Purulent and Constrictive Pericarditis Arising from a Staphylococcal Lumbar Infection

Abstract

A 78-year-old man was admitted to the hospital for evaluation and treatment of anasarca secondary to staphylococcal purulent pericarditis. One month earlier, he had undergone a lumbar laminectomy that was complicated postoperatively by a lumbar abscess that was not clinically apparent. The infection subsequently spread to the pericardium. Despite aggressive therapy that included therapeutic pericardiocentesis, drainage of pericardial fluid, and prolonged intravenous antibiotic therapy, the patient returned 2 weeks after discharge from the hospital with complicating constrictive pericarditis. Pericardiectomy was performed, resulting in complete relief of the patient's symptoms.

Purulent (or bacterial) pericarditis is a rare, life-threatening condition that is most often a result of the spread of a contiguous pulmonary, intracardiac, or chest wall infection. Herein, we describe the case of a patient in whom staphylococcal purulent pericarditis arising from a remote source was aggressively treated but was rapidly complicated by constrictive pericarditis.

Case Report

In July 2000, a 78-year-old man with hypertension, hyperlipidemia, and gastro-esophageal reflux disease was referred to our institution for surgical evaluation of chronic, debilitating back pain. Severe, symptomatic spinal stenosis was diagnosed, and a lumbar laminectomy was then performed. No notable postoperative complications were documented.

One month later, the patient presented at the hospital with body edema, associated shortness of breath, and easy fatigability. The edema had started 3 weeks earlier in his feet and had since progressed upward to his abdomen. The patient had no orthopnea, paroxysmal nocturnal dyspnea, chest pain, palpitations, or syncope. Review of his medical records revealed no history of hepatic or renal disease. The patient was admitted to the hospital for further evaluation.

Except for mild tachycardia, the patient had no unusual vital signs. Physical examination revealed anasarca, marked by 4+ pitting edema in both the upper and lower extremities and obvious edema of the abdominal wall. Evaluation of jugular venous pulses was hampered by both the patient's obesity and the anasarca. There was no evident pulsus paradoxus. Examination of the patient's surgical lumbar wound revealed some mild tenderness and a mildly purulent, slightly green discharge. All other sites of prior intravenous line insertions showed no signs of infection.

Laboratory tests revealed a white blood cell (WBC) count of 11,500/mm ³, a mildly left-shifted differential WBC count, a slightly elevated serum creatinine of 2.0 mg/dL, and no notable proteinuria. Results of a thyroid panel, screening collagen vascular disease profile, and electrocardiogram were normal. A chest radiograph, however, revealed mild pulmonary edema, small bilateral pleural effusions, and a cardiomediastinal silhouette larger than that seen on a chest radiograph 3 years earlier.

A transthoracic echocardiogram revealed a large circumferential pericardial effusion and a mitral inflow pattern that suggested early tamponade (Fig. 1). The patient underwent pericardiocentesis, which yielded approximately 800 cc of bloody fluid. Subsequent cultures of pericardial fluid, as well as cultures of fluid from the patient's lumbar wound, produced methicillin-sensitive Staphylococcus aureus. A peripherally inserted central catheter was placed, and the patient was discharged from the hospital on prolonged intravenous antibiotic therapy for the S. aureus infection. The antibiotic sensitivities of the 2 separate wound and pericardial isolates were subsequently found to be identical, suggesting a common focus of infection.

Fig. 1 A) Two-dimensional subcostal echocardiogram shows a large circumferential pericardial effusion. B) Pulsed-wave Doppler analysis shows respiratory variation of the mitral inflow pattern, with an excessive decrease in flow on inspiration.

Approximately 2 weeks after leaving the hospital, the patient presented again with recurrent shortness of breath, along with persistent edema of the lower extremities that was no longer improving. A repeat echocardiogram revealed near resolution of the peri-cardial effusion but also the development of para-doxical septal motion and a mitral inflow pattern consistent with constrictive pericarditis. Cardiac magnetic resonance imaging revealed a moderately thickened pericardium and an abnormal diastolic bounce in the mid and distal septum (Fig. 2). Subsequent right heart catheterization showed hemodynamics consistent with constrictive pericarditis (that is, diastolic pressure equalization, a right atrial M pattern with a prominent Y descent, and a right ventricular “dip and plateau” pattern) (Fig. 3). The patient underwent pericardiectomy, which resulted in complete relief of symptoms. Histopathologic analysis of the resected specimen revealed a constrictive pericardium marked by extensive fibrous thickening and fibrinous exudates.

