Targeting inflammation with colchicine does not prevent atrial fibrillation after major non-cardiac thoracic surgery

Comment on ‘Effect of colchicine on perioperative atrial fibrillation and myocardial injury after non-cardiac surgery in patients undergoing major thoracic surgery (COP-AF): an international randomised trial’, which was published in The Lancet, https://doi.org/10.1016/S0140-6736(23)01689-6. The Colchicine for the Prevention of Perioperative Atrial Fibrillation (COP-AF) was an investigator-initiated, randomized, double-blind, placebo-controlled trial aiming to test the efficacy of low-dose colchicine in reducing the occurrence of clinically important perioperative atrial fibrillation (AF) and myocardial injury after major non-cardiac thoracic surgery (MINS).1 Patients aged 55 years or older scheduled for major non-cardiac thoracic surgery without history of documented AF were randomized in a 1:1 ratio to receive colchicine .5 mg twice daily or matching placebo, starting within 4 h before surgery and continuing for 10 days. The primary endpoint was clinically important perioperative AF during 14 days of follow-up, defined as an episode of AF (or atrial flutter) resulting in angina, heart failure, or symptomatic hypotension or requiring treatment with a rate-controlling drug, anti-arrhythmic drug, or electrical cardioversion. The coprimary endpoint, which was added ∼4 years after enrolment start, was MINS, defined as myocardial infarction or elevated postoperative troponin related to myocardial ischaemia. Safety outcomes were a composite of sepsis or infection and non-infectious diarrhoea. From February 2018 to June 2023, a total of 3209 patients (mean age, 68 years; 48% female) were enrolled at 45 institutions in 11 countries and randomized to colchicine (n = 1608) or placebo (n = 1601). The most common procedures were pulmonary resections (99%), and a thoracoscopic approach was used in 75% of cases. The clinical indication for surgery was confirmed or presumed malignancy in 93% of cases. During follow-up, 15% of patients discontinued the study medication (17% in the colchicine group vs. 13% in the placebo group). In the colchicine group, the most common reason for study drug discontinuation was non-infectious diarrhoea (29% of cases). The incidence of clinically important perioperative AF was not significantly different between the colchicine group and the placebo group [6.4% vs. 7.5%, respectively; hazard ratio (HR) .85; 95% confidence interval (CI), .65–1.10; P = .22], as was the incidence of MINS (18.3% vs. 20.3%, respectively; HR .89; 95% CI, .76–1.05; P = .16). The composite secondary outcome of all-cause death, non-fatal MINS, and non-fatal stroke occurred in 18.7% of patients in the colchicine group and in 20.9% of patients in the placebo group (HR, .88; 95% CI, .75–1.03). The incidence of venous thromboembolism was low (.7%), but was numerically higher in the colchicine group than in the placebo group (.9% vs. .4%, respectively; HR 2.50; 95% CI, .97–6.44). The composite safety outcome of sepsis or infection occurred in 6.4% of patients in the colchicine group and in 5.2% of those in the placebo group (HR 1.24; 95% CI, .93–1.66). A significant increase in the incidence of non-infectious diarrhoea was observed in the colchicine group as compared with the placebo group (8.3% vs. 2.4%, respectively; HR 3.64; 95% CI, 2.54–5.22). The median length of hospital stay was 5 days, without any difference between the two groups. Perioperative AF is the most common type of secondary AF and the most frequent complication after cardiac surgery, occurring in approximately one-third of patients, particularly after heart valve replacement (up to 50% of cases).2 Perioperative AF is relatively frequent also after non-cardiac surgery (10%–20%), with higher rates among patients undergoing more invasive surgical procedures, including lung transplant, pneumonectomy, and oesophagectomy.3 Regardless of the type of surgery, periprocedural AF is associated with adverse cardiovascular (CV) outcomes, as well as with prolonged hospital stay and increased health care expenses.4 Although the pathogenesis of periprocedural AF is multifactorial, inflammation seems to represent a key mechanism.5 Interestingly, higher inflammatory biomarker concentrations, including C-reactive protein, are associated with increased occurrence of periprocedural AF.6 Previous studies explored the use of colchicine as prophylaxis of