Mediastinoscope-controlled parasternal fenestration of the pericardium: definitive surgical palliation of malignant pericardial effusion

The number of patients with secondary malignant pericardial effusion is steadily increasing due to the improving efficacy of complex cancer treatment modalities. Metastases to the myocardium have reportedly been found in up to 1.23% of all autopsies [1]. The incidence of malignancy associated with pericardial involvement ranges from 8 to 20% in other autopsy studies [2–4]. The tumorous infiltration or carcinosis of the pericardium could cause pericardial effusion in up to one-third of cases of malignancy, thus potentially interfere with the otherwise desirable oncological treatment [4].

The existing surgical methods for the management of pericardial fluid are well-established but are not without limitations in providing symptomatic relief of malignant pericardial effusion. Pericardiocentesis and the Larrey-Fontenelle approaches [5, 6] using a subxyphoideal fenestration have a recurrence rate ranging between 43 and 69% after pericardiocentesis [3, 7] and 9 and 16% after pericardial drainage [8–11]. The video-assisted thoracoscopic surgical (VATS) pericardio-pleural fenestration is highly efficient but requires anaesthesia with contralateral single-lung ventilation which is compromising in patients with an already reduced cardio-pulmonary reserve [8, 12, 13]. The transdiaphragmatic pericardial fenestration was developed as an alternative method, but the pericardio-peritoneal window might be blocked by the adjacent abdominal structures and is a two-cavity procedure [14, 15]. The desire to overcome the relative limitations detailed above led us to explore an alternative approach.

Having extensive experience in collar mediastinoscopy, we utilized the existing armamentarium of the Carlens procedure [16] in a Chamberlain approach in order to achieve pericardial decompression in the form of a pericardiopleural window [17–19].

The patient was placed in a 10–30 degree supine Fowler position. Anaesthesia was introduced with single lumen endotracheal tube. A transverse skin incision was made above the sternal end of the third or fourth rib, a little bit longer than the expected length of its cartilaginous part. The laterality of approach (i.e. left or right) was decided based upon the pleural involvement, as tumor-free pericardiopleural reflection is required. In the case of synchronous pleural carcinosis the opposite side was our choice. A left-sided approach was preferred when both pleural cavities were suitable for the procedure. After blind splitting the pectoral muscles the sternal edge of the rib was resected. The internal thoracic artery and vein were identified and secured in order to avoid annoying bleeding in the course of later manipulation. The pleural cavity was opened close to its mediastinal reflection. A limited exploration of the thoracic cavity was undertaken introducing a rigid mediastinoscope (Karl Storz GMBH, Tuttlingen, Germany), and biopsies were performed if suspicious parenchyma/pleural lesions were present (Figure 1). Keeping aside the lung with a long retractor a holding stitch secured pericardial access ventrally to the phrenic nerve. Lifting up this stitch, a window an at least 3 cm in diameter was created on the pericardium using scissors to manipulate within or parallel to the main tube of the mediastinscope (Figure 2, 3). The effluent pericardial liquid was emptied with mild suction and a 20 Ch marked silicone chest tube was inserted into the pleural cavity. The chest tube was kept under suction (50–100 mbar) for 48 to 72 hours and was removed following chest X-ray control. Bacterial and cytological examination of the pericardial fluid and the histological examination of the pericardium specimen completed the procedure. The pain in the first postoperative days was alleviated with parenteral Tramadol and Diclofenac administration.

The average procedure time of MCPF was 31 minutes (min. 20, max. 50, SD 9,5) and the average hospital stay was 9.4 days (min. 7, max. 11, SD 1,3). Considering the type of the primary tumor, there were 11 lung cancer, 6 breast, no gastrointestinal, 2 haematologic malignancies. In 3 cases out of 22 parasternally fenestrated patients, a malignant origin was not verified.

