statin use beneficial in chronic liver disease

 A recent article found that statins decreased liver fibrosis progression and lowered the risk of hepatocellular carcinoma and hepatic decompensation in patients with chronic liver disease (see statin dec risk HCC if chr liver dz JAMAIntMed2025 in dropbox,, or doi:10.1001/jamainternmed.2025.0115)

 

Details:

-- 16,501 participants were reviewed from the Research Patient Data Cape Registry, an aggregation of clinical data from 10 hospitals within the Mass General Brigham healthcare system in Boston, a database of roughly 4 million patients, with data from 2000 to 2023 on patients at least 40 years old with chronic liver disease (CLD) and a baseline Fibrosis-4 (FIB-4) score of 1.3 or higher

   -- 6750 females (40.9%) and 9751 males (59.1%), including 3610 statin users (meaning being on statins for 180 days between their first CLD diagnosis and statin initiation) and 12,891 nonusers (no statin use for the 180-day period)

        -- exclusion criteria included prior hepatocellular carcinoma (HCC) diagnosis, history of hepatic decompensation, HIV coinfection, or prior liver transplant

-- mean age 59.7 years, 78% white/7% Black/4% Asian

-- BMI <25 in 13%, 25-30 in 21%, at least 30 in 30%

-- chronic liver disease cause: hepatitis B (HBV)infection 5%, hepatitis C (HCV) infection 20%, autoimmune hepatitis 2%, alcoholic hepatitis 17%, (metabolic dysfunction-associated steatotic liver disease (MASLD, formally nonalcoholic fatty liver disease, NAFLD) 31%, cryptogenic cirrhosis 8%, primary biliary cholangitis 4%; cirrhosis in 38%

-- medical comorbidities: diabetes 30%, hypertension 60%, coronary artery disease 27%, dyslipidemia 50% (this was not defined further), chronic kidney disease 10% (not further defined), stroke 13%

-- median FIB-4 score 2.1; with intermediate FIB-4 score (1.3-2.67) in 65%, high FIB-4 score (> 2.67) in 35%

    -- of note, patients who were on statins were reasonably equivalent in terms of CLD causes, but there were disproportionately more on statins in the setting of diabetes, coronary artery disease, peripheral vascular disease, dyslipidemia, chronic kidney disease, stroke

-- medications: metformin 15%, aspirin 35%, oral anti-HBV meds 30% of those with HBV, oral anti-HCV meds 15% of those with HCV

-- lab test: platelet counts 190, ALT 33, AST 37, albumin 4.1, total bilirubin 0.7

-- abdominal imaging examinations per year: <1 in 32%, 1-2 in 12%, >2 in 19%

 

-- Statins were classified as lipophilic statins (atorvastatin, fluvastatin, lovastatin, pitavastatin, and simvastatin) or hydrophilic statins (rosuvastatin and pravastatin)

-- researchers determined the cumulative daily defined dose (cDDD) of the statins, reflecting the calculated cumulative statin exposure over the entire observation period and defined as <30 (reference), 30-599, and 600 or higher

-- FIB-4 reflects hepatic fibrosis and was calculated and stratified into low, intermediate, or high levels

    -- FIB-4 is a calculation involving the patient’s age, AST and ALT levels, and platelet count, and can be easily determined by https://www.mdcalc.com/calc/2200/fibrosis-4-fib-4-index-liver-fibrosis

-- Especially given the profound differences in those patients who were on statins versus not on statins, as noted above, the researchers employed inverse probability of treatment weighting (IPTW) based on propensity scores, a statistical methodology used to equalize disparate groups

 

-- primary outcome: incident hepatocellular carcinoma (HCC)

-- secondary outcome: hepatic decompensation, as determined by ICD codes for variceal bleeding, ascites, hepatic encephalopathy, spontaneous bacterial peritonitis, and hepato-renal syndrome

    -- they also assessed transitions in the FIB-4 scores between those on or not on statins

 

Results:

-- 10 year cumulative incidence of HCC:

    -- statin users: 3.8%

    -- non-statin users: 8.0%

        -- difference: -4.2% (-5.3% to -3.1%)

        -- 33% decrease in statin users, adjusted sub hazard ratio (aSHR) 0.67 (0.59-0.76)

-- hepatic decompensation:

    -- statin users: 10.6%

    -- non-statin users: 19.5%

        -- difference: -9.0% (-10.6% to -7.3%)

        -- 22% decrease in statin users: aSHR 0.78 (0.67-0.91)

 

-- Lipophilic statins (atorvastatin, fluvastatin, lovastatin, pitavastatin, and simvastatin) were associated with further reductions in HCC and hepatic decompensation scores, versus hydrophilic statins (rosuvastatin and pravastatin); however, only 680 participants were on the hydrophilic statins vs 2930 on the lipophilic ones:

    -- 10-year cumulative incidence of HCC: in hydrophilic versus lipophilic statins:

        -- hydrophilic statins, comparing statin users with nonusers: 4.1%

        -- lipophilic statins, comparing statin users with nonusers: 8.0%

            -- absolute difference: -3.9% (-6.2 to -1.6)

