Pericardial tamponade caused by Actinomyces after bronchoscopy

November 24, 2008

Actinomyces odontolyticus isa rare cause of pleuropericardialinfection, with only 1 caseidentified in the literature. Inthat instance, the infectionwas believed to be secondaryto gastric surgery. We present apatient with pericarditis andpericardial tamponade causedby A odontolyticus. The infectionoccurred after an ultrasound-guided subcarinalbronchoscopic needle biopsyperformed for a suspicious mediastinalmass found on a CTscan of the chest. We describethe case presentation, the microbiologyand treatment of Aodontolyticus infection, andthe classic features of pericarditisand cardiac tamponade.

Actinomyces odontolyticus is a rare cause of pleuropericardial infection, with only 1 case identified in the literature. In that instance, the infection was believed to be secondary to gastric surgery. We present a patient with pericarditis and pericardial tamponade caused by A odontolyticus. The infection occurred after an ultrasound- guided subcarinal bronchoscopic needle biopsy performed for a suspicious mediastinal mass found on a CT scan of the chest. We describe the case presentation, the microbiology and treatment of A odontolyticus infection, and the classic features of pericarditis and cardiac tamponade.

 

THE CASE

A 50-year-old man with a history of Barrett esophagus presented to our emergency department (ED) with a chief concern of epigastric pain that had started at 3 AM on the day of admission. The patient described the pain as sharp and focal. It was exacerbated by inspiration and movement and was relieved by sitting forward. He also reported having a temperature of 38.8°C (102°F) for 1 week. He denied shortness of breath, nausea, vomiting, diaphoresis, and any recent history of sick contacts or travel.

Laboratory test results showed leukocytosis, with a white blood cell (WBC) count of 24.9 X 109/L (88.6% neutrophils). The patient's troponin and serum myoglobin levels were normal. A chest radiograph revealed no focal consolidation, infiltrate, or effusion and a normal cardiac silhouette. The ECG showed diffuse, concave ST-segment elevations (Figure), while the echocardiogram showed an ejection fraction of 60% (±5%) with concentric left ventricular (LV) hypertrophy, a normal LV chamber size, no evidence of systolic or diastolic dysfunction, and a small pericardial effusion.

Figure – This ECG shows diffuse, concave ST-segment elevations. These findings suggest pericarditis.

Acute pericarditis was diagnosed, and the patient was admitted for treatment and close observation. He was initially given a course of intravenous ciprofloxacin, 400 mg/d, and intravenous ketorolac, 30 mg every 6 hours. On the morning of the second hospital day, the patient became acutely short of breath, hypotensive, and tachycardic. His heart rate rose from 114 to 120 beats per minute, and his blood pressure dropped from 108/49 to 85/50 mm Hg. His oxygen saturation remained stable at 98% on room air. An emergent echocardiogram revealed a small LV cavity, normal right atrial size with an inferior vena cava that lacked inspiratory collapse, and a moderate-sized to large circumferential pericardial effusion.

The patient was immediately taken to the cardiac catheterization laboratory for emergent pericardiocentesis; 560 mL of cloudy, yellow fluid was removed, and a pericardial drainage catheter was placed. Ciprofloxacin was discontinued, and the patient was given vancomycin, 1000 mg/d, while the pericardial fluid was sent for bacterial and fungal culture, cell count, and cytology.

Cell count of the fluid revealed 120,000 WBCs and 2620 red blood cells, and cytology showed no malignant cells; bacterial culture grew Actinomyces odontolyticus. An infectious disease physician was consulted, and the patient's antibiotic therapy was switched to intravenous ceftriaxone, 2 g/d. Vancomycin and ketorolac were discontinued. Despite placement of the pericardial drainage catheter, the pericardial effusion continued to accumulate, which required a subxiphoid pericardial window to be performed.

Before his ED visit, the patient had been seen by his primary care physician for hematuria. As part of the initial workup, a CT scan of the abdomen and pelvis was ordered. The CT scan did not identify a cause of the hematuria but did reveal 2 indeterminant right lower lobe pulmonary nodules. A subsequent CT scan of the chest revealed an enlarged subcarinal lymph node with a borderline axillary lymph node and unchanged right lower lobe pulmonary nodules.

