LBBB: The ECG Patterns and Cardiac Function in Patients With and Without Coronary Artery Disease

et al.) An inverse relationship between the QRS duration and left ventricular ejection fraction has been observed in LBBB patients in the presence of mechanical asynchrony. (Roshan, Pati et al. 2008) This finding provides the basis for cardiac resynchronization therapy (CRT) in heart failure patients. (Strauss, Selvester et al.) It is unclear, however, if such a relationship is more pronounced when underlying CAD exists.

The second part of this study, therefore, was to examine the correlation between QRSd and LV ejection fraction (EF) in patients with LBBB, and to determine if this relationship is altered by CAD.

Materials and methods
A retrospective study in accordance with compliance guidelines was conducted after obtaining the approval of the Institutional Review Board of the Lankenau Medical Center. Patients' identifiable information was removed for the data analysis.
The LBBB diagnosis was first made by the automated interpretation of the ECG system. Then it was over-read by a cardiologist to confirm the diagnosis. LBBB ECGs were retrieved from the ECG library and from patients who underwent cardiac catheterization between January 1 st 2005 and June 30 th 2010. Electronic medical records were reviewed to extract demographic and clinical information. A cardiac catheterization database containing patients who had undergone diagnostic coronary angiography, percutaneous coronary intervention (PCI) and/or coronary artery bypass grafting (CABG) was interrogated. Patients were classified as having CAD if they had history of PCI, CABG or if they had ≥50% luminal diameter stenosis of a major coronary artery on angiography.

ECG evaluation
To determine whether LBBB ECG morphology is different between CAD and non-CAD groups, the resting supine 12-lead ECG (AC filter 60Hz, 25 mm/sec, 10 mm/mV) was evaluated for six aspects: 1) Pathologic Q waves. 2) QRS duration (QRSd), amplitude and axis. 3) Among patients with leftward QRS axis  -45 degrees, QRS morphology was further evaluated for the presence of left anterior fascicular block (LAFB) pattern (Figure 1) ECG from a 69-year-old man with CAD and EF 31% demonstrating QRS-ST-T pattern typical to divisional LBBB: QRSd = 188 ms. The LAFB pattern is shown as QRS axis > -45 degree, rS in II, III and aVF with deflection of SIII > SII; and broad and notched R wave in I and qR in aVL with RaVL > RI; Poor R progression with rS in V6 and T wave opposite to the direction of QRS complex in precordial leads. Fig. 1. Poor R progression V1-6 in a CAD patient with LBBB www.intechopen.com 4) Low QRS voltage in the limb leads was considered if QRS amplitude was  0.5 mV in six limb leads. 5) Poor R wave progression was diagnosed if the S wave was the dominant deflection in the QRS complex, i.e. in rS waveform, from right to left precordial leads and if the ratio of R/S was < 1 in V1-V4 or beyond. 6) Presence of ischemic ST-T changes such as ST elevation accompanying T wave inversion, or ST depression with upright T wave in the same lead. For patients with injury or ischemic ST-T changes, short term follow-up ECGs were evaluated to determine whether these ST-T changes were resolved.

QRSd-EF relationship
Cardiac output was estimated by one of two methods: 1) 2-D echocardiography using modified Simpson's biplane method, or 2) post-stress, nuclear perfusion imaging the single photon emission computed tomography (gated SPECT). The QRSd was obtained from automated interpretation of each ECG tracing and QRSd-EF curve was plotted.

Statistical analysis
Comparisons of CAD and non-CAD groups were performed using NCSS 2007 biomedical statistical software (Kaysville, UT). A two-sample t test was applied for descriptive analysis in quantitative measures showing normal distribution of data points. Nonparametric tests Mann-Whitney U or Wilcoxon Rank Tests were used for difference in medians if the normality was rejected. A chi square (X2) statistic was used to compare categorical variables. For the ECG changes that were more prevalent in one group than the other, sensitivity, specificity, positive and negative predictive values were calculated for the assessment of predictive accuracy. Univariate linear regression analysis was performed when appropriate. A p value <0.05 was considered statistically significant.

