Noninvasive tests may fail to detect liver fibrosis, especially among Black adults
Click Here to Manage Email Alerts
Key takeaways:
- Noninvasive tests for liver fibrosis screening performed worse among Black adults vs. other racial-ethnic groups.
- Using the fibrosis-4 index first to screen cirrhosis may lead to providers missing 48% of cases.
Vibration-controlled transient elastography and serum-based noninvasive tests show a significant discrepancy in detecting liver fibrosis, according to study findings published in Obesity.
“The fibrosis-4 (FIB-4) index performs poorly overall to screen the general population with cardiometabolic risk factors for significant liver fibrosis when compared to vibration-controlled transient elastography,” Meagan Gray, MD, assistant professor in the division of gastroenterology and hepatology, department of medicine, at University of Alabama at Birmingham, told Healio. “Furthermore, differences in performance across different racial and ethnic groups may continue to promote health disparities.”
Researchers obtained data from 6,359 adults who participated in the National Health and Nutrition Examination Survey from 2017 to 2020 and had blood samples collected and a vibration-controlled transient elastography performed at the same appointment. Adults were considered to have metabolic dysfunction-associated steatotic liver disease (MASLD), which is the new nomenclature replacing nonalcoholic fatty liver disease (NAFLD), if they had a controlled attenuation parameter of 288 dB/m or more and at least one cardiometabolic risk factor. Advanced fibrosis was defined as a liver stiffness of 9.7 kilopascal (kPa) or higher and cirrhosis was liver stiffness of 13.6 kPa or more. Anthropometric and laboratory measurements were used to calculate noninvasive test scores, including the FIB-4 index, NAFLD fibrosis score and aspartate aminotransferase (AST) to platelet ratio index. NAFLD liver fat score and fatty liver index were also calculated for all participants to estimate hepatic steatosis.
MASLD rates varied by race and ethnicity, with 42% of Hispanic adults having MASLD compared with 36.8% of white adults, 30.1% of Asian adults and 26% of Black adults. Black and Asian adults also had the lowest levels of advanced liver fibrosis and cirrhosis among racial-ethnic groups.
Racial-ethnic disparities with noninvasive tests
Fatty liver index and NAFLD liver fat score overestimated hepatic steatosis compared with controlled attenuation parameter results in all racial-ethnic groups. The overall agreement between fatty liver index and controlled attenuation parameter results for diagnosing hepatic steatosis was 73.4% and NAFLD liver fat score agreed with controlled attenuation parameter results 73.2% of the time.
The three noninvasive tests to estimate liver fibrosis performed poorly at detecting advanced liver fibrosis. NAFLD fibrosis score outperformed FIB-4 index and AST to platelet ratio index. FIB-4 index and AST to platelet ratio index performed worse among Black adults than other racial-ethnic groups.
For predicting cirrhosis, NAFLD fibrosis score performed slightly better than the other two noninvasive tests. All three scores performed worse at predicting cirrhosis among Black adults compared with other racial ethnic groups. FIB-4 index and AST to platelet ratio index performed worse the NAFLD fibrosis score for predicting advanced fibrosis and cirrhosis among Black adults vs. other racial-ethnic groups.
“There is substantial evidence that race and ethnicity affect the development and progression of MASLD,” Fernando Bril, MD, assistant professor in the division of endocrinology, diabetes and metabolism, department of medicine, University of Alabama at Birmingham, told Healio. “However, despite this well-accepted fact, none of the noninvasive tests include race or ethnicity as a covariable in their calculation. Our results suggest that we may need to move toward including this information when noninvasively assessing the risk of liver fibrosis. This will allow us to improve the performance of these scores and make them more generalizable, similar to what happens with other formulas commonly used in clinical practice, such as estimated glomerular filtration rate.”
FIB-4 index misses fibrosis, cirrhosis cases
Of the study group, 78.7% were considered at-risk of liver disease based on having prediabetes, type 2 diabetes, elevated liver enzymes or at least two cardiometabolic risk factors. Noninvasive tests performed similarly among adults at risk for liver disease as they did with the full study group.
If sequential testing were performed with vibration-controlled transient elastography only conducted for adults with a FIB-4 index score of 1.3 or higher, providers would miss 56% of adults with advanced fibrosis and 48% of those with cirrhosis.
“We suspected that the FIB-4 results would vary across race and ethnicities, but did not expect such poor performance of the FIB-4 across all patients,” Gray said. “This test is now being widely recommended as a first-line screening test for liver fibrosis for patients in primary care clinics. The results are worrisome that we may be missing many patients with significant liver disease by using this as the initial screening test of choice.”
Bril said researchers need to examine developing new noninvasive tests or modifying old ones to include race and ethnicity as a covariate.
“It will be important to do this in studies that have histology as the gold standard,” Bril said.
For more information:
Fernando Bril, MD, can be reached at fbril@uab.edu.