            -- aSHR: 23% decrease with lipophilic statins, 0.79 (0.63- 0.99)

    -- 10-year cumulative incidence of hepatic decompensation in hydrophilic versus lipophilic statins:

        -- hydrophilic statins, comparing statin users with nonusers: 7.9%

        -- lipophilic statins, comparing statin users with nonusers: 19.5%

            -- absolute difference: - 11.6% (-14.4% to -8.9%)

            -- aSHR: 42% decrease with lipophilic statins, 0.58 (0.42-0.88)

 

-- statin use duration:

    -- 10-year cumulative incidence of HCC: comparing patients with at least 600 cDDD (which was better than for those with cDDD of 30-599) vs nonusers:

        -- 3.5% versus 8.0% for nonusers

        -- absolute difference -4.5% (-5.6% to -3.2%)

            -- aSHR: 40% decrease with lipophilic statins, 0.60 (0.52-0.70)

    -- 10-year cumulative incidence of hepatic decompensation: comparing patients with at least 600 cDDD (which was better than for those with cDDD of 30-599) vs nonusers:

        -- 9.1% versus 19.5% for nonusers

        -- absolute difference -10.4% (-12.2% to -8.7%)

            -- aSHR: 36% decrease with lipophilic statins, 0.64 (0.51-0.80)

 

-- assessment of serial FIB-4 data, available in 7038 patients:

    -- those with intermediate baseline FIB-4 scores who transitioned to the high FIB-4 group:

        -- statin users: 14.7% (13.0%-16.6%)

        -- non-statin users: 20.0% (18.6%-21.5%)

    -- those with high baseline FIB-4 scores who transitioned to lower FIB-4 groups:

        -- transitioning to the intermediate group:

            -- statin users: 31.8% (20.0%-35.9%)

            -- non-statin users: 18.8% (17.2%-20.6%)

        -- transitioning to the low-risk group:

            -- statin users: 7.0% (5.2%-9.6%)

            -- non-statin users: 4.3% (3.5%-5.2%)

 

Subgroup analysis

Cirrhosis:

    -- patients with cirrhosis (n= 6734):

        -- 10 year incidence of HCC:

            -- statin users: 6.3%

            -- statin nonusers: 10.2%

                -- 18% better with statins, aSHR 0.82 (0.69- 0.98)

        -- 10 year incidence of hepatic decompensation

            -- statin users: 17.7%

            -- statin nonusers: 29.3%

                -- 20% better with statins, aSHR 0.80 (0.65-0.98)

     --patients without cirrhosis (n= 9767)

        -- 10 year incidence of HCC:

            -- statin users: 2.4%

            -- statin nonusers: 6.2%

                -- 38% better with statins, aSHR 0.62 (0.51 - 0.75)

        -- 10 year incidence of hepatic decompensation

            -- statin users: 6.8%

            -- statin nonusers: 12.0%

                -- 24% better with statins, aSHR 0.76 (0.59-0.98)

 

    -- patients with dyslipidemia (not defined) (n= 5746):

        -- 10 year incidence of HCC:

            -- 57% better with statins, aSHR 0.43 (0.33-0.56)

    -- patients without dyslipidemia (n= 10755):

        -- 10 year incidence of HCC:

            -- 24% better with statins, aSHR 0.76 (0.65-0.90)

 

    -- patients with MASLD (n= 5104):

        -- 10 year incidence of HCC:

            -- 28% better with statins, aSHR 0.72 (0.56-0.93)

 

    -- patients with other causes of chronic liver disease (n= 5104):

        -- 10 year incidence of HCC:

            -- 32% better with statins, aSHR 0.68 (0.60-0.78)

                -- every CLD type was better with statins (the non-statistically significant ones had relatively small numbers of patients):

                    -- statistically significant ones included alcohol, MASLD, hepatitis B and C viruses

                    -- strong trend towards statistical significance: primary biliary cirrhosis

                    -- no apparent difference: autoimmune hepatitis

 

    -- Other factors on exploratory analysis associated with decreased HCC risk:

        -- aspirin users: 28% better with statins, aSHR 0.72 (0.61-0.75); P= 0.006 for interaction

        -- metformin users: 53% better with statins, aSHR 0.47 (0.32-0.70); P=0.01 for interaction

        -- FIB-4: transitioning from intermediate to high risk scores: 14.7% of statin users versus 20% of nonusers, P<0.001 for interaction

        -- statins were statistically more effective in those with diabetes, chronic kidney disease, strokes, and those without peripheral vascular disease (the latter having very small number of patients)

        -- FIB-4: transitioning from high baseline FIB-4 to intermediate scores: 31.8% of statin users versus 7.0% of nonusers, P<0.001 for interaction

        -- also, statins statistically better in both men and women, and in those with BMI at least 25

 

Commentary:

-- there may well be a lingering concern that statins, known to occasionally increase hepatic transaminases, might be deleterious in those with underlying chronic liver disease. However, several studies in the past have found that this is not the case, that patients with various viral or nonviral hepatitis actually seem to do better.

