NAFLD: A Practical Approach to Diagnosis and Staging
NAFLD: A Practical Approach to Diagnosis and Staging
A diagnosis of NAFLD requires that there is evidence of hepatic steatosis on imaging or histology, and other causes of liver disease or steatosis have been excluded. NAFLD is usually asymptomatic, so diagnosis usually follows the incidental finding of abnormal liver enzymes or steatosis on imaging. If abnormal LFTs are present, this is usually mildly raised transaminases (ALT> aspartate transaminase (AST)) and/or gamma-glutamyltransferase. However, ~80% of patients have normal-range ALT levels (males <40 IU/L and females <31 IU/L), and even if elevated, the ALT typically falls (and AST may rise) as fibrosis progresses to cirrhosis. ALT values do not correlate with histological findings and are unhelpful in both the diagnosis of NAFLD and determining disease severity. Clinicians overly rely on abnormal liver enzymes to identify patients with NAFLD, so patients with significant liver disease can be overlooked, potentially missing opportunities for intervention. It has been repeatedly shown that 70–80% of subjects with central obesity and 50–80% of patients with type 2 diabetes have evidence of NAFLD on imaging. Therefore, a new approach is needed to use metabolic risk factors to identify subjects with NAFLD/NASH rather than relying on liver enzyme abnormalities.
In patients with abnormal LFTs, alternative causes of liver disease (or cofactors) should be excluded, including alcohol excess, drug-induced liver injury, viral hepatitis, autoimmune liver disease, haemochromatosis, coeliac disease and Wilson's disease (in patients <45 years old). Autoantibodies are also frequently detected at a low titre in subjects with NAFLD (antinuclear antibody (ANA) ≥1:160 and/or antismooth muscle antibody (ASMA) ≥1:40) and are usually associated with normal IgG levels and do not generally indicate autoimmune hepatitis. Raised ferritin levels are common in NAFLD and usually reflect underlying inflammatory activity or insulin resistance. A transferrin saturation <45% rules out haemochromatosis. If there is uncertainty about the diagnosis of NAFLD, then a liver biopsy should be considered.
The NAFLD Liver Fat Score is calculated using the presence of the metabolic syndrome, type 2 diabetes, fasting serum insulin, fasting serum AST and the AST/ALT ratio (AAR). In a cohort of 470 patients, a score greater than −0.640 predicted NAFLD with a sensitivity of 86% and specificity of 71%. Using cut-off scores of −1.413 and ≥1.257 gave 95% sensitivity for the prediction of NAFLD (with 52% and 51% specificity, respectively). However, this score does not distinguish between the different stages of NAFLD.
Once suspected clinically, fatty infiltration of the liver can be confirmed with imaging. Ultrasonography is widely used as a first-line investigation for hepatic steatosis that provides a qualitative assessment of fatty infiltration of the liver. Ultrasound is very effective in diagnosing steatosis where >33% of hepatocytes are steatotic but can be unreliable with lesser degrees of steatosis. Therefore, the finding of a normal liver on ultrasound does not rule out mild fatty infiltration of the liver. Other imaging modalities such as CT or MRI can also detect hepatic steatosis, but they are not routinely used in the assessment of steatosis. MRI and proton magnetic resonance spectroscopy (MRI/H-MRS) are the most accurate non-invasive measures of steatosis.
Controlled Attenuation Parameter (CAP) is a new ultrasound-based technique to measure steatosis simultaneously with assessment of liver stiffness using transient elastography (TE) (discussed below) that is likely to increase in use in coming years. In the largest study conducted to date, CAP accurately detected steatosis (area under receiver operating characteristic (AUROC) were 0.80 for mild, 0.86 for moderate and 0.88 for severe steatosis) independent of fibrosis in 615 patients with hepatitis C. However, few patients with NAFLD have been studied, and the diagnostic thresholds have not yet been clearly defined in NAFLD.
