Impact of Cigarette Smoking on Hematoma Expansion and Functional Outcomes in Patients with Intracerebral Hemorrhage

Abstract

Background—Hematoma expansion is a predictor of mortality and functional outcomes in patients with
intracerebral hemorrhage. The effect of cigarette smoking on hematoma expansion and functional outcomes
has not been established.
Methods—Retrospective analysis of patients with intracerebral hemorrhage, recruited in multicenter clini‐
cal trials, was conducted by using the Virtual International Stroke Trials Archive information regarding
smoking status, baseline, 24-hour and 72-hour hematoma volumes, and a 90-day modified Rankin Scale
score. Patient demographics, comorbidities, admission National Institutes of Health Stroke Scale and Glas‐
gow Coma Scale scores, baseline and follow-up hematoma volumes and the rates of hematoma expansion
(defined as the increase in hematoma volume of either ≥6 mL or >33% on follow-up CT scans), hematoma
retraction (defined as the decrease in hematoma volume of either >5 mL or 30% on follow-up CT scans),
and functional outcomes were compared between active, former, and nonsmokers. Functional independence
was defined by a modified Rankin scale score <3 at 90 days postoccurrence. Multivariable logistic models
were used to compare hematoma expansion and functional outcomes between the groups.
Results—A total of 119 (14%) out of the 860 patients who were analyzed were active cigarette smokers,
123 (14%) were former smokers, and 618 (72%) were nonsmokers. Active cigarette smokers were younger
(59 vs. 67 years, p<0.0001) and more likely male compared to nonsmokers (79% vs. 56%%, p < 0.0001).
No statistically significant differences were observed in the rates of hematoma expansion at 24 hours
between active cigarette smokers and nonsmokers [adjusted odds ratio (aOR), 1.68; 95% confidence inter‐
val (CI), 0.75–3.38; p = 0.20] and active cigarette smokers and former smokers (aOR 1.73; 95% CI, 0.54–
5.48; p = 0.1), and in the rates of hematoma expansion at 72 hours between active cigarette smokers and
nonsmokers (aOR 1.48; 95% CI, 0.96–2.28; p = 0.17), and active cigarette smokers and former smokers
(aOR 1.23; 95% CI, 0.71–2.16; p = 0.19). There was no significant difference in the rates of functional
independence (aOR 0.79; 95% CI 0.33–1.94, p = 0.62), excellent outcome (aOR 1.02; 95% CI 0.36–2.81, p
= 0.98), and mortality (aOR 0.72; 95% CI 0.21–2.58, p = 0.61) at 90 days between active cigarette smokers
and nonsmokers. Due to the small sample size (27 and 119 patients with active smoking status on hema‐
toma expansion analysis at 24 and 72 hours, respectively), the possibility of type II error and, therefore,
false negative result of association between smoking status and the rate of hematoma expansion was not
excluded.

Conclusion—Although the point estimates suggest an association between active cigarette smoking and
hematoma expansion, wide confidence intervals could not express a beneficial, harmful, or neutral effect.
Future research with a larger sample size is warranted to reliably investigate this association.