Adults who smoke or who are genetically predisposed to smoking behaviors are more likely to experience a serious type of stroke called subarachnoid hemorrhage (SAH), according to new research published today in Stroke, a journal of the American Stroke Association, a division of the American Heart Association.
The results of this study provide important evidence that there is a causal link between smoking and the risk of SAH.
SAH is a type of stroke that occurs when a blood vessel on the surface of the brain ruptures and bleeds into the space between the brain and the skull. It mainly affects middle-aged adults and has high rates of complications and death.
“Previous studies have shown that smoking is associated with higher risks of SAH, yet it has been unclear if smoking or another confounding condition such as high blood pressure was a cause of the stroke,” said senior author of the study Guido Falcone, M.D., Sc.D., M.P.H, an assistant professor of neurology at Yale School of Medicine in New Haven, Connecticut.
“A definitive, causal relationship between smoking and the risk of SAH has not been previously established as it has been with other types of stroke.”
To determine whether there is a causal effect of smoking and SAH, researchers analyzed the genetic data of 408,609 people from the UK Biobank, ages 40 to 69 at time of recruitment (2006-2010).
Incidence of SAH was collected throughout the study, with a total of 904 SAHs occurring by the end of the study.
Researchers developed a genetic risk scoring system that included genetic markers associated with risk of smoking and tracked smoking behavior data, which was collected at the time each participant was recruited.
Researchers found that:
- the relationship between smoking and SAH risk appeared to be linear, with those who smoked half a pack to 20 packs of cigarettes a year having a 27% increased risk;
- heavier smokers, those who smoked more than 40 packs of cigarettes a year, were nearly three times more at risk for SAH than those who did not smoke; and,
- people who were genetically predisposed to smoking behaviors were at a 63% greater risk for SAH.
“Our results provide justification for future studies to focus on evaluating whether information on genetic variants leading to smoking can be used to better identify people at high risk of having one of these types of brain hemorrhages,” said Julian N. Acosta, M.D., neurologist, postdoctoral research fellow at the Yale School of Medicine and lead study author.
“These targeted populations might benefit from aggressive diagnostic interventions that could lead to early identification of the aneurysms that cause this serious type of bleeding stroke.”
Researchers say while their findings suggest a more pronounced and harmful effect of smoking in women and adults with high blood pressure, they believe larger studies are needed to confirm these results.
Their analysis is also limited by the type of data used in the UK Biobank, which, like all large information resources, rely on standardized treatment codes from medical charts, whereas smaller studies are focused on more detailed health records and information for each individual.
Subarachnoid hemorrhage (SAH) is a serious disease with a high rate of unfavorable outcomes [1,2,3,4]. The rate of treatment of unruptured aneurysms (UIAs) has been increased with a purpose to reduce incidence of SAH [4,5,6].
However, the effect of treating UIAs on the incidence SAH has been low [7], compared with that of reducing the prevalence of smoking [7,8]. In any case, the majority of diagnosed UIAs never rupture during the remainder of the patient’s lifetime even in the case of patients of working age [9,10].
Since the prevalence of aneurysms is not decreasing [11], the risk of rupture is likely decreasing, resulting in lower SAH incidence rates [7,8]. The indications for treatment of UIAs are already challenging because of a lack of prospective studies of natural history of UIAs [12].
Since such studies are not possible to complete nowadays without treatment selection, interest has been focused on studies of aneurysm growth, assuming this to be an important prognostic factor for aneurysm rupture [13,14,15]. Most UIA growth studies have also been derived from patient populations with treatment selection bias with low rupture rates. UIA growth has been shown to increase risk of aneurysm rupture [14,15].
According to two recent meta-analyses of aneurysm growth, female sex, cigarette smoking at baseline, aneurysm size, posterior circulation aneurysms, and possibly age, hypertension, and aneurysm multiplicity may increase the risk of UIA growth, while prior SAH and a family history of aneurysms do not [13,14]. Aneurysm growth rates were lower in prospective and high-quality studies [13].
A large meta-analysis of individual data from 1507 patients in 10 cohorts showed UIA growth in 257 cases (17%) during 5782 patient-years of follow-up [16]. Predictors for aneurysm growth during a mean follow-up of 3.8 years were Earlier SAH (decreased risk), Location of the aneurysm, Age >60 years, Population, and Sze and Shape of the aneurysm (ELAPSS score).
The 3-year growth rate ranged between <5% and >42% and the 5-year growth rate between <9% and >60%, depending on the risk factors. The results between ELAPSS score and previous meta-analyses differed between each other for sex, age, previous SAH, populations, and some aneurysm locations. Cigarette smoking was not studied in the ELAPSS study.
Most of the cohorts were from selected populations where the patients with UIAs were those left for conservative follow-up because of an estimated low rupture risk, or a limited life expectancy or high treatment risks [16]. UIA may, however, be unchanged >20 years, then grow a little and rupture [15,17,18].
An almost lifelong follow-up cohort of UIAs [9,15] seen at our institution was not subject to treatment selection bias and had been considered to be the highest quality study and to have the least source of bias of any prospective natural history studies of both UIA rupture and growth [12,13,14].
The aim here is to investigate the validity of the ELAPSS score in predicting the long-term UIA growth risk and to search for a new score from variables known at baseline which could better predict the UIA growth rate.
Discussion
The growth score of UIAs obtained from this prospective study with an almost lifelong follow-up and a very low treatment selection bias suggests that growth of UIAs among patients of working age can be estimated quite reliably with only three variables at baseline. Female sex, patient age (<40 years), and cigarette smoking determine the risk of UIA growth significantly better than does the ELAPSS score.
Simultaneously with the discovery of increasing numbers of UIAs, it is likely that the indications for treatment will decrease, because of the decline in the prevalence of smoking and the incidence of SAH [7,8] and because of the discovery of incidental UIAs in older people more than previously.
These patients are also less likely to be smokers or to have higher treatment risks because of age. The cost-effectiveness ratio of UIA treatment is thus increasing. To obtain more reliability in the treatment decision, follow-up angiographic investigations have become more prevalent to search for UIAs that grow. Growing of UIAs correlates with later rupture, particularly in a long-term follow-up [14,15,17,18].
According to the results of two recent meta-analyses of aneurysm growth, female sex, cigarette smoking at baseline, aneurysm size, posterior circulation aneurysms, and possibly age, hypertension, and aneurysm multiplicity may increase the risk of UIA growth, while prior SAH and a family history of aneurysms do not [13,14].
There was a substantial heterogeneity for risk factors between studies, except in the case of female sex, smoking at baseline, and hypertension. High data quality (low risk of study bias), Japanese and Finnish populations, and prospective study design were associated with decreased growth risk [13].
The results of these meta-analyses were subject to treatment selection of UIAs and follow-ups were mostly short extending only approximately for 3–5 years. Because of treatment selections most UIA growth studies have shown low rupture rates [13,14]. High risk UIAs (large UIAs in younger patients and smokers) are usually treated already at baseline without any follow-up [12].
The present study with an almost lifelong follow-up of UIA patients also included patients with a high UIA rupture risk as was seen in the relatively high UIA growth rate among patients of working age. Expectedly, