Chelation With Multivitamins for Coronary Disease
Between September 10, 2003, and October 4, 2010, 1,708 patients were randomized: 421 to EDTA chelation infusions + high-dose oral multivitamins, 418 to EDTA chelation infusions + oral placebo, 432 to placebo infusions + high-dose oral multivitamins, and 437 to placebo infusions + oral placebo. The median duration of follow-up was 55 months (interquartile range, 26–60) overall. There was no significant difference in length of follow-up across all 4 groups.
Baseline characteristics were similar among the 4 randomized factorial groups (Table I). Patients were 65 (59–72) years old, 18% were female, and 9% were minorities. The qualifying MI had occurred 4.6 (1.6–9.2) years prior to enrollment. There was a high prevalence of diabetes (37%); prior coronary revascularizations (83%); and postinfarct, guideline-recommended medication use of aspirin (84%), β-blocker (72%), and statin (73%).
The 5-year Kaplan-Meier event rate estimate for the primary end point was 31.9% in the chelation + high-dose vitamin group, 33.7% in the chelation + placebo vitamin group, 36.6% in the placebo infusion + active vitamin group, and 40.2% in the placebo infusions + placebo vitamin group (Figure 1, A; Table II). The primary end point by treatment group occurred in 139 (32%) of the placebo infusion + placebo vitamin group and 108 (26%) of patients in the chelation + high-dose vitamin group (Figure 1, B). The groups with one active therapy had an intermediate number of events and were not statistically significantly different from the placebo-placebo group. The comparison of active infusion + active vitamins with placebo infusions + placebo vitamins was significant (HR 0.74, 95% CI 0.57–0.95, P = .016). The absolute difference in 5-year Kaplan-Meier estimated event rates between placebo-placebo and active-active groups was 8.3%, and the number needed to treat (NNT) to prevent 1 event over 5 years was 12.
(Enlarge Image)
Figure 1.
A, Kaplan-Meier curves (4 factorial groups, primary end point). B, Kaplan-Meier curves placebo/placebo versus active/active (primary end point).
The principal secondary end point, cardiovascular death, MI, or stroke, occurred in 58 (13%) of the placebo infusions + placebo vitamin group, 57 (14%) of the chelation + placebo vitamin group, 55 (13%) of the placebo infusion + active vitamin group, and 39 (9%) of patients in the chelation + high-dose vitamin group. The comparison of active infusion + active vitamins with placebo infusions + placebo vitamins favored chelation + vitamins (HR 0.66, 95% CI 0.44–0.99, P = .046). The groups with one active therapy had an intermediate number of events and were not statistically significantly different from the placebo-placebo group.
There were no significant differences in adherence to IV infusions or to oral vitamins between groups (Table III). Consent withdrawal at some point during the trial was reported in 289 patients. A greater frequency of consent withdrawals occurred among patients randomized to placebo infusions.
Serious adverse events were documented in 55 patients (13%) of the EDTA chelation and high-dose vitamin group, 45 (11%) of the EDTA chelation and placebo vitamin group, 69 (16%) of the placebo infusion and high-dose vitamin group, and 58 (13%) of the placebo infusion and placebo vitamin group (P = .170).
In the 633 patients with diabetes, the 5-year Kaplan-Meier event rate estimates for the primary end point was 29.1% in the chelation + high-dose vitamin group, 36.1% in the chelation + placebo vitamin group, 48.1% in the placebo infusion + active vitamin group, and 47.3% in the placebo infusions + placebo vitamin group (Figure 2, A). The primary end point by treatment group occurred in 56 (38%) of the placebo infusion + placebo vitamin group and 36 (23%) of patients in the chelation + high-dose vitamin group (HR 0.49, 95% CI 0.33–0.75, P < .001, 5-year NNT = 5.5) (Figure 2, B). The factorial groups receiving only one active treatment had intermediate treatment benefit not statistically significantly different from double placebo.
(Enlarge Image)
Figure 2.
A, Kaplan-Meier curves (4 factorial groups, primary end point, diabetes). B, Kaplan-Meier curves placebo/placebo versus active/active (primary end point, diabetes).
Results
Between September 10, 2003, and October 4, 2010, 1,708 patients were randomized: 421 to EDTA chelation infusions + high-dose oral multivitamins, 418 to EDTA chelation infusions + oral placebo, 432 to placebo infusions + high-dose oral multivitamins, and 437 to placebo infusions + oral placebo. The median duration of follow-up was 55 months (interquartile range, 26–60) overall. There was no significant difference in length of follow-up across all 4 groups.
Baseline Characteristics
Baseline characteristics were similar among the 4 randomized factorial groups (Table I). Patients were 65 (59–72) years old, 18% were female, and 9% were minorities. The qualifying MI had occurred 4.6 (1.6–9.2) years prior to enrollment. There was a high prevalence of diabetes (37%); prior coronary revascularizations (83%); and postinfarct, guideline-recommended medication use of aspirin (84%), β-blocker (72%), and statin (73%).
Factorial Treatment Comparisons (Overall Group)
The 5-year Kaplan-Meier event rate estimate for the primary end point was 31.9% in the chelation + high-dose vitamin group, 33.7% in the chelation + placebo vitamin group, 36.6% in the placebo infusion + active vitamin group, and 40.2% in the placebo infusions + placebo vitamin group (Figure 1, A; Table II). The primary end point by treatment group occurred in 139 (32%) of the placebo infusion + placebo vitamin group and 108 (26%) of patients in the chelation + high-dose vitamin group (Figure 1, B). The groups with one active therapy had an intermediate number of events and were not statistically significantly different from the placebo-placebo group. The comparison of active infusion + active vitamins with placebo infusions + placebo vitamins was significant (HR 0.74, 95% CI 0.57–0.95, P = .016). The absolute difference in 5-year Kaplan-Meier estimated event rates between placebo-placebo and active-active groups was 8.3%, and the number needed to treat (NNT) to prevent 1 event over 5 years was 12.
(Enlarge Image)
Figure 1.
A, Kaplan-Meier curves (4 factorial groups, primary end point). B, Kaplan-Meier curves placebo/placebo versus active/active (primary end point).
The principal secondary end point, cardiovascular death, MI, or stroke, occurred in 58 (13%) of the placebo infusions + placebo vitamin group, 57 (14%) of the chelation + placebo vitamin group, 55 (13%) of the placebo infusion + active vitamin group, and 39 (9%) of patients in the chelation + high-dose vitamin group. The comparison of active infusion + active vitamins with placebo infusions + placebo vitamins favored chelation + vitamins (HR 0.66, 95% CI 0.44–0.99, P = .046). The groups with one active therapy had an intermediate number of events and were not statistically significantly different from the placebo-placebo group.
Treatment Adherence
There were no significant differences in adherence to IV infusions or to oral vitamins between groups (Table III). Consent withdrawal at some point during the trial was reported in 289 patients. A greater frequency of consent withdrawals occurred among patients randomized to placebo infusions.
Safety
Serious adverse events were documented in 55 patients (13%) of the EDTA chelation and high-dose vitamin group, 45 (11%) of the EDTA chelation and placebo vitamin group, 69 (16%) of the placebo infusion and high-dose vitamin group, and 58 (13%) of the placebo infusion and placebo vitamin group (P = .170).
Diabetes Analyses
In the 633 patients with diabetes, the 5-year Kaplan-Meier event rate estimates for the primary end point was 29.1% in the chelation + high-dose vitamin group, 36.1% in the chelation + placebo vitamin group, 48.1% in the placebo infusion + active vitamin group, and 47.3% in the placebo infusions + placebo vitamin group (Figure 2, A). The primary end point by treatment group occurred in 56 (38%) of the placebo infusion + placebo vitamin group and 36 (23%) of patients in the chelation + high-dose vitamin group (HR 0.49, 95% CI 0.33–0.75, P < .001, 5-year NNT = 5.5) (Figure 2, B). The factorial groups receiving only one active treatment had intermediate treatment benefit not statistically significantly different from double placebo.
(Enlarge Image)
Figure 2.
A, Kaplan-Meier curves (4 factorial groups, primary end point, diabetes). B, Kaplan-Meier curves placebo/placebo versus active/active (primary end point, diabetes).
SHARE
