Compounds with a graded row

Ordered by rung, strongest first. The rung rates the study design; the chip beside it says what the study measured. Both are needed.

1. Dulaglutide

   human RCT Direct outcome Approved for this condition

   Who was studied

   adults aged 50 or over with type 2 diabetes, most without prior cardiovascular events, n=9901, median 5.4 years (REWIND, subcutaneous dulaglutide 1.5 mg weekly)

   REWIND met superiority on its primary composite of non-fatal myocardial infarction, non-fatal stroke or cardiovascular death: hazard ratio 0.88 (95% CI 0.79-0.99), p=0.026, with events in 12.0 percent against 13.4 percent on placebo. It is the one trial in this group whose population was mostly primary prevention rather than established cardiovascular disease, which is the reason its absolute event reduction is small even though the relative result reached superiority.

   Source PubMed 1 primary source read for this row

2. Exenatide

   human RCT Direct outcome Not approved for this condition

   Who was studied

   adults with type 2 diabetes, 73.1 percent with previous cardiovascular disease, n=14752, median 3.2 years (EXSCEL, exenatide extended-release 2 mg weekly; the corpus record's approved brand BYETTA is the twice-daily formulation of the same peptide)

   EXSCEL met non-inferiority and did not meet superiority. The primary composite of cardiovascular death, non-fatal myocardial infarction or non-fatal stroke gave a hazard ratio of 0.91 (95% CI 0.83-1.00), p<0.001 for non-inferiority but p=0.06 for superiority. The finding is that exenatide did not increase cardiovascular risk, which is a different claim from reducing cardiovascular events; no cardiovascular indication is on the label.

   Source PubMed 1 primary source read for this row

3. Liraglutide

   human RCT Direct outcome Approved for this condition

   Who was studied

   adults with type 2 diabetes at high cardiovascular risk, n=9340, median 3.8 years (LEADER, subcutaneous liraglutide up to 1.8 mg daily)

   LEADER was a regulator-mandated non-inferiority safety trial that went on to meet superiority on its primary composite of cardiovascular death, non-fatal myocardial infarction or non-fatal stroke: hazard ratio 0.87 (95% CI 0.78-0.97), p<0.001 for non-inferiority and p=0.01 for superiority, with events in 13.0 percent on liraglutide against 14.9 percent on placebo. The population was people with type 2 diabetes and established cardiovascular disease or high risk; there is no trial of this size in people without diabetes on file.

   Source PubMed 1 primary source read for this row

4. Lixisenatide

   human RCT Direct outcome Not approved for this condition

   Who was studied

   adults with type 2 diabetes and a myocardial infarction or hospitalisation for unstable angina within the previous 180 days, n=6068, median 25 months (ELIXA, subcutaneous lixisenatide daily)

   ELIXA met non-inferiority and showed no benefit. The primary composite was four-point: cardiovascular death, myocardial infarction, stroke or hospitalisation for unstable angina, and the hazard ratio was 1.02 (95% CI 0.89-1.17) against a pre-specified non-inferiority margin of 1.3. The trial establishes that lixisenatide did not increase cardiovascular events after an acute coronary syndrome; it does not show any reduction in them.

   Source PubMed 1 primary source read for this row

5. Semaglutide

   human RCT Direct outcome Approved for this condition

   Who was studied

   adults with type 2 diabetes, n=3297, 104 weeks (SUSTAIN-6, subcutaneous semaglutide 0.5 or 1.0 mg weekly); and adults aged 45 or over with established cardiovascular disease and BMI 27 or above but no diabetes, n=17604, mean 39.8 months (SELECT, semaglutide 2.4 mg weekly)

   SELECT was a superiority trial and met it: the primary composite of cardiovascular death, non-fatal myocardial infarction or non-fatal stroke occurred less often on semaglutide, hazard ratio 0.80 (95% CI 0.72-0.90, p<0.001), in people with cardiovascular disease and overweight or obesity but no diabetes. The earlier SUSTAIN-6 was designed and powered as a non-inferiority safety trial in type 2 diabetes and met non-inferiority (p<0.001); its hazard ratio of 0.74 (95% CI 0.58-0.95) also crossed the superiority threshold, but that test was not what the trial was sized for. FDA has since added a cardiovascular indication to the semaglutide 2.4 mg label on the strength of SELECT.

   Source PubMed 2 primary sources read for this row

6. Testosterone

   human RCT Direct outcome Not approved for this condition

   Who was studied

   men aged 45-80 with hypogonadism (testosterone below 300 ng/dL) and pre-existing or high-risk cardiovascular disease, n=5246, mean treatment 21.7 months and mean follow-up 33.0 months (TRAVERSE, transdermal testosterone gel)

   TRAVERSE was a cardiovascular SAFETY trial, designed and powered for non-inferiority with an upper limit of 1.5, and it met that: hazard ratio 0.96 (95% CI 0.78-1.17) for the composite of cardiovascular death, non-fatal myocardial infarction or non-fatal stroke. This is a finding that testosterone replacement in hypogonadal men did not raise cardiovascular event rates. It is not evidence that testosterone treats or prevents cardiovascular disease, and no trial testing that question is on file. Testosterone is approved for replacement in male hypogonadism only, and carries no cardiovascular indication.

   Source PubMed 1 primary source read for this row

7. Tirzepatide

   human RCT Direct outcome Not approved for this condition

   Who was studied

   adults with type 2 diabetes and atherosclerotic cardiovascular disease, mean age 64.1 years, 29 percent women, n=13299 randomised (6586 tirzepatide, 6579 dulaglutide in the modified intention-to-treat analysis), median treatment about 46.9 months (SURPASS-CVOT)

   SURPASS-CVOT was an active-comparator non-inferiority trial against dulaglutide, not a placebo trial. Tirzepatide met non-inferiority on the primary composite of cardiovascular death, myocardial infarction or stroke, hazard ratio 0.92 (95.3% CI 0.83-1.01) against a margin of 1.05, p=0.003 for non-inferiority; superiority over dulaglutide was not met, p=0.09. No placebo-controlled cardiovascular outcome trial of tirzepatide is on file, and there is no cardiovascular indication on the label.

   Source PubMed 1 primary source read for this row

8. Apelin

   human RCT Surrogate marker Not approved for this condition

   Who was studied

   18 patients with NYHA class II-III chronic heart failure, 6 patients undergoing diagnostic coronary angiography, and 26 healthy volunteers; single-session acute studies. The peptides infused were (Pyr1)apelin-13 (intrabrachial, and systemic 30-300 nmol/min) and apelin-36 (intracoronary bolus), not apelin generically

   These were randomised, double-blind, placebo-controlled acute physiology studies measuring forearm blood flow by plethysmography, coronary blood flow by Doppler wire, left ventricular pressure by pressure wire, and cardiac output by bioimpedance. They record a haemodynamic response over minutes to hours in fewer than 20 heart failure patients: vasodilation, raised cardiac index, lowered mean arterial pressure. No study measured any cardiovascular outcome, symptom or event, and no trial of chronic dosing in heart failure is on file.

   Source PubMed 1 primary source read for this row

9. Estradiol

   human RCT Surrogate marker Not approved for this condition

   Who was studied

   healthy postmenopausal women, n=643, median 5 years, randomised to oral 17-beta-estradiol 1 mg daily or placebo and stratified by time since menopause: early (under 6 years) or late (10 years or more) (ELITE)

   ELITE measured the rate of change in carotid-artery intima-media thickness, an imaging surrogate, not cardiovascular events. Progression was slower on estradiol in the early-postmenopause stratum (0.0044 vs 0.0078 mm/year, p=0.008) but not in the late stratum (0.0100 vs 0.0088 mm/year, p=0.29), interaction p=0.007 - the timing hypothesis. The cardiac CT endpoints pointed the other way: coronary-artery calcium, total stenosis and plaque did not differ from placebo in either stratum. No cardiovascular event outcome was measured, estradiol has no cardiovascular indication, and its label carries a boxed warning covering cardiovascular disorders.

   Source PubMed 1 primary source read for this row

10. Adrenomedullin

    human case series Surrogate marker Not approved for this condition

    Who was studied

    7 patients with congestive heart failure and 7 healthy subjects given intravenous human adrenomedullin at 0.05 microg/kg/min in a single acute session, with a further 7 heart failure and 6 healthy subjects on placebo. The abstract does not describe randomisation or blinding

    The measurements were acute haemodynamic, renal and hormonal: mean arterial pressure, heart rate, cardiac index, pulmonary capillary wedge pressure, urine volume, urinary sodium and plasma aldosterone, over the course of one infusion in 7 heart failure patients. In those patients the blood-pressure fall was smaller than in healthy subjects (about 8 mmHg) and heart rate rose about 5 bpm. No study measured a cardiovascular outcome, symptom or event, and no chronic dosing trial is on file.

    Source PubMed 1 primary source read for this row

11. Angiotensin (1-7)

    human case series Mechanistic only Not approved for this condition

    Who was studied

    8 patients with heart failure already treated with an ACE inhibitor; angiotensin-(1-7) infused into the brachial artery, alone and with bradykinin, in a single acute session. The abstract does not describe randomisation or blinding

    Recorded, and not evidence for this condition

    What was measured was forearm blood flow by venous occlusion plethysmography in 8 people - a local vascular response, not a cardiovascular outcome. Angiotensin-(1-7) produced no vasodilation on its own and did not alter the response to bradykinin; the authors concluded it was biologically inactive in this circulation. No study measured cardiovascular disease itself - no events, no symptoms, no cardiac function endpoint - so the link to the condition is an argument from mechanism rather than a result.

    Source PubMed 1 primary source read for this row