Fluctuating Electrocardiographic Changes Predict Poor Outcomes After Acute Subarachnoid Hemorrhage

INTRODUCTION

The incidence of subarachnoid hemorrhage (SAH) from intracranial aneurysm is estimated to be 6-28 per 10,000 population.¹ SAH produces significant morbidity and mortality in patients, and only an estimated one-third of patients with SAH will survive functionally. Unique pathophysiologic changes occur after the rupture of an intracranial aneurysm. Although many of these changes occur in the brain, distant effects such as cardiac abnormalities are also commonly seen. Electrocardiogram (ECG) abnormalities, including changes in ST segment and T wave, prolonged QT interval (QTc), and P wave abnormalities, occur in 50%-100% of patients during the acute stage of SAH.^2,3 ECG changes usually occur during the first 42 hours after SAH, but their duration is variable, and ECG usually returns to normal by 6 weeks after the event.⁴ Evidence suggests that cardiac dysfunction is an independent risk factor for death and disability in aneurysmal SAH.^5,6

The ECG changes associated with aneurysmal SAH can be confusing to clinicians because they may be of cardiac origin or neutrally mediated electrophysiologic effects.⁷ Ibrahim and Macdonald demonstrated that ECG changes can occur with angiographic vasospasm after SAH.⁸ However, Huang et al reported that in patients with SAH, the occurrence of nonspecific ST segment or T wave changes and prolonged QTc assessed early in the emergency department are independently associated with in-hospital mortality.⁹

Despite the link between poor outcome and cardiac abnormalities, the outcomes of individual patients with cardiac dysfunction are quite variable. We noted that some patients in the operating room had dynamic changes in their ECG during surgical clipping of the cerebral aneurysm. Consequently, we designed this study to prospectively analyze whether these dynamic changes were associated with worse outcomes in patients with aneurysmal SAH. Our hypothesis was that the dynamic changes in the ECG during the perioperative period can predict outcome for patients with aneurysmal SAH.

RESULTS

Twenty patients (12 females and 8 males) 18-70 years of age (mean 47.2 ± 12.8 years) with aneurysmal SAH within 48 hours of admission were included in the study (Table 1). Patients were classified into 2 groups according to the presence of ECG abnormalities at any time during the study period. Seven (35%) patients had normal ECGs throughout the entire study, while 13 patients (65%) had some ECG abnormalities during the study period. Seven (53.8%) had T wave abnormalities, 6 (46.2%) had ST segment changes, 4 (30.8%) had QTc segment prolongation, 4 (30.8%) had sinus tachycardia, 3 (23.1%) had U waves present, 3 (23.1%) had sinus bradycardia, 3 (23.1%) had frequent ventricular premature contractions, and 1 patient (7.7%) had a nonspecific change in ECG. Four patients (30.8%) had ECG changes that fluctuated among the preoperative, intraoperative, and postoperative ECG time points (ie, the type of abnormality changed between time points).

There were no significant differences between patients with ECG abnormalities vs patients without ECG abnormalities in age; weight; and history of diabetes mellitus, smoking, and hypertension (P>0.05). Baseline imaging showed 17 patients (85%) had aneurysms in the anterior circulation (middle cerebral, posterior communicating artery, or anterior communicating artery), and 3 patients (15%) had posterior circulation artery aneurysms. Fifteen patients (75%) had large or giant aneurysms. Seven patients (35%) had a baseline Hunt and Hess clinical grade of 1, 5 patients (25%) had grade 2, 2 patients (10%) had grade 3, 3 patients (15%) had grade 4, and 3 patients (15%) had grade 5 (Table 2).

As shown in the Figure, during the 3-day preoperative period, the incidence of ECG abnormalities was higher among patients with poor Hunt and Hess clinical grades than among patients with good clinical grades on the first day of hospitalization (100% vs 30%, respectively; P<0.01), on the second day (85.7% vs 16.6%, respectively; P<0.01), and on the third day of hospitalization (84.6% vs 28.6%, respectively; P<0.05). Postoperatively, no difference was detected in the incidence of ECG changes between patients with low clinical grades and patients with high clinical grades.

• Download figure
• Open in new tab
• Download powerpoint

Figure.

Percentages of patients with good clinical grade (Hunt and Hess grades 1 and 2) and poor clinical grade (Hunt and Hess grades 3, 4, and 5) aneurysmal subarachnoid hemorrhage who developed electrocardiographic (ECG) abnormalities during the 3-day preoperative period. The differences between groups were statistically significant on all 3 days (*P<0.01, **P<0.05).

The incidence of ECG changes increased in patients who had lower grades on the Glasgow Coma Scale, meaning that neurologic status was worse among patients in the abnormal ECG group compared to patients in the normal ECG group.

We further categorized the patients depending on their score on the Glasgow Coma Scale. Patients in category I had a Glasgow Coma Scale score of 15, while patients in categories II, III, and IV had Glasgow Coma Scale scores <15. All the patients in the normal ECG group were in category I or II; all the patients in categories III and IV had ECG changes.

We compared the incidence of ECG abnormalities between patients in category I and patients in categories II, III, and IV (Table 3). On the first day of hospitalization, the incidence of ECG changes was 30% among patients in category I and 100% among patients in other categories. This difference was statistically significant (P<0.01). On the second day of hospitalization, the incidence was 16.6% among patients in category I and 85.7% among patients in the other categories. This difference was also statistically significant (P<0.01). On the third day of admission, the incidence was 28.6% among patients in category I and 84.6% among patients in the other categories, and this difference too was statistically significant (P<0.05). One day postoperatively, however, the incidence of ECG abnormalities was 60% among patients in category I and 66.6% among patients in the other categories. This difference was not statistically significant (P>0.05).

