Clin. Cardiol.

29, 363368 (2006)

Prognostic Implications of PR-Segment Depression in Inferior Leads in Acute Inferior Myocardial Infarction
MAN-HONG JIM, M.D., CHUNG-WAH SIU, M.D., ANNIE ON-ON CHAN, M.D., PH.D., RAYMOND HON-WAH CHAN, M.D., STEPHEN WAI-LUEN LEE, M.D., CHU-PAK LAU, M.D. Department of Medicine, Queen Mary Hospital, Hong Kong, P.R.C.

Summary

Background: Concurrent atrial ischemia is usually overlooked in acute myocardial infarction (MI) due to its subtle electrocardiographic (ECG) changes, lack of clear-cut clinical picture, and prognostic significance. PR-segment depression in the inferior leads is a simplified ECG sign for detecting possible underlying atrial ischemia. Hypothesis: The purpose of this study was to document the incidence, clinical characteristics, and prognostic implications of this ECG sign in the setting of acute inferior MI. Methods: Demographics, clinical characteristics, and outcomes of 463 consecutive patients presenting with acute inferior MI were reviewed. The in-hospital ECG was examined by two independent reviewers. The results were then compared between those with and without ECG sign. Results: Profound PR-segment depression 1.2 mm in inferior leads was found in 9 of 463 (1.9%) patients. Patients with atrial ischemia tended to present earlier (2.4 2.6 vs. 7.0 8.2 h, p = 0.000) and had a higher frequency of first-degree atrioventricular block (77.8 vs. 30.6%, p = 0.028) and supraventricular arrhythmias (55.5 vs. 20.2%, p = 0.022). Of greater importance, it was significantly associated with an increased rate of cardiac free-wall rupture (33.3 vs. 2.0%, p = 0.001) and in-hospital mortality (44.4 vs. 11.7%, p = 0.015). Conclusion: Profound PR-segment depression 1.2 mm in inferior leads was associated with a complicated hospital

course and poor short-term outcome in acute inferior MI. These patients were at high risk for the development of atrioventricular block, supraventricular arrhythmias, and cardiac free-wall rupture.

Key words: PR-segment depression, atrial infarction, in-hospital mortality, cardiac free-wall rupture

Introduction
In acute myocardial infarction (MI), the electrocardiographic (ECG) sign of PR-segment depression could signify extensive atrial ischemia or infarction, which is usually overlooked because of lack of a clear-cut clinical picture and uncertain prognostic implications.17 To date, no consensual diagnostic criteria of atrial infarction have been derived.2, 812 Indeed, the categorization into atrial infarction may not be meaningful as there are no other antemortem means to verify the accuracy of those criteria. From a practical point of view, a simple ECG sign that can predict clinical outcome may even be more relevant. PR-segment depression 1.2 mm in inferior leads is a simplified, truncated, and modified ECG sign derived from Lius criteria (Table I). It was chosen as the prognostic indicator in this study because it was measurable and simple to use with a strong electrophysiologic basis. In acute inferior MI, the infarct-related artery is either the left circumflex or the right coronary artery, both of which provide blood supply to the atria. The aim of this study was to document the incidence, clinical characteristics, and prognostic implications of concomitant PR-segment depression 1.2 mm in inferior leads in a cohort of patients with acute inferior MI.

Address for reprints: Dr. Man-Hong Jim Associate Consultant Grantham Hospital 125 Wong Chuk Hang Road Hong Kong, P.R.C. e-mail: jimmanh2002@yahoo.com Received: February 15, 2006 Accepted with revision: April 28, 2006

Methods
Study Population

From January 1997 to December 2004, hospital records and outpatient files of all patients admitted to the coronary care

