Multicenter study of lipoprotein(a) and apolipoprotein(a) phenotypes in patients with end-stage renal disease treated by hemodialysis or continuous ambulatory peritoneal dialysis.

Numerous studies have investigated lipoprotein(a) (Lp(a)) plasma concentrations in patients with ESRD, a patient group with an enormous risk for atherosclerosis. The reported differences in Lp(a) between controls and patients vary from a decrease of 49% to an increase of more than 1,000%. However, data are not consistent, mostly because of problems with statistical analysis, and only limited data are available for patients treated by continuous ambulatory peritoneal dialysis (CAPD). To estimate the significance of Lp(a) in ESRD and to demonstrate the statistical pitfalls concerning Lp(a) in case-control studies, a large multicenter study including 702 patients treated by either hemodialysis (HD) (N = 534) or CAPD (N = 168) was conducted, and results were compared with results from 256 healthy controls. Both patient groups showed significantly elevated Lp(a) levels in comparison with controls: 23.4 +/- 25.0 mg/dL (P < 0.005; HD) and 34.6 +/- 38.4 mg/dL (P < 0.0001; CAPD) versus 18.4 +/- 22.8 mg/dL (controls). CAPD patients showed significantly higher Lp(a) values than did patients treated by HD (P < 0.001). The difference between the two treatment groups possibly reflects an overproduction of Lp(a) to compensate for protein losses in CAPD patients. Both treatment groups included significantly more patients with Lp(a) values greater than the 75th percentile (25.6 mg/dL) of the control group (33.9 and 41.7% for HD and CAPD, respectively; P < 0.005). The higher Lp(a) values in patients were not explained by differences in isoform frequencies and the increase in Lp(a) was apolipoprotein(a) type specific: only patients with high-molecular-weight apolipoprotein(a) isoforms showed a significant elevation in Lp(a) levels. The increased plasma concentrations of Lp(a) may contribute to the high risk for atherosclerosis in ESRD, especially in patients treated by CAPD. Finally, it is believed that small sample sizes are responsible for the diverging results in Lp(a) literature.