Br HeartJ 1984; 51: 205-10

Constrictive pericarditis as a complication of coronaryartery bypass surgery

PAULO RIBEIRO, RALPH SAPSFORD, TOM EVANS,* GEORGE PARCHARIDIS,fCELIA OAKLEYFrom the Deparment of Clinical Cardiology and Cardiothoracic Surgery, Hammersmith Hospital, Lnfdon

suMMARY Although it is now recognised as a rare complication of cardiac surgery, constrictivepericarditis was diagnosed in three patients after coronary artery bypass surgery. The time intervalbetween cardiac surgery and the development of constrictive features varied from two to six weeks.All three patients presented with severe congestive heart failure. Haemodynamic findings were

characteristic of constrictive pericarditis. Pericardial thickening detected by computed tomographyin one patient was useful in establishing a definite diagnosis. One of the patients had a serousconstrictive effusive pericarditis, and surgical pericardial drainage was needed. The other patientunderwent pericardiectomy with preservation of the grafts. The diagnosis of constrictive pericarditisshould be considered in patients presenting with unexplained right sided heart failure after cardiacsurgery.

Constrictive pericarditis is an uncommon disease butan important one to recognise as it is potentially cur-able. Bacterial infection of the pericardium after car-diac surgery is a rare but recognised cause of constric-tive pericarditis,I but the development of constrictionwithout overt preceding infection has only recentlybeen reported.2-5

Constrictive pericarditis can be induced experimen-tally by trauma and by injection of autologous bloodin the pericardial sac.6 Traumatic haemopericardiumwith subsequent development of constrictive pericar-ditis has been reported.7 It is surprising that constric-tive pericarditis does not complicate cardiac surgerymore often.We report the clinical presentation, diagnosis, and

management of three patients who developed con-strictive pericarditis as a complication of coronaryartery bypass surgery.

*Present address: Royal Free Hospital, Pond Street, Hampstead, London NW3.tPresent address: University of Thessaloniki, 82 Egnaia Street, Greece.

Requests for reprints to Dr C M Oakley, Hammersmith Hospital,Royal Postgraduate Medical School, Du Cane Road, London W12OHS.

Accepted for publication 27 October 1983

Case reports

CASE 1A 67 year old Greek man gave a three year history oftypical exertional angina. He was otherwise in goodhealth. There was a history of hypertension, and hehad diabetes mellitus controlled by glibenclamide.

Fig. 1 Chestx rayfilm ofcase) before pericardectomy showingmild cardiomegaly and clear lungfields.

205

on May 24, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.51.2.205 on 1 February 1984. D

ownloaded from

206

Physical examination was unremarkable, and both theelectrocardiogram and chest x ray examinationsshowed no abnormality (Fig. 1). Cardiac catheterisa-tion showed good left ventricular function withinferior hypokinesia. Coronary angiography detecteda normal left main stem and three vessel disease. InMarch 1981 he had saphenous bypass grafts placed inthe left anterior descending, the marginal branch ofthe circumflex, and the right coronary arteries. Thepericardium was left open. His postoperative coursewas uneventful, but in retrospect his electrocardiog-ram on discharge showed low voltage and poor prog-ression of the R wave. Chest x ray examinationshowed a normal heart size with a left pleural effu-sion. Six weeks later in Greece he started to complainof tiredness, breathlessness, and ankle swelling. Hewas apyrexial, and a pericardial rub was heard. Chestx ray films showed cardiomegaly. Postcardiotomysyndrome was diagnosed. He was given diuretics andsteroids and his condition improved. His heart sizedecreased on chest x ray examination (Fig. 1), but hissymptoms recurred.

Six months after the operation he was referredback to London. On physical examination his venouspressure was 15 cm with an M shaped contour. Therewas no pulsus paradoxicus, and an early third heartsound was heard. His chest x ray film was clear, and

ECG

yr

LVV

40CRV/

20

mmHgFig. 2 Haemodynamicfindings in case 1. Right vtwnicular(RV) and left ventricular (LV) end diastolic pressures aresimilarly raised with a dip-plateau contour. Note the early dipapproaching the zero level (mm Hg).

