Hemophilus influenzae purulent pericarditis in children

J

THoRAc CARDIOVASC SURG

79:933-936, 1980

Hemophilus influenzae purulent pericarditis

in children Diagnostic and therapeutic considerations Purulent pericarditis is an unusual complication of infection in infancy and has been associated with an extremely high mortality rate. Early diagnosis followed by combined antibiotic therapy and surgical drainage of the pericardium has markedly improved survival. Between April, 1975, and February, 1979, nine patients with purulent pericarditis secondary to Hemophilus influenzae type B were treated at the Oklahoma Children's Memorial Hospital. In every case signs and symptoms of congestive heart failure were present, and a pericardia I effusion was demonstrated by echocardiography and confirmed by pericardiocentesis. The organism was identified with countercurrent immunoelectrophoresis and antibiotic sensitivity determined by rapid beta lactamase assay. All patients were treated with a combination of parenteral antibiotics and open surgical drainage of the pericardium. There were no deaths and all patients demonstrated marked improvement following operation. Follow-up echocardiography revealed no evidence of pericardial effusion or signs of constriction in any patient.

James E. Cheatham, Jr., M.D., R. Nathan Grantham, M.D., Marvin D. Peyton, M.D., Webb M. Thompson, M.D., Eugene F. Luckstead, M.D., Jerry D. Razook, M.D., and Ronald C. Elkins, M.D., Oklahoma City, Okla.

Purulent pericarditis is an unusual complication of infection in infancy that may be secondary to contiguous spread or hematogenous dissemination. The mortality rate without treatment approaches 100% and the mortality rate from medical treatment or surgical treatment alone remains unacceptably high. '- 4 Early diagnosis with the use of echocardiography and pericardiocentesis followed by combined medical and surgical management has markedly improved survival. 3-6 Although several organisms including staphylococcus and meningococcus are known to cause purulent pericarditis, Hemophilus infiuenzae type B is being isolated with increasing frequency. 3, 6-8 This paper presents the findings in a series of consecutive patients having purulent pericarditis, all nine of whom had H. influenzae as the etiologic organism, to emphasize the diagnostic and From the Division of Thoracic Surgery and Pediatric Cardiology, The University of Oklahoma Health Sciences Center, Oklahoma City, Okla. Received for publication May 21, 1979. Accepted for publication Nov. 14, 1979. Address for reprints: Ronald C. Elkins, M.D., Division of Thoracic Surgery, P. O. Box 25606, Oklahoma City, Okla. 73125.

surgical considerations in the management of this potentially lethal disease process. Patients Between April, 1975, and February, 1979, nine patients with purulent pericarditis secondary to H. influenzae were treated at the Oklahoma Childrens Memorial Hospital. Their ages ranged from 5 weeks to 4 years. Seven (78%) of the patients were less than I year of age, six (67%) of the patients were girls, and seven (78%) of the patients were white. All of the patients had a history of a prodromal illness 6 weeks to 2 days before the onset of symptoms (Table I). In addition, two patients had associated meningitis. Four of the patients had been treated previously with antibiotics. There were no patients with either chronic illness or a known immunologic disorder. Physical findings at the time of admission are listed in Table II. Tachypnea, tachycardia, and hepatomegaly were present in all patients. Six patients were febrile and five patients had muffled heart sounds. Two patients demonstrated a pericardial friction rub, No patients had pulsus paradoxus. Diagnostic findings are shown in Table III. All pa-

0022-5223/801060933+04$00.4010 © 1980 The C. V. Mosby Co.

933

934

The Journal of Thoracic and Cardiovascular Surgery

Cheatham et al.

