GHENT, Belgium -- Noninvasive serum N-glycan profiling might unmask undiagnosed hepatocellular carcinoma in HBV-infected patients with cirrhosis, investigators here said.
GHENT, Belgium, Aug. 9 -- Noninvasive serum N-glycan profiling might unmask undiagnosed hepatocellular carcinoma in HBV-infected patients with cirrhosis, investigators here said.
Analysis of two different N-glycans found that one was elevated in patients with hepatocellular carcinoma and the other was elevated in patients with cirrhosis, Cuiying Chen, Ph.D., of Ghent University, and colleagues reported in the August issue of Hepatology.
The log ratio of the two peaks distinguished between patients with liver cancer, cirrhosis or fibrosis, and healthy controls, they said.
About 60% to 80% of liver cancers are preceded by cirrhosis, so screening cirrhosis patients for early-stage carcinoma could reduce mortality, the authors noted.
Imaging techniques used to diagnose hepatocellular carcinoma cannot distinguish between it and benign hepatic lesions. Serum ?-fetoprotein (AFP) is widely used in the diagnosis, but the sensitivity and specificity of the test vary greatly.
Because serum N-linked glycoproteins are synthesized by the liver and B-lymphocytes, changes in serum total N-glycans could reflect changes in liver or B-lymphocyte physiology, the authors hypothesized.
Altered N-linked sugar chains occur in various types of tumors, providing a rationale to evaluate serum N-glycan fingerprinting as a means to diagnose cancer in patients with cirrhosis induced by hepatitis-B infection, they said.
So they examined the diagnostic potential of N-glycan profiling in 440 HBV-infected patients from four hospitals in China. The study population comprised 143 patients with liver fibrosis, 80 patients with cirrhosis alone, and 147 patients with cirrhosis and hepatocellular carcinoma. Liver cancer was diagnosed by various means, including biopsy, autopsy, surgical specimens, imaging studies, and AFP measurement.
The investigators examined the N-glycan profile by means of DNA sequencer-assisted fluorophore-assisted carbohydrate electrophoresis (DSA-FACE). Blood samples from 130 healthy volunteers served as a control. Each peak in the profile from the patients was quantified and compared against the peaks of the control group.
To be able to detect cancer against a cirrhosis background, "we focused on identifying glycan structures whose abundance would not increase in patients with cirrhosis but would be elevated in [cancer] patients," the authors stated.
They found that one peak (NA3Fb, peak 9) was significantly elevated in patients with carcinoma (P<0.0001). A second peak (NA2FB, peak 7) was significantly lower in cancer patients than in patients with cirrhosis (P<0.0001).
Moreover, they said, the log ratio (peak 9/peak 7) was significantly elevated in patients with cancer compared with cirrhosis patients, fibrosis patients, and controls (P<0.0001).
By means of receiver operating characteristic curve analysis, Dr. Chen and colleagues determined that the log ratio (peak 9/peak 7), which they have named the GlycoHCCTest, could distinguish hepatocelllular cancer patients from those with cirrhosis with 81% accuracy.
The accuracy of the test compared favorably with the 78% diagnostic accuracy of AFP, they noted. Moreover, a cutoff value of -0.34 identified liver cancer with 88% specificity and 57% sensitivity, similar to the values associated with AFP at a cutoff of 100 ng/ml. The test generated a constant value in the patients with fibrosis, regardless of stage, indicating specificity for cancer.
The concentration of the two glycans and the log ratio (peak 9/peak 7) predicted tumor stage.
Dr. Chen and colleagues subsequently defined their test as the ratio of the two glycans. According to information provided by the Flanders Interuniversity Institute of Biotechnology (a research institute affiliated with Ghent and three other Belgian universities), the N-glycan profile test detects about 50% of hepatocellular carcinomas that are missed by AFP. Using the tests in tandem results in diagnostic accuracy exceeding that of either test alone.