Customize your JAMA Network experience by selecting one or more topics from the list below.
Knudtson ML, Wyse DG, Galbraith PD, et al. Chelation Therapy for Ischemic Heart DiseaseA Randomized Controlled Trial. JAMA. 2002;287(4):481–486. doi:10.1001/jama.287.4.481
Context Chelation therapy using EDTA is an unproven but widely used alternative
therapy for ischemic heart disease.
Objective To determine if current EDTA protocols have a favorable impact on exercise
ischemia threshold and quality of life measures in patients with stable ischemic
Design Double-blind, randomized, placebo-controlled trial conducted between
January 1996 and January 2000.
Setting Participants were recruited from a cohort of cardiac catheterization
patients and the practices of cardiologists in Calgary, Alberta.
Participants We screened 3140 patients, performed a qualifying treadmill test in
171, and enrolled 84. Entry criteria included age at least 21 years with coronary
artery disease proven by angiography or a documented myocardial infarction
and stable angina while receiving optimal medical therapy. The required treadmill
test used a gradual ramping protocol and patients had to demonstrate at least
1-mm ST depression.
Interventions Patients were randomly assigned to receive infusion with either weight-adjusted
(40 mg/kg) EDTA chelation therapy (n = 41) or placebo (n = 43) for 3 hours
per treatment, twice weekly for 15 weeks and once per month for an additional
3 months. Patients in both groups took oral multivitamin therapy as well.
Main Outcome Measure Change from baseline to 27-week follow-up in time to ischemia (1-mm
Results Thirty-nine patients in each group completed the 27-week protocol. One
chelation patient had therapy discontinued for a transient rise in serum creatinine.
The mean (SD) baseline exercise time to ischemia was 572 (172) and 589 (176)
seconds in the placebo and chelation groups, respectively. The corresponding
mean changes in time to ischemia at 27 weeks were 54 seconds (95% confidence
interval [CI], 23-84 seconds; P<.001) and 63 seconds
(95% CI, 29-95 seconds; P<.001), for a difference
of 9 seconds (95% CI, −36 to 53 seconds; P
= .69). Exercise capacity and quality of life scores improved by similar degrees
in both groups.
Conclusion Based on exercise time to ischemia, exercise capacity, and quality of
life measurements, there is no evidence to support a beneficial effect of
chelation therapy in patients with ischemic heart disease, stable angina,
and a positive treadmill test for ischemia.
Ischemic heart disease continues to be the leading cause of death and
disability among North American adults. Testimonials of symptomatic improvement
frequently lead patients with ischemic heart disease to seek alternative therapies
that have not been scrutinized in clinical trials. One such therapy is the
repeated intravenous administration of the chelating agent EDTA in combination
with oral vitamins and minerals. Two small randomized controlled clinical
trials showed no benefit of EDTA in patients with peripheral arterial disease,1,2 and the few available trials in ischemic
heart disease are uninformative.3-5
Although the recent review by Ernst6 concluded
that chelation therapy for coronary heart disease should be considered "obsolete,"
many patients continue to seek alternative therapy.7
There are no current data on the number of patients seeking chelation
therapy. In 1993, Grier and Meyers8 estimated
that more than 500 000 people in the United States are treated with
EDTA therapy each year. In a recent Canadian study, 8% of patients who had
undergone cardiac catheterization and responded to a survey had used chelation
therapy.9 Assuming 8% of the 1.25 million US
residents who have undergone cardiac catheterization10
have tried chelation therapy, we project that 100 000 have tried
chelation therapy. Estimating a cost of $4000 for the usual series of treatment
sums to an annual expenditure of approximately $400 million. The actual amount
is likely higher because these estimates do not include all the cardiac patients
who do not undergo catheterization and all other noncardiac patients who seek
We undertook a randomized, double-blind, placebo-controlled clinical
trial of chelation therapy using the American College for Advancement in Medicine
(ACAM) protocol11 to determine the efficacy
of EDTA with respect to exercise ischemia threshold, symptoms, and quality
of life in patients with stable ischemic heart disease.
