The core finding of this study was that NRT has no significant impact on VS. NRT was associated with higher 90-day mortality, poorer outcomes and prolonged hospitalization in light smokers. In contrast, moderate to heavy smokers appeared to have some benefit from NRT, exhibiting significantly lower rates of atrial fibrillation and numerically shorter hospitalization times. NRT had no impact on smoking cessation, regardless of smoking amount.
NRT and Vasospasm
The definition and diagnosis of VS varies in the literature and one can assume differences between studies that are caused by different methods to establish the presence of VS. We therefore scored our patients both for radiological/ultrasonographical VS, clinical VS (which many denote as DCI) and radiologically ischemic lesions (here defined as DCI). NRT was not associated with VS among moderate to heavy smokers, regardless of how it was measured. In light smokers, radiological/ultrasonological VS was more often absent without NRT. However, it is unlikely that a finding of no VS versus light or moderate VS has any clinical implication or effect on outcome. Our findings concur with those from other reports that did not find an association between NRT and increased frequency or severity of angiographic VS [5, 38, 32], or clinical VS [5]. In these studies, VS was either diagnosed using digital subtraction angiography (DSA) [5] or TCD [38, 5], though Panos et al. did not specify which modality was applied for assessing VS [32]. Carandang et al. even reported less clinical VS with NRT, where clinical VS was defined as any radiological narrowing of a vessel (either > 25% from baseline or > 50% from normal) with an associated clinical deterioration not attributable to any other cerebral or non-cerebral complications with the latter corresponding to what we presently denoted as clinical VS [5]. Moreover, Seder et al. is the only study to mention DCI in relation to NRT, in which they did not find DCI more often with NRT [38]. DCI was in that study defined as ‘clinical deterioration or cerebral infarction due to vasospasm’[38].
There are reports of nicotine exerting both vasodilatory and vasoconstrictive effects. Administration of nicotine has been associated with increased cerebral blood flow in chronic smokers [34, 9, 40]. Given that vasoconstriction and endothelial dysfunction contribute to the development of VS [1, 8, 18], it seems reasonable to assume that nicotine’s vasodilatory effects could counteract these mechanisms. Nicotine may hence reduce the severity of hypoperfusion, potentially lowering the risk and extent of DCI [7, 18]. On the other hand, chronic smoking has also been associated with reduced cerebral blood flow, primarily due to nicotine’s inhibitory effect on nitric oxide (NO) synthesis, a mechanism implicated in the pathogenesis of VS [9, 20, 46, 48]. However, most studies have not stratified their findings by smoking amount, making it difficult to determine whether the observed vasoactive effects were influenced by cumulative nicotine exposure. This lack of differentiation complicates the interpretation of results and highlights the need for a more nuanced understanding of nicotine’s impact on the cerebral circulation.
One may argue that nicotine’s vasoactive properties are highly dependent on prior nicotine exposure, which could explain the different trends observed in light smokers compared to moderate to heavy smokers. Elbejjani et al. found that higher pack-years were associated with significantly increased cerebral blood flow among current smokers [9]. A potential explanation for this finding is that prolonged exposure to nicotine may induce vascular adaptations that shift nicotine’s effects toward vasodilation when administered in low doses [34, 9, 40]. However, what constitutes a “low dose” is relative to prior nicotine exposure—what may be considered a low dose for moderate to heavy smokers could be excessive for light smokers. This could explain why NRT had a slight association with VS in light smokers, as their lower baseline nicotine levels may have made them more susceptible to the unwanted effects of nicotine. Conversely, in moderate to heavy smokers, the same NRT doses may have instead promoted nicotine’s vasodilatory effects, mitigating VS risk [9]. Still, if nicotine exerted a pronounced vasodilatory effect, one would expect a significantly lower incidence of VS among moderate to heavy smokers receiving NRT. Since this was not the case, it could be argued that the administered dose was sufficient to partially mitigate VS formation rather than completely counteract it. This raises further questions about the threshold at which nicotine transitions from a vasoconstrictive to a vasodilatory agent, which likely depends on both prior exposure levels and the acute dosing regimen.
It seems worthwhile to point out that nicotine is not synonymous with tobacco smoke. While nicotine is one of the key vasoactive components in tobacco, it is only one of many substances that may contribute to the development of VS and DCI. Some studies challenge the idea that nicotine alone is integral to the formation of VS [1, 8, 5], suggesting that other toxic components in tobacco smoke may play a more significant role [5, 49]. These include carbon monoxide and polycyclic aromatic hydrocarbons [10], among more than 7000 other chemicals present in tobacco smoke [41]. Both carbon monoxide and polycyclic aromatic hydrocarbons are associated with increased oxidative stress and inflammation [35, 16, 30], which in turn may alter cerebrovascular tone and promote endothelial dysfunction [30, 16]. One should acknowledge the impact of the noxious components found in tobacco smoke when discussing the relationship between NRT and VS, because nicotine when administered as NRT is not equal to nicotine interacting with possibly 7000 other chemicals in tobacco smoke. While nicotine itself can exert vasoactive effects that depend on prior nicotine exposure [9, 20, 46, 48], its role in the development of VS appears to be modulated by the presence of these additional toxic compounds. It is plausible that nicotine acts synergistically with these substances in tobacco smoke to drive endothelial dysfunction, sympathetic activation, and vasoconstriction, ultimately leading to VS in smokers. This potential synergy could in part explain why NRT was not associated with increased VS in moderate to heavy smokers, as nicotine alone may not exert the same degree of vasoconstrictive influence as it does when combined with the other constituents of tobacco smoke.
NRT and complications other than VS
Presently, NRT was associated with less development of atrial fibrillation in moderate to heavy smokers. aSAH can trigger both arrhythmias, including atrial fibrillation, and ECG disturbances, with an estimated 50–100% of patients experiencing the latter in the acute phase [15, 14]. Nicotine withdrawal is associated with dysregulation of the hypothalamic-pituitary-adrenal axis end thereby with the stress-response system [19]. This stress may increase the vulnerability for atrial fibrillation. This effect would be strongest in those used to the highest nicotine blood levels. NRT did not seem to increase the risk of atrial fibrillation in aSAH in light smokers. A few reports proposed a link between atrial fibrillation and NRT, but they were conducted on patients who smoked concurrently with NRT use or who used NRT improperly [37, 44, 28]. None of our patients smoked during their hospital admission.
NRT did not lead to an increased occurrence of respiratory failure and thromboembolism. This phenomenon was observed irrespective of smoking amount and aligns with prior studies on side effects of NRT [28]. During the first day of smoking cessation, carbon monoxide in the blood decreases to normal levels which leads to swelling of erythrocytes in smokers and thereby enhances the risk of thromboembolism during that time frame. It is uncertain if this is due to nicotine or rather other compounds in smoke. Our results do not support any notion of reduced thromboembolic risk with nicotine substitution. A study on NRT and post-operative complications found no overall increase in post-operative complications, including respiratory failure [43]. Approximately 24 hours after smoking cessation, the respiratory cilia regain better function and there is a higher production and transportation capacity of mucus which would cause increased coughing in awake smokers. In patients with reduced consciousness this can lead to secretory stagnation and increased respiratory problems. If nicotine plays a significant role in this effect, its replacement could theoretically reduce time on mechanical respiratory support in poor grade patients. We actually observed shorter times for ICU stays and respiratory support in our moderate to heavy smokers that received NRT. It is, however, questionable whether this is attributable to a nicotine effect in the lungs or due to lower stress levels with NRT, allowing extubation at an earlier point of time.
NRT and outcome
NRT was associated with poor outcome in our light smokers. Since NRT was not associated with DCI, it seems likely that NRT affected outcome through other mechanisms than those related to DCI. While no clinical studies have directly addressed this, animal and experimental models suggest that chronic nicotine exposure may alter the cerebral microcirculation and induce microscopic thromboemboli [22, 52, 11, 2]. This microvascular dysfunction would, however, also occur in moderate to heavy smokers in whom we did not see a negative effect from NRT on outcome. If NRT is able to exacerbate microcirculatory dysfunction independent of VS, the lack of this effect in moderate to heavy smokers raises questions. A plausible explanation is that the impact of nicotine on cerebral vessels is influenced by prior nicotine exposure, as chronic exposure could alter nicotinic receptor expression and downstream receptor-mediated effects in ways that modify vascular response.
If nicotine has vasodilatory effects on intracerebral vessels, it will also lead to an increase in cerebral blood volume and thereby ICP. This effect would be most prominent in the light smokers that are administered relatively high doses of nicotine. Increased ICP would be managed according to protocol so that one can assume that the ICP did not reach levels that would impact outcome negatively. It could, however, prolong hospitalization times as we saw in our light smokers.