364
TABLE I

Clin. Cardiol. Vol. 29, August 2006

Lius diagnostic criteria of atrial infarction

Electrocardiography

Major criteria 1. PR-segment elevation > 0.5 mm in leads V5 and V6 with reciprocal PR-segment depression in leads V1 and V2 2. PR-segment elevation > 0.5 mm in lead I with reciprocal depression in leads II and III 3. PR-segment depression > 1.5 mm in precordial leads and 1.2 mm in leads I, II, and III associated with any atrial arrhythmias Minor criteria 1. Abnormal P waves: M-shaped, W-shaped, irregular, or notched

Electrocardiograms were performed every 6 h on the first 2 days, and then once daily. Cardiac arrhythmias were documented by telemetry. The ECGs were magnified two-fold, and measurements were performed with calipers. PR-segment depression was defined as 1.2 mm with reference to the preceding TP segment in all inferior leads (Figs. 1A, 2A). If no clearcut TP segment could be defined, the vertical distance from the beginning to the end of PR segment was measured (Figs. 1B, 2B). The serial ECGs of each patient were examined separately by two investigators blinded to the clinical information. A consensus was reached in the case of disagreement.
Clinical Definitions

unit at our hospital with a discharge diagnosis of acute inferior MI were retrieved. Acute inferior MI was diagnosed if there was (1) typical chest pain lasting for 30 min, (2) ST-segment elevation of 1 mm in any two of the inferior leads (II, III, and aVF), or (3) increase of serum creatinine kinase (CK) > 2 times the upper limit of normal. The demographics, clinical information, treatment strategy, disease complications, and clinical outcome data were reviewed. Death during the follow-up period was identified by a computer system linked with all major government hospitals in Hong Kong.

Advanced atrioventricular block was defined as either second- or third-degree atrioventricular block. Cardiac free-wall rupture was defined as hemodynamic collapse or sudden death due to electromechanical dissociation associated with pericardial effusion on echocardiography.
Statistical Analysis

The data were presented as the mean value 1 standard deviation. Differences between groups with respect to continuous variables were compared by the two-sided t-test. Two-sided

PR-segment depression

II

II 1mV

III

III

aVR

aVR

(A)

0.2 s

1mV aVL aVL

PR-segment depression

aVF aVF (A) (B) (B) 0.2 s

FIG. 1 Profound PR-segment depression in inferior leads with (A) and without (B) clear-cut TP segment in acute inferior myocardial infarction. Note also ST-segment elevation in inferior leads.

FIG. 2 Measurement of PR-segment depression with (A) and without (B) clear-cut TP segment.

M.-H. Jim, et al.: PR-segment depression and in-hospital complications

TABLE II Comparison of clinical presentations and outcome in patients with and without inferior PR-segment depression PR-segment depression n = 9 (1.9%) Age Male sex (%) Smoking history (%) Diabetes mellitus (%) Hypertension (%) Hyperlipidemia (%) Echocardiogarphic findings Mean left ventricular ejection fraction (%) Pericardial effusion (%) Arrhythmias First-degree heart block (%) Advanced heart block (%) Supraventricular arrhythmias (%) Ventricular arrhythmias (%) Complications Cardiac rupture (%) Clinical outcome In-hospital mortality (%)
a p < 0.05.

365

Controls n = 454 (98.1%) 64.9 13.1 338 (74.4) 251 (55.3) 135 (29.7) 197 (43.3) 212 (46.7) 48 11 13 (2.7) 139 (30.6) 80 (17.6) 92 (20.2) 29 (6.4) 9 (2.0) 53 (11.7)

p Value 0.5 0.22 0.19 0.36 0.47 0.89 0.16 < 0.0001 a 0.028 a 0.26 0.022 a 0.6 0.001 a 0.015 a

67.9 10.4 5 (55.5) 3 (33.3) 4 (44.4) 5 (55.5) 4 (44.4) 41 13 4 (44.4) 7 (77.8) 3 (33.3) 5 (55.5) 1 (11.1) 3 (33.3) 4 (44.4)

chi-square or Fishers exact test (whenever an expected cell value was < 5) was used for analysis of categorical variables. A p value of < 0.05 was considered statistically significant.

