Adverse drug reactions of non-statin antihyperlipidaemic drugs in China from 1989 to 2019: a national database analysis

Abstract

Objective

This study aims to understand the adverse drug reactions (ADRs) for non-statin antihyperlipidaemic drugs included in the China Anti-hyperlipidemic Drug Database.

Design

An approach of Chinese national database analysis was employed to screen clinical trials involving non-statin antihyperlipidaemic drugs from 1989 to 2019.

Setting

The database was provided by the China National Medical Products Administration Information Centre.

Participants

In total, 117 clinical studies with 8800 patients were selected from 2650 clinical trials of the Anti-hyperlipidemic Drug Database.

Interventions

The non-statin antihyperlipidaemic drugs were divided into three groups: (1) fibrates (fenofibrate, gemfibrozil, bezafibrate, etofylline clofibrate); (2) nicotinic acid and derivatives (niacin, acipimox) and (3) others (probucol, cholestyramine).

Results

The results of this study show that first, gastrointestinal symptoms were the most common reactions (6.975%), which account for approximately 50% of the reported cases with ADRs. Second, cholestyramine (16.418%) and gemfibrozil (13.158%) were the most common gastrointestinal side effect-causing non-statin antihyperlipidaemic drugs, which account for one-third of the population. Third, niacin (7.879%) and gemfibrozil (5.000%) were the most likely cause of liver disease symptoms. Finally, niacin (10.909%) and acipimox (18.847%) were the major non-statin antihyperlipidaemic drugs with skin symptoms.

Conclusion

This study revealed that gastrointestinal symptoms were the most common ADRs of fibrates, probucol and cholestyramine in the Chinese population. For nicotinic acid and derivatives, the ADRs of skin symptoms were the most common in China.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The authorised national database was used to analyse the adverse drug reactions (ADRs) of non-statin antihyperlipidaemic drugs in China from 1989 to 2019.

• The difference in the ADR ratios between fibrates was compared in the Chinese population.

• Omega-3 fatty acids, proprotein convertase subtilisin/kexin type 9 inhibitor inhibitors, adenosine triphosphate-citrate lyase inhibitors and ezetimibe were not included in the study.

Introduction

Lipoproteins consist of esterified and unesterified cholesterol, triglycerides (TGs), phospholipids and apolipoproteins.¹ There are six major lipoproteins in blood: chylomicrons, very low-density lipoprotein, intermediate-density lipoprotein, low-density lipoprotein (LDL), lipoprotein(a) and high-density lipoprotein (HDL).¹ Lipoproteins in plasma are responsible for energy utilisation, lipid deposition, steroid hormone production and bile acid formation through transporting lipid to tissues.¹

Dyslipidaemia, a significant risk factor for cardiovascular disease (CVD) events, is correlated with elevated LDL cholesterol and TG levels.^{1 2} Among lipid-lowering drugs (LLDs), statins are the primary prescription and the cornerstone of therapy, and the others are ezetimibe, bile acid sequestrants and proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors.² TG-lowering drugs (TLDs), such as fibrates and niacin, have a mild LDL-lowering action but are not suggested to use as add-on drugs to statin therapy.²

Among the non-statin antihyperlipidaemic drugs, ezetimibe is the most frequently prescribed and is associated with a low incidence of life-threatening liver failure.^1–4 Bile acid sequestrants are not absorbed and, therefore, do not cause systemic side effects; however, they may cause gastrointestinal complaints and severe hypertriglyceridemia.² PCSK9 inhibitors are highly effective at lowering LDL levels and are generally well tolerated.^2–6 Fibrates and niacin are useful in certain patients with severe hypertriglyceridemia.² Fibrates are generally well tolerated with mild adverse effects, such as gastrointestinal disturbances and skin rashes, while nicotinic acid has been associated with an increased frequency of serious adverse effects.¹

Using a national database, we found that gastrointestinal symptoms were the most common statin adverse drug reaction (ADR) for dyslipidaemia in Chinese clinical trials from 1989 to 2019, followed by hepatic disease and muscle symptoms.⁷ However, there was a lack of studies on systematic and long-term analysis of adverse drug reactions (ADRs) of non-statin antihyperlipidaemic drugs in China. In this study, we aimed to assess the clinical trials with non-statin antihyperlipidaemic drugs to clarify the ADRs in the Chinese population.

