The prognostic effect of ST-elevation in lead aVR on coronary artery disease, and outcome in acute coronary syndrome patients: a systematic review and meta-analysis

Background Rapid diagnosis of coronary artery disease has an important role in saving patients. The aim of this study is to evaluate if aVR lead ST-elevation (STE) can predict LM/3VD, left main (LM) disease, and three-vessel disease (3VD), outcome in acute coronary syndrome (ACS) patients. Methods In this systematic review and meta-analysis, 45 qualified studies were entered. Scopus, Pub med, Google scholar, Web of science, Cochrane library were searched on 12 November 2021. Results This systematic review includes 52,175 participants. In patients with STE, the total odds ratios for LM, 3VD, and LM/3VD were 5.48 (95% CI 3.88, 7.76), 2.21 (95% CI 1.78, 3.27), and 6.21 (95% CI 3.49, 11,6), respectively. STE in lead aVR was linked with in-hospital death (OR = 2.99, CI 1.90, 4.72) and 90-day mortality (OR = 3.09, CI 2.17, 4.39), despite the fact that it could not predict 30-day mortality (OR = 1.11, CI 0.95, 1.31). The STE > 1 mm subgroup had the highest sensitivity for LM (0.9, 95% CI 0.82, 0.98), whereas the STE > 0.5 mm (0.76, 95% CI 0.61, 0.90) subgroup had the highest sensitivity for LM/3VD. The appropriate cut-off point with highest specificity for LM/3VD and LM was STE > 1.5 mm (0.80, 95% CI 0.75, 0.85) and STE > 0.5 mm, respectively (0.75, 95% CI 0.67, 0.84, I2 = 97%). Conclusion The odds of LM and LM/3VD were higher than 3VD in ACS patients with STE in lead aVR. Also, STE > 0.5 mm was the best cut-off point to screen LM/3VD, whereas for LM diagnosis, STE > 1 mm had the highest sensitivity. Furthermore, LM/3VD had a higher overall specificity than LM. Supplementary Information The online version contains supplementary material available at 10.1186/s40001-022-00931-5.


Introduction
One of the main reasons of death in the worldwide adult population is ischemic heart disease (IHD) that imposes a significant financial burden on the health care system [1,2]. Almost, 40% of patients with IHD present with acute coronary syndrome (ACS), which includes acute non-ST elevation myocardial infarction (NSTEMI), unstable angina, and ST-elevation myocardial infarction (STEMI). ACS will rise exponentially in the coming years as result of the rising prevalence of some risk factors such as diabetes, obesity as well as increasing the life expectancy of the worldwide population [3]. A significant proportion of ACS patients have left anterior coronary artery (LAD), left main coronary artery stenosis (LMCA), or both of them [4]. Obstruction in these arteries critically decrease coronary flow, which impair left ventricular function, leading to adverse outcomes and intraoperative complications, so early detection of these lesions is critical. Patients with ACS (LM/3VD) are at high risk for short-term and long-term cardiovascular side effects [5]. Despite remarkable progress in medical diagnosis, the electrocardiogram (ECG) is still the primary diagnostic tool in ACS patients. Valuable information is accrued of ECG in order to early detection of damaged coronary artery area, myocardial ischemia, lesion identification, and extent of infarction. Additionally, ECG can help predict possible complications [6].
Lead aVR has been long forgotten until recent years unlike the other 11 leads. Although recent examinations have stated that ST-segment change analysis in lead aVR provides helpful data on the coronary angiographic anatomy and risk classification in ACS [7]. Previous studies have shown that an increase in the ST-segment in the aVR lead (STE-aVR) might be helpful in order to diagnose left main disease or three-vessel disease (3VD) [4,5,8], despite some limitations including the selection bias, the retrospective nature of the studies, and the small sample size. Therefore, the aVR lead changes are not included in the American Heart Association's latest scientific statement on ECG interpretation [10]. The aim of this systematic review and meta-analysis was to investigate the diagnostic role of STE-aVR in ACS patients and its association with LM disease and 3VD.

