23. Statins in Sepsis & ARDS (HARP-2 and SAILS)

Quick Recap

Cross-cutting protocol — a deliberately concise entry addressing a mechanistically plausible therapy that has been tested twice, definitively, and found not to work — with a genuinely interesting mechanistic wrinkle underneath the shared negative headline: the two statins tested (simvastatin, lipophilic; rosuvastatin, hydrophilic) appear to diverge, not converge, when examined more closely in specific subpopulations, including a signal of harm with one agent that the other does not share.

1. Definition

Statins in sepsis/ARDS: the proposed use of HMG-CoA reductase inhibitors — beyond their established lipid-lowering indication — as an anti-inflammatory, "pleiotropic-effect" adjunctive therapy for sepsis and ARDS, based on preclinical and observational data suggesting statins modulate inflammatory and endothelial pathways relevant to these conditions' pathophysiology, independent of their cholesterol-lowering action.

Two large, definitive RCTs, testing two different agents: HARP-2 (simvastatin, a lipophilic statin) and SAILS (rosuvastatin, a hydrophilic statin) — both large, multicenter, placebo-controlled trials specifically designed to test this hypothesis rigorously, following years of preclinical enthusiasm and smaller observational signals.

2. Pathophysiology

Statins possess well-documented "pleiotropic" effects beyond HMG-CoA reductase-mediated cholesterol lowering — anti-inflammatory and anti-proliferative properties that modify pathways plausibly relevant to ARDS pathogenesis (endothelial dysfunction, inflammatory cascade activation) — providing genuine, preclinically-supported biological rationale for testing this drug class in a disease process where cholesterol-lowering itself is not the therapeutic goal.

A specific, genuinely important mechanistic wrinkle distinguishing the two tested agents: rosuvastatin (hydrophilic) has been specifically found, in post hoc SAILS analysis, to be independently associated with a rise in plasma IL-18 (an inflammatory marker itself positively associated with mortality), and a murine acute lung injury model demonstrated that pravastatin (a hydrophilic statin similar to rosuvastatin) increased pulmonary IL-1β and IL-18 and enhanced macrophage inflammasome activity, dependent on mitochondrial reactive oxygen species generation — a concrete, biologically specific mechanism by which a hydrophilic statin specifically might paradoxically worsen inflammation in this context, distinct from the lipophilic simvastatin tested in HARP-2. This is not merely "statins don't work" but a more nuanced, mechanistically grounded finding that different statins within the same drug class may have genuinely divergent, even opposite, effects in critically ill patients — worth understanding as a specific pharmacological lesson, not simply folding both agents into an undifferentiated "statins failed" conclusion.

3. Immediate Stabilization (ABCDE) — Not Applicable

This protocol addresses a specific adjunctive therapy question rather than acute stabilization; cross-reference ARDS and Sepsis Bundles & Source Control protocols for the underlying management framework.

4. Focused History — Not a Primary Component

Relevant history is limited to prior statin therapy for an established cardiovascular indication (a distinct question from initiating statins for sepsis/ARDS — continuation of a pre-existing, indicated statin during critical illness is a separate clinical question this protocol does not address) and baseline cholesterol status, relevant to the subgroup findings discussed in Section 11.

5–10. [Not Applicable / Not a Primary Component for This Focused, Deliberately Concise Protocol]

Given this protocol's narrow scope — a specific, largely resolved "this mechanistically plausible therapy does not work as hypothesized" question — the standard examination, syndrome-identification, differential-diagnosis, severity-scoring, and investigation sections are not separately applicable in the way they are for disease-specific or diagnostic protocols elsewhere in this library. Cross-reference ARDS and Sepsis Bundles & Source Control protocols for the relevant underlying disease management framework.