Results
II

II

From January 1997 to December 2004, 463 patients were admitted with acute inferior MI. Nine patients were found to have PR-segment depression satisfying the diagnostic criteria, reflecting an incidence of 1.9%. The remaining 454 patients without significant PR-segment depression served as controls.
Clinical Presentation (Table II)

III

III

The mean age of patients was 67.9 10.4 years, with a slight male predominance (55.6%). Smoking history was found in three of nine (33.3%), diabetes mellitus in four of nine (44.4%), hypertension in five of nine (55.5%), and hyperlipidemia in four of nine (44.4%) patients. There was no statistical difference in the demographics and prevalence of cardiovascular risk factors compared with controls. All patients presented < 12 h from onset of symptoms, and eight of the nine (88.9%) patients presented in 6 h. The average time of presentation was 2.4 2.6 h, which was significantly earlier than that for the controls (7.0 8.2 h, p = 0.000). Concurrent PR-segment depression was present in the precordial leads (leads V13) in six of nine (66.7%) patients. Compared with ST-segment elevation, PR-segment depression resolved faster in five patients and at the same rate in two patients (Fig. 3 A, B). PR-segment depression disappeared after a

aVR

aVR

aVL

aVL

aVF

aVF

(A)

(B)

FIG. 3 Resolution of PR-segment depression with (A) and without (B) residual ST-segment elevation.

366

Clin. Cardiol. Vol. 29, August 2006

mean of 4.0 2.5 h in these seven patients. The remaining two could not be differentiated, as both ECG changes were still persistent upon the time of death. The mean heart rate on presentation was 111 27 beats/min; the median heart rate was 112 beats/min. The mean sum of ST-segment elevation in all inferior leads was 7.1 3.6 mm, which was not significantly different from the controls (7.5 4.2 mm).
Treatment and Clinical Course (Table II and III)

First-degree atrioventricular block was seen in seven of nine (77.8%) patients, compared with 139 of 454 (30.6%) patients serving as controls (p = 0.028). Five of the seven patients had prolonged first-degree heart block lasting > 24 h. Advanced atrioventricular block was documented in three of nine (33.3%) patients. Two had conversion to first-degree heart block within 24 h, while the remaining patient remained pacemaker dependent for 1 week. Supraventricular arrhythmias occurred in five of nine (55.5%) patients: three had atrial tachycardia while two had atrial fibrillation. The difference was also statistically significant (55.5 vs. 20.3%, p = 0.022). However, there was no difference in the incidence of ventricular arrhythmias between the two groups (11.1 vs. 6.4%, p = 0.6). Four of the nine patients with PR-segment depression were treated with thrombolytic therapy, two with primary angioplasty, and three were managed conservatively. Coronary angiography was subsequently performed in six patients. The left circumflex artery was the culprit vessel in three patients (two with proximal obstruction and one with distal obstruction) and the right coronary artery in the remaining three patients (one with proximal obstruction and two with distal obstruction). The incidence of single- and multivessel disease was equal. Mean peak creatine kinase (CK) level was 4534 3216 U/l, significantly higher than in the controls. Mean left ventricular ejection fraction was 41 13%. Pericardial effusion, found on the first day of the infarct, was detected in four of nine (44.4%) patients. On the other hand, 13 patients in the controls had pericardial effusion, 11 had PR-segment depression < 0.5 mm, and only 2 had PR-segment depression between 0.51.2 mm.

Cardiac free-wall rupture developed in three of nine (33.3%) patients, resulting in in-hospital death in all. Two died within 24 h of presentation and one on Day 3 after primary angioplasty. PR-segment depression was significantly associated with free-wall rupture compared with controls (33.3 vs. 2.0%, p = 0.001). One other patient died of heart failure on Day 6. None of the nine control patients with cardiac rupture showed any significant PR-segment deviation. The presence of PR-segment depression was associated with increase in inhospital mortality (44.4 vs. 11.7%, p = 0.015). During followup, one more patient died at 3 months; the remaining four patients survived >1 year.