The patient was referred to a pulmonologist, and bronchoscopy with ultrasound-guided subcarinal lymph node biopsy was performed. This was done 3 weeks before his ED visit. Cytology results were positive for malignant cells consistent with a metastatic adenocarcinoma.

Further imaging was performed to search for the primary focus of the suspected metastatic adenocarcinoma. A positron emission tomography scan showed no hypermetabolic activity in the right lower lobe nodules and no evidence of abnormal activity in the skeletal system, liver, or adrenal glands. An MRI scan of the brain showed 2 abnormal enhancing foci with a questionable third lesion suggestive of metastatic disease. Unfortunately, a primary tumor site was not identified.

The patient's current hospital course was complicated by atrial fibrillation with rapid ventricular response that required amiodarone and by persistent postoperative pleural effusions believed to be secondary to the pericardial window that eventually required bilateral pleurodesis. Despite this, the patient was discharged in stable condition and instructed to complete a 6-week course of ceftriaxone, 2 g/d. An echocardiogram obtained before discharge showed only a trivial amount of pericardial fluid.

We believe the patient's pericardial infection with A odontolyticus and subsequent pericardial tamponade were the result of direct inoculation during the bronchoscopic subcarinal needle biopsy.

DISCUSSION

Bronchoscopic biopsy
Although bronchoscopy and transbronchial needle aspiration (TBNA) can be used to sample mediastinal lymph nodes, its greatest utility is in the biopsy of subcarinal nodes. Aspiration of these nodes is safe and easy to perform, because the bronchoscope does not require flexion or extension during the procedure and because of the paucity of large vessels in this location.1,2 This method has been enhanced by the introduction of endobronchial ultrasonography, especially when it is done with a double-channel scope.3

The complication rate of TBNA is minimal when the proper technique is used and the appropriate precautions are taken.4 The most common complications are bacteremia, fever, oozing of blood, and inadvertent puncture of mediastinal structures.4 We believe that the latter occurred in our patient.

Bronchoscopy can cause infection in a number of ways, especially since the procedure involves a foreign body traversing many of the patient's natural defense barriers.5 A study by Srinivasan and associates6 revealed an increase in the incidence of Pseudomonas aeruginosa infections, which were found to be associated with loose biopsy port caps that allowed the organisms to be sheltered within the bronchoscopes and then to disseminate during procedures. Overall, however, such occurrences are rare.

Actinomyces odontolyticus
In our patient, cultures of the pericardial fluid grew A odontolyticus. This bacterial strain was first isolated by Batty7 in 1958 from persons with advanced dental caries. In ensuing years, 23 different cases have been reported by the CDC, with pulmonary, cardiopulmonary, and mediastinal sites of involvement. 8 Most of the affected patients were immunocompromised.

Only 19 cases of pericardial actinomycosis were reported between 1950 and 1990; in these cases, Actinomyces israelii was the main causative agent.9 There have been reports of intrathoracic infections caused by A odontolyticus, usually occurring in immunocompromised patients (Table 1). However, to our knowledge, there has been only 1 reported case of pericardial infection caused by A odontolyticus, which was hypothesized to be the result of laparotomy with surgical resection of a noninvasive gastric adenocarcinoma.10

A odontolyticus is an anaerobic, facultative capnophilic, gram-positive, nonmotile, non–acid-fast, nonsporulating bacterium. Rods may be short to medium length, resembling diphtheroids or propiobacteria. They may be arranged in palisades on the Gram stain.

These bacteria normally present as irregular, small, whitish colonies on blood agar that show a dark red pigment on maturation, which may take 2 to 14 days. A odontolyticus may be differentiated from other Actinomyces species when results reveal a negative oxidase, negative catalase, reduction of nitrate to nitrite, inability to grow at a pH of 5.5, and development of pink pigment on blood agar (after 7 to 10 days of incubation).