Although frequently not required for diagnosis, a liver biopsy is the definitive investigation for NAFLD and provides an assessment of hepatic steatosis, hepatocellular injury, inflammation and fibrosis. The presence of hepatocyte ballooning degeneration in association with steatosis is the key histological feature that distinguishes NASH from simple steatosis. The 'NAFLD activity score' (NAS) is the most widely used histological grading and staging system for NAFLD (Table 3). The SAF score (encompassing an assessment of steatosis (S), activity (A) and fibrosis (F)) has been introduced more recently, which may be more accurate in identifying NASH. However, the majority of patients with NAFLD can be diagnosed and staged adequately using non-invasive strategies. Liver biopsy should be used for subjects where there is diagnostic uncertainly or if non-invasive staging is indeterminate.
Knowledge of whether a patient has simple steatosis or NASH is very important prognostically. Subjects with simple steatosis have a good long-term prognosis with low rates of liver-related morbidity and mortality, and therefore do not require specific liver-related treatment. However, patients with NASH can progress to cirrhosis and therefore should be more actively managed to try and prevent disease progression. Unfortunately there is no widely available simple blood test or imaging modality that can differentiate simple steatosis from NASH. Clinically, the risk of steatohepatitis increases with the number of metabolic risk factors. Seventy per cent of centrally obese patients with hypertension and diabetes have steatohepatitis on liver biopsy. Therefore, until effective blood tests are available, metabolic risk factor profiling could be used to identify patients for further investigations. The search for an accurate biomarker of NASH is an active area of clinical research, and there have been some recent advances (see ref. 38 for a comprehensive review).
NASH is associated with increased apoptosis, so serum markers of apoptosis may be valuable in distinguishing NASH from simple steatosis. During apoptosis, caspases are activated and cleave various substrates, including cytokeratin-18 (CK-18), a major intermediate filament protein in hepatocytes. Hepatocyte apoptosis releases cleaved CK-18 fragments to the bloodstream that can be detected with an ELISA. Studies have demonstrated that the M30 antibody can identify patients with NASH with reasonable accuracy. Feldstein et al showed that the level of plasma-cleaved CK-18 fragments was an independent predictor of NASH with an AUROC for a diagnosis of NASH compared with 'not' or 'borderline' NASH of 0.83. In that study, cleaved CK-18 fragments were 71% sensitive and 85% specific for NASH at a cut-off of 279 U/L. Another CK-18 assay (M65 ELISA) is available that detects both cleaved and intact plasma CK-18 fragments and as a result is a marker of cell death by apoptosis and necrosis. Further validation is required to establish whether this assay is more accurate in differentiating NASH from simple steatosis. Although CK-18 is still being evaluated, some units are starting to incorporate this test into investigation algorithms for patients with NAFLD.
Another potentially interesting biomarker of NASH is the serum marker of matrix turnover, terminal peptide of procollagen III (PIIINP). In a study of 172 patients, PIIINP differentiated between simple steatosis and NASH with reasonable accuracy in patients with both mild and advanced fibrosis (AUROC 0.77–0.82 in patients with F0–2 fibrosis and 0.82–0.84 in patients with F0–3 fibrosis). Moreover, PIIINP was accurate in identifying patients with either NASH or advanced fibrosis (AUROC 0.85–0.87). While this needs to be validated, PIIINP might offer a useful test to identify the highest risk patients for further investigation.
Staging fibrosis is essential in all patients with NAFLD to identify subjects with advanced fibrosis who are at risk of liver-related complications.
