A prospective study on the prevalence of MASLD in people with type‐2 diabetes in the community. Cost effectiveness of screening strategies

As screening for the liver disease and risk‐stratification pathways are not established in patients with type‐2 diabetes mellitus (T2DM), we evaluated the diagnostic performance and the cost‐utility of different screening strategies for MASLD in the community.


| INTRODUC TI ON
Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most common cause of abnormal liver function tests (LFTs) worldwide, with a global estimated prevalence of 30%. 1 MASLD is also expected to become the leading cause of end-stage liver disease in the coming decades. 2 Histologically, MASLD encompasses a spectrum of disorders from steatosis with or without hepatocellular injury and/ or inflammation (Metabolic dysfunction-associated Steato-Hepatitis, MASH) and a variable degree of fibrosis through to cirrhosis. 3Fibrosis stage represents the strongest predictor of clinical outcomes-liver and non-liver-related-in these patients. 4From a clinical perspective, the presence of type 2 diabetes mellitus (T2DM), is an independent predictor of advanced fibrosis in patients with MASLD, 5 with a greater prevalence of advanced disease in diabetic compared to nondiabetic individuals, especially in younger ages. 6ven the high prevalence-estimated at 55.5% 7 -and severity of MASLD in the diabetic population, there is a major interest in early detection of the disease, especially in primary care, 8 where diagnosing MASLD is perceived as a clinical challenge, with specific concerns on performing risk-stratification among patients. 9Both the AASLD 10 and the EASL guidelines 11 recommend screening for MASLD in highrisk risk groups (i.e., patients with metabolic syndrome) following a 2-tier system.Specifically, patients should be stratified using noninvasive markers of fibrosis such as Fibrosis score-4 (FIB-4) and/or NAFLD fibrosis score in primary care, followed by Enhanced liver fibrosis test (ELF) and/or transient elastography (TE) in a specialist setting.Such strategy relies heavily on scores, which were derived from a tertiary care setting, and whose diagnostic accuracy in primary care is still unclear. 12Furthermore, there is still a debate on whether screening might be cost-effective in this population. 8Finally, despite the most recent diabetes management guidelines suggesting to screen diabetics for MASLD, 13 the overall awareness among diabetologists remains low. 14 this study, we aimed to establish the prevalence of MASLD in patients with T2DM in primary care.Moreover, we tested the performance of non-invasive markers and further developed a riskstratification pathway.We also built a Markov model simulating MASLD screening and assessed the cost-utility of different screening strategies for MASLD in the diabetic community.

| Study population
This single-centre, cross-sectional study prospectively recruited consecutive patients with T2DM in primary care and community clinics from the North-West London general practitioner (GP) network.Inclusion criteria were the ability to give informed consent, age >18 years and presence of T2DM, as defined by medical history or recent 2 h post-challenge plasma glucose ≥11.1 mmol/L.Patients were excluded if they had known the liver disease.

| Screening for liver disease and MASLD
All the patients were screened for liver disease and the presence of MASLD with blood tests (full liver screen), imaging ultrasound (US) and TE, with liver stiffness measurement (LSM) and Controlled Handling Editor: Luca Valenti the cost-utility analysis, ICER was £2480/QALY for TE, £2541.24/QALYfor ELF and £2059.98/QALYfor FIB-4.

Conclusion:
Screening for MASLD in the diabetic population in primary care is costeffective and should become part of a holistic assessment.However, traditional screening strategies, including FIB-4 and ELF, underestimate the presence of significant liver disease in this setting.

K E Y W O R D S
liver fibrosis, metabolic dysfunction-associated steatotic liver disease, primary care, screening

Lay summary
Metabolic dysfunction-associated steatotic liver disease (MASLD), a common disease where excessive fat accumulated in the liver and may result in cirrhosis and heart attacks, is a highly prevalent, yet largely underappreciated liver condition that is closely associated with metabolic disease and type 2 diabetes mellitus (T2DM).Yet, a strategy to understand who is at risk of developing this disease and suffering liver damage is lacking.
In this study, we describe the prevalence of advanced liver disease in diabetics in primary care and we define an easy way to screen diabetics for MASLD.We demonstrate that, among diabetics, education level is associated with a greater risk of having liver disease.Moreover, we demonstrate that screening for fatty liver in primary care using non-invasive markers of fibrosis, is cost-effective and should be offered to all the diabetics in the community.attenuation parameter (CAP) score.Moderate-to-high cardiovascular risk was defined based on Qrisk2 score as ≥20%.Further details on screening procedures are given in Supplementary material.

