Pericarditis, or inflammation of the pericardium, is most often caused by viral infection.

It can also develop as a result of bacterial or other infection, autoimmune disease, renalfailure, injury to the mediastinal area, and the effects of certain drugs (notably hydralazine and procainamide). The clinical features of pericarditis depend on its cause, as wellas the volume and type of effusion. Patients with uncomplicated pericarditis have pleu-ritic-type chest pain that radiates to the left shoulder and may be relieved by leaningforward. Chest radiographs, Doppler studies, and laboratory tests confirm the diagno-sis and provide information about the degree of effusion. In most patients, pericarditisis mild and resolves spontaneously, although treatment with a nonsteroidal anti-inflam-matory drug or a short course of a corticosteroid may be helpful. When a large peri-cardial effusion is produced, cardiac function may be compromised, and cardiac tamponade can occur. In patients with longstanding inflammation, the pericardiumbecomes fibrous or calcified, resulting in constriction of the heart. Drainage or surgicalintervention may be necessary in patients with complicated pericarditis. (Am Fam Physi-cian 2002;66:1695-702. Copyright 2002 American Academy of Family Physicians
Pericarditis is inflammation of thepericardial sac surrounding theheart and the origins of the greatvessels. The condition is mostoften caused by a viral infectionand generally resolves in a few weeks with nosequelae. Pathophysiology and Classification Depending on the underlying cause, theacute inflammatory response in pericarditismay produce serous fluid,pus,or dense fibrin-ous material. Viral pericarditis usually resultsin small accumulations of serous fluid thatresolve spontaneously or require minimaltherapeutic intervention. Even large volumesof serous fluid (up to 250 mL) may not cause significant clinical signs or symptoms if accu-mulation is gradual.The effusion may only beapparent as an enlarged cardiac silhouette onthe chest radiograph. Conversely,largeacuteaccumulationsofpericardialfluidmaycause intrapericardialpressure to rise,thereby impeding filling of theright side of the heart through the superiorvenacavaandinferiorvenacava. Acutely,thissituationcanresultincardiactamponade.Ifthepericarditisprocess continuesandthefluidorganizesintoathickened(evencalcified)coating,the resultantconstrictivepericarditismaymimicrestrictivecardiomyopathy. Thus,pericarditismaybeclassifiedasacute,subacute,or chronic,depending on the under-lyingpathophysiologicprocess . Hemodynamic Considerations Changes in the jugular veins and the rela-tionship of those changes to respiration can beextremely useful in differentiating pericarditisfrom other causes of chest pain Etiology of Pericarditis Idiopathic (nonspecific, probably viral)Infectious causesViruses: coxsackievirus A and B, hepatitis viruses, human immunodeficiencyvirus, measles virus, mumps virus, varicella virusBacteria: gram-positive and gramnegative organisms; rarely, Mycobacteriumspecies

(tuberculosis)Fungi (most often in immunocompromised patients): Blastomyces dermatitidis ,Candida species, Histoplasma capsulatum Noninfectious causesAcute myocardial infarction*Renal failure closed procedures and pacemaker implantation(puncture of myocardium)Drugs : hydralazine (Apresoline), procainamide (Pronestyl), phenytoin(Dilantin), isoniazid (e.g., Nydrazid); with rifampin (Rifamate),phenylbutazone, dantrolene (Dantrium), doxorubicin (Adriamycin, Rubex),methysergide (Sansert), penicillin, mesalamine (Rowasa) Pericardial effusion may occur on the second to fourth day after acutemyocardial infarction as a reaction to underlying necrotic myocardium in up to25 percent of patients. It can also develop weeks to months after myocardial infarction in patients with postmyocardial infarction syndrome (Dressler s syn-drome); in this situation, pericarditis occurs because of a late autoimmune reac-tion stimulated by the entry of necrotic myocardial tissue into the circulation,where it acts as an antigen. Pericarditis in uremia indicates a need for dialysis; pericarditis may also beassociated with hemodialysis. Listed in relative order of frequency. . descentduringventricular systole). During inspiration, lungexpansion causesnegativeintrathoracicpres-surethatistransmittedtothepericardium andcardiac chambers, leading to dilation of thechambers. As a result, the right chambers of theheartfillwithbloodfromthesuperior venacavaandtheinferiorvenacava,andtheleftchambersfillbecauseofforward flowinthepulmonaryveins.Even with a moderately large effusion,mildacute pericarditis usually has no hemody-namic impact on the heart. Hence, jugularvenous profiles are generally normal.Conversely,cardiactamponadeandcon-strictivepericarditisleadto importantanddis-tinctivehemodynamicchanges.Becausetheheartisconstricted byfluidorarigidfibrinousor calcified pericardium, negative intratho-racic pressureduringinspirationisnottrans-mittedtothepericardialsac. Consequently,intrapericardial and intracardiac pressuresremain elevated,andintrapericardialpressureequalsorexceedsrightatrialpressure.In cardiactamponade,thissituationresultsinearlyrightventriculardiastoliccollapse andlate diastolic right atrial collapse,because pres-suresarelowestin diastole.The descentisbluntedbecausetherightventriclecannotexpand,butthe descentremainsnormal.Incardiactamponade,someblood(thoughlessthan normal)stillflowstowardtherightsideoftheheartduringinspiration.Becausethe car-diacchambersareinhibitedfromdilating,theinflowcausestheintraventricular septumtoshifttotheleft,whichfurtherdecreasesleftventricularcapacity.The hemodynamic changes can cause sys-tolic blood pressure to fall by 10 mm Hg orm ore during i nspirati on. Thi s situati on,termed pulsusparadoxus, istheclassichall-mark of cardiac tamponade.Pulsus paradoxusis detectable in 70 to 80 percent of patientswith cardiac tamponade and in about onethird of patients with pericarditis. In constrictive pericarditis,no forward flowoccurs from the superior vena cava and theinferior vena cava during inspiration. Thisresults in Kussmauls sign, which is a para-doxic inspiratory

swelling of the neck veins.Because blood flow to the right side of theheart does not increase,no septal shift occurs,and pulsus paradoxus is not common in con-strictive pericarditis. Clinical Features Patients usually recall a nonspecific pro-drome of malaise, fever, and chest pain, espe-cially in viral or idiopathic pericarditis. Thechest pain is typically pleuritic and radiates tothe left shoulder and,characteristically,the lefttrapezius musculature. The pain is exacer-bated by inspiration or recumbency andrelieved to some extent by leaning forward.Patients may have additional symptoms,depending on the etiology of the pericarditisand the presence of complications.Clinical signs may be subtle. A pericardialfriction rub may or may not be present.Largeeffusions are unlikely to produce a pericardialrub. With large, slowly accumulating effu-sions, there may be dullness to percussion in Pericarditis Classification of Pericarditis Acute pericarditis (<6 weeks)EffusiveFibrinousSubacute pericarditis (>6 weeks to 6 months)Chronic pericarditis (>6 months)EffusiveAdhesiveEffusive-adhesiveConstrictive

edema are evi-dent. Clinical features may mimic those of restrictive cardiomyopathy. Diagnosis The diagnosis of pericarditis and its compli-cations requires a high index of suspicion.Clinical features and the probability of a causeof pericarditis may assist in recogni-tion. Potential confirmatory studies includethe electrocardiogram (ECG),the chest radio-graph, Doppler studies, and selected labora-tory tests. A suggested approach to the diag-nosis of pericarditis is provided in

the scapular region because of compression of the left lung (Ewarts sign).As previously noted,acute,rapid accumula-tionoffluidinthe pericardiumcausessignsofacutehemodynamiccompromisein cardiactamponade. Patients with this conditiondevelop tachycardia,hypotension,pulsus para-doxus, and distended neck veins. Althoughthesepatientsappeartobeinacutedistress(similar to cardiac failure and pulmonary edema),their lungs are clear on auscultation.In chronic constrictive pericarditis, thene ck ve ins are markedly di ste nded, butpatients do not appear distressed and showno evidence of pulsus paradoxus. Duringinspiration, the neck veins become distended(normally, they collapse). The jugular venouspulse shows a sharp descent because of rapid right ventricular filling in early diastole.A sharp heart sound called a pericardialknock is heard in early diastole.Unless otherproblems are present, the lungs are usually clear on auscultation. Because of congestionof the inferior vena cava, the liver is enlarged,and ascites and peripheral leg

