a b s t r a c t

Objectives: In acute inferior ST-segment elevation myocardial infarction , multiple criteria have been pro- posed to predict the culprit artery based on the 12-lead electrocardiogram . We assessed the utilities of 11 traditional and 2 new criteria to devise a new ECG algorithm to localize the culprit artery in acute inferior STEMI. Methods: We analyzed electrocardiographic and angiographic findings of 194 consecutive patients with acute in- ferior STEMI to devise a new ECG algorithm, further validated in another cohort of 80 patients with acute inferior STEMI.

Results: In derivation cohort, the 2 new criteria including (1) ST-segment depression in lead I equal to half of that in lead aVL and (2) equal ST-segment elevation in leads II, III, and aVF did not prove useful. The most powerful electrocardiographic criteria were (1) the ratio of ST elevation in lead III to that in lead II, (2) the ratio of ST de- pression in lead I to that in lead aVL, and (3) ST changes in lead I; these formed a 3-step algorithm. Application of this algorithm suggested the location of the culprit artery in 192 of 194 patients (nearly 99%) in the derivation cohort. In validation cohort, the algorithm possessed a sensitivity and specificity of 100% and 89%, respectively, for predicting the right coronary artery and 89% and 100%, respectively, for predicting the left Circumflex artery. Conclusions: A new 3-step algorithm based on 12-lead ECG is proposed to localize the culprit artery at the bedside of acute inferior STEMI patients before primary percutaneous coronary intervention, allowing immediate deci- sions about therapy.

(C) 2016

Introduction

Acute inferior ST-segment elevation myocardial infarction is a fairly heterogeneous entity, which is usually caused by the occlusion of the right coronary artery ; less often, the left circumflex artery (LCx) or, rarely, the left anterior descending artery (LAD) may be the cause [1]. The severity of inferior STEMI depends on the infarct- related artery and its size [2]. It can be limited to the posterobasal, dia- phragmatic, or posterolateral segments; or it can involve 2 or 3 seg- ments. Nearly 50% of patients with acute inferior STEMI have specific hemodynamic and bradycardic complications, usually due to the total

? Sources of grants: This work was supported by grants from The National Natural Science Foundation of China (nos. 81100132 and 81170138), the Fundamental Research Funds for the Central Universities (no. 1191320117), and the First Affiliated Hospital of XJTU Fund (No. 2015yk26).

* Corresponding author at: Department of Cardiology, First Affiliated Hospital of Xi’an Jiaotong University, No. 277 Yanta West Rd, Xi’an 710061, Shaanxi, PR China.

E-mail address: [email protected] (N. Guo).

¹ They contributed equally to this work

occlusion of the proximal RCA [3,4]. It is important to identify high- risk subgroups of Inferior myocardial infarction with increased mortali- ty and morbidity [5-10].

The electrocardiogram (ECG) is an easily obtained, noninvasive method to diagnose acute myocardial infarction on admission. Timely identification of culprit coronary artery using ECG could provide clini- cally important information for early risk stratification to augment deci- sion making and tailor reperfusion therapy. Previous studies [11-26] have proposed ECG criteria based on ST-segment “ups and downs” var- iation to identify the culprit coronary artery after acute inferior STEMI. The purpose of this study is to devise an effective algorithm to localize culprit coronary artery in acute inferior STEMI.

Materials and methods

Study population

In the present study, all patients with acute inferior STEMI admitted to our coronary care unit and treated with primary percutaneous coro- nary intervention (PCI) between January 1, 2010, and December 31, 2011, were assessed as derivation cohort; and patients admitted

http://dx.doi.org/10.1016/j.ajem.2016.06.005

0735-6757/(C) 2016

electrocardiographic characteristic”>between January 1, 2012, and October 31, 2012, as validation cohort. Pa- tients were included if they had been admitted within 12 hours from symptom onset, had sinus rhythm on admission ECG, and had the diag- nosis of first acute inferior myocardial infarction defined as (1) chest pain or discomfort for more than 30 minutes, (2) ST elevation greater or equal to 1 mm in 2 or more inferior leads (II, III, or aVF), (3) transient in- crease in cardiac enzymes to more than 2-fold the normal laboratory value, and (4) the development of new Q waves. Patients with electro- cardiographic evidence of right or Left bundle-branch block, left ventric- ular hypertrophy, accelerated ventricular rhythm, paced rhythm or severe artifacts, history of previous myocardial infarction, or previous Coronary artery bypass grafting or PCI were excluded from the study. The patient with the culprit artery of LAD verified by coronary angiogra- phy was also excluded (n = 1). Informed consent was given by all pa- tients before their inclusion. The study protocol was approved by the clinical research Ethics Committee of the First Affiliated Hospital of Xi’an Jiaotong University. Finally, 194 patients comprised the derivation cohort, and 80 patients were in the validation cohort.