periprocedural AF, yielding conflicting results.7,8 In the Effect of Low-dose ColchiciNe on the InciDence of Atrial Fibrillation in Open Heart Surgery Patients (END-AF) trial, low-dose colchicine was not effective in preventing periprocedural AF in 152 patients undergoing cardiac surgery.7 In contrast, a substudy of the Colchicine for the Prevention of the Post-Pericardiotomy Syndrome (COPPS) in 336 patients showed a beneficial effect of 1-month colchicine therapy in reducing the incidence of perioperative AF, resulting in shorter duration of hospitalization and rehabilitation stay after cardiac surgery.8 The COP-AF was the first randomized clinical trial (RCT) testing the hypothesis that targeting inflammation with colchicine in patients undergoing major non-cardiac thoracic surgery may prevent perioperative AF and MINS.9 However, colchicine did not show any significant benefit in reducing the incidence of both outcomes in this clinical setting. Of note, the observed 1.1% absolute reduction in periprocedural AF was about one-third of the expected 3.4% reduction. This finding, together with the lower-than-expected event rate in the placebo group (7.5% instead of 11%), reduced the statistical power of the study. Therefore, despite the addition of MINS as an independent coprimary endpoint and sample size recalculation (with an increase from 2800 to 3200 patients) several years after trial initiation, the trial was clearly underpowered to detect a more realistic clinical benefit than the expected 30% reduction in periprocedural AF. It is possible that the absence of a predefined protocol for postoperative rhythm monitoring (based on routine practices at each site), with only ∼50% of patients having an electrocardiogram (ECG) on postoperative Day 3, might have contributed to the relatively low occurrence of periprocedural AF. The lower-than-expected incidence of periprocedural AF might have also been related to the relatively low CV risk profile of the trial population (<10% of patients had a history of CV disease) and to the type of surgical intervention. While CV risk factors and comorbidity burden contribute to the risk of periprocedural AF, surgery-related factors (including direct organ manipulation), autonomic factors (e.g. increased catecholamine levels or vagal tone, frequently undergoing fluctuations during surgery), atrial pressure, and volume overload are other important determinants, which could possibly prevail over inflammation during major thoracic surgery.9 An unanticipated finding was the observation of a two-fold higher incidence of deep venous thrombosis or pulmonary embolism in the colchicine group compared to the placebo group, a finding that had also been reported in a previous trial.10 As the rate of this adverse event was low (<1%) and there are no clear mechanistic explanations, this finding should be interpreted with caution and verified in future colchicine trials. Of note, the incidence of sepsis or infection was numerically higher in the colchicine group, with a relative risk increase of ∼25% compared to placebo on intention-to-treat analysis, which might have been even higher (and potentially significant) using on-treatment analysis, given the high drop-out rate in the colchicine group. An expected side effect of colchicine was diarrhoea, which was over three times more frequent than in the control group, and represented the most common cause of study drug discontinuation in the colchicine group (∼29% of cases of colchicine discontinuation). These downsides need to be considered when managing patients undergoing major non-cardiac surgery who have important comorbidity burden, as colchicine-induced diarrhoea can alter intestinal absorption of other medications.11 In conclusion, colchicine does not provide a significant benefit in patients undergoing major non-cardiac thoracic surgery in terms of prevention of periprocedural AF and myocardial injury, and its use cannot therefore be recommended in this setting. Other potential preventive strategies are currently under investigation in ongoing RCTs.12 Whether specific subgroups of high-risk patients undergoing major non-cardiac surgery, such as those with increased markers of inflammation, might benefit from colchicine treatment remains to be determined. R.V. received consulting or lecturing fees from Abbott Vascular, Abiomed, Amgen, Daiichi Sankyo, Medtronic, and Terumo, outside the submitted work. L.G. has nothing to disclose.