- Email: [email protected]
Pericarditis of Acute Myocardial Infarction
Pericarditis of Acute Myocardial Infarction* John C . Toole, M.D.,OO and Mark E. Silverman, M.D.?
Forty patients with acute m y o c a d d infarction and pericarditis (AMI-P) were encountered over a three-year period. The incidence of AMI-P was 7.2 percent (40 of 554 patients). Fifty consecutive patients with acute transmural infarction withoat pericarditis (AMI-C) were used as a control group. There were no signYcant differences between the AMI-P and AMI-C groups regarding age, sex, infarct location, hospital stay or mortality. Painful symptoms of pericarditis were experienced by 37 patients (92 percent), all of whom had developed symptoms by the fourth hospital day. The pericardial friction rub lasted three days or le& in 34 patients (85 percent), but an occasional rub coald be heard for up to eight daya Twenty patients with AMI-P (50 percent) devdoped plenral effusiom and/or parenchymal pulmonary infiltrates. Twentyeight AMI-P patients (70 percent) were thought to have had congestive heart m u r e (CHF) on
the bssis of their symptoms and physical findings. Radi* graphic examination could confirm only 13 cases of CHF among the 28 patients in whom the diagnosis was made clinically. Glucocorticoids were given parenterally to 31 of the 37 patients (84 percent) who had symptomatic pericarditis and was felt to be effective in ameliorating painful symptoms. Followup data was obtained on 28 of the 32 surviving patients. Five patients (15 percemt) had seven episodes of the postmyocardial infarction ayndrome (PMIS). Pericarditis is generally a shortlived complication of acute myocardial infarction. Pleural and parenchymal pulmonary abnormalities are common and probably account for the tendency to "overdiagnose" CHF in patients with AMI-P. PMIS appears to occur more frequently in patients who have had pericarditis at the time of the acute myocardial infarction.
I
mitted to the Piedmont Hospital, coronary care unit (CCU) between January 1971 and January 1974 were reviewed. Since January 1971 a comprehensive coronary care unit data sheet has been compiled on each patient during hospitalization. These data sheets provided prospective information about the type, onset and course of complications, including arrhythmias, that occurred during the hospitalization. In addition, problem-oriented hospital records were reviewed to c o d r m that the data sheet was complete. Patients were excluded from the study if they did not meet one or more of the following criteria for acute myocardial infarction: ( 1 ) development of abnormal Q waves, or evolutionary ST-T wave changes on the electrocardiogram; ( 2 ) diagnostic history; or ( 3 ) compatible history and elevation of creatine phosphokinase, serum glutamic oxalacetic transmainase or lactic dehydrogenase enzyme levels. A diagnosis of pericarditis was accepted when at least two physicians heard a two or three-component pericardial rub. A single component rub was not accepted. Almost all patients had phonocardiographic documentation of the pericardial rub. All patients were seen at least daily by a CCU director and a cardiac fellow in addition to their personal physician. Patients were routinely examined for pericardial rubs in the left lateral, supine, and sitting positions in deep inspiration and full expiration. The patients were also examined repeatedly by nurse-CCU specialists who were trained to do detailed physical examinations and to obtain important subjective information. The nurse's subjective and objective findings were recorded in detail at the end of each shift, using a problem-oriented approach. This provided a thricedaily account of symptoms and physical findings so that the clinical case could be carefully studied. All patients had a standard 12-lead electrocardiogram and a portable upright chest x-ray film on admission to the CCU.
n 1908 William Osler commented, "The existence of pericarditis is very frequently overlooked. Its course in many cases is so insidious and the signs so indefinite that its existence is not suspected, and a For editorial comment, see p x e 632
careful examination is therefore not made . . ."l The diagnosis of pericarditis has continued to be difEcult, particularly following acute myocardial infarction when recurrent myocardial infarction or pulmonary embolus may complicate the differential diagnosis. In the past three years we have studied 40 patients with pericarditis complicating acute myocardial infarction. A review of the clinical presentation and course in these patients has disclosed several features that have been helpful in establishing an early diagnosis and in avoiding inappropriate therapy. The clinical records of all patients with pericarditis ad'From the Department of Medicine, Division of Cardiology, Emory University School of Medicine and Piedmont Hospital, Atlanta, Ga. 'Fellow in Medicine ( Cardiology ) , Emory University School of Medicine. ?Associate Professor of Medicine ( Cardiology), Emory School of Medicine. was supported by NHLl training grant number HE 5653. Manuscript received August 28; revision accepted October 15. Re rint requests: Department of Medidne, Emosy Uniuersity S&Z of Medicine, Atlanta 30303
% ::
CHEST, 67: 6, JUNE, 1975
PERlCARDlTlS OF ACUTE MYOCARDIAL INFARCTION 647
Subsequent electrocardiograms and chest x-ray films were made as deemed necessary by the physicians. All x-ray films were read by one of a three-man group of experienced radiologists who were aware only that the patient was in the
ccu.
For the purpose of this study two categories of congestive heart failure were used. Patients were said to have "clinical congestive heart failure" if the attending physician felt heart failure was present and used digitalis and/or diuretics regardless of the x-ray film findings. An additional class of congestive heart failure, "radiographic congestive heart failure," was said to be present if the radiologist reported pulmonary venous distention, perihilar haze, or alveolar fluid. The findings in fifty consecutive patients admitted during the study period with transmural infarction on electrocardiogram and without pericarditis were analyzed as a control group ( AMI-C) .
There were 554 patients admitted to the CCU during the study period who fulfilled criteria for acute myocardial infarction. Of the 554 patients with acute myocardial infarction, diagnosis was confirmed in 393 (70 percent) by the development of electrocardiographic Q waves or evolutionary ST-T wave changes. The remaining 161 patients were confirmed as having acute myocardial infarction because of a classic or compatible history and serial enzyme changes. Forty (7.2 percent) of the 554 patients with acute myocardial infarction had associated pericarditis (AMI-P) detected during hospitalization. Table 1 reveals no significant difEerence in age, sex, hospital stay or infarct location between the study and control groups.
