Hyperlipidaemia in primary biliary cholangitis: treatment, safety and efficacy

Abstract

Primary biliary cholangitis (PBC) is an autoimmune liver disease associated with altered lipoprotein metabolism, mainly cholesterol. Hypercholesterolaemia, a major modifiable risk factor for cardiovascular disease in the general population, occurs in 75%–95% of individuals with PBC. The impact of hypercholesterolaemia on cardiovascular risk in PBC, however, is controversial. Previous data have shown that hypercholesterolaemia in PBC is not always associated with an increase in cardiovascular events. However, patients with PBC with cardiovascular risk factors may still warrant cholesterol-lowering therapy. Treatment of hypercholesterolaemia in PBC poses unique challenges among primary care providers due to concerns of hepatotoxicity associated with cholesterol-lowering medications. This review summarises the current understanding of the pathophysiology of hypercholesterolaemia in PBC and its pertinent cardiovascular risk. We will also discuss indications for treatment and the efficacy and safety of available agents for hypercholesterolaemia in PBC.

Introduction

Dyslipidaemia in primary biliary cholangitis (PBC) is common. Although most patients with PBC have normal to slightly elevated triglyceride, hypercholesterolaemia occurs in 75%–95% of patients with PBC.^1–3 Hypercholesterolaemia, especially elevated low-density lipoprotein cholesterol (LDL-C), is a well-established modifiable risk factor for cardiovascular disease (CVD) in the general population.⁴ Whether or not hypercholesterolaemia carries the same risk for CVD in the PBC population is controversial.^5–7 Part of the confusion stems from the multifaceted effect of PBC on serum lipids with both increases in the levels of LDL, which increases cardiovascular risk, and in the levels of adiponectin and lipoprotein X (Lp-X), a circulating lipid particle with density similar to LDL that has been suggested to confer cardioprotection.^{8 9} Further adding to the confusion, PBC therapies such as ursodeoxycholic acid (UDCA) and fibrates impact lipid metabolism as well. UDCA increases cholesterol absorption and fibrates modulate bile acid and cholesterol transportation.¹⁰ Therefore, it is essential to fully understand lipid metabolism in patients with PBC to tailor therapies.

In addition to confusion surrounding cardiovascular risk due to hypercholesterolaemia in PBC, concern exists regarding the use of lipid-lowering medications in patients with PBC. Since many medications are metabolised by the liver, the worry has been that reduced metabolic clearance could increase drug concentrations and induce toxicity in a liver already compromised by PBC. Early reports that statins were especially hepatotoxic have not been borne out.¹¹ Given the paucity of evidence, confusion surrounding the significance of hypercholesterolaemia in PBC, as well as the complicated consideration regarding therapy, neither the European nor American liver societies make any strong recommendations regarding monitoring or treatment.^{12 13} In order to fill this gap, we review the safety and efficacy of lipid-lowering medications in patients with PBC and their possible interactions with PBC treatments.

Pathophysiology

The pathogenesis of hypercholesterolaemia in PBC is complex. Plasma cholesterol levels increase through intestinal absorption of dietary and biliary cholesterol, endogenous cholesterol synthesis, which occurs primarily in the liver, and secretion of very low-density lipoprotein (VLDL) particles into the circulation.¹⁴ On the other hand, cholesterol levels are decreased by bile acid synthesis, biliary excretion of cholesterol and hepatic uptake of cholesterol via the LDL receptor (LDLR).¹⁴

In the setting of cholestasis, reduced bile acid secretion impairs micellar formation and leads to decreased intestinal cholesterol absorption. This stimulates intrahepatic cholesterol synthesis. At the same time the removal of LDL from the circulation is decreased due to both decreased LDLR activity as well as altered LDL composition.¹⁵ Although these responses may be impaired in late disease due to the loss of functioning hepatocytes, the end result of these processes results in increased serum cholesterol levels.