Results
LBBB was found in 3.7% (134/3582) of all patients screened. Among LBBB patients, 72% (96/134) had evidence of CAD. Patients with CAD were older (76  9 vs. 65  13 years, p <0.001) and more commonly of male gender (53% vs. 43%, p < 0.05). Table 1 demonstrates that in patients with LBBB, leftward QRS axis, QS wave in V1, poor R wave progression, and broad/notched/slurred R wave in V6 are prevalent in both CAD and non-CAD groups. RS morphologies (rS, rs or Rs) in V6 was the second most common waveform (Figure 1) next to broad/notched/slurred R wave in this lead, seen in 40% of CAD and 26% of non-CAD patients. Though it is less frequent, low QRS voltage is more common in CAD patients. Ischemic ST-T changes were infrequently seen in the CAD group. Among all ECG variables assessed, none reached statistical significance except T wave inversion in V4, present in 13.5% in CAD and in none of the non-CAD patients (p < 0.05). Clinical assessment of those with T wave inversion in V4 revealed that those patients were indeed having acute coronary events at the time the ECGs were taken. Since acute ischemic ST-T changes were infrequently seen in the CAD group, the sensitivity and negative predictive value (NPV) were low though the specificity and positive predictive value (PPV) were relatively high (

Discussion
CAD is one of the most common findings in patients with LBBB. Recognizing CADassociated ECG changes is of clinical importance. Among all the criteria proposed, (Havelda, Sohi et al. 1982;Fesmire 1995;Sgarbossa, Pinski et al. 1996) the Sgarbossa criteria (Sgarbossa, Pinski et al. 1996) seem very promising with good predictive values reported by the investigators. Since it is mainly targeted at ST-T changes at an acute or sub-acute phase of myocardial infarction, when applied to identify both acute and chronic ischemia, a low sensitivity was reported (Shlipak, Lyons et al. 1999;Gunnarsson, Eriksson et al. 2001).

QRS-ST-T Morphology CAD Non-CAD p Value
Our study examined ECG changes associated with acute coronary syndromes (ACS) as well as in patients with stable CAD. Consistent with previous studies, there were no QRS-ST-T pattern differences between the patients with and without CAD.
Regardless of etiology poor R wave progression is highly prevalent in our study cohort (Table 1, Figure 1).
This ECG wave pattern reflects the right ventricle to septum and to left ventricle activation sequence in LBBB. Reversal of this pattern has been used in evaluating the effectiveness of CRT. (Sweeney, van Bommel et al.) Although broad/notched/slurred R waves are predominant in V6, we found that RS morphology (rS, rs or Rs) is the second most common waveform in V6, seen in 40% of CAD patients and 26% of non-CAD patients with LBBB. rS in V6, is part of poor R progression and can be well explained by the right to left activation sequence. (Sweeney, van Bommel et al.) 120 Depending on the blocking sites LBBB can be divided into pre-divisional, divisional and intramural subtypes. (Runge, Dorner et al. 1973;Jazayeri, Caceres et al. 1989;Childers, Lupovich et al. 2000) Pre-divisional block is at the level of left bundle trunk, divisional is at fascicular level and intramural at the intramyocardial Purkinje network. The lower level block is often associated with myocardial involvement and worse prognosis. In our study cohort, 21% of CAD and 24% non-CAD patients had LAFB ECG pattern, indicating the presence of divisional LBBB with complete blockage at LAF and partial blockage at left posterior fascicular branch level. Together they account for 42.6% of LBBB patients with leftward QRS axis.
Ischemic ST-T changes were only seen in patients with ACS. Since the majority of CAD patients do not have ACS at the time ECGs are taken, the sensitivity and negative predictive value (NPV) are low though the specificity and positive predictive value (PPV) are high ( Table 2).
The second part of this study examined the correlation between QRS duration and LVEF in LBBB patients with and without CAD. LBBB patients are more predisposed to having LV www.intechopen.com systolic dysfunction. (Littmann and Symanski 2000;Talreja, Gruver et al. 2000) LBBB patients also have a higher mortality (Huvelle, Fay et al.) when compared to patients who also have LV systolic dysfunction without LBBB. (Abdel-Qadir, Tu et al.) This in part can be attributed to the higher than normal incidence of CAD, hypertension, cardiomyopathy and valvular heart disease in patients with LBBB.(Abdel-Qadir, Tu et al.) These co-morbidities may be responsible for the LBBB itself. However, LBBB results in alteration of both the systolic and diastolic properties of the LV. The alterations in the diastolic filling time and septal wall motion cause both systolic and diastolic dysfunction (Grines, Bashore et al. 1989), therefore reduce the global cardiac performance. This effect, as shown in this study, is possibly exaggerated by and is positively proportional to the QRS duration. Moreover, we demonstrated that the negative correlation between QRS duration and LVEF is present in both patients with and without CAD, further proof of the negative outcome independently caused by LBBB.

Conclusions
Other than acute ischemic ST-T changes seen in patients with ACS, the overall LBBB morphology is not different in patients with and without CAD. Increased QRS duration in LBBB has a significant negative correlation with LVEF in both CAD and non-CAD patients.