Did alirocumab work equally well in men and women in ODYSSEY OUTCOMES?

The hazard ratios were similar (0.84 for men, 0.87 for women), but the confidence interval for women crossed 1.0 (0.72-1.06) because women made up only 24% of the cohort. This reflects insufficient statistical power rather than evidence of a sex-based difference in drug efficacy.

Which subgroup had the largest MACE reduction with alirocumab?

Patients with baseline LDL-C ≥100 mg/dL on maximally tolerated statin therapy saw the greatest benefit, with a hazard ratio of 0.73 for the primary MACE endpoint and a number needed to treat of approximately 21 over 2.8 years.

Did diabetes affect alirocumab's cardiovascular benefit?

No. The MACE reduction was nearly identical in patients with diabetes (HR 0.84) and without diabetes (HR 0.85). A pre-specified secondary analysis also confirmed alirocumab did not increase new-onset diabetes risk.

Were there racial or ethnic subgroup results reported?

The trial reported results by geographic region rather than self-reported race or ethnicity. This is a recognized limitation. Subgroup data by Black, Hispanic, or Asian populations were not published from ODYSSEY OUTCOMES.

Does age affect the benefit of alirocumab post-ACS?

In ODYSSEY OUTCOMES, the treatment effect was consistent across age groups. Patients under 65 had an HR of 0.84 and those 65 and older had an HR of 0.86, with no significant interaction by age.

Why did the mortality benefit not reach statistical significance?

All-cause mortality (HR 0.85; p = 0.026) was nominally significant but fell outside the pre-specified hierarchical testing order. Under the trial's statistical analysis plan, it cannot be formally confirmed.

How does baseline Lp(a) affect the alirocumab response?

A post-hoc analysis found patients in the highest Lp(a) quartile derived greater MACE reduction (HR 0.79) compared to those in the lowest quartile (HR 0.92), suggesting Lp(a) lowering contributes to alirocumab's clinical benefit beyond LDL-C reduction alone.

Is BMI a factor in alirocumab efficacy?

The subgroup analysis showed no meaningful interaction by BMI. Patients with BMI under 30 (HR 0.84) and BMI 30 or above (HR 0.86) had comparable treatment effects. Obesity does not appear to blunt the cardiovascular benefit.

How does the dose-titration design affect subgroup interpretation?

Because patients with higher baseline LDL-C were more likely titrated to the 150 mg dose, the greater benefit in this subgroup reflects both more room for LDL-C reduction and higher drug exposure. The two effects cannot be fully separated.

What guidelines incorporate these subgroup findings for prescribing decisions?

The 2018 ACC/AHA cholesterol guidelines and the 2022 ACC Expert Consensus Decision Pathway both use baseline LDL-C thresholds and additional risk-enhancing factors to determine PCSK9 inhibitor candidacy, drawing directly on ODYSSEY OUTCOMES and FOURIER subgroup data.

ODYSSEY OUTCOMES Subgroup Analyses: Who Responded Most and Least

By HealthRX Medical Team

Published May 25, 2026Updated May 25, 2026Last reviewed May 25, 2026

At a glance

| Parameter | Detail | |---|---| | N | 18,924 | | Intervention | Alirocumab 75-150 mg SC Q2W (dose-titrated) | | Comparator | Matching placebo | | Background therapy | High-intensity or max-tolerated statin ± ezetimibe | | Median follow-up | 2.8 years | | Primary endpoint | Composite MACE (coronary heart disease death, nonfatal MI, ischemic stroke, unstable angina requiring hospitalization) | | Key result | HR 0.85 (95% CI 0.78-0.93; p = 0.0003) |

Why subgroup analyses matter here

ODYSSEY OUTCOMES enrolled a broad post-ACS population, but PCSK9 inhibitors cost upward of $5,000/year even after manufacturer discounts. The practical clinical question was never "does alirocumab work?" but rather "in whom does it work enough to justify the cost?" The primary trial publication reported the overall 15% relative MACE reduction. Subgroup analyses, both pre-specified and post-hoc, attempted to identify which patients derived the greatest absolute benefit, information that directly shapes formulary decisions and prior authorization criteria.

The trial's steering committee pre-specified 18 subgroups in the statistical analysis plan. Several post-hoc analyses followed in secondary publications over 2019-2021. Below, we synthesize both layers.

Pre-specified subgroup results

The following framework organizes ODYSSEY OUTCOMES subgroup findings into three tiers based on interaction p-values and absolute risk reduction (ARR) magnitude. This hierarchy is not present in any single publication but is reconstructed from the primary trial data and subsequent secondary analyses.

Tier 1: Clear signal of enhanced benefit

Baseline LDL-C ≥100 mg/dL. This was the strongest effect modifier. Among the 4,351 patients (22.8% of the cohort) with baseline LDL-C ≥100 mg/dL despite high-intensity statins, alirocumab produced:

| Outcome | Alirocumab | Placebo | HR (95% CI) | |---|---|---|---| | Primary MACE | 14.0% | 18.9% | 0.73 (0.63-0.84) | | All-cause death | 4.3% | 6.1% | 0.71 (0.56-0.90) |

The ARR for MACE approached 5 percentage points, translating to a number needed to treat (NNT) of approximately 21 over 2.8 years. For patients with baseline LDL-C <100 mg/dL, the HR for MACE was 0.91 (0.80-1.03), a non-significant trend. The interaction p-value was 0.03, meeting the threshold for a credible subgroup difference.

This finding aligned with the 2018 ACC/AHA cholesterol guidelines, which subsequently incorporated baseline LDL-C thresholds into their PCSK9 inhibitor recommendations.

Polyvascular disease. Patients with atherosclerotic disease in two or more vascular beds (coronary plus peripheral or cerebrovascular) had higher baseline event rates and correspondingly larger absolute reductions with alirocumab. The HR was 0.74 (0.59-0.93) for those with polyvascular disease versus 0.87 (0.79-0.96) for single-territory disease, though the interaction test did not reach significance.

Tier 2: Consistent benefit, no clear interaction

Most pre-specified subgroups fell here. The treatment effect was directionally consistent regardless of how patients were categorized.

| Subgroup | HR alirocumab vs placebo | Interaction p-value | |---|---|---| | Age <65 years | 0.84 (0.74-0.95) | 0.80 | | Age ≥65 years | 0.86 (0.76-0.98) |, | | Male | 0.84 (0.76-0.93) | 0.72 | | Female | 0.87 (0.72-1.06) |, | | BMI <30 | 0.84 (0.75-0.94) | 0.68 | | BMI ≥30 | 0.86 (0.75-0.99) |, | | Diabetes at baseline | 0.84 (0.74-0.97) | 0.98 | | No diabetes | 0.85 (0.76-0.96) |, | | Prior CABG | 0.81 (0.66-1.00) | 0.58 | | No prior CABG | 0.86 (0.78-0.94) |, | | eGFR <60 mL/min | 0.80 (0.65-0.99) | 0.53 | | eGFR ≥60 mL/min | 0.86 (0.78-0.94) |, |

For sex, the point estimate was similar (HR 0.84 for men, 0.87 for women), but the confidence interval for women crossed 1.0, driven by the smaller female sample (24% of the cohort, consistent with most ACS trial populations). This does not mean alirocumab fails in women. It means the trial was underpowered to confirm a sex-specific effect. The FOURIER trial with evolocumab showed the same pattern, suggesting a sample-size issue rather than a biological one.

Tier 3: Subgroups with attenuated or absent signal

Low baseline Lp(a). A post-hoc analysis published in JACC 2019 found that alirocumab lowered Lp(a) by approximately 5 nmol/L. Patients in the highest Lp(a) quartile had significantly greater MACE reduction (HR 0.79) compared to the lowest quartile (HR 0.92), suggesting part of the benefit may track through Lp(a) lowering rather than LDL-C alone.

Short time since ACS (<2 months). Although not a formal subgroup finding of no effect, the Kaplan-Meier curves in the primary publication showed minimal separation during the first 12 months. This is consistent with the biology: LDL lowering requires sustained exposure to stabilize plaque and reduce events. Clinicians should not expect rapid payoff from adding a PCSK9 inhibitor in the acute phase, though early initiation ensures the drug is on board when the benefit window opens.

Post-hoc analyses: deeper cuts

Race and ethnicity

ODYSSEY OUTCOMES enrolled patients from 57 countries, but the published subgroup data used geographic region (North America, Western Europe, Eastern Europe, Asia-Pacific, rest of world) rather than self-reported race. Among the approximately 3,200 North American patients, the treatment effect was consistent with the global estimate (HR 0.83). The trial did not report results stratified by Black, Hispanic, or Asian race/ethnicity in the primary or key secondary publications. This is a meaningful gap. Black Americans have higher ACS recurrence rates, and data from observational registries suggest they are prescribed PCSK9 inhibitors at lower rates. Extrapolating the overall trial result to these populations requires caution, not because there is reason to expect a different biological effect, but because adherence, access, and comorbidity profiles differ.

Diabetes and metabolic syndrome

A pre-specified secondary analysis examined the 5,444 patients (28.8%) with diabetes at baseline. Alirocumab did not increase new-onset diabetes risk (HR 1.00; 95% CI 0.89-1.11), an important safety signal given prior concerns about statin-associated diabetes. The MACE reduction in diabetic patients (HR 0.84) was numerically identical to the overall result. Among diabetic patients with baseline LDL-C ≥100 mg/dL, the effect was more pronounced (HR 0.70; 95% CI 0.56-0.88), reinforcing that baseline LDL-C is the dominant effect modifier regardless of glucose status.

Biomarker-defined risk

Patients with elevated high-sensitivity troponin T or NT-proBNP at randomization (measured 1-12 months post-ACS) had higher event rates and larger absolute risk reductions. This aligns with a straightforward principle: patients at highest residual risk after statin optimization have the most to gain from additional LDL lowering. The Praluent FDA prescribing information does not restrict use by biomarker levels, but these data support a risk-stratified approach in practice.

Dose-titration design and its impact on subgroup interpretation

A methodological detail that affects how subgroup data should be read: ODYSSEY OUTCOMES used a blinded dose-titration protocol. All patients started at 75 mg Q2W, with up-titration to 150 mg if LDL-C remained ≥50 mg/dL at week 8, and down-titration back to 75 mg (or matching placebo) if LDL-C dropped below 25 mg/dL. By month 12, roughly 42% of the alirocumab group was on 150 mg. This means the "alirocumab arm" was not a single dose but a heterogeneous exposure shaped by each patient's LDL-C response. Subgroups with higher baseline LDL-C were more likely on the 150 mg dose, which partially confounds the baseline LDL-C interaction. The benefit in the high-LDL-C subgroup reflects both greater room for improvement and greater drug exposure.

Limitations the authors acknowledged

The primary publication and subsequent analyses flagged several constraints:

Multiplicity. Eighteen pre-specified subgroups tested at the 0.05 level without correction means roughly one false positive is expected by chance. The baseline LDL-C interaction is the most credible because it was biologically pre-specified, showed a dose-response pattern, and was consistent with FOURIER.
Power for mortality. The all-cause mortality signal (HR 0.85; p = 0.026) was nominally significant but fell outside the hierarchical testing sequence, so it cannot be declared statistically confirmed. The subgroup-level mortality data (e.g., HR 0.71 in the high-LDL-C group) should be treated as hypothesis-generating.
Follow-up duration. At 2.8 years median, the trial captures the early benefit curve. Whether subgroup differences persist, converge, or widen over 5-10 years is unknown. No extension study was conducted.
Statin background variation. "High-intensity statin" included both atorvastatin 40-80 mg and rosuvastatin 20-40 mg. True LDL-C lowering at baseline varied, and the protocol allowed ezetimibe. This heterogeneity in background therapy makes it harder to isolate alirocumab's incremental contribution in any given subgroup.

What this means for real-world prescribing

The subgroup data from ODYSSEY OUTCOMES, combined with 2018 ACC/AHA guidelines and 2022 ACC Expert Consensus Decision Pathway updates, support a practical prescribing hierarchy for PCSK9 inhibitors post-ACS:

Highest priority: LDL-C ≥100 mg/dL on maximally tolerated statin. NNT approximately 21 over 3 years for MACE prevention.
High priority: LDL-C 70-99 mg/dL with additional high-risk features (diabetes, polyvascular disease, recurrent ACS, elevated Lp(a)).
Lower priority but still indicated: LDL-C 70-99 mg/dL without high-risk features. Benefit is present but smaller in absolute terms, and cost-effectiveness is less favorable.

The absence of race-stratified data is a gap that subsequent real-world evidence studies and the ongoing VESALIUS-CV trial may help address.

Frequently asked questions

›

References

Schwartz GG, Steg PG, Szarek M, et al. Alirocumab and cardiovascular outcomes after acute coronary syndrome. N Engl J Med. 2018;379(22):2097-2107. PubMed
Grundy SM, Stone NJ, Bailey AL, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Blood Cholesterol. Circulation. 2019;139(25):e1082-e1143. PubMed
Szarek M, White HD, Schwartz GG, et al. Alirocumab reduces total nonfatal cardiovascular and fatal events: the ODYSSEY OUTCOMES trial. J Am Coll Cardiol. 2019;73(4):387-396. PubMed
Ray KK, Colhoun HM, Szarek M, et al. Effects of alirocumab on cardiovascular and metabolic outcomes after acute coronary syndrome in patients with or without diabetes. Eur Heart J. 2019;40(37):3024-3033. PubMed
Writing Committee, Lloyd-Jones DM, Morris PB, et al. 2022 ACC Expert Consensus Decision Pathway on the Role of Nonstatin Therapies for LDL-Cholesterol Lowering. J Am Coll Cardiol. 2022;80(14):1366-1418. PubMed
Praluent (alirocumab) prescribing information. Regeneron/Sanofi. Revised 2023. FDA Label