Thirty-seven of 40 patients (92 percent) with AMI-P complained of painful symptoms that were attributed to pericarditis. Twenty-six (65 percent) Table 1--Clinical Profile of the Patients with Acute Myocardial Infarction with (AMI-P) and without (AMIC) Pencmditis AMI-P Patients, no. Mean age, yr Standard deviation Age range, yr Male Female Infarct Location by ECG Anterior Inferior Posterior Mixed Hospital stay, days Hospital mortality
648 TOOLE, SILVERMAN
O/,
AMI-C
%
Table %Time of Onset o f Painful Symptoms of Pericarditir in Symptomatic Patients Symptom Onset, No. Hospital Day, (On admission) 1
2 3
4 No symptoms
Patients, No. 3 9 15 9
1
3
patients with AMI-P described their discomfort as "sharp" and 11 (27 percent ) experienced an "aching sensation" as their main symptom. Three (7.5 percent) patients denied any painful symptoms, but two of the three did complain of restlessness. The pain was most frequently located in the substernal area; however, five patients ( 12.5 percent) complained of the greatest discomfort in the shoulders and left arm. Movements that enhanced the chest pain were noted in 31 of the 37 symptomatic patients with AMI-P. Any form of thoracic motion, most frequently deep breathing, would aggravate the pain in all 31 patients. Painful symptoms of pericarditis were noted most frequently on the first, second, and third day after infarction (Table 2). Three patients had typical symptoms and a pericardial rub present on admission. One of our patients developed painful symptoms of acute pericarditis on the fourth day after infarction, but none developed painful pericarditis after the fourth day. The onset of the pain of pericarditis preceded, by a period of 4 to 48 hours, the detection of pericardial rub in 26 of the 37 symptomatic patients. In nine patients the rub was discovered simultaneously with the onset of the pericarditis pain; however, threehad pericarditis at the time of admission. A pericardial rub was heard prior to the onset of pain in only two patients, with an interval of less than 24 hours in each case. Symptoms were of short duration, lasting three days or less in 33 of the 37 symptomatic patients. Similarly, in 34 patients (85 percent) the rubs lasted three days or less, although three patients had a persistent rub heard for seven or eight days. The patients with prolonged pericardial friction rubs did not have prolonged painful symptoms.
Heart Failure The incidence of radiographically demonstrable congestive heart failure was remarkably similar in the AMI-P group (37 percent) ( 15 of 40 patients), and the AMI-C group (38 percent) (19 of 50 paCHEST, 67: 6, JUNE, 1975
AMI-P (40 M E N T S )
AMI-C (!W WillENTSl
0
NOCHF
FIGURE1. Venn diagram demonstrating relationship of radiographic and clinical heart failure in AMI-P and AMIC groups. Low incidence of radiographic confirmation of clinical diagnosis of heart failure (13 of 28) in AMI-P group contrasts with correlation (19 of 20) in AMI-C group (P< .01).
tients) (Fig 1).There was excellent correlation between the clinical and radiographic diagnosis of congestive heart failure within the AMI-C group. The AMI-P group of 40 patients contained 28 patients in whom a diagnosis of congestive heart failure had been made clinically. Only 13 of these 28 patients had x-ray film confirmation of the clinical diagnosis. The remaining 15 patients had no evidence of congestive heart failure radiographically. Two patients with AMI-P had radiographic CHF which was not suspected clinically (Fig 1).
Chest X-ray Film Findings Table 3 indicates the incidence, location and type of abnormality found on chest x-ray film in the AMIP and AMI-C groups. Twenty of 40 patients (50 percent) with AMI-P developed pulmonary infiltrates andlor pleural ehsions, contrasted with only 7 of 50 patients ( 14 percent) with AMI-C ( p <.05). There was a strong tendency toward infiltrates or effusions involving the left hemithorax in patients with pericarditis. Isolated involvement of the right side was present in only one patient. Ex-
FIGURE 2. Left lower lobe pulmonary infiltrate and effusion occurring during second day of pericarditis of acute myocardial infarction.
amples of minimal and marked changes are shown in Figures 2 and 3.
E bctrocardwgram All patients with AMI-P fulfilled World Health Organization criteria for transmural myocardial inf a r c t i ~ n The . ~ observed ST-T wave changes were generally attributed to the acute myocardial infarction. An occasional patient, however, showed additional transient ST segment elevation that was pre-
Table L R a d i o g r a p h i c Findings in Patients with AM1 with and without Pericarditis AMI-P
%
AMI-C %
Patients, No. 40 Abnormal x-ray film findings 25 62.5 CHF on x-ray film 15 37.5 Abnormalities, Location R* L** B t Infiltrate only 0 3 1 Effusion only 0 5 3 Infiltrate & effusion 1 6 1 CHF only 0 0 4 *Right cheat. **Left chest. tBoth left and right cheat.
CHEST, 67: 6, JUNE, 1975
FIGURE 3. Large bilateral pleural effusions occurring during pericarditis of acute myocardial infarction. Note relatively clear lung fields above effusions.
PERlCARDlTlS OF ACUTE MYOCARDIAL INFARCTION 849
Table &Episodes
of Significant Arrhythmias in Patients with Acute Myocardial Infarction, with and without Pericarditis, Detected by Constant Electrocardiographic Monitoring
AMI-P
1
Patients, No. Arrhythmia Frequent APB Atrial flutter Atrial fibrillation Atrial tachycardia Frequent VPB VT-VF 2' A-V block
%
AMI-C % 50
40 4
10
2
4
12
30
2
4
4
10 12 10
5 7 5
10 14 10
5 4
sumably due to pericarditis. Twenty-two of the 40 patients (55 percent) with AMI-P had 29 significant arrhythmias (Table 4). There were 12 instances of supraventricular tachyarrhythmias in the AMI-P group compared to only two episodes in the control group ( .10 < p >.05). There were no other significant differences in the incidence or type of arrhythmias between the study and control groups.
Mortality The hospital mortality was 20 percent ( 8 of 40 patients) in the AMI-P group and 12 percent ( 6 of 50 patients) in the AMI-C group ( p > .05). The mortality for all 554 patients with acute myocardial infarction was 9.5 percent for the three years of the study. Six patients with AMI-P died between the 3rd and 30th hospital day of resistant ventricular arrhythmias. One patient died of cardiac rupture on day nine and another died of intractable heart failure on day 30. All six deaths in the AMI-C group occurred within the first three days of hospitalization. Postmortem examination was performed on six of the eight patients in the study group. All six patients had pathologic evidence of acute transmural infarction and three demonstrated a ventricular aneurysm. The average age of patients dying with AMI-P (68.7 years) was greater than that of the AMI-C group (61.8 years), but not significantly so ( p > .05).
Anticoagulant Drugs and Complications Eleven patients (27.5 percent) with AMI-P received anticoagulant therapy with heparin and/or warfarin sodium ( Coumadin ) during the period of pericarditis. Two of these patients developed hypotension and pulsus paradoxus greater than 20 mrn Hg. In both patients pericardiocentesis was attempted, but no pericardial fluid could be obtained. The signs of tamponade resolved 12 to 24 hours after cessation of the anticoagulant therapy. One patient made an uneventful recovery and was feeling well at followup. The other patient died 16 days later after several cardiac arrests necessitating closed chest cardiac message. At autopsy an organizing hemopericardium was found. Pericardial effusion was not apparent radiographically in either of these two cases. There were no suspected instances of pericardial tamponade in the 29 patients with AMI-P who where not receiving anticoagulants, nor in any of the 50 control cases.