Ribeiro, Sapsford, Evans, Parchanidis, Oakley

his liver was palpable 4 cm below the costal margin.There was ascites and ankle swelling. An electrocar-diogram showed sinus rhythm with low voltage. Car-diac catheterisation showed a very small vigorouslycontracting left ventricle and no change in the nativecoronary anatomy. There were three patent grafts. Aright ventricular angiogram was normal. Thehaemodynamic measurements were typical of constr-iction with similar high diastolic pressures in bothventricles. An early dip and plateau were seen in thepressure tracings of both ventricles (Fig. 2). Atpericardiectomy a mass of dense fibrous tissue filledthe anterior mediastinum and was adherent to thepericardium. The anterior and lateral surfaces of thepericardium were excised in between the grafts,which were not disturbed. His symptoms and con-strictive features resolved slowly but completely overthe following six months.

CASE 2A 55 year old man sustained an anterior myocardialinfarct in April 1982. Since then he complained oftypical exertional angina. He smoked 50 cigarettes aday. There was no history of hypertension. Diabetesmellitus was diagnosed in 1981 and controlled withoral hypoglycaemic agents. He was admitted forinvestigation in December 1982. Physical examina-tion was unremarkable. An electrocardiogram showedsinus rhythm and signs of an old inferior myocardialinfarction. Chest x ray films were normal. Left ven-tricular angiography showed inferior dyskinesia withoverall good function. Coronary angiogram showed aleft main stem stricture and three vessel disease.Coronary bypass graft surgery was performed amonth after admission; four saphenous grafts wereinserted. His postoperative course was uneventful,and he was discharged 10 days later taking aspirin anddipyridamole. At the time his electrocardiogramshowed minor ST-T changes and slightly lower vol-tage than before the operation. Chest x ray examina-tion showed the heart to be at the upper limit of nor-mal size.A month later he complained of sickness, tiredness,

breathlessness, and ankle swelling over the previoustwo weeks. On physical examination he was apyrexial,and the venous pressure was 15 cm with pulsus para-doxicus of 15 mm Hg. His chest x ray film was clear.Auscultation detected an early third heart sound.There was gross ascites and ankle swelling. An elec-trocardiogram showed very low voltage. Chest x rayexamination indicated slight cardiomegaly with bilat-eral pleural effusions. At cardiac catheterisationhaemodynamic measurements were typical of con-striction (Fig. 3). The left ventricular -angiogram-showed a small vigorously contracting chamber. Thenative coronary arteries were unchanged. There were

on May 24, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.51.2.205 on 1 February 1984. D

ownloaded from

Constrictive pericarditis as a complication of coronary artery bypass surgeryECG

,.

0 mmHgFig. 3 Haemodynamicfindings in case 2. Right ventricular(RV) and left ventricular (LV) end diastolic pressures aresimilarly raised. The early dip does not approach the zero level(mm Hg).

three patent grafts. Echocardiography showed apericardial effusion. Attempted pericardial tap wasunsuccessful. He was referred for radical pericardec-tomy. At operation the pericardium was thickened,and severe adhesions to the mediastinum and sternumwere found. A pericardial effusion with blood clotswas drained. Because of dense adhesions between thefour grafts and the pericardium an attempt at radicalpericardiectomy was abandoned. During the followingday the gross ascites and ankle swelling subsided. Tendays later his condition was considerably improved.An electrocardiogram still showed low voltage. Twomonths later M mode and cross sectional echocardiog-raphy showed pericardial thickening (Fig. 4).

CASE 3A 47 year old man had a history of myocardial infarc-tion complicated by ventricular fibrillation in 1980.Since then he had exertional angina. He smoked 12cigarettes a day. He was admitted for investigation inJune 1982. Physical examination showed no abnor-malities. Electrocardiogram showed signs of oldinferior and anterior myocardial infarction. A heartsize at the upper limit of normal and clear lung fieldswere evident on the chest x ray films. Cardiaccatheterisation showed overall good left ventricularfunction, but there was anteroapical dyskinesia.Coronary angiography indicated two vessel disease.He was referred for coronary bypass surgery and hadtwo saphenous grafts three days after admission. Hispostoperative recovery was uneventful. At discharge,

his electrocardiogram showed minor ST-T changesand lower voltage than before the operation. Theheart shadow was unchanged on x ray examination.A month later when seen as an outpatient he com-

plained of feeling that his head and neck were con-gested and of tiredness, breathlessness, and ankleswelling during the last few days. On physical exami-nation his venous pressure was 8 cm with an Mshaped pulse contour. His chest was clear, and heartsounds were normal. There was pitting oedema overboth ankles. An electrocardiogram showed low vol-tage. On chest x ray examination the heart size wasunchanged, but small bilateral pleural effusions wereseen. Ventilation perfusion scan, superior vena cava-gram, and left leg venogram were normal. Echocar-diogram showed no pericardial effusion. Computedtomography of the thorax indicated pericardialcalcification and thickening (Fig. 5). At cardiaccatheterisation the right ventricular angiogramshowed a small, but well contracting chamber. Rightsided haemodynamic measurements were characteris-tic of constriction. Constrictive pericarditis after car-diac surgery was diagnosed. As the patient was unwil-ling to undergo further surgery it was decided to deferpericardiectomy.He was seen three months later whenhis symptoms and clinical features were unchanged.