For statistical analysis we used the “Statistica 9.0” software package (Stat Soft Inc. USA). The mean values and the standard deviation (SD) were calculated for age, operation time, hospital stay and survival. The confidence interval was 95%. The Kaplan-Meier survival curve was made also with the Statistica 9.0 software package. The result was controlled with the “Matlab R2010b” software package (MathWorks Inc. USA). This calculation also shows the lower and upper confidence bounds for the cumulative distribution of survival (Figure 4). These bounds were calculated using Greenwood’s formula. In the cases of those two patients who were still alive at the end of the examined period, the survival times were used in the calculation and are shown in the curve as censored data. The comparison of survivals was analyzed with Student’s two-sample t-test. The critical value of “t” was given at a level of 95% (significance level: p < 0.05) with n1 + n2-2 degrees of freedom and was compared with the calculated “t” value. The statistical analyses were calculated with an assumed normal distribution. Due to the low number of patients, the result is only of informative value and would require an even larger test sample.

There were no operative deaths in the group of 22 MCPF patients. Transient intraoperative dysrhythmia was detected in 3 cases (14%). We lost one patient (4.5%) in the postoperative hospital stay. This end-stage cachectic lung-cancer patient died due to a cascade of inflammations. The mean hospital stay of our patients was 9.4 days (SD 1.3).

The postoperative pain was easily relievable in spite of the small partial subperiosteal parasternal rib resection. Usually after two to four days our patients did not require additional continuous therapy with painkillers because of postoperative pain. Our patients recovered without difficulty – considering their original complaints and physical incapacity caused by their malignant disease.

We detected transient recurrence of the pericardial fluid in one patient (4.5%) 14 days after the MCPF. The echocardiography revision after 24 hours did not detect an echolucent zone around the heart, and the recidive pericardial effusion disappeared spontaneously.

The demand for a definitive palliative treatment of neoplastic pericardial effusion led us to develop an alternative method of pericardial fenestration.

The aim of the present retrospective study was to publish the methodology of mediastinoscope-controlled parasternal fenestration of the pericardium (MCPF) that we have developed and applied. The pericardio-pleural fenestration with parasternal approach is an easy to perform, straightforward procedure. It is easy demonstrable, trainee-friendly, and does not require sophisticated manual skill. It requires no special tools. Single-lung ventilation is not necessary in the course of the operation. This strictly palliative method is minimally invasive, cost efficient, and provides immediate relief from complaints and symptoms caused by the pericardial effusion. The pericardial fluid produced empties into the pleural cavity across the pericardio-pleural window and will be absorbed on the bigger pleural surface. In our 10 year series of operations with 22 patients there was no prolonged recurrence of pericardial effusion, so the MCPF fulfilled what we had expected with a low hospital mortality rate.

Unfortunately, the idea presented is not exclusively new. A method not significantly different from ours was first described by Calvin in 1971. He suggested the parasternal mediastinotomy as a possible clinical application for pericardial fenestration [20]. We have found one article of a case report using the same approach for solving the problem of tension pneumo-pericardium [21] and another about pericardioscopy performed with the same armamentarium [22]. However our Pub Med search was unable to reveal any publication that discussed an application of parasternal pericardial fenestration in cases of neoplastic pericardial effusion.

The existing methods for solving the clinical consequences of pericardial fluid are well-known and safely performed. We do not offer our method instead of them!

However the malignant pericardial effusion and its relatively high recurrence rate is a special problem. The MCPF is a possible choice in such special cases. The mortality was low and the recurrence rate tended to zero in our patients treated with MCPF. Single-lung ventilation was not necessary in the course of the operation. The postoperative care and the mobilization of our patients were without difficulty and the postoperative pain was easily relievable in spite of the small partial subperiosteal parasternal rib resection. In our everyday practice we definitely prefer MCPF in cases of MPE to VATS or subxyphoideal fenestration.

On the other hand we do not think it a mistake to perform MCPF in non-malignant cases, but considering its advantages we recommend this “new-old method” especially for the definitive surgical palliation of pericardial effusion with malignant origin.

IT: MD, chief consultant surgeon and consultant general thoracic surgeon in Semmelweis Teaching Hospital, Miskolc, Hungary, GSz: MD, PhD, head of the Department of General and Thoracic Surgery, Semmelweis Teaching Hospital, Miskolc, Hungary, TFM: MD, PhD, professor, head of the Department of Thoracic Surgery, University of Pecs, Surgery Clinic, Pecs, Hungary.