The composition of serum cholesterol in PBC is altered as well given the presence of Lp-X. Lp-X, strictly speaking, is not a lipoprotein particle, but a vesicle rich in phospholipids and unesterified cholesterol and low in cholesterol ester and triglycerides. Lp-X is nearly absent of apolipoprotein B (ApoB); rather it carries albumin in the aqueous core. In the healthy liver, lecithin-cholesterol acyltransferase (LCAT) converts the free cholesterol in Lp-X into the cholesterol esters found in LDL.¹⁶ In PBC, decreased LCAT function leads to changes in lipoprotein composition, decreased LDL and increased Lp-X.¹⁶ Lp-X prevents LDL oxidation, thus protecting endothelial cells and slowing atherosclerosis.⁸ However, two distinct lipoprotein patterns have been observed; early and intermediate histological stages of PBC are associated with mild elevations of VLDL and LDL, with increased high-density lipoprotein (HDL). This is in contrast to advanced stage PBC with marked elevations in LDL with Lp-X and decreased HDL.¹⁷

In addition to Lp-X, alterations in other lipoproteins likely confer cardioprotection. HDL levels are elevated in all but end-stage PBC.^{17 18} Apolipoprotein A1, a major protein constituent of HDL, contains two major classes of apolipoprotein A1-containing lipoproteins: LpA1 and LpA1:A2. LpA1 may also have antiatherogenic properties¹⁹ and is found in greater concentrations compared with LpA1:A2 in patients with PBC compared with controls.¹⁸ Adiponectin, which is also associated with protection against atherosclerosis, is also increased in patients with PBC compared with controls.¹⁰ On the other hand, lipoprotein(a), which is considered an independent risk factor for CVD, is not elevated or even reduced in patients with PBC.²⁰

Triglyceride levels are normal to somewhat elevated in PBC.¹ This is thought to be due to decreased lipoprotein lipase levels in PBC.

Cardiovascular risk

While some studies have shown impaired cardiac function²¹ and increased cardiovascular risk in PBC,²² the majority of the evidence indicates that hypercholesterolaemia in PBC does not increase cardiovascular risk in the absence of concomitant metabolic syndrome.⁷ One study found autonomic dysfunction in the form of impaired capacity of the left ventricle to respond to orthostasis in patients with PBC compared with controls; however, there was no difference in cardiac structure, baseline function or cardiovascular outcomes.²¹ One other study demonstrated increased risk of incident coronary artery disease (CAD) in PBC.²² However this study compared individuals hospitalised in Sweden for immune-mediated diseases (including PBC) with the general, non-hospitalised population, essentially demonstrating hospitalised patients have increased risk of CAD. Similarly, a single meta-analysis of four studies, which found an increased risk ratio of CAD in PBC, relied heavily on the Swedish study with 52.3% weight.²³ Numerous other prospective and retrospective cohort studies have not shown increased incidence of CVD compared with the general population³ or age-matched and sex-matched controls.^{1 5} In addition, hypercholesterolaemia in PBC has not been associated with subclinical atherosclerosis as determined by carotid intima-media thickness compared with patients with PBC without hypercholesterolaemia.⁶

Patients with PBC and hypertension do appear to have significantly increased risk of cardiovascular events and risk of mortality compared with the general population.^{3 24} Furthermore, patients with PBC with metabolic syndrome have significantly more cardiovascular events than patients with PBC without metabolic syndrome.²⁴ This is significant as metabolic syndrome is present in 30% of patients with PBC.²⁴

Management

Initial evaluation

Lipid-lowering therapy should be individualised based on CVD risk assessment and comorbidities (figure 1). All patients with PBC should be assessed for traditional risk factors for CVD such as tobacco use, diabetes, hypertension or elevated body mass index (BMI), and have a lipid panel including ApoB-100 levels checked. ApoB-100 has a linear correlation with LDL-C and was demonstrated as a surrogate of controlled levels of lipids during treatment.²⁵ Conventional lipid panel does not distinguish Lp-X from LDL-C, making it impossible to use traditional LDL-C measurement for risk stratification in PBC. Since ApoB-100 is not present in Lp-X, it may be helpful to differentiate LDL-C from Lp-X. Agarose gel electrophoresis and nuclear magnetic resonance spectroscopy (NMR)-based lipoprotein profiling are potential alternative methods to distinguish Lp-X from LDL, but are limited by cost and availability.²⁶ Since reliance on LDL-C from conventional lipid panels may lead to overtreatment of hypercholesterolaemia in PBC, we propose the use of ApoB-100 for risk stratification in the context of PBC.