Followup information regarding recurrent pencarditis was obtained in 28 of the 32 patients with AMI-P who survived the initial hospital course. The period of followup ranged from two months to three years. Five of the surviving 32 patients ( 15 percent ) with AMI-P had seven episodes of pericarditis not associated with evidence of reinfarction at intervals of from one to four months following myocardial infarction. These episodes consisted of anterior chest pain consistent with pericarditis associated with a pericardial rub. Three additional patients with the onset of the pericarditis within three days of the acute myocardial hfarction developed large pleural effusions, which persisted from two to five weeks after infarction. The magnitude and duration of the pleural effusions, in the absence of congestive heart failure, suggests that these patients may have had an early onset of the postmyocardial infarction syndrome (PMIS ).
Treatment Corticosteroids were given to 31 of the 37 (83 percent ) symptomatic patients in attempts to relieve the pain of pericarditis. An initial dose equivalent to 40-60 mg of prednisone was given intravenously. This single intravenous dose produced satisfactory control of symptoms in 9 of the 31 treated patients (29 percent). The remaining 22 patients, although improved by the initial injection, still had sufficient symptoms to warrant a three to five day course of orally administered glucocorticosteroid therapy.
In 1872, Baurnler reported a series of patients with pericarditis, three of whom most likely had pericarditis of acute myocardial infarction. His description noted "well marked symptoms" which lasted only a few days and consisted primarily of pleuritic chest pain frequently radiating into the shoulders and arms. Recent studies have confirmed that generally the symptoms and signs of AMI-P occur within the first four days after infarction, and a rub is rarely heard after the seventh The current
650 TOOLE, SILVERMAN
CHEST, 67: 6, JUNE, 1975
study supports previous observations on incidence, time course and mortality associated with AMI-P. There are, however, important features of AMI-P which have received little attention and which deserve emphasis. There was some variability in the character of pain experienced by patients with AMI-P, and almost all patients had pain which was exacerbated by thoracic motion. Many patients did not volunteer this information, but supplied it only when asked specifically about the effect of movement on their pain. The recognition of this typical pain pattern, especially on the second or third hospital day, led us to diagnose pericarditis very early in its course. The knowledge that symptoms generally preceded an audible rub permitted us on occasion to diagnose and treat pericarditis before a con6rmatory rub was heard. We were thus able to reassure the patients and allay anxiety concerning another heart attack. There was a high incidence of abnormal chest x-ray film findings in our patients with AMI-P. Twenty of our 40 patients (50 percent) with AMI-P had a pulmonary infiltrate and/or pleural effusion during the acute phase of their illness. Niarchos et a14 reported a high incidence ( 13/18) of abnormal chest x-ray film findings in patients with AMI-P. Only 3 of the 18 patients, however, had pericardial or pleural effusion and none had pulmonary infiltrate^.^ Thadani et a15 reported a 7.7 percent incidence of pleural effusion in patients with AMI-P and only a 1 percent incidence in the control g r o ~ p Barman .~ et a17 found no pleural or pericardial effusion in an analysis of 106 cases of AMI-P. The studies of Niarchos4 and ThadaniQuggest a slightly increased incidence of pleural effusion in patients with AMI-P, but with a frequency several orders of magnitude below that of the present study. A definitive explanation for the pulmonary findings cannot be given; however, there are several possibilities to consider. It has been suggested that the observed pulmonary changes are purely the result of congestive heart f a i l ~ r eFifty-five .~ percent of our patients with AMI-P and pulmonary abnormalities had radiographic findings of congestive heart failure; however, the remaining 45 percent showed no such evidence. The AMI-C group demonstrated the same incidence of radiographic congestive heart failure as did the AMI-P group, yet there were no patients with pleural effusion and only seven (14 percent) with pulmonary infiltrates in the AMI-C group. Furthermore, there is a tendency for congestive heart failure not to produce unilateral left-sided pleural effusion,1° yet the pleural and parenchymal findings in our patients had a strong predilection to occur on the left. While it is quite possible that CHEST, 67: 6, JUNE, 1975
congestive heart failure contributes to the pulmonary changes observed in some patients with AMI-P, for the above reasons we feel it is unlikely to be the only factor in their genesis. Another possibility is that the pleuritic pain experienced by patients with AMI-P is indirectly responsible for the observed changes in x-ray film findings. The rapid, shallow respiratory pattern adopted by most patients with pleuritic pain could conceivably produce basilar atelectasis. It would, however, be difficult to explain the predominantly unilateral infiltrates and the pleural effusions on the basis of decreased thoracic motion. The PMIS is known to occur occasionally early in the course of acute myocardial infarction,"-l4 and pulmonary infiltrates and effusions are frequent features of this syndrome.15 Although an early onset of PMIS may account for the radiologic abnormalities we observed, it still does not explain the mechanism by which they occur. Furthermore, it would imply a much higher frequency for the PMIS than has previously been de~cribed.~ The x-ray film findings in idiopathic pericarditis, PMIS, and the postcommissurotomy syndrome are very similar.15These abnormalities, with the exception of pericardial effusion, are also similar to the abnormalities seen in patients with AMI-P.16 The radiographic similarities between the various types of pericarditis may be more than a coincidence. It is conceivable that regardless of inciting events or agents, pericarditis could produce the pulmonary abnormalities by direct extension of the inflammatory process to the adjacent pleura and lung. In the patients with AMI-P, congestive heart failure was "overdiagnosed" when clinical criteria alone were used ( Fig 1 ) . Pain-related tachycardia, tachypnea, and basilar rilles resulting from the pulmonary infiltrates were probably responsible for the erroneous diagnosis of congestive heart failure in most cases. Early in the course of AMI-P a short, early diastolic friction rub could easily be interpreted as a ventricular gallop, adding more spurious data to support the diagnosis of congestive heart failure. These pitfalls can be avoided if the chest x-ray film is carefully studied before making a diagnosis of congestive heart failure in a patient with AMI-P. The incidence of the PMIS was high in our surviving patients with AMI-P. In addition to the five patients with typical PMIS, three may have had AMI-P blending imperceptibly into the PMIS. Even if these three cases are excluded, the 15 percent incidence of the PMIS is much higher than the 1 to 4 percent incidence cited for all patients with acute myocardial infarction. 4 . 1 7 PERlCARDlTlS OF ACUTE MYOCARDIAL INFARCTION 651
