SERUM LIPOPROTEIN(A) CONCENTRATIONS AND APOLIPOPROTEIN(A) PHENOTYPES IN HEMODIALYSIS, CHRONIC AMBULATORY PERITONEAL DIALYSIS AND POST-TRANSPLANT PATIENTS

Abstract

Serum lipoprotein(a) [Lp(a)] concentrations and apolipoprotein(a) apo(a) phenotypes were determined in 81 hemodialysis (HD) patients, 37 chronic ambulatory peritoneal dialysis (CAPD) patients, 25 post-transplant patients and 99 healthy subjects as the reference group. The CAPD patients had significantly higher serum Lp(a) concentration than HD patients, but both had significantly increased Lp(a) levels as compared with the reference group and post-transplant patients. When all studied groups were divided into two subgroups with at least one low molecular weight (LMW) isoform and with only one high molecular weight (HMW) isoform, they presented a similar distribution. (Pearson's chi-squared = 2,78; df = 3; p = NS). The median serum Lp(a) levels were significantly increased with HMW class versus the reference group and post-transplant patients. In CAPD patients, the LMW phenotypes showed significantly increased median serum Lp(a) concentrations versus the reference group, but they were not statistically elevated in HD patients. In the post-transplant patients, LMW and HMW phenotypes did not differ as compared to the reference group. The elevated Lp(a) levels in HD and CAPD groups were explained by apo(a) type-specific, but not by differences in, isoform frequencies. We conclude that HD and CAPD patients had increased Lp(a) levels compared with the reference group, whereas elevated Lp(a) concentrations were observed mainly in patients with HMW apo(a) phenotypes. Patients after renal transplantation showed a correction of Lp(a) levels mainly in HMW phenotypes. The LMW status corresponding to high Lp(a) levels and apo(a) isoforms could be used together with Lp(a) levels with other risk factors to assess in uremic patients the predisposition to coronary artery disease.

• Lp(a)
• Apolipoprotein(a) phenotypes
• Renal failure
• Transplant

INTRODUCTION

Elevated plasma levels of lipoprotein(a) (Lp(a)) cause an increased risk of coronary heart disease (CHD) and have been confirmed as an independent risk factor for this disease [[1]]. Lp(a) consists of a low-density lipoprotein particle linked by a disulfide bridge to apolipoprotein(a) [apo(a)], a glycoprotein that varies considerably in size among individuals [[2]], [[3]]. There is a high degree of homology between plasminogen and apo(a) [[4]]. The former protein contains five subunits called kringles (I to V), whereas the latter is characterized by repetition of a widely differing number of the plasminogen-like kringle IV, explaining the interindividual size variation [[4]]. This size polymorphism of apo(a) is determined by the apo(a) gene on chromosome 6, which contains a corresponding number of repeats of the kringle IV encoding sequence [[3]], [[4]]. Individuals with low-molecular-weight (LMW) isoforms have average to high Lp(a) concentrations, whereas those with high-molecular-weight (HMW) isoforms express low Lp(a) serum concentrations [[2]], [[3]], [[4]]. The plasma Lp(a) concentration correlates inversely with the molecular weight of Lp(a), and consequently also with the size of the apo(a) gene [[3]], [[4]], [[5]].

The aim of the present study was to investigate and compare the serum Lp(a) levels in relation to apo(a) phenotypes in hemodialysis (HD), chronic ambulatory peritoneal dialysis (CAPD) and post-transplant patients with chronic renal failure.