Materials and methods

Data sources

This study obtained all authorised data on published clinical trials of antihyperlipidaemic drugs in China from the Anti-hyperlipidemic Drug Database (https://www.ncmi.cn/phda/dataDetails.do?id=CSTR:A0006.11.A0008.201905.000498-V1.0) in the National Population Health Data Center, which was provided by the Information Center of the China National Medical Products Administration.

Data description

The Anti-hyperlipidemic Drug Database encompasses a broad range of data, including 26 columns such as the quantity of clinical trials, total cases, sex-specific cases, cases in the treatment and control groups, age span, average age, associated disease and complications, Chinese generic and trade names of drugs, English names of drugs, dosage forms, specifications, approval numbers, manufacturers, classifications, medication purposes, usage and dosage, combined medication, efficacy evaluation criteria, comparison of therapeutic effects and clinical indications, ADRs reported in the clinical trial, other reported ADRs, information sources and simplified code in pinyin.⁷ The database included 2650 clinical trials, both controlled and non-controlled, from 1989 to 2019 (updated to 24 October 2019). The non-statin antihyperlipidaemic drugs were used to treat or prevent CVD in the clinical trial of the database. The single-armed trials were defined as the treatment group. The detail of the data description was described in our previous study.⁷

Inclusion and exclusion criteria

The clinical trials with non-statin antihyperlipidaemic drugs, including controlled and single-arm trials, were selected from the database by manually comparing the columns ‘Chinese generic name of drugs’, ‘Chinese trade name of drugs’ and ‘English name of drugs’. The non-statin antihyperlipidaemic drugs were defined based on the 2019 European Society of Cardiology/European Atherosclerosis Society Guidelines for the management of dyslipidaemias.¹ Probucol was also defined as one of the non-statin antihyperlipidaemic drugs.⁸ We conducted a comprehensive review of available clinical trial data for non-statin antihyperlipidaemic drugs in the Anti-hyperlipidemic Drug Database. However, we were unable to find any clinical trials for PCSK9 inhibitors and adenosine triphosphate-citrate lyase inhibitors.⁹ Additionally, ezetimibe was only used as an add-on drug in the database. Omega-3 fatty acids, such as eicosapentaenoic acid and docosahexaenoic acid,¹ which are also known as n-3 fatty acids, were not examined in this study due to the challenge in defining them within the Anti-hyperlipidemic Drug Database. Then, we also removed the trials with incomplete ADR information. The trial-based data selection procedure is shown in figure 1.

Figure 1.

The procedure of trial-based data selection.

Definition and classification of ADRs

The definition and classification of ADRs were described in our previous publication.⁷ In brief, the ADR type and rate were assessed by analysing the column ‘reported ADR of the clinical trial’, and each ADR term in the column was selected for counting. One case reported ‘loose stool’ in the gemfibrozil-treated group; this was difficult to define and excluded. The case number of ADRs was calculated using the patient number described in the column. In addition, ADR severity was ignored for the analyses because there was no description in most of the trials. The ADR classification included symptoms related to the muscles, liver disease, gastrointestinal, neurological, cardiovascular, skin, urinary and other areas, as shown in table 1.

Table 1.

The classification of the adverse drug reactions (ADRs)

Number	Classification	Included ADRs
1	Gastrointestinal symptoms	Gastrointestinal reaction, abdominal distention, abdominal pain, gastric distention, diarrhoea, constipation, stomach ache, upper abdominal discomfort, belching/acid reflux, nausea, vomit, reduced appetite, inappetence, anorexia, weight loss, stool times↑, stool scarce
2	Liver disease symptoms	Abnormal liver function, hepatalgia, liver damage, alanine aminotransferase↑, aspartate aminotransferase↑
3	Muscle symptoms	Muscle pain, muscle soreness, creatine kinase↑, lactate dehydrogenase↑
4	Neurological symptoms	Dizziness/headache, fatigue, lethargy
5	Skin symptoms	Skin rash, itchy skin, allergy, facial flushing, skin pain, skin fever
6	Cardiovascular symptoms	Myocardial infarction
7	Urinary symptoms	Urea ammonia↑, urine ketone+, proteinuria
8	Others	Undefined

Patient and public involvement

This study using the Anti-hyperlipidemic Drug Database in China was designed and conducted without patient and public involvement.

