Effect of pemafibrate on LDL‐cholesterol properties in type 2 diabetes and hypertriglyceridemia by lipid management status: PARM‐T2D post hoc analysis

ABSTRACT

Although pemafibrate attracts attention for reducing LDL‐C in patients with high baseline LDL‐C, its comprehensive effects remain unclear in patients with type 2 diabetes. This post hoc analysis of the PARM‐T2D study investigated the effects of pemafibrate on LDL‐C concentration and particle size in 156 subjects with type 2 diabetes and hypertriglyceridemia, categorized by baseline LDL‐C levels. While triglycerides and remnant lipoprotein cholesterol significantly improved across all groups over 52 weeks, the impact on LDL‐C was multifaceted and dependent on baseline status. In participants with low baseline LDL‐C, concentrations increased without a rise in ApoB, suggesting an improvement in particle size. Conversely, in those with high baseline LDL‐C, both LDL‐C and ApoB decreased, reflecting favorable modifications according to baseline LDL‐C control status. These findings suggest that pemafibrate exerts beneficial, status‐dependent effects on LDL‐C properties, potentially reducing cardiovascular risk by improving LDL particle quality and quantity in high‐risk patients with diabetes.

INTRODUCTION

A high serum low‐density lipoprotein‐cholesterol (LDL‐C) concentration has been identified as a key risk factor for atherosclerotic cardiovascular disease (ASCVD) in many epidemiological studies worldwide. ¹ However, despite aggressive lipid‐lowering strategies that target serum LDL‐C levels using statins, there is still a high incidence of ASCVD. ² Smaller LDL particle sizes are associated with a three‐fold higher incidence of ASCVD. ³ Therefore, in addition to the total LDL‐C concentration, the LDL particle size and the small‐dense LDL (sdLDL) concentration also needs to be well managed.

Pemafibrate is a potent and selective synthetic agonist at the PPARα nuclear receptor that reduces plasma triglyceride concentrations by increasing the lipoprotein lipase activity and reducing hepatic VLDL‐cholesterol, apolipoprotein B (ApoB) levels and triglyceride production. ⁴ In the PROMINENT trial of the administration of pemafibrate to individuals with type 2 diabetes (T2D) whose LDL‐C was fairly well controlled, pemafibrate increased their serum LDL‐C concentrations versus placebo, ⁵ possibly because of an increase in the large LDL‐C concentration, but not that of small LDL‐C, as shown in a phase 2 trial. ⁶ On the basis of this, we conducted a post hoc analysis of data from a phase 4 trial performed in individuals with T2D and a wide range of LDL‐C management statuses to clarify the relationship between the total serum LDL‐C concentration and particle size after pemafibrate therapy commenced.

MATERIALS AND METHODS

We performed a post hoc analysis of data from a multicenter prospective observational study (the PARM‐T2D study), in which individuals with T2D and hypertriglyceridemia aged ≥20 years commenced pemafibrate treatment at 0.2–0.4 mg/day were followed for 52 weeks. ⁷ The participants generally continued their concomitant medications for dyslipidemia, other than fibrates. For this post hoc analysis, data from participants who newly initiated pemafibrate without prior fibrate therapy and had no missing relevant lipid data were analyzed. These selected participants were allocated to four subgroups on the basis of their baseline LDL‐C status (Group 1, LDL‐C <80 mg/dL; Group 2, 80 ≤LDL‐C <100 mg/dL; Group 3, 100 ≤LDL‐C <120 mg/dL; Group 4, ≥120 mg/dL), and the changes in lipid management‐related parameters were compared over 52 weeks.

The study protocol was approved by the Institutional Review Board of Hokkaido University Hospital Clinical Research and Medical Innovation Center (approval number 018‐0440) and was registered with the University Hospital Medical Information Network (registration number UMIN000037385). The study was conducted in accordance with the principles of the Declaration of Helsinki and its amendments. All the subjects provided their written informed consent before participation.