Meagan Gray, MD, can be reached at meaganegray@uabmc.edu.
Perspective
Back to Top
The authors examine the important issue of potential racial-ethnic disparities in the performance of commonly used diagnostic panels in the NHANES 2017-2020 cohort. Of particular relevance are the results related to fibrosis-4 (FIB-4) index, derived from age, plasma aspartate aminotransferase (AST) and alanine transaminase (ALT), and platelet count (www.mdcalc.com/calc/2200/fibrosis-4-fib-4-index-liver-fibrosis), because FIB-4 has been adopted by all medical societies as first-line screening, especially to identify adults at high-risk of advanced fibrosis (stage F3) or cirrhosis (stage F4), such as those with metabolic syndrome and/or type 2 diabetes. This recommendation is based on its extensive validation, low-cost and potential to be broadly applied across health care settings, reasonable specificity and negative predictive value and recent data as predictive of future liver (even cardiovascular) outcomes.
Complaints of poor performance of the FIB-4 are well-known and need to be balanced with the lack of prospective, long-term cost-effectiveness studies to support recommending proprietary blood or imaging tests across large populations at risk. Taken together, FIB-4 is a reasonable first step for risk-stratification. An early diagnosis is important to promote lifestyle intervention and pharmacologic treatments used for obesity or type 2 diabetes, such as a GLP-1 receptor agonist or pioglitazone, to reverse or halt steatohepatitis progression (Genua I, et al. Diabetes Spectr. 2024;doi:10.2337/dsi23-0012).
Within this context, the work by Bril and Gray is a welcome call to fully understand the role of ethnicity in the performance of FIB-4, in addition to known caveats of its poor sensitivity and age-dependent performance (best between age 35 to 65 years). The authors report that while FIB-4 performed similarly among adults of white and Hispanic ancestry, and possibly even better among non-Hispanic Asians, it appeared to underperform in non-Hispanic Black adults. Of note, FIB-4 did not perform worse in detecting advanced fibrosis (stage F3-F4 vs. stage F0-F2) in the overall population or in those at higher risk across ethnic groups, but it did so in non-Hispanic Black adults with cirrhosis (F4).
Overall rates of cirrhosis were low, ranging from 1.2% in Asian and 1.5% in non-Hispanic Black adults (about 25 patients) to 2.2% and 2.8% in white and Hispanic adults, respectively. Why?
The small number of events calls for further confirmation of these findings. However, in this study, performance of FIB-4 was maintained in all ethnic groups with cirrhosis, except in non-Hispanic Black adults, where it worsened, causing the apparent diagnostic discordance. In cirrhosis, a decrease in platelets due to portal hypertension and hypersplenism, further increases FIB-4, even as ALT and AST decline. FIB-4 may have underperformed among non-Hispanic Black adults because they usually have less steatosis, a driver of plasma aminotransferase elevation and often a higher FIB-4, even when steatohepatitis is present (Bril F, et al. Diabetes Care. 2018;doi:10.2337/dc17-1349).
More likely, this discordance is explained by the fact that liver stiffness measurement by transient elastography, used in the study as the “gold-standard” to which FIB-4 was compared, may read higher levels of fibrosis with worsening obesity. Non-Hispanic Black adults had the highest BMI among ethnic groups, possibly leading to higher liver stiffness measurements and falsely higher rates of cirrhosis not tracked by FIB-4. That no patient had a platelet count below 150 per microliter of blood, typical of cirrhosis, appears to reinforce this hypothesis for non-Hispanic Black adults.
This study offers important lessons. First, that non-Hispanic Black adults are not immune to developing advanced fibrosis and cirrhosis, an entrenched perception by clinicians that leads to underdiagnosis and undertreatment.
Second, having diabetes overall doubled the prevalence of advanced fibrosis and tripled that of cirrhosis, irrespective of racial-ethnic background. This calls clinicians to have a high level of suspicion about the risk for cirrhosis in any ethnic group with diabetes.
Third, non-Hispanic Black adults are still a multiethnic and heterogeneous group with likely a variable genetic risk, simplistically placed into a single category. Thus, a “one fits all” approach should be avoided.
Fourth, socioeconomic differences exacerbate racial-ethnic disparities and should always be explored as these impact disease prevalence and outcomes. Finally, practice guidelines are not meant to exempt practitioners from using clinical judgment.
Thus, interpreting FIB-4 should always be done within the context of a patient’s medical history, socioeconomic context and overall risk for steatohepatitis. The presence of obesity or diabetes should always tip the clinician to more strongly consider whether a secondary test to assess cirrhosis risk may be needed. Future guidelines may consider imaging as a first-line test in very high-risk groups, such as those with obesity and type 2 diabetes, but stronger evidence is needed to this end. The work by Bril and Gray is an invitation toward a more individualized and thoughtful approach to cirrhosis prevention.
Collapse
Click Here to Manage Email Alerts