| Screening strategies and identification rates
In this study, five screening strategies were compared against the standard of care: (1) US plus LFTs, (2) FIB-4, (3) NAFLD fibrosis score, (4) ELF and ( 5) TE.Standard of care was derived from previously published economic evaluations of MASLD screening where this entailed LFTs or no screening 15,16 (Supplementary Table S1).As part of the standard of care, abnormal LFTs were assumed to prompt referral to the hospital, with a 65% specificity and 35% sensitivity for liver fibrosis. 17 the first tier of each strategy, patients were divided into two groups: no disease/MASLD without significant fibrosis versus MASLD with significant and advanced fibrosis.No disease and MASLD without significant fibrosis were considered the same group for this analysis as the management would be similar and would not trigger a referral to secondary care compared to MASLD with significant and advanced fibrosis that triggers a referral to specialist care. 11In strategy 1 (US plus LFTs), MASLD with significant fibrosis was defined as evidence of steatosis and features of chronic liver disease on ultrasound, plus elevated LFTs.In strategy 2, significant fibrosis was defined as FIB-4 >1.3 and in strategy 3, as NAFLD fibrosis score >−1.45.In strategy 4 (ELF) significant fibrosis was defined as ELF≥9.8 and in strategy 5 (TE), as LSM ≥8.1 kPa. 11

| Decision-analytic model
We developed a decision tree to characterise the risk stratification and diagnostic performance of each of the primary care screening strategies evaluated.We adapted previously published Markov models 15,17 to characterise the subsequent health states and disease pathways of patients based on their initial primary care screening risk stratification (Figure 1).Patients with advanced stages of liver disease could progress to health states that reflect end-stage liver disease, including decompensated cirrhosis (DC), hepatocellular carcinoma (HCC), liver transplant (LT) and death. 15With a diagnosis present (status of significant liver disease and/or compensated cirrhosis), there is a probability that the management of the patient is modified to reduce the risk of progression to CC, decompensation or death.In this case, progression rates of SLD and CC groups were assumed as slower, compared to those whose diagnosis was missed at screening (false negatives). 15Details on model input parameters are given in Supplementary material.

| Model outcomes
The cost-utility analysis in the base-case was conducted over a lifetime horizon and generated the cost per quality-adjusted life year (QALY) gained.A discount rate of 3.5% per year was applied to outcomes and costs, as recommended by the NICE guidelines. 18 calculated the average cost-effectiveness and the incremental cost-effectiveness ratio (ICER) compared to the standard of care. 15,17Life expectancy, lifetime costs and the number of correct diagnoses were also estimated.According to NICE guidelines, a cost-effectiveness threshold (CET) of £20 000/QALY gained was set for the base-case analysis as per previous studies. 19 probabilities, utility values, costs and screening ratios) were varied to determine the impact of their variability on cost-effectiveness results.Sensitivity analysis ranges and probabilistic distributions were derived from previous literature and are reported in detail in Supplementary Tables S4-S8.

| Study population
Between April 2019 and January 2021, a total of 300 consecutive patients with T2DM were enrolled from the North-West London GP network.Overall, 287 patients underwent the whole screening procedure, while 13 did not complete the screening and were excluded (Supplementary Figure S1).The study population was diverse in terms of ethnic background and also diverse in terms of severity of T2DM and anti-diabetic treatments (Tables 1 and   2).The success rate for performing TE in this population was 99% (286/287).
Of note, 42% of the patients with 8.1 kPa ≤LSM ≤12.1 kPa and 38% of patients with LSM ≥12.1 kPa had normal LFTs.
When the CAP score was used to define steatosis, the overall prevalence of MASLD was 67% (195/287), the prevalence of significant liver disease (LSM ≥8.1 kPa), was 16% (48/287), while the prevalence of advanced fibrosis (LSM ≥12.1 kPa), was 11% (33/287) in the whole population.Of note, 3% (9/287) had elevated CAP score but no evidence of steatosis on the US.However, only those with a positive ultrasound were considered as having steatosis.

| Prevalence of cirrhosis
The prevalence of newly diagnosed cirrhosis secondary to MASLD was 3% (8/287; 6 with clinical diagnosis and 2 based on histology) in the whole diabetic population and 5% (8/184) in the MASLD subgroup (Supplementary Figure S1).The number needed to treat/ screen (NNT) in this population was 4.56 (3.38-7).Due to the COVID-19-related restrictions, only 11 patients underwent a liver biopsy among those with elevated LSM (as per standard of care): all the biopsied cases had liver fibrosis stage ≥2 according to the CRN scoring system.

