A case report of ventricular dysfunction post pericardiocentesis: stress cardiomyopathy or pericardial decompression syndrome?

We report a case of transient biventricular dysfunction post therapeutic pericardiocentesis, with classic features of stress cardiomyopathy (SCM). In our patient, the clinical and echocardiographic features were more in keeping with Takotsubo-type SCM than pericardial decompression syndrome (PDS). Our case is instructive in challenging our understanding of the aetiology of LV dysfunction complicating pericardiocentesis, and in highlighting the importance of careful clinical evaluation (altered heart rate and dyspnoea) in suspecting acute LV dysfunction after initial clinical improvement with pericardial aspiration. Electronic supplementary material The online version of this article (doi:10.1186/s12947-015-0026-3) contains supplementary material, which is available to authorized users.


Background
We report a case of reversible biventricular dysfunction following successful pericardiocentesis with classic features of stress or "Takotsubo" cardiomyopathy (SCM). Reports of SCM after pericardiocentesis are rare [1], as distinct from so-called pericardial decompression syndrome (PDS) which encompasses a spectrum of features of cardiac decompensation after large volume pericardiocentesis, including pulmonary oedema, adult respiratory distress syndrome, severe bi-ventricular failure and cardiogenic shock [2]. Our case is instructive in challenging our understanding of the aetiology of LV dysfunction complicating pericardiocentesis, and in highlighting the importance of careful clinical observations (heart rate and dyspnoea) in suspecting acute LV dysfunction after initial clinical improvement with pericardiocentesis.

Case report
A 62-year-old male presented with progressive dyspnoea for 10 days. He had a background of stage IV metastatic non-small lung carcinoma treated for 6 months with non-cardiotoxic chemotherapy (carboplatin and gemcitabine), and recently commenced on target therapy (Erlotinib). Clinical examination revealed signs consistent with cardiac tamponade, including significant pulsus paradoxus, tachycardia (heart rate 101), tachypnea (respiratory rate 25), elevated jugular venous pressure and muffled heart sounds. He was normotensive at 130/90mmHg. The patient was extremely anxious and spontaneously expressed concern about his imminent death.
His electrocardiogram (ECG) (Fig. 1) demonstrated electrical alternans and bedside transthoracic echocardiography (TTE) revealed a large pericardial effusion with features of cardiac tamponade, including diastolic compression of both right atrium and ventricle (Fig. 2, Additional file 1: Video 1 and Additional file 2: Video 2) and large mitral inflow variation (Fig. 3). Urgent pericardiocentesis was performed with a restricted aspiration of only 600 ml drained initially over the first hour, and a total drainage of 1.8 l of heavily blood-stained pericardial fluid over 36 h. During initial aspiration of pericardial fluid there was immediate symptomatic relief and haemodynamic improvement (heart rate [HR] decreased to 80/min, respiratory rate [RR] decreased to 15 breaths/min and BP increased to 150/70 mmHg).
Overnight (9 h post procedure) the patient developed chest discomfort, dyspnea, tachycardia (HR 110) and tachypnoea (RR 24). TTE the next morning showed no re-accumulation of pericardial fluid, but detected new severe impairment in function of both ventricles, with akinesis of the apex and peri-apical region (Figs. 4 and 5, Additional file 3: Video 3 and Additional file 4: Video 4). Biomarkers demonstrated a rise in highly sensitive troponin from 8 to 224ng/L, but creatinine kinase did not rise significantly (107 to 116U/L). ECG after chest pain demonstrated resolution of the electrical alternans, with new loss of R waves in the anterior leads (Fig. 6).
Based on a presumptive diagnosis of SCM, angiotensin converting enzyme inhibitor and long acting betablocker were commenced, chemotherapy withheld and the patient discharged for early clinical and echocardiographic review. Serial follow up TTEs showed normalization of bi-ventricular function after two weeks (Figs. 7 and 8, Additional file 5: Video 5 and Additional file 6: Video 6), and restoration of R waves on subsequent  Apical four chamber view on initial presenation demonstating large pericardial effusion with tamponade causing compression of right heart chambers (red arrows). LV function was normal prior to pericardiocentesis ECGs ( Fig. 9). Subsequent computed tomography examination showed normal coronary arteries with a calcium score of zero and no evidence of LAD laceration or dissection.
The patient presented three months later with reaccumulation of pericardial effusion and tamponade. Therapeutic pericardiocentesis was performed with 500 ml of blood stained pericardial fluid drained immediately, with 1.9 L in total over 36 h. On this presentation he was relaxed and well adjusted in regards to his diagnosis. No LV dysfunction was detected on serial follow-up echocardiograms after the second pericardiocentesis ( Fig. 10).