In this situation, the differential diagnosisincludes acute myocardial infarction and nor-mal-variant repolarization abnormality. It isparticularly important to distinguish peri-carditis from acute myocardial infarction,because thrombolytic therapy could have dis-astrous effects in patients with pericarditis.Characteristic features of acute pericarditis,acute myocardial infarction, and early repo-larization are summarized in incardiactamponade,theECGshowselec-trical alternans as the heart floats in relationtotherecordingleads.Chronicconstrictivepericarditis presentswithlowvoltageoftheQRScomplexanddiffuseflatteningorinver-sion of the T waves.Atrial fibrillation occurs inonethirdofpatientswith pericardialdisease. CHEST RADIOGRAPHY Ifmorethan250mLoffluidhasaccumu-lated,thecardiacsilhouetteisusually enlargedonthechestradiograph.Smalleraccumulationsmayappearnormal.In constrictivepericarditis,calcificationmaybeseenaroundtheheart. DOPPLER STUDIES Thediagnostictestofchoiceforlargeeffu-sions, cardiac tamponade, and constrictivepericarditisistwo-dimensionalDopplerecho-cardiography. This imaging modality candemonstratemoderateorlargeeffusions. Incardiac tamponade,Doppler examination may showthecharacteristic swingingmotionoftheheartthatgivesrisetoelectricalalternans.Doppler examination can also show thereversal of inflow velocities in the tricuspidand mitral valves during inspiration and expiration.Divergence of right and left ventricularsystolicpressuresis demonstratedincardiactamponade and constrictive pericarditis.Doppler studies differentiate pericarditis fromrestrictivecardiomyopathy.

Symptoms

Chest pain , caused by the inflamed pericardium rubbing against the heart. o Usually relieved by sitting up and leaning forward o Pleuritis type: a sharp, stabbing pain o May radiate to the neck, shoulder, back or abdomen o Often increases with deep breathing and lying flat, and may increase with coughing and swallowing Breathing difficulty when lying down Need to bend over or hold the chest while breathing Drycough Ankle, feet and leg swelling(occasionall y) Anxiety

There may be other signs of fluid in the pericardium (pericardial effusion). If the disorder is severe, there may be crackles in the lungs, decreased breath sounds, or other signs of fluid in the space around the lungs (pleural effusion). If fluid has accumulated in the pericardial sac, it may show on:

Fatigue

Signs and tests When listening to the heart with a stethoscope, the health care provider can hear a typical sound called a pericardial rub. The heart sounds may be muffled or distant.

Chest x-ray Echocardiogram Chest MRI scan Heart MRI or heart CTscan Radionuclide scanning These tests show enlargement of the heart from fluid collection in the pericardium, and signs of inflammation. They may also show scarring andcontrac ture of the pericardium (constrictive pericarditis). Other findings vary depending on the cause of pericarditis.

AnECG is abnormal in 90% of patients with acute pericarditis. ECG changes generally evolve during the disease process, and they may mimic the ECG changes of a heart attack. To rule out heart attack, serial cardiac marker levels (CK -MB and troponin I) may be ordered. Other laboratory tests may include: Pericardiocentesis , with chemical analysis and pericardial fluid culture Treatment The cause of pericarditis must be identified, if possible. in most types of pericarditis, it is necessary to treat the pain withanalgesics (pain killers). The inflammation of the pericardium is treated with antiinflammatory drugs(NSAIDS) such as aspirin and ibuprofen. Iin some cases, corticosteroids may be prescribed. Diuretics may be used to remove excess fluid accumulated in the pericardial sac. If the buildup of pericardial fluid makes the heart function poorly or produces cardiac tamponade, it is necessary to drain the fluid from the sac. This procedure, called pericardiocentesis, may be done using an echocardiography-guided needle or surgically in a minor procedure. Bacterial pericarditis must be treated with antibiotics. Fungal pericarditis is treated with antifungal agents. If the pericarditis ischronic, recurrent, or causes constrictive pericarditis, cutting or removing part of the pericardium may be recommended.

Expectations (prognosis) Pericarditis can range from mild cases that resolve on their own to life-threatening cases complicated by significant fluid buildup around the heart and poor heart function. The outcome is good if the disorder is treated promptly. Most people recover in 2 weeks to 3 months. Complications

Arrhythmias , such as atrial fibrillation. When pericarditis accompanies myocarditis, other arrhythmias may be present, such as supraventricular tachycardia (SVT) or complete heart block.

Cardiac tamponade

Constrictive pericarditis, where inflammation of the pericardial sac results in fibrosis and thickening of the pericardium with adhesions (sticky scars) between the pericardium and the heart. The pericardium creates a rigid "case" around the heart, which can severely limit the ability of the heart to fill with blood. Patients with constrictive pericarditis may develop heart failure, which responds poorly to treatment. Constrictive pericarditis must be differentiated from a chronic heart condition called restrictive cardiomyopathy, which produces symptoms and signs similar to constrictive pericarditis.