MATERIAL AND METHODS

Eighty-one HD patients, 37 CAPD patients and 25 post-transplant patients were selected from the Clinical Nephrology Medical School and Department of Nephrology, Kardynal³ St. Wyszyñski Technical Hospital, Lublin. The reference group included 99 healthy subjects. All studied patients were without diabetes, liver disease, active inflammatory disease, malignancy, obesity or glucose intolerance. The diagnosis of the patients is presented in Table 1. HD and CAPD patients were dialyzed as presented in our previous paper [[6]]. The post-transplant patients were tested 6–43 months after transplantation and had received cyclosporin, prednisolone, and azathioprine but not antihypertensive medications. In patients and the reference group, blood samples were drawn after a 14-h overnight fasting. Routine laboratory parameters were determined using a Roche Cobas Integra analyzer (Roche Diagnostic Systems, Basel, Switzerland). Serum triglycerides (TG), total cholesterol (TC), LDL-cholesterol (LDL-C), HDL-cholesterol (HDL-C), apolipoprotein AI (apoAI) and apolipoprotein B (apoB) were determined as described previously [[7]]. The selected routine clinical and laboratory parameters of the studied and reference groups are shown in Table 2.

Table 1. Diagnosis of Patients

	HD Patients (n = 81)	CAPD Patients (n = 37)	Post-transplant Patients (n = 25)
Glomerulonephritis	M-17, F-18	M-5, F-5	M-8, F-7
Interstitial nephritis	M-5, F-3	M-3, F-3	M-3, F-1
Obstructive nephropathy	M-5, F-9	M-4, F-2	M-2
Hypertensive nephrosclerosis	M-4, F-2	M-3, F-3	M-1, F-1
Polycystic disease	M-4, F-5	M-3, F-1	F-1
Pyelonephritis	M-3, F-6	M-2, F-3	M-1

M—male; F—female.

Table 2. Clinical and Laboratory Parameters in Reference Group, HD, CAPD, and Post-transplant Patients

	Reference Group (n = 99)	HD (n = 81)	CAPD (n = 37)	Post-transplant (n = 25)
Age	35.5 ± 17.7	49.7 ± 14.3	51.7 ± 14.5	33.4 ± 15.2
Gender	50F, 49M	39F, 42M	17F, 20M	15F, 10M
BMI	25.8 ± 3.1	23.1 ± 2.1	25.3 ± 1.9	24.2 ± 2.3
Time dialysis month		40 ± 28.1	43 ± 25.1
Proteinuria g/day	Negative	0.74 ± 0.55	1.52 ± 0,71	Negative
Kt/V		1.23 ± 0.37	2.15 ± 0.22
Serum .0 albumin (mg/dl)	4.42 ± 0.37	3.92 ± 0.512^a	3.68 ± 0.502^a	4.43 ± 0.50
Creatinine (mg/dl)	0.90 ± 0.23	9.91 ± 2.2923^ab	11.6 ± 2.9523^ab	1.68 ± 0.592^a
Urea (mg/dl)	24.5 ± 5.9	150.8 ± 30.223^ab	122.7 ± 35.023^ab	41.1±12.02^a
Uric acid (mg/dl)	4.27 ± 1.27	5.76 ± 1.072^a	5.40 ± 1.022^a	5.27 ± 1.232^a
Na mmol/l	140.3 ± 2.49	140.0 ± 4.2	139.3 ± 2.9	141.9 ± 1.5
K mmol/l	4.39 ± 0.34	5.29 ± 1.192^a	4.40 ± 0.65	4.42 ± 0.50
P (mmol/l)	3.62 ± 0.58	6.08 ± 1.8823^ab	5.76 ± 1.4723^ab	3.65 ± 0.77

^a—vs. control,

^b—vs. post-transplant patients.

BMI-body mass index, Kt/V—index of effectiveness of dialysis, M—male, F—female.