Table I. Preceding illness in nine patients Illness

I

No. ofpatients

Upper respiratory tract infection Cellulitis/abscess Pneumonia Urinary tract infection Otitis

Technique

5 2 2 I I

Findings

No. ofpatients

Tachypnea Tachycardia Hepatomegaly Fever Muffled heart sounds Pericardia! friction rub Diarrhea Pulsus paradoxus

9 9 9 6 5 2 2

8 I 7 2 3

Table V• Follow-up data on nine patients

I

--------

Follow-up

o

Table III. Diagnostic findings in nine patients

1. Chest roentgenogram: Cardiomegaly Pleural effusion Pneumonia 2. Echocardiogram: Effusion (echo dense) 3. ECG: Elevated ST segment Decreased voltage Left ventricular hypertrophy 4. Pericardiocentesis: Positive 5. C1E-done in eight cases: H. infiuenzae type B 6. Beta lactamase s-done in seven cases: Negative (ampicillin sensitive)

No. of patients

1. Approach: Left anterior thoracotomy Subxiphoid 2. Procedure: Pericardiotomy with tube Partial Pericardiectomy with tube Thoracostomy tubes in addition to mediastinal tubes

Table II. Physical findings in nine patients on admission

Findings

Table IV. Surgical technique employed in nine patients

I No. of patients 9 3

2 9 4 3 2

9 8 7

tients had roentgenologic evidence of cardiomegaly and three had pleural effusions. In each case there was echocardiographic evidence of pericardial effusion, which was described as echo dense. All patients had abnormal electrocardiograms, with ST-segment elevation in four patients and decreased voltage in three patients. Periocardiocentesis yielded purulent material in all patients. Countercurrent immunoelectrophoresis was positive for H. infiuenzae in all eight cases in which it was performed. This was confirmed in each case by positive cultures obtained from blood, cerebrospinal fluid, and pericardial fluid. No patient had

1. Length: Longest 2 yr, 4 mo Shortest 6 mo Average I yr, 5 mo 2. Chest roentgenogram: Within normal limits Mild cardiomegaly 3. Echocardiogram: Signs of constriction Effusion 4. ECG: Normal

No. ofpatients

7

2

o

o 9

ampicillin-resistant H. infiuenzae as tested by the rapid beta lactamase assay. All patients were treated with a combination of parenteral antibiotics and open surgical drainage. Ampicillin and chloramphenicol were used initially in seven patients. One patient received ampicillin plus gentamicin and another received ampicillin plus oxacillin. Secondary antibiotics were stopped as soon as sensitivities were obtained. Table IV outlines the techniques used in surgical drainage. Eight patients underwent a limited left anterior thoracotomy with resection of a portion of the fifth intercostal cartilage. Drainage was carried out through a subxiphoid incision in the other patient. Careful exploration of the pericardium to remove adhesions and break up loculated areas was performed in all cases. Pericardial drainage by tube pericardiostomy was performed in seven of the patients with a left tube thoracostomy performed in one patient for drainage of a pleural effusion. Two patients had thick, fibrinous material within the pericardium that required partial pericardiectomy (phrenic nerve to phrenic nerve), removal of epicardial peel, placement of pericardial tube, and left tube thoracostomy, for a left pleural effusion.

Volume 79 Number 6

Purulent pericarditis in children

93 5

June, 1980

All patients had marked improvement in the symptoms of cardiac decompensation and sepsis following operation and institution of appropriate intravenous antimicrobial therapy. There were no deaths and all patients were discharged within 3 weeks. Postoperative follow-up, which averaged 1 year, 5 months, reveals all patients to be asymptomatic with normal electrocardiograms. Seven patients had normal chest roentgenograms and two had mild cardiomegaly. Echocardiograms showed no residual effusion and no evidence of constriction (Table V). Discussion