Normally distributed continuous data are expressed as mean ± standard deviation, nonnormally distributed continuous data as median (interquartile range), and categorical data as number (%). Comparisons between multiple groups were made using one‐way repeated measures anova, followed by Tukey's post hoc test; the Friedman test followed by Dunn's post hoc test; or the chi‐squared test. Changes from baseline are expressed as mean or median (95% confidence intervals), and group comparisons were performed using the Kruskal–Wallis test, followed by Dunn's post hoc analysis. Relationships between variables were assessed using Pearson's correlation analysis. Because the effects of pemafibrate on metabolic parameters may be affected by concomitant statin use, analysis of covariance (ancova) was also performed to adjust for this confounding factor. P < 0.05 was considered to represent statistical significance. Data were analyzed using GraphPad Prism v.8.4.2 (GraphPad Software, San Diego, CA, USA) or JMP v.18.2.2 (SAS Institute, Inc., Cary, NC, USA).

RESULTS

Participants, who commenced pemafibrate, were not currently treated with any fibrate, and had complete relevant data were studied (n = 156) (Figure 1). When the characteristics of the participants were compared according to their baseline LDL‐C concentration (Table 1), no significant differences in the baseline triglyceride concentrations of the groups were identified. The participants in Group 3 had higher BMIs, and Group 4 consisted of younger participants and the incidence of concomitant statin use in this group was notably lower. As shown in Figure 2a, there was no significant total change in LDL‐C concentration during the study. However, as expected, the triglyceride and RLP‐C concentrations significantly improved (Figure 2b,c). In the category‐specific analysis, the LDL‐C concentration increased in Group 1, without an increase in ApoB, resulting in an increase in the LDL‐C/ApoB ratio, which suggests an improvement in LDL‐C particle size. This trend was more pronounced in subgroups with higher triglyceride concentrations, possibly because remnant‐rich lipoprotein was effectively metabolized as a result of pemafibrate action (Figure 2d). In contrast, there were decreases in ApoB, reflecting the number of LDL‐C particles, in Groups 2–4. Notably, in Group 4, the LDL‐C concentration also decreased, reflecting favorable modifications according to baseline LDL‐C control status (Figure 2a,e,f). Given the presumed close relationship between the changes in LDL‐C and triglyceride concentrations, we assessed this using correlation analysis. However, the change in LDL‐C concentration did not significantly correlate with the changes in triglyceride or RLP‐C concentrations (Figure S1A,B). Finally, the effect of statin use on the change in LDL‐C concentration was evaluated, but no significant effect was identified (Figure S2).

Figure 1.

Flow diagram for the selection of participants in the present post hoc analysis. From the pemafibrate cohort of the PARM‐T2D study (n = 275), 156 participants were included in the present post hoc analysis, after excluding those with prior conventional fibrate use, those switching from other fibrates, and those with missing data. These study participants were allocated to four subgroups on the basis of their baseline LDL‐C concentrations. Groups 1, 2, 3, and 4 had baseline LDL‐C concentrations of <80 mg/dL, 80 ≤LDL‐C <100 mg/dL, 100 ≤LDL‐C <120 mg/dL, and ≥120 mg/dL, respectively.

Table 1.