RV

V.-

Fig. 4 M mode echocardiogram of the left ventricle in case 2showing parallel motion of the visceral and parietal pericardiallayers, implying that they are adherent (arrows). PW, posteriorwall; IVS, interventricular septum; RV, right ventricle.

207

oliooo- 00,0,

t p

on May 24, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.51.2.205 on 1 February 1984. D

ownloaded from

Ribeiro, Sapsford, Evans, Parcharidis, Oakley

Fig. 5 Serial chest computed tomograms (case 3) showing extensive pericardial calcification (arrows).

Discussion

Constrictive pericarditis as a consequence of cardiacsurgery was first recognised at necropsy8 and seems tobe a surprisingly rare complication of cardiac surgery.A few cases have been reported after coronary arterybypass surgery,235 prosthetic valve replacement,5 9and atrial septal repair.'0 The diagnosis should besuspected in patients presenting with right sided heartfailure after operation, even if there is associated leftventricular failure.2One of the reasons for the late emergence of con-

striction as a complication of cardiac surgery is thedifficulty in recognising it in patients who have hadsurgery for congenital or valvular heart disease andwho often have other reasons for raised venous fillingpressures. The symptoms and clinical features in our

patients, who had only coronary disease, were typicalof constrictive pericarditis." The pathologicalfindings varied from constrictive effusive pericarditisin the patient (case 1) with pericardial thickening,adhesions, and a pericardial effusion containing bloodclots to those of established chronic constrictivepericarditis as described by Paul Wood.12The main differential diagnosis in patients present-

ing with signs of constriction early after cardiacsurgery is between constrictive effusive pericarditisand late cardiac tamponade. Haemodynamic studiesin serous effusive constrictive pericarditis show aprominent dip plateau ventricular pressure.'3 Theearly dip does not approach the zero level (mm Hg) asin chronic constrictive pericarditis as a result of theeffusive element.13 14 In cardiac tamponade the earlydiastolic dip is absent.'4 15 This is explained by the

208

on May 24, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.51.2.205 on 1 February 1984. D

ownloaded from

Constrictive pericarditis as a complication of coronary artery bypass surgery

tense effusion that prevents filling of the ventriclesfrom early diastole. Furthermore, the venous pressurewave form in serous effusive constrictive pericarditisshows equal systolic and diastolic descents (xy),13whereas in cardiac tamponade the y descent isabsent.'3 15 Computed tomography is diagnostic inshowing pericardial thickening or effusion as well astheir degree and location. In addition these scansshow the haemodynamic consequences with a smallnarrow left ventricle with a sinusoidal septal con-tour when constriction is severe.'6 Perioperative rightventricular infarction should also be considered. Thecharacteristic haemodynamic findings, right ventricu-lar angiography or MUGA scan, or cross sectionalechocardiography will exclude it.'7 18The pathogenesis of postoperative pericardial con-

striction and the reasons for its rarity are not fullyunderstood. Residual blood elements within thepericardial sac, especially in those areas of majorinvolvement, have been found at cardiac surgery inmost patients with chronic constrictive pericarditis ofvarious aetiologies. '9 Even more important are Cohenand Greenberg's findings of organised haematoma,loculated clots, and unclotted viscous blood duringpericardectomy within the pericardial space inpatients with constrictive pericarditis after cardiacsurgery.9 Injection of autologous blood into thepericardial cavity failed to produce constrictivepericarditis in animals. Nevertheless, when thepericardium was temporarily dried and blood wasinjected into the pericardium of rabbits severe adhe-sions between pericardium and epicardiumdeveloped.20 Furthermore, in dogs haemopericar-dium as a result of trauma was followed by clinicaland pathological evidence of constrictive pericarditis.6This suggests that blood and trauma may be impor-tant factors in the pathophysiology of constrictivepericarditis after cardiac surgery. In humans, con-strictive pericarditis after traumatic haemopericar-dium has been well documented.21 22 To support thisconcept, there is experimental evidence of decreasedpericardial mesothelial fibrinolytic activity aftermechanical injury.23 Alternatively, a clinically unde-tectable low grade infection may be a contributivefactor. A role for the postcardiotomy syndrome in thepathogenesis of constriction is another possibility.The high incidence of this syndrome,24 however,makes it difficult to establish a link with the rareoccurrence of constrictive pericarditis after cardiacsurgery.