The Lp(a) levels were measured using a monoclonal anti-Lp(a) antibody technique by the ELISA system, which detects natural and recombinant Lp(a) (Boehringer Mannheim). No cross-reaction with serum components other than Lp(a) was found. Sensitivity was >70 µg/dl, the intra-assay and inter-assay coefficients of variation were less than 6.2 and 9.3%, respectively. The apo(a) phenotype was determined by the immunoblotting technique using commercially available kits including PhastSystem (Pharmacia Biotech, Uppsala, Sweden) and Lp(a) phenotype kit (ImmunoAG, Vienna, Austria). The serum sample was reduced and was applied to SDS-PAGE in a 4–15% gradient polyacrylamide gel. The separated proteins were transferred to Immobilon P (Sigma) by diffusion. Then they were treated with a blocking buffer and incubated with a sheep antihuman Lp(a) as the first antibody and with rabbit antisheep IgG(Fc)-alkaline phosphate-conjugate as the second antibody. Subsequent treatment with substrate revealed the Lp(a) bonds. The apo(a) phenotype standard consisted of B, S1, S3 and S4 isoforms whose kringle IV repeats were 14, 19, 23, 27 and 35 respectively [[3]]. The low-molecular weight (LMW) group included all phenotypes with at least one of the isoforms B, S1 or S2, which corresponded to the number of kringle IV repeats; <B and S1 corresponded to the 11–19 K-IV repeats and S2 corresponded to the 20–22 K-IV repeats. The high molecular weight (HMW) group however, comprised subjects with only S3, S4 isoforms or with the null type. The number of kringle IV repeats is 23–25 K-IV repeats, 26–28 K-IV repeats, 29–31 K-IV repeats, and >31 K-IV repeats.

STATISTICAL ANALYSIS

Statistical analysis was performed using one-way analysis of the ANOVA variance and multiple comparisons for assessment of the mean ± S.D. The nonparametric Mann–Whitney U test was applied to discriminate differences of Lp(a) levels between the patient and reference group. Pearson's chi-squared test was applied to compare apo(a) phenotype frequencies between the patients and reference group taking into consideration two subgroups formed by combining subjects with LMW and HMW phenotypes. Spearman's rank correlation coefficient was used to evaluate the degree of association between variables. Statistical significance of the variables was established at the level p<0.05.

RESULTS

Comparisons of lipid and lipoprotein parameters and lipid and lipoprotein ratios are presented in Table 3. HD and CAPD patients had atherogenic lipid and lipoprotein profiles. However, CAPD patients had significantly higher concentrations of TC, LDL-C and apoB, than HD patients, but HDL-C and apoAI levels were significantly decreased for both groups compared to the reference group and post-transplant patients. All studied patients had a significantly increased concentration of TG, but CAPD patients had statistically higher TG levels than HD and post-transplant patients as compared to the reference group. The TC/HDL-C and LDL-C/HDL-C ratios were significantly increased in HD and CAPD patients. On one hand, the HDL-C/apoAI ratio was decreased in HD and CAPD patients, but on the other hand, the apoAI/apoB ratio was decreased only in CAPD patients.

Table 3. Lipids, Apolipoproteins, Lp(a), and Lipoprotein Ratios in Reference Group, HD, CAPD, and Post-transplant Patients

Parameters	Reference Group (n = 47)	HD (n = 72)	CAPD (n = 37)	Post-transplant (n = 23)
TG (mg/dl)	87.9 ± 22.1	136.7 ± 33.556^ab	181.2 ± 59.15^a	131.3 ± 51.256^ab
TC (mg/dl)	175.6 ± 37.1	182.1 ± 44.16^b	229.8 ± 32.85^a	209.9 ± 44.2
LDL-C (mg/dl)	99.8 ± 33.3	110.0 ± 35.16^b	142.4 ± 33.55^a	111.3 ± 43.6
HDL-C (mg/dl)	58.7 ± 12.5	38.7 ± 11.35, 7^b	43.4 ± 11.55, 7^b	57.6 ± 12.5
ApoAI (mg/dl)	155.8 ± 20.5	129.4 ± 20.65, 7^b	125.8 ± 17.05, 7^b	151.6 ± 25.5
ApoB (mg/dl)	88.9 ± 19.1	90.1 ± 22.16^b	120.7 ± 19.05, 7^b	92.8 ± 25.4
TC/HDL-C	3.32 ± 0.78	5.01 ± 1.725^a	5.43 ± 1.825, 7^b	3.64 ± 1.18
LDL-C/HDL-C	2.01 ± 0.64	3.02 ± 1.355^a	3.40 ± 1.285, 7^b	2.08 ± 0.92
HDL-C/apoAI	0.36 ± 0.04	0.30 ± 0.055, 7^b	0.30 ± 0.055, 7^b	0.35 ± 0.05
ApoAI/apoB	1.78 ± 0.44	1.59 ± 0.51	1.15 ± 0.325^a	1.69 ± 0.64

