QRS duration correlates with poor prognosis in patients with left bundle branch block (LBBB), but the importance of left-axis deviation (LAD) is not well established. To determine if LAD confers a mortality risk in patients with LBBB, a single-center, retrospective, population-based cohort study was conducted. Included were all patients at 1 hospital with LBBB on electrocardiography from 1995 to 2005 over a 17-year follow-up period (n = 2,794, median follow-up duration 20 months, interquartile range 6 to 64). Half of all patients with LBBB had LAD. The all-cause mortality rate in the entire cohort was 15%. LAD was not associated with mortality, either as a single outcome (odds ratio [OR] 1.1, 95% confidence interval [CI] 0.88 to 1.3, p = 0.50) or in time-to-event analysis (p = 0.40). Significant risk factors for mortality included high creatinine (OR 1.2, 95% CI 1.1 to 1.3), low hemoglobin (OR 1.2, 95% CI 1.1 to 1.3), history of atrial fibrillation (OR 1.6, 95% CI 1.3 to 2.1), electrocardiographic evidence of previous infarct (OR 1.5, 95% CI 1.2 to 1.9), and history of ventricular tachycardia (OR 1.4, 95% CI 1.0 to 1.9). On bivariate analysis, LAD was associated with atrial fibrillation, ventricular tachycardia, age, and congestive heart failure. Patients with LBBB who converted from normal axis to LAD had significantly higher mortality in time-to-event analysis (p = 0.02). In conclusion, in patients with LBBB, LAD does not confer significant mortality risk. However, those with normal axis who developed LAD during the study period had significantly higher mortality. Perhaps when LBBB and LAD develop concurrently, there is no increased risk over baseline LBBB development, but it may herald a worse prognosis if LAD develops against the background of previous LBBB, from an unknown mechanism.
Ventricular conduction delay (VCD) afflicts a large proportion of patients with cardiomyopathy from heart failure (HF), hypertension, or coronary artery disease and is a marker of poor prognosis. However, data are conflicting regarding the relative importance of frontal plane axis in patients with conduction delay. In patients with HF with right bundle branch block, there is no influence of axis shift on the clinical or echocardiographic progression of heart disease. Early data using pathologic specimens of left bundle branch block (LBBB), additionally, suggest no significant difference in those with left-axis deviation (LAD). Furthermore, the most common mechanism of LAD is left anterior fascicular block (LAFB), which itself has an uncertain influence on cardiovascular prognosis. One study of only 200 patients showed LAFB and LBBB with LAD to confer mortality risk, but several small case-control studies have demonstrated equivocal risk. LAFB causes LAD because of resulting left ventricular activation solely via the left posterior fascicle, but LAFB alone cannot be the explanation for LAD in patients with complete LBBB. Proposed mechanisms of LAD in patients with LBBB include anatomic considerations (right bundle branch insertion or cardiomegaly) and changes in myocyte mass (inferoposterior infarction or left ventricular hypertrophy). We conducted a large, single-center, cohort study to test the hypothesis that LAD in patients with existing LBBB confers increased risk for mortality from any cause compared with those with LBBB and normal axis.
Methods
We conducted a retrospective, observational cohort study in all patients at our hospital (University of Pennsylvania Health System, Philadelphia, Pennsylvania) who underwent electrocardiography for any reason from 1995 to 2005. LBBB, LAD, and other electrocardiographic variables were queried from the computerized records of all electrocardiograms obtained in our health system, which are stored in the MUSE database (GE Healthcare, Milwaukee, Wisconsin). All electrocardiograms were overread by clinical cardiologists and electrophysiologists employed at the hospital to ensure accuracy of diagnosis. Clinical parameters were obtained for a 17-year follow-up period from the Penn Data Store (Philadelphia, Pennsylvania). The study was approved by the University of Pennsylvania Institutional Review Board. The date of the index electrocardiographic study was used to calculate time-to-event analysis. The primary end point was subsequent mortality from any cause, as documented by the electronic medical record; if no death was recorded, the time end point was the most recent clinical encounter. In a prespecified subgroup analysis, we compared those who entered our study with LBBB and normal axis with those who then developed LAD during the course of the study. LBBB was defined as QRS interval >120 ms, delayed intrinsicoid deflection in lead V 6 , and rS or QS deflection in leads V 1 and V 2 ; LAD was defined as R axis <−30°; and right-axis deviation (RAD) was defined as R axis >90°.
