N Engl J Med 2019;380:752-762
Background: Growing body of evidence suggests the role of inflammation in atherosclerosis. The CANTOS trial, discussed earlier, provided further evidence for that hypothesis. In the CANTOS trial, canakinumab a monoclonal antibody that neutralizes interleukin-1β led to fewer myocardial infarctions compared to placebo.
Cardiology Trial’s Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
The Cardiovascular Inflammation Reduction Trial (CIRT) sought to test the hypothesis that an alternative approach to reduce inflammation using low-dose methotrexate, also lowers cardiovascular events compared to placebo. Low-dose methotrexate was chosen since it’s widely used, inexpensive and was shown, in observational studies, to lower cardiovascular events in patients with rheumatoid arthritis and psoriatic arthritis compared to other drugs or placebo.
Patients: Patients were enrolled if they had history of myocardial infarction or multivessel coronary artery disease and had either type II diabetes mellitus or metabolic syndrome. Patients had to be medically stable and had completed any planned revascularization. Exclusion criteria were many and included patients with chronic liver disease, chronic inflammatory conditions, chronic infectious diseases, HIV positive, chronic use of immunosuppressive therapy or women of childbearing potential.
Baseline characteristics: The trial randomized 2,391 patients to the low-dose methotrexate arm and 2,395 patients to the placebo arm. The average age of patients was 66 years with 81% being men. The median body mass index was 31 kg/m2. About 61% had prior myocardial infarction, 34% had diabetes, 32% had metabolic syndrome and 13% had congestive heart failure. The median total cholesterol was 141 mg/dL, median LDL was 68 mg/dL and the median hs-CRP was 1.5 mg/L.
Procedures: The study had a run-in phase in which patients were administered 1 mg of oral folic acid daily in addition to oral methotrexate once weekly, initially at 5 mg, then increased to 10 mg, and finally to 15 mg. This phase was conducted in an open-label manner, and patients were excluded from further participation if they experienced adverse events, including laboratory abnormalities related to the treatment.
Patients who were able to tolerate the 15 mg dose of methotrexate for 2 consecutive weeks without adverse events, were randomized in 1:1 fashion to receive methotrexate at a dose of 15 mg or placebo in a double-blinded fashion. All patients took folate daily. At 4 months, the dose of methotrexate was increased to 20 mg. Symptoms and laboratory variables were assessed every 2 months and the dose of methotrexate, or placebo, was adjusted as needed. Randomization was stratified based on the type of index event, time since the index event and metabolic syndrome alone or diabetes.
Endpoints: Initially, the primary endpoint of the trial was a composite endpoint of the first occurrence of nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death. Approximately 5 years into the trial, the composite end point was expanded to include hospitalization for unstable angina that led to urgent coronary revascularization. This increased the statistical power and allowed for smaller sample size. Death from any cause was included as a secondary endpoint.
Statistical analysis was performed based on the intention-to-treat principle. Based on the new composite primary endpoint, it was estimated that a total of 5,500 patients and 634 primary endpoint events would give the trial 90% power to detect 23% relative reduction of the new composite primary endpoint, with methotrexate compared to placebo. The data safety and monitoring board recommended early termination of the trial for futility.
Results: Out of 6,158 patients who entered the open label run-in phase, 997 (16.2%) patients were excluded due to symptoms or adverse events, 129 (2.1%) patients were excluded due to laboratory values out of range, and 4,786 (77.7%) patients were randomized.
At 8 months, methotrexate compared to placebo, did not result in greater reduction in hs-CRP, interleukin-1β, or interleukin-6.
After a median follow-up time of 2.3 years, the rate of the final composite primary endpoint was similar in both treatment groups [8.4% (4.13%/ year) with methotrexate vs 8.6% (4.31%/ year) with placebo, HR: 0.96, 95% CI: 0.79 – 1.16; p= 0.67]. There was no significant difference in death from any cause [4.0% (1.80%/ year) vs 3.5% (1.55%/ year)], cardiovascular death [2.0% (0.92%/ year) vs 1.8% (0.80%/ year)], nonfatal myocardial infarction [4.7% (2.29%/ year) vs 4.8% (2.32%/ year)], nonfatal stroke [1.2% (0.55%/ year) vs 1.3% (0.60%/ year)], or hospitalizations for unstable angina that led to urgent revascularization [1.7% (0.81%/ year) vs 2.1% (1.01%/ year)].
Mouth sores and oral pain were more common with methotrexate [4.0% (1.95%/ year) vs 2.3% (1.13%/ year); p= 0.001]. Non-basal-cell skin cancer was also more common with methotrexate [1.3% (0.61%/ year) vs 0.4% (0.20%/ year); p= 0.002]. Serious infections [4.6% (2.24%/ year) vs 5.1% (2.47%/ year); p= 0.50] and serious bleeding [1.3% (0.63%/ year) vs 1.0% (0.50%/ year); p= 0.44] were not significantly different between both treatment groups.
Data on subgroup analysis was not provided in the main paper or supplement.
Conclusion: In patients with history of myocardial infarction or multivessel coronary artery disease, low-dose methotrexate did not reduce cardiovascular events compared to placebo.
The authors proposed several reasons why the results of CIRT contrast the results of CANTOS in which canakinumab, a monoclonal antibody targeting interleukin-1β was used. First, CANTOS limited enrollment to patients with elevated hs-CRP while CIRT did not require that. Consequently, the median hs-CRP was significantly higher in CANTOS at 4.2 mg/L compared to 1.5 mg/L in CIRT. Furthermore, canakinumab in CANTOS lowered interleukin-6 and hs-CRP levels by 35-40% compared to placebo, while methotrexate in CIRT did not lower interleukin-6, hs-CRP or interleukin-1β. This suggests that the effectiveness of reducing atherosclerosis risk may vary depending on the specific pathway targeted. Other notable findings are higher LDL levels in CANTOS compared to CIRT (82 mg/ dL vs 68 mg/ dL). In addition, all patients had history of myocardial infarction in CANTOS compared to 61% in CIRT.
Another important teaching point from this trial is the use of a run-in period with the active drug, methotrexate. Although employing a run-in period is beneficial for assessing adherence and eliminating early dropouts, it can introduce selection bias. This can occur by excluding patients who experience adverse events from the study drug, methotrexate in this case, and thereby favor the balance of benefits and harms in favor of the study drug.
Cardiology Trial’s Substack is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
Get full access to Cardiology Trial’s Substack at
cardiologytrials.substack.com/subscribe
