Inflammation and postoperative complications after major surgery: a bi-directional two-sample Mendelian randomization study

This bi-directional Mendelian randomization study reveals that while down-regulated IL-6 signaling may causally reduce perioperative stroke risk, the divergent effects of inflammatory markers between acute postoperative and chronic all-cause outcomes suggest distinct underlying mechanisms, and previously reported associations between CRP and delirium likely stem from horizontal pleiotropy rather than direct causation.

The Gist

Imagine your body is a bustling city. When you undergo major surgery, it's like a massive construction project happening in the middle of downtown. This project inevitably causes some chaos: traffic jams, dust clouds, and emergency sirens. In medical terms, this is inflammation.

For years, doctors have noticed that when this "construction chaos" (inflammation) gets too loud, the city often suffers other problems: heart rhythm issues (atrial fibrillation), confusion (delirium), strokes, or infections at the construction site. The big question has always been: Is the chaos causing the problems, or is it just a witness to them?

This study is like hiring a time-traveling detective to solve the mystery. Instead of just watching what happens, the researchers used Mendelian Randomization (MR). Think of this as looking at people's "genetic blueprints" (their DNA) which were set in stone at birth. These blueprints act like a natural lottery: some people are genetically wired to have a slightly louder "construction crew" (higher inflammation) than others, regardless of their diet, stress, or lifestyle.

By comparing these "genetically loud" people to "genetically quiet" people, the researchers could see if the loudness actually caused the city to crash, or if they were just correlated.

Here is what they found, broken down into simple stories:

1. The "IL-6" Signal: A Tale of Two Cities

The researchers focused heavily on a specific messenger in the body called IL-6. Think of IL-6 as a fire alarm.

• In the "All-Cause" City (Chronic Disease): When they looked at people over their whole lives, they found that people with a "quieter" IL-6 alarm were much less likely to get Atrial Fibrillation (an irregular heartbeat). It's like saying, "If you turn down the fire alarm sensitivity, you get fewer false alarms and fewer heart issues over a lifetime." This suggests that for chronic heart issues, calming this specific alarm might actually help.
• In the "Post-Op" City (Surgery): Here, things got weird. When they looked specifically at people who had just had major surgery, the "quieter" IL-6 alarm seemed to protect against strokes. However, this protective effect didn't show up for heart attacks or general heart issues in the same way it did for the chronic disease group.
  • The Takeaway: The "fire alarm" (IL-6) behaves differently during the immediate chaos of surgery compared to the slow burn of aging. What helps in the long run might not be the exact same thing that helps right after a surgery.

2. The "CRP" Red Herring: The Smoke vs. The Fire

Another marker they studied was CRP (C-Reactive Protein). Imagine CRP as the thick black smoke rising from the construction site.

• The Old Theory: Observational studies (just watching people) suggested that if you could somehow "clean the smoke" (lower CRP), you would prevent delirium (sudden confusion after surgery).
• The Detective's Finding: The genetic study showed that this was a trick! When they looked at the root cause of the smoke (the genes that make CRP), they found that cleaning the smoke didn't actually stop the confusion.
  • The Metaphor: It's like seeing a lot of smoke and assuming the smoke is causing the fire. But the study proved that the smoke is just a symptom of the fire, not the arsonist. Lowering the smoke (CRP) won't stop the confusion; the smoke is just a sign that something else is going on.

3. The Big Picture: Why This Matters

The study concluded that for most post-surgery complications, simply trying to "dial down" the body's general inflammation system might not work the way we hope.

• The "One-Size-Fits-All" Trap: We can't just assume that what works for preventing heart disease over 20 years will work for preventing a stroke the day after surgery. The body's reaction to a sudden, massive trauma (surgery) is a different beast than the slow, steady wear and tear of aging.
• The Future: Instead of trying to turn off the whole "construction crew" (which might stop the healing process), doctors might need to look at the specific "local neighborhoods" (tissues) where the trouble is happening.

Summary in a Nutshell

• The Mystery: Does inflammation cause bad things after surgery?
• The Tool: We used DNA as a time-traveling detective to find the truth.
• The Twist:
  • IL-6 (The Alarm): Might help prevent strokes after surgery and heart issues in the long run, but the rules change depending on whether it's an acute injury or a chronic disease.
  • CRP (The Smoke): Is likely just a bystander. Lowering it won't stop confusion; it's just a sign that the body is stressed.
• The Lesson: We need to stop treating inflammation as a single, simple switch we can flip. The body is complex, and the "noise" after surgery is a unique event that requires a unique solution, not just a generic anti-inflammatory pill.

Technical Summary

1. Problem Statement

Observational studies have consistently reported associations between systemic inflammation (proxied by markers like CRP, IL-6, etc.) and postoperative complications such as acute kidney injury (AKI), atrial fibrillation (AF), delirium, myocardial infarction (MI), stroke, and surgical site infections (SSI). However, these observational associations are prone to confounding and reverse causation, making it unclear if inflammation is a causal driver of these complications or merely a biomarker of physiological stress.

Furthermore, it is unknown whether the causal architecture of acute postoperative complications differs from that of chronic (all-cause) development of these same conditions. If inflammation is causal, it could suggest immunomodulatory therapies as treatment strategies. This study aims to rigorously test these causal hypotheses using Mendelian Randomization (MR).