Categorical variables are presented as frequencies and percentages and continuous variables as mean ± SD if normally distributed and median (interquartile range) if skewed. Associations of LAD with clinical variables were assessed using Pearson’s correlation coefficients when normally distributed and Spearman’s correlation when skewed. Study groups were compared using Student’s t tests for continuous variables and chi-square tests for dichotomous variables. Skewed data were compared using Wilcoxon’s rank-sum test (bivariate associations) or Mann-Whitney tests (group differences). Kaplan-Meier curves were constructed using death as the primary event and censoring by either loss to follow-up or conclusion of the observation period. The Mantel-Haenszel log-rank test was used to evaluate for significant difference across curves, and Cox proportional hazards were calculated for individual time points. All p values are 2 tailed, with p <0.05 indicating statistical significance. Analysis was performed using JMP version 10.0 (SAS Institute Inc., Cary, North Carolina).
Results
There were 572,383 electrocardiograms obtained in 177,863 patients in our health system during the study period. We identified 3,897 individual patients with ≥1 electrocardiogram showing LBBB (2.2%). Of these patients, clinical data were available for 2,794 (71.6%), who were included in further analyses. The median follow-up time was 20 months (interquartile range 6 to 64).
Baseline characteristics of the cohort are listed in Table 1 . The overall all-cause mortality rate was 15%, or 9% per year of follow-up. Table 2 lists all clinical parameters according to vital status. Mortality was significantly predicted by previous myocardial infarction (odds ratio [OR] 1.5, 95% confidence interval [CI] 1.2 to 1.9, p = 0.0023), high creatinine (OR 1.2, 95% CI 1.1 to 1.2, p = 0.0001), high blood urea nitrogen (OR 1.03, 95% CI 1.02 to 1.04, p = 0.0001), low hemoglobin (OR 1.2, 95% CI 1.1 to 1.3, p = 0.0001), and any history of atrial fibrillation (OR 1.6, 95% CI 1.3 to 2.1, p = 0.0001) or ventricular tachycardia (OR 1.4, 95% CI 1.01 to 1.91, p = 0.047). Unexpectedly, low cholesterol (total or low-density lipoprotein cholesterol) was also associated with mortality (see Table 2 ).
| Parameter | n | Value |
|---|---|---|
| Age (yrs) | 2,794 | 72 (63–79) |
| Women | 1,152 | 41% |
| Chronic renal insufficiency | 293 | 15% |
| Diabetes mellitus | 692 | 19% |
| Tobacco use | 146 | 3.9% |
| Coronary artery disease ∗ | 548 | 15% |
| Any previous myocardial infarction | 436 | 11% |
| Anterior wall infarction | 98 | 2.5% |
| Inferior wall infarction | 161 | 4.3% |
| Heart failure | 1,301 | 46% |
| Dyslipidemia † | 752 | 27% |
| Atrial fibrillation | 575 | 15% |
| Ventricular tachycardia | 256 | 6.6% |
| LAD | 1,959 | 50% |
| Creatinine (mg/dl) | 1,702 | 1.2 (0.9–1.6) |
| Blood urea nitrogen (mg/dl) | 1,138 | 22 (15–32) |
| Potassium (mEq/L) | 1,655 | 4.3 ± 0.6 |
| Total cholesterol (mg/dl) | 252 | 150 (127–187) |
| High-density lipoprotein (mg/dl) | 79 | 46 ± 14 |
| Low-density lipoprotein (mg/dl) | 114 | 80 (62–107) |
| Triglycerides (mg/dl) | 377 | 117 (87–165) |
| Hemoglobin (mg/dl) | 16,811 | 11.7 (10.3–13.2) |
∗ Coronary artery disease was coded as a clinical diagnosis including International Classification of Diseases, Ninth Revision, code 412 at any visit during the study period.
† Dyslipidemia was coded as any lipid disorder including International Classification of Diseases, Ninth Revision, codes 272, 272.0, 272.1, and 272.2 at any visit during the study period.
| Parameter | Dead (n = 431) | Living (n = 2,363) | OR | 95% CI | p Value |
|---|---|---|---|---|---|
| Age (yrs) | 71.5 (63–78) | 72 (63–79) | NS | ||
| Women | 181 (42%) | 967 (41%) | NS | ||
| Renal insufficiency | 62 (24%) | 229 (20%) | NS | ||
| Diabetes | 115 (38%) | 575 (39%) | NS | ||
| Tobacco use | 17 (8.3%) | 129 (13%) | 0.62 | 0.370–1.05 | 0.07 |
| Coronary disease | 86 (31%) | 461 (34%) | NS | ||
| Any previous infarction | 89 (17%) | 334 (12%) | 1.5 | 1.2–1.9 | 0.0023 |
| Heart failure | 210 (53%) | 1088 (55%) | NS | ||
| Dyslipidemia | 90 (32%) | 659 (42%) | 0.64 | 0.490–0.840 | 0.0012 |
| Atrial fibrillation | 121 (23%) | 431 (15%) | 1.6 | 1.30–2.10 | 0.0001 |
| Ventricular tachycardia | 50 (9%) | 195 (7%) | 1.4 | 1.01–1.90 | 0.047 |
| LAD | 195 (45%) | 1047 (44%) | 1.1 | 0.880–1.30 | NS |
| Creatinine (mg/dl) | 1.6 (1.1–2.6) | 1.1 (0.9–1.6) | 1.18 | 1.10–1.20 | 0.0001 |
| Blood urea nitrogen (mg/dl) | 31 (20–48) | 21 (15–30) | 1.03 | 1.02–1.04 | 0.0001 |
| Potassium (mEq/L) | 4.3 (3.9–4.7) | 4.2 (3.9–4.5) | NS | ||
| Total cholesterol (mg/dl) | 130 (106–169) | 153 (129–187) | 0.99 | 0.980–0.990 | 0.008 |
| High-density lipoprotein (mg/dl) | 39 (30–46) | 45 (40–55) | 0.95 | 0.890–1.010 | 0.06 |
| Low-density lipoprotein (mg/dl) | 62 (42–84) | 80.5 (62–109) | 0.98 | 0.960–0.990 | 0.05 |
| Triglycerides (mg/dl) | 116 (83–175) | 117 (89–164) | NS | ||
| Hemoglobin (mg/dl) | 11.0 (9.6–12.6) | 11.9 (10.4–13.3) | 0.84 | 0.730–0.930 | 0.0001 |
Overall survival for the cohort stratified by LAD is shown in Figure 1 . There was no difference in mortality rate across the study period (log-rank p = 0.36). At any of the prespecified time points of 6, 12, 24, 48, and 60 months from index electrocardiography, there was no difference in mortality rates between the 2 groups (hazard ratios all cross unity, p >0.05).
Bivariate correlations with LAD were calculated ( Table 3 ). Patients with LAD were older and more likely to have HF, atrial fibrillation, and ventricular tachycardia (p <0.05 for all). Although the overall death rate was higher in the population with LAD (16.3% vs 14.4%), this was not statistically significant. We additionally examined the subgroup of patients with normal axis upon entering the study but who developed LAD on follow-up electrocardiography. Characteristics of this population are listed in Table 4 . The overall survival curves, compared with those with LBBB and normal axis, are shown in Figure 2 . There was significantly higher mortality in the group with axis shift compared with those who retained a normal axis (log-rank p = 0.02). The most significant hazard ratio for mortality is at the 6-month time point (hazard ratio 0.57, 95% CI 0.34 to 0.94, p = 0.03). Adjustment for age and gender did not significantly change this relation (data not shown).
| Bivariate Correlate | Normal Axis (n = 1,323) | LAD (n = 1,467) | p Value |
|---|---|---|---|
| Age (yrs) | 66.5 ± 14 | 69.7 ± 13 | <0.001 |
| Women | 42% | 40% | NS |
| Chronic renal insufficiency | 158 (7.9%) | 135 (7.8%) | NS |
| Diabetes mellitus | 375 (19%) | 317 (18%) | NS |
| Smoking | 88 (4.4%) | 58 (3.3%) | NS |
| Coronary disease | 278 (14%) | 270 (16%) | NS (0.06) |
| Previous myocardial infarction | 210 (5%) | 226 (6%) | NS |
| Heart failure | 676 (34%) | 625 (36%) | 0.010 |
| Dyslipidemia | 418 (21%) | 334 (19%) | NS |
| Atrial fibrillation | 226 (6%) | 349 (9%) | <0.001 |
| Ventricular tachycardia | 111 (2.8%) | 145 (3.7%) | 0.035 |
| Creatinine (mg/dl) | 1.2 (0.9–1.7) | 1.2 (0.9–1.6) | NS |
| Blood urea nitrogen (mg/dl) | 21 (15–32) | 22 (16–33) | NS |
| Potassium (mEq/L) | 4.26 ± 0.6 | 4.26 ± 0.6 | NS |
| Total cholesterol (mg/dl) | 150 (129–189) | 149 (125–181) | NS |
| High-density lipoprotein (mg/dl) | 47.8 ± 13 | 44.6 ± 13 | NS |
| Low-density lipoprotein (mg/dl) | 85 ± 40 | 85 ± 38 | NS |
| Triglycerides (mg/dl) | 122 (90–166) | 116 (84–163) | NS |
| Hemoglobin (mg/dl) | 11.6 (10.2–13.2) | 11.9 (10.3–13.3) | NS |
| Deaths | 240 (16.3%) | 191 (14.4%) | NS |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