Electrocardiogram

The standard 12-lead ECG, recorded on admission of all patients using a paper speed of 25 mm/s and a standardization of 1 mV/10 mm, was analyzed by 2 investigators who were blinded to coronary an- giography findings. Electrocardiograms were recorded using Cardiofax ECG-9130K or ECG-9020P machine (Nihon Kohden Corporation, Tokyo, Japan). Any disagreement between the investigators was re- solved by consensus. The TP segment was used as the isoelectric line; the PR segment was used when the P wave and the T wave merged. The J point was determined for each lead independently. Both ST eleva- tion and ST depression were measured at 80 ms after the J point in all leads. The utilities of 11 traditional electrocardiographic criteria previ- ously reported in the prediction of RCA or LCx occlusion in acute inferior STEMI were evaluated in the derivation cohort listed as follows:

Lead I: ST-segment depression >=0.5 mm; ST isoelectric; ST-segment elevation >= 0.5 mm [13,15].

Ratio of ST-segment elevation in lead III to that in lead II: N 1, 1, b 1 [13,26,27].

?ST-segment depression in leads V1-V3 vs ?ST-segment eleva- tion in leads II, III, and aVF: N 1, <= 1 [14].

Lead V1: ST-segment depression; ST isoelectric; ST-segment eleva- tion [23].

Lead aVR: ST-segment depression >= 1 mm; ST-segment elevation, ST isoelectric, ST-segment depression b 1 mm [20].

Lead aVL: ST-segment depression; ST isoelectric; ST-segment eleva-

tion [15].

Ratio of ST-segment depression in lead V3 to ST-segment elevation in lead III: <=1.2, N 1.2 [19].

Ratio of ST-segment depression in lead I to that in lead aVL: <= 1, N 1 [11,16].

ST-segment depression in lead aVR >= aVL [22].

ST-segment elevation in leads V5 or V6 >= 1 mm [1].

Maximum ST-segment depression in leads V1-V3 (ST-segment de- pression >= 1 mm in >= 1 precordial lead with the sum of ST-segment depression greater in leads V1 to V3 than in leads V4 to V6) [28].

In clinical practice, an equal ST-segment elevation in leads II, III, and aVF was often present in patients with acute inferior STEMI caused by occlusion of LCx, whereas ST-segment depression in lead I equal to half of that in lead aVL was observed in patients with occlusion of RCA. Therefore, the utility of the above 2 new electrocardiographic criteria unreported was also tested in the derivation population.

The performance of the new algorithm formed by the most powerful criteria was tested in the validation population.

Coronary angiography

The culprit coronary lesion was defined as the most severe lesion and/or the lesion with local dissection or thrombus. The films were interpreted by 2 experienced investigators without knowledge of the ECG findings. In case of a discrepancy, a third investigator reviewed the coronary angiography. Patients were classified into 2 groups accord- ing to the site of the culprit coronary lesion documented by coronary an- giography as follows: group RCA and group LCx.

Statistics

SPSS version 19.0 (SPSS, Chicago, IL) was used for the statistical anal- ysis. Data were expressed as mean +- SD for continuous variables and as number and percentage for categorical variables. For comparison of continuous variables, the analysis of variance was used. For comparison of categorical variables, the ?² test or the Fisher exact test was used. P b .05 was considered to be statistically significant.

Results

Baseline characteristics of the derivation cohort

The derivation population consisted of 194 patients (171 men and 23 women) with an age of 59 +- 11 years. On coronary angiography, the culprit artery was shown to be the RCA in 166 patients and the LCx in 28 patients. RCA was a dominant artery in 120 (62%) patients, dominant LCx in 12 (6%), and co-dominant circulation in 62 (32%) pa- tients. There was a complete agreement in coronary angiographic read- ings between the 2 investigators. There were no significant differences between the RCA group and the LCx group concerning the baseline de- mographic and clinical characteristics (Table 1).

Electrocardiographic characteristics

The relationship between ECG criteria and culprit artery is presented in Tables 2 and 3. There was a complete agreement in ECG readings be- tween the 2 investigators. The sensitivity, specificity, and positive and negative predictive values (PPV and NPV) of the traditional ECG criteria and 2 new ones to differentiate RCA occlusion from LCx occlusion in

Table 1

Baseline characteristics of the derivation cohort

	LCx group (n = 28)	RCA group (n = 166)	P
Female	5 (17.9%)	18 (10.8%)	.254
Age (y)	57.3 +- 11.7	58.9 +- 11.4	.908
Time from symptom onset to ECG (h)	6.5 +- 0.5	6.1 +- 0.2	.289
Time from onset of symptoms to ECG <= 6 h	20 (71.4%)	133 (80.1%)	.547
Time from onset of symptoms to CAG	7.0 +- 0.3	6.8 +- 0.4	.398
HR on admission (beat per min)	72.5 +- 14.1	70.7 +- 15.8	.977
SBP on admission (mm Hg)	124 +- 20.8	118 +- 24.2	.648
DBP on admission (mm Hg)	77.9 +- 14.1	74.3 +- 14.7	.790
Diabetes mellitus	5 (17.9%)	46 (27.7%)	.135
Hypertension	11 (39.3%)	83 (50.0%)	.111
Ischemic heart disease	4 (14.3%)	14 (8.4%)	.114
Family history	6 (21.4%)	36 (21.7%)	.959
Smoking	18 (64.3%)	115 (69.3%)	.123
Hyperlipidemia	2 (7.1%)	12 (7.2%)	.761
LVEF (%)	54.4 +- 8.2	57.7 +- 10.5	.826
Peak CK level (U/L)	2492.7 +- 1379.1	2805.8 +- 1904.1	.132
Peak CK-MB level (U/L)	237.5 +- 148.4	253.3 +- 168.1	.319
Killip class N 1	5 (17.9%)	45 (27.1%)	.798
Multivessel disease	15 (53.6%)	78 (47.0%)	.996
In-hospital death	0	2 (1.2%)	.999