Figure 1.

Suggested algorithm for the evaluation of hypercholesterolaemia in PBC. Treatment of hypercholesterolaemia is recommended in patients with PBC with pre-existing cardiovascular disease, diabetes, primary hyperlipidaemia (ApoB-100 >120 mg/dL) and in those with cardiovascular risk factors (tobacco use, hypertension) and ApoB-100 >90 mg/dL. ApoB, apolipoprotein B; PBC, primary biliary cholangitis.

We suggest treatment of hypercholesterolaemia in PBC in patients with pre-existing CVD, diabetes or primary hypercholesterolaemia (an ApoB-100 level >120 mg/dL based on the Canadian Cardiovascular Society guidelines, the only such organisation to make specific recommendations based on ApoB-100 levels).²⁷ In patients with PBC without pre-existing CVD, diabetes or primary hypercholesterolaemia, but additional cardiac risk factors (ie, tobacco use, diabetes, hypertension, elevated BMI), we suggest treatment in those with an ApoB-100 level >90 mg/dL, corresponding to a true LDL level >116 mg/dL.²⁸

Specific recommendations regarding dosing are difficult to make in light of sparse data. The most recent American College of Cardiology/American Heart Association guidelines recommend high-intensity statins (expected to reduce LDL-C by ≥50%; atorvastatin 40–80 mg daily or rosuvastatin 20–40 mg daily) or moderate-intensity statins (expected to reduce LDL-C by 30%–50%; atorvastatin 10–20 mg daily, rosuvastatin 5–10 mg daily or simvastatin 20–40 mg daily) based on risk stratification.²⁹ Since data on risk stratification within PBC are not available and most studies have only examined moderate-intensity statins (atorvastatin 10–20 mg daily or simvastatin 20–40 mg daily) (table 1), we generally start at these doses and uptitrate as clinically indicated. To assess response and compliance, we recommend repeating a lipid panel with ApoB-100 level in 3 months with an expected reduction in ApoB-100 of 30%–50%.²⁹ In those patients who do not achieve expected reduction in ApoB-100 and are compliant with the medication, we suggest uptitration of the statin with use of a proprotein convertase subtilisin kexin 9 (PCSK9) inhibitor as second-line therapy in those intolerant of a statin or in whom ApoB-100 reduction is less than expected despite statin uptitration. Figures 1 and 2 illustrate our proposed algorithm for the evaluation and treatment of hypercholesterolaemia in PBC. Table 1 summarises different studies investigating lipid-lowering therapies in PBC and chronic liver disease.

Table 1.

Summary of studies on lipid-lowering therapies in primary biliary cholangitis

Study	Intervention	Design	Outcomes	Safety
Statins
Cash et al 2	Simvastatin 20 mg daily vs placebo	RCT, n=11 (simvastatin), n=10 (placebo)	Significant decrease in TC and LDL-C. No change in vascular compliance.	One patient discontinued due to myalgias. No worsening in liver disease or biochemical markers.
Abu Rajab and Kaplan32	Various statins in different doses	Retrospective cohort, n=58	30% decrease in TC.	No adverse effects reported. No increase in ALT.
Stojakovic et al 33	Atorvastatin 10 mg daily	Single-arm trial, n=19	Significant decrease in TC and LDL-C. No change in vascular compliance.	AP increased in one patient. Myositis in one patient.
Stanca 200855	Atorvastatin in various doses	Retrospective cohort	Significant decrease in TC and LDL-C.	One patient discontinued treatment due to myalgias.
Del Puppo et al 31	Simvastatin 40 mg daily, UDCA 600 mg daily	Crossover trial, n=6	Greater and more rapid reduction in TC in simvastatin compared with UDCA.	No adverse effects reported.
Fibrates
Corpechot et al 37	Bezafibrate 400 mg daily vs placebo	RCT, n=50 (bezafibrate), n=50 (placebo)	Significant decrease in TC, LDL-C and AP.	6% increase in transaminases vs 2% in placebo. Myalgias (20%) more common than in placebo (10%). Greater increase in creatinine (5%) than in placebo (−3%).
Hegade et al 39	Fenofibrate (median dose 200 mg daily)	Retrospective cohort	Significant decrease in AP.	No adverse events reported.
Hosonuma et al 36	Bezafibrate 400 mg daily + UDCA vs UDCA alone	RCT	Significant decrease in AP and Mayo score.	One patient discontinued due to renal dysfunction and one due to myalgias. Higher creatinine levels in bezafibrate compared with placebo.
Cheung et al 40	Fenofibrate 200 mg daily + UDCA vs UDCA alone	Retrospective cohort, n=13 (fenofibrate), n=14 (UDCA alone)	Greater biochemical response and less decompensation and transplant. Decreased TG.	Seven patients discontinued treatment. Most common abdominal pain and myalgias.
Han 201256	Fenofibrate 200 mg four times a day + UDCA	Single-arm trial, n=22	Significant decrease in AP, TC and TG.	No adverse effects reported.
PCSK9 inhibitors
Gibbs et al 48	Evolocumab 140 mg daily	Open-label parallel group, n=8 (healthy), n=8 (mild hepatic impairment), n=8 (moderate hepatic impairment)	Similar LDL-C reductions in all three groups.	No safety risks identified.
UDCA
Poupon et al 49	UDCA 13–15 mg/kg/day vs placebo	RCT, n=17 (UDCA), n=15 (placebo)	Significant decrease in TC and LDL-C.	No adverse events reported.
Balan et al 50	UDCA vs placebo	RCT	Significant decrease in TC.	No adverse events reported.