Statistics

The non-statin anti-hyperlipidaemic drugs were divided into three groups: (1) fibrates (fenofibrate, gemfibrozil, bezafibrate, etofylline clofibrate), (2) nicotinic acid and derivatives (niacin, acipimox); (3) others (probucol, cholestyramine). The ADR data of the eight drugs were analysed, including the number of clinical trials, the number of single-arm and controlled trials (with the control or experimental arms), the total number of patients, the number of patients in single-arm trials or the control or experimental arms of controlled trials and the number and percentage of patients with all adverse reactions in response to statin therapy. The patients treated with non-statin antihyperlipidaemic drugs (single-arm trials and the experimental groups of controlled trials) were used for ADR calculation, and the percentage of each ADR was also calculated. All ADR terms reported between trials were assumed to be identical in the Anti-hyperlipidemic Drug Database, and we supposed no duplicates of the patients who reported ADRs in the database. The ‘% of n’ was calculated by analysing the ADR cases in the total patients treated with non-statin antihyperlipidaemic drugs by using Microsoft Excel software. The free software environment R (R Core Team, 2022) was used to conduct a one-way analysis of variance with the Tukey honest significant differences (HSD) post hoc test to compare the differences in ADRs between fibrates.

Results

In the ADR analysis of non-statin antihyperlipidaemic drugs in China from 1989 to 2019, 6667 patients from 117 clinical trials, including single-arm trials and an experimental group in the controlled trials, were treated via fibrates (fenofibrate, gemfibrozil, bezafibrate, etofylline clofibrate), nicotinic acid and derivatives (niacin, acipimox) and others (probucol, cholestyramine), and the number of included clinical trials of different drugs is shown in table 2. In total, 3732 patients from 68 clinical trials were treated with fenofibrate, 760 patients from 16 clinical trials were treated with gemfibrozil, 531 patients from 10 clinical trials were treated with bezafibrate, 209 patients from four clinical trials were treated with etofylline clofibrate, 165 patients from one clinical trial were treated with niacin, 642 patients from eight clinical trials were treated with acipimox, 561 patients from nine clinical trials were treated with probucol and 67 patients from one clinical trial were treated with cholestyramine (table 2).

Table 2.

Analysis of the included clinical trials with non-statin anti-hyperlipidaemic drugs in China from 1989 to 2019

The clinical trials with non-statin anti-hyperlipidaemic drugs	Single-arm trials		Controlled trials
			Number of trials	Control group	Experimental group	The patients treated with non-statin lipid-lowering drugs
	Number of trials	Number of patients		Number of patients	Number of patients	Number of trials	Number of patients
Fibrates
Fenofibrate	39	2458	29	1134	1274	68	3732
Gemfibrozil	10	410	6	259	350	16	760
Bezafibrate	2	50	8	449	481	10	531
Etofylline clofibrate	2	90	2	105	119	4	209
Subtotal	53	3008	45	1947	2224	98	5232
Nicotinic acid and derivatives
Niacin	1	165	0	0	0	1	165
Acipimox	7	612	1	60	30	8	642
Subtotal	8	777	1	60	30	9	807
Others
Probucol	6	466	3	88	95	9	561
Cholestyramine	0	0	1	38	67	1	67
Subtotal	6	466	4	126	162	10	628
Total	67	4251	50	2133	2416	117	6667

To simplify the ADRs of non-statin antihyperlipidaemic drugs, we classified all ADRs reported in the selected trials into eight groups. Among the reported cases with ADRs (844 of 6667; 12.659%), gastrointestinal symptoms were the most common (465 of 6667; 6.975%), followed by hepatic disease (164 of 6667; 2.460%), skin (153 of 6667; 2.295%), neurological (30 of 6667; 0.450%), urinary (15 of 6667; 0.225%), muscle (14 of 6667; 0.210%), others (2 of 6667; 0.030%) and cardiovascular (1 of 6667; 0.015%), as shown in online supplemental table 1.

