Quackwatch Home Page
Planned clinical trials of ethylene-diamine-tetra-acetic acid (EDTA) chelation therapy by the National Center for Complementary and Alternative Medicine and others call for investigation of chelation's biochemistry and pharmacology, its toxicity, and the history of claims made for it.
EDTA, known to reduce serum levels of polyvalent metals by chelation, was proposed in the late 1950s for removal of calcium from atherosclerotic plaques. Proponents now claim that EDTA can remove toxic heavy-metal ions and that it can neutralize or reduce oxygen free radicals. A review of atherosclerosis pathophysiology and EDTA chemistry reveals that (1) EDTA chelation effectiveness is implausible; (2) the preponderance of evidence shows ineffectiveness; and (3) EDTA augments oxidative reactions involving iron instead of inhibiting them, resulting in increased likelihood of production of oxygen free radicals rather than neutralization of them, as claimed.
Further investigation of this therapy for atherosclerosis and degenerative diseases may be ethically questioned.
The initial claim for EDTA chelation (1950s) was based on an assumption that the structure of arterial plaque depended on its calcium content. Calcium was compared to rivets in a steel structure, the removal of which would cause the arterial plaque to disintegrate, thus enlarging the lumen's diameter and increasing blood flow. This mechanism was likened to that of a drain cleaner cleaning a water pipe [1-4].
By the mid-1970s the mechanism of calcium removal had been challenged, as no proof had been presented. Advocates invoked parathormone (PTH) to explain perceived benefits, theorizing that when ionic calcium was removed from serum by EDTA chelation it was replaced by calcium from bone, the loss of which was said to stimulate PTH secretion, which promoted remineralization of bone. The calcium for bone remineralization was said to be supplied through "gradual transfer" of calcium from hardened arterial tissue and plaque. This was said to soften the arteries and cause the plaque to disintegrate [1,5].
By the early 1980s, basing theory on more recent knowledge of free-radical effects on tissue, proponents posited a new and current theory of free-radical neutralization. The proposed mechanism was that toxic metals such as iron and copper, released by clot lysis at arterial injury sites, generate free radicals that oxidize fatty acids to lipid peroxides, which then would generate new free radicals. This chain of oxidation reactions would cause arterial cell membrane damage and plaque formation. This mechanism was more in line with then-current thinking on the pathophysiology of atherosclerosis. Advocates then claimed that through EDTA binding, iron would become chemically nonreactive, cease catalyzing free-radical production, and thus curb the pathological processes that cause atheromas .
A more recent chelation advocate theory states that atheromas are benign tumors that arise when artery cells mutate as a result of free-radical damage to their DNA .
The modern consensus of the genesis and pathology of atherosclerosis differs from both older and recent chelation theories. In brief, arterial atheromas begin as low-density lipoproteins (LDL), cross the endothelial cell layer of the artery at a point of local injury or oxidant damage, and are deposited in the subendothelial layer. Monocytes, attracted to the injured area and subendothelial layers, engulf LDL and become foam cells, collectively forming a fatty streak on the arterial wall. The accumulation of foam cells ruptures the arterial endothelial cell layer, and platelets aggregate at the site and release growth factors, stimulating smooth-muscle cells to proliferate. During repair, cells produce collagen and form a fibrous, collagen-rich cap over the site (plaque). This plaque contains cholesterol, lipid particles, fibrous protein, and debris. The plaque enlarges as the process repeats. Calcium deposition inside the atheroma is a late event, and is not intrinsic to the genesis or maintenance of the atheroma's structure. Calcium rarely occupies more than a small fraction of the plaque's volume .