All continuous variables are presented as mean +- SD. All categorical variables are pre- sented as number and percentage. CAG, coronary angiography; HR, heart rate; SBP, systol- ic blood pressure; DBP, diastolic blood pressure; LVEF, left ventricular ejection fraction; CK, creatine kinase; CK-MB, creatine kinase-MB.

Table 2

Relationship between traditional ECG criteria and culprit artery

Table 4

Sensitivity, specificity, and PPV, and NPV of the traditional ECG criteria for predicting the culprit artery

Traditional ECG criteria LCx group RCA group P

	(n = 28)	(n = 166)			Traditional ECG criteria for RCA	Sensitivity	Specificity	PPV	NPV
1. Lead I					1. STD I >= 0.5 mm, isoelectric	100%	29%	89%	100%
STD >= 0.5 mm	2 (7.1%)	140 (84.3%)	b.001		2. Ratio of STE in lead III vs lead II >= 1	100%	64%	94%	100%
ST isoelectric	18 (64.3%)	26 (15.7%)	b.001		3.?STD in leads V1-V3 vs ?STE in leads	60%	89%	97%	28%
STE >= 0.5 mm 2. Ratio of STE in lead III vs lead	8 (28.6%) II	0	b.001		II, III, and aVF <= 1 4. STE in lead V1	57%	96%	99%	28%
N 1	1 (3.6%)	160 (96.4%)	b.001		5. STE in lead aVR, isoelectric, STD b1 mm	92%	43%	91%	48%
1	9 (32.1%)	6 (3.6%)	b.001		6. Ratio STD V3/STE III <= 1.2	42%	86%	95%	20%
b1	18 (64.3%)	0	b.001		7. Ratio STD in lead I/STD in lead aVL <= 1	87%	93%	99%	54%
3. ?STD in leads V1-V3 vs ?STE in leads II, III, and aVF			8. Maximum STD in leads V1-V3			42%	89%	96%	21%

N 1	25(89.3%)	66 (39.8%)	b.001
<= 1	3 (10.7%)	100 (60.2%)	b.001

Lead V1

Traditional ECG criteria for LCx
1. STE I >= 0.5 mm	29%	100%	100%	89%
2. Ratio of STE in lead III vs lead II b 1	64%	100%	100%	94%
3.?STD in leads V1-V3 vs ?STE in leads	89%	60%	28%	97%

STD	3 (10.7%)	69 (41.6%)	b.001
ST isoelectric	24 (85.7%)	2 (1.2%)	b.001
STE	1 (3.6%)	95 (57.2%)	b.001
Lead aVR STD >= 1 mm	12 (42.9%)	13 (7.8%)	b.001
STE, ST isoelectric, STD b 1 mm	16 (57.1%)	153 (92.2%)	.03

Lead aVL

STD	23 (82.1%)	124 (74.7%)	.77
ST isoelectric	0	42 (25.3%)	b.001
STE	5 (17.9%)	0	b.001

Ratio STD in lead V3/STE in lead III

II, III, and aVF N 1

4. STD in lead V1, isoelectric	96%	57%	28%	99%
5. STD in lead aVR >= 1 mm	43%	92%	48%	91%
6. Ratio STD V3/STE III N 1.2	86%	42%	20%	95%
7. Ratio STD in lead I/STD in lead aVL N 1	93%	87%	54%	99%
8. STD in lead aVR >= aVL	14%	99%	80%	98%
9. STE in lead V5 or V6 >= 1 mm	82%	97%	82%	97%

<= 1.2	4 (14.3%)	69 (41.6%)	.02	4. Discussion
N 1.2	24 (85.7%)	97 (58.4%)	.03

Ratio STD in lead I/STD in lead aVL

<= 1 2 (7.1%) 144 (86.7%) b.001

N 1 26 (92.9%) 22 (13.3%) b.001

STD in lead aVR >= aVL 4 (14.3%) 1 (0.6%) b.001

STE in lead V5 or V6 >= 1 mm 23 (82.1%) 5 (3.0%) b.001

Maximum STD in leads V1-V3 3 (10.7%) 69 (41.6%) b.001

STD, ST-segment depression; STE, ST-segment elevation.

inferior acute STEMI are shown in Tables 4 and 5, respectively. Among the traditional and new ECG criteria, the most powerful criteria included

(1) the ratio of ST elevation in lead III to that of lead II, (2) the ratio ST depression in lead I to that of lead aVL, and (3) ST changes in lead I. The algorithm shown in Fig. 1 correlated the best with angiographic findings with the highest accuracy to predict a culprit artery. With this algorithm, we were able to identify the culprit artery in nearly 99% (192/194) of patients. The other 2 patients had occlusion of a proximal, dominant LCx. Figs. 2 and 3 present examples of electrocardiographic changes found in patients with occlusion in the RCA or LCx that corresponded to the sequential algorithm.