Discussion
Our patient developed biventricular apical dysfunction following successful and judicious pericardiocentesis, with features typical of stress or "Takotsubo cardiomyopathy". The case is instructive for its comparison with PDS and the clinical pattern of initial improvement followed by deterioration respectively due to pericardial aspiration and myocardial pathology.
In light of the timing of onset of biventricular impairment immediately post procedure PDS is an important differential diagnosis. Other differentials such as laceration to the ventricle or left anterior descending (LAD) coronary artery were clinically unlikely. The former was  Accordingly, we reviewed the literature describing SCM and PDS. Whereas SCM has been rarely reported after pericardiocentesis, much has been published on PDS. The incidence of PDS or new left or right systolic dysfunction has been reported to range from 5 % to 36 % of patients post pericardiocentesis [3,4], especially after malignant pericardial effusions. Although the first case report of PDS in 1983 noted APO with preserved LV function [5], most subsequent reports describe severe impairment of left, right or bi-ventricular function, which may be segmental or global (Tables 1 and 2).
A number of mechanisms have been proposed to explain the pathogenesis of LV systolic dysfunction in PDS.  Acute withdrawal of exaggerated sympathetic drive during relief of tamponade may trigger paradoxical haemodynamic instability [5]. Mechanical, inter-ventricular volume mismatch may also contribute, with sudden relief of pericardial constraint leading to abrupt, disproportionate increase in RV volume and a paradoxical rise in pulmonary artery pressure, resulting in raised LV end diastolic pressure and transient left heart failure [5][6][7][8][9]. Others have proposed myocardial stunning from coronary perfusion mismatch with acute distension of cardiac chambers after decompression [6,10,11]. Taken together, it is likely that a combination of hormonal and mechanical pathophysiologic mechanisms contribute to LV dysfunction and the final clinical sequelae in PDS.
The classic echocardiographic feature in SCM is transient LV apical ballooning, although other segmental patterns have been described [12,13]. A stressor leading to sympathetic overdrive and excessive catecholamine release is the currently accepted trigger in the development of SCM [12]. The catecholamine surge precipitates 1) 'peripheral arterial vasospasm leading to increased afterload and transient increase in LV end-systolic pressure' , 2) 'acute multiple coronary artery vasospasm leading to myocardial ischaemia' , and 3) direct catecholamine-       β-adrenoceptor -mediated myocardial stunning in the apex [14]. These three pathophysiologic pathways are thought to contribute to the ischaemia, morphologic features and potential haemodynamic sequelae that can be seen in SCM.
More recent case reports have made reference to LV apical ballooning related to PDS as similar to SCM [10,11,[15][16][17], and have postulated the physiological stressor being cardiac tamponade along with emotional stress [16]. It is therefore possible that the transient ventricular systolic dysfunction in PDS is actually a variant form of stress cardiomyopathy. We carefully reviewed 25 cases of heart failure post pericardiocentesis in the literature (Tables 1 and 2), and we believe that seven cases (two considered to be SCM [1,18] by the authors and five classified as PDS [10,16,17,19,20]) could be considered to have echocardiographic features of SCM.
SCM has relatively characteristic clinical presentation, with rise of cardiac enzymes [21], and often associated with ischaemic ECG changes (up to 44 % of those with SCM have T-wave inversion and 41 % ST elevation [13,21]). The clinical manifestations in PDS are more variable, ranging from asymptomatic in some to severe low cardiac output states in others. The primary clinical symptom in PDS has been reported as dyspnoea (Table 1). This is in contrast to chest pain being predominant in SCM (69-83 % of presentations) [13,21]. In the majority of cases of PDS in the literature ( Table 2) there was no cardiac enzyme rise, and ischaemic type changes on ECG were seen in a minority (seven of twenty five cases). In all the cases where ischaemic ECG changes where present except for one, there was concomitant apical and peri-apical regional wall motion abnormality, which could be classified as SCM also.
Generally SCM has a benign course, with recovery of LV function and good prognosis [12], whilst PDS has poorer outcomes and increased mortality [4]. Reports of PDS suggested normalization of LV dysfunction in 12 of 25 cases classified as PDS. Of the 12 cases that did recover LV function, four had LV impairment with classic SCM pattern of LV impairment on echocardiogram [16,17,19,20]. The normalisation of LV function in our patient 2 weeks subsequently is more in keeping with SCM.
Current literature has not specifically addressed risk factors for the development of ventricular dysfunction after pericardiocentesis. In our patient, the malignant nature of the effusion, the presence of tamponade and larger size of pericardial effusion [4], may have increased his predisposition to develop ventricular dysfunction. Amount and rate of fluid removed on initial decompression are also associated with development PDS [4,5], however there are no guidelines regarding the maximum amount of pericardial fluid that can be drained immediately. There is consensus to stop initial drainage with improvement of symptoms or hemodynamic parameters, followed by gradual decompression through indwelling catheter [5].
Our patient's apical systolic dysfunction post pericardiocentesis was associated with chest discomfort, transient loss of R waves and rise in cardiac enzymes are typical of classic SCM. The clinical sequence of HR and RR improving immediately post decompression and then increasing again hours after the procedure, was a useful clinical marker of myocardial dysfunction, prompting investigation which identified new ventricular impairment. It is likely that the frequency of transient LV dysfunction is underestimated in these patients.

