All that glitters ain’t gold

Propofol is commonly used due to its favorable pharmacokinetics, including rapid onset and elimination. However, all that glitters ain’t gold. In 2001, an official warning on a possible increased mortality due to propofol administration in pediatric populations was published into the Food and Drug Administration website (1). Subsequently, two randomized controlled trials (RCTs) showed a statistically significant reduction in survival with a propofol-based anesthesia (2,3). Widespread use despite a warning signal of mortality risks forms a sufficient background to perform further investigations. Indeed, our meta-analyses (4,5) found an increased mortality risk of propofol (5).

Our meta-analysis of RCT (5) suggested a potentially detrimental effect of propofol administration on survival in critically ill and perioperative settings, which generated an intense debate among clinicians (6). We included 252 RCTs comprising more than 30,000 patients. The propofol group had a 5.2% (760/14,754) mortality compared with 4.3% (682/16,003) in the comparator [risk ratio (RR), 1.10; 95% confidence interval (CI): 1.01–1.20; P=0.03]. Particularly, we found a pronounced effect in the cardiovascular subgroup (4.6% vs. 2.8%; RR, 1.46; 95% CI: 1.16–1.89; P=0.004). These findings led us to conclude, “Propofol may reduce survival in perioperative and critically ill patients. This needs careful assessment of the risk versus benefit of propofol compared to other agents while planning for large, pragmatic multicentric randomized controlled trials to provide a definitive answer.”

A 2024 Annals of Translational Medicine editorial (7) reignited the debate by presenting a series of limitations concerning our meta-analysis. We now provide a point-by-point reply.

First, we extracted data according to the intention-to-treat principle. This is a conservative approach, mitigating possible inappropriate magnifications of effect measures, which may occur in per-protocol analysis. When extracting data with the per-protocol approach, the mortality increase in patients randomized to receive propofol was still significant in the cardiovascular population (RR, 1.36; 95% CI: 1.06–1.76).

Second, we performed a simple meta-analysis (with both frequentist and Bayesian statistics) of RCTs comparing using or not using propofol. We aimed to document the effects of propofol on survival and not to compare all available hypnotics and anesthetics through a network meta-analysis. Notably, propofol was associated with an increased mortality when compared with any comparator in the overall analysis, and subgroup analyses considering individual comparators exhibited consistent effect estimates and directions.

Third, we correctly excluded the MYRIAD trial (8) together with several other RCTs performed on this topic, such as the SPICE III trial, which did not fulfill our eligibility criteria. The two arms of our meta-analysis were transparently anticipated in the PROSPERO registration (CRD: 42022323143), are clearly defined in the methods section and could be summarized using simple PICOS framework: in critically ill and perioperative patients (P), does a propofol-based anesthesia and/or prolonged sedation (I) compared with any different hypnotic agent (C) reduces survival (O) in randomized controlled trials (S)? Particularly, in both the MYRIAD and SPICE III RCTs, propofol administration was allowed in both groups by the study protocols and was therefore administered in 98% and 80% of the control group patients, respectively. We also would like to underline that a post-hoc analysis of the MYRIAD trial (9) demonstrated a significantly increased 1-year death due to cardiac cause in total intravenous anesthesia group, and that in the SPICE III study, patients aged over and randomized in the propofol group had significantly reduced survival. Again, these two studies do not fulfill the criteria to be included in our meta-analysis, but they further highlighted the potential detrimental effects of propofol in some populations.

Fourth, we performed the analyses using a fixed effects model, which was prespecified in the PROSPERO registration (CRD: 42022323143). According to the Cochrane handbook for systematic review of interventions, the application of a fixed-effects model over a random-effects one is preferred whenever statistical heterogeneity, measured using the Cochran’s Q statistic and displayed as I² (%), is low (heterogeneity was 0% in our analyses!). Importantly, several small randomized trials with trends towards to an increased mortality in the propofol group were never published for unknown reasons as documented by the trim and fill analysis. When using the random-effects model to our database enriched with these 24 unpublished RCTs we found again a significant increased mortality in patients receiving propofol.