3.3. New algorithm in the validation cohort

Demographic and clinical characteristics of the derivation and vali- dation cohorts were similar (Table 6). In validation cohort, the new al- gorithm located the culprit artery correctly in 78 of 80 patients (97.5%) with a sensitivity and specificity of 100% and 89%, respectively, for predicting the RCA and 89% and 100%, respectively, for predicting the LCx (Table 7). The other 2 patients had occlusion of a proximal, dominant LCx.

The major findings of the present study were that (1) among 13 ECG criteria to differentiate RCA occlusion vs LCx occlusion in acute inferior STEMI, the most powerful ones were the ratio of ST elevation in lead III to that of lead II, the ratio ST depression in lead I to that in lead aVL, and ST changes in lead I and (2) a new 3-step algorithm was proposed to discriminate the culprit artery at the bedside of acute inferior STEMI patients.

Patients with inferior myocardial infarction are a heterogeneous group depending on the location of the culprit lesion. The inferior left ventricular wall may be supplied by distal branches from the RCA or by obtuse marginal branches of the LCx; and thus, the culprit vessel in inferior myocardial infarction may be either the RCA or the LCx. Rarely, acute inferior myocardial infarction may result from occlusion of the LAD branch [28,29], which is the terminal portion of a “wraparound” LAD, but this was not the case in any of our patients. Among the 274 pa- tients with acute inferior STEMI in the current study, the culprit artery was RCA in 228 patients and LCx in 46 patients, a ratio of 5:1, consistent with previous studies of 2.2:1 to 7.0:1 [12,13,16,26,30]. Inferior myocar- dial infarction with total occlusion of the proximal RCA often has hemo- dynamic complications or arrhythmias [3,4], where there is a chance for Right ventricular infarction which makes nitroglycerin relatively contra- indicated. Although coronary angiography is the best means to recog- nize the culprit artery in acute inferior STEMI, deciding which is the culprit can be difficult when both the RCA and LCx are severely diseased. In such circumstances, early recognition of whether the culprit artery is the RCA or LCx by ECG on admission may facilitate management and allow particular complications to be avoided [5-10,31].

Table 5

Sensitivity, specificity, PPV, and NPV of the new ECG criteria for predicting the culprit artery

Table 3

Relationship between new ECG criteria and culprit artery

Sensitivity Specificity PPV NPV

New ECG criteria for RCA

1. STD I = 1/2 STD aVL 41% 96% 99% 22%

New ECG criteria	LCx group (n = 28)	RCA group (n = 166)	P	2. STE II = III = aVF New ECG criteria for LCx	0	68%	0	10%
1. STD in lead I = 1/2 STD in lead aVL	1 (3.6%)	68 (41.0%)	b.001	1. STD I = 1/2 STD aVL	4%	59%	2%	78%
2. STE in lead II = III = aVF	9 (32.1%)	0	b.001	2. STE II = III = aVF	32%	100%	100%	90%

Fig. 1. Sequential electrocardiographic algorithm to predict the culprit artery in acute inferior ST elevation myocardial infarction. STD, ST-segment depression; STE, ST-segment elevation.

Inferior wall leads

In most patients, the myocardial distribution of the RCA is slightly rightward in the frontal plane, and consequently, the current of injury resulting from its occlusion will be reflected more in lead III than in lead II. Conversely, the distribution of the LCx is slightly leftward in

the frontal plane, and the current of injury from its closure will be seen more in lead II than in lead III. Greater ST elevation in lead III than in II has been shown to indicate RCA-related inferior STEMI [13,26]. Our study also confirmed the validation of ST-segment eleva- tion in lead II and III to differentiate RCA occlusion from LCx occlusion in acute inferior STEMI. Higher ST-segment elevation in lead II than

Fig. 2. Electrocardiographic tracing of a patient with acute inferior ST elevation myocardial infarction with occlusion of the RCA. Higher ST elevation in lead III than in lead II and deeper ST depression in lead aVL than in lead I identified the RCA as the culprit artery.

Fig. 3. Electrocardiographic tracing of a patient with acute inferior ST elevation myocardial infarction with occlusion of the LCx. Higher ST elevation in ST elevation in lead II than in lead III identified the LCx as the culprit artery.

lead III predicted the occlusion of LCx with 64% sensitivity and 100% specificity. The new criterion of equal ST segment elevation in leads II, III, and aVF is specific for LCx occlusion but with low sensitivity.