Discussion
PR-segment depression is a reflection of underlying inflammatory conditions of the heart, for example, acute pericarditis,13 pericardial effusion,14 or atrial ischemia. Acute pericarditis or Dresslers syndrome is a late complication after acute MI and unlikely to be the cause, as all ECG changes were detected very early after onset of symptoms. Early pericardial effusion preceding rupture could be the other explanation, but in most of the patients with pericardial effusion (64.8%) PR-segment depression was < 0.5 mm. On the other hand, there was no pericardial effusion in more than half of PR-segment depressions 1.2 mm. Thus, pericardial effusion could possibly contribute, but not be the sole cause of profound PR-segment depression. In the present study, profound PR-segment depression probably represented a subgroup of patients who had extensive atrial ischemia or infarction in addition to acute inferior MI. The early appearance of PR-segment deviation after ligation of the atrial branches has been confirmed in experimental animal models.15, 16 Some investigators even suggest that, in humans, the PR-segment changes appear before ST-segment elevation in acute MI and persist for a few hours.1 In accordance with these animal studies, PR-segment depression occurred early and persisted for an average of 4.0 h only. Most studies on atrial infarction were autopsy studies, and there is no major clinical study. We found that the incidence of profound PR-segment depression was 1.9% among patients with acute inferior MI who survived to hospitalization. In addition, all presented almost within 6 h, and this ECG sign lasted only for 4 h. It is possible that some patients with atrial infarction may have died before admission or were missed if they presented late. The true incidence of atrial infarction is likely to be higher. Tachycardia may interfere with the measurement of PR segment. In normal persons during exercise, the onset of PRsegment depression coexists with ST-segment depression so that the PR and ST segments form an imaginary arc. This physiologic change is masked by the presence of ST-segment elevation in acute MI. To date, there has been no formula for the correction of PR-segment magnitude in response to changes in heart rate. The mean and median heart rate of PRsegment depression was approximately 110 beats/min. This heart rate is not likely to exert any significant effect on the PR-

TABLE III Comparison of peak creatine kinase (CK) level and treatment received in patients presented within 12 h with and without inferior PR-segment depression PR-segment depression n = 9 (2.4%) Treatment Thrombolytics Primary angioplasty Conservative Peak CK level (U/l)
a p < 0.05.