Patients were recruited from the Alberta Provincial Project for Outcome
Assessment in Coronary Heart Disease (APPROACH) cohort of cardiac catheterization
patients12 and the practices of community cardiologists
in Calgary. Participants had to be aged 21 years or older and have coronary
artery disease proven by coronary angiography or a documented myocardial infarction
and stable angina while receiving optimal medical therapy. To qualify for
randomization patients were required to have a treadmill test, using a gradual
ramping protocol, demonstrating at least 1 mm of horizontal or downsloping
ST-segment depression from the isoelectric line 80 milliseconds after the
J point. The study protocol required detection of ST-segment depression between
2 and 14 minutes from the onset of exercise.
Exclusion criteria included planned revascularization, previous chelation
therapy, evidence of heart failure, inability to walk on the treadmill, resting
electrocardiographic (ECG) changes that would interfere with ischemic assessment,
abnormal renal or liver function, or untreated lipid abnormality at the time
Treadmill testing was done at baseline and at 15 and 27 weeks after
randomization. The protocol began with a level of exercise equivalent to 2
metabolic equivalents (METs) and increased slowly every 10 to 15 seconds,
reaching an equivalent of 13 METs at 14 minutes. A 12-lead ECG was recorded
every 20 seconds. Maximum oxygen consumption (O2max) and
anaerobic thresholds were determined by continuous measurement of expired
gases using a gas analyzer (MedGraphics model CPX/D; MedGraphics Corporation,
St Paul, Minn) calibrated online.
Patients were randomized in blocks of 10. Investigators were blinded
to treatment assignment. The hospital pharmacy assigned the randomized therapy
and prepared solutions for blinded administration of infusions. The 500-mL
infusion solution of 5% dextrose in water for the active treatment containing
disodium EDTA (Endrate; Abbott Laboratories, Abbott Park, Ill) was weight
adjusted (40 mg/kg), with a maximum total dose for each treatment of 3 g.
Each treatment solution also contained 750 mg of magnesium sulfate, 5 g of
ascorbic acid, and 5 g of sodium bicarbonate (titrated to physiologic pH)
in the 5% dextrose. Lidocaine, 80 mg, was added to relieve pain at the administration
site. In the placebo infusion solution the EDTA was replaced by 20 mL of 0.9%
sodium chloride. The infusion solutions were indistinguishable by color and
labeling. The infusion solution was administered over 3 hours to minimize
the potential unblinding effect of infusion-related adverse effects. All patients
received treatments twice weekly for 15 weeks and once monthly for an additional
3 months, for a total of 33 treatments. In accordance with the ACAM protocol,
patients in both groups took oral multivitamin therapy, 2 tablets 3 times
daily as tolerated, except on treatment days. All patients were seen at the
University of Calgary Cardiovascular Risk Reduction Clinic and had treatment
of their risk profile optimized according to American Heart Association guidelines
(including management of diet, lipid levels, angina, stress, and exercise).
Dipstick urine testing was performed at each visit. Urinalysis and serum
creatinine were measured at every fifth visit. Hematology, electrolyte, and
cholesterol panels were measured at baseline and at 15 and 27 weeks. The study
nurse supervised patients throughout the duration of therapy, and hourly pulse
and blood pressure measurements were obtained.
Exercise parameters and quality of life questionnaires were collected
at 15 and 27 weeks after randomization. The primary end point was the change
in time to reach at least 1 mm of ST-segment depression at the 27-week evaluation.
Patients who did not achieve ischemic changes at 27 weeks had the test period
truncated at 14 minutes, and 14 minutes was recorded as their "time to ischemia."
Functional reserve was also measured by determination of O2max
and time to reach anaerobic threshold. Quality of life instruments included
the Duke Activity Status Index,13 Health Status
Survey Short Form-36,14 and Seattle Angina
All patients were followed up for 1 year from randomization. During
this time, all clinical events were tabulated, including death, myocardial
infarction, coronary artery bypass graft surgery, and percutaneous coronary
All patients signed an informed consent form. The Conjoint Ethics Committee
of the University of Calgary and the Calgary Regional Health Authority approved
this study and its consent form. All clinical events were reported to an independent
safety monitoring committee.