Lateral limb leads

Additional limb lead criteria involve careful analysis of I and aVL. An injury vector leftward enough to cause ST-segment elevation in lead I is common with LCx occlusion but rare with RCA occlusion. Patients with LCx-related inferior infarction less frequently show reciprocal ST de- pression in lead aVL and more often show an isoelectric or a raised ST- segment in leads I and aVL compared with patients with RCA-related in- ferior infarction [16,17]. Hasdai et al [15] reported that absence of recip- rocal ST depression in lead aVL indicated injury of the anterosuperior base of the heart typically caused by LCx occlusion proximal to the first obtuse marginal branch. The aVL lead faces the high-lateral seg- ment of the left ventricle and is the only lead truly reciprocal to the in- ferior wall. Inferior myocardial infarction caused by RCA occlusion has greater ST-segment depression in lead aVL than in lead I [16]. We also tested the utility of ST-segment variation in lead I and aVL to detect cul- prit coronary lesion location in acute inferior STEMI. Our study revealed that ST elevation in lead I >= 0.5 mm was a specific criterion for diagnos- ing LCx as the culprit vessel with 100% specificity, consistent with previ- ous studies [20]. The conventional RCA-related criterion of ST depression in lead aVL did not work in our population. However, the

Table 6

Baseline characteristics of the validation cohort

criterion of the ratio ST depression in lead I to that in lead aVL <= 1 pos- sessed high specificity and sensitivity to recognize the RCA occlusion. The new criterion of ST segment depression in lead I equal to half of that in lead aVL was not useful.

Lead aVR

The current of injury with RCA occlusion is more or less perpendicu- lar to the axis of lead aVR, whereas the current of injury resulting from occlusion of the LCx has a mean vector that forms an obtuse angle with the axis of aVR. Therefore, significant ST-segment depression in aVR is more likely to occur with LCx occlusion. In a recent study in a small group of patients, the presence and amount of ST-segment depression in lead aVR were found to predict LCx involvement during acute inferior myocardial infarction [20]. In our study, ST depression in aVR >= 1 mm was more common in patients with LCx-related acute inferior myocar- dial infarction (42.9% vs 7.8%), albeit with low sensitivity and specificity. Sun et al [32] reported that ST changes in lead aVR were associated with an excellent specificity (94%) and a good sensitivity (70%) in differenti- ating LCx from RCA. The disparity may be due to different population and ECG recording time.

ST elevation, ST depression b 1 mm, or isoelectric ST in lead aVR as a marker of RCA occlusion was present in 153 patients (92.2%) of the RCA group in the current study. Nair and Glancy [20] reported a sensitivity of 96% and a specificity of 80% for this criterion to recognize RCA occlusion in acute inferior myocardial infarction. In the current study population, 16 (57.1%) patients of the LCx group had ST elevation or isoelectric ST in lead aVR. Owing to its low sensitivity and specificity, this criterion was a

Validation cohort (n = 80)

Derivation cohort P

(n = 194)

poor predictor to differentiate RCA occlusion and LCx occlusion. Baptista et al [33] made similar deductions. The disagreement between these re-

fied in a larger study.

Female	9 (11.3%)	23 (11.9%)	.254	sults should be clari
Age (y)	57.6 +- 12.0	58.7 +- 11.4	.908
Time from symptom onset to ECG (h)	6.8 +- 0.4	6.2 +- 0.2	.589	4.4. Precordial leads
Time from onset of symptoms to ECG <= 6 h	54 (67.5%)	153 (78.9%)	.247

Some studies have reported that anterior precordial ST-segment de-

Time from onset of symptoms to CAG	7.1 +- 0.4	6.8 +- 0.4	.288
HR on admission (beat per min)	73.7 +- 15.8	71.0 +- 15.6	.778

SBP on admission (mm Hg)	122 +- 23.2	119.4 +- 23.7	.708	pression was significantly more likely to occur when the LCx was the
DBP on admission (mm Hg)	76.6 +- 15.2	74.8 +- 14.6	.812	culprit artery rather than the right, and increasing degrees of coronary
Diabetes mellitus	22 (27.5%)	51 (26.3%)	.735	disease did not alter this relation [17]. Whereas right ventricular infarc-

tion tends to elevate ST-segment in leads V1-V3. The lack of precordial ST-segment depression had a high NPV for excluding the LCx as the cul-

Hypertension	45 (56.3%)	94 (48.5%)	.311
Ischemic heart disease	6 (7.5%)	18 (9.3%)	.214
Family history	10 (12.5%)	42 (21.6%)	.115

Smoking	50 (62.5%)	133 (68.6%)	.623	prit artery, although specificity was somewhat limited [17]. Others have
Hyperlipidemia	7 (8.8%)	14 (7.2%)	.861	found that a discordant pattern, ST-segment elevation in lead V1 and

ST-segment depression in lead V2, is an important electrocardiographic

LVEF (%)	54.4 +- 11.5	57.2 +- 10.2	.753
Peak CK level (U/L)	2500.8 +- 1765.1	2760.0 +- 1828.3	.321
Peak CK-MB level (U/L)	240.5 +- 160.2	247.5 +- 165.3	.519
Killip class N 1	18 (22.5%)	50 (25.8%)	.898