Different surgical techniques are used to avoid fluidaccumulation within the pericardium. In our institu-tion the pericardium is left open to permit drainageinto the pleural spaces and mediastinum. Further-more, a pericardium not properly protected duringsurgery will shrink and if closed may lead to constric-

tion.25 Even when the pericardium is not closed con-strictive pericarditis may develop.9 Douglas et alfound frequent blood clots around the heart when thepericardium was left open in those cases in whichreoperation was necessary for excessive bleeding.26Most postoperative bleeding originates from struc-tures outside the pericardium.25 This would favourclosing the pericardium. Cunningham et al in theirseries found no intrapericardial clot when reoperationfor excessive bleeding was performed when thepericardium was closed.25 Also, far fewer pericardialeffusions were present when the pericardium was leftclosed.25

Pericardial irrigation with povidone-iodine solutionhas been implicated as a contributory factor in thepathogenesis of postoperative constrictive pericar-ditis.2 None of our patients was exposed to this solu-tion. The available experimental evidence and theabsence of adhesions after peritoneal irrigation makeit unlikely.2 Interestingly, constrictive pericarditis hasbeen reported after pericardial irrigation with Dakin'ssolution.27

Medical treatment with diuretics and steroids hasbeen successful in resolving constrictive signs in twoinstances after cardiac surgery,5 probably by facilitat-ing reabsorption of pericardial fluid and thus revers-ing the effusive element in serous effusive constrictivepericarditis. Constrictive features were relieved bysurgical drainage in one of our patients; radicalpericardectomy may be needed in the future. Saphen-ous vein grafts have been nicked during attempts atpericardiectomy.5 Our patient had three patent graftswith dense adhesions to the pericardium. Attempts atradical pericardiectomy were abandoned.

Radical pericardiectomy is the treatment of choicein constrictive pericarditis. It entails removal of virtu-ally the entire parietal pericardium from all cardiacsurfaces. Delayed recovery after partial pericardiec-tomy has been attributed to myocardial atrophy dur-ing long periods of constriction.28 Somerville sug-gested that inadequate decortication was responsiblefor the delayed improvement and persistence of con-strictive symptoms.28 This concept has been sup-ported by the report of eight successful reoperationsfor constrictive pericarditis with various otheraetiologies after inadequate pericardial decortica-tion.28

After coronary artery bypass graft surgery radicalpericardectomy may be impossible without sacrificingthe grafts. Incomplete pericardiectomy, as carried outin our first patient, was adequate, but it would be verydifficult to achieve if needed in our second patient.The likely course of events, as constrictive pericarditisdevelops, suggests the need for early recognition withsurgical removal as soon as possible of allintrapericardial blood clots and fluid with as much

209

on May 24, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.51.2.205 on 1 February 1984. D

ownloaded from

210

pericardium as can be excised.

CONCLUSIONSThere is evidence that blood accumulation within thepericardial space plays an important part in thepathophysiology of constrictive pericarditis secondaryto cardiac surgery. Possibly with secondary failure ofthe mechanisms of dissolving blood clots by thepericardial mesothelial cells.The diagnosis should be considered in patients pre-

senting with right sided heart failure after cardiacsurgery. Computed tomography is the best availablemethod of screening postoperatively patients in whomthere is a suspicion of developing constriction.Haemodynamic studies will confirm the diagnosis.

Radical pericardiectomy is the treatment of choice.Medical treatment or surgical drainage may, however,partially resolve the constrictive signs when an effu-sive element is also present in constrictive pericarditisso avoiding the potential risk of graft damage duringdecortication.

References1 Beger HG, Bucherl ES. Perikardektomie nach akuter

postoperative Perikarditis: kasuistischer Beitrag. Thorax-chirurgie 1969; 17: 83-7.

2 Marsa R, Mehta S, Willis W, Bailey L. Constrictivepericarditis after myocardial revascularization. Am JCardiol 1979; 44: 177-83.

3 Brown DF, Older T. Pericardial constriction as a latecomplication of coronary bypass surgery. J Thorac Car-diovasc Surg 1977; 74: 61-4.