A sample size of 40 per group was chosen to provide 90% power to detect
a 60-second difference in mean change in exercise time from baseline to the
27-week follow-up, assuming an SD of 80 seconds within each group. The 60-second
difference was based on a minimally important difference determined by a consensus
of Calgary cardiologists. Statistical analysis was performed using S-plus,
version 6.0 (Mathsoft, Seattle, Wash). Categorical variables were analyzed
with the χ2 or Fisher exact test, as appropriate. Continuous
variables were examined with paired and unpaired t
tests. Graphical examination of the data showed that normal assumption was
viable. All reported significance levels are 2-sided, and P<.05 was set as the significance level. All analyses of exercise
and quality of life data were conducted using last-observation-carried forward.
A total of 3140 patients (Figure 1)
were screened and 171 of these agreed to undergo a qualifying exercise test.
Eighty-four patients met the treadmill test criteria, consented, and were
randomized between January 1996 and January 2000. Baseline characteristics
according to treatment assignment are shown in Table 1. There were no important differences between the groups.
Of the 84 patients randomized, 78 completed treatment, the final treadmill
test, and the final quality of life assessments (39 in each group). Four placebo
patients and 2 chelation patients were unable to complete the treatment phase
At baseline, mean (SD) treadmill test times to ischemic ECG changes
were 572 (172) seconds in the placebo and 589 (176) seconds in the chelation
groups. Both groups were able to significantly (P<.001)
increase their exercise time to ischemia at the 27-week treadmill test (Table 2). Changes in exercise measurements
of functional reserve (time to anaerobic threshold and O2max)
are shown in Table 2. The magnitude
of the increases in time to ischemic changes and to anaerobic threshold were
not statistically different in the 2 groups. The increase in O2max was not significant at the 27-week treadmill test in the placebo
group but the increase in the chelation group was significant (P = .03). However, the difference between these 2 results was not significant
(P = .46).
The changes in quality of life scores between baseline measurement and
those obtained at the 27-week evaluation are shown in Table 2. There was a tendency for modest increases in quality of
life scores in both groups with significant but similar improvements in the
exertional capacity component of the Seattle Angina Questionnaire. Differences
between the groups were not significant.
Clinical events are presented on an intention-to-treat basis (all 84
patients included). The duration of follow-up was 1 year from randomization
for each patient. There were no deaths during that time. One patient in the
placebo group had a documented myocardial infarction and 6 other patients
were admitted at least once for worsening angina. Four of these 7 patients
had angioplasty and none had coronary artery bypass graft (CABG) surgery for
these events, although 1 other patient had elective surgery (CABG was planned
by the cardiologist after randomization without investigators' knowledge).
There was 1 myocardial infarction in the chelation group and 9 patients were
admitted at least once for worsening angina. None of these had angioplasty
or CABG surgery associated with these events.
One of the chelation patients was withdrawn from therapy because of
an elevation in serum creatinine. During the first 10 treatments the patient's
serum creatinine level increased from 1.5 to 2.1 mg/dL (129 to 186 µmol/L,
respectively). Treatment was stopped and the serum creatinine level decreased
to 1.6 mg/dL (138 µmol/L) after 10 weeks. No other cause for the elevation
in creatinine was found. In addition to the nonischemic events shown in Figure 1 leading to discontinuation of therapy,
3 additional placebo patients were hospitalized for nonischemic events: gout,
lumbar back pain from a herniated disk, and gastrointestinal bleeding. These
events did not interfere with completion of the treatment phase. There were
no electrolyte results out of normal range during the study.
The main finding of this study was that chelation therapy had no beneficial
effect on exercise time to ischemia, functional reserve for exercise, and
quality of life in patients with proven ischemic heart disease, stable angina,
and evidence of ischemia on treadmill examination. Accordingly, chelation
therapy remains unproven in the treatment of ischemic heart disease.