AP, alkaline phosphatase; LDL-C, low-density lipoprotein cholesterol; PCSK9, proprotein convertase subtilisin kexin 9; RCT, randomised controlled trial; TC, total cholesterol; TG, triglycerides; UDCA, ursodeoxycholic acid.

Figure 2.

Suggested algorithm for the treatment of hypercholesterolaemia in primary biliary cholangitis. We suggest initiation of moderate-dose statin (atorvastatin 10–20 mg daily or simvastatin 20–40 mg daily) as first-line therapy in all patients requiring treatment. In those patients intolerant of statins or in whom adequate ApoB-100 reduction cannot be achieved with statins alone, we suggest a PCSK9 inhibitor as second-line therapy. ApoB, apolipoprotein B; PCSK9, proprotein convertase subtilisin kexin 9.

Statins

Statins work by inhibiting 3-hydroxy-3-methylglutaryl coenzyme A reductase, the rate-limiting enzyme in cholesterol synthesis, and are the mainstay in the treatment of hypercholesterolaemia in the general population. Statin therapy is effective in reducing LDL-C levels and major coronary events in patients at risk in the general population.³⁰ However, initial reports of liver injury limited their use in chronic liver disease.¹¹ In addition, already decreased cholesterol synthesis and dysregulated cholesterol metabolism in PBC led to concern as to the efficacy of statins in PBC.¹⁵ However, subsequent studies have demonstrated their safety in chronic liver disease and their effectiveness in the treatment of hypercholesterolaemia in PBC.³¹

One initial study of simvastatin 40 mg daily in six patients with PBC confirmed its effectiveness in lowering total cholesterol with a 34% decrease in 30 days.³¹ A larger retrospective study, in which 58 patients with PBC on a variety of statins were followed for a mean of 41 months, did not show any increase in alanine aminotransferase levels or other significant adverse events.³² In a randomised controlled trial of 20 mg simvastatin or placebo in 21 patients with PBC for 12 months, serum LDL levels were significantly lower with no elevation in ALT or hepatic decompensation.²

Two studies have looked at low-moderate dose of atorvastatin (10–20 mg) in patients with PBC and hypercholesterolaemia. They similarly showed that atorvastatin had no effects on liver enzymes or Mayo score and reduced the total cholesterol by 28% and LDL by 35%,³³ a modest reduction compared with general population in real-life data.

