Community-acquired MRSA pericarditis

, , , , ,
Cardiology Review® Online, January 2008, Volume 26, Issue 1

Only 2 cases of community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) pericarditis have been reported in the English literature. Over the last 15 years, CA-MRSA has emerged as an increasingly common pathogen that is genetically and epidemiologically different from hospital-acquired MRSA (HA-MRSA).

Staphylococcus aureus

S. aureus

Methicillin-resistant (MRSA) is a specific strain of that has developed antibiotic resistance to beta-lactams such as methicillin, oxacillin, nafcillin, and cephalosporins. The first reports of MRSA occurred in the early 1960s, when MRSA was primarily a nosocomial infection occurring in healthcare-associated facilities (to be termed hospital-acquired MRSA or HA-MRSA for the balance of this paper).1 This resistance has become increasingly prevalent worldwide as both a nosocomial and a community-associated pathogen. Community-acquired MRSA (CA-MRSA) tends to cause illness in people who have not had a medical procedure (such as surgery, hospitalization, or catheter placement) or been hospitalized within the past year. Community-acquired MRSA currently accounts for up to one third of adult MRSA infections, and frequently presents as skin and soft-tissue infections, but patients may also develop pneumonia, septicemia, endocarditis, or osteomyelitis.2 Infection of the pericardium with MRSA is extremely rare; there have been only 5 reported cases. We report a case of CA-MRSA infection of the pericardial sac in a young woman with a history of intravenous (IV) drug use; this is only the third reported case of CA-MRSA pericarditis.

Case history

A 34-year-old African-American female was initially admitted to a local hospital, where she presented with complaints of intermittent, nonexertional, nonradiating left-sided chest pain for 3 months. A high-resolution computed tomography image of the thorax showed evidence of pulmonary embolus, and the patient was started on anticoagulation. The patient left the hospital against medical advice the day after admission. During the interim she ingested cocaine and heroin and was re-admitted that same night with continued complaints of chest pain and shortness of breath. On re-admission the patient was febrile with a leukocytosis. She was empirically started on piperacillin-tazobactam and levofloxacin, and was continued on the heparin drip for pulmonary embolus. Blood, urine, and sputum cultures were collected and showed no growth of bacteria. The patient had a left upper-extremity antecubital infection at a previous IV access site that was also cultured and grew MRSA. The initial antibiotics were stopped and the patient was started on vancomycin. A transthoracic echocardiogram (TTE) showed normal left ventricular size and function, left ventricular hypertrophy, moderate pericardial effusion, and no evidence of vegetations. The patient was transferred to our institution on her seventh day of admission.

On transfer, the patient complained of persistent sternal chest pain that worsened with movement. It was described as being sharp in nature, 10/10 in intensity with no radiation. She had no associated diaphoresis, nausea, or vomiting. The patient reported that she was having mild episodes of chest pain over the past couple of months, which were relieved by ibuprofen. On exam she was febrile with a temperature of 101°F, tachycardic, and with a heart rate of 114 beats per minute (BPM). No jugular venous pulsation was appreciated; distant heart sounds were heard on cardiac exam with a regular rhythm and normal S1 and S2. Her respiratory exam was clear to auscultation with poor inspiratory effort and shallow breaths. An initial electrocardiogram (ECG) noted a heart rate of 120 BPM with PR depression, diffuse ST-segment elevation, and electrical alternans. A repeat ECG done a few days later showed diffuse T-wave flattening.

Figure.Transthoracic echocardiogram demonstrating a large effusion with fibrinous

strands within the pericardium.

The patient was admitted to the cardiac intensive care unit (ICU) and was treated for presumed pericarditis with ibuprofen. A repeat TTE was obtained, which showed a large pericardial effusion with evidence for hemodynamic compromise including subtle diastolic flattening of the right ventricle, respirophasic change of the mitral inflow (>25%), and a dilated inferior vena cava (Figure). The patient was taken to the cardiac catheterization laboratory for pericardiocentesis, and 650 cc of serosanguineous fluid was removed from the pericardial space. The intrapericardial pressure was elevated both pre- and postcatheterization (25 mm Hg). A pigtail catheter was inserted and left in the pericardial space for further drainage. The patient tolerated the procedure well, and reported that her breathing and chest pain had improved significantly postprocedure.

The pathology report on the pericardial fluid showed no malignant cells and was negative for tuberculosis (TB) and fungal elements. The fluid contained predominantly acute inflammatory exudates and grew MRSA, the susceptibilities of which are shown in the Table. The susceptibility pattern is consistent with CA-MRSA.

Table. Sensitivities of MRSA.


Vitek susceptibility















Over the course of the next 2 days, the pigtail catheter drained an additional 300 cc of serosanguineous fluid (which was also sent for cytology), and 3 days later the catheter was removed. A TTE obtained 2 days after catheter removal showed minimal pericardial effusion and significant respirophasic variation, suggesting possible constrictive physiology. Given the TTE findings, we were concerned for TB as a possible cause for the patient's cardiovascular finding, and so she was placed in respiratory isolation. A TB quanteferon was sent, a purified protein derivative test given, and 3 sets of induced sputum were sent for acid-fast bacilli smear, and all tests came back negative. We continued to monitor the patient over the next few days and obtained another TTE on the seventh day postpericardiocentesis. Trivial pericardial effusion was seen with no significant respiratory variation of mitral flow. The patient had improved clinically, remained asymptomatic, and was hemodynamically stable prior to her discharge. She was treated with a 6-week course of IV vancomycin.