EDTA is an amino acid complex with a high affinity for divalent and
trivalent cations such as lead, magnesium, zinc, iron, and calcium. Conventional
chelation therapy involves multiple infusions of EDTA to chelate lead, iron,
copper, calcium, and other metal ions in a redox inactive state. Chelated
metal ions are then excreted from the body in the urine. For this reason EDTA
has been used as a chelating agent in clinical situations in which these elements
are found in excess.16
Because calcium is often found in atheromatous plaques, early proponents
hypothesized that EDTA might be effective in treating ischemic heart disease
by liberating plaque calcium with a subsequent favorable change in the properties
of the plaque.3,4,17,18
Other hypotheses possibly accounting for the symptomatic improvement reported
by many patients with ischemic heart disease include a free radical–scavenging
function, inhibition of lipid oxidation (antioxidant), reduction of total
body iron stores, cell membrane stabilization, arterial dilation due to possible
calcium channel blocking actions, or stimulation of prostacyclin production
and improvement in arterial wall elasticity.3,4,17-21
Oxidized cholesterol plays an important role in endothelial function and the
formation of atherosclerotic plaque.22 There
is at least some evidence, albeit controversial, that increased total body
iron stores are associated with increased ischemic heart disease.23 Therefore, some of these hypotheses about chelation
having a potential mechanism for benefit in ischemic heart disease have plausibility.
In the absence of studies confirming such effects and, more importantly, confirming
a definite clinical benefit of chelation therapy, it remains possible that
anecdotally reported improvements are simply due to the spontaneous fluctuations
in symptoms frequently seen in ischemic heart disease.24-27
In our trial, the 1-minute increase in exercise time to ischemia and the improvement
in the exertional capacity component of the Seattle Angina Questionnaire in
both groups is consistent with a combination of placebo24-27
and training effects28-30
commonly seen in studies of angina patients. Another potential explanation
for improvement is that both groups were treated with optimal risk reduction
Chelation therapy is practiced and promoted as a form of complementary
or alternative medicine in many developed countries. Additional vitamins and
mineral supplements are recommended for patients undergoing chelation therapy.
In our study, both groups received multivitamins; we cannot exclude the possibility
that these supplements could be partially responsible for the improvement
that we saw in both groups.
In the literature, numerous authors have reported positive results in
Very few randomized clinical trials have been published on the effects of
chelation therapy, and those that have been published were performed in patients
with peripheral arterial disease.1,2
Olszewer et al32 published a small trial of
10 men with peripheral arterial disease in which improvement was demonstrated
in walking distance after 20 treatments, but there were only 5 patients in
each group and therapy was not blinded. Guldager and colleagues1
published a randomized, double-blind, placebo-controlled trial of 153 male
patients with peripheral arterial disease (75 EDTA and 78 placebo), and during
the 6-month follow-up no effect of EDTA on walking time or ankle-brachial
blood pressure index was demonstrated. That trial has been criticized for
the high dropout rate (123 completed 6-month follow-up). van Rij et al2 published the results of a similar randomized, double-blind,
placebo-controlled trial of walking time and ankle-brachial blood pressure
indices in 32 patients (15 EDTA and 17 placebo) with claudication, which also
showed no effect of EDTA therapy.
There is even less evidence in patients with ischemic heart disease.
Kitchell et al5 conducted a placebo-controlled,
double-blind, crossover study of 9 patients with coronary heart disease and
assessed performance on a treadmill. The authors documented that 2 of 4 EDTA-treated
patients benefited at 12 weeks but only 2 patients volunteered to be treated
in the second phase, and neither patient showed improvement. No statistical
analyses were presented in that study.
As with all randomized clinical trials, our results can be applied to
fit only a similar population to that studied: patients with stable angina
who are not candidates for revascularization and can exercise on a treadmill.
Our study showed that following 33 treatments with EDTA therapy, there was
no evidence of any benefit compared with placebo in either objective measurements
of exercise capacity or in measurements of patient-perceived well-being. One
patient receiving EDTA had a transient increase in serum creatinine. There
was no difference in the number of clinical ischemic events, but our study
was not powered to detect any such differences. According to our findings,
the use of chelation therapy to increase ischemic threshold and improve quality
of life cannot be supported for patients with ischemic heart disease. Larger
trials with a broader range of patients will be needed to assess the safety
and impact of EDTA therapy on clinical event rates.