Serum Hyaluronic Acid (HA): A Potential Reflexive Test for Further Evaluating Individuals with Chronic Hepatitis B and C, NASH and Other Liver Pathologies
Why am I writing this paper?
In underwriting, we are seeing an increasing number of cases of chronic hepatitis C and nonalcoholic fatty liver disease (especially nonalcoholic steatohepatitis, NASH) wherein the proposed insured has not had a sufﬁcient clinical workup to determine if advanced liver pathology (signiﬁcant ﬁbrosis, cirrhosis) is present.
This is a key determination regarding insurability and the absence of this information has forced us to take uncertain actions…certainly insuring some who are bad risks while postponing others who are insurable.
Serum hyaluronic acid (also known as hyaluronate, hyaluron or simply HA) appears to be a potentially affordable marker for the noninvasive detection of high grade ﬁbrosis and cirrhosis in these and other liver pathologies.
What is the clinical perspective on this issue?
“Progressive hepatic ﬁbrosis with the development of cirrhosis is a feature of almost all chronic liver diseases.”
Nezam H. Afdahl
Chief of Hepatology
Harvard Medical School
The American Journal of Gastroenterology
“One of the major problems facing the hepatology and gastroenterology community is how to best evaluate and manage the increasing numbers of patients with hepatitis C virus… However, little progress has been made in improving either our ability to determine the degree of hepatic injury, particularly ﬁbrosis, or to predict the risk of disease progression for the individual patient. This information still requires an old fashioned liver biopsy.”
Nezam H. Afdahl
“Diagnosing Fibrosis in Hepatitis C: Is the Pendulum Swinging From Biopsy to Blood Tests?”
Are there a number of alternative choices out there for this role?
Yes, several multi-test protocols have been recommended, such as a panel dubbed FIBROTEST. However, these proﬁles require as many as 5 or more test components, some of which may be expensive or difﬁcult to access for our purposes.
Of all the individual markers feasible for our purposes, HA appears to be the best.
What is the basic viewpoint on HA held by its advocates?
“Serum HA is currently the best available marker of hepatic ﬁbrogenesis.”
University of Sydney
New South Wales, Australia
Journal of Gastroenterology and Hepatology
“The greatest clinical utility of HA may be its ability to exclude patients with signiﬁcant ﬁbrosis and cirrhosis….an HA level of < 60 ng/L was found to exclude patients with cirrhosis and signiﬁcant hepatic ﬁbrosis with predictive values of 99% and 93%, respectively.”
Nezam H. Afdhal
The American Journal of Gastroenterology
What is HA?
It is a protein, manufactured by liver stellate cells and “is considered to be a factor contributing to ﬁbrogenesis…the increase in HA production may reﬂect the…the synthesis of extrecellular matrix due to inﬂammation.” (1)
“Elimination of hyaluronic acid from the circulation is a speciﬁc receptor-mediated mechanism on hepatic sinusoidal endothelial cells. These specialized cells are thought to play an important role in the maintenance of liver function.” (2)
In other words, it plays a key role in setting the stage for cirrhosis.
How is HA measured?
There are several methods reported, including radioimmunoassay (RIA) and sandwich ELISA.
RIA involves the handling of radioactive materials and is being phased out.
We use an ELISA test now for HIV-1 screening, so this methodology has been deemed affordable in our realm.
This underwriter has tried to track down prevailing prices for performing HA clinically, only to discover that it is not readily available (perhaps not available at all) from clinical labs in the United States (3) However, it is available, and apparently used regularly, in Germany and perhaps elsewhere in Europe. (4)
I will continue this effort to get more information on testing methodologies and projected costs.
Are there special considerations that limit this test?
Not that I have discovered thus far.
There was a question of the testing being superior when done fasting, but another paper maintains that post-prandial testing is adequate. (5)
HA is not affected by pregnancy. It also does not appear to vary by patient age. (6) However, we have no information as yet regarding the potential impact of smoking, obesity (which is a prevalent component in nonalcoholic fatty liver disease), exercise and other potential confounders.
Does HA correlate with disease progression?
Xu et al. in Shanghai report that HA levels increased with progression of ﬁbrosis-to-cirrhosis. (7)
It does not appear that HA correlates with necroinﬂammatory disease, only with its sequelae (ﬁbrosis and cirrhosis). On the other hand, the aminotransferases (ALT, AST) are more reliable in the inﬂammation and cell death phases of chronic liver disease than they are as markers for ﬁbrosis and cirrhosis, so HA would seem to be a logical potential reﬂexive test for subjects with elevated aminotransferases who have proven or strongly suspected chronic liver pathology.
What do we know so far about HA in chronic hepatitis C?