The American College of Advancement in Medicine (ACAM), an organization of physicians advocating use of EDTA chelation, developed the following protocol for EDTA administration . An intravenous infusate of 500 to 1000 ml of Ringer lactate or 10% fructose solution contains 50 mg EDTA per kilogram body weight, heparin, magnesium chloride, lidocaine, pyridoxamine, B-complex vitamins (including vitamin B12), and gram (usually 5 gm) amounts of vitamin C (ascorbate). The solution is infused slowly over 3.5 to 4 hours, 1 to 3 times a week. The initial series is about 30 infusions, with the possibility of additional treatments later. Minor changes have occurred over the past 20 years, including use of magnesium-containing EDTA, but no pharmacologically significant change has occurred. Adjunctive therapy consists of gram doses of oral ascorbate; vitamin E; mineral supplements of magnesium, calcium, potassium, copper, iron, chromium; and vitamins at doses ranging 3 to 5 times the recommended daily intake (RDI). Optional oral supplements include pancreatic enzymes, thyroid extract, estrogen, fiber, dimethylglycine, and iodine.
Lifestyle modifications include exercise program, stress reduction, caffeine and tobacco avoidance, alcohol limitation, and nutritional counseling.
Current charges are in the range of $100 per infusion, with an average of 30 treatments. There are about 200 physician practitioners in the United States practicing EDTA chelation for degenerative disease.
EDTA was first synthesized and patented in 1938 . It is a highly negatively charged ringlike compound with 4 negative charges that hold polyvalent metal ions in a highly stable, water-soluble form. The strength with which each metal is held varies; in order of decreasing strength as follows: iron+++, mercury++, copper++, aluminum++, nickel++, lead++, cobalt++, iron++, zinc++, cadmium++, manganese++, magnesium++, and calcium++. The relative amount of each ion held by EDTA depends on each element's relative affinity, the strength of its bonding to tissue and transport proteins, and its concentration in plasma.
In general, a preparation of EDTA containing an ion with lesser affinity will exchange that ion for one with greater affinity. For instance, EDTA has a lesser affinity for sodium than for calcium. Thus, disodium EDTA will exchange its sodium ions for a calcium ion, and reduces hypercalcemia of malignancy in emergency situations [7,8]. Calcium-sodium EDTA, another commercial form, will exchange its calcium for a metal higher on the affinity scale and is approved for use in heavy-metal poisoning such as lead poisoning.
The postulated reason for lead binding to EDTA in preference to mercury, for which EDTA has a greater affinity, is that mercury is more tightly bound to tissue ligands by -SH bonds, or that it exists in compartments not available to EDTA. Copper and iron, also held more powerfully to EDTA are also more strongly bonded to tissue and transport proteins -- ferritin and transferrin for iron, ceruloplasmin for copper . Therefore, although EDTA binds some copper and iron, it is not useful for reducing overload states.
EDTA-metal complexes are excreted by the kidney with a biological half-life of 20 to 60 minutes, Rapid infusion of disodium EDTA may cause acute hypocalcemia and tetany, weakening of cardiac muscle contraction, and arrhythmias. However, slower infusion (<15 mg/mm) is tolerated with no symptom or immediate toxicity .
Chelation advocate theories of calcium metabolism are also inconsistent with scientific consensus. Normal calcium levels are necessary for normal functioning of almost all tissues. Serum calcium levels are normally maintained at between 9 and 11.0 mg/dl. Bone is the major calcium storage and active pool site, containing over 99% of total body calcium. Soft tissues contain 0.6%. Plasma contains 0.03% (approximately 350 mg) and extravascular fluid contains 0.07% (approximately 700 mg). Plasma calcium level is maintained by PTH and 1,25 dihydroxyvitamin D-3 (calcitonin). These hormones regulate calcium absorption from the gut, phosphate and calcium reabsorption in the kidney, and calcium mobilization from bone.
Remineralization of bone uses calcium from plasma. A fall in plasma calcium triggers PTH secretion. PTH effect increases calcium reabsorption by the kidney, and renal synthesis of calcitonin, increasing calcium absorption from the gut.