Controls n = 371 (97.6%) 244 (65.3) 29 (8.2) 98 (26.4) 2620 2289

p Value 0.19 0.1 0.13 0.015 a

4 (44.4) 2 (22.2) 3 (33.3) 4535 3216

M.-H. Jim, et al.: PR-segment depression and in-hospital complications

367

segment magnitude. Tachycardia was very uncommon in the controls; the higher mean heart rate in the PR-segment depression group may actually represent a bias, denoting enhanced atrial irritability. The voltage change involving PR-segment deviation is very subtle. A large territory of the atrium needed to be jeopardized before any significant change could be detected. Profound PR-segment depression was associated with higher peak CK level, implying more extensive myocardial injury. This finding was not supported by the angiographic findings, which showed the same incidence in the pattern of culprit vessel (left circumflex vs. right coronary artery), culprit location (proximal vs. distal obstruction), and number of diseased vessels (single- vs. multivessel) involvement. However, the small number of index cases precluded any meaningful angiographic interpretation. Ischemia involving the anterior part of the atrium around the atrioventricular node area may have accounted for the high incidence of all degrees of atrioventricular block, which tended to last longer than was observed in those reflex mediated in the absence of atrial infarction. The large area of irritable atrial muscle predisposes to arrhythmogenesis. The reported incidence of supraventricular arrhythmias in atrial infarction ranges from 61 to 74%,6, 7, 17 similar to the incidence of 55.5% in this study. The resultant atrial tachyarrhythmias were characterized by sudden onset and termination,18 they further compromise the hemodynamics in the setting of acute MI. Free-wall rupture is an uncommon but life-threatening condition. It was reported in 2.2 to 2.7% of patients, complicating acute MI and carrying a grave prognosis.1921 In the present study, PR-segment depression was associated with an exceptionally high rate of free-wall rupture of 33.3% compared with 2.0% in those without PR-segment deviation. The site of rupture was uncertain as no postmortem study was performed. Given the low incidence of ventricular rupture in acute inferior MI, we believe that the increased risk of freewall rupture could be attributed to atrial rupture. The extensive necrosis and transmural involvement of the thin wall atria probably explained the high free-wall rupture rate in atrial infarction. Indeed, atrial rupture was reported in 4.5% of the autopsy series of atrial infarction.9 The presence of profound PR-segment depression was associated with increased in-hospital mortality in the setting of acute inferior MI. From a physiologic point of view, atrial infarction closely resembles atrial fibrillation in the sense that both have loss of atrial kick, which contributes to deterioration of hemodynamics. Atrial fibrillation is a well-known poor prognostic marker of both short- and long-term mortality in acute MI.9, 22 In this study, the increase in in-hospital mortality was clearly attributed to an increase in the free-wall rupture rate. Whether uncomplicated atrial infarction per se increases short-term mortality is unknown. Among the survivors, all except one lived beyond 1 year. The patient number in this study was too small to give a definite answer on the effect of atrial ischemia on long-term prognosis. The relationship of PR-segment depression and PR interval is not a linear one. After the P wave, the down-stroke continues

to form the steep part of the curve, followed by a plateau phase. The majority of the magnitude of PR-segment depression was reached within the initial 60 ms; the plateau part only contributed the terminal 0.10.2 mm. Therefore, it is not surprising to see the association of PR-segment depression with atrioventricular block, since the latter simply allows more time to reach the diagnostic criteria. No specific formula could be used for correction of the PR interval. It is interesting to see the effect of lowering the diagnostic criteria to 1.0 mm. By doing this, 14 more patients fell into the zone, including 1 more with cardiac rupture and 4 more suffering in-hospital mortality, bringing the incidence, cardiac rupture rate, and in-hospital mortality to 5.0, 21.7, and 34.8%, respectively. The cardiac rupture rate (21.7 vs. 1.8%, p = 0.001) and in-hospital mortality (34.8 vs. 11.1%, p = 0.02) were still significantly worse than those in the controls.
Limitations

The small number of patients with profound PR-segment depression is the major drawback in the present study. Tachycardia may interfere with measurement of the PR segment, and this effect could not be totally excluded. Without autopsy in all deaths, the exact site of cardiac rupture cannot be ascertained but, based on reported pathologic studies, is likely to be from the atrium. This low incidence may be due to the inherent high mortality associated with this condition. It is possible that other ECG signs may be more sensitive, but the present sign is simple and effective for identifying a high-risk cohort. As PR segment tended to resolve spontaneously, it is possible that this ECG sign may not be seen in patients surviving inferior MI who presented late. This may have a variable effect on the incidence and clinical outcome of atrial ischemia observed in this study. Nevertheless, the poor outcome in patients with this ECG sign remained relevant.

Conclusion
Profound PR-segment depression 1.2 mm in inferior leads was found in 1.9% of patients with acute inferior MI. This ECG sign represents a subgroup of patients with extensive atrial ischemia. Patients usually presented very early, mostly within 6 h. This sign was associated with a high frequency of atrioventricular block and supraventricular arrhythmias. More important, it was associated with the development of cardiac rupture, with high in-hospital mortality.