Most cases of infection caused by A odontolyticus resemble those caused by A israelii. They mainly affect the chest, pelvis, and abdomen and, less frequently, the bones, joints, and CNS. The disease predominantly affects men and occurs most frequently in the fifth decade of life. In 97% of 181 patients with actinomycosis, the clinical presentation included abdominal disease, abscess, dental disease, pulmonary disease, intracranial infection, or incidental swelling or mass.8

Actinomycosis usually responds to intravenous benzylpenicillin, 1.2 to 1.8 g every 3 hours for 3 to 6 weeks, followed by oral penicillin V, 2 to 4 g every day for 6 to 12 months. Other effective treatments include amoxicillin and ampicillin. In patients who have penicillin allergy, erythromycin, clindamycin, and chloramphenicol are reasonable alternatives.

Treatment usually results in complete resolution of the disease, but there may be some residual scarring and fibrosis of the affected tissue. More recent reports have suggested that intravenous agents such as ceftriaxone and imipenem are also effective and have the benefit of once-daily dosing and the ability to be given at home.11

Cardiac tamponade
The pericardium is a fibrous sac composed of 2 layers, the visceral and parietal pericardium. The visceral layer is made up of a single layer of mesothelial cells and adheres to the epicardial surface of the heart. The parietal layer is usually 2 mm thick and is composed of elastin and collagen fibers. The reflection of the visceral pericardium is located near the origins of the great vessels and is contiguous with the parietal pericardium. Usually, the pericardial space within these 2 layers contains 50 mL of serous fluid.

Any cause of pericardial effusion or hemorrhage into the pericardial space may result in tamponade (Table 2). One study revealed that tamponade occurred in 61% of patients with purulent, tuberculous, or neoplastic pericarditis and in 14% of patients with idiopathic pericarditis.12 It has been described in fewer than 1% of patients with acute myocardial infarction (MI) who were treated with thrombolytics and in even fewer patients who underwent percutaneous coronary intervention.12,13 Tamponade also may result from free wall rupture as a complication of acute MI and from persistent pericardial effusions after coronary artery bypass graft or valve repair.14,15

 

The list of possible bacterial causes of pericarditis is diverse. A odontolyticus should be considered in any patient with pericarditis who presents with a recent history of abdominal or thoracic surgery.

Cardiac tamponade is a life-threatening manifestation of the accumulation of pericardial fluid and must be recognized quickly to provide appropriate management. The history and physical examination are paramount. Dyspnea is the most common presenting symptom of cardiac tamponade. Other symptoms include cough, air hunger, anorexia, fatigue, and chest discomfort.

In rare cases, patients with tamponade have presented with repetitive yawning. The underlying pathophysiology of this is unknown. One theory suggests that yawning may reverse microatelectasis associated with low lung volumes from an expanding pericardium, while others argue that it is the result of irritation of the phrenic nerve.16

Physical findings include decreased intensity of heart sounds, elevated jugular venous pressures, and falling systolic blood pressure (Beck triad). Catheter readings of right atrial pressure may display a normal x descent with an attenuated y descent because of attenuated right ventricular (RV) filling during late diastole. Patients commonly present with tachycardia, although the absence of a rapid heart rate should not exclude the possibility of tamponade.

Patients with subacute pericardial tamponade caused by hypothyroidism may have bradycardia. These patients may also present with pulsus paradoxus (a drop in systolic blood pressure greater than 10 mm Hg with inspiration), which is the result of reduced chamber size of the left ventricle during inspiration. Inspiration causes increased negative intrathoracic pressure that results in increased venous return to the right atrium and, subsequently, to the right ventricle. This increased volume in patients with reduced ventricular compliance is caused by increased intrapericardial chamber pressures from fluid accumulation; it results in septal deviation into the left ventricle, decreasing end-diastolic volume and stroke volume with a subsequent reduction of cardiac outflow during inspiration.

It should be noted that the absence of pulsus paradoxus does not necessarily exclude tamponade17 and that many diseases may actually present with pulsus paradoxus other than tamponade (Table 3).

 

Cardiac tamponade is associated with multiple imaging and ECG abnormalities that may lead to its definitive diagnosis. The ECG may show a sinus tachycardia, low voltage that is defined as a QRS complex amplitude of less than 5 mm in all limb leads and less than 10 mm in all precordial leads, and electrical alternans defined as beat-to-beat variation of the QRS complex caused by the swinging of the heart in the pericardial fluid. The ECG also may display signs consistent with pericarditis, as was the case in our patient.