| Advanced liver disease is more prevalent in the deprived population
In terms of socio-economic status, those with MASLD and significant fibrosis lived in more deprived neighbourhoods according to their median education rank (18 789 vs. 23 148, p = .03)(Supplementary Table S9).Similarly, those with MASLD and advanced fibrosis lived in more deprived neighbourhoods according to their median education rank (18 793 vs. 23 162, p = .05).Conversely, there was no difference in terms of the other deprivation scores: income, employment, health deprivation and disability, barriers to housing and service, and crime.
TA B L E 1 Characteristics of the study population and differences between patients with and without MASLD.
There was no difference in terms of the prevalence of MASLD, significant or advanced liver fibrosis.
When compared to age-matched men, menopausal women showed smaller waist (103 vs. 108 cm, p = .03)and hip circumference (111 vs. 113 cm, p =.018).Menopausal women also showed significantly lower ALT (26 vs. 33 IU/L, p = .003)and CAP score (293 vs. 313 dB/m, p = .013)compared to age-matched men.There was no difference in terms of the prevalence of MASLD, significant or advanced fibrosis.

| Cost-effectiveness analysis
The cost-effectiveness analysis was based on the performance characteristics (positive predictive values and negative predictive values) for the identification of patients with significant and advanced fibrosis from the study population (Supplementary Figures S2-S6).
Overall, screening for MASLD by any of the strategies analysed improved the rate of diagnosis by 8%-15%.All screening strategies were associated with QALY gains, ranging from 121 to 149 years, with TE (148.73 years) resulting in the most substantial gains, followed by FIB-4 (134.07 years), ELF (131.68 years) and NAFLD fibrosis score (121.25 years).The ICER of TE compared to the standard of care was £2480 per QALY gained (Table 4).
The ICER was most sensitive to variations in progression rates (effect of early diagnosis on disease progression), screening test sensitivity and specificity and model time horizon.Nevertheless, when transition probabilities, utilities, screening treatment effect and cost inputs were modified, we found a >99% probability of MASLD screening tests being cost-effective compared to standard of care in all evaluated scenarios (Figure 2, Supplementary Tables S4-S8).When sensitivity and specificity of each screening test were varied in a range between 20% and 100%, the ICER remained cost-effective below £3260 in all scenarios (Supplementary Tables S4-S8).Although all screening strategies were found to be costeffective compared to standard of care in the base-case, when the time horizon was decreased from 40 years (lifetime) to 5 years, only FIB-4 remained cost-effective within the NICE cost-effectiveness threshold criteria.

