Coronary sinus lactate measured shortly after reperfusion may help identify patients more likely to develop postoperative atrial fibrillation after coronary artery bypass grafting (CABG). In a prospective cohort of 94 patients, Verma and colleagues found that a concentration above 3 mmol/L five minutes after aortic cross-clamp removal was associated with substantially higher odds of treated postoperative atrial fibrillation. The result is promising, but the small study does not establish a validated treatment threshold or demonstrate that lactate-guided prophylaxis improves outcomes.
Study Design and Patient Population
The investigators analyzed prospectively collected data from patients undergoing isolated CABG at Brigham and Women’s Hospital between December 2024 and February 2026. Patients with a history of atrial fibrillation, concomitant valve procedures, operations without aortic cross-clamping, and those receiving del Nido cardioplegia were excluded. The final cohort reflected the practice of two surgeons at one academic center.
Myocardial protection used antegrade and retrograde 8:1 whole-blood cardioplegia, with repeat doses at 20-minute intervals. Patients were allowed to cool passively, typically to 34°C. These details matter because lactate patterns and their relationship with postoperative rhythm may differ with other myocardial protection strategies, operations, or patient populations.
When Coronary Sinus Lactate Was Sampled
Blood was collected through the retrograde coronary sinus cannula before cardiopulmonary bypass (CPB), approximately five minutes before cross-clamp removal, and five minutes after removal. For the pre-removal sample, the left internal mammary artery was briefly opened to provide sufficient coronary blood volume for sampling. Right atrial samples were obtained before CPB and after cross-clamp removal. Samples were analyzed immediately.
The primary endpoint was electrocardiographically documented atrial fibrillation treated with amiodarone during the index hospitalization. Treatment was given for episodes lasting more than one hour or symptomatic episodes of any duration. Continuous telemetry continued until discharge. Thus, the primary endpoint represented clinically treated arrhythmia rather than every detected episode of atrial fibrillation.
Key Findings After CABG Reperfusion
Twenty-four patients, or 26%, met the primary endpoint. Another 11 had atrial fibrillation documented on telemetry without meeting the amiodarone-treatment criteria. Across the cohort, median coronary sinus lactate rose from 1.0 mmol/L before CPB to 2.8 mmol/L before cross-clamp removal, then decreased to 2.1 mmol/L after reperfusion.
| Reported measure | Finding |
|---|---|
| Patients with post-reperfusion lactate >3 mmol/L | 13 of 94 patients (14%) |
| Treated atrial fibrillation with lactate >3 mmol/L | 62% |
| Treated atrial fibrillation with lactate ≤3 mmol/L | 20% |
| Adjusted odds ratio for lactate >3 mmol/L | 11.7 (95% CI, 1.80–76; P = .010) |
| Sensitivity / specificity of the threshold | 33% / 93% |
| Positive / negative predictive value | 62% / 80% |
The adjusted analysis accounted for age, number of distal anastomoses, CPB duration, cross-clamp duration, and earlier lactate measurements. Older age was also associated with greater odds of the primary outcome. Earlier coronary sinus lactate values, CPB time, and cross-clamp time were not independently associated with that endpoint in the reported model.
Interpreting the Size of the Association
The adjusted odds ratio of 11.7 should not be described as an 11.7-fold increase in absolute risk. Odds and probabilities differ, especially when an outcome is common. The observed proportions—62% versus 20%—provide a more direct description of the group difference, while the very wide confidence interval signals substantial uncertainty in the adjusted estimate.
The threshold had high specificity but low sensitivity. It identified a small group with frequent treated arrhythmia, yet missed approximately two-thirds of patients who developed the primary outcome. A value at or below 3 mmol/L therefore did not exclude postoperative atrial fibrillation. Predictive values also depend on the underlying event rate and may change in another institution.
Myocardial Metabolism and Perfusion Relevance
The association was strongest after reperfusion rather than during the preceding ischemic period. Patients who remained free of the treated endpoint generally showed a decline in coronary sinus lactate after cross-clamp removal, whereas those who developed it had more persistent elevations. This pattern is consistent with a possible relationship between postoperative rhythm vulnerability and metabolic recovery, but the study did not establish a molecular mechanism.
Post-reperfusion lactate correlated with longer CPB duration, higher body mass index, and a greater number of distal anastomoses. These findings may reflect interacting patient, procedural, and metabolic factors. They do not demonstrate that a specific perfusion technique, cardioplegia formulation, or change in bypass duration would prevent arrhythmia.
For perfusion teams, the practical research implication is that sample location and timing deserve explicit documentation. A routine systemic lactate concentration should not automatically be substituted for this coronary sinus measurement. Coronary sinus concentration alone also does not establish net myocardial lactate production without corresponding inflow measurements and an appropriate metabolic assessment.
Right Atrial Samples and Model Performance
Right atrial and coronary sinus lactate after reperfusion were strongly correlated (r = 0.88). The reported right atrial model had a C-statistic of 0.82, compared with 0.87 for the coronary sinus model. The overlapping confidence intervals and small cohort limit conclusions about whether coronary sinus sampling offers a clinically meaningful advantage.
The complete adjusted model’s C-statistic of 0.87 should not be attributed to lactate alone. In additional analyses, discrimination was 0.68 for age alone, 0.63 for post-reperfusion lactate alone, and 0.77 when the two were combined. These internally derived estimates require external validation before use as a clinical prediction tool.
Limitations and Next Steps
The study included only 94 patients and 24 primary events, with just 13 patients above the proposed lactate threshold. The cutoff was selected from the same dataset using restricted cubic splines, which increases the possibility of an optimistic result. Multiple predictors relative to the number of events create additional concern about model stability despite the investigators’ use of regularization and bootstrap assessment.
The treatment-based endpoint could vary with local amiodarone practice, and monitoring ended at discharge. A sensitivity analysis including all detected postoperative atrial fibrillation retained an association with elevated lactate, but does not eliminate concerns about sample size or generalizability. All patients meeting the primary endpoint were male, further limiting conclusions across patient groups.
Patients received perioperative beta-blockade when feasible; the study did not test a lactate-triggered prophylaxis strategy. Larger multicenter cohorts should prospectively validate sampling methods, thresholds, and model calibration. A separate interventional study would be needed to determine whether acting on the measurement improves outcomes. The authors reported no conflicts of interest.
Original Article
Verma A, Shin B, Hirji S, et al. Coronary sinus lactate at reperfusion: A predictor of atrial fibrillation after coronary artery bypass grafting. JTCVS Open. 2026;33:101971. doi:10.1016/j.xjon.2026.101971.
This independently written educational summary is based on the original open-access article. The source PDF is unchanged. The cover is an AI-generated editorial illustration.