Chest radiographs may reveal an enlarged cardiac silhouette. However, this may not be present in acute tamponade where fluid accumulation may be minimal. Echocardiography is perhaps the most specific test used to confirm cardiac tamponade. The acquisition of an echocardiogram is an American College of Cardiology/American Heart Association (ACC/ AHA) class I recommendation; the ACC/AHA recommends echocardiography in the evaluation of any patient who has suspected pericardial disease.18 Findings that are suggestive of tamponade include collapse of the right atrium and right ventricle.

Right atrial collapse that occurs at end-diastole is caused by increased pericardial pressure and is highly specific and sensitive for pericardial tamponade when it persists through at least one-third of the cardiac cycle. RV end-diastolic collapse is more specific but less sensitive. In patients with cor pulmonale or RV hypertrophy, RV collapse may be absent because of elevated end-diastolic pressure.19

Left atrial collapse occurs in 25% of patients with tamponade and is a very specific finding. LV collapse is less common because of the increased muscularity of the myocardium. Echocardiography may also reveal less than 50% reduction in diameter of the inferior vena cava during inspiration, indicating an elevated central venous pressure.

The differential diagnosis of tamponade is limited and includes MI (especially involving the right ventricle), aortic dissection, and heart failure. These diagnoses should not be overlooked when considering the possibility of tamponade. An appropriate laboratory and imaging workup can easily help in differentiating between these diagnoses.

The approach to tamponade largely depends on the clinical status of the patient. In patients with tamponade who are hemodynamically stable, serial echocardiograms and careful monitoring may be performed. In patients who have hemodynamic compromise, a fluid bolus should be given to maintain an adequate blood pressure while preparing for urgent pericardiocentesis. These patients will also benefit from close monitoring in a cardiac critical care unit. Medications that have inotropic properties may be necessary to maintain adequate mean arterial pressure in these patients.

Severe cases, such as myocardial free wall rupture with subsequent hemorrhagic pericardial tamponade, require urgent surgical intervention. Surgery is also necessary when catheter pericardiocentesis does not adequately drain an active pericardial effusion.

Pericardiocentesis in the nonemergent setting is typically performed with the use of image guidance via echocardiography, fluoroscopy, or even CT, since this allows for proper visualization of the needle as well as the best anatomical site closest to the fluid. This may be done either at the bedside or, typically, in the catheterization laboratory if the diagnosis is uncertain.

Generally there are 2 approaches: paraxiphoid or apical. If the clinical picture warrants an emergent procedure and imaging is not readily available, the typical site of entry is via a paraxiphoid approach, where the needle is inserted between the xiphoid process and the left costal margin and directed toward the left shoulder.20

In the apical approach, the needle is aimed internally. Once the skin is penetrated, the needle is advanced until the pericardium is pierced and fluid is aspirated. A sheathed needle can be placed and, in the case of prolonged drainage, a pigtail catheter can be placed through this. Once output from the drain decreases to less than 25 to 50 mL/d, the catheter can safely be removed.

Antibiotic therapy is warranted in patients with purulent pericarditis. Empirical therapy should be broad-based and take into account the immune status of the patient as well as the environment in which the infection may have been contracted (nosocomial vs communityacquired). Both gram-positive and gram-negative bacterial pathogens should be covered. Once the infectious agent has been identified, treatment may be tailored to that particular pathogen.

SUMMARY

Pericardial complications associated with bronchoscopic biopsy are quite rare. However, the possibility of introducing oral cavity bacteria exists and, in the case presented here, we believe that A odontolyticus was introduced during the bronchoscopic needle biopsy. The presence of necrotic subcarinal tumor may have facilitated the seeding of the organism. The patient had classic features of pericarditis and pericardial tamponade. Prompt recognition and drainage with appropriate antibiotic coverage led to clinical improvement and the eventual discharge of this patient.

References:

REFERENCES


1.