Simple Non-invasive Markers of Fibrosis. Hepatocellular dysfunction and portal hypertension result from advancing hepatic fibrosis. This may be reflected in 'routine' blood tests such as liver function tests (low albumin), full blood count (thrombocytopenia) and coagulation profile (prolonged prothrombin time). These tests provide an indirect measure of fibrosis and are potentially appealing non-invasive markers of fibrosis as they are inexpensive and are performed in all patients with liver disease. With increasing liver fibrosis the ALT typically falls and the AST remains stable or rises, and as a result the AAR increases and can be a useful simple method of identifying patients with advanced fibrosis. Previous studies identified a cut-off >1 for the AAR as a diagnostic test for cirrhosis. However, a lower cut-off of >0.8 is more sensitive in patients with NAFLD. In our own study, an AAR <0.8 had high predictive ability to exclude advanced fibrosis (AUROC 0.83, sensitivity 74%, specificity 78%, negative predictive value (NPV) 93%), but its positive predictive ability was limited (PPV 44%). Although the AAR is reasonably accurate alone, its accuracy is enhanced when combined with other clinical and biochemical features and as a result is incorporated into other non-invasive scores (see Table 4 ).
The BARD score ( Table 4 ) is a simple test using the body mass index (BMI), AAR and presence of type 2 diabetes mellitus. A score <2 has excellent NPV of 95–97%, which reliably excludes advanced fibrosis. However, in a typical NAFLD cohort, a large proportion of patients with mild disease have a score of ≥2 due to obesity and diabetes, which limits its utility in clinical practice.
The NAFLD fibrosis score ( Table 4 ; http://www.nafldscore.com/) is a validated scoring system that comprises six routinely measured parameters. Advanced fibrosis can be reliably excluded (NPV 93%) using the low cut-off score (<−1.455) and diagnosed with high accuracy (PPV 90%) using the high cut-off score (>0.676). These results have been validated in other studies.
The FIB-4 score ( Table 4 ; http://gihep.com/calculators/hepatology/fibrosis-4-score/ ) although derived in patients with hepatitis C and HIV coinfection, appears to be one of the most useful non-invasive tests for diagnosing advanced fibrosis in NAFLD. For stage 3–4 fibrosis, a score <1.3 has a 90% NPV and a score >2.67 has an 80% PPV, with 72% of patients scoring below 1.3 or above 2.67. Other studies have confirmed that the FIB-4 score is slightly better than other non-invasive tests in diagnosing advanced fibrosis in NAFLD, including in subjects with normal range ALT levels.
All these simple non-invasive tests for fibrosis have good NPVs and can therefore exclude advanced fibrosis in patients with NAFLD who have low scores. As they can be calculated in all patients with routine blood tests, they offer an excellent method of identifying patients with mild disease who can be managed in primary care. However, the PPVs for these tests are modest (ranging from 27 to 79%), meaning that clinicians should consider further investigation to look for advanced fibrosis in patients with an intermediate or high score for their chosen test.
Fibroscan: Fibrotic livers have reduced elasticity due to the deposition of fibrous tissue in the hepatic parenchyma. TE (Fibroscan) gives a 'liver stiffness measurement' (LSM) using pulsed-echo ultrasound as a surrogate marker of fibrosis. The LSM correlates well with the degree of hepatic fibrosis in a range of liver diseases, including NAFLD. In a study of 246 patients with biopsy-proven NAFLD, TE achieved high AUROCs for the detection of ≥stage 2 fibrosis, ≥stage 3 fibrosis and cirrhosis (0.84, 0.93 and 0.95, respectively) and performed better than a number of simple non-invasive scores in the staging of fibrosis. In that study, TE had a high NPV of 96% for ≥stage 3 fibrosis at a cut-off of 7.9 kPa but only modest PPV (52% at 7.9 kPa and 72% at 9.6 kPa). A low LSM reliably excludes advanced fibrosis, but the optimum cut-offs for clinical use are yet to be determined.
However, there are significant limitations to using TE in NAFLD. Results may be invalid in older patients (>52 years) and those with central obesity (BMI >35 kg/m) or type 2 diabetes. For obese patients, the Fibroscan XL probe has been developed that is associated with fewer LSM failures (1.1% vs 16%) than the M probe and was accurate for the diagnosis of ≥F2 fibrosis and cirrhosis (AUROC 0.83 and 0.94, respectively). However, even with the XL probe, 10% of patients with a BMI >28 kg/m have a difference of ≥2 fibrosis stages between TE and liver biopsy.
