Serum lipoprotein (a) (Lp[a]) has been associated with coronary artery atherosclerosis. Its association with restenosis after percutaneous transluminal coronary angioplasty (PTCA) has not been previously studied. Serum levels of Lp(a), in addition to other lipoproteins, and their components using standard assays, were determined in subjects undergoing cardiac catheterization within 10 months after PTCA. Clinical (e.g., sex, diabetes, angina class) and angiographic (e.g., PTCA percent diameter reduction) factors were not different between the group without (diameter reduction less than 50%; group A) and the group with (diameter reduction greater than or equal to 50%; Group B) restenosis. Total cholesterol, triglycerides, high- and low-density lipoprotein cholesterol, apolipoprotein A-I, apolipoprotein B and Lp(a) were compared. Univariate predictors of restenosis were serum triglycerides (2.50 +/- 1.07 mmol/liter for group A vs 1.72 +/- 0.79 +/- mmol/litre for group B, p = 0.008), and Lp(a) (median: 7.0 mg/dl [range 0 to 44] for group A vs 19 mg/dl [range 1 to 120] for group B; p = 0.006). Stepwise logistic regression revealed the only significant independent predictor of restenosis to be serum Lp(a) (p = 0.018). Each quintile of Lp(a) was associated with a progressively higher risk of restenosis, with the highest quintile (40 to 120 mg/dl) having an odds ratio of 11 (95% confidence interval 9 to 13) compared with the lowest quintile (0 to 3.9 mg/dl) (p = 0.033). A serum Lp(a) of greater than 19 mg/dl was associated with an odds ratio of 5.9 (95% confidence interval 4.6 to 7.2) (restenosis rates of 58% in the group with 0 to 19 mg/dl and 89% in the group with 19 to 120 mg/dl; p = 0.006).(ABSTRACT TRUNCATED AT 250 WORDS)