings, the abbreviated protocol reached 82.6% sensitivity, which is not only comparable to the complete MRI, but also considerably greater than that reported in the literature for ultrasound-based surveillance.4,5
Resources such as the abbreviated MRI procedure may be useful for improving surveillance sensitivity and identifying patients requiring follow-up with more costly and time-intensive imaging techniques. With the potential to reduce the current 20 minute to 40 minute procedure time to a 15-minute time slot and 5-minute scan, the authors estimated that switching from a complete MRI to an abbreviated MRI surveillance program may afford a total cost savings of nearly 31%.3 Overall, the study by Marks and colleagues concluded that the abbreviated MRI technique may be beneficial for centers currently utilizing full-length MRI for HCC surveillance.3 The authors went on to stress that additional studies are still necessary to determine the relative cost-to-benefit ratio of utilizing this technique compared with the conventional ultrasound approach.3Until alternative surveillance modalities, such as the abbreviated MRI, are thoroughly validated and widely available, methods for improving ultrasound sensitivity are of critical importance. In the absence of alternative modalities, several groups have investigated the utility of serum biomarkers, such as AFP, as potential surveillance tools. By itself, however, AFP has proved suboptimal in detecting HCC.6—8
“There have been false-positive results, as elevated AFP also occurs in chronic liver disease and other malignancies,” explained Chang and colleagues in their recent report.1 False-negative results may also occur when using this biomarker, as not all HCCs secrete AFP. “This is particularly veritable for early- stage HCCs, which are the targets of surveillance,” added the authors.1
Due to these potential limitations, the American Association for the Study of Liver Diseases (AASLD) and the European Association for the Study of the Liver (EASL) recently discouraged the use of AFP as a surveillance modality for HCC.9—11 Despite this, interest remains as to whether the biomarker may still provide some benefit when used as a complement to conventional ultrasound.1,12
Since ultrasound is a subjective, highly operator- dependent technique, Chang explained that “the detection of increased AFP levels during surveillance may alert the surveyor and thus trigger a more intense survey.”1
In light of the controversy over the benefits of AFP in HCC surveillance, the performance of these modalities in a real-world practice setting was examined through the comparison of ultrasound-based surveillance or AFP measurement alone versus a combined approach.1
Investigators retrospectively evaluated patients with cirrhosis who were screened for HCC at a tertiary referral hospital in Southern Taiwan between January 2002 and July 2010. The median duration of follow-up in the study was 4.75 years.1
Of the 1597 patients analyzed over the follow-up period, 22.7% developed HCC. With a cutoff value of 20 ng/mL, the use of AFP measurement alone resulted in a sensitivity of 52.9% and a specificity of 93.3%. Comparatively, the sensitivity and specificity of ultrasound surveillance were 92.0% and 74.2%, respectively.1
Interestingly, investigators found that by combining ultrasound and AFP, the sensitivity of ultrasound alone was significantly improved from 92.0% to 99.2%. Unfortunately, specificity was simultaneously lowered from 74.2% to 68.3%.1
Further investigation demonstrated, however, that requiring an increase in AFP levels greater than twice the nadir during the previous year in addition to the 20 ng/mL threshold was able to increase the specificity of combined ultrasound and AFP surveillance from 68.3% to 71.5%. In addition, the improved sensitivity of 99.2% was maintained.1
“Our results support the concept that an increased AFP level is a risk factor for HCC development,” Chang and colleagues wrote. Moreover, compared with individuals who did not develop HCC, patients developing the disease exhibited greater AFP levels within the 12 months leading up to HCC onset.1
These findings were similar to results published in Cancer Epidemiology, Biomarkers, and Prevention by Amit G. Singal, MD, and colleagues from the University of Texas Southwestern Medical Center and the University of Michigan, in May 2012. This study demonstrated a significantly increased HCC surveillance sensitivity when combining ultrasound and AFP in a real-world clinical setting, with a minimal loss of specificity versus ultrasound alone.12The improved sensitivity of combined ultrasound and AFP versus ultrasound alone remained significant in the subsets of patients with hepatitis B (HBV) or hepatitis C virus (HCV) infection, as well as in HBV and HCV co-infected individuals.1 In contrast, the finding did not hold true for cirrhotic patients who were negative for both HBV and HCV.1 Investigators concluded that the benefits of combined ultrasound and AFP screening were not as marked in patients with nonviral cirrhosis.1 As most of these patients had alcohol- induced cirrhosis, the authors concluded that “AFP measurement is not as effective as a complementary method to ultrasound for HCC surveillance in western countries, where most liver cirrhosis is related to alcoholism.”1 Despite the AASLD and EASL guidance that AFP should not be used as a surveillance modality, the findings of Chang and colleagues illustrate that use of this biomarker may still be relevant as a complement to ultrasound.1 The combined approach may be able to increase surveillance sensitivity and improve HCC detection in patients with cirrhosis, particularly among those with hepatitis-driven disease.1
“Despite considerable efforts in developing new serum biomarkers for detecting or diagnosing HCC,” noted Chang, “none have been proven superior to AFP.”1,13 As such, the authors concluded that “until a superior alternative biomarker is developed, AFP measurement should be used in combination with ultrasound for HCC surveillance… [especially in] those with viral hepatitis-induced cirrhosis.”1HCC is the most common form of cancer affecting the liver, driven mainly by viral hepatitis infection or alcohol-induced cirrhosis. This malignancy is the third leading cause of cancer deaths worldwide, with more than 500,000 individuals currently affected.14 Predicted to be a major health concern in future decades, data from the Surveillance Epidemiology and End Results program indicate that HCC is the fastest-growing cause of cancer mortality in the United States.15—17 In fact, disease incidence has more than doubled during the last two decades, with a near twofold increase in the HCC mortality rate over the last 5 years alone.14
Patients diagnosed at advanced stages are often precluded from potentially curative treatment options such as surgical resection or liver transplantation, making early detection a priority for improved prognosis.18 Due to the benefits of early detection on patient survival,7 the AASLD and the EASL currently encourage HCC surveillance as a measure to promote early detection in high-risk patients with chronic liver disease.9—11
As a low-cost and readily accessible technique, most organizations recommend the use of ultrasound for HCC surveillance. Unfortunately, per-patient sensitivity is reportedly insufficient with this method, falling below 65% when detecting small or early-stage HCCs.4,5 As a result, the search for alternative methods to improve HCC surveillance is an important area of ongoing research.