Acoustic Radiation Force Impulse: Another imaging technique that has the potential for the non-invasive assessment of fibrosis is acoustic radiation force impulse (ARFI). This technique uses conventional B-mode ultrasonography to generate an ultrasonic pulse and measure the response of the liver tissue as shear wave velocity. The median velocity measured by ARFI increases with the degree of fibrosis. In one study of 54 patients with NAFLD, the AUROC for the diagnosis of stage 3 or 4 fibrosis was 0.97. Although further validation is necessary, this technique is becoming increasingly available on ultrasound machines and has the potential to stage liver fibrosis at the time of liver ultrasound.
Commercial Non-invasive Fibrosis Tests: The Enhanced Liver Fibrosis (ELF) test is a commercial panel of markers of matrix turnover: tissue inhibitor of matrix metalloproteinase 1 (TIMP1), hyaluronic acid and PIIINP. This test performs slightly better than the NAFLD fibrosis score for diagnosing moderate fibrosis (AUROC 0.90 vs 0.86) and severe fibrosis (AUROC 0.93 vs 0.89), but combining the two tests gives an AUROC of 0.93 for moderate fibrosis and 0.98 for severe fibrosis.
Fibrotest (FT) is a commercial panel of biochemical markers of fibrosis that is widely used in France. In NAFLD, FT can diagnose advanced fibrosis with modest accuracy (AUROCs 0.75–0.86 for stage 2–4 fibrosis and 0.81–0.92 for stage 3–4 fibrosis). Using a FT cut-off of 0.30 gives a 90% NPV for advanced fibrosis (sensitivity 77%), and a FT cut-off of 0.70 had a 73% PPV for advanced fibrosis (specificity 98%). However, this test is not widely available in the UK.
Diagnosis and Staging of NAFLD/NASH
Clinical Features and Blood Tests
A diagnosis of NAFLD requires that there is evidence of hepatic steatosis on imaging or histology, and other causes of liver disease or steatosis have been excluded. NAFLD is usually asymptomatic, so diagnosis usually follows the incidental finding of abnormal liver enzymes or steatosis on imaging. If abnormal LFTs are present, this is usually mildly raised transaminases (ALT> aspartate transaminase (AST)) and/or gamma-glutamyltransferase. However, ~80% of patients have normal-range ALT levels (males <40 IU/L and females <31 IU/L), and even if elevated, the ALT typically falls (and AST may rise) as fibrosis progresses to cirrhosis. ALT values do not correlate with histological findings and are unhelpful in both the diagnosis of NAFLD and determining disease severity. Clinicians overly rely on abnormal liver enzymes to identify patients with NAFLD, so patients with significant liver disease can be overlooked, potentially missing opportunities for intervention. It has been repeatedly shown that 70–80% of subjects with central obesity and 50–80% of patients with type 2 diabetes have evidence of NAFLD on imaging. Therefore, a new approach is needed to use metabolic risk factors to identify subjects with NAFLD/NASH rather than relying on liver enzyme abnormalities.
In patients with abnormal LFTs, alternative causes of liver disease (or cofactors) should be excluded, including alcohol excess, drug-induced liver injury, viral hepatitis, autoimmune liver disease, haemochromatosis, coeliac disease and Wilson's disease (in patients <45 years old). Autoantibodies are also frequently detected at a low titre in subjects with NAFLD (antinuclear antibody (ANA) ≥1:160 and/or antismooth muscle antibody (ASMA) ≥1:40) and are usually associated with normal IgG levels and do not generally indicate autoimmune hepatitis. Raised ferritin levels are common in NAFLD and usually reflect underlying inflammatory activity or insulin resistance. A transferrin saturation <45% rules out haemochromatosis. If there is uncertainty about the diagnosis of NAFLD, then a liver biopsy should be considered.