2. Methodology

The authors employed a bi-directional, two-sample Mendelian Randomization (MR) design.

• Data Sources:
  • Exposures (Inflammatory Markers): Genetic instruments (SNPs) were sourced from 9 Genome-Wide Association Studies (GWAS) involving up to 204,402 European participants. Markers included IL-6, IL-6R, CRP, MCP-1, and others.
  • Outcomes (Postoperative): Summary statistics from case-control GWAS within the UK Biobank. Cases were defined as patients undergoing major surgery who developed a specific ICD-10 coded complication within 30 days. Controls underwent major surgery without the complication.
  • Outcomes (All-cause/Chronic): Parallel analyses were conducted using all cases of the same ICD-10 phenotypes in the UK Biobank (non-surgical context) to compare acute vs. chronic pathways.
• Instrument Selection:
  • Cis-variants: SNPs within ±1Mb of the protein-coding gene (primary instruments for specific causal hypotheses).
  • Trans-variants: Genome-wide SNPs (used to characterize broader genetic architecture but susceptible to horizontal pleiotropy).
• Statistical Analysis:
  • Primary Method: Inverse-Variance Weighted (IVW) regression.
  • Sensitivity Analyses: MR-Egger, Weighted Median, Weighted Mode, and MR-PRESSO (to detect and correct for outliers/pleiotropy).
  • Validation: Steiger filtering, heterogeneity tests (Cochran's Q), and pleiotropy tests (Egger intercept).
  • Correction: Bonferroni correction applied for multiple testing.
  • Reverse MR: Tested if postoperative complications causally influence baseline inflammatory marker levels.

3. Key Contributions

• Differentiation of Acute vs. Chronic Pathways: The study explicitly compares the causal effects of inflammation on postoperative (acute) events versus all-cause (chronic) events, revealing significant divergences in causal architecture.
• Disentangling Pleiotropy: By comparing cis-restricted analyses with genome-wide analyses, the study distinguishes between direct causal effects and associations driven by horizontal pleiotropy (specifically regarding CRP and delirium).
• Specific Causal Evidence for IL-6: Provides robust genetic evidence for the role of IL-6 signaling in specific cardiovascular outcomes, while clarifying its context-dependent effects.

4. Key Results

A. Postoperative Complications (Acute Setting)

• IL-6 Signaling: Genetically proxied down-regulation of IL-6 signaling (via IL6R cis-variants) showed a robust protective effect against postoperative stroke (OR 0.27, 95% CI 0.11–0.69, p=0.006). There was suggestive evidence of protection against postoperative AF and MI, though these did not survive strict multiple testing corrections in all contexts.
• CRP and Delirium: Genome-wide analysis suggested a protective effect of CRP on postoperative delirium. However, this effect disappeared completely when restricted to cis-variants (OR 1.1, p=0.8). This indicates the initial association was driven by horizontal pleiotropy, not direct causation.
• Other Markers: No strong causal evidence was found for other inflammatory markers (e.g., MCP-1, IL-18) regarding postoperative AKI, SSI, or other complications after sensitivity analyses.
• Reverse MR: No significant evidence that postoperative complications causally alter baseline inflammatory marker levels (except a negligible effect of delirium on CRP).

B. All-Cause Outcomes (Chronic Setting)

• Divergence in IL-6 Effects: The protective effect of IL-6 down-regulation on stroke and MI seen in the postoperative analysis was absent in the all-cause analysis.
• Consistency in AF: In contrast, IL-6 down-regulation showed a robust, consistent protective effect against all-cause AF across all three instruments studied (p < 0.009).
• Infection: Genetically proxied IL-6 down-regulation was associated with an increased risk of all-cause skin/subcutaneous infections (OR ~1.41), suggesting a complex, tissue-specific role for IL-6 in infection susceptibility.
• CRP and Delirium: Similar to the postoperative analysis, the genome-wide protective association between CRP and all-cause delirium was abolished in cis-restricted analysis, confirming horizontal pleiotropy.

5. Significance and Conclusions

• Causal Architecture Differs: The study concludes that the inflammatory mechanisms driving acute postoperative complications differ significantly from those driving chronic disease. For instance, IL-6 signaling appears critical in chronic AF but its role in acute postoperative stroke/MI is distinct or less pronounced in the chronic context.
• CRP is a Biomarker, Not a Cause: The findings strongly suggest that the association between CRP and delirium is non-causal, likely reflecting shared upstream pathways (pleiotropy). This reinforces the view of CRP as a biomarker of inflammation rather than a therapeutic target for delirium.
• Therapeutic Implications: The results do not support the broad application of systemic immunomodulatory therapies (e.g., IL-6 blockade) for the prevention of general postoperative complications. The lack of causal signals in the acute setting, combined with the increased risk of infection seen in the chronic IL-6 analysis, suggests that systemic inhibition may be ineffective or potentially harmful in the perioperative period.
• Future Directions: Research should focus on local tissue responses and the immediate perioperative period rather than relying solely on lifelong genetically predicted baseline inflammation levels, as the latter may not capture the acute reactivity to surgical stress.

Limitations Noted: The study was likely underpowered for specific rare postoperative complications due to sample size constraints in UK Biobank. Additionally, genetic instruments reflect lifelong baseline levels, which may not accurately model the acute inflammatory reactivity triggered by surgery.