4 Rubio PA, Farrell EM, Goens RW. Severe adhesive con-strictive pericarditis after coronary artery bypass. SouthMed 1980; 73: 90.

5 Kutcher MA, King SB III, Alimurung BN, Craver JM,Logue RB. Constrictive pericarditis as a complication ofcardiac surgery: recognition of an entity. Am J Cardiol1982; 50: 742-8.

6 Sbokos CG, Karayannacos PE, Kontaxis A, Kambylaf-kas J, Skalkeas GD. Traumatic hemopericardium andchronic constrictive pericarditis. Ann Thorac Surg 1977;23: 225-9.

7 Glenn EE. Traumatic constrictive pericarditis.Journal ofthe Missouri State Medical Association 1940; 37: 7-11.

8 Simon JS, Pluth JR. Constrictive pericarditis. AnnThorac Surg 1976; 21: 440-1.

9 Cohen MV, Greenberg MA. Constrictive pericarditis;early and late complications of cardiac surgery. Am JCardiol 1979; 43: 657-61.

Ribeiro, Sapsford, Evans, Parcharidis, Oakley10 Lange RL. Compressive cardiac and circulatory disor-

ders, clinical and laboratory correlation. Am Heart J71967; 74: 419-30.

11 Fowler NO. Constrictive pericarditis: new aspects. Am J

Cardiol 1982; 50: 1014-7.12 Wood P. Chronic constrictive pericarditis. AmJ Cardiol

1961; 7: 48-61.13 Hancock EW. On the elastic and rigid forms of constric-

tive pericarditis. Am Heart J 1980; 100: 917-23.14 Shabetai R, Fowler NO, Guntheroth WG. The

hemodynamics of cardiac tamponade and constrictivepericarditis. Am J Cardiol 1970; 26: 480-9.

15 Hirschmann JV. Pericardial constriction. Am Heart J1978; 96: 110-22.

16 Rienmuller R, Doppman JL, Reichardt B, Strauer, BE.Diagnostic signs of constrictive pericardial diseases bycomputed tomography [Abstract]. J7ournal of theAmerican College of Cardiology 1983; 1/2: 737.

17 Rotman M, Ratliff NB, Hawley J. Right ventricularinfarction; a haemodynamic diagnosis. Br Heart J 1974;36: 941-4.

18 D'Arcy BJ, Gondi B, Nanda NC, Gatewood RP, BiddleT. Real time two-dimensional echocardiography in rightventricular infarction [Abstract]. Am J Cardiol 1980; 45:436.

19 Effler DB. Chronic constrictive pericarditis treated withpericardiectomy. Am J Cardiol 1961; 7: 62-8.

20 Cliff WJ, Grob6ty J, Ryan GB. Postoperative pericardialadhesions. J Thorac Cardiovasc Surg 1973; 65: 744-50.

21 Warburg E. Traumatic armour heart. Am Heart 1955;49: 633-6.

22 Goldstein S, Yu PN. Constrictive pericarditis after bluntchest trauma. Am Heart J 1965; 69: 544-50.

23 Porter JM, Ball AP, Silver D. Mesothelial fibrinolysis. JThorac Cardiovasc Surg 1971; 62: 725-30.

24 Engle MA, Ito T. The postpericardiotomy syndrome.Am J Cardiol 1961; 7: 73-82.

25 Cunningham JN Jr, Spencer FC, Zeff R, Williams CD,Cukingnan R, Muflin M. Influence of primary closure ofthe pericardium after open-heart surgery on the fre-quency of tamponade, postcardiotomy syndrome, andpulmonary complications. J Thorac Cardiovasc Surg1975; 70: 119-25.

26 Douglas JS Jr, King SB III, Hatcher CR Jr, Jones EL,Logue RB. Late cardiac tamponade after open heartsurgery: a problem of differential diagnosis [Abstract].Am J Cardiol 1975; 35: 133.

27 Beck CS. The effect of surgical solution of chlorinatedsoda (Dakin's solution) in the pericardial cavity. ArchSurg 1929; 18: 1659-71.

28 Somerville W. Constrictive pericarditis with specialreference to the change in natural history brought aboutby surgical intervention. Circulation 1968; 38 (suppl 5):102-11.

on May 24, 2021 by guest. P

rotected by copyright.http://heart.bm

j.com/

Br H

eart J: first published as 10.1136/hrt.51.2.205 on 1 February 1984. D

ownloaded from