By Michael H. Crawford, MD, Editor

SYNOPSIS: Researchers studied patients with malignant pericardial effusion treated with pericardiocentesis and then anti-inflammatory agents if signs of adhesions or constriction were observed in post-drainage echocardiograms. Compared to NSAIDs and steroids, the authors found colchicine administration for 60 days reduced the rate of subsequent all-cause mortality and recurrent pericardial effusion.

SOURCE: Kim SR, Kim EK, Cho J, et al. Effect of anti-inflammatory drugs on clinical outcomes in patients with malignant pericardial effusion. J Am Coll Cardiol 2020;76:1551-1561.

Pericardial effusion (PE) is a common complication of malignancies and may be caused by metastases to the pericardium, an immune reaction to chemotherapy or pericardial damage resulting from radiotherapy. Despite the poor prognosis of many of these patients, therapy to preserve cardiac hemodynamics so that therapy of the malignancy can proceed is important. Pericardiocentesis (PCC) with placement of a drain performed under echocardiographic guidance is a safe and effective treatment for avoiding or treating pericardial tamponade. However, over time, adhesions and pericardial constriction (PC) can develop, which complicates cancer therapy. Since anti-inflammatory agents may reduce the reaction that leads to constriction, Kim et al investigated the incidence of PC and the effect of colchicine on long-term outcomes after PCC for malignancy-associated PE. After eliminating patients with no follow-up echoes, the authors identified 376 cancer patients who underwent PCC from 2007 to 2018 and followed them for a minimum of 24 months. After drainage fell below 30 mL/day, the catheter was removed, and a follow-up echo was performed within three days. Patients with adhesions or evidence of constriction after catheter removal were treated with an anti-inflammatory agent per their physician’s choice (colchicine, NSAIDs, or steroids). Researchers gave colchicine for one to three months, ibuprofen was tapered over two weeks, and steroids were administered depending on the patient’s response to therapy.

The primary outcome was a composite of all-cause mortality and re-PCC or pericardial window for recurrent PE. Men made up 56% of the patients, and the mean age of all patients was 57 years. Advanced lung cancer invading the pericardium was the most common diagnosis. About two-thirds underwent chemotherapy, and 27% radiotherapy. Pericardial tamponade was evident in 87%. The initial PCC was successful 97% of the time. Over the two-year follow-up, 26% of patients developed recurrent PE, and 46% developed CP. Colchicine was used in 24% of patients for a mean of 63 days. Initial clinical characteristics were not different in the colchicine group vs. the rest of the patients, but colchicine-treated patients exhibited more post-PCC adhesion and constriction and post-PCC NSAID use. The colchicine group was much less likely to reach the primary endpoint (adjusted HR, 0.60; 95% CI, 0.45-0.81; P = 0.001) compared to the non-colchicine group. This result persisted after propensity score matching (HR, 0.55; 95% CI, 0.37-0.82; P = 0.003). Also, the results were consistent across multiple subgroups. The authors concluded that in patients with malignant PEs undergoing PCC, colchicine administration for one to three months improves clinical outcomes.


The usual treatment for significant PEs in cancer patients is pericardial drainage, with the catheter left in place until fluid drainage is at or near zero. This occurs mainly to relieve symptoms or prevent impending cardiac tamponade. This aspect of the Kim et al study is not novel, but little formal study data on its effectiveness exist. It is encouraging that pericardial drainage was successful and complication-free 97% of the time. There was only one death despite four cases of cardiac perforation and two cases of pneumothorax. Median duration of indwelling catheter was four days, but most of the fluid was drained in the first day. The post-catheter removal echocardiogram showed adhesions in 71% and constriction in 37%. Thus, the unique aspect of this study was the treatment with anti-inflammatory agents in the patients who showed evidence of adhesions and constriction after drainage.

The principle weakness of the study was that the choice of specific anti-inflammatory was at the treating physician’s discretion. Colchicine and NSAIDs were the most common agents administered, but NSAIDs were only given for a short time, presumably to treat symptoms. About 20% of the patients received steroids. This analysis focused on comparing the colchicine-treated patients to those who did not receive colchicine. It showed the use of colchicine was associated with significantly less all-cause mortality and repeat pericardial drainage over the two-year follow-up. Steroid use was associated with an increase in mortality and subsequent PCC, but this could be because sicker patients were treated with steroids. In fact, in a multivariate analysis, steroid use was no longer significant (P = 0.07). NSAID use did not affect short- or long-term outcomes. Inspection of the Kaplan-Meier curves showed the benefits of colchicine therapy waned after one year, which perhaps is not surprising in these patients.

Why is this study important? The one-year survival of advanced cancer patients with PE in the Kim et al study was 30%, which is higher than that observed in older studies. Such patients are living longer today thanks to better cancer treatments, so abrogating the problem of recurrent PE or PC is of value to the patient. Also, because such patients are living longer, advanced complications like PE are becoming more frequent. Thus, using a relatively well-tolerated drug like low-dose colchicine for up to three months seems justified. Some physicians prefer a surgical or percutaneous pericardial window for malignant PE, but these procedures carry higher complication rates than PCC. In terminal cases with recurrent PE, pericariodesis can be considered, but this often is a painful procedure. Of course, this is a one-center, observational experience, so it can only be hypothesis-generating. Considering the alternatives, it seems reasonable to use now in appropriate patients.