Fifth, Hauquiert et al. (7) stated that, “A significant counterpoint to Kotani et al.’s (5) stance is provided by a large RCT published by Pasin et al. in 2015 (4)”, without realizing that this manuscript is (I) a large meta-analysis of randomized trials not a large RCT; (II) it was performed by our group (Dr. Pasin is an author of both meta-analyses); (III) it represents the roots of the 2023 meta-analysis; and (IV) the results are perfectly consistent to the 2023 Kotani et al. update.

Sixth, we presented mortality data at the longest follow-up available. When limiting the analysis to the few studies reporting 30-day mortality, would have delayed the dissemination of these clinically relevant findings which have implication for millions of patients every year. Longest follow-up data extraction and analysis strategy is an important validated technique to increase sample size in critically ill settings (10).

Finally, we remark that this meta-analysis would not make any specific practical recommendation against the use of propofol, but rather warning on this possible detrimental effect and the need for further research (6). In fact, we suggested paying more attention to some good clinical practice measures, already performed in many centers, such as the implementation of hypnotic agents rotation and the application of propofol sparing strategies while waiting for a definitive answer.

Acknowledgments

None.

Footnote

Provenance and Peer Review: This article was a standard submission to the journal. The article did not undergo external peer review.

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://atm.amegroups.com/article/view/10.21037/atm-25-35/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Canadian Medical Association. FDA issues warning on propofol (Diprivan). CMAJ 2001;164:1608.
2. Likhvantsev VV, Landoni G, Levikov DI, et al. Sevoflurane Versus Total Intravenous Anesthesia for Isolated Coronary Artery Bypass Surgery With Cardiopulmonary Bypass: A Randomized Trial. J Cardiothorac Vasc Anesth 2016;30:1221-7. [Crossref] [PubMed]
3. de la Gala F, Piñeiro P, Reyes A, et al. Postoperative pulmonary complications, pulmonary and systemic inflammatory responses after lung resection surgery with prolonged one-lung ventilation. Randomized controlled trial comparing intravenous and inhalational anaesthesia. Br J Anaesth 2017;119:655-63. [Crossref] [PubMed]
4. Pasin L, Landoni G, Cabrini L, et al. Propofol and survival: a meta-analysis of randomized clinical trials. Acta Anaesthesiol Scand 2015;59:17-24. [Crossref] [PubMed]
5. Kotani Y, Pruna A, Turi S, et al. Propofol and survival: an updated meta-analysis of randomized clinical trials. Crit Care 2023;27:139. [Crossref] [PubMed]
6. Vincent JL. Harmful effects of propofol? The Editor’s standpoint. Crit Care 2023;27:279. [Crossref] [PubMed]
7. Hauquiert B, Blackman S, Perriens E, et al. Propofol curtails survival in perioperative and critically ill patients by a relative reduction of 10%: should propofol be abandoned? Ann Transl Med 2024;12:98. [Crossref] [PubMed]
8. Landoni G, Lomivorotov VV, Nigro Neto C, et al. Volatile Anesthetics versus Total Intravenous Anesthesia for Cardiac Surgery. N Engl J Med 2019;380:1214-25. [Crossref] [PubMed]
9. Zangrillo A, Lomivorotov VV, Pasyuga VV, et al. Effect of Volatile Anesthetics on Myocardial Infarction After Coronary Artery Surgery: A Post Hoc Analysis of a Randomized Trial. J Cardiothorac Vasc Anesth 2022;36:2454-62. [Crossref] [PubMed]
10. Roth D, Heidinger B, Havel C, et al. Different Mortality Time Points in Critical Care Trials: Current Practice and Influence on Effect Estimates in Meta-Analyses. Crit Care Med 2016;44:e737-41. [Crossref] [PubMed]