Inflammatory responses following stroke, including aSAH, are also relevant when considering outcome differences based on smoking amount. The literature on nicotine’s role in post-stroke inflammation is conflicting. Some studies conducted on animal models have found that low dose nicotine exerts a protective effect on neuronal tissue by attenuating cytotoxicity, minimizing tissue injury and reducing cell death in the case of cerebral ischemia [29, 51]. Nicotine has also been implicated in neuroplasticity and induction of neurotrophic factors, potentially improving cognitive function [3, 47]. On the other hand, there are also studies that associate nicotine with exacerbated inflammatory responses following stroke, including brain edema and reperfusion injuries [2, 24, 39]. Both chronic and acute exposure of nicotine were associated with pro-inflammatory mechanisms [2, 24, 39]. Notably, the levels of nicotine exposure in these studies vary widely, complicating the interpretation of dose-dependent effects. This variability underscores the importance of considering prior nicotine exposure when assessing the vascular and inflammatory consequences of NRT, as pre-existing receptor adaptations and systemic responses may modulate its effects.
Exposure to nicotine through cigarette smoke is different from the continuous supply provided by a nicotine patch. When smoking, there would be a nighttime break of nicotine exposure which is not the case with a nicotine patch. Smoking 5 cigarettes/day leads to approximately 1.5mg x 5 of nicotine entering the body, which is far less than the daily dose of a patch of 14-21mg. Light smokers may hence be exposed to higher doses of nicotine than they are used to which could lead to higher blood pressure and heartrate, nausea and vomiting, headache, as well as hyperventilation. In contrast to that, in moderate to heavy smokers, the nicotine patch is a more adequate substitution of the nicotine deficit arising with smoking cessation upon hospitalization.
If assuming a positive effect of NRT, one may also argue for the opposite, i.e. not a situation of too much nicotine but too little nicotine contributing to poorer outcome. Nicotine lowers stress and pain by reducing dopamine in the brain, an action that seems to be dependent on the endogenous opioid system [19]. Substances that counteract the opioid system, such as Naloxone, would hinder nicotine’s stress reduction. In aSAH patients that are able to swallow, we often use painkillers with the combination of oxycodone/naloxone, which could then either reduce or nullify nicotine’s protective effects against stress and withdrawal symptoms, thereby potentially worsening delirium.
Carandang et al. [5] have not seen a negative impact on outcome with NRT, however, their NRT group had less severe aSAH, whereas our NRT patients had a higher severity. Moreover, they scored outcome at discharge, whereas we scored it at 90 days and we expressed outcome with the mRS which is more nuanced than the Glasgow Outcome Score that Carandang et al. used [5]. Panos et al. found that NRT was not associated with increased VS when controlling for aSAH [32], and Seder et al. reported that NRT was associated with lower mortality [38]. However, neither stratified their patients by smoking amount, complicating the interpretation of NRT on VS and mortality.
NRT has been considered a useful tool in permanent smoking cessation. Still, in the setting of NRT in acute aSAH, there was no difference between the non-NRT and NRT group regarding smoking cessation. This may be attributable to the patients being forced into smoking cessation and not having chosen to quit themselves. While some patients report to have lost the craving for cigarettes after their aSAH, others fall back into their usual habits once they return to their home and may even smoke more due to inability to work and boredom.
Strengths and limitations
The retrospective character of the study is a clear limitation; however, data are from a prospective database assumed to contain relatively high quality data. Our study includes more than twice the number of those included in earlier studies (that also were retrospective). The single-centre design of the study limits its external validity; especially outcome is a complex multifactorial construct that would be affected by institutional treatment protocols and possibly modify the effect of NRT. A strength is that we also took into consideration the intensity of chronic nicotine exposure prior to aSAH, which unveiled somewhat different risks in light versus moderate to heavy smokers. Still, the amount of cigarettes smoked may have been underreported by patients. Furthermore, precise data on the duration of nicotine exposure prior to ictus was not available, which is unfortunate as it could have provided valuable insights into potential dose-dependent effects of nicotine on VS, DCI, and functional outcomes. A strength is that none of the included patients actively smoked during hospitalization, minimizing the confounding effects of tobacco smoke. This is particularly important, as previous studies examining the relationship between nicotine/NRT, VS and extra-cerebral complications have not always distinguished between the effects of nicotine itself and those of tobacco smoke.
We did not systematically assess nicotine withdrawal symptoms, preventing us from evaluating the efficacy of NRT in mitigating withdrawal-related complications. We also had no measurements of delirium or important medical variables like heart rate, blood pressure, and ICP levels available within this study. Finally, NRT dosing and selection for NRT was not standardized and was decided by the treating physician, introducing the possibility of treatment bias. Future studies should preferably be carried out in a randomized controlled design.