Recent data from a systematic review and meta-analysis showed that statin use in patients with chronic liver disease is probably associated with lower risk of hepatic decompensation and mortality, and thus lowering the risk of hepatic disease progression.³⁴ Moreover, the presence of chronic liver disease or Child-Pugh A is not a contraindication for statin use. However, the likelihood of benefit in the use of statins for primary prevention in patients with a 10-year cardiovascular event risk of less than 10% and without traditional risk factors is small, and there is uncertainty in individual risk prediction.³⁵

Statins should be approached with caution in decompensated cirrhosis. Inconsistent data are available in the use of statins in decompensated cirrhosis,³⁴ and its utility is questionable in the late stages of PBC or patients with severe cholestasis. There may be some benefits, but if used dose should be reduced and patients should be monitored closely and frequently. Whether decompensated patients with Child B and C cirrhosis are at risk for increased hepatotoxicity or rhabdomyolysis is uncertain.

Fibrates

Fibrates, or peroxisome proliferator-activated receptor (PPAR) agonists, were originally designed to treat hypertriglyceridaemia; however, interest in their use to treat PBC has increased in recent years. Two recent studies have shown that bezafibrate in combination with UDCA is effective in decreasing alkaline phosphatase and Mayo risk score compared with UDCA alone or with placebo in those refractory to UDCA monotherapy.³⁶ Bezafibrate also resulted in a 26% reduction in serum LDL compared with placebo.³⁷ However, bezafibrate also resulted in an increase in creatinine and increased myalgias compared with placebo, which may require closer monitoring.³⁷

Fenofibrate has also been studied in patients with PBC with incomplete biochemical response to UDCA. In small studies, fenofibrate exhibits modest activity in decreasing total cholesterol and triglycerides.³⁸ However, 26% of the patients had an early abandonment of the treatment due to drug intolerance; 15% experienced abdominal pain and myalgia and 8% stopped taking the medication due to an increase in liver enzymes.^{39 40} In terms of renal disease, there was no difference from their baseline creatinine in the 5 years of follow-up.³⁹

Fenofibrate plus UDCA has also been compared with bezafibrate plus UDCA.⁴¹ In this study, 14 patients received fenofibrate 80 mg daily plus UDCA and 7 patients received bezafibrate 400 mg daily plus UDCA for 48 weeks. Both had similar improvement in alkaline phosphatase levels; however, fenofibrate led to more significant LDL reduction.⁴¹ The authors hypothesise this is due to stronger PPARα agonist activity.

While statin therapy is generally considered effective for the prevention of cardiovascular events in the general population,³⁰ the benefit of fibrates is more controversial. The Action to Control Cardiovascular Risk in Diabetes study evaluated the combination of simvastatin plus fenofibrate in 5518 patients with diabetes and did not demonstrate improvement in cardiovascular events.⁴² A subsequent meta-analysis of fibrates did show a 10% relative risk reduction (95% CI 0 to 18) in major cardiac events, but did not improve cardiovascular mortality.⁴³ In light of these data, despite the benefit of fibrates in the treatment of PBC, it is difficult to recommend fibrate use over statins for the reduction of cardiovascular risk.

Ezetimibe

Ezetimibe is a cholesterol absorption inhibitor targeting jejunal enterocytes. In the general population, it is effective at reducing LDL levels and risk of cardiovascular events.⁴⁴ However, in patients with PBC, poor micellar formation already inhibits cholesterol absorption, hypothetically attenuating the lipid-lowering effect of ezetimibe.¹⁴ One retrospective study evaluating statin use also included five patients on ezetimibe alone. They did not note any increase in ALT or other side effects; however, they did not comment on its effectiveness at lowering cholesterol.³² In the absence of better data, the routine use of ezetimibe in patients with PBC cannot be recommended.

Combination therapy

The combination of statin-fibrate has not been shown to reduce cardiovascular mortality or events compared with statin alone in the general population,⁴⁵ and especially given concerns for increased liver injury we do not recommend the combination in PBC. Combination statin-ezetimibe is effective at further lowering LDL and reducing cardiovascular events compared with statin plus placebo⁴⁶; however, given concerns for lack of ezetimibe efficacy in cholestatic liver disease,¹⁴ there are insufficient data to recommend combination statin-ezetimibe in PBC.