     -- Another recent study of 184 patients having cirrhosis and severe portal hypertension and high-risk varices but who had a suboptimal response to carvedilol found that adding simvastatin vs placebo was associated with a decrease in the portal pressure hepatic venous pressure gradient: https://journals.lww.com/hep/abstract/9900/hemodynamic_effects_of_carvedilol_plus_simvastatin.1072.aspx

 

-- the presumptive mechanism for improvement in CLD outcomes by statins is through several of the different physiologic effects of statins, including their being anti-inflammatory, anti-fibrotic, antioxidant, and improving endothelial function. these benefits may be more impressive with the use of atorvastatin in some studies.

 

-- This study has the advantage of looking at a large number of patients with multiple causes of chronic liver disease, finding:

    -- statin use was associated with decreased likelihood of the very important outcomes of progression to HCC and hepatic decompensation

    -- the higher the cumulative dose of statins was associated with increased benefit, suggesting that prolonged statin use is better

    -- those on statins were more likely to have improved calculated FIB-4 scores, a marker of hepatic fibrosis, being more likely to transition from high to intermediate or to low FIB-4 scores and less likely to remain in the high-risk group; those not on statins tended to have progressively worse FIB-4 calculations

    -- and, unlike prior studies, this study found the benefit of statins incorporating large numbers of patients from 10 different hospitals with many different causes of liver disease, with access to lots of granular data allowing them to calculate FIB-4 scores and their changes over time, and also to control for more potentially important factors (for example, low-dose aspirin had been found to decrease the risk of HCC and liver-related mortality in patients with hepatitis B or C infections: https://www.nejm.org/doi/10.1056/NEJMoa1912035 ). and this was also found in subgroup analysis in this current study with a 28% reduction of HCC

    -- another finding was though there was clear benefit from statins, the lipophilic ones outperformed the hydrophilic ones

 

Limitations:

-- this was an observational study, so the results can only be interpreted as associations and not causality

    -- for example, many potentially relevant social factors were not included, such as socioeconomic status, health care access, medical literacy

    -- and, we do not know why clinicians prescribed the statins to individual patients or did not prescribe them. did some not prescribe them because they were afraid to cause hepatic injury, or perhaps they did prescribe them because they had seen some of the older studies (and these clinicians were perhaps more versed in other aspects of HCC/hepatic decompensation management as well)? These issues could introduce a selection bias at the start of the study:

        -- they did pretty sophisticated propensity-score statistical analysis to attempt to equalize the groups (which is certainly helpful)

        -- but this still is different from having a randomized controlled trial, where the basic selection bias is the inclusion and exclusion criteria. but in this and many studies, RCTs are remarkably complex and largely undoable. it is hard to invite patients for a many year study and randomize them to statin vs placebo. For example, would patients with diabetes, who should be on statins, be excluded from the study (which would eliminate huge numbers of patients and make the results much less useful); this same concern would apply for many of the other chronic diseases (CKD, etc)? or should they include these patients and make the results less interpretable since there was not clinically useful randomization?

-- the group studied was hospital-based and may not be generalizable to others

-- this study excluded those with HIV, limiting generalizability of the results to this group

-- this was largely a group of white individuals (about 80%) with mean age of 60, again perhaps limiting generalizability to others (breakdown to those <60 and >60 had similar results on subgroup analysis, though subgroup analyses are also less statistically robust)

-- one frequent and extremely important issue in these types of data-mining studies is that the patient characteristics reflect only the initial assessment at the beginning of the study and does not include changes during the course of the study.  did some develop coronary artery disease? did medications change (eg metformin/aspirin, or perhaps beginning therapy for hepatitis B/C change?, etc)? did they start or increase or stop drinking alcohol?

-- this study did not provide information on cholesterol levels. We do know that lower LDL cholesterol levels improve cardiovascular outcomes. Would higher doses of statins help with preventing liver complications?  It would be clinically useful to know the optimal statin dose in treating patients, though this would require further studies. just looking at big buckets of cumulative statin intake (<30, 30-599, >600) is really not granular enough

 

so,

-- there have been several studies of the safety of statins in patients with some types of hepatitis, and these studies have pretty much all (to my knowledge) found some benefit. this study suggested that patients with hepatic inflammation from an array of causes and array of severity (eg 38% had cirrhosis) do seem to benefit from statins, and more so from the lipophilic ones

-- this consistent finding does run counter to our clinical concern that statins, known to cause increases in inflammatory hepatic enzymes in some people, does not seem to be a concern:

    --  several high-quality studies over the past few decades have attested to statin safety

    -- this study suggests patients with chronic liver disease from many different causes should perhaps be on the list of medical problems benefiting from statins, and that we clinicians should have a very low threshold to at least consider statins as an appropriate therapy for these individuals

    -- it would be useful to have information on the etiologies of CLD not included in this analysis (hemochromatosis, etc) or with inadequate numbers of patient to be statistically valid

-- the study also raises again the potential benefit of aspirin. It would be clinically useful to have a real risk/benefit analysis of statins vs aspirin vs both in patients with chronic liver disease

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them