11. Evidence-Based Management — Two Definitive Negative Trials, With a Genuinely Interesting Subgroup Divergence

HARP-2 (2014) — Simvastatin, Lipophilic

• HARP-2 trial (McAuley et al., NEJM 2014): a multicenter RCT across 40 ICUs in the UK and Ireland, n=540, enrolling intubated, ventilated ARDS patients (any cause, within 48 hours of ARDS onset) — randomized to simvastatin 80 mg daily vs. placebo for up to 28 days
• Primary outcome (ventilator-free days to day 28): no significant difference — 12.6±9.9 days (simvastatin) vs. 11.5±10.4 days (placebo), mean difference 1.1 days (95% CI −0.6 to 2.8, p=0.21)
• 28-day mortality: no significant difference (RR 0.8, 95% CI 0.6–1.1)
• A genuine methodological strength specifically noted in independent appraisal: the trial's heterogeneous population (all ARDS causes, not restricted to sepsis-related ARDS) was specifically highlighted as enhancing generalizability of the negative result

SAILS (2014) — Rosuvastatin, Hydrophilic

• SAILS trial: a multicenter RCT across 44 hospitals, n=745, enrolling patients with sepsis-related ARDS specifically (a narrower population than HARP-2's all-cause ARDS) — randomized to rosuvastatin 40 mg loading dose then 20 mg daily vs. placebo for up to 28 days
• Primary outcome: in-hospital mortality or mortality to study day 60 — no significant benefit, consistent with HARP-2's overall negative finding

The Concordant Headline Conclusion

• Independent commentary reviewing both trials together concluded directly: "the findings of this study, taken along with the results from the SAILS trial, suggest that there is little value in using statins to treat ARDS" — a clear, converging negative conclusion across two large, independently conducted, methodologically rigorous trials testing two different agents in overlapping but not identical populations

The Genuinely Interesting Divergence Underneath the Shared Negative Headline

• Inflammatory subphenotype re-analysis of HARP-2: a post hoc analysis using biomarker-based phenotyping (sTNFR1, IL-6, platelet counts, vasopressor use — the same "hyper-inflammatory vs. hypo-inflammatory" ARDS subphenotyping framework independently derived from re-analysis of the ARMA, ALVEOLI, and FACTT trials) found that simvastatin improved 28-day survival specifically in the "hyper-inflammatory" phenotype — an absolute risk reduction of 13%, compared to essentially no effect (0.01% AR difference) in the "hypo-inflammatory" phenotype — though this ventilator-free-days difference, while clinically meaningful in magnitude, did not reach statistical significance in this subgroup analysis, and represents hypothesis-generating, not confirmatory, evidence
• A specific, striking cholesterol-level secondary analysis, examining SAILS and HARP-2 together: patients with low baseline cholesterol (lowest quartile: <69 mg/dL in SAILS, <44 mg/dL in HARP-2) were sicker at enrollment (higher APACHE III and shock prevalence in SAILS; higher SOFA score and vasopressor use in HARP-2) — and critically, the effect of statin therapy diverged by agent within this specific, low-cholesterol, sicker subgroup:
• In SAILS (rosuvastatin): low-cholesterol patients receiving rosuvastatin were more likely to die — odds ratio 2.23 (95% CI 1.06–4.77, p=0.02; interaction p=0.02) — a real, statistically significant harm signal specific to this hydrophilic statin in this specific, sicker, low-cholesterol subgroup
• In HARP-2 (simvastatin): low-cholesterol patients receiving simvastatin showed a trend toward lower mortality — odds ratio 0.44 (95% CI 0.17–1.07, p=0.06; interaction p=0.22) — not reaching statistical significance in this smaller cohort, but pointing in the opposite direction from rosuvastatin's harm signal in the analogous subgroup
• The secondary analysis's own, carefully stated conclusion: "despite the very low levels of cholesterol, simvastatin therapy seems safe and may reduce mortality in this group, though rosuvastatin was associated with harm" — a genuinely important, agent-specific divergence that a simple "statins don't work in ARDS" summary would completely obscure

Practical Synthesis

Both HARP-2 and SAILS, as large, rigorous, primary RCTs, found no overall benefit from statin therapy in ARDS — this remains the clear, definitive, headline conclusion, and neither simvastatin nor rosuvastatin should be initiated as a routine ARDS or sepsis-specific adjunctive therapy. However, this protocol treats the underlying subgroup and mechanistic data as genuinely worth understanding, not merely a footnote to a settled negative result: the two tested statins appear to behave differently, not identically, in specific patient subpopulations — rosuvastatin (hydrophilic) showed a real, statistically significant harm signal in sicker, low-cholesterol patients, plausibly mediated by IL-18/inflammasome-related mechanisms distinct from simvastatin's (lipophilic) profile, while simvastatin showed a non-significant trend toward benefit in an analogous low-cholesterol subgroup and a more substantial, though still not definitively confirmed, benefit signal in the hyper-inflammatory ARDS subphenotype specifically. This is a genuinely instructive example of a broader principle relevant across critical care pharmacology: a negative overall RCT result for a drug class does not necessarily mean every agent within that class behaves identically, and cholesterol/inflammatory-phenotype-based patient stratification may be a more clinically meaningful way to think about this specific evidence base than a blanket "statins failed" conclusion — though none of these subgroup findings have been independently, prospectively confirmed as a basis for targeted clinical use, and should be understood as hypothesis-generating rather than practice-changing.