Haemophilus aphrophilus Purulent Pericarditis and Tamponade: Case Report Case Report A 62-year-old previously healthy man was admitted to a regional hospital after a 3-week prodrome of malaise and intermittent fevers, with temperatures higher than 39C (102.2F). In the 3 days before admission, he had sharp, intermittent, nonexertional retrosternal chest pains. Twenty-four hours after admission, he became acutely hypotensive. Fluid resuscitation was instituted, and he was transferred to our tertiary care medical center. At our institution, findings on physical examination of the patient demonstrated normal blood pressure and jugular veins that collapsed normally with inspiration. He had poor dentition, advanced gingivitis, dental caries, and pyorrhea. No cardiac murmur or pericardial friction rub was present. There were no cutaneous, ocular, or oral stigmata of bacterial endocarditis present. Selected laboratory findings included leukocytosis (18 x 103/L) with 29% band forms; an elevated erythrocyte sedimentation rate (58 mm/h); and an alkaline phosphatase level of 438 IU/L. Electrocardiography showed normal sinus rhythm with low voltage. Roentgenograms of the chest (frontal and lateral views) showed cardiomegaly but no pulmonary infiltrates or pleural effusions. Two-dimensional echocardiography demonstrated an estimated 50 mL of free-flowing pericardial fluid without evidence of valvular vegetation, cardiac insufficiency, or myocardial abscess. The effusion was unchanged on repeated echocardiography 24 hours later. Results of serologic testing for Bartonella henselae, Bartonella quintana, Coccidioides immitis, Aspergillus fumigatus, Epstein-Barr virus, Cryptococcus neoformans, Mycoplasma pneumoniae, Blastomyces dermatitidis, and Coxiella burnetii did not support a diagnosis of an acute infection. Cultures of 2 serial blood specimens drawn at admission grew H aphrophilus. Based on the positive cultures, a presumptive diagnosis of subacute bacterial endocarditis (SBE) was made. The patient was treated with intravenous ceftriaxone, 2 g/d. The in vitro sensitivity to this antibiotic was confirmed using the Kerby-Bauer disk diffusion method with bacteria plated on Haemophilus test medium. Pericardiocentesis was not attempted because of the small and unchanging volume of fluid demonstrated on sequential echocardiograms. During the next 48 hours, the patient's fever resolved and his chest pain subsided. Follow-up blood cultures remained sterile. One week into his hospitalization, the patient had an acute pulmonary embolism, which was attributed to a thrombus associated with a central venous catheter that had been placed on the day of admission. A CT scan of the chest revealed pericardial thickening enhanced by the contrast medium, consistent with pericardial inflammation (Figure 1). The patient received heparin therapy for the pulmonary embolus before being discharged on a regimen of intravenous ceftriaxone and warfarin sodium. On the day of discharge, he underwent total tooth extraction to reduce his future risk of recurrent bacter- emia and endocarditis. Five days later, the patient was readmitted with rapidly evolving symptoms of dyspnea and circulatory collapse. Findings on the physical examination now revealed distended jugular veins and a palpable pulsus paradoxus. Transthoracic echocardiography revealed compression of the ventricles and inferior vena caval engorgement, consistent with the clinical diagnosis of acute cardiac tamponade. An emergency median sternotomy and pericardiectomy were performed, with decortication of an adherent fibrinous pericardial peel. Approximately 500 mL of frankly purulent, gelatinous fluid were removed from the pericardial space. Gram stain of the pericardial fluid demonstrated 41 leukocytes and 31 gram-variable coccobacilli, morphologically consistent with H aphrophilus; cultures of the pericardial exudate remained sterile. Follow-up echocardiography revealed no residual abnormalities. The patient's condition improved rapidly, and he was discharged to complete 6 weeks of therapy with oral ciprofloxacin.