Characteristics of the study sample

Variable	Entire cohort n = 156	Group 1 n = 37	Group 2 n = 42	Group 3 n = 40	Group 4 n = 37	Overall P‐value
Age (years)	61.1 ± 13.6	64.5 ± 11.8	65.8 ± 13.5	59.2 ± 12.6	54.4 ± 13.6**†††	<0.001
Sex (male, %)	97 (62.2%)	29 (78.4)	26 (61.9)	24 (60.0)	18 (48.6)	0.069
BMI (kg/m2)	27.7 ± 4.5	25.9 ± 4.1	27.8 ± 4.4	29.1 ± 4.5**	27.9 ± 4.5	0.017
HbA1c (%)	7.22 ± 0.91	7.42 ± 1.00	7.03 ± 0.73	7.13 ± 0.82	7.32 ± 1.07	0.213
TG (mg/dL)	199 (156, 275)	194 (169, 262)	181 (143, 235)	225 (173, 324)	220 (147, 294)	0.065§
LDL‐C (mg/dL)	99.7 ± 27.5	65.7 ± 12.4	89.3 ± 5.9***	108.6 ± 5.6***†††	135.9 ± 16.4***†††‡‡‡	<0.001
HDL‐C (mg/dL)	51.8 ± 12.0	49.1 ± 13.9	52.4 ± 11.3	52.3 ± 10.1	53.4 ± 12.6	0.439
Apo A1 (mg/dL)	141.7 ± 22.2	140.0 ± 29.2	139.8 ± 19.2	144.5 ± 20.1	142.4 ± 19.8	0.758
Apo B (mg/dL)	90.7 ± 17.6	73.5 ± 11.8	82.9 ± 8.9**	96.7 ± 8.7***†††	110.0 ± 15.0***†††‡‡‡	<0.001
Apo E (mg/dL)	3.84 ± 1.81	3.68 ± 1.97	3.35 ± 1.37	4.04 ± 1.72	4.33 ± 2.07	0.086
RLP‐C (mg/dL)	7.1 (4.4, 10.6)	7.1 (4.4, 9.6)	5.2 (3.5, 7.5)	8.1 (4.8, 12.5)†	8.7 (5.6, 12.8)††	0.002§
LDL‐C/Apo B	1.09 ± 0.18	0.91 ± 0.17	1.09 ± 0.11***	1.13 ± 0.12***	1.25 ± 0.15***†††‡‡	<0.001
Statin use (%)	110 (70.5)	31 (83.1)	36 (85.7)	28 (70.0)	15 (40.5)	<0.001

Values are expressed as mean ± SD or median (interquartile range). Group 1: baseline LDL‐C <80 mg/dL, Group 2: 80 mg/dL ≤baseline LDL‐C <100 mg/dL, Group 3: 100 mg/dL ≤baseline LDL‐C <120 mg/dL, Group 4: baseline LDL‐C ≥120 mg/dL. Groups was compared by one‐way repeated‐measures anova, followed by Tukey's post hoc test, or the Friedman test following by Dunn's post‐hoc test. §The Friedman test was used for TG. *P < 0.05, **P < 0.01, and ***P < 0.001 vs. Group 1. †P < 0.05, ††P < 0.01, and †††P < 0.001 vs. Group 2. ‡‡P < 0.01 and ‡‡‡P < 0.001 vs. Group 3. Apo, apolipoprotein; BMI, body mass index; HbA1c, glycated hemoglobin; HDL‐C, high‐density lipoprotein‐cholesterol; LDL‐C, low‐density lipoprotein‐cholesterol; RLP‐C, remnant lipoprotein‐cholesterol; TG, triglyceride.

Figure 2.

Changes in key parameters involved in lipid metabolism. (a–c and e–f) Mean changes from baseline in the (a) LDL‐C, (b) triglyceride, (c) RLP‐C, (e) ApoB, and (f) LDL‐C/ApoB in the subgroups stratified by the baseline LDL‐C concentration Groups 1, 2, 3, and 4 had baseline LDL‐C concentrations of <80 mg/dL, 80 ≤LDL‐C <100 mg/dL, 100 ≤LDL‐C <120 mg/dL, and ≥120 mg/dL, respectively. The triglyceride and RLP‐C data were logarithmically transformed before analysis. (d) Change in LDL‐C concentration from baseline, stratified by the baseline LDL‐C and triglyceride concentrations. Bars are presented as mean ± SD. *P < 0.05, **P < 0.01, *** P < 0.001 vs. baseline (Student's t‐test). One‐way anova, followed by post hoc analysis, was used for intergroup analyses. ApoB, apolipoprotein B; LDL‐C, low‐density lipoprotein‐cholesterol; RLP‐C, remnant lipoprotein‐cholesterol; TG, triglyceride.