Conclusion
We report a case of transient biventricular dysfunction post pericardiocentesis, with classic features of SCM. LV dysfunction post pericardiocentesis and in PDS is more prevalent than previously thought, and some previous reports of PDS may also be potentially considered as SCM complicating pericardiocentesis. In addition to judicious and gradual decompression to avoid ventricular dysfunction or PDS, patients undergoing therapeutic pericardiocentesis should have careful haemodynamic monitoring, as changes in parameters such as heart rate and respiratory rate can raise suspicion of acute LV impairment.

Consent
Written informed consent was unable to be obtained from the patient for publication of this Case report and any accompanying images, as he has passed away. His next of kin are not contactable after their subsequent return to their home country of China. Professor L. Kritharides, Head of Department, approves the publication of this report, with all patient identifiers kept confidential and material presented solely for educational purposes arising from the clinical encounter.

Additional files
Additional file 1: Video 1. Apical four chamber view demonstrating large pericardial effusion on presentation, with tamponade and diastolic compression of right heart chambers. Note the LV function is normal.
Additional file 2: Video 2. Parasternal long view demonstrating large pericardial effusion on presentation, with tamponade and diastolic compression of right heart chambers. Note the LV function is normal.
Additional file 3: Video 3. Apical four chamber view performed on the following morning after the patient experienced chest discomfort. The main finding is severe biventricular impairment, with akinesis of the apex and periapical areas. There is small residual pericardial effusion.
Additional file 4: Video 4. Parasternal long view performed on the following morning after the patient experienced chest discomfort. The main finding is severe biventricular impairment, with akinesis of the apex and periapical areas. There is small residual pericardial effusion.