12–20. [Organ Support / Therapy / Consultation / Monitoring / Bundle / Complications / Escalation / Discharge / Documentation — Not Separately Applicable]

Given this protocol's focused, negative-evidence scope, these sections are addressed collectively rather than separately: current evidence does not support initiating statin therapy specifically for sepsis or ARDS treatment. Patients already on a statin for an established cardiovascular indication should generally continue it per standard critical illness medication-continuation principles (a distinct question from initiating statins for sepsis/ARDS, not addressed by this protocol's evidence base). No specific complications, monitoring framework, or escalation pathway is applicable to a therapy that is not currently recommended for initiation in this context; the one specific, quantified risk worth documenting if rosuvastatin is being continued in a critically ill patient with low baseline cholesterol for an unrelated indication is the harm signal identified in SAILS's secondary analysis (Section 11), which may warrant closer monitoring or reconsideration in that specific clinical scenario.

21. Key Guidelines

• No major sepsis or ARDS guideline recommends initiating statin therapy as an adjunctive treatment, consistent with both HARP-2 and SAILS's clear negative primary findings

22. Landmark Trials

23. Controversies

• The central, genuinely instructive point this protocol emphasizes: "statins don't work in ARDS" is an accurate summary of both trials' primary, definitive findings, but it obscures a real, mechanistically grounded divergence between the two specific agents tested — rosuvastatin's hydrophilic-statin-specific harm signal in sicker, low-cholesterol patients (with a plausible IL-18/inflammasome mechanism) is a materially different finding from simvastatin's neutral-to-potentially-favorable signal in the same clinical scenario, and collapsing both into an undifferentiated class conclusion loses genuinely useful information.
• None of the subgroup/phenotype-specific signals (hyper-inflammatory phenotype benefit with simvastatin; low-cholesterol harm with rosuvastatin) have been independently, prospectively confirmed — both remain hypothesis-generating post hoc findings from trials that were, in their primary, pre-specified analyses, clearly negative; this protocol presents them as genuinely interesting, biologically plausible signals worth future investigation, not as a basis for current targeted clinical practice.
• Statin continuation (for patients already on therapy for an established indication) during critical illness is a conceptually distinct question from the initiation-for-treatment question this protocol addresses — this protocol's negative findings should not be conflated with, or used to argue against, standard practice around continuing a pre-existing, indicated statin during an ICU admission.

24. References

1. National Heart, Lung, and Blood Institute ARDS Clinical Trials Network. Rosuvastatin for sepsis-associated acute respiratory distress syndrome (SAILS). N Engl J Med. 2014;370(23):2191-2200.
2. McAuley DF, Laffey JG, O'Kane CM, et al; HARP-2 Investigators. Simvastatin in the acute respiratory distress syndrome (HARP-2). N Engl J Med. 2014;371(18):1695-1703.
3. Calfee CS, Delucchi K, Sinha P, et al. Acute respiratory distress syndrome subphenotypes and differential response to simvastatin: secondary analysis of a randomised controlled trial. Lancet Respir Med. 2018;6(9):691-698.
4. Effect of total cholesterol and statin therapy on mortality in ARDS patients: a secondary analysis of the SAILS and HARP-2 trials. Crit Care. 2023;27:123.
5. Baseline plasma IL-18 may predict simvastatin treatment response in patients with ARDS: a secondary analysis of the HARP-2 randomised clinical trial. [journal citation per source]. 2022.
6. Targeted treatment of acute respiratory distress syndrome with statins — a commentary on two phenotype stratified re-analyses of randomized controlled trials. J Thorac Dis. 2019.
7. The Washington Manual of Critical Care, 4th ed. 2025 — ARDS pharmacotherapy chapter.
8. ICU Protocols: A Step-wise Approach, 2nd ed. — relevant ARDS content.