Vansteenkiste J, Lacquet LM, Demedts M, et al. Transcarinal needle aspiration biopsy in the staging of lung cancer.

Eur Respir J.

1994;7:265-268.

2.

Oki M, Saka H, Kumazawa A, et al. The role of transcarinal needle aspiration in the diagnosis and staging of lung cancer: computed tomographic correlates of a positive result.

Respiration.

2004;71:523-527.

3.

Kanoh K, Miyazawa T, Kurimoto N, et al. Endobronchial ultrasonography guidance for transbronchial needle aspiration using a double-channel bronchoscope.

Chest.

2005;128:388-393.

4.

Dasgupta A, Jain P, Minai OA, et al. Utility of transbronchial needle aspiration in the diagnosis of endobronchial lesions.

Chest.

1999;115:1237-1241.

5.

Prakash UB. Does the bronchoscope propagate infection?

Chest.

1993;104:552-559.

6.

Srinivasan A, Wolfenden LL, Song X, et al. An outbreak of Pseudomonas aeruginosa infections associated with flexible bronchoscopes.

N Engl J Med.

2003;348:221-227.

7.

Batty I. Actinomyces odontolyticus, a new species of actinomycete regularly isolated from deep carious dentine.

J Pathol Bacteriol.

1958;75:455-459.

8.

Brown JR. Human actinomycosis. A study of 181 subjects.

Hum Pathol.

1973;4:319-330.

9.

Fife TD, Finegold SM, Grennan T. Pericardial actinomycosis: case report and review. Rev Infect Dis. 1991;13:120-126.

10.

Litwin KA, Jadbabaie F, Villanueva M. Case of pleuropericardial disease caused by Actinomyces odontolyticus that resulted in cardiac tamponade.

Clin Infect Dis.

1999;29:219-220.

11.

Read RC. Orocervical and esophageal infection. In: Cohen J, Powderly WG, eds.

Cohen & Powderly: Infectious Disease.

2nd ed. New York: Elsevier; 2004.

12.

Fasseas P, Orford JL, Panetta CJ, et al. Incidence, correlates, management, and clinical outcome of coronary perforation: analysis of 16,298 procedures.

Am Heart J.

2004;147:140-145.

13.

Patel MR, Meine TJ, Lindblad L, et al. Cardiac tamponade in the fibrinolytic era: analysis of >100,000 patients with ST-segment elevation myocardial infarction.

Am Heart J.

2006;151:316-322.

14.

Sanchez CE, Koshal VB, Antonchak M, Albers AR. Survival from combined left ventricular free wall rupture and papillary muscle rupture complicating acute myocardial infarction.

J Am Soc Echocardiogr.

2007;20:905.e1-3.

15.

Krantz MJ, Lee JK, Spodick DH. Repetitive yawning associated with cardiac tamponade.

Am J Cardiol.

2004;94:701-702.

16.

Meurin P, Weber H, Renaud N, et al. Evolution of the postoperative pericardial effusion after day 15: the problem of late tamponade.

Chest.

2004;125:2182-2187.

17.

Chiles CD, Menon V. Echocardiographic tamponade in severe left ventricular dysfunction: the impact of small pericardial effusion and the absence of pulsus paradoxicus.

J Am Soc Echocardiogr.

2004;17:78-79.

18.

Cheitlin MD, Armstrong WF, Aurigemma GP, et al. ACC/AHA/ASE 2003 guideline update for the clinical application of echocardiography-summary article: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/ASE Committee to Update the 1997 Guidelines for the Clinical Application of Echocardiography.

J Am Coll Cardiol.

2003;42:954-970.

19.

Leimgruber PP, Klopfenstein HS, Wann LS, Brooks HL. The hemodynamic derangement associated with right ventricular diastolic collapse in cardiac tamponade: an experimental echocardiographic study.

Circulation.

1983;68:612-620.

20.

Minai OA, Dasgupta A, Mehta AC. Transbronchial needle aspiration of central and peripheral lesions. In: Bolliger CT, Mathur PN, eds.

Progress in Respiratory Research-Interventional Bronchoscopy.

Basel, Switzerland: Karger; 1999.