A good amount, based on clinical studies of cohorts of HCV patients:
“In chronic hepatitis C, serial hyaluronic acid levels have been advocated as a means to monitor disease progression and to limit the need for follow-up biopsy.”
Martin G. Phillips
Department of Gastroenterology
Norfolk and Norwich University Hospital, UK
European Journal of Gastroenterology and Hepatology
Guechot (Paris) evaluated 326 untreated HCV patients. They found HA superior to other markers. Sensitivity was 64.5% and speciﬁcity was 91.2% for discriminating signiﬁcant ﬁbrosis from no ﬁbrosis/mild ﬁbrosis. For cirrhosis, sensitivity was 79.2% and speciﬁcity was 89.4%. The authors concluded that “HA should be preferred when monitoring liver ﬁbrosis in patients with chronic viral hepatitis C.” (8)
Wong (UK) looked at 130 asymptomatic HCV carriers (a risk we often see, if only by inference). ALT had no correlation with ﬁbrosis/cirrhosis. HA levels went from means of 17 in stages 1 + 2 ﬁbrosis to 30 in stage 3 and 350 in stage 4 and cirrhosis, using their criteria. They advocated using HA “…to monitor patients at risk for progressive ﬁbrosis…” (9)
McHutchinson (Scripps Clinical, California) followed 486 chronic HCV patients.
Predictive values were in the 90%+ range. For ﬁbrosis, mean values were 179 micrograms/L if present and 62 if absent. For cirrhosis, 382 if present and 110 if absent. Both had p values of 0.001, which means these are highly signiﬁcant differences. (10)
Murakawi (Japan) tested 169 chronic HCV patients and found HA to be the #1 test in terms of “…grading patients in chronic hepatitis C.” Using a cut-off of 60 ng/mL provided a sensitivity of 77% and a speciﬁcity of 74%. (11)
Jara (University of Miami) compared 130 subjects with chronic HCV to 50 controls:
|Stage 1 ﬁbrosis||
|Stage 2 ﬁbrosis||
|Stage 3 ﬁbrosis||
An HA < 70 excluded 96% of cases free of advanced (stage 3) ﬁbrosis or cirrhosis. (12)
How does HA function in cases of medically-treated chronic HCV?
On balance, as well as in untreated cases (above).
Guechot (Paris) followed 52 patients before and after interferon therapy. (13)
- If ﬁbrosis progressed, HA rose signiﬁcantly
- If ﬁbrosis regressed, HA fell
- If ﬁbrosis remained stable, so did HA
- ALT had absolutely no correlation with ﬁbrosis/cirrhosis
Ishibashi (Japan) found that HA reached normal levels within a year in subjects in sustained complete remission after Rx. (14)
Patel (Duke University Medical School, US) reported that sustained virological responders had “a signiﬁcantly reduced HA compared to non-responders.” (15)
How does HA stack up in chronic hepatitis B (HBV)?
Chen (Guangzhou, China) tested 350 chronic HBV patients and said that HAA was “…the best marker for diagnosing compensated liver cirrhosis.” He and his colleagues tested 350 subjects and a cutoff of 119 ng/mL provided a sensitivity of 87% and a speciﬁcity of 68%, with a 95% negative predictive value. (16)
Portuguese researchers looked at 111 consecutive chronic HBV carriers (a scenario we encounter often) and found that “…serum hyaluron concentration correlates with hepatic ﬁbrosis…” (17)
Park in Korea reported on 100 cases of chronic HBV, wherein a cutoff of 77 ng/mL had a sensitivity of 81% and a speciﬁcity of 87% as a marker for ﬁbrosis or worse. (18)
What was discovered in studies with patients who have a mixture of chronic liver diseases?
Ramadori (Germany) examined 133 patients with various chronic liver pathologies, as compared to 22 with acute hepatitis and 50 disease-free controls.
HA was elevated in 32% with chronic active hepatitis C and 91% with proven cirrhosis. It was also elevated in the patients with acute hepatitis (whom we would postpone), but not in the controls. They found the HA results to be statistically signiﬁcant. (19)
Murakawi (Japan) used the sandwich ELISA in 115 chronic hepatitis patients. HAA levels were closely correlated with the severity of ﬁbrosis. HA was the best performing of the ﬁbrosis markers they studied. (20)
Oberti (France) evaluated 243 patients with chronic liver disease. At a cutoff of > 59 ng/mL, HA had a sensitivity of 97% and a speciﬁcity of 73% for compensated cirrhosis and it was labeled “…the most sensitive screening variable for screening.” (21)
Plevris (Scotland) evaluated 221 patients with various chronic liver pathologies.
This is how the mean levels of HA turned out: (22)
|Necroinﬂammatory disease only||
To avoid redundancy, we will not cite two additional papers wherein the same results were shown. (23)
Is HA effective in NASH?