Multivessel disease	34 (42.5%)	93 (47.9%)	.996	Table 7
Dominant RCA	56 (70.0%)	120 (61.9%)	.061	Sensitivity, specificity, PPV, and NPV of the new ECG algorithm for predicting the culprit
Dominant LCx	5 (6.3%)	12 (6.2%)	.978	artery in the derivation and validation cohort

Co-dominant circulation Culprit RCA	19 (23.8%) 62 (77.5%)	62 (32.0%) 166 (85.6%)	.082 .313		Sensitivity	Specificity	PPV	NPV
Culprit LCx	18 (22.5%)	28 (14.4%)	.215	New algorithm for RCA in derivation cohort	100%	93%	99%	100%
In-hospital death	1 (1.3%)	2 (1.0%)	.999	New algorithm for LCx in derivation cohort	93%	100%	100%	99%

All continuous variables are presented as mean +- SD. All categorical variables are pre- sented as number and percentage.

New algorithm for RCA in validation cohort	100%	89%	97%	100%
New algorithm for LCx in validation cohort	89%	100%	100%	97%

pattern for the diagnosis of right ventricular infarction [34,35]. Tsuka et al [23] found that ST segment in lead V1 may be elevated when the RCA is occluded because it is influenced by the current of injury from right ventricular infarction. ST-segment elevation in V1 on admission in patients with acute Q-wave Inferior wall myocardial infarction indi- cates an RCA lesion associated with a larger infarct size and a higher in- cidence of major in-hospital arrhythmias. ST-segment depression in both V1 and V2 is the reciprocal change resulting from ST-segment ele- vation in the posterior wall of the left ventricle and is typical of LCx oc- clusion [30]. In our study, a sensitivity of 57%, specificity of 96%, and positive predictive value (PPV) of 99% were obtained for ST elevation in lead V1, making it a significant criterion to predict RCA as the occlud- ed artery.

When ST elevation in lead V5 or lead V6 was applied to the LCx group to assess their diagnostic accuracy, 13 (46.4%) patients had ST el- evation in lead V5 that were all accompanied with ST elevation in lead V6. And a total of 23 (82.1%) patients presented with ST elevation in lead V6. A sensitivity of 82%, specificity of 97%, and PPV of 82% were ob- tained for ST elevation in lead V5 or V6. In the current study, ST eleva- tion in lead V6 was a more accurate criterion than ST elevation in lead V5. Wong et al [24] reported a sensitivity of 59%, specificity 76%, and PPV of 91% for the electrocardiographic criterion ST elevation in lead V5 and a sensitivity of 67%, specificity 75%, and PPV of 32% for the ECG criterion ST elevation in lead V6.

Additional leads

ST-segment elevation in the right precordial chest leads (V3R, V4R, and V5R) especially V4R has been shown to be a reliable marker of right ventricular involvement in inferior acute myocardial infarction and thus RCA involvement [36]. Nevertheless, it is usually an early and transient phenomenon, and sometimes, it disappeared within 2 hours after the onset of chest pain [37]. Patients with left dominant coronary circulation will sustain a right ventricular infarct with occlusion of LCx, and although less common, occlusion of the LAD may result in infarction of the anterior wall of right ventricle [38]. Sometimes, there are no changes in leads V3R to V5R due to the presence of concomitant lateral or posterior involvement [18,39]. Posterior injury (95%) was detected mostly during LCx occlusion. The additional posterior leads (V7, V8, and V9) have been found to possess high specificity of 99% to detect LCx occlusion with low sensitivity of 49% [40]. On the other hand, inferoposterior wall myocardial infarctions involve either RCA or LCx. Most people are RCA dominant and have the posterior descending ar- tery as a branch of the RCA. Therefore, some patients are left coronary artery dominant and have the posterior descending artery branch off the LCx artery to supply the posterior aspect of the left ventricle. Anoth- er important disadvantage of a diagnosis based on these leads is that they are often not recorded in the initial evaluation of patients with chest discomfort. Therefore, apart from using right-sided precordial leads and posterior leads, attempts to recognize the culprit artery on a standard 12-lead ECG have been made.

Study limitations

The number of patients was relatively small and from a single center. Larger studies from multicenters can provide further information on the diagnostic value of the Standard ECG in patients with inferior myocardi- al infarction.

5. Conclusion

The present study confirmed the utilities of (1) the ratio of ST elevation in lead III to that in lead II, (2) the ratio of ST depression in lead I to that in lead aVL, and (3) ST changes in lead I to locate the culprit artery in inferior STEMI. A new 3-step algorithm based on 12-lead ECG is proposed to localize the culprit artery at the bedside of acute inferior

STEMI patients before Primary PCI, allowing immediate decisions about therapy.