To provide additional clarity regarding the variation in ADRs among non-statin antihyperlipidaemic drugs, we collated all ADRs reported in the clinical trials and categorised them by different groups, as presented in online supplemental table 1. The ADRs of nicotinic acid and derivatives (214 of 807; 26.518%) were the most common, followed by fibrates (598 of 5232; 11.430%) and others (32 of 628; 5.096%) in online supplemental table 1. For different drugs, the ADRs of acipimox (179 of 642; 27.882%) were the most common, followed by niacin (35 of 165; 21.212%), gemfibrozil (151 of 760; 19.868%), cholestyramine (11 of 67; 16.418%), fenofibrate (386 of 3732; 10.343%), bezafibrate (50 of 531; 9.416%), etofylline clofibrate (11 of 209; 5.263%) and probucol (21 of 561; 3.743%), as shown in online supplemental table 2.

For fibrates, the rates of gastrointestinal symptoms were found to differ between each drug (online supplemental table 2). Gemfibrozil (100 of 760; 13.158%) caused gastrointestinal side effects most commonly, followed by fenofibrate (250 of 3732; 6.699%), bezafibrate (29 of 531; 5.461%) and etofylline clofibrate (6 of 209; 2.871%) in online supplemental table 2. In the ADRs related to liver disease symptoms, there was a slight difference in values between each fibrate. Gemfibrozil (38 of 760; 5.000%) was the most commonly followed by fenofibrate (98 of 3732; 2.626%), bezafibrate (10 of 531; 1.883%) and etofylline clofibrate (0 of 209; 0.000%), as shown in online supplemental table 2. Table 3 displays the ADRs comparison between fibrates. Our findings revealed that there were differences in that: (1) the gastrointestinal, liver, neurological and total ADRs between gemfibrozil and fenofibrate; (2) the neurological ADRs between gemfibrozil and bezafibrate; (3) the urinary ADRs between gemfibrozil and bezafibrate (table 3).

Table 3.

Comparison of the adverse drug reactions (ADRs) between fibrates

Types of ADRs	ANOVA (p-value)	Tukey HSD post-hoc test (p-value)
		Clo-Bez	Fen-Bez	Gem-Bez	Fen-Clo	Gem-Clo	Gem-Fen
GI	0.0312*	0.9684215	0.9994088	0.1707789	0.9314059	0.2051420	0.0264983*
Liver	0.015*	0.8781187	0.9512123	0.0858456	0.6235084	0.0642730	0.0293338*
Muscle	0.255	N/A	N/A	N/A	N/A	N/A	N/A
Neurological	0.000199***	1.0000000	0.9999931	0.0105393*	0.9999992	0.1077446	0.0000823***
Skin	0.476	N/A	N/A	N/A	N/A	N/A	N/A
CVD	0.934	N/A	N/A	N/A	N/A	N/A	N/A
Urinary	0.0476*	0.3943866	0.0363779*	0.4578223	0.9999989	0.9283683	0.6268774
Others	0.934	N/A	N/A	N/A	N/A	N/A	N/A
Total	0.00685**	0.7550618	0.9684340	0.2132317	0.8354962	0.0735552	0.0055756**

*p<0.05; **p<0.01; ***p<0.001.

ADR, adverse drug reaction; ANOVA, analysis of variance; Bez, bezafibrate; Clo, etofylline clofibrate; CVD, cardiovascular disease; Fen, fenofibrate; Gem, gemfibrozil; GI, gastrointestinal; HSD, honest significant differences; N/A, not applicable.

