1. IntroductionBy using lipoprotein electrophoresis (LPEP) and ultracentrifugation, Fredrickson, Levy, and Lees (FLL) separated clinical dyslipidemia into five phenotypes (types I–V) of hyperlipoproteinemia (HLP) [1], each phenotype represented a characteristic abnormality of specific lipoproteins. Although the World Health Organization (WHO) adopted this classification [2], the current lipid guidelines [3,4] have been simplified to focus on the elevation of low-density lipoprotein cholesterol (LDL-C) in clinical practice. Recently, Sniderman et al. validated a diagnostic apolipoprotein B (apoB) algorithm derived from patients with genetically verified dyslipidemia to determine the FLL phenotypes of HLP by evaluating the conventional lipid profiles together with the plasma apoB levels [5]. The apoB algorithm can provide valuable information to indicate which lipoprotein abnormalities are involved: chylomicrons (types I and V), very low-density lipoproteins (VLDLs) (types IIb, IV, and V), low-density lipoproteins (LDLs) (types IIa, IIb), or remnant lipoproteins (type III) [6,7,8]. According to the apoB algorithm, type IIa (HLP2a) and IIb (HLP2b) are identified by apoB ≥ 120 mg/dL together with variable plasma TG levels, while type I (HLP1), III (HLP3), IV (HLP4), and V (HLP5) are identified by apoB 5]. The use of the apoB algorithm may assist in assessing the possible pathogenic mechanisms of each FLL phenotype and the residual cardiovascular risk of patients who have elevated TG-rich atherogenic lipoproteins but a normal plasma LDL-C level [5].HTG is associated with a decrease in high-density lipoprotein cholesterol (HDL-C) and an increase in atherogenic lipoproteins. Previous epidemiological and Mendelian randomization studies have demonstrated that the risk of atherosclerotic cardiovascular disease (ASCVD) remains high in patients with HTG despite statin therapy [9,10,11,12]. In addition, moderate to severely elevated plasma TG is associated with an increased risk of pancreatitis [3]. Based on the available evidence, the current American lipid guidelines have recommended the identification of HTG as a “risk-enhancing factor” of ASCVD [3], and the use of TG-lowering medications is also recommended in patients with moderate to severe HTG [3]. Of the HTG-associated phenotypes of HLPs, HLP4 and HLP5 are both referred to as polygenic lipid disorders complicated with various degrees of metabolic derangements accompanied by a marked increase in plasma VLDLs [1,2,5]. Specifically, the identification of HLP5 is separated from HLP4 by the presence of chylomicrons in the fasting blood, which can be determined by LPEP [1,5,6]. HLP4 is the most common FLL phenotype with HTG, whereas the prevalence of HLP5 in the adult population is uncommon [6]. Additionally, it is difficult to differentiate HLP5 from HLP4 by conventional blood lipid tests without the apoB algorithm and LPEP, leading clinicians to easily ignore HLP5 in clinical practice [6]. Until now, evidence showing the management of patients with HLP5 remains scarce. The present study aimed to report real-world data on the management of patients with HLP5 and persistent HTG. 4. Discussion

The present study showed that the demographic characteristics of patients with HLP5 were quite different from those of patients with HLP4. Despite an aggressive use of statin and TG-lowering medications followed by a marked reduction in plasma TC, TG, and non-HDL-C, patients with HLP5 still stayed in persistent HTG and had an incremental change in plasma LDL-C without a significant alteration in plasma apoB compared with patients with HLP4. To the best of our knowledge, the present study is the first evidence showing a real-word data on the management of patients with HLP5.