References

1. Bairey CN, Shah PK, Lew AS, Hulse S. Electrocardiographic differentiation of occlu- sion of the left circumflex versus the right coronary artery as a cause of inferior acute myocardial infarction. Am J Cardiol 1987;60:456-9.
2. Bates ER. Reperfusion therapy in inferior myocardial infarction. J Am Coll Cardiol 1988;12:44A-51A.
3. Berger PB, Ryan TJ. Inferior myocardial infarction. High-risk subgroups. Circulation 1990;81:401-11.
4. Gacioch GM, Topol EJ. Sudden paradoxic clinical deterioration during angioplasty of the occluded right coronary artery in acute myocardial infarction. J Am Coll Cardiol 1989;14:1202-9.
5. Bates ER, Clemmensen PM, Califf RM, Gorman LE, Aronson LG, George BS, et al. Pre- cordial ST segment depression predicts a worse prognosis in inferior infarction de- spite reperfusion therapy. The Thrombolysis and Angioplasty in Myocardial Infarction (TAMI) study group. J Am Coll Cardiol 1990;16:1538-44.
6. Edmunds JJ, Gibbons RJ, Bresnahan JF, Clements IP. Significance of anterior ST de- pression in inferior wall acute myocardial infarction. Am J Cardiol 1994;73:143-8.
7. Peterson ED, Hathaway WR, Zabel KM, Pieper KS, Granger CB, Wagner GS, et al. Prognostic significance of precordial ST segment depression during inferior myocar- dial infarction in the thrombolytic era: results in 16, 521 patients. J Am Coll Cardiol 1996;28:305-12.
8. Shah PK, Pichler M, Berman DS, Maddahi J, Peter T, Singh BN, et al. Noninvasive iden- tification of a high risk subset of patients with acute inferior myocardial infarction. Am J Cardiol 1980;46:915-21.
9. Sugiura T, Iwasaka T, Takehana K, Nagahama Y, Hasegawa T, Inada M. Precordial ST segment depression in patients with Q wave inferior myocardial infarction: role of infarction-associated pericarditis. Am Heart J 1993;125:672-5.
10. Zehender M, Kasper W, Kauder E, Schonthaler M, Geibel A, Olschewski M, et al. Right ventricular infarction as an independent predictor of prognosis after acute inferior myocardial infarction. N Engl J Med 1993;328:981-8.
11. Assali AR, Herz I, Vaturi M, Adler Y, Solodky A, Birnbaum Y, et al. Electrocardiograph- ic criteria for predicting the culprit artery in inferior wall acute myocardial infarc- tion. Am J Cardiol 1999;84:87-9 [A88].
12. Bayram E, Atalay C. Identification of the culprit artery involved in inferior wall acute myocardial infarction using electrocardiographic criteria. J Int Med Res 2004;32: 39-44.
13. Chia BL, Yip JW, Tan HC, Lim YT. Usefulness of ST elevation II/III ratio and ST devia- tion in lead I for identifying the culprit artery in inferior wall acute myocardial in- farction. Am J Cardiol 2000;86:341-3.
14. Fiol M, Cygankiewicz I, Carrillo A, Bayes-Genis A, Santoyo O, Gomez A, et al. Value of electrocardiographic algorithm based on “ups and downs” of ST in assessment of a culprit artery in evolving inferior wall acute myocardial infarction. Am J Cardiol 2004;94:709-14.
15. Hasdai D, Birnbaum Y, Herz I, Sclarovsky S, Mazur A, Solodky A. ST segment depres- sion in lateral limb leads in inferior wall acute myocardial infarction. Implications re- garding the culprit artery and the site of obstruction. Eur Heart J 1995;16:1549-53.
16. Herz I, Assali AR, Adler Y, Solodky A, Sclarovsky S. New electrocardiographic criteria for predicting either the right or left circumflex artery as the culprit coronary artery in inferior wall acute myocardial infarction. Am J Cardiol 1997;80:1343-5.
17. Kontos MC, Desai PV, Jesse RL, Ornato JP. Usefulness of the admission electrocardio- gram for identifying the infarct-related artery in inferior wall acute myocardial in- farction. Am J Cardiol 1997;79:182-4.
18. Kosuge M, Ishikawa T, Morita S, Ebina T, Hibi K, Maejima N, et al. Posterior wall in- volvement attenuates predictive value of ST-segment elevation in lead V4R for right ventricular involvement in inferior acute myocardial infarction. J Cardiol 2009;54: 386-93.
19. Kosuge M, Kimura K, Ishikawa T, Hongo Y, Mochida Y, Sugiyama M, et al. New elec- trocardiographic criteria for predicting the site of coronary artery occlusion in infe- rior wall acute myocardial infarction. Am J Cardiol 1998;82:1318-22.
20. Nair R, Glancy DL. ECG discrimination between right and left circumflex coronary ar- terial occlusion in patients with acute inferior myocardial infarction: value of old criteria and use of lead aVR. Chest 2002;122:134-9.
21. Senaratne MP, Weerasinghe C, Smith G, Mooney D. Clinical utility of ST-segment de- pression in lead AVR in acute myocardial infarction. J Electrocardiol 2003;36:11-6.
