Pericardial Window Surgery,
London

Why does the echocardiogram, not the CT scan, decide whether an effusion matters?

Most pericardial effusions in cancer patients are picked up incidentally, on a staging or restaging CT performed for another reason. The scan report notes the effusion and often gives a size, and a referral follows from there. Referrals therefore frequently arrive with cross-sectional imaging and no echocardiogram, simply because no echocardiogram had been indicated until that moment.

CT does a good deal of work here and the 2025 ESC guidelines recommend it for exactly that — pericardial thickness, calcification, loculated or posteriorly trapped collections, pericardial masses, and the state of the lungs and pleura alongside, which changes the operative plan. What it does not settle is haemodynamics, and haemodynamics is what a pericardial window is an operation about.

Effusion depth on a scan correlates poorly with significance in either direction, because tamponade turns on the rate of accumulation and the compliance of the sac rather than on volume: a rapidly accumulating 200 mL can be dangerous where a litre collected over months is tolerated. The consequence in practice is that large incidental effusions are sometimes escalated harder than they need to be, and smaller acute ones are occasionally underestimated.

The assessment that resolves it is the standard echocardiographic one — right-sided chamber collapse, respiratory variation across the atrioventricular valves, inferior vena caval plethora. A transthoracic echocardiogram is therefore arranged before the surgical decision is made, and where a referral arrives with CT alone that is the first thing organised. It is worth doing before the referral where it is practical, because it usually determines whether the case is urgent, elective, or not surgical at all.

Is a pericardial window better than draining the fluid with a needle?

For durability, yes. For survival, no — and it is worth setting that out plainly, because the literature is often quoted selectively in one direction or the other.

Pericardiocentesis is the correct first response to tamponade. It is fast, it can be performed without general anaesthesia, and it converts an emergency into a stable situation. Its weakness is recurrence. In cancer patients, drainage alone has long been associated with re-accumulation in the great majority within three months, and it is that pattern of repeated taps — each one an admission, each one an interruption to systemic treatment — that a window is designed to end.

The comparative evidence is retrospective throughout; no randomised trial exists. A 2025 systematic review of 12 studies and 3,721 patients found re-accumulation markedly more common after pericardiocentesis than after a surgical window (odds ratio 3.81, 95% CI 1.99–7.31), with no significant difference in in-hospital mortality (OR 0.99, 95% CI 0.68–1.46). A 2026 meta-analysis restricted to the first occurrence of malignancy-related effusion, covering six studies and 1,369 patients, found procedural failure significantly higher with pericardiocentesis (OR 2.99, 95% CI 1.10–8.14) but more deaths reported in the surgical window group (OR 0.68, 95% CI 0.46–0.99).

That last finding should be read carefully rather than dismissed. In retrospective series, patients who reach an operating theatre are systematically different from those who do not: they are the ones whose effusion recurred, whose disease is more advanced, or who were sick enough that a definitive procedure seemed warranted. A mortality signal of that shape is much more likely to reflect selection than harm from the operation itself — and the largest dedicated surgical series reported no procedure-related deaths at all. But the honest position is that a window has not been shown to prolong life, and it should not be offered on that basis.

The reasonable reading is that a window is the answer to a recurrent effusion, not to a first one. That is also where the guidelines place it: the 2025 ESC guidelines recommend pericardiocentesis for tamponade and for suspected neoplastic or bacterial pericarditis, and give a surgical pleuro-pericardial window a Class I recommendation for relapsing pericardial effusion despite medical therapy.

In practice that resolves into a simple division of labour. A first presentation is drained percutaneously under cardiology, with fluid sent for cytology. Where the effusion then recurs — and particularly where the underlying disease is malignant — the case comes to surgery for a window. The one indication that goes to surgery from the outset is pyopericardium, because a thick fibrinous exudate will not pass through a needle or a pigtail catheter any more than empyema will drain through a chest tube; it has to be evacuated under vision, and the pericardial space debrided. That reasoning is the same one that governs surgical drainage in the pleural space, and it is why the 2025 guidelines single out purulent effusion for surgical drainage rather than percutaneous.

What stops the effusion coming back after a window?

The commonest reason an effusion recurs after surgery is that the window was too small.

A short pericardiotomy behaves like a drainage tract rather than a window. The cut edges of pericardium fall back against the epicardium, fibrin and adhesion seal them over days to weeks, and the communication closes. What follows is a second effusion in a patient who has already had an operation for the first — and the inference usually drawn is that the disease was aggressive, rather than that the opening was inadequate.

The working rule is a window about the size of a fifty pence piece — roughly three centimetres across — taken as a full-thickness disc of pericardium rather than a slit or a flap. At that size the edges cannot appose, the defect cannot close by healing, and the pericardial space stays permanently open to the pleural cavity, where fluid produced afterwards is reabsorbed instead of accumulating under pressure. It is the difference between decompressing a space once and changing how it drains.