^a—vs. control,

^b—vs. CAPD patients,

^c—vs. post-transplant.

Serum Lp(a) concentrations were significantly increased in HD and CAPD patients as compared to the reference group and post-transplant patients, but CAPD patients had significantly higher Lp(a) levels than HD patients (Table 4). There were differences in the percentage of patients’ (>30 mg/dl) serum Lp(a) levels in all studied groups of patients. The CAPD patients had a higher percentage of serum Lp(a) concentrations (>30 mg/dl) (46%) than HD patients (25%), but the reference group and post-transplant patients had similar percentages of serum Lp(a) levels (>30 mg/dl) (15% and 12%) (Table 5). It was shown that the median serum Lp(a) levels were significantly increased with HMW isoform class in HD and CAPD patients as compared to the reference and post-transplant patients. Moreover, only LMW phenotypes in CAPD patients had significantly elevated median serum Lp(a) concentrations compared to the reference group, but not in HD and post transplant patients. In post-transplant patients, HMW and LMW phenotypes did not differ as compared to the reference group (Table 4).

Table 4. Lp(a) Concentration in the Reference Group, HD, CAPD, and Post-transplant Patients According to apo(a), LMW, and HMW Phenotypes

	Reference Group n = 99	HD n = 81	CAPD n = 37	Post-transplant n = 25
Lp(a) (mg/dl)	8.6 (0.9–99)	16.0 (2.0–140)8910^abc	27.0 (2.0–140)89^ab	9.5 (0.9–90)
LMW (mg/dl)	26.7 (11.7–99)	38.0 (8.2–140)	48.0 (14.0–140)8^a	23.3 (11.5–90)
HMW (mg/dl)	5.0 (0.9–34)	13.2 (2.0–49)89^ab	15.0 (2.0–38)89^ab	4.5 (0.9–26)

^a—vs. reference group,

^b—vs. post-transplant patients,

^c—vs. CAPD patients.

Median (min − max), Mann–Whitney U test.

Table 5. Frequencies of apo(a) LMW and HMW Phenotypes in the Reference Group, HD, CAPD, and Post-transplant Patients

	Reference gr. n = 99		HD n = 81		CAPD n = 37		Post-transpl. n = 25
	No.	%	No.	%	No.	%	No.	%
LMW13^a	30	30.3	24	29.6	16	43.2	7	28
HMW14^b	69	69.7	57	70.4	21	56.7	18	72
Patients with Lp(a) >30 mg/dl (%)		15		25		46		12

Pearson's chi-squared test of LMW vs. HMW phenotypes (x² = 2.78; df = 3; p = NS).

^a—All subjects with the number of kringle IV repeats; 11–19 K-IV repeats, 20–22 K-IV repeats.

^b—All subjects with the number of kringle IV repeats; 23–25 K-IV repeats, 26–28 K-IV repeats, 29–31 K-IV repeats, and >31 K-IV repeats.

Table 5 shows the frequency of the distribution of the LMW and HMW apo(a) phenotypes. When all studied groups were divided into two subgroups with at least one LMW isoform and with only one HMW isoform, it was clear that the studied groups presented a similar distribution (chi-squared = 2.617; df = 3; p = NS), although CAPD patients had a higher percentage of LMW phenotypes (43.2%) than HD patients (29.6%), post-transplant patients (28%) and the reference group (30%). No correlation was observed between Lp(a) levels and albumin in HD patients (r = 0.023, p<0.8), but in CAPD patients, a tendency to inverse correlation (r = −0.26, p<0.1) was observed.