Early diagnosis and treatment with appropriate antibiotics combined with surgical drainage of the pericardium are imperative in the management of children with purulent pericarditis. Diagnosis is difficult because the classic signs of pericarditis may be absent. I, 2 The importance of a high index of suspicion cannot be overemphasized, as the signs of cardiac decompensation from any cause are frequently indistinguishable from those of cardiac tamponade due to purulent pericardial fluid. Congestive heart failure from congenital heart disease, myocarditis, or cardiomyopathy may present with findings similar to pericardial tamponade. Pulsus paradoxus was a prominent physical finding in the series of infants reported by Gersony and McCracken 2 and Symbus and associates," but this finding was not present in any of our patients. Likewise, pericardial friction rub, the hallmark of pericarditis, was not a consistent finding in our patients nor in those reported by Okorama, Perry, and Scott. 1 Although all electrocardiograms were abnormal, there were no diagnostic electrocardiographic changes. Hepatomegaly, tachypea, and tachycardia were found in all patients in this series as weII as cardiomegaly on chest roentgenography. These findings occurring in a septic patient should arouse the suspicion of effusive pericarditis. Echocardiographic diagnosis of septic pericarditis has been reported." This simple noninvasive technique permits rapid evaluation of the pericardial space without risk to the patient. The echocardiogram may suggest a purulent process because of the increased density of the effusion, but pericardiocentesis should be performed to ascertain the character of the fluid. Identification of the causative organism has in the past depended on Gram stain and culture, which usuaIIy requires 24 to 48 hours for reliable results. In our series, the technique of countercurrent immunoelectrophoresis (CIE) as reported by Shackelford, CampbeII, and Feigin'? was used to rapidly identify the infective

organism. This technique, which requires only a few hours, can be applied to any body fluid. At our institution, the immunoprecipitin technique employing antesera for H. influenzae type b, Neisseria meningitidis, Streptococcus pneumoniae, and beta hemolytic Streptococcus is used. The only organism reported to cause purulent pericarditis that would be missed by this method is Staphylococcus. There are occasional false negative results with CIE, so that all patients should also have cultures performed. 11 A cross-reactivity between H. influenzae type b and Escherichia coli has also been reported;" Ampicillin sensitivity of H. infiuenzae can be rapidly determined by the rapid beta lactamase assay. 13 Although purulent pericarditi s caused by ampicillin-resistant H. infiuenzae has not been reported, it is recommended that chloramphenicol in addition to ampicillin be given until the antimicrobial sensitivity of the organism is obtained, because of the appearance of increasing numbers of ampicillin-resistant H. influenzae type b organisms in patients with various other types of H. influenzae disease.!' The importance of surgical drainage of the pericardium in addition to antibiotic therapy of purulent pericarditis has been emphasized, but the method of pericardial drainage remains controversial. 1-3, 5. 9. 15 Pericardiocentesis often gives relief of pericardial tamponade, with improvement in the patient's condition, and has been used by some authors as a definitive drainage procedure.v 16. 17 Reports of early pericardial constriction associated with intrapericardial abscess and loculation of purulent material have led us to perform pericardiostomy in all our patients. 18 - 20 It is doubtful whether repeated pericardiocentesis would have aIIowed sufficient drainage in our patients who required partial pericardiectomy. Dai and Pate " reported only one death in their series with the use of a subxiphoid approach; however, Symbas and associates" preferred a left thoracotomy with transpleural drainage following a partial pericardiectomy. Our choice of a limited left anterior thoracotomy aIIows for the simplicity of the subxiphoid approach without compromising the operative exposure for thorough exploration of the pericardial space and for pericardiectomy, if necessary. The incision is also weII tolerated by these ill children. It has been our preference, for those patients who have only liquid pus in the pericardium, to drain the fluid through a pericardiostomy tube and to perform partial pericardiectomy in patients who have fibrinous adhesions or multiple loculated abscesses within the pericardial space, because we feel that constrictive pericarditis is more likely to develop in these patients. Acute constrictive pericarditis has been reported to

The Journal of

93 6 Cheatham et al.