For nicotinic acid and derivatives (online supplemental table 2), skin symptoms were the most common ADRs of both acipimox (121 of 642; 18.847%) and niacin (18 of 165; 10.909%). The ADRs related to gastrointestinal and liver disease symptoms revealed marked differences between nicotinic acid and derivatives. The gastrointestinal ADRs of acipimox (48 of 642; 7.477%) were higher than niacin (4 of 165; 2.424%), in contrast to the hepatic ADRs of niacin (13 of 165; 7.879%) were higher than acipimox (3 of 642; 0.467%) in online supplemental table 2.

For other non-statin antihyperlipidaemic drugs (online supplemental table 2), gastrointestinal symptoms were the most common ADRs of both cholestyramine (11 of 67; 16.418%) and probucol (17 of 561; 3.030%). Few other types of ADRs were reported.

Discussion

Among the TLDs, our results show that gastrointestinal and liver disease symptoms were the major ADRs of fibrates, including fenofibrate, gemfibrozil, bezafibrate and etofylline clofibrate, in Chinese clinical trials between 1989 and 2019. Skin symptoms were the most common ADRs of nicotinic acid and derivatives, and there were some concerns about the hepatic ADRs of niacin. For the LLDs, gastrointestinal symptoms were the most common ADRs of cholestyramine as well as the antioxidant probucol.

The ADRs of fibrates in China

The Cochrane systematic review shows that fibrates are not associated with an increased risk for adverse effects, based on very low‐quality evidence.¹⁰ The reported ADRs of fibrates include increases in serum creatinine and transaminase levels, the occurrence of pancreatitis and venous thrombotic events but none of the occurrence of myositis/rhabdomyolysis.¹⁰ Our finding revealed that the ADRs of fibrates (11.430%; n=5232) were similar to that of non-statin antihyperlipidaemic drugs (12.659%; n=6667) but slightly higher than the ADRs of statins (7.460%; n=37 828) in Chinese population.⁷ It was supposed that there was a lower risk of statin ADRs in the primary prevention of hyperlipidaemia in China, despite the incidence rate of statin and non-statin antihyperlipidaemia needs further investigation.

Fenofibrate is a widely used fibric acid derivative with enhanced potency and specificity of action on lipids.¹¹ European experiences with fenofibrate show that adverse effects are relatively low in frequency (6%; n=7145) in the European clinical trials and manifested as gastrointestinal effects, muscle pain, skin problems and sweating or dizziness.¹¹ US experience in clinical trials with fenofibrate mirrors the European experience, and three types of adverse effects occur more commonly in fenofibrate patients versus placebo: skin reactions, neurologic effects and musculoskeletal reactions.¹¹ Japan’s experience in the clinical trial with fenofibrate shows that the incidences of ADRs are 14.1% (n=85) in the fenofibrate 100 mg/day group and 26.4% (n=140) in the fenofibrate 200 mg/day group.¹² In this study, China’s experience in clinical trials with fenofibrate showed that gastrointestinal symptoms (6.699%) were the most common ADRs, followed by hepatic (2.626%) and neurological symptoms (0.429%). Taken together, the usage of fenofibrate led to different ADRs in different countries/regions, suggesting that the clinical choice of fibrates may be different based on the patient population. The dosage of fibrates should be further clarification of ADR occurrence in different populations.

For gemfibrozil, the small-scale New Zealand data reveal that gemfibrozil treatment is stopped more frequently (12.2%; n=200) due to a possible adverse reaction compared with bezafibrate (5.4%) and simvastatin (4.8%), and approximately half of the adverse reactions leading to the withdrawal of gemfibrozil are gastrointestinal.¹³ In this study, the ADRs of gemfibrozil (19.868%; n=760) were higher than that of fibrates (11.430%; n=5232) in China population, especially the ADRs of gastrointestinal symptoms (13.158%; n=760). These findings suppose that gemfibrozil may not be the first-line fibrates because of the higher risk of ADRs.