DISCUSSION

Our studies clearly showed that CAPD patients had more atherogenic lipid and lipoprotein profiles than did HD patients. In contrast, in post-transplant patients, an improvement in these profiles [[8]], [[9]] was observed. We also showed in these studies that CAPD patients had statistically higher concentrations of Lp(a) than HD patients [[10]]. Moreover, both HD and CAPD patients had significantly greater Lp(a) levels than the reference group and post-transplant patients. The CAPD patients had a higher percentage (>30 mg/dl) of serum Lp(a) levels (46%) than HD patients (25%), post-transplant patients (12%) and the reference group (15%), in agreement with some previously published data [[11]].

In our results, we did not observe a correlation of serum Lp(a) levels with other lipid parameters that might suggest that the Lp(a) metabolism is independent of the other lipoproteins, at least in patients with renal disease. We also did not find a correlation between Lp(a) concentration and serum albumin levels in all studied groups [[11]], although it was reported that there was an inverse correlation between serum Lp(a) and albumin levels [[8]], [[10]]. Intriguingly, we found that frequencies of apo(a) isoforms LMW and HMW in all studied groups did not differ. It was demonstrated that elevated Lp(a) plasma concentrations in ESRD patients were non-genetic and secondary to renal insufficiency, because restoration of kidney function completely normalized Lp(a) levels [[9]], [[11]].

In agreement with the Kronenberg et al. [[8]] and Milionis et al. [[11]] results, we demonstrated that HD and CAPD patients with HMW apo(a) isoforms showed a significant increase in concentrations of Lp(a) compared to the reference group and post-transplant patients. However, in contrast to Kronenberg et al. [[8]], but in agreement with Milionis et al. [[11]] and Wanner et al. [[12]], we observed in CAPD patients with LMW apo(a) phenotypes higher serum Lp(a) levels than the reference group. It was suggested that a high degree of proteinuria or protein loss through the peritoneal membrane is related to an increased hepatic synthesis of Lp(a) of all apo(a) isoforms classes [[11]], [[13]]. Renal failure seems to influence Lp(a) metabolism.

It was observed that the healthy human kidney is able to remove a large amount of Lp(a) from renal circulation [[14]]. Furthermore, it was demonstrated that in most cases, ACTH induces a shift toward expression of the apo(a) isoform of lower molecular weight. This probably reflects inhibited production of high-molecular weight isoforms [[15]].

Recently, it was reported that patients with LMW phenotypes and a more pronounced atherosclerosis preload develop a “galloping” atherosclerosis after commencement of renal insufficiency or HD or CAPD. It was found that the apo(a) phenotype is more predictive for coronary artery disease than Lp(a) concentration [[16]]. Gazzaruso et al. [[17]] showed that apo(a) isoforms of low MW were the only significant markers of severity of coronary atherosclerotic disease; therefore, the role of apo(a) phenotypes as predictors of coronary lesions severity is very impressive.

CONCLUSION

The HD and CAPD patients had increased Lp(a) concentrations compared with the reference group, which were regarded as apo(a) types-specific, whereas elevated Lp(a) levels were observed mainly in patients with HMW apo(a) phenotypes. Patients after renal transplantation showed a correction of Lp(a) levels mainly in HMW phenotypes. The LMW status corresponding to high Lp(a) levels and apo(a) isoforms could be used together with Lp(a) levels and with other risk factors to assess in uremic patients the predisposition to coronary artery disease.

ACKNOWLEDGMENT

This work was supported by Prof. L. Janicka and Prof. A. Ksiażek with the Department of Clinical Nephrology, University of Medicine, Lublin and Head of Kardynal S. Wyszyñski Technical Hospital, Lublin.