The NAFLD Liver Fat Score is calculated using the presence of the metabolic syndrome, type 2 diabetes, fasting serum insulin, fasting serum AST and the AST/ALT ratio (AAR). In a cohort of 470 patients, a score greater than −0.640 predicted NAFLD with a sensitivity of 86% and specificity of 71%. Using cut-off scores of −1.413 and ≥1.257 gave 95% sensitivity for the prediction of NAFLD (with 52% and 51% specificity, respectively). However, this score does not distinguish between the different stages of NAFLD.
Imaging Assessment of Steatosis
Once suspected clinically, fatty infiltration of the liver can be confirmed with imaging. Ultrasonography is widely used as a first-line investigation for hepatic steatosis that provides a qualitative assessment of fatty infiltration of the liver. Ultrasound is very effective in diagnosing steatosis where >33% of hepatocytes are steatotic but can be unreliable with lesser degrees of steatosis. Therefore, the finding of a normal liver on ultrasound does not rule out mild fatty infiltration of the liver. Other imaging modalities such as CT or MRI can also detect hepatic steatosis, but they are not routinely used in the assessment of steatosis. MRI and proton magnetic resonance spectroscopy (MRI/H-MRS) are the most accurate non-invasive measures of steatosis.
Controlled Attenuation Parameter (CAP) is a new ultrasound-based technique to measure steatosis simultaneously with assessment of liver stiffness using transient elastography (TE) (discussed below) that is likely to increase in use in coming years. In the largest study conducted to date, CAP accurately detected steatosis (area under receiver operating characteristic (AUROC) were 0.80 for mild, 0.86 for moderate and 0.88 for severe steatosis) independent of fibrosis in 615 patients with hepatitis C. However, few patients with NAFLD have been studied, and the diagnostic thresholds have not yet been clearly defined in NAFLD.
Liver Biopsy for NAFLD
Although frequently not required for diagnosis, a liver biopsy is the definitive investigation for NAFLD and provides an assessment of hepatic steatosis, hepatocellular injury, inflammation and fibrosis. The presence of hepatocyte ballooning degeneration in association with steatosis is the key histological feature that distinguishes NASH from simple steatosis. The 'NAFLD activity score' (NAS) is the most widely used histological grading and staging system for NAFLD (Table 3). The SAF score (encompassing an assessment of steatosis (S), activity (A) and fibrosis (F)) has been introduced more recently, which may be more accurate in identifying NASH. However, the majority of patients with NAFLD can be diagnosed and staged adequately using non-invasive strategies. Liver biopsy should be used for subjects where there is diagnostic uncertainly or if non-invasive staging is indeterminate.
Differentiating Steatosis From Steatohepatitis Without Liver Biopsy
Knowledge of whether a patient has simple steatosis or NASH is very important prognostically. Subjects with simple steatosis have a good long-term prognosis with low rates of liver-related morbidity and mortality, and therefore do not require specific liver-related treatment. However, patients with NASH can progress to cirrhosis and therefore should be more actively managed to try and prevent disease progression. Unfortunately there is no widely available simple blood test or imaging modality that can differentiate simple steatosis from NASH. Clinically, the risk of steatohepatitis increases with the number of metabolic risk factors. Seventy per cent of centrally obese patients with hypertension and diabetes have steatohepatitis on liver biopsy. Therefore, until effective blood tests are available, metabolic risk factor profiling could be used to identify patients for further investigations. The search for an accurate biomarker of NASH is an active area of clinical research, and there have been some recent advances (see ref. 38 for a comprehensive review).