| DISCUSS ION
Non-alcoholic fatty liver disease has now become the leading cause of chronic liver disease in Western countries and the fastest-growing indication for liver transplantation in the United States. 20Defining and implementing models of care has been identified as an area of priority for tackling MASLD worldwide. 21Specifically, there is need for clearly defined, pragmatical referral management pathways, which are based on clinical context and shared with local primary care providers.Being a high-risk group for advanced liver disease, 6,22 TA B L E 3 Predictive factors for the presence of significant liver disease in the whole diabetic population.patients with T2DM represent an ideal target for MASLD screening in primary care.
In this study, we studied a cohort of patients with diabetes who were screened for MASLD and other liver diseases in primary care, without any a priori selection.This cohort includes patients with a wide range of antidiabetic treatments, comorbidities, ranges of glycaemic control and length of disease.Furthermore, conducting this study in North-West London, provided us with a very diverse in terms of ethnic and social background, which is a bonus compared to other studies in the field.Overall, the prevalence of MASLD based on the US was 64%, while the prevalence of significant liver disease was 17%, advanced liver disease 11% and cirrhosis 3% in the whole cohort.In a recently published work, in diabetic patients over 50 years old in the community and endocrinology clinics, their results are similar to our cohort with the prevalence of MASLD, advanced fibrosis and cirrhosis at 65%, 14% and 6%. 23 our cohort, visceral obesity, education attainment and AST were the main clinical predictors for the presence of significant and advanced fibrosis in primary care.Overall, despite education level being a well-known risk factor for other chronic liver diseases, 24 this is the first work demonstrating clearly that education level is an important determinant of liver disease in the general diabetic TA B L E 4 Base-case cost-effectiveness analysis of MASLD screening strategies versus standard of care (baseline screening).According to the latest published EASL guidelines, patients with T2DM should be screened for MASLD using a two-tier system, that is, FIB-4 and/or ELF in primary care, followed by TE in a specialist setting. 11Nevertheless, standard of care for diagnosing MASLD among GPs still relies on ultrasound and LFTs, possibly due to limited awareness on the disease and/or screening policies. 25,26In this cohort, despite AST being a predictive factor against the liver disease, up to 42% of the patients with 8.1 kPa ≤LSM≤12.1 kPa and 38% of the patients with LSM ≥12.1 kPa had normal LFTs at screening.
Risk stratification should not rely on LFTs, as they both under-and over-estimate the severity of liver disease in MASLD. 26Moreover, according to the results from this study, applying FIB-4 with a cut-off of 1.3 in this population would miss up to 38% of the patients with significant liver disease and these would mainly be younger patients with normal LFTs.These results are in line with recently published data and highlight the limitation of the use of FIB-4 in primary care. 27milarly, when applying a cut-off of ELF ≥9.8, up to 59% of those with significant liver disease would be missed at screening.Despite recent literature highlighting gender-related differences in MASLD phenotypes, 28 in this population, there was no difference in terms of false negative rates between men and women in this population.
Of note, recent evidence has raised the concern that currently used non-invasive markers may underestimate liver disease in diabetics and that more evidence in primary care is needed. 10,29Nevertheless, it is worth noting that when FIB-4 and ELF were used as standards, LSM underestimated the presence of significant fibrosis in 28% and 62% of the patients, respectively.
Though cost-effectiveness data in screening for MASLD in patients with T2DM in the community is emerging, 8 there is still a debate about the appropriate screening strategy.It is of great importance to identify patients with a high risk of progressive disease, as this would lead to a reduction in progression rates to end-stage liver disease and associated healthcare burden.Moreover, not only lifestyle intervention could delay or reverse fibrosis progression 30,31 but also pharmacotherapies will soon be available.Furthermore, as more severe forms of SLD are also associated with the greatest additional risk of cardiovascular events, 32 screening tools which identify advanced SLD may by proxy also identify those at higher risk for acute cardiovascular events, further extending the utility of screening within this scenario.
In this study, we present a cost-effectiveness analysis for screening for MASLD based on a real-life population of patients with T2DM in primary care.MASLD screening in people with T2DM improved diagnostic outcomes and was cost-effective in all evaluated scenarios under a CET at £20 000.Overall, TE was the screening strategy associated with the greatest clinical gains (148.73QALYs).These results are in line with published work 27 and emphasise that screening for MASLD is cost-effective compared to standard of care defined as abnormal LFTs or even the combination of US and LFTs. 15,17,33,34Nevertheless, previous Key input parameters with the highest level of uncertainty (i.e., transition F I G U R E 1 Markov model for the cost-effectiveness analysis.CC, compensated cirrhosis (clinical diagnosis of cirrhosis); DC, decompensated cirrhosis; FN, false negatives (MASLD with LSM ≥8.1 kPa who were false negatives at screening); HCC, hepatocellular carcinoma; MLD, mild liver disease (no MASLD or MASLD with LSM≤8 kPa); MASLD, metabolic-dysfunction associated steatotic liver disease; SLD, significant liver disease (MASLD with LSM ≥8.1 kPa who were true positives at screening); T2DM, type 2 diabetes mellitus; +, diagnosed; −, undiagnosed.
population.Clinicians managing patients with T2DM should be aware of the risks associated with poor education and incorporate this knowledge into their patient clinical management.Multidisciplinary teams should ensure that families with poor literacy have an adequate understanding of them being at higher risk for liver disease.

F I G U R E 2
Cost-effectiveness acceptability curve.Red and blue lines represent the cost-effectiveness acceptability of MASLD screening strategies 1-5 and standard of care, respectively.Streategy 1 ultrasound plus liver function tests; strategy 2 FIB-4, strategy 3 NAFLD fibrosis score; strategy 4 ELF; strategy 5: Transient elastrography.Each dot on the graph shows the probability of each of the strategies (1-5) being cost-effective (Y-axis) at a given cost-effectiveness threshold (X-axis).(1) Point A shows the point at which both strategies have 50% probability of being cost-effective; (2) point B shows the point at which scenario 1 has 100% probability of being cost-effective.