Clinical and Inflammatory Effects of Dietary L-Arginine in Patients With Intractable Angina Pectoris Arnon Blum, MD, Reuven Porat, MD, Uri Rosenschein, MD, Gad Keren, MD, Arie Roth, MD, Shlomo Laniado, MD, and Hylton Miller, MB N itric oxide (NO) is primarily synthesized by en- dothelial cells and is a potent vasodilator. L- arginine, a precursor of the enzyme nitric oxide syn- thase (NOS) 1 has been extensively studied in animal models and in human beings. L-arginine enhances NO bioavailability in the vascular endothelium, 2 attenu- ates enhanced monocyte-vascular endothelium adhe- siveness, and restores endothelium-dependent vasodi- lation. 3 L-arginine can be given intra-arterially, 4 intra- venously, 5 and orally. 6 The advantage of the oral route is the longer half-life and longer term effect, which are both short (1 minute) in the intra-arterial and the intravenous routes. This study assesses the effects of oral L-arginine on the clinical and inflammatory state of patients with coronary artery disease and intractable angina pectoris. ••• The inclusion criteria were coronary artery disease documented by coronary angiography, angina pectoris functional class IV, previous revascularization (coro- nary angioplasty or coronary artery bypass surgery), and a thallium perfusion scan that demonstrated a reversible perfusion defect in 1 segment. The exclu- sion criteria included significant heart failure (New York Heart Association more than or equal to class III), chronic disease, and malignancy. This study was a prospective, case-controlled, pilot study. Ten men were enrolled, all had undergone coronary angiogra- phy and angioplasty, and 9 had undergone coronary artery bypass surgery before enrollment. All suffered angina pectoris class IV and had frequent attacks of angina at rest and at night despite large doses of blockers, calcium channel blockers, nitrates, and as- pirin; no further intervention could be offered to these patients. The average age was 72 10 years (range 48 to 80); all were men with hypercholesterolemia (well treated), 6 were hypertensive, and 4 had diabetes mellitus (noninsulin dependent). Mean left ventricular ejection fraction was 56 13%. Each patient had a clinical examination before starting the study, and monthly afterwards for 3 months. Patients were re- quested to fill out The Seattle Angina Questionnaire 7,8 before treatment and after 1 and 3 months of treat- ment. Blood samples were taken at enrollment and at each of the 3 monthly visits. The serum was separated by centrifugation and stored at -70° C until analyzed in 1 batch. Serum concentration of soluble intercellu- lar adhesion molecule-1 (sICAM-1), vascular cell ad- hesion molecule-1 (sVCAM-1), E-selectin, P-selectin, and L-selectin were measured by monoclonal anti- body based enzyme-linked immunosorbent assay kits (British Biotechnology Products Ltd., Abdington, Ox- fordshire, United Kingdom and Takara Biomedicals, Kurume, Japan). Briefly, known concentrations (for a standard curve) of adhesion molecules and serum samples (100 l) were added to 96-well plates pre- coated with the first monoclonal antibody and incu- bated for 1 to 2 hours at 37°C. After washing, a second antibody labeled with horseradish peroxidase was added for additional inclusion (1 hour at 37°C) and the plates were vigorously washed again. Substrate solu- tions were then added for a short incubation period (30 minutes) followed by the addition of 1 N H 2 SO 4 , and the reaction stopped. Absorbance was read at 492 nm. Serum concentrations of cellular adhesion molecules were calculated based on the standard curve. Circu- lating tumor necrosis factor-, interleukin 1-, and interleukin-6 were measured by specific non– cross- reactive radioimmunoassay as described. 9-11 Cytokine concentrations were read from a log plot of percent specific binding versus the log concentration of seri- ally diluted standard cytokines from the linear portion of the curve (between 35% to 75% specific binding). The lower limit of detection of the assays were 40 pg for tumor necrosis-, and 78 pg for interleukin 1- and interleukin-6. Data were expressed as mean SD. For data comparisons the nonparametric test for 2 matched samples and the paired Student’s t test were used. All patients took L-arginine 9 g daily for 3 months. No one reported side effects from the L-arginine (con- stipation, diarrhea, hypotension, abdominal cramps). Seven patients improved clinically (from angina pec- toris functional class IV to functional class II); their improvement was persistent and continued as long as L-arginine was taken. Discontinuing L-arginine sup- plementation (after 3 months) caused a “rebound phe- nomena,” and they deteriorated to functional class IV. One patient improved to functional class III, and 2 patients remained in functional class IV. After 1 month of L-arginine, cell adhesion mole- cule and cytokine levels remained stable throughout 3 months of the study. Individual data are presented in Table I. In 7 patients who had a significant clinical improvement, P-selectin level decreased from 338 40 pg/ml to 304 31 pg/ml (p = 0.01), ICAM level decreased from 315 57 pg/ml to 274 36 pg/ml (p = 0.08), and E-selectin changed from 47 10 to 44 11 (p = 0.09). VCAM and L-selectin levels From the Cardiology Branch, National Heart, Lung, and Blood Insti- tute, National Institutes of Health, Bethesda, Maryland; and Cardiol- ogy Department and Internal Medicine B, Tel-Aviv Sourasky Medical Center and Tel-Aviv University, Tel-Aviv, Israel. Dr. Miller’s address is: Catheterization Laboratory, Tel-Aviv Medical Center, 6 Weizman St., Tel-Aviv, 64239, Israel. E-mail: himiller@tasmc.health.gov.il. Manu- script received November 19, 1998; revised manuscript received and accepted January 14, 1999. 1488 ©1999 by Excerpta Medica, Inc. 0002-9149/99/$–see front matter All rights reserved. PII S0002-9149(99)00129-0