Thoracic and Cardiovascular Surgery

occur following purulent pericarditis. 18-20 Previous antibiotic therapy, corticosteroids, and delay in diagnosis have been implicated as causative factors. 18 Four of our patients had received previous antibiotic therapy and none had received corticosteroids. Two of our patients clearly had a fibrinous form of the disease, but nothing in their preoperative courses explained this finding. Because of the potential of developing this complication, all patients should be observed closely for several years. Echocardiography is an excellent means of detecting evidence of constriction, and postoperative echocardiograms have been obtained in all of our patients without signs of constriction. 4 With the treatment regimen just described, an excellent therapeutic result was obtained in all nine patients. They are all presently asymptomatic without any restriction in daily routine. Conclusion H. infiuenzae type B purulent pericarditis continues to carry a high mortality rate if not appropriately diagnosed and managed. Any septic child with evidence of cardiac decompensation must be viewed with a high index of suspicion for the possibility of purulent pericarditis. Diagnosis is made with echocardiography, pericardiocentesis, and countercurrent immunoelectrophoresis (eIE). Antibiotic therapy should be initiated, and then open surgical drainage should be established through a limited left anterior thoracotomy with thorough exploration of the pericardial space and pericardial drainage. This series demonstrates that if these steps are followed, the mortality rate from this disease process can be markedly improved. REFERENCES Okorama EO, Perry LW, Scott LP: Acute bacterial pericarditis in children. Report of 25 cases. Am Heart J 90:709-713, 1975 2 Gersony WM, McCracken GH Jr.: Purulent pericarditis in infancy. Pediatrics 40:224-232, 1967 3 Benzing G III, Kaplan S: Purulent pericarditis. Am J Dis Child 106:289-294, 1963 4 Pieroni DR, Park SC, Holbrook PR, Houghton PB: Echocardiographic diagnosis of septic pericarditis in infancy. J Pediatr. 82:689-691, 1973

5 Rubin RH, Moellering RC Jr: Clinical microbiologic and therapeutic aspects of purulent pericarditis. Am J Med 59:68-78, 1975 6 Echeverria P, Smith EWP, Ingram D, Sade RM, Gardner P: Hemophilus inftuenzae B pericarditis in children. Pediatrics 56:808-818, 1975 7 Crossley K, Bieos T, Joffe CD: Hemophilus influenzae pericarditis. A report of two cases in adults with a summary of the literature. Am Heart J 85:246-251,1973 8. Gotoff SP, Fousek MD: Purulent pericarditis and cardiac tamponade due to Hemophilus infiuenzae type B. J Pediatr. 61:576-581, 1962 9 Symbas PN, Ware RE, DiOrie PA, Hatcher CR Jr: Purulent pericarditis. A review of diagnostic and surgical principles. South Med J 67:46-48, 1974 10 Shackelford PG, Campbell J, Feigin RD: Countercurrent immunoelectrophoresis in the evaluation of childhood infections. J Pediatr 85:478-481, 1974 II Ingram DL, Anderson P, Smith DH: Countercurrent immunoelectrophoresis in the diagnosis of systemic diseases caused by Hemophilus inftuenzae type B. J Pediatr 81:1156-1158, 1972 12 Lampe RM, Chottipitayasunondh T, Sunakorn P: Detection of bacterial antigen in pleural fluid by counterimmunoelectrophoresis. J Pediatr 88:557-560, 1976 13 Thornesberry L, Kirven LA: Ampicillin resistance in Hemophilus infiuenzae as determined by a rapid test for beta lactamase production. Antimicrob Chemother 6: 653-654, 1974 14 American Academy of Pediatrics, Committee on Infectious Diseases: Ampicillin-resistant strains of Hemophilus infiuenzae type B, August, 1974 15 Dai ND, Pate JW: Purulent pericarditis in South Vietnam. Report of 16 cases. South Med J 67: 1306-1307, 1972 16 Santhanakrishnan BR, Baliga RK, Raju VB: Purulent pericarditis. Indian Pediatr 9:686-690, 1972 17 Weir EK, Joffe HS: Purulent pericarditis in children. An analysis of 28 cases. Thorax 32:438-443, 1977 18 Rubenstein 11, Goldblatt A, Daggett WM: Acute constriction complicating purulent pericarditis in infancy. Am J Dis Child 124:591-594, 1972 19 Strauss AW, Santa-Maria M, Goldring D: Constrictive pericarditis in children. Am J Dis Child 129:822-826, 1975 20 Caird R, Conway N, McMillan IKR: Purulent pericarditis followed by early construction in young children. Br Heart J 35:201-203, 1973