Selection process and eligibility criteria
All articles were divided into three groups. Then, three authors (by E.K, A.M, and M.B) screened the article base on title, abstract, and keywords independently. Studies fulfilling the entire inclusion criteria entered in the study. Besides, there was no limitation in terms of language of article. Eligible criteria: (1) cohort, cross-sectional, and case-control studies were enrolled; (2) the enrolled studies were adopted from articles with acute coronary syndrome (ACS) study population; (3) studies reported odds ratio (OR) or sensitivity /specificity to predict LM or 3VD or LM/3VD or death base on the size of aVR ST-elevation. Some studies did not report any OR, although they had essential data for calculation of OR. Consequently, they were included in the study.
Extraction process and quality assessment E.K, A.M, and M.B. did extraction process and quality assessment of article independently. Checklist used to assess the quality of studies was Appendix 2: MINORS Criteria. Non-comparative studies and comparative studies have 8 and 12 criteria, respectively. The items were scored in this way: (1) not reported = 0, (2) report but inadequate = 1, (3) completely reported = 2. The total ideal score was 16 for non-comparative studies and 24 for comparative studies [10].

Analysis
OR was used as a common correlation index to assess the strength of the relationship. Forest plots were drawn to examine the modified ORs along with their confidence intervals. The meta-analyses were performed using the fixed-effects or random-effects method to estimate the summary OR and 95% of CI using the inverse-variance weights and the DerSimonian-Laird estimator, respectively. The heterogeneity was evaluated by Q-Cochran test at the significance level of 0.05 and index I 2 among studies. For I 2 ≥ 50% and P ≤ 0.05, heterogeneity was considered statistically significant. Meta-regression and subgroup analysis were performed to identify the source of heterogeneity. Subgroup analysis was done based on different sizes of STE in lead aVR and time of death. Publication bias was assessed the publication bias. In the funnel plot, ORs were plotted against the inverse of the square of the standard error. All analyses were done using STATA 14.0 software. All P values were two-tailed. Also, the significant level of p value was less than 0.05.

Study selection and characteristics
Six-hundred fourteen related studies were extracted initially. Duplicate articles (n = 318) and studies that could not fulfill the study inclusion criteria (n = 264) were excluded after title, abstract and full-text screening (Fig. 1). Finally, 45 qualified articles were entered in this study [5,. The total participants of the included studies were 52,175. All of the eligible studies were performed on both men and women. Table 1 summarizes the characteristics and scores of eligible researches.

Main analysis
Subgroup analysis was performed for LM, 3VD and LM/3VD based on the size of STE in lead aVR (Figs. 2, 3 and 4). For LM, the overall OR was 5.48 (95% CI 3.88, 7.76). STE > 0.5 mm groups had higher OR compared with STE > 1 mm and 0.5 < STE < 1 mm (Fig. 2) and the heterogenicity between the studies was significant (I 2 = 63.8%, p = 0.000). For 3VD, the overall OR was
To investigate the possible causes of heterogeneity, meta-regression analysis was performed based on sex, country, total sample size, mean age and publish year. There was not any significant source of heterogenicity except for total sample size in LM odds ratio (p = 0.011), LM/3VD odds ratio (p = 0.002) and LM npv     (p = 0.045). Additionally, publish year was also meaningful for LM/3VD odds ratio (p = 0.007).