22. Tierala I, Nikus KC, Sclarovsky S, Syvanne M, Eskola M, Group HS. Predicting the cul- prit artery in acute ST-elevation myocardial infarction and introducing a new algo- rithm to predict infarct-related artery in inferior ST-elevation myocardial infarction: correlation with coronary anatomy in the HAAMU trial. J Electrocardiol 2009;42:120-7.
23. Tsuka Y, Sugiura T, Hatada K, Nakamura S, Yuasa F, Iwasaka T. Clinical significance of ST-segment elevation in lead V1 in patients with acute inferior wall Q-wave myocar- dial infarction. Am Heart J 2001;141:615-20.
24. Wong TW, Huang XH, Liu W, Ng K, Ng KS. New electrocardiographic criteria for identifying the culprit artery in inferior wall acute myocardial infarction- usefulness of T-wave amplitude ratio in leads II/III and T-wave polarity in the right V5 lead. Am J Cardiol 2004;94:1168-71.
25. Zhan ZQ, Wang W, Dang SY, Wang CQ, Wang JF, Cao Z. Electrocardiographic charac- teristics in angiographically documented occlusion of the dominant left circumflex artery with acute inferior myocardial infarction: limitations of ST elevation III/II ratio and ST deviation in lateral limb leads. J Electrocardiol 2009;42:432-9.
26. Zimetbaum PJ, Krishnan S, Gold A, Carrozza II JP, Josephson ME. Usefulness of ST- segment elevation in lead III exceeding that of lead II for identifying the location of the totally occluded coronary artery in inferior wall myocardial infarction. Am J Cardiol 1998;81:918-9.
27. Saw J, Davies C, Fung A, Spinelli JJ, Jue J. Value of ST elevation in lead III greater than lead II in inferior wall acute myocardial infarction for predicting in-hospital mortal- ity and diagnosing right ventricular infarction. Am J Cardiol 2001;87:448-50 [A446].
28. Birnbaum Y, Wagner GS, Barbash GI, Gates K, Criger DA, Sclarovsky S, et al. Correlation of angiographic findings and right (V1 to V3) versus left (V4 to V6) precordial ST-segment depression in inferior wall acute myocardial infarction. Am J Cardiol 1999;83:143-8.
29. Correale E, Battista R, Martone A, Pietropaolo F, Ricciardiello V, DiGirolamo D, et al. Electrocardiographic patterns in acute inferior myocardial infarction with and with- out right ventricle involvement: classification, Diagnostic and prognostic value, masking effect. Clin Cardiol 1999;22:37-44.
30. Mak KH, Chia BL, Tan AT, Johan A. Simultaneous ST-segment elevation in lead V1 and depression in lead V2. A discordant ECG pattern indicating right ventricular in- farction. J Electrocardiol 1994;27:203-7.
31. Mukharji J, Murray S, Lewis SE, Croft CH, Corbett JR, Willerson JT, et al. Is anterior ST depression with acute transmural inferior infarction due to posterior infarction? A vectorcardiographic and scintigraphic study. J Am Coll Cardiol 1984;4:28-34.
32. Sun TW, Wang LX, Zhang YZ. The value of ECG lead aVR in the differential diagnosis of acute inferior wall myocardial infarction. Intern Med 2007;46:795-9.
33. Baptista SB, Farto e Abreu P, Loureiro JR, Thomas B, Nedio M, Gago S, et al. Electro- cardiographic identification of the infarct-related artery in acute inferior myocardial infarction. Rev Port Cardiol 2004;23:963-71.
34. Geft IL, Shah PK, Rodriguez L, Hulse S, Maddahi J, Berman DS, et al. ST elevations in leads V1 to V5 may be caused by right coronary artery occlusion and acute right ven- tricular infarction. Am J Cardiol 1984;53:991-6.
35. Morgera T, Alberti E, Silvestri F, Pandullo C, Della Mea MT, Camerini F. Right precor- dial ST and QRS changes in the diagnosis of right ventricular infarction. Am Heart J 1984;108:13-8.
36. Wellens HJ. The value of the Right precordial leads of the electrocardiogram. N Engl J Med 1999;340:381-3.
37. Braat SH, Brugada P, den Dulk K, van Ommen V, Wellens HJ. Value of lead V4R for recognition of the infarct coronary artery in acute inferior myocardial infarction. Am J Cardiol 1984;53:1538-41.
38. Kinch JW, Ryan TJ. Right ventricular infarction. N Engl J Med 1994;330:1211-7.
39. Kosuge M, Kimura K, Ishikawa T, Hongo Y, Shigemasa T, Sugiyama M, et al. Implica- tions of the absence of ST-segment elevation in lead V4R in patients who have infe- rior wall acute myocardial infarction with right ventricular involvement. Clin Cardiol 2001;24:225-30.
40. Khaw K, Moreyra AE, Tannenbaum AK, Hosler MN, Brewer TJ, Agarwal JB. Improved detection of posterior myocardial wall ischemia with the 15-lead electrocardiogram. Am Heart J 1999;138:934-40.