HLP5 is scarcely investigated, with a prevalence of 0.13-0.15% in the adult population, while HLP4 is the most common phenotype of HLP with a prevalence of 20.5-24.1% [6]. It is impossible to separate HLP5 from HLP4 based on conventional blood lipid tests, which usually results in a misdiagnosis of HLP5. An important distinguishing feature to identify HLP5 is the presence of chylomicrons together with an elevation in VLDLs in the fasting blood samples, whereas chylomicrons are absent in HLP4 [1,5]. The findings in the present study implicate that the use of LPEP in combination with the apoB algorithm may increase the diagnostic accuracy and assist in the differentiation between HLP4 and HLP5. Additionally, blood samples collected following at least 10 h of fasting could increase the accuracy of identification in plasma chylomicrons [1,5,6]. The pathogenic mechanisms of HLP5 remains not fully understood. It has been reported that patients with HLP5 may carry heterozygous mutations in LPL, GHIHBP1, APOC2, APOA5, or LMF1 genes, or burdens of single nucleotide polymorphisms in multiple TG-associated genes (i.e., DOCK7, ANGPTL3, GALNT2, KLHL8, AFF1, MAP3K1, ANKRD55, MLXIPL, NAT2, TRIB1, JMJD1C, FADS1, FADS2, FADS3, APOA1, APOC3, APOA4, CAPN3, FRMD5, CETP, SUGP1, PLTP, GCKR, CYP26A1, or CILP2) accompanied by secondary metabolic or environmental factors (i.e., obesity, alcohol drinking, or diabetes) compared with patients with HLP4, resulting in a further decreased activity in lipoprotein lipase (LPL), reduced clearance of chylomicrons and VLDLs, and further elevation of plasma TG levels [16,17].As plasma TG levels rise with a decrease in LPL activity, the circulating VLDL particles increase, accompanied by inefficient lipolysis and impaired clearance [18]. These VLDLs and their remnant particles enriched in apolipoprotein E and apolipoprotein C-III (apoC-III) may retain in the arterial wall, contributing to the initiation and progression of atherosclerosis, vascular inflammation, and prothrombotic effects [19]. HTG also results in the formation of small, dense LDLs that have a high atherogenicity because of their longer residence time in circulation and bioactive inflammatory lysolipids [20]. Moderate to severe HTG also leads to an increased risk of pancreatitis [15]. The excess of TG hydrolyzed by pancreatic lipase released in the vascular bed of the pancreas leads to high concentrations of free fatty acids, which overwhelm the binding capacity of albumin and self-aggregate to micellar structures with detergent properties, thereby exhibiting a toxic injury to pancreatic acinar cells. Additionally, the elevated levels of TG-rich lipoproteins increase the viscosity of blood and thus impair the blood flow in the pancreas, which enhances the toxicity of free fatty acids and triggers ischemic injury within the pancreas [21].In addition to difficult recognition and identification of HLP5 in clinical scenarios, the findings of the present study also highlight that the clinical treatment guidelines for patients with HLP5 still remain lacking. Weight loss is considered the most effective nonpharmacological intervention to lower plasma TG levels [15]. Although the response may be variable, a 5-10% reduction in body weight is associated with a 20% decrease in plasma TG [22]. Dietary recommendations should be individualized based on fasting plasma TG levels. In most circumstances, referral to a registered dietician nutritionist is necessary. Generally, the elimination of added sugar, restriction of fat to less than 15% of daily calories, and complete abstinence of alcohol are strongly advised for patients who have a plasma TG level ≥1000 mg/dL [15], as observed in patients with HLP5 in the present study. The initiation or intensification of statin therapy, together with the use of omega-3 fatty acids [15,23,24], is considered in high-risk patients with HTG who have histories of ASCVD, diabetes, or have a 10-year ASCVD risk >5%. According to the Reduction of Cardiovascular Events with Icosapent Ethyl Intervention Trial (REDUCE-IT) [24], which enrolled 8179 patients who mostly had received statin therapy, patients who were randomized to 4 g of omega-3 fatty acids (in the form of icosapent ethyl) had a 21.6% reduction in plasma TG levels from baseline and a reduced risk of composite cardiovascular events compared with patients who were randomized to the placebo. Fenofibrate is also effective to lower plasma TG; nevertheless, it did not exhibit definite clinical benefits in reducing adverse cardiovascular events. [15,25,26] In the Action to Control Cardiovascular Risk in Diabetes (ACCORD) lipid trial, fenofibrates, in addition to statin therapy, showed a 20-30% reduction in plasma TG among patients with type 2 DM, but yielded no significant additional benefit in the primary outcome of nonfatal MI, nonfatal stroke, or death from cardiovascular causes [25]. Similarly, the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study also showed that patients with type 2 DM randomized to fenofibrate therapy had a further 29% reduction in plasma TG, without a significant reduced risk in coronary events compared with those patients randomized to the placebo over a 5-year follow-up [26].The cost of omega-3 fatty acids was not reimbursed by our national health insurance, leading to a very limited use in the present study. Despite an aggressive use of statin and fenofibrate resulting in a significant reduction of plasma TG from baseline, patients with HLP5 in the present study still had an incremental change in plasma LDL-C and persistent HTG, which led to an increased risk of ASCVD and pancreatitis. It has been reported that the paradoxical increase in LDL-C following TG-lowering treatment is more frequently observed in patients who have lower LDL-C, higher TG, and higher non-HDL-C before treatment [27,28]. The use of fenofibrate may enhance the activities of LPL and hepatic lipase, thereby promoting the catabolism of VLDLs and lipolytic conversion of VLDLs to LDLs [28,29], which may explain the findings in the present study. In the present study, we also observed that patients with HLP5 had a greater reduction in VLDL cholesterol (VLDL-C) compared with patients with HLP4 (Figure S1), implicating an improvement in LPL and hepatic lipase activities following treatment. Nevertheless, the paradoxical increase in LDL-C following TG-lowering treatment may lead clinicians and patients to doubt not only the efficacy of treatment, but also the potential increased risk of ASCVD. In addition to an incremental change in plasma LDL-C, patients with HLP5 did not have a marked reduction in plasma apoB, despite having significant reductions in plasma TG and non-HDL-C. These findings support that TG-lowering medications in combination with the simultaneous use or intensification of statin therapy should be implemented in patients with HLP5, although they were younger and had less comorbidities deemed to have a lower risk of ASCVD compared with patients with HLP4. Moreover, a personized risk assessment, including family history of premature ASCVD, chronic inflammatory diseases, other biomarkers (i.e., lipoprotein (a) and high-sensitivity C-reactive protein), and the presence of subclinical atherosclerosis detected by carotid ultrasound scan and coronary artery calcium scanning, should be also considered to guide TG-lowering treatment and physician–patient discussion [15].Collectively, the development of novel pharmacological therapies is needed to treat patients with HLP5. Some medications targeting apoC-III, volanesorsen, and olezarsen have shown impressive TG-lowering effects. Volanesorsen exhibited a marked reduction in plasma apo C-III and TG (−84.2% and −76.5%, respectively) in patients with familial chylomicronemia syndrome [30]. However, a significant increase in plasma LDL-C (135%) following volanesorsen treatment was observed, potentially based on the modification of lipoprotein metabolism and increased lipolytic conversion of VLDLs to LDLs, as mentioned before [28,29]. Treatment with olezarsen was shown to provide a significant reduction in plasma TG (-56% to -60%) with variable LDL-C changes (-6% to 16%) in patients with ASCVD, familial chylomicron syndrome, and severe HTG [31]. Further clinical trials are necessary to explore the roles of pharmacological inhibition in apoC-III in patients with HLP5.

The present study was subjected to some limitations. First, the patient population was relatively small because of its single-center study design and the uncommon prevalence of patients with HLP5. In addition, the apoB algorithm was not previously validated in Taiwanese cohorts; nevertheless, the use of LPEP in the present study may assist the identification of each HLP phenotype. We also acknowledge that the detection of fasting chylomicronemia might be influenced by commercial analyzers; however, the methods used in the present study have been widely described. Second, potential changes in diet and body weight of patients during the study period, which might influence the plasma lipid levels of patients, were not surveyed. Furthermore, we did not routinely conduct LPEP in patients with HLP5 to observe the changes in the plasma chylomicrons following treatment. Third, the use of omega-3 fatty acids was very limited under the current regulation of Taiwan’s national health insurance, so the effects of omega-3 fatty acids were not shown in the present study. Finally, the present analysis did not observe and measure the patients’ clinical cardiovascular outcomes. In addition, the database of this present study did not measure and evaluate the presence of preclinical atherosclerosis (e.g., carotid intima-media thickness) and proinflammatory cytokines (e.g., high-sensitivity C-reactive protein). Future prospective and multicenter studies with a longer duration of observation are needed.