Discussion
Fast diagnosis of cardiovascular disease plays a critical role in rescuing of patients, especially patients with coronary artery disease [28]. aVR lead that is often ignored in clinic, pose some key date [55].
Our finding demonstrated STE in lead aVR can predict LM and LM/3VD with higher odds ratio than 3VD (OR LM : 5.48, CI [3.88, 7.76], OR LM/EVD = 6.21, CI [3.49, 11.06], OR 3VD = 2.41, CI [1.78, 3.27]). Also, Misumida N et al. declared STE in lead aVR as an independent diagnosis factor of LM/3VD in non-STEMI (OR = 2.99, CI [1.79, 4.98]). In addition, Rathi N et al. from Pakistan represented that the number of LM/3VD patients in STE-aVR group was significantly higher than non-STE aVR group (51 (62.96%) vs 9 (29.03%), p < 0.0001) [41]. In a single cohort study from Mazandaran province, there was not any meaningful difference in both groups (STE-aVR vs non-STE aVR) in terms of LM/3VD. By the way, they pointed that STE in lead aVR related to severity of atherosclerosis, however their sample size was small [38]. Besides, another study with larger sample size (n = 472) could not found significant relationship between angiography result and STE in lead aVR [28]. Moreover, a systematic review and meta-analysis recently demonstrated that not only STE in AVR is related to LM but also the degree of elevation is effective, which is consistent with our results (OR STE> [5]. Moreover, another study with a large sample size (n = 15,315) reported that there was not any significant relationship between STE in lead aVR and 30-day mortality in adjusted model [49]. By contrast, a Spanish study pointed increase in the number of death related to STE in lead aVR significantly (p = 0.04). Also, another retrospective cross-sectional study represented that the chance of mortality in patients with STE upper than 1 mm was 7. 72 (CI [ 3.07, 19.42, P < 001) [18]. Besides, one study from New Zealand declared that mortality rate in inferior myocardial infarction was along with STE in lead aVR in adjusted model (hazard ratio = 5.87, CI [2.09-16.5]) [12]. In our study, the chance of in-hospital and 90-day mortality increased unlike 30 [30]. One study from Iran had similar results for detection of LM (sensitivity = 100, specificity = 33.5%) [24]. However, their cut-off point was 1 mm in order to consider STE in lead aVR, that was upper than prior study (cut-off = 0.5 mm). Besides, another study considered 0.5 mm elevation as a significant STE and had similar sensitivity (80%) and higher specificity (92.3%) [27]. Hussien A et al. declared that sensitivity and specificity were 77% and 65% when the cut-off point was considered 0.5 mm for detecting LM/3VD. Also, when they set higher cut-off point (> 1.5 mm), sensitivity and specificity reach to 14% and 98%, respectively. Likewise, cut-off point of 0.5 mm and 1.5 mm had highest NPV and PPV, respectively (78%, 91%) [29]. Kosuge M et al. followed the same pattern. Thus, 0.5 mm STE had highest sensitivity and NPV (91%, 99%, respectively) and 1.5 mm STE had highest specificity and PPV (98%, 58%, respectively) for diagnosis of LM/3VD. [31]. In this regard, the results of Misumida N et al. 's study were concordant with previous studies in this regard. A systematic review and meta-analysis showed overall sensitivity of LM and LM/3VD was 39% and 40%, respectively. Moreover, the overall specificity of LM and LM/3VD was 86% and 82%, respectively [57]. Our results represented overall sensitivity of LM and LM/3VD was 77% and 52%, respectively.
And also, the overall specificity of LM/3VD was higher than LM (89% Vs 71%). Furthermore, STE ≥ 0.5 mm and STE ≥ 1 mm had the highest sensitivity for LM/3VD and LM (sensitivity LM = 90%, sensitivity LM/3VD = 76%). Additionally, cut-off points of 1 mm STE in lead aVR had the highest NPV (94%) and PPV (53%) with regard to LM. However, cut-off points of 0.5 mm and 1 mm STE in lead aVR had the highest NPV (89%) and PPV (75%) in terms of LM/3VD, respectively.

Limitation
In this study, we were not able to access the full text of some studies that might change our result.

Conclusion
STE in lead aVR increases the risk of LM and LM/3VD more than 3VD. Furthermore, STE ≥ 0 0.5 mm and STE ≥ 1 mm were the best cut-off points to screen patients in terms of LM/3VD and LM, respectively. Additionally, the overall specificity of LM/3VD was greater than LM.