DISCUSSION

Yamashita et al. reported that pemafibrate reduced LDL‐C levels in individuals with hypertriglyceridemia by inhibiting cholesterol synthesis in the liver and cholesterol absorption from the intestinal tract, and that this lowering effect was more pronounced in those with high baseline LDL‐C concentrations. ⁸ We have obtained similar results in a high‐risk T2D cohort with more marked abnormalities in lipid metabolism. Similarly, Tanigawa et al. showed that pemafibrate had an LDL‐C‐lowering effect in individuals with MASLD, ⁹ but also that it altered the LDL particle profile, as in the present participants with T2D. Given that both the quantity and quality of LDL‐C are closely related to the pathogenesis of ASCVD, these effects of pemafibrate would be beneficial in individuals with T2D who are at high risk for ASCVD. It has been reported that adding pemafibrate to statin‐based therapy significantly reduces the sdLDL‐C concentration in subjects with T2D. ¹⁰ In this respect, a strength of the present study was that it enabled the estimation of LDL‐C particle size within the context of LDL‐C control status, revealing that pemafibrate improved LDL‐C quality particularly in participants with low baseline LDL‐C levels, while reducing LDL‐C quantity in those with high baseline levels.

The present study has also yielded interesting results that differ from those of the previously reported PROMINENT trial. Specifically, in the low‐LDL‐C subgroup with relatively mild hypertriglyceridemia, pemafibrate administration led to an increase in LDL‐C level without a concomitant rise in ApoB. Although the investigators in the PROMINENT trial attributed the simultaneous increases in the LDL‐C and ApoB levels to an increase in the total number of ApoB‐containing particles secondary to greater heavy remnant conversion ⁸ —our findings suggest that the pemafibrate‐induced increase in LDL‐C concentration does not necessarily imply a greater burden of ApoB‐containing particles but may instead reflect cholesterol enrichment within the small baseline remnant pool particles.

This study has several limitations, including its post hoc design with potential statistical power constraints, the lack of sdLDL‐C measurements, and the evaluation being confined to the pemafibrate‐treated population.

In summary, the present subanalysis has revealed that pemafibrate has multifaceted effects on LDL‐C properties that depend on the LDL‐C management status of individuals with T2D and inadequate triglyceride management. Furthermore, pemafibrate may have a beneficial effect on lipid status by altering LDL particle quality.

FUNDING INFORMATION

The publication fee (Article Processing Charge) for this article was funded by Kowa Company Limited.

DISCLOSURE

Hiroshi Nomoto and Akinobu Nakamura have received honoraria for lectures and research funding from Kowa Company, Ltd. The other authors declare no conflict of interest. The design of the study, collection, analysis, and interpretation of all data and drafting of the manuscript were conducted solely by the authors.

Approval and registration of the research protocol: The protocol for this research project has been approved by a suitably constituted Ethics Committee of the institution and it conforms to the provisions of the Declaration of Helsinki. Institutional Review Board of Hokkaido University Hospital Clinical Research and Medical Innovation Center (approval number 018‐0440) and was registered with the University Hospital Medical Information Network (registration number UMIN000037385).

Informed consent: All the subjects provided their written informed consent before participation.

Supporting information

Figure S1. Changes in lipid parameters. (A) Relationship between the changes in the LDL‐C and log triglyceride concentrations. (B) Relationship between the changes in the LDL‐C and log RLP‐C concentrations. The subjects in Groups 1, 2, 3 and 4 were classified according to their baseline LDL‐C concentration: <80 mg/dL, 80 ≤LDL‐C <100 mg/dL, 100 ≤LDL‐C <120 mg/dL, and ≥120 mg/dL, respectively. Correlation analysis was performed using Pearson's method. LDL‐C, low‐density lipoprotein‐cholesterol; RLP‐C, remnant lipoprotein‐cholesterol.

Figure S2. Changes in the LDL‐C concentration from baseline for statin users and nonstatin users. Bars represent least square mean ± standard error. The differences between the two groups are shown as the estimated treatment difference (ETD) (95% confidence interval). The data were adjusted using ancova (the covariates were age, sex, pemafibrate dose, glycated hemoglobin concentration, body mass index, and LDL‐C concentration; and all except the pemafibrate dose were baseline values). LDL‐C, low‐density lipoprotein‐cholesterol.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.