Abstract
WE ARE GOING DOWN THE WRONG PATH! THE CURRENT AASLD GUIDANCE MISCLASSIFIES MAJORITY OF PATIENTS WITH MASLD AND SIGNIFICANT FIBROSIS
Background:
Metabolic dysfunction-Associated Steatotic Liver Disease (MASLD) has extremely high prevalence in the US and globally with increasing incidence of morbidity and mortality in the population. Effective strategies to identify patients at risk of future complications, namely those with stage 2 (F2) or more advanced fibrosis represent the critically important first step in mitigating the public health impact of MASLD. Essentially all of the society guidelines, including the latest guidance from AASLD, recommend using FIB-4 <1.3 to define low risk subjects not requiring further evaluation. We evaluate the diagnostic performance of FIB-4 in detecting F2 and F3 in the US general population, in comparison to the SAFE (Steatosis-Associated Fibrosis Estimator, Hepatology 2023;77:256) score, recently developed to detect F2 in primary care.
Methods:
We queried the most recent (2017-2020) National Health and Nutrition Examination Survey (NHANES) data, which included transient elastography. We selected adult (≥18 years) examinees with a controlled attenuation parameter (CAP) score ≥274 dB/m and without viral hepatitis or significant alcohol use. Individuals with incomplete elastography results or other missing data were excluded. Liver stiffness (LSM) of 8 kPa was taken as the threshold for F2 and 12 kPa for F3. The FIB-4 and SAFE scores were calculated using the published formulae.
Results:
The NHANES data projected to 73 million(M) US civilian adults with MASLD, of whom 12.2M (16.6%) had LSM ≥8kPa and 4.3M (5.9%) had LSM ≥12kPa.
As seen in the Table, FIB-4 classified 55.5M, 16.7M and 1.2M US adults as low- (<1.3), intermediate-, and high-risk (>2.67), respectively. Of those, 15.7%, 17.8% and 44.5% had LSM ≥8kPa and 4.9%, 7.5%, and 30.4% had LSM≥12kPa, respectively. More strikingly, of subjects with LSM ≥8kPa, 8.7M (71%) subjects would have FIB-4 <1.3 and thus be excluded from further assessment. Similarly, of subjects with LSM ≥12 kPa, 2.7M (63%) would have FIB-4 <1.3.
The SAFE score classified 34.4M, 26.0M and 13.0M subjects as low-(<0), intermediate-, and high-risk (≥100), respectively. Of those, 8.2%, 17.7% and 36.7% had LSM ≥8kPa and 1.8%, 5.2%, and 18.2% LSM ≥12kPa, respectively. SAFE performed better than FIB-4 in excluding fibrosis in the low-risk category. Of subjects with LSM ≥8kPa, 23% would have SAFE<0 and of those with LSM ≥12 kPa, 14% would have SAFE<0.
As in other validation studies, SAFE had higher Area Under the Receiver Operating Characteristic curve in detecting F2 than FIB-4 (0.70 versus 0.58).
Conclusion:
When applied to population-based data representative of the US, the current guidance recommendation based on the FIB-4 score would miss the majority of patients with F2 and even F3. On the other hand, the SAFE score, designed to detect F2 in primary care, more accurately risk-stratify subjects with MASLD in the population.