It would appear that the patient who has had pericarditis from any of several causes remains at risk for recurrent attacks. For this reason, it is especially important to document pericarditis occumng during acute myocardial infarction and record it in the medical record, or problem list, along with the other complications of acute myocardial infarction. Ready availability of this information would be of great assistance in the management of a patient who presents with pleuritic chest pain and a pulmonary infiltrate several weeks following AMI-P. Pulmonary embolus or infarction is an obvious diagnostic possibility in this situation, but the PMIS should also be given strong consideration. The distinction between the PMIS and pulmonary infarction has special significance because of the reported risk of pericardial hemorrhage in patients with PMIS who are also on anticoagulant It has been our policy not to prescribe anticoagulants for such patients unless lung scans or pulrnonary angiograms demonstrate strong evidence of pulmonary embolus. anticoagulation is According to many authoritie~,~ not contraindicated in AMI-P. Niarchos,* however, reported significant hemopericardium in 2 out of 22 patients with AMI-P and we suspected two cases of tamponade, one of which showed hemopericardium at autopsy 16 days later. All patients, in both series, with suspected or proved pericardial tamponade were receiving anticoagulant therapy at the time of the suspected hemorrhage. Thus, there appears to be a small (1.5 percent) but definite risk in using anticoagulants in patients with AMI-P. Our current policy is to discontinue anticoagulant therapy at the onset of pericarditis if the only indication for it is prophylaxis against venous thromboembolism. When a specific indication for anticoagulation exists (such as phlebitis or pulmonary embolus ), anticoagulation is continued with careful monitoring of blood pressure, pulsus paradoxus and neck veins for signs of pericardial tamponade. All of our patients with AMI-P had definite electrocardiographic documentation of a transmural myocardial infarction. It is conceivable that pericarditis could occur with a nontransmural infarction; however, most authors"' have not found this to be the case. Both this study and the literature suggest that patients who present with pericarditis and do not evolve electrocardiographic changes of a transmural infarction are unlikely to have AMI-P as the etiology of their pericarditis. The pain of pencarditiscan mimic myocardial infarction in every detail including radiation down one or both arms and it may be impossible to separate idiopathic pericarditis from AMI-P on the basis of history alone. 652 TOOLE, SILVERMAN
Since almost all of our symptomatic patients (31 to 37) received parenterally administered stexoids, we can make no objective assessment of the therapeutic efficacy of steroids in AMI-P. Our strong impression is that a significant number of patients obtained dramatic relief from constant pain within 30 minutes after the steroid injection. Many of these patients had responded poorly or not at all to high doses of intravenous narcotics. In fact, severe pain which did not respond to opiates was often the first clue to the development of pericarditis. The use of steroids allowed us to avoid completely or greatly decrease the dosage of narcotics, thereby avoiding drug-induced respiratory depression or hypotension. The patients who continued to have some painful symptoms after the initial steroid injection were given steroids orally in tapering dosage over a three to five day period. We observed no relapses of symptomatic pericarditis following steroid withdrawal. A single instance of aggravated diabetic control was the only complication attributable to the steroid therapy. Early reportslg suggested an increased mortality in AMI-P compared to acute myocardial infarction without pericarditis. More recent information," including the current study, indicates no statistically significant increase in mortality for the AMI-P group. An unexplained finding in our patients with AMI-P was the high incidence of late hospital deaths. All late deaths followed a prolonged and complicated course, during which time the patient remained severely ill. There were no sudden or unexplained deaths to suggest that this group might benefit from extended cardiac monitoring. ACKNOWLEDGMENTS: We thank Dr. Robert C. Schlant for editorial assistance and encouragement in the preparation of this manuscript. We also thank the nursing staff of the Coronary Care Unit, Piedmont Hospital, without whom this study could not have been conducted.
1 Osler W: Modem Medicine: Its Theory and Practice (Osler W, ed) (vol 4 ) . Philadelphia, Lea & Febiger, 1908, p 58 2 World Health Organization: WHO Report Euro 5010. Ischemic Heart Disease Registers. Report of Working Group. Copenhagen, World Health Organization, 1968 3 Blumer G : Pericarditis epistenocardia. JAMA 107:178181,1936 4 Niarchos AP, McKendrick CS: Prognosis of pericarditis after acute myocardial infarction. Br Heart J 35:49-54, 1973 5 Thadani V, Chopra MA, Aber CP, et al: Pericarditis after acute myocardial infarction. Br Med J 2:135-37, 1971 6 Lichstein E, Liu H, Gupta P: Pericarditis complicating acute myocardial infarction: incidence of complications and significance of electrocardiogram on admission. Am Heart J 87:246-252, 1974
CHEST, 67: 6, JUNE, 1975
7 Barman PC, Krishnaswami V, Geraci AR: Pericarditis in acute myocardial infarction. NY State J Med 72:645-48, 1973 8 Yan V: Pericarditis in acute myocardial infarction. Heart & Lung 3:247-251, 1974 9 Weiser NJ, Kantor M, Russell HK, et al: Post myocardial infarction syndrome: Nonspecificity of pulmonary manifestations. Circulation 25:643-650, 1962 10 Meszaros WT: Lung changes in left heart failure. Circulation 47 :859-871, 1973 11 Stein I, Prussin G: Early appearance of the post myocardial infarction syndrome. Angiography 20:26268, 1969 12 Ananthasubrarnaniam G, Muthukrishman TV: Pericarditis in myocardial infarction. Indian Heart J 23:302-05, 1971 13 Weiser NJ, Kantor M, Russell HK: Post myocardial in-
farction syndrome. Circulation 20:371-380, 1959 14 Kossowsky WA, Epstein PJ, Levine RS: Post myocardial infarction syndrome: An early complication of acute my* cardial infarction. Chest 63:35-40, 1973 15 Dressler W: Idiopathic recurrent pericarditis. Am J Med 18:591601,1955 16 Maude1 W, Johnson EC: Pleuropericardial effusion following myocardial infarction. Am Heart J 53: 145-49, 1957 17 Dressler W: The post myocardial infarction syndrome: A report of 44 cases. Arch Intern Med 103:28-42, 1959 18 Dressler W: Management of pericarditis secondary to myocardial infarction. Progr Cardiovasc Dis 3:134-140, 1980 19 Levine SA: Coronary thrombosis: Its various clinical features. Medicine (Baltimore) 8:245-418, 1929
The Penguin
'
The penguin spends more than half of each year on islands. Although he is a bird, he cannot fly. His element is the sea. He can swim almost as swiftly a s a shark. Fish is his only diet-a variety of small sardine, peculiarly rich in oil. When night arrives, he floats in the attitude of a duck and sleeps. He propels himself, when swimming, entirely by means of his flippers. A transparent film comes over his eyes the moment he goes under the water. In February and September the birds will mate and in the next six or eight weeks are making for their island to lay their eggs. They dig in a sloping direction a sort of tunnel to a depth of one, two or three feet. The two
CHEST, 67: 6, JUNE, 1975
penguins that mated take their turns in the hatching of the egg for about twelve hours each for four weeks. Two or three days after the first egg, there comes another; there may be a third or even a fourth. The ibis and the gull take not only eggs but also young chicks. The chicks are fed every twenty minutes. The parent's digestive machinery turns into oil the food collected and this oil is brought back into the beak. Then the young chicks insert their own small beaks into the parent's larger beak and so draw the nourishment. Kearton C: The Island of Penguins. New York, McBride, 1931
PERlCARDlTlS OF ACUTE MYOCARDIAL INFARCTION 653
Forty patients with acute m y o c a d d infarction and pericarditis (AMI-P) were encountered over a three-year period. The incidence of AMI-P was 7.2 percent (40 of 554 patients). Fifty consecutive patients with acute transmural infarction withoat pericarditis (AMI-C) were used as a control group. There were no signYcant differences between the AMI-P and AMI-C groups regarding age, sex, infarct location, hospital stay or mortality. Painful symptoms of pericarditis were experienced by 37 patients (92 percent), all of whom had developed symptoms by the fourth hospital day. The pericardial friction rub lasted three days or le& in 34 patients (85 percent), but an occasional rub coald be heard for up to eight daya Twenty patients with AMI-P (50 percent) devdoped plenral effusiom and/or parenchymal pulmonary infiltrates. Twentyeight AMI-P patients (70 percent) were thought to have had congestive heart m u r e (CHF) on
the bssis of their symptoms and physical findings. Radi* graphic examination could confirm only 13 cases of CHF among the 28 patients in whom the diagnosis was made clinically. Glucocorticoids were given parenterally to 31 of the 37 patients (84 percent) who had symptomatic pericarditis and was felt to be effective in ameliorating painful symptoms. Followup data was obtained on 28 of the 32 surviving patients. Five patients (15 percemt) had seven episodes of the postmyocardial infarction ayndrome (PMIS). Pericarditis is generally a shortlived complication of acute myocardial infarction. Pleural and parenchymal pulmonary abnormalities are common and probably account for the tendency to "overdiagnose" CHF in patients with AMI-P. PMIS appears to occur more frequently in patients who have had pericarditis at the time of the acute myocardial infarction.