PCSK9 inhibitors

PCSK9 inhibitors, such as alirocumab and evolocumab, are the newest therapies approved for the treatment of hypercholesterolaemia. PCSK9 binds to the LDLR, preventing LDL-C clearance. PCSK9 inhibitors block PCSK9 activity at the LDLR, thereby increasing hepatic uptake of LDL and decreasing serum LDL levels. Evolocumab has been shown to lower total cholesterol more than the combination of high-dose statins and ezetimibe.⁴⁷ In a meta-analysis consisting of 25 randomised control trials, treatment with evolocumab was not associated with liver enzymes abnormalities compared with placebo.⁴⁷ Furthermore, no safety risks were identified in 16 patients with mild and moderate liver disease (Child-Pugh A and B) compared with healthy adults.⁴⁸ No studies are available in terms of safety for evolocumab in PBC or decompensated cirrhosis.

Ursodeoxycholic acid

UDCA is the mainstay therapy for PBC and has also been shown to ameliorate hyperlipidaemia, mainly total cholesterol. UDCA lowers total cholesterol, LDL-C and VLDL with no effect on HDL-C or total triglycerides.^{49 50} The change is strongly correlated with serum bilirubin indicating the mechanism may be related to improvement in the underlying liver disease or direct modulation of cholesterol metabolism, improving cholestasis.^{49 50} Since Lp-X accumulates in cholestasis, lipid-lowering of UDCA in PBC may be driven by improvements in cholestasis. UDCA is a safe medication in cholestatic liver disease with limited side effects.⁵¹

Obeticholic acid

Obeticholic acid (OCA) is a farnesoid X receptor ligand recently approved for the treatment of PBC with inadequate response to UDCA. Concerns regarding increased LDL and total cholesterol were raised among patients with non-alcoholic steatohepatitis treated with OCA.⁵² Decreased total cholesterol due to HDL reduction has been noted in patients with PBC, although the effect on LDL has been mixed, with some reporting no change and others an increase.^{53 54} None has confirmed whether any potential increase in LDL is from Lp-X or true LDL-C. Until further data accumulate regarding the lipid effects of OCA in patients with PBC, it is reasonable to check a lipid panel with ApoB-100 level 3 months after initiation of therapy.

Conclusion

Hypercholesterolaemia is common in patients with PBC. In general, this does not appear to confer increased cardiovascular risk due to alterations in lipid composition leading to increased cardioprotective Lp-X, HDL and adiponectin with decreased atherogenic LDL-C and lipoprotein(a). A subset of patients with PBC with other risk factors such as tobacco use or metabolic syndrome remain at risk of CVD. In this group, measuring ApoB-100 may be helpful in determining in whom to start lipid-lowering therapy. Statins appear to be safe and effective in PBC and should be used as first-line therapy. Fibrates have the added benefit of treating underlying PBC, although the cardioprotective effect of statins is likely superior, and should not be used solely for treatment of hypercholesterolaemia if the underlying PBC is otherwise controlled on UDCA. PCSK9 inhibitors are also likely safe and effective and may be used in those unable to tolerate statins or whose disease is well controlled with UDCA alone and do not require fibrates. Long-term studies are needed to demonstrate the benefit of lipid-lowering therapy on cardiovascular mortality in PBC and better define treatment goals using ApoB-100 levels or NMR-measured LDL.

Significance of this study.

What is already known on this topic?

• Patients with primary biliary cholangitis (PBC) have significantly higher total cholesterol than the general population due in part to the lipoprotein-like particle lipoprotein X.

• Despite hypercholesterolaemia, patients with PBC are not at increased cardiovascular risk compared with the general population.

• Patients with PBC with components of the metabolic syndrome such as hypertension are at increased risk of cardiovascular events.

What this study adds?

• We provide a suggested strategy for monitoring and treating hypercholesterolemia in PBC based on current uncerstanding of cardiovascular risk and treatment safety and efficacy in this patient population

How might it impact on clinical practice in the foreseeable future?

• Statins are safe and effective at lowering low-density lipoprotein (LDL) cholesterol in patients with PBC, but should be used carefully in decompensated cirrhosis.

• Fibrates may be effective at treating PBC, but are not as effective in lowering LDL cholesterol or improving cardiovascular mortality.

• Proprotein convertase subtilisin kexin 9 inhibitors are also safe and effective in chronic liver disease and are preferred second-line therapy over fibrates or ezetimibe in patients with PBC well-controlled on ursodeoxycholic acid.