For bezafibrate, the Japan clinical study shows that the bezafibrate ADRs are experienced by 5.1% (n=5978), and the most common ADRs are increased blood creatine phosphokinase (0.8%), blood creatinine (0.8%), blood urea (0.5%), renal impairment (0.3%) and aspartate aminotransferase (0.3%).¹⁴ The severe ADRs are experienced by 0.28% (n=5978), mainly including stroke (0.03%), renal impairment 0.03%), elevated blood urea nitrogen (0.03%), abnormal hepatic function (0.02%), pneumonia 0.02%), rhabdomyolysis (0.02%) and hypoglycaemia (0.02%).¹⁴ In the Chinese population, the overall ADR ratio was 9.416% (n=531), which was similar to that of the Japanese population. However, we found that the gastrointestinal ADRs of bezafibrate (5.461%) in China were much higher than that in Japan. There were no severe ADRs reported in the Chinese population, especially muscle symptoms. We suggested that fibrate therapy may result in different ADR profiles in different countries/regions not only because of race but also eating habits and lifestyles. In addition, because of the lower sample number of etofylline clofibrate (n=209), we supposed that bezafibrate might be with a lower risk of ADR occurrence than the other included fibrates in this study and potentially beneficial for the primary prevention of hyperlipidaemia in China.

In our study, we observed that the incidence of hepatic ADRs associated with fibrates was 2.791% (n=5232), whereas the incidence of fenofibrate was 2.626% (n=3732), which was lower compared with gemfibrozil (5.000%; n=760) in China. On the other hand, a clinical trial conducted in Japan revealed that the occurrences of adverse events related to the liver were 15.3% (n=85) in patients who received a daily dose of 100 mg fenofibrate and 24.3% (n=140) in those who received a daily dose of 200 mg fenofibrate.¹² Moreover, aspartate aminotransferase, alanine aminotransferase and γ-glutamyl transferase levels showed significant increases in the fenofibrate groups compared with the placebo group.¹² Furthermore, our results demonstrated that, in China, the incidence of urinary ADRs related to fibrates was 0.229% (n=5232). On the other hand, a study conducted in Japan reported that the use of fenofibrate led to a marked increase in serum creatinine and a decrease in estimated glomerular filtration rate, especially in the 200 mg/day group, compared with the placebo group. This suggests that liver and kidney function should be monitored when using fibrates in Chinese patients, but not to a great extent. In addition, a previous study suggested that myopathy, characterised by muscle pain, may develop after a few days or prolonged use of all fibrates.¹⁵ Our study showed that the occurrence of muscle-related ADRs due to fibrates was not common (0.268%; n=5232) among Chinese individuals. Additionally, a clinical trial conducted in Japan supports our findings, as it revealed that the incidence of rhabdomyolysis/myopathy was low and comparable to placebo in all fenofibrate dose groups.¹² These results provide evidence for the safe use of fibrates in the Asian population regarding myotoxicity.

The ADRs of nicotinic acid and derivatives in China

The safety of niacin in the US Food and Drug Administration adverse event reporting database indicates that niacin increased the number of side effects, specifically including flushing (moderate‐quality evidence), pruritus (moderate‐quality evidence), rash (moderate‐quality evidence), headache (moderate‐quality evidence) and gastrointestinal symptoms (moderate‐quality evidence).¹⁶ Dermal and gastrointestinal reactions are the most common ADRs of niacin, and truncal and facial flushing are reported in 90% to 100% of treated patients in large clinical trials.¹⁷ Niacin-induced vasodilation may be caused by interacting with a G-protein-coupled receptor 109A on Langerhans’ cells in the epidermis, resulting in the activation of prostaglandins.¹⁸ In this study, we found that the ratio of skin symptoms of nicotinic acid and derivatives was 17.224% (n=807) in China, and 10.909% (n=165) and 18.847% (n=642) for niacin and acipimox respectfully. We supposed that the proportion of skin-related ADRs caused by niacin may be relatively low in the Chinese population, although that was already the highest among all non-statin hypolipidemic drugs in this study.