I
mitted to the Piedmont Hospital, coronary care unit (CCU) between January 1971 and January 1974 were reviewed. Since January 1971 a comprehensive coronary care unit data sheet has been compiled on each patient during hospitalization. These data sheets provided prospective information about the type, onset and course of complications, including arrhythmias, that occurred during the hospitalization. In addition, problem-oriented hospital records were reviewed to c o d r m that the data sheet was complete. Patients were excluded from the study if they did not meet one or more of the following criteria for acute myocardial infarction: ( 1 ) development of abnormal Q waves, or evolutionary ST-T wave changes on the electrocardiogram; ( 2 ) diagnostic history; or ( 3 ) compatible history and elevation of creatine phosphokinase, serum glutamic oxalacetic transmainase or lactic dehydrogenase enzyme levels. A diagnosis of pericarditis was accepted when at least two physicians heard a two or three-component pericardial rub. A single component rub was not accepted. Almost all patients had phonocardiographic documentation of the pericardial rub. All patients were seen at least daily by a CCU director and a cardiac fellow in addition to their personal physician. Patients were routinely examined for pericardial rubs in the left lateral, supine, and sitting positions in deep inspiration and full expiration. The patients were also examined repeatedly by nurse-CCU specialists who were trained to do detailed physical examinations and to obtain important subjective information. The nurse's subjective and objective findings were recorded in detail at the end of each shift, using a problem-oriented approach. This provided a thricedaily account of symptoms and physical findings so that the clinical case could be carefully studied. All patients had a standard 12-lead electrocardiogram and a portable upright chest x-ray film on admission to the CCU.
n 1908 William Osler commented, "The existence of pericarditis is very frequently overlooked. Its course in many cases is so insidious and the signs so indefinite that its existence is not suspected, and a For editorial comment, see p x e 632
careful examination is therefore not made . . ."l The diagnosis of pericarditis has continued to be difEcult, particularly following acute myocardial infarction when recurrent myocardial infarction or pulmonary embolus may complicate the differential diagnosis. In the past three years we have studied 40 patients with pericarditis complicating acute myocardial infarction. A review of the clinical presentation and course in these patients has disclosed several features that have been helpful in establishing an early diagnosis and in avoiding inappropriate therapy. The clinical records of all patients with pericarditis ad'From the Department of Medicine, Division of Cardiology, Emory University School of Medicine and Piedmont Hospital, Atlanta, Ga. 'Fellow in Medicine ( Cardiology ) , Emory University School of Medicine. ?Associate Professor of Medicine ( Cardiology), Emory School of Medicine. was supported by NHLl training grant number HE 5653. Manuscript received August 28; revision accepted October 15. Re rint requests: Department of Medidne, Emosy Uniuersity S&Z of Medicine, Atlanta 30303
% ::
CHEST, 67: 6, JUNE, 1975
PERlCARDlTlS OF ACUTE MYOCARDIAL INFARCTION 647
Subsequent electrocardiograms and chest x-ray films were made as deemed necessary by the physicians. All x-ray films were read by one of a three-man group of experienced radiologists who were aware only that the patient was in the
ccu.
For the purpose of this study two categories of congestive heart failure were used. Patients were said to have "clinical congestive heart failure" if the attending physician felt heart failure was present and used digitalis and/or diuretics regardless of the x-ray film findings. An additional class of congestive heart failure, "radiographic congestive heart failure," was said to be present if the radiologist reported pulmonary venous distention, perihilar haze, or alveolar fluid. The findings in fifty consecutive patients admitted during the study period with transmural infarction on electrocardiogram and without pericarditis were analyzed as a control group ( AMI-C) .
There were 554 patients admitted to the CCU during the study period who fulfilled criteria for acute myocardial infarction. Of the 554 patients with acute myocardial infarction, diagnosis was confirmed in 393 (70 percent) by the development of electrocardiographic Q waves or evolutionary ST-T wave changes. The remaining 161 patients were confirmed as having acute myocardial infarction because of a classic or compatible history and serial enzyme changes. Forty (7.2 percent) of the 554 patients with acute myocardial infarction had associated pericarditis (AMI-P) detected during hospitalization. Table 1 reveals no significant difEerence in age, sex, hospital stay or infarct location between the study and control groups.
Thirty-seven of 40 patients (92 percent) with AMI-P complained of painful symptoms that were attributed to pericarditis. Twenty-six (65 percent) Table 1--Clinical Profile of the Patients with Acute Myocardial Infarction with (AMI-P) and without (AMIC) Pencmditis AMI-P Patients, no. Mean age, yr Standard deviation Age range, yr Male Female Infarct Location by ECG Anterior Inferior Posterior Mixed Hospital stay, days Hospital mortality
648 TOOLE, SILVERMAN
O/,
AMI-C
%
Table %Time of Onset o f Painful Symptoms of Pericarditir in Symptomatic Patients Symptom Onset, No. Hospital Day, (On admission) 1
2 3
4 No symptoms
Patients, No. 3 9 15 9
1
3
patients with AMI-P described their discomfort as "sharp" and 11 (27 percent ) experienced an "aching sensation" as their main symptom. Three (7.5 percent) patients denied any painful symptoms, but two of the three did complain of restlessness. The pain was most frequently located in the substernal area; however, five patients ( 12.5 percent) complained of the greatest discomfort in the shoulders and left arm. Movements that enhanced the chest pain were noted in 31 of the 37 symptomatic patients with AMI-P. Any form of thoracic motion, most frequently deep breathing, would aggravate the pain in all 31 patients. Painful symptoms of pericarditis were noted most frequently on the first, second, and third day after infarction (Table 2). Three patients had typical symptoms and a pericardial rub present on admission. One of our patients developed painful symptoms of acute pericarditis on the fourth day after infarction, but none developed painful pericarditis after the fourth day. The onset of the pain of pericarditis preceded, by a period of 4 to 48 hours, the detection of pericardial rub in 26 of the 37 symptomatic patients. In nine patients the rub was discovered simultaneously with the onset of the pericarditis pain; however, threehad pericarditis at the time of admission. A pericardial rub was heard prior to the onset of pain in only two patients, with an interval of less than 24 hours in each case. Symptoms were of short duration, lasting three days or less in 33 of the 37 symptomatic patients. Similarly, in 34 patients (85 percent) the rubs lasted three days or less, although three patients had a persistent rub heard for seven or eight days. The patients with prolonged pericardial friction rubs did not have prolonged painful symptoms.
Heart Failure The incidence of radiographically demonstrable congestive heart failure was remarkably similar in the AMI-P group (37 percent) ( 15 of 40 patients), and the AMI-C group (38 percent) (19 of 50 paCHEST, 67: 6, JUNE, 1975
AMI-P (40 M E N T S )
AMI-C (!W WillENTSl
0
NOCHF
FIGURE1. Venn diagram demonstrating relationship of radiographic and clinical heart failure in AMI-P and AMIC groups. Low incidence of radiographic confirmation of clinical diagnosis of heart failure (13 of 28) in AMI-P group contrasts with correlation (19 of 20) in AMI-C group (P< .01).
tients) (Fig 1).There was excellent correlation between the clinical and radiographic diagnosis of congestive heart failure within the AMI-C group. The AMI-P group of 40 patients contained 28 patients in whom a diagnosis of congestive heart failure had been made clinically. Only 13 of these 28 patients had x-ray film confirmation of the clinical diagnosis. The remaining 15 patients had no evidence of congestive heart failure radiographically. Two patients with AMI-P had radiographic CHF which was not suspected clinically (Fig 1).
Chest X-ray Film Findings Table 3 indicates the incidence, location and type of abnormality found on chest x-ray film in the AMIP and AMI-C groups. Twenty of 40 patients (50 percent) with AMI-P developed pulmonary infiltrates andlor pleural ehsions, contrasted with only 7 of 50 patients ( 14 percent) with AMI-C ( p <.05). There was a strong tendency toward infiltrates or effusions involving the left hemithorax in patients with pericarditis. Isolated involvement of the right side was present in only one patient. Ex-
FIGURE 2. Left lower lobe pulmonary infiltrate and effusion occurring during second day of pericarditis of acute myocardial infarction.
amples of minimal and marked changes are shown in Figures 2 and 3.
E bctrocardwgram All patients with AMI-P fulfilled World Health Organization criteria for transmural myocardial inf a r c t i ~ n The . ~ observed ST-T wave changes were generally attributed to the acute myocardial infarction. An occasional patient, however, showed additional transient ST segment elevation that was pre-
Table L R a d i o g r a p h i c Findings in Patients with AM1 with and without Pericarditis AMI-P
%
AMI-C %
Patients, No. 40 Abnormal x-ray film findings 25 62.5 CHF on x-ray film 15 37.5 Abnormalities, Location R* L** B t Infiltrate only 0 3 1 Effusion only 0 5 3 Infiltrate & effusion 1 6 1 CHF only 0 0 4 *Right cheat. **Left chest. tBoth left and right cheat.
CHEST, 67: 6, JUNE, 1975
FIGURE 3. Large bilateral pleural effusions occurring during pericarditis of acute myocardial infarction. Note relatively clear lung fields above effusions.
PERlCARDlTlS OF ACUTE MYOCARDIAL INFARCTION 849
Table &Episodes
of Significant Arrhythmias in Patients with Acute Myocardial Infarction, with and without Pericarditis, Detected by Constant Electrocardiographic Monitoring
AMI-P
1
Patients, No. Arrhythmia Frequent APB Atrial flutter Atrial fibrillation Atrial tachycardia Frequent VPB VT-VF 2' A-V block
%
AMI-C % 50
40 4
10
2
4
12
30
2
4
4
10 12 10
5 7 5
10 14 10
5 4
sumably due to pericarditis. Twenty-two of the 40 patients (55 percent) with AMI-P had 29 significant arrhythmias (Table 4). There were 12 instances of supraventricular tachyarrhythmias in the AMI-P group compared to only two episodes in the control group ( .10 < p >.05). There were no other significant differences in the incidence or type of arrhythmias between the study and control groups.
Mortality The hospital mortality was 20 percent ( 8 of 40 patients) in the AMI-P group and 12 percent ( 6 of 50 patients) in the AMI-C group ( p > .05). The mortality for all 554 patients with acute myocardial infarction was 9.5 percent for the three years of the study. Six patients with AMI-P died between the 3rd and 30th hospital day of resistant ventricular arrhythmias. One patient died of cardiac rupture on day nine and another died of intractable heart failure on day 30. All six deaths in the AMI-C group occurred within the first three days of hospitalization. Postmortem examination was performed on six of the eight patients in the study group. All six patients had pathologic evidence of acute transmural infarction and three demonstrated a ventricular aneurysm. The average age of patients dying with AMI-P (68.7 years) was greater than that of the AMI-C group (61.8 years), but not significantly so ( p > .05).
Anticoagulant Drugs and Complications Eleven patients (27.5 percent) with AMI-P received anticoagulant therapy with heparin and/or warfarin sodium ( Coumadin ) during the period of pericarditis. Two of these patients developed hypotension and pulsus paradoxus greater than 20 mrn Hg. In both patients pericardiocentesis was attempted, but no pericardial fluid could be obtained. The signs of tamponade resolved 12 to 24 hours after cessation of the anticoagulant therapy. One patient made an uneventful recovery and was feeling well at followup. The other patient died 16 days later after several cardiac arrests necessitating closed chest cardiac message. At autopsy an organizing hemopericardium was found. Pericardial effusion was not apparent radiographically in either of these two cases. There were no suspected instances of pericardial tamponade in the 29 patients with AMI-P who where not receiving anticoagulants, nor in any of the 50 control cases.