NASH is associated with increased apoptosis, so serum markers of apoptosis may be valuable in distinguishing NASH from simple steatosis. During apoptosis, caspases are activated and cleave various substrates, including cytokeratin-18 (CK-18), a major intermediate filament protein in hepatocytes. Hepatocyte apoptosis releases cleaved CK-18 fragments to the bloodstream that can be detected with an ELISA. Studies have demonstrated that the M30 antibody can identify patients with NASH with reasonable accuracy. Feldstein et al showed that the level of plasma-cleaved CK-18 fragments was an independent predictor of NASH with an AUROC for a diagnosis of NASH compared with 'not' or 'borderline' NASH of 0.83. In that study, cleaved CK-18 fragments were 71% sensitive and 85% specific for NASH at a cut-off of 279 U/L. Another CK-18 assay (M65 ELISA) is available that detects both cleaved and intact plasma CK-18 fragments and as a result is a marker of cell death by apoptosis and necrosis. Further validation is required to establish whether this assay is more accurate in differentiating NASH from simple steatosis. Although CK-18 is still being evaluated, some units are starting to incorporate this test into investigation algorithms for patients with NAFLD.
Another potentially interesting biomarker of NASH is the serum marker of matrix turnover, terminal peptide of procollagen III (PIIINP). In a study of 172 patients, PIIINP differentiated between simple steatosis and NASH with reasonable accuracy in patients with both mild and advanced fibrosis (AUROC 0.77–0.82 in patients with F0–2 fibrosis and 0.82–0.84 in patients with F0–3 fibrosis). Moreover, PIIINP was accurate in identifying patients with either NASH or advanced fibrosis (AUROC 0.85–0.87). While this needs to be validated, PIIINP might offer a useful test to identify the highest risk patients for further investigation.
Non-invasive Staging of Liver fibrosis
Staging fibrosis is essential in all patients with NAFLD to identify subjects with advanced fibrosis who are at risk of liver-related complications.
Simple Non-invasive Markers of Fibrosis. Hepatocellular dysfunction and portal hypertension result from advancing hepatic fibrosis. This may be reflected in 'routine' blood tests such as liver function tests (low albumin), full blood count (thrombocytopenia) and coagulation profile (prolonged prothrombin time). These tests provide an indirect measure of fibrosis and are potentially appealing non-invasive markers of fibrosis as they are inexpensive and are performed in all patients with liver disease. With increasing liver fibrosis the ALT typically falls and the AST remains stable or rises, and as a result the AAR increases and can be a useful simple method of identifying patients with advanced fibrosis. Previous studies identified a cut-off >1 for the AAR as a diagnostic test for cirrhosis. However, a lower cut-off of >0.8 is more sensitive in patients with NAFLD. In our own study, an AAR <0.8 had high predictive ability to exclude advanced fibrosis (AUROC 0.83, sensitivity 74%, specificity 78%, negative predictive value (NPV) 93%), but its positive predictive ability was limited (PPV 44%). Although the AAR is reasonably accurate alone, its accuracy is enhanced when combined with other clinical and biochemical features and as a result is incorporated into other non-invasive scores (see Table 4 ).
The BARD score ( Table 4 ) is a simple test using the body mass index (BMI), AAR and presence of type 2 diabetes mellitus. A score <2 has excellent NPV of 95–97%, which reliably excludes advanced fibrosis. However, in a typical NAFLD cohort, a large proportion of patients with mild disease have a score of ≥2 due to obesity and diabetes, which limits its utility in clinical practice.
The NAFLD fibrosis score ( Table 4 ; http://www.nafldscore.com/) is a validated scoring system that comprises six routinely measured parameters. Advanced fibrosis can be reliably excluded (NPV 93%) using the low cut-off score (<−1.455) and diagnosed with high accuracy (PPV 90%) using the high cut-off score (>0.676). These results have been validated in other studies.
The FIB-4 score ( Table 4 ; http://gihep.com/calculators/hepatology/fibrosis-4-score/ ) although derived in patients with hepatitis C and HIV coinfection, appears to be one of the most useful non-invasive tests for diagnosing advanced fibrosis in NAFLD. For stage 3–4 fibrosis, a score <1.3 has a 90% NPV and a score >2.67 has an 80% PPV, with 72% of patients scoring below 1.3 or above 2.67. Other studies have confirmed that the FIB-4 score is slightly better than other non-invasive tests in diagnosing advanced fibrosis in NAFLD, including in subjects with normal range ALT levels.