Purulent pericarditis is an uncommon disease that can be caused by a wide variety of organisms. Although this report describes an uncommon pathogen in an unusual clinical setting, we believe that the case reinforces several principles about the optimal management of purulent pericarditis in general. In 1940, Khairat[1] reported the first case of human disease caused by H aphrophilus in a patient with SBE. Since that time, this organism has been implicated in a number of disease states, including pneumonia, brain abscess, fasciitis, cholecystitis, sinusitis, and soft tissue infections. However, we could find no previous instances of H aphrophilus and purulent pericarditis in the Medline database or in textbooks on the subject. The presence of H aphrophilus in gingival scrapings from healthy volunteers indicates that it is a normal component of the human oral flora. Colonization of the mouth, teeth, and upper respiratory tract of healthy hosts has also been documented.[2] The oropharynx is a clinically significant reservoir of this organism and other HACEK species (Haemophilus, Actinobacillus, Cardiobacterium, Eikenella, and Kingella species), and periodontal disease and recent dental procedures have been cited as risk factors for endocarditis caused by these organisms. Features typical of HACEK endocarditis include an indolent course; large, friable vegetations seen on echocardiography and pathologic specimens; and a higher frequency of major embolic disease than is seen in other forms of endocarditis.[3] Because of its insidious progression, diagnosis of HACEK endocarditis can be delayed [4] for up to 6 months, with the median approximately 3 months, according to one large case series. The pathogenesis of purulent pericarditis has been described by Rubin and Moellering,[5] who analyzed 26 cases and concluded that purulent pericarditis was usually a sequela of another intrathoracic illness. Contiguous spread to the pericardial space from adjacent pneumonia and mediastinal infection following thoracic surgery were the most common antecedents, although the infection appeared to have spread hematogenously in some cases. Several other case series support these observations, including one in which 93% of the cases of pericarditis were preceded by pneumonia.[6] The organisms most commonly associated with purulent pericarditis include Streptococcus pneumoniae (33%); Staphylococcus aureus (23%); Streptococcus pyogenes (11%); Neisseria meningitidis (4%), Haemophilus influenzae (3%), and other gram-negative organisms (19%); anaerobes (2%); and other organisms (5%).[7] Classic reviews of purulent pericarditis stress the importance of early pericardial drainage and document the poor outcomes when antibiotic therapy is attempted without simultaneous surgical intervention.[8] Several factors influenced the delay in reaching the correct diagnosis. First, our patient's clinical presentation was indolent, spanning several weeks. This is in sharp contrast to the fulminant illness that is characteristic of most cases of purulent pericarditis. Second, the stability of the effusion on serial echocardiograms argued against this diagnosis. Third, the positive blood culture results were interpreted incorrectly as indicating HACEK endocarditis, despite the absence of vegetations on the echocardiograms. Applying the Duke criteria for the diagnosis of endocarditis,[8] our patient would have been classified as "possible endocarditis," given 1 major and 1 minor criterion (culture positive for typical organism; fever, with a temperature higher than 39C [102.2F]). In hindsight, clinical data that suggested the correct diagnosis included the pleuropericardial chest pain, the lowvoltage ECG, cardiomegaly seen on the radiographs, the absence of vegetation on serial echocardiograms, and the presence of pericardial effusion. Perhaps most important, the presence of pericardial thickening and inflammation on the chest CT scan might have suggested the diagnosis earlier. The slow clinical progression of the illness and rapid improvement with antibiotic therapy decreased our suspicion of purulent pericarditis early in his evaluation. It is likely that the slow pace of the illness, so uncharacteristic of most cases of purulent pericarditis, resulted from infection with a relatively avirulent organism.

Our current interpretation of the data is that the pericardial infection was likely established before our patient's initial admission to the ICU. It is theoretically possible that the dental extraction performed before the completion of his antibiotic therapy may have provoked bacteremia with the very pathogen that we were attempting to eradicate. However, the fact that he was receiving appropriate antibiotic therapy at the time argues strongly against this possibility. It would also be logical to consider whether the acute decompensation resulted from the use of anticoagulant therapy, causing hemorrhage into the inflamed pericardial space. However, the absence of blood in the pericardial fluid effectively refutes this hypothesis. While it is difficult to draw any firm conclusions about the usefulness of CT versus echocardiography for the diagnosis of purulent pericardi- tis based on this experience, it is noteworthy that the contrast-enhanced CT scan could have suggested the diagnosis earlier. The majority of published data regarding the use of CT scans for diagnosing purulent pericarditis remain anecdotal,[9,10] although one case series concluded that CT was a helpful [11] adjunct to echocardiography in the case of pericarditis. Surgical intervention proved to be definitive in this case, despite treatment with antibiotics with good in vitro activity against H aphrophilus. This reinforces the principle that successful therapy depends on early recognition of this disease and prompt surgical drainage, with antibiotic therapy playing a secondary role. Moreover, this case emphasizes the fact that the onset of purulent pericarditis may occasionally be subtle, particularly when caused by a relatively avirulent pathogen