Followup information regarding recurrent pencarditis was obtained in 28 of the 32 patients with AMI-P who survived the initial hospital course. The period of followup ranged from two months to three years. Five of the surviving 32 patients ( 15 percent ) with AMI-P had seven episodes of pericarditis not associated with evidence of reinfarction at intervals of from one to four months following myocardial infarction. These episodes consisted of anterior chest pain consistent with pericarditis associated with a pericardial rub. Three additional patients with the onset of the pericarditis within three days of the acute myocardial hfarction developed large pleural effusions, which persisted from two to five weeks after infarction. The magnitude and duration of the pleural effusions, in the absence of congestive heart failure, suggests that these patients may have had an early onset of the postmyocardial infarction syndrome (PMIS ).
Treatment Corticosteroids were given to 31 of the 37 (83 percent ) symptomatic patients in attempts to relieve the pain of pericarditis. An initial dose equivalent to 40-60 mg of prednisone was given intravenously. This single intravenous dose produced satisfactory control of symptoms in 9 of the 31 treated patients (29 percent). The remaining 22 patients, although improved by the initial injection, still had sufficient symptoms to warrant a three to five day course of orally administered glucocorticosteroid therapy.
In 1872, Baurnler reported a series of patients with pericarditis, three of whom most likely had pericarditis of acute myocardial infarction. His description noted "well marked symptoms" which lasted only a few days and consisted primarily of pleuritic chest pain frequently radiating into the shoulders and arms. Recent studies have confirmed that generally the symptoms and signs of AMI-P occur within the first four days after infarction, and a rub is rarely heard after the seventh The current
650 TOOLE, SILVERMAN
CHEST, 67: 6, JUNE, 1975
study supports previous observations on incidence, time course and mortality associated with AMI-P. There are, however, important features of AMI-P which have received little attention and which deserve emphasis. There was some variability in the character of pain experienced by patients with AMI-P, and almost all patients had pain which was exacerbated by thoracic motion. Many patients did not volunteer this information, but supplied it only when asked specifically about the effect of movement on their pain. The recognition of this typical pain pattern, especially on the second or third hospital day, led us to diagnose pericarditis very early in its course. The knowledge that symptoms generally preceded an audible rub permitted us on occasion to diagnose and treat pericarditis before a con6rmatory rub was heard. We were thus able to reassure the patients and allay anxiety concerning another heart attack. There was a high incidence of abnormal chest x-ray film findings in our patients with AMI-P. Twenty of our 40 patients (50 percent) with AMI-P had a pulmonary infiltrate and/or pleural effusion during the acute phase of their illness. Niarchos et a14 reported a high incidence ( 13/18) of abnormal chest x-ray film findings in patients with AMI-P. Only 3 of the 18 patients, however, had pericardial or pleural effusion and none had pulmonary infiltrate^.^ Thadani et a15 reported a 7.7 percent incidence of pleural effusion in patients with AMI-P and only a 1 percent incidence in the control g r o ~ p Barman .~ et a17 found no pleural or pericardial effusion in an analysis of 106 cases of AMI-P. The studies of Niarchos4 and ThadaniQuggest a slightly increased incidence of pleural effusion in patients with AMI-P, but with a frequency several orders of magnitude below that of the present study. A definitive explanation for the pulmonary findings cannot be given; however, there are several possibilities to consider. It has been suggested that the observed pulmonary changes are purely the result of congestive heart f a i l ~ r eFifty-five .~ percent of our patients with AMI-P and pulmonary abnormalities had radiographic findings of congestive heart failure; however, the remaining 45 percent showed no such evidence. The AMI-C group demonstrated the same incidence of radiographic congestive heart failure as did the AMI-P group, yet there were no patients with pleural effusion and only seven (14 percent) with pulmonary infiltrates in the AMI-C group. Furthermore, there is a tendency for congestive heart failure not to produce unilateral left-sided pleural effusion,1° yet the pleural and parenchymal findings in our patients had a strong predilection to occur on the left. While it is quite possible that CHEST, 67: 6, JUNE, 1975
congestive heart failure contributes to the pulmonary changes observed in some patients with AMI-P, for the above reasons we feel it is unlikely to be the only factor in their genesis. Another possibility is that the pleuritic pain experienced by patients with AMI-P is indirectly responsible for the observed changes in x-ray film findings. The rapid, shallow respiratory pattern adopted by most patients with pleuritic pain could conceivably produce basilar atelectasis. It would, however, be difficult to explain the predominantly unilateral infiltrates and the pleural effusions on the basis of decreased thoracic motion. The PMIS is known to occur occasionally early in the course of acute myocardial infarction,"-l4 and pulmonary infiltrates and effusions are frequent features of this syndrome.15 Although an early onset of PMIS may account for the radiologic abnormalities we observed, it still does not explain the mechanism by which they occur. Furthermore, it would imply a much higher frequency for the PMIS than has previously been de~cribed.~ The x-ray film findings in idiopathic pericarditis, PMIS, and the postcommissurotomy syndrome are very similar.15These abnormalities, with the exception of pericardial effusion, are also similar to the abnormalities seen in patients with AMI-P.16 The radiographic similarities between the various types of pericarditis may be more than a coincidence. It is conceivable that regardless of inciting events or agents, pericarditis could produce the pulmonary abnormalities by direct extension of the inflammatory process to the adjacent pleura and lung. In the patients with AMI-P, congestive heart failure was "overdiagnosed" when clinical criteria alone were used ( Fig 1 ) . Pain-related tachycardia, tachypnea, and basilar rilles resulting from the pulmonary infiltrates were probably responsible for the erroneous diagnosis of congestive heart failure in most cases. Early in the course of AMI-P a short, early diastolic friction rub could easily be interpreted as a ventricular gallop, adding more spurious data to support the diagnosis of congestive heart failure. These pitfalls can be avoided if the chest x-ray film is carefully studied before making a diagnosis of congestive heart failure in a patient with AMI-P. The incidence of the PMIS was high in our surviving patients with AMI-P. In addition to the five patients with typical PMIS, three may have had AMI-P blending imperceptibly into the PMIS. Even if these three cases are excluded, the 15 percent incidence of the PMIS is much higher than the 1 to 4 percent incidence cited for all patients with acute myocardial infarction. 4 . 1 7 PERlCARDlTlS OF ACUTE MYOCARDIAL INFARCTION 651