“HA had a signiﬁcant positive correlation with the stage of ﬁbrosis in patients with NASH. Serum HA increased the level of prediction when added to an existing liver diagnostic model for the presence of severe ﬁbrosis.”
34,4,Supplement 1(2003):503A(abstract #715)
Hashimoto (Tokyo) found that a cut-off of 51 ng/mL provided a sensitivity of 88% and a speciﬁcity of 97% in distinguishing grades 3 and 4 ﬁbrosis from subjects with lesser degrees of liver damage in nonalcoholic fatty liver disease. (24)
Although most cases of primary biliary cirrhosis are uninsurable, does HA serve a useful purpose here, too?
Guechot (Paris) found a strong correlation between degree of ﬁbrosis, as did both Nyberg (Sweden) and Remmel (Estonia). (25)
Is there an equally strong link between HA and advanced chronic alcoholic liver disease?
Every bit as strong.
"…serum hyaluronic acid increases at an early stage of ALD, rising continuously as liver disease progresses. Because of its tight correlation with the degree of ﬁbrosis and a high sensitivity for detecting early histological changes due to alcohol, serum hayluronic acid concentrations should be considered an important marker for noninvasive prediction of ﬁbrosis and ﬁnally cirrhosis in alcoholics.”
Flex Stickel et al.
University of Erlangen-Nuremberg
European Journal of Gastroenterology and Hepatology
Pares reports an excellent correlation, paralleling extent of liver damage in 45 chronic alcoholics in Barcelona. (26)
Tsutusami (Japan) discovered that “serum HA concentrations in patients actively drinking with ALD were signiﬁcantly higher (p <0.001) than those with non-ALD.” After 4 weeks abstinence, the HA values normalized. (27)
Does HA have any association with non-liver diseases of interest to us in underwriting?
I found one study where HA was of use in detecting progression of disease in a cohort of 82 patients with rheumatoid arthritis. The authors said that “baseline HA level correlated with measures of disease activity, including swollen and tender joints…ESR, CRP…” After a year, those who developed joint erosions has signiﬁcantly higher levels. (28)
Does this link between liver disease and RA compromise its use in underwriting?
Not really. We should be able to distinguish the RA cases…and RA is probably underpriced in most manuals!
Does one assessment sum up these ﬁndings?
Try this one:
Christophe Pilette et al. in France compared available ﬁbrosis markers. In 243 patients with chronic liver diseases, they found the correlations of all other tests were inferior to the correlation of HA, concluding their paper by saying that “hyaluronate is the serum marker which is the most highly correlated with liver ﬁbrosis.” (29)
How do the statistical benchmarks (sensitivity, speciﬁcity, positive and negative predictive values) sort in terms of the potential underwriting value of HA?
The answer lies in what we do.
We sort risks.
We do not make diagnoses or treat the sick!
If we could rule out the presence of uninsurable degrees of ﬁbrosis and the presence of cirrhosis, we would be better able to assess risk. Clearly, the negative predictive values cited above are very encouraging in this regard.
This point is made by Phillips from a clinical perspective:
“…a low hyaluronic acid level in combination with a low clinical probability of ﬁbrotic disease could reassure clinicians that a biopsy is not required.” (30)
Our existing alternatives – basically, the aminotransferases – are incredibly inferior to HA in terms of separating these risks in the absence of (recent) biopsy on an MD report.
How might we use HA in underwriting, assuming it is available and affordable?
It is NOT a suitable candidate for a screening test.
At least one major company has shown that screening for hepatitis C has signiﬁcant protective value. Being able to get a handle on the extent of disease in screened cases would allow us to postpone the worst cases for further clinical assessment and perhaps make offers on the remainder.
HA would be the PERFECT “conﬁrmatory” test when we have a positive anti-HCV or Hepatitis B surface antigen in subjects reﬂexed due to liver enzyme elevations. By doing serology and HA, we would be better able to integrate the ﬁndings with what else is known and make better offers wherever possible.
We have nothing at this time to get a ﬁrm underwriting handle on the rising number of nonalcoholic liver disease cases unless – as is the distinct exception – a biopsy is done. This is especially true for the most advanced state within this category, better known as NASH. The AST:ALT ratio helps in this regard but it would be enhanced with HA testing on the same specimen.
What is the next step?
Petition the industry labs to further investigate the potential for using HA as a reﬂexive test, especially as regards:
- The availability of technology (which should not be a problem if the test is done widely in Germany)
- Whether the test can be done in-house or would have to be shipped out to a specialty lab
- Assuming the test can be done, determining at what price it could be made available as a reﬂexive test
I hope you found this to be of some interest.
Maybe we have something here.