All these simple non-invasive tests for fibrosis have good NPVs and can therefore exclude advanced fibrosis in patients with NAFLD who have low scores. As they can be calculated in all patients with routine blood tests, they offer an excellent method of identifying patients with mild disease who can be managed in primary care. However, the PPVs for these tests are modest (ranging from 27 to 79%), meaning that clinicians should consider further investigation to look for advanced fibrosis in patients with an intermediate or high score for their chosen test.
Fibroscan: Fibrotic livers have reduced elasticity due to the deposition of fibrous tissue in the hepatic parenchyma. TE (Fibroscan) gives a 'liver stiffness measurement' (LSM) using pulsed-echo ultrasound as a surrogate marker of fibrosis. The LSM correlates well with the degree of hepatic fibrosis in a range of liver diseases, including NAFLD. In a study of 246 patients with biopsy-proven NAFLD, TE achieved high AUROCs for the detection of ≥stage 2 fibrosis, ≥stage 3 fibrosis and cirrhosis (0.84, 0.93 and 0.95, respectively) and performed better than a number of simple non-invasive scores in the staging of fibrosis. In that study, TE had a high NPV of 96% for ≥stage 3 fibrosis at a cut-off of 7.9 kPa but only modest PPV (52% at 7.9 kPa and 72% at 9.6 kPa). A low LSM reliably excludes advanced fibrosis, but the optimum cut-offs for clinical use are yet to be determined.
However, there are significant limitations to using TE in NAFLD. Results may be invalid in older patients (>52 years) and those with central obesity (BMI >35 kg/m) or type 2 diabetes. For obese patients, the Fibroscan XL probe has been developed that is associated with fewer LSM failures (1.1% vs 16%) than the M probe and was accurate for the diagnosis of ≥F2 fibrosis and cirrhosis (AUROC 0.83 and 0.94, respectively). However, even with the XL probe, 10% of patients with a BMI >28 kg/m have a difference of ≥2 fibrosis stages between TE and liver biopsy.
Acoustic Radiation Force Impulse: Another imaging technique that has the potential for the non-invasive assessment of fibrosis is acoustic radiation force impulse (ARFI). This technique uses conventional B-mode ultrasonography to generate an ultrasonic pulse and measure the response of the liver tissue as shear wave velocity. The median velocity measured by ARFI increases with the degree of fibrosis. In one study of 54 patients with NAFLD, the AUROC for the diagnosis of stage 3 or 4 fibrosis was 0.97. Although further validation is necessary, this technique is becoming increasingly available on ultrasound machines and has the potential to stage liver fibrosis at the time of liver ultrasound.
Commercial Non-invasive Fibrosis Tests: The Enhanced Liver Fibrosis (ELF) test is a commercial panel of markers of matrix turnover: tissue inhibitor of matrix metalloproteinase 1 (TIMP1), hyaluronic acid and PIIINP. This test performs slightly better than the NAFLD fibrosis score for diagnosing moderate fibrosis (AUROC 0.90 vs 0.86) and severe fibrosis (AUROC 0.93 vs 0.89), but combining the two tests gives an AUROC of 0.93 for moderate fibrosis and 0.98 for severe fibrosis.
Fibrotest (FT) is a commercial panel of biochemical markers of fibrosis that is widely used in France. In NAFLD, FT can diagnose advanced fibrosis with modest accuracy (AUROCs 0.75–0.86 for stage 2–4 fibrosis and 0.81–0.92 for stage 3–4 fibrosis). Using a FT cut-off of 0.30 gives a 90% NPV for advanced fibrosis (sensitivity 77%), and a FT cut-off of 0.70 had a 73% PPV for advanced fibrosis (specificity 98%). However, this test is not widely available in the UK.
Source...