It would appear that the patient who has had pericarditis from any of several causes remains at risk for recurrent attacks. For this reason, it is especially important to document pericarditis occumng during acute myocardial infarction and record it in the medical record, or problem list, along with the other complications of acute myocardial infarction. Ready availability of this information would be of great assistance in the management of a patient who presents with pleuritic chest pain and a pulmonary infiltrate several weeks following AMI-P. Pulmonary embolus or infarction is an obvious diagnostic possibility in this situation, but the PMIS should also be given strong consideration. The distinction between the PMIS and pulmonary infarction has special significance because of the reported risk of pericardial hemorrhage in patients with PMIS who are also on anticoagulant It has been our policy not to prescribe anticoagulants for such patients unless lung scans or pulrnonary angiograms demonstrate strong evidence of pulmonary embolus. anticoagulation is According to many authoritie~,~ not contraindicated in AMI-P. Niarchos,* however, reported significant hemopericardium in 2 out of 22 patients with AMI-P and we suspected two cases of tamponade, one of which showed hemopericardium at autopsy 16 days later. All patients, in both series, with suspected or proved pericardial tamponade were receiving anticoagulant therapy at the time of the suspected hemorrhage. Thus, there appears to be a small (1.5 percent) but definite risk in using anticoagulants in patients with AMI-P. Our current policy is to discontinue anticoagulant therapy at the onset of pericarditis if the only indication for it is prophylaxis against venous thromboembolism. When a specific indication for anticoagulation exists (such as phlebitis or pulmonary embolus ), anticoagulation is continued with careful monitoring of blood pressure, pulsus paradoxus and neck veins for signs of pericardial tamponade. All of our patients with AMI-P had definite electrocardiographic documentation of a transmural myocardial infarction. It is conceivable that pericarditis could occur with a nontransmural infarction; however, most authors"' have not found this to be the case. Both this study and the literature suggest that patients who present with pericarditis and do not evolve electrocardiographic changes of a transmural infarction are unlikely to have AMI-P as the etiology of their pericarditis. The pain of pencarditiscan mimic myocardial infarction in every detail including radiation down one or both arms and it may be impossible to separate idiopathic pericarditis from AMI-P on the basis of history alone. 652 TOOLE, SILVERMAN
Since almost all of our symptomatic patients (31 to 37) received parenterally administered stexoids, we can make no objective assessment of the therapeutic efficacy of steroids in AMI-P. Our strong impression is that a significant number of patients obtained dramatic relief from constant pain within 30 minutes after the steroid injection. Many of these patients had responded poorly or not at all to high doses of intravenous narcotics. In fact, severe pain which did not respond to opiates was often the first clue to the development of pericarditis. The use of steroids allowed us to avoid completely or greatly decrease the dosage of narcotics, thereby avoiding drug-induced respiratory depression or hypotension. The patients who continued to have some painful symptoms after the initial steroid injection were given steroids orally in tapering dosage over a three to five day period. We observed no relapses of symptomatic pericarditis following steroid withdrawal. A single instance of aggravated diabetic control was the only complication attributable to the steroid therapy. Early reportslg suggested an increased mortality in AMI-P compared to acute myocardial infarction without pericarditis. More recent information," including the current study, indicates no statistically significant increase in mortality for the AMI-P group. An unexplained finding in our patients with AMI-P was the high incidence of late hospital deaths. All late deaths followed a prolonged and complicated course, during which time the patient remained severely ill. There were no sudden or unexplained deaths to suggest that this group might benefit from extended cardiac monitoring. ACKNOWLEDGMENTS: We thank Dr. Robert C. Schlant for editorial assistance and encouragement in the preparation of this manuscript. We also thank the nursing staff of the Coronary Care Unit, Piedmont Hospital, without whom this study could not have been conducted.
1 Osler W: Modem Medicine: Its Theory and Practice (Osler W, ed) (vol 4 ) . Philadelphia, Lea & Febiger, 1908, p 58 2 World Health Organization: WHO Report Euro 5010. Ischemic Heart Disease Registers. Report of Working Group. Copenhagen, World Health Organization, 1968 3 Blumer G : Pericarditis epistenocardia. JAMA 107:178181,1936 4 Niarchos AP, McKendrick CS: Prognosis of pericarditis after acute myocardial infarction. Br Heart J 35:49-54, 1973 5 Thadani V, Chopra MA, Aber CP, et al: Pericarditis after acute myocardial infarction. Br Med J 2:135-37, 1971 6 Lichstein E, Liu H, Gupta P: Pericarditis complicating acute myocardial infarction: incidence of complications and significance of electrocardiogram on admission. Am Heart J 87:246-252, 1974
CHEST, 67: 6, JUNE, 1975
7 Barman PC, Krishnaswami V, Geraci AR: Pericarditis in acute myocardial infarction. NY State J Med 72:645-48, 1973 8 Yan V: Pericarditis in acute myocardial infarction. Heart & Lung 3:247-251, 1974 9 Weiser NJ, Kantor M, Russell HK, et al: Post myocardial infarction syndrome: Nonspecificity of pulmonary manifestations. Circulation 25:643-650, 1962 10 Meszaros WT: Lung changes in left heart failure. Circulation 47 :859-871, 1973 11 Stein I, Prussin G: Early appearance of the post myocardial infarction syndrome. Angiography 20:26268, 1969 12 Ananthasubrarnaniam G, Muthukrishman TV: Pericarditis in myocardial infarction. Indian Heart J 23:302-05, 1971 13 Weiser NJ, Kantor M, Russell HK: Post myocardial in-
farction syndrome. Circulation 20:371-380, 1959 14 Kossowsky WA, Epstein PJ, Levine RS: Post myocardial infarction syndrome: An early complication of acute my* cardial infarction. Chest 63:35-40, 1973 15 Dressler W: Idiopathic recurrent pericarditis. Am J Med 18:591601,1955 16 Maude1 W, Johnson EC: Pleuropericardial effusion following myocardial infarction. Am Heart J 53: 145-49, 1957 17 Dressler W: The post myocardial infarction syndrome: A report of 44 cases. Arch Intern Med 103:28-42, 1959 18 Dressler W: Management of pericarditis secondary to myocardial infarction. Progr Cardiovasc Dis 3:134-140, 1980 19 Levine SA: Coronary thrombosis: Its various clinical features. Medicine (Baltimore) 8:245-418, 1929
The Penguin
'
The penguin spends more than half of each year on islands. Although he is a bird, he cannot fly. His element is the sea. He can swim almost as swiftly a s a shark. Fish is his only diet-a variety of small sardine, peculiarly rich in oil. When night arrives, he floats in the attitude of a duck and sleeps. He propels himself, when swimming, entirely by means of his flippers. A transparent film comes over his eyes the moment he goes under the water. In February and September the birds will mate and in the next six or eight weeks are making for their island to lay their eggs. They dig in a sloping direction a sort of tunnel to a depth of one, two or three feet. The two
CHEST, 67: 6, JUNE, 1975
penguins that mated take their turns in the hatching of the egg for about twelve hours each for four weeks. Two or three days after the first egg, there comes another; there may be a third or even a fourth. The ibis and the gull take not only eggs but also young chicks. The chicks are fed every twenty minutes. The parent's digestive machinery turns into oil the food collected and this oil is brought back into the beak. Then the young chicks insert their own small beaks into the parent's larger beak and so draw the nourishment. Kearton C: The Island of Penguins. New York, McBride, 1931
PERlCARDlTlS OF ACUTE MYOCARDIAL INFARCTION 653