The incidence of CK elevation in patients with abnormal ECGs was 69% compared to 0% in patients without ECG abnormalities (P<0.05). The peak elevation occurred 3-4 days after onset of hemorrhage. The mean value of CK-MB was not significantly different in the 2 groups (P>0.05).

Patients in the abnormal ECG group were hemodynamically less stable than patients with normal ECGs. Blood pressure was initially elevated in 5 patients (38.5%) with abnormal ECGs compared to 1 patient (14.3%) in the normal ECG group. Hypotension developed in 3 patients in the abnormal ECG group (23.1%). Heart rate and central venous pressure remained within normal range in all patients.

Outcome at the time of discharge was not related to the mere presence of an abnormality on ECG. The incidence of poor outcome was higher in the group with ECG abnormalities, but the trend was not statistically significant (P>0.05). The presence of fluctuating ECG changes during the study period was a strong predictor of poor outcome. All 4 patients who had fluctuating ECG changes had a poor outcome (100%) compared to patients who had ECG abnormalities but no fluctuation (33.3%, P<0.05) (Table 4).

DISCUSSION

The association of aneurysmal SAH with electrocardiographic abnormalities has been recognized for many years. In their experiments with animals, Offerhaus and van Gool attributed ECG changes and subendocardial lesions similar to those induced by aneurysmal SAH to increased levels of catecholamines.¹⁰ Hawkins and Clower proposed a direct tissue toxic effect of norepinephrine after finding elevated levels of norepinephrine in myocardial tissue post–aneurysmal SAH.¹¹ On the other hand, Karch and Billingham showed that a contraction band necrosis represents the most likely pathologic substrate of myocardial injury in aneurysmal SAH.¹² Other studies have linked cardiac dysfunction and poor outcome.^5,6 Coghlan and colleagues concluded that bradycardia, relative tachycardia, and nonspecific ST segment and T wave abnormalities are strongly and independently associated with 3-month mortality after SAH.¹³

In our cohort of patients, we found an association between ECG changes and the severity of the injury (assessed by Hunt and Hess grade), premorbid hypertension, and CK elevation although not all associations reached statistical significance. We believe that the myocardial enzyme elevation is not from acute myocardial infarction but rather indicates low-grade myocardial damage caused by sustained sympathetic stimulation from cerebral lesions. This myocardial damage may not influence the wall motion or short-term outcome. In 2012, the VISION¹⁴ team reported the peak postoperative troponin T measurement during the first 3 days after surgery was significantly associated with 30-day mortality. We believe postoperative cardiac enzyme measurements in patients who have ECG changes after SAH can enhance risk stratification after surgery.

The unique finding in this study that has not been reported previously in the literature is the contribution of dynamic changes in the ECG to the prognosis for good recovery from aneurysmal SAH. In our group, all the patients who had ECG changes that fluctuated from one abnormal change to another had a poor outcome. The etiology of this finding is not clear but may open the door to further study into the pathogenesis of cardiac changes in aneurysmal SAH. The animal and human studies to date have implicitly assumed that the injury to the myocardium occurs at the time of the hemorrhage and that the ECG changes we see are the sequelae of this early event. Our study may suggest that injury to the cardiac conduction system may be an ongoing process in some patients, meaning that interventions may be available to counteract the detrimental neurologic changes. Conversely, these data could suggest that the initial injury to the heart is greater in patients who later develop fluctuating ECG findings and have poor outcomes. This finding is not well supported by the literature in myocardial ischemia in which the electric abnormalities are typically monomorphic and localized to the area of infarction. Further study is necessary to determine the mechanisms that lead to fluctuating ECG changes in patients with aneurysmal SAH. In our study, patients with dynamic changes in ECG had worse outcomes than patients with fixed ECG abnormalities. Consequently, we suggest that patients with dynamic changes should be vigilantly followed with continuous ECG monitoring.

It is unclear if and why the period around surgery is important for ECG changes. On the one hand, our study was constructed around the surgical intervention and may only reflect the changes that occur early after SAH. In contrast, surgery could act as a kind of stress test that uncovers autonomic instability after SAH. Also, it is possible that some process around surgery affects some patients and not others irrespective of SAH, although the association with poor outcome would suggest that the association is not trivial.

This study has several limitations. This was an observational study with a small cohort of patients. The advantage of the patient population is that the treatment of the aneurysm was uniform. In current practice, >50% of ruptured aneurysms are treated with coil embolization. In the university hospital where this study was performed, coil embolization was not a treatment option. A second limitation is that outcome was measured at the time of discharge. Follow-up for several months may be needed to confirm outcomes. Unfortunately, follow-up information for this study cohort was not available. The clinical utility of the variability of ECG abnormalities needs to be validated in a larger cohort of patients with longer follow-up.

CONCLUSION

The results of our study suggest that patients with aneurysmal SAH who have fluctuating ECG abnormalities have poor outcomes compared to patients with consistently abnormal ECGs. We believe that elevated cardiac enzymes and fluctuating ECGs after surgery strongly predict poor outcomes for patients after SAH.