This is the substantive reason to excise rather than incise, and it is the technical point on which recurrence turns. It is consistent with what the published series show: where windows are properly made, re-intervention is uncommon — fewer than one in ten patients in the largest surgical cohort — and of those who survived a year, close to nine in ten never needed the pericardium touched again. The excised disc is also the histology specimen, so an adequate window and an adequate biopsy are the same manoeuvre.

What is recovery like after a pericardial window?

The operation itself is short — usually under an hour by either route. A drain stays in place until output settles, typically two to four days, and is removed on the ward without a further anaesthetic.

Length of stay is determined far more by the underlying illness than by the operation. A patient admitted electively with a recurrent effusion and otherwise reasonable performance status will often be home within a few days. A patient admitted acutely unwell with advanced cancer, or one who required pericardial drainage before surgery, will stay longer, and that stay usually reflects the systemic illness rather than the chest.

Breathlessness generally improves immediately and noticeably, because the component of it that was mechanical — a compressed heart, a compressed lung, or both — has been removed. The improvement that comes from decompressing the pericardium is one of the more reliable symptomatic gains in thoracic surgery.

Systemic anti-cancer treatment can be resumed once the wounds have healed, and the timing is agreed directly with the treating oncologist rather than dictated by the surgery. This matters: the operation exists in large part to protect the continuity of systemic treatment, and published series suggest that patients who receive systemic chemotherapy alongside a pericardial window do better than those who receive the operation alone. Recurrence requiring further pericardial intervention is uncommon once a window has been made adequately, for the reasons set out above.

Histology and cytology from the pericardium are reported within days and go directly to the referring team. As set out below, the result carries prognostic weight alongside its diagnostic value.

What is the outlook after a pericardial window for a malignant effusion?

The great majority of pericardial effusions that reach an operation are malignant, and the prognosis of the underlying disease is poor. There is no useful way to write about this operation without saying that plainly.

In the largest published surgical series — 166 consecutive cancer patients undergoing a pericardial window at a single centre over twelve years — one-year overall survival after the operation was 31.4%. Lung cancer accounted for 61.4% of that cohort and breast cancer for 9.6%; seven in ten patients had stage IV disease. That survival figure describes the cancer, not the operation. There were no procedure-related deaths in the series, and fewer than one in ten patients required any further pericardial intervention.

Within that population the outlook is not uniform, and the differences are large enough to change how a case is discussed. One-year survival was 45.1% where the pericardial cytology and biopsy were negative for malignant cells, against 20.6% where they were positive. Disease that was not stage IV carried an independently better outcome after adjustment (hazard ratio 0.54, 95% CI 0.35–0.84), as did negative cytology (hazard ratio 0.45, 95% CI 0.31–0.67); NYHA class IV breathlessness at presentation carried a worse one. The primary cancer type made no significant difference once those factors were accounted for, although haematological malignancy did comparatively well.

The practical consequence is that a pericardial window is not a terminal gesture in every patient who has one. Where systemic treatment is effective — a lung adenocarcinoma with a targetable driver mutation, a lymphoma, a chemosensitive tumour — a functioning window removes the effusion as a recurring emergency and allows that treatment to continue uninterrupted. Patients in that position can live for many months, and sometimes considerably longer, with the pericardium no longer part of the clinical problem. That is the argument for operating rather than tapping repeatedly, and it is a narrower argument than "surgery improves survival", which the evidence does not support.

It is also why the cytology result is reported back promptly rather than simply filed with the histology: it is one of the few pieces of prognostic information this operation generates, and it arrives at a point where treatment decisions are still open.

When is a pericardial window done for a non-cancerous effusion?

Less often, but the indication is well established and in some respects clearer than in advanced malignancy, because the patient's overall prognosis is not the limiting factor.

The 2025 ESC guidelines give a surgical pleuro-pericardial window a Class I recommendation in patients with relapsing pericardial effusion despite medical therapy, and recommend surgical pericardial drainage where percutaneous drainage is not feasible or where the effusion is purulent. Pyopericardium is the clearest of those: the exudate is thick and fibrinous, it will not pass a needle or a pigtail catheter, and it needs evacuating and debriding under vision — both to clear the infection and to reduce the risk of subsequent constriction. In practice the benign indications seen most often are recurrent effusion following viral or idiopathic pericarditis that has failed anti-inflammatory treatment including colchicine, effusion in end-stage renal disease, effusion associated with connective tissue disease, post-cardiac-surgery effusion that has recurred after drainage, and late effusion following thoracic radiotherapy.

Two points are worth stating explicitly. First, a large asymptomatic effusion is not by itself an indication for surgery; the decision rests on symptoms, haemodynamics and the trajectory over serial echocardiograms, not on a measurement. Second, where the underlying process is inflammatory, medical treatment is the treatment — surgery addresses the fluid and does not modify the disease, and an effusion that is still actively inflammatory may warrant a further attempt at medical control first. Where a benign effusion has genuinely relapsed despite adequate therapy, a window is durable and the operation is well tolerated.

Simple pericardial cysts are a different entity again — congenital, usually asymptomatic, found incidentally in the visceral mediastinum, and resected robotically or thoracoscopically only where they are symptomatic or diagnostically uncertain. They are covered on the mediastinal surgery page.