Fig. 2 Gadolinium-enhanced magnetic resonance imaging (MRI) scan shows a thickened pericardium (arrows). The MRI cinematography reveals an abnormal diastolic bounce of the mid and distal septum consistent with constriction (arrowhead).

Real-time motion image is available at www.texasheartinstitute.org/ho283.html.

Fig. 3 Results of right heart catheterization. A) Right atrial hemodynamics reveal an elevation of right atrial pressures equal to that of left ventricular diastolic pressures. Note also the prominent Y-descent displayed in the right atrial pressure tracing. B) Right ventricular hemodynamics reveal a “dip and plateau” pattern consistent with pericardial constriction.

Discussion

Hematogenous spread of a remote infection to the pericardial space is unusual. Herein, we report a case in which a lumbar abscess disseminated hematog-enously to the pericardium, which subsequently led to purulent constrictive pericarditis.

Purulent pericarditis is a rare finding; proper diagnosis and treatment require a high index of suspicion. This potentially life-threatening condition most of-ten results from the contiguous spread of infection (often as a complication of pneumonia and concomitant empyema). It can also occur after chest surgery or chest wall infections. ¹ Purulent pericarditis is often diagnosed late in its course and, as in this case, when tamponade is threatening. The condition is formally diagnosed when purulent material is drained from the pericardial space or when bacteria are cultured from the pericardial fluid.

Before antibiotics, the pathogens Pneumococcus and Streptococcus, with their tendency to infect contiguous lung tissues, often gave rise to purulent pericarditis. However, the widespread use of antibiotics has changed the microbiologic spectrum of purulent pericarditis. Staphylococcus aureus (a species well known for infecting lumbar surgical wounds and for hema-togenous dissemination) is now the most frequent cause of purulent pericarditis. ¹ Other (less commonly implicated) species include gram-negative, fungal, and other atypical organisms. ¹ In a recent retrospective review of 15 cases of acute purulent pericarditis, Brook and Frazier ² noted the frequent involvement of anaerobic species.

In our patient, the purulent pericarditis was a consequence of S. aureus infection. In a previous series of patients with acute bacterial pericarditis, Majid and Oma ³ reported that 6 of 12 cases (50%) were secondary to S. aureus infection but that only 1 of those arose from a remote focus of septic osteomyelitis or septic arthritis. Sixteen years before that, Rubin and Moellering ⁴ had reported similar results in their series of patients with acute bacterial pericarditis: 8 of 26 cases (31%) were secondary to S. aureus infection, and most of those were, predictably, the result of contiguous or intracardiac spread.

Constrictive pericarditis is a well-known sequela of purulent pericarditis. Its onset is typically insidious, developing months or even years after an initial epi-sode of pericarditis. ⁵ Limiting their study to cases of constrictive pericarditis as a sequela of postsurgical pericarditis, Killian and colleagues ⁶ reported that the average interval between the original cardiac surgery and the diagnosis of constrictive physiology was 23.4 months (range, 1–204 months). In the present case, however, constrictive pericarditis was diagnosed only 2 weeks after the initial diagnosis and treatment of bacterial pericarditis (perhaps because of the virulence of the patient's staphylococcal infection). Other reported complications of staphylococcal pericarditis include the development of a mycotic aneurysm of the proximal aorta and metastatic septic emboli. ⁴

Despite aggressive treatment of purulent pericarditis (pericardiocentesis, drainage of the infected pericardial fluid, and prolonged intravenous antibiotic therapy), constrictive pericarditis often complicates the patient's condition. Some studies have suggested that prophylactic pericardiectomy be performed in patients with purulent pericarditis to prevent even-tual pericardial constriction. ³ Others have suggested that intrapericardial fibrinolysis be used to minimize the risk of constrictive pericarditis and to avoid, if possible, the necessity of pericardiectomy. ^7,8 Intrapericardial fibrinolysis might facilitate the dissolution of inflammatory and infectious substances and enhance antibiotic penetration. In accordance with that hypothesis, Defouilloy and coworkers ⁹ have recently reported the successful use of intrapericardial streptokinase, both prophylactically and acutely, in preventing constriction as a complication of bacterial pericarditis. ⁹