Sporadic reports of hepatotoxicity with niacin remain arising.¹⁹ A Caucasian woman in her early 70s developed acute liver failure after her home dose of immediate‐release niacin was replaced with an extended‐release formulation during an inpatient hospital stay.²⁰ The US case report indicates that a male teenager was transferred to the paediatric intensive care unit for evaluation of acute liver failure.²¹ There is a direct toxic effect similar to acute hepatic necrosis in niacin therapy, and the severity of toxicity seems to be dose dependent and more commonly involves sustained release preparations.¹⁸ Our findings also revealed that the hepatic ADRs of niacin (7.879%; n=165) were higher than all the other antihyperlipidaemic drugs in these studies, especially acipimox (0.467%; n=642). We suggest that long-term usage of niacin in the treatment of hyperlipidaemia may need to monitor hepatic function in China. Besides hepatoxicity, significant elevations of serum glucose and uric acid are occasionally seen with nicotinic acid therapy.¹⁷ In the Chinese population, we do not find reports of diabetes—or gout-related ADRs.

For nicotinic acid derivatives, the gastrointestinal ADRs (7.477%; n=642) were higher than other classifications of ADRs in acipimox therapy in the Chinese population. Moreover, an Italy clinical trial provides that heartburn, epigastric pain and flushing, in this order, are the most common complaints in patients on active treatment with acipimox, and overall ADR cases are 26 out of the follow-up 102 subjects.²² The Austria clinical trial indicates that treatment with acipimox (8 out of 82) leads to moderate adverse events of transient characters, such as skin reactions and gastric disturbances.²³ Taken together, the gastrointestinal symptoms of acipimox in China need more attention in clinical use.

The ADRs of other non-antihyperlipidaemic drugs in China

Gastrointestinal ADRs are most frequently observed in adult patients treated with cholestyramine, including constipation, abdominal pain and unpalatability, and large doses of cholestyramine may be associated with malabsorption of fat-soluble vitamins.¹⁷ We also found one clinical trial of cholestyramine in China with the highest gastrointestinal ADRs (16.418%; n=67) than all the other non-statin antihyperlipidaemic drugs in this study. Based on the pharmacological mechanism of cholestyramine, we believed that such a result is reasonable in spite of the sample number was small.

Adverse effects reported with probucol administration clinically are few, and the most common effects are gastrointestinal in nature and include transient diarrhoea, flatulence, nausea and abdominal pain.¹⁷ Our findings also revealed that gastrointestinal ADRs of probucol (3.030%; n=561) were the most common of all the other types of ADRs in China. However, prolonging the corrected QT (QTc) interval is the primary concern with the ADR of probucol, possibly resulting from its accumulation in high concentrations in adipose tissue and its effects on reducing HDLs.¹⁷ There were at least two clinical trials that reported the ADRs of tachyarrhythmias and the magnitude of QTc prolongation associated with probucol therapy, respectfully from the US and Korea (with a part of patients from China).^{24 25} Even if our findings showed that no cardiovascular ADRs of probucol (0%; n=561) in China population, we suggested that changes in QTc prolongation in probucol therapy needed further investigation, especially in the Asian population.

Limitations

As stated in our prior research, the constraints of the Anti-hyperlipidemic Drug Database were present.³ Stratification was very difficult to perform for sensitivity analysis. Furthermore, we found that the sample sizes of non-statin antihyperlipidaemic drugs were different. There was only one clinical trial of niacin and cholestyramine included in this study, spite these drugs were used rarely. Although we did not provide the dosage and duration of the included trials, our findings could offer more clinical evidence regarding the safety of non-statin antihyperlipidaemic drugs for the treatment of dyslipidaemia. Furthermore, our results suggest that there could be variations in the ADRs among fibrates. Nonetheless, since the number of clinical trials varies across different fibrates, further evidence is required to validate our findings.

Conclusions

Based on the findings of this study, we concluded that skin symptoms were the most common ADRs of nicotinic acid and derivatives, while gastrointestinal symptoms were the most common ADRs during the treatment of fibrates, acipimox, cholestyramine and probucol in the Chinese population from 1989 to 2019. These results highlight the importance of conducting further investigation on ADRs related to the long-term use of non-statin antihyperlipidaemic drugs in China.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

Not applicable.