References
1. 2. James MR: Atrial myocardial infarction. Arch Intern Med 1984;144: 573574 Cushing EH, Feil HS, Stanton EJ, Wartman WB: Infarction of the cardiac auricles (atria): Clinical, pathological and experimental studies. Br Heart J 1942;4:1734 Bean WB: Infarction of the heart. Ann Intern Med 1938;12:7194 Wartman WB, Saunders JC: Localization of myocardial infarcts with respect to the muscle bundles of the heart. Arch Pathol Lab Med 1950;50: 321364

3. 4.

368
5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

Clin. Cardiol. Vol. 29, August 2006

McCain FH, Kline EM, Gilson JS: A clinical study of 281 autopsy reports on patients with myocardial infarction. Am Heart J 1957;39:263272 Gardin JM, Singer DH: Atrial infarction: Importance, diagnosis and localization. Arch Intern Med 1981;141:13451348 Lazar EJ, Goldberger J, Peled H, Sherman M, Frishman WH: Atrial infarction: Diagnosis and management. Am Heart J 1988;116:10581063 Liu CK, Greenspan G, Piccirilo RT: Atrial infarction of the heart. Circulation 1961;23:331338 Soderstrom N. Myocardial infarction and mural thrombosis in the atria of the heart. Acta Med Scan Suppl 1948;217:1114 Dilelsi AJ, Piky HA, Eynon HK: Auricular infarction: Reports of two cases. Ann Intern Med 1952;36:640647 Freundlich J, Sereno LR: Auricular infarction. Am Heart J 1959;57: 654660 Hellerstein HK: Atrial infarction with diagnostic electrocardiographic findings. Am Heart J 1948;36:422430 Baljepally R, Spodick DH: PR-segment deviation as the initial electrocardiographic response in acute pericarditis. Am J Cardiol 1998;81(12): 15051506 Kudo Y, Yamasaki F, Doi Y, Sugiura T: Clinical correlates of PR-segment depression in asymptomatic patients with pericardial effusion. J Am Coll Cardiol 2002;39(12):20002004 Sanders A: Experimental localized auricular necrosis: Electrocardiographic study. Am J Med Sci 1939;198:690694 Abramson DI, Fenichel NM, Shookhof C: A study of the electrical activity of the auricles. Am Heart J 1938;15:471481 17. Nielson FE, Andersen HH, Gram-Hansen P, Sorensen HT, Klausen IC: The relationship between ECG signs of atrial infarction and the development of supraventricular arrhythmias in patients with acute myocardial infarction. Am Heart J 1992;123:6972 18. Van Durme JP, Bossaert L, Vermeire P, Pannier R: Intra-artrial electrocardiogram in the diagnosis of atrial infarction. Arch Mal Coeur 1972;65(7): 887889 19. Tanaka K, Sato N, Yasutake M, Takeda S, Takano T, Takana S: Clinical course, timing of rupture and relationship with coronary recanalization therapy in 77 patients with ventricular free wall rupture following acute myocardial infarction. J Nippon Med Sch 2002;69:481488 20. Moreno R, Lopez de Sa E, Lopez-Sendon JL, Garcia E, Soriano J, Abeytua M, Elizaga J, Botas J, Rubio R, Moreno M, Garcia-Fernandez MA, Delcan JL: Frequency of left ventricular free-wall rupture in patients with acute myocardial infarction treated with primary angioplasty. Am J Cardiol 2000; 85:757760.A8. 21. Moreno R, Lopez-Sendon JL, Garcia E, Perez de Isla L, Lopez de Sa E, Ortega A, Moreno R, Rubio R, Soriano J, Abeytua M, Garcia-Fernandez MA: Primary angioplasty reduces the risk of left ventricular free wall rupture compared with thrombolysis in patients with acute myocardial infarction. J Am Coll Cardiol 2002;39:598603 22. Pizzetti F, Turazza FM, Franzosi MG, Barlera S, Ledda A, Maggioni AP, Santoro L, Tognoni G: GISSI-3 Investigators. Incidence and prognostic significance of atrial fibrillation in acute myocardial infarction: The GISSI-3 data. Heart 2001;86(5):527532