Discussion:
The etiology of pericardial effusion and cardiac tamponade is often idiopathic. Cubero et al states the three major causes of large pericardial effusions were idiopathic pericarditis, tuberculous, and neoplastic origin in descending order of incidence. Similar results were found in Sagristà-Sauleda et al, attributing the most common cause of large pericardial effusions to idiopathic pericarditis, followed by iatrogenic, and malignant with the latter having the highest likelihood of becoming cardiac tamponade [9,10]. Among the majority of data pertaining to etiology, drug-induced pericardial effusion seems to be either excluded or dismissed under the umbrella of idiopathic.
Management of pericardial effusion and cardiac tamponade is case-dependent and essentially remains the same with either pericardiocentesis, pericardial window or pericardiectomy. However, rare etiologies in certain clinical scenarios may warrant a different approach to management. The progression of a pericardial effusion into a cardiac tamponade could impose catastrophic clinical outcomes, especially in patients with co-morbid conditions such as malignancy that may hinder any invasive intervention and augment the hemodynamic instability of cardiac tamponade [11,12]. Markiewicz et al found that predictors of long-term survival relied on the prognosis of the primary illness, regardless of the treatment modality. This poses a challenge since the majority of pericardial effusion and cardiac tamponade cases are attributed to an idiopathic etiology, as previously mentioned.
The diagnosis of cardiac tamponade continues to rely on echocardiography, which apart from identifying the specific signs, also aims to evaluate the size, distribution, presence of intrapericardial adhesions or clot, the feasibility of pericardiocentesis, and the evolution of the pericardial effusion. Sanjay et al concluded that the sensitivity, specificity and positive predictive value of RV diastolic collapse is 93%, 100% and 100%, respectively. On the other hand, Rifkin et al states that although sensitivity of RV diastolic collapse increased in relation to the severity of cardiac tamponade, its specificity remained higher and constant at 88% whether the tamponade was early, moderate or severe. RV collapse takes place in early diastole, and as the cardiac tamponade progresses, it could span the entirety of diastole [15-17]. Although our patient’s echocardiography showed tamponade physiology, the classic beck’s triad of muffled heart sounds, presence of JVD, and hypotension did not completely manifest and the patient’s only recurrent complaint was shortness of breath requiring minimal oxygen supplementation by nasal cannula. However, decompensation risk remained very high. In practice, the clinical symptoms and signs of cardiac tamponade are a good compass of suspicion, but they are not reliable without echocardiographic confirmation, especially since they could be present in other overlapping conditions such as acute or chronic obstructive airway disease, pulmonary embolism, tension pneumothorax, large pleural effusions, effusive constrictive pericarditis, RV infarction, restrictive cardiomyopathy, extreme obesity, or tense ascites. In addition, the combination of classic and non-classic cardiac tamponade signs and symptoms are most likely to be absent in gradually developing cardiac tamponade [18]. Thus, given our patient’s sustained and relative hemodynamic stability, her pericardial effusion most likely developed in a gradual manner over days to weeks following the change in her ET regimen by the introduction of anagrelide, allowing the pericardium to stretch and accommodate the large effusion. The only identifiable event was the discontinuation of hydroxyurea and the initiation of anagrelide three weeks prior to presentation, which is consistent with enough lead-time for a gradual, large pericardial effusion to develop and be reflected with the patient’s progressive feeling of being unwell since the change. It is important to note that the patient had previously been placed on an anagrelide trial with an intolerant outcome described by the patient as pleuritic chest pain, shortness of breath and palpitations. The recent switch to anagrelide was a re-trial by the patient’s primary hematology team in an attempt to reevaluate tolerability in order to reduce her sustained, hydroxyurea-refractory thrombocytosis at the time.
Pericardial effusion was reported as a rare adverse reaction of anagrelide during its trials. Its incidence among its current users is unknown and the mechanism behind it is unclear [13]. In conditions indicated to receive anagrelide such as our patient’s ET, management of drug-induced cardiovascular events becomes a challenge in the setting of withdrawing the insulting medication. ET is known to put patients at risk for both thromboembolic as well as bleeding events. The thromboembolic events stem from abnormalities in platelet morphology, including its membrane, as well as derangements in arachidonic acid metabolism, and the bleeding nature of the disease is a result of qualitative platelet dysfunction as well as acquired Von Willebrand disease secondary to diminished large Von Willebrand multimers [8, 13].
Thus, at baseline, ET patients are at a higher risk for cardiovascular thromboembolic events in addition to a low threshold of bleeding, potentially complicating the course of, in our case, pericardial effusion and demonstrating an obstacle in the face of required invasive procedures. In chronic, stable pericardial effusions of large volume as well as cardiac tamponade, a complete benefit versus harm stratification that addresses both the myeloproliferative disorder and the cardiovascular pathology must be considered before initiating a medical decision concerning either one. Our case represents a complex scenario with multidisciplinary collaboration that could become more frequent as anti-platelet aggregation therapies become the dominantly preferred approach to hematologic disorders in patients at risk of cardiovascular disease and its escalation.