Streptococcus pneumoniae

S. aureus

S. pneumoniae

Bacteria gain access to the pericardium through 5 mechanisms: hematogenous spread, direct extension from an intrathoracic focus, extension from a suppurative subdiaphragmatic focus, direct infection from trauma or surgery, and extension from a myocardial focus. and are the most common bacterial pathogens. Both spread hematogenously from other sites of infection in the body; can also extend directly from a concurrent pneumonia. Nevertheless, bacteria seldom cause pericarditis. The etiology of acute pericarditis is viral or idiopathic in 9 of 10 patients.3 Bacteria, myocardial infarction, dissecting aortic aneurysm, trauma, neoplasm, radiation, uremia, cardiothoracic surgery, autoimmune disorders, and drugs combine to make up the remaining 10%.4 Infection from MRSA is even rarer, with only 5 cases reported in the English literature.5 Two of these have been associated with CA-MRSA; this patient represents the third reported case of CA-MRSA pericarditis.

S. aureus

Methicillin was first introduced in the early 1960s, and the first case reports of MRSA surfaced 1 year later.1 Since then, MRSA has become the leading cause of HA-MRSA infections in the United States, and more than 50% of all isolates in ICUs are now methicillin resistant.

In the 1980s, the first cases of MRSA infections were reported in Canadian and Australian aboriginals (ie, those with no exposures to hospitals or nursing home facilities), and so these cases were therefore said to be community-acquired. In the 1990s, CA-MRSA was recognized as an emerging cause of infection. Community-acquired MRSA differs from HA-MRSA epidemiologically, and has a predilection to infect hospitalized patients and different organ systems.2

S. aureus

The biological basis of MRSA is as follows. The mecA gene confers resistance to methicillin, and allows to grow in its presence. It resides on a mobile genetic element, the staphylococcal cassette chromosome (SCCmec), which is transmitted from bacteria to bacteria through horizontal transfer.6 With the 5 known MRSA clones, there are 5 known SCCmec chromosomes. Hospital-acquired MRSA isolates contain SCCmec 1,2,3 and are multidrug resistant. Community-acquired MRSA, on the other hand, contains SCCmec 4 and is resistant only to methicillin, other beta-lactams, and erythromycin. The organisms are susceptible to bactrim, clindamycin, and in some geographic regions, the fluoroquinolones.7

Community-acquired MRSA contains toxins that are unique; the most important of these is the Panton-Valentine leukocidine (PVL) gene, which induces pore formation in leukocyte membranes, leading to tissue necrosis.6,7 This gene explains the predilection of CA-MRSA to cause skin and soft-tissue infections, and explains why, in the rare cases in which it causes pneumonia, that necrotizing pneumonitis is the end result.

Most cases of CA-MRSA have been reported in patients younger than 2 years and in younger patients with no comorbidities.8 The mortality rate associated with CA-MRSA is also lower than that of HA-MRSA. This may be due to a more healthy patient population or to the fact that only rarely is sepsis reported.

Our patient was a young female with no comorbidities, and no history of prolonged hospital admission or exposure to a hospital. She did, however, use IV drugs, which served as a likely nidus for infection. In addition, she had an antecubital infection that could have served as the source.

In our review of the literature, we found a paucity of articles discussing MRSA pericarditis, and even fewer reporting CA-MRSA isolates in the pericardial fluid. A letter to the editor described 5 surgical patients who had died from MRSA sepsis in hospital.9 Postmortem exams revealed the presence of pericarditis, both purulent and nonpurulent. Although the letter does not mention whether pericardial cultures were obtained from these postmortem samples, we presume that the pericarditis followed directly from MRSA septicemia and could therefore be expected to contain the bacteria.

We found a single published English-language case report of HA-MRSA pericarditis in a patient following a protracted hospital course for necrotic-hemorrhagic pancreatitis.10 The patients described in both of these items have many common factors that relate to their developing MRSA pericarditis; namely, that their pericardial infections were preceded by an extended hospital stay with invasive surgery, presence of IV catheters, multiple courses of antibiotics, and the known presence of MRSA bacteremia.

Our case differs from these reports in that our patient had not been hospitalized nor had invasive medical procedures prior to her presentation. Further, she had negative blood cultures throughout her stay. The only known source for MRSA was a superficial antecubital skin wound, which is likely a sequela of her IV drug use or an IV line infection. Although we did not obtain genotyping on the isolate to evaluate for the presence of the PVL gene or a mecA subtype, the susceptibilities of the organism were consistent with that expected for a CA-MRSA isolate.

The scenario we present is similar to that of a case report in which a 20-year-old man developed CA-MRSA pericarditis as a complication of a mycotic pseudoaneurysm of the ascending aorta.11 Both patients were relatively young and otherwise healthy, but both had known skin infections that were the likely source of the MRSA. Both patients had no evidence of bacteremia during their hospital stay, and the pericardial fluid was serosanguinous rather than grossly purulent. A significant difference is that in the previously reported case, the patient's pericarditis was secondary to seeding of a pseudoaneurysm, which provided a concentrated source of the bacteria from which direct seeding of the pericardial sac could occur.

Our case is consistent with, and serves to highlight, the emerging profile of CA-MRSA as a virulent bacterium with a predilection towards aggressive skin infections in young, otherwise healthy patients. Community-acquired MRSA has not been traditionally associated with sepsis or severe disease. As the incidence of CA-MRSA continues to increase in the general population, this may change. Increased cognizance of its virulence and spectrum is therefore warranted.