Soft-tissue calcium is kept within physiological limits by exchange with ions in the extracellular fluid. There is no normal physiological mechanism by which soft tissue releases calcium for bone remineralization, and there is no process that selectively releases calcium from arterial wall plaques while leaving normal tissues untouched. Thus, there is no selective action of EDTA or other mechanism on arterial wall calcium. EDTA calcium mobilization through the action of PTH is almost entirely from bone .
Clarke et al. reported in 1956 that a series of patients felt
better after EDTA treatment for occlusive peripheral vascular
disease, but there were no controls . In 1960 Meltzer reported
on angina patients with EDTA, finding subjective benefit, but
no evidence of objective
Benefit . In 1963 Kitchell retrospectively analyzed 28 of Meltzer's EDTA-treated patients and found that the improvements initially reported were short-lived and that there was no alteration of disease progress or of the degree of pathology .
More recent proponent studies reviewed by Chappell were reported to have shown improvement of various indicators, including sense of well-being, angina, and macular degeneration. None of the studies was a randomized, controlled trial .
Another retrospective study reviewed 22,765 patients with a variety of degenerative diseases. Eighty-seven percent of patients were reported to have shown some measurable improvement . A later analysis of unpublished data showed similar results . These claims were rebutted by Margolis, who showed the studies were reported in a non-peer-reviewed journal created as an organ for physicians who perform EDTA chelation. Margolis found some studies were retrospective, some showed control outcomes equal to treated patients, and some studies broke blinds and codes prematurely .
Early use produced complications because of high EDTA concentration -- larger than 3 gm/L -- and too-rapid infusion. Early deaths occurred from renal tubular necrosis, cardiac arrhythmias,and acute hypocalcemia. Lesser complications included allergic reactions, hypotension, and marrow depression . The trace metal most dramatically lost as a result of EDTA chelation is zinc.
Allain found that 24 hours after an infusion of EDTA the urine of human subjects contained 15 times the normal amount of zinc .
In another study of 60 patients for metal excretion with EDTA, lead and zinc were found to be in 25-fold higher concentrations in 24-hour urine collections after the first infusion. Zinc plasma concentration was reduced by 34% by 5 to 9 weeks . In contrast, as shown elsewhere in this report, disodium EDTA chelation drops the serum calcium concentration acutely, but results in negligible long-term loss. The zinc and other losses probably occur through exchange of the sodium-calcium EDTA first formed, with serum zinc and other metals higher on the EDTA affinity scale.
Ferric iron has 2 electrons in its outermost, or N, shell and 14 electrons in its M shell. This configuration confers a characteristic of accepting 3 pairs of electrons from other ions. As long as 1 pair of these electrons is left unbound, ionic iron remains highly reactive .
When iron is dissolved in water at pH of 7 or higher, its 3 pairs of electrons are bound to three OH- groups of water. The resulting Fe(OH)3 or ferric hydroxide is insoluble and precipitates. In contrast, when ionic iron is chelated with EDTA's 4 charged sites, only 2 pairs of electrons are bound. Ferric ions thus remain in solution, and the remaining electron pair is left free to be involved in oxidation reactions, generating free radicals. Therefore, EDTA-chelated ionic iron does not stop free radical generation, but by keeping ionic iron dissolved, it magnifies the production of free radicals,.
EDTA infusion solutions usually contain megadoses of ascorbate, furthering potential for free radical formation, as shown by Herbert . Normally, most iron in the body is bound to protein and produces only a limited number of free radicals. Ascorbate causes release of iron from storage and transport protein, increasing the amount of ionic iron available for EDTA chelation, and increasing the potential for free-radical production . Pharmacologic amounts of ascorbate also increase iron absorption from the gut, producing even more available ionic iron. A number of studies have demonstrated deleterious effects on DNA synthesis from EDTA with and without ascorbate . Several studies have demonstrated teratogenesis [23-26].
Recent concern over interference with drug action by oxidizing and reducing agents has led to a series of experiments demonstrating enhancement of redox reactions by EDTA and demonstrating the paradoxical pro-oxidant properties of ascorbate.
Green first described oxidant qualities of EDTA in the presence of iron  The substrate was epinephrine, which was found to have been oxidized. Pro-oxidant characteristics of ascorbate have also been studied and recorded [21-27]. In a series of redox reactions (Haber-Weiss/Fenton reactions) oxygen and an electron donor -- such as hemoglobin, cytochrome, and a redox metal such as iron -- interact with hydrogen peroxide to oxidize the electron donor substrate. Red blood cell membrane changes similar to those of thalassemia and hemoglobin degradation with Heinz body formation (de-natured hemoglobin chains) have been produced by these reactions .
EDTA- iron-reducing agent reactions as described by Green create hydroxyl-free radical (OH-o) and other reactive oxygen species that oxidize LDL and cell mem-branes. This EDTA-induced free-radical oxidant reaction was demonstrated on DNA by Dervan, who used it to develop standard methods for induction of single-stranded DNA breaks .
Since EDTA does not enter cells, its free-radical-generating action in vivo would be limited to the extracellular compartment with greatest potential action on LDL and cell membranes -- especially vascular endothelial cells. The presence of ascorbate in the infusate magnifies the conditions for the above reactions. Thus, instead of protecting against and neutralizing metallic free radicals, EDTA in presence of iron and ascorbate produces free radicals and potentially induces the changes that it is intended to prevent.
These interactions are apparently not commonly appreciated, although known to organic chemists and experimental biologists for decades. Because of the short exposure time to EDTA during chelation infusions, and because of the long time needed to develop even microscopic changes, gross EDTA-induced changes may not be detectable in short-term investigations. In addition, the advanced age of most EDTA recipients makes it makes it unlikely that an increase incidence or severity of major disease would be manifest in survey data. However, it is likely that oxidant reactions do occur during infusion, that existing disease is worsened, and that younger patients may eventually show evidence of acceleration of their disease.
Few animal studies have been done -- an unusual situation for a therapy advocated to be applied in humans beneficially and universally. In one study of atherosclerotic, obese, diabetic rats (JCR:LA-cp strain), EDTA chelation was found to have no benefit on arterial lesions, but produced in a 74% increase in triglyceride levels .
A review of literature data to 1984 concluded, "[For] Chelation therapy with intravenous injections of edetate sodium . . . promoted to treat coronary and other arterial atherosclerosis . . . evidence . . . is lacking." Based on reviews of the world medical literature, these same conclusions were reached by the U.S. Food and Drug Administration, the National Institutes of Health, the National Academy of Sciences and National Research Council, the California Medical Association, the American Medical Association, the Centers for Disease Control, the American Heart Association, the American College of Physicians, the American Academy of Family Practice, the American Society for Clinical Pharmacology Therapeutics, the American College of Cardiology, and the American Osteopathic Association .
Randomized, blinded clinical trials since then have been negative. In 1985 and 1986, Diehm et at. treated 45 patients with high-grade arterial obstruction and intermittent claudication with either EDTA chelation or Bencyclan, an anticoagulant. Measures included ability to perform pain-free walking exercises, blood flow, red cell viscosity, erythrocyte aggregation, triglyceride, and cholesterol levels. Measurements made during the 4-week treatment period and for 3 months after treatment stopped showed both treatment and control groups improved equally. Diehm concluded that improvement was due to motivation from his strong interest in subjects' well-being and their motivation to perform .
In 1987 Hopf reported on 16 patients with angiographic evidence of coronary heart disease, randomized to an EDTA chelation group and a non-EDTA group. Patients were infused with 500 ml of solution at 3-day intervals for 20 infusions. Both groups showed the same degree of subjective improvement and performed ergometric weightlifting tests equally well. Clinical exercise testing showed no reduction in ischemic reaction in either group, both showing slight progression. EDTA chelation had no effect on the disease process [32,33].
In 1992, 153 patients with intermittent claudication entered a double-blind, randomized, placebo-controlled trial of EDTA chelation. Each group received 20 infusions of either EDTA or placebo for 5 to 9 weeks, attempting to replicate the method used by Olszewer and Carter in 1990. Pain-free and maximal walking distances were similar for the EDTA and placebo groups, and there were no long-term therapeutic effects after 3 and 6 months .
Van Rij et al. randomized 15 patients with intermittent claudication to EDTA chelation and 17 patients to control infusions. No overall difference was found in 35 different measures of function and symptoms. Only 2 of 18 measures of lifestyle showed clinically significant improvement .
A recent trial at the University of Alberta, in cooperation with the ACAM, randomized 84 patients with coronary artery disease (CAD) to EDTA chelation or dummy infusions twice a week for 15 weeks, followed by 3 monthly injections. Serial treadmill tests showed no difference between groups .
Ernst performed a systematic review of all reported clinical trials of peripheral vascular disease and found no evidence for effect greater than placebo and advised that it be considered obsolete . In a subsequent overview of all clinical investigations for coronary heart disease through 1999, Ernst found no evidence for effectiveness, and recommended that chelation "be discarded in favor of therapies of proven effectiveness." 
Stoichiometric features of EDTA chelation predict that the maximum effect, if any were to occur, would be negligible. Since EDTA holds calcium 1 mole to 1, 1 gm of EDTA holds 0.120 gm of calcium; 3 gm EDTA, the amount in each infusion, holds 0.36 gm of calcium. A 70 kg person's total body calcium is about 1.7 kg or 1700 gm. Calcium removed by one infusion is 0.360/1700, or about 0.0002 of total body calcium. A 1-month course of EDTA chelation would remove at most 0.001-0.003 of the calcium in arterial plaques. In order to remove only 10% of the calcium in any structure, one would require 530 daily treatments, 5 days per week, for 2 years. The cost, at $100 per infusion, would be $50 000 .
Meanwhile the intestinal tract absorbs 0.5-1.0 gm of calcium daily from the diet, while renal conservation occurs as well, if compensatory mechanisms function normally, no net calcium loss from EDTA chelation therapy would occur.
The same principle applies regarding neutralization of free radicals. Given a 4-hour infusion, and a 75-year (675 000-hour) lifetime, a course of 20 EDTA chelations could remove a maximum of only 1/7000 of the free radicals one is exposed to in a lifetime .
The U.S. Federal Trade Commission in 1998 cobtained a consent order barring the ACAM not to advertise EDTA chelation therapy as effective. However, the agreement does not bind individual physicians, who may advertise and practice chelation. The ACAM and supporters succeeded in having 10 states pass "Access to Medical Treatment Acts" (AMTA), the latest being Colorado and Minnesota. These bills forbid state medical boards from disciplining physicians for using ineffective treatments as long as patients agree to receive them. The province of Ontario passed a similar bill, and other bills are before other state and provincial legislative bodies. A similar U.S. federal bill has appeared over the last 6 years but has not passed.
Specific experiments are needed for EDTA-iron oxidant activity on lipoproteins, cell membranes, and other oxidation substrates in other test systems. In view of the negative clinical studies done so far, the theoretical implausibility of effectiveness, and the biochemical evidence of oxidant damage, there seems to be no indication for pursuing clinical trials. Institutional review boards will have to question the ethics of EDTA use clinically and experimentally.
Dr. Green is President, Zol Consultants, and formerly Professor of Biochemistry at the Sloan-Kettering Cancer Institute, New York. Dr. Sampson is Clinical Professor of Medicine, Emeritus, at the Stanford University School of Medicine and Editor of the Scientific Review of Alternative Medicine. This article was published in the journal's Winter 2002 issue (© 2002 Prometheus Books, all rights reserved). Correspondence concerning this article should be addressed to Wallace Sampson, MD, 841 Santa Rita Avenue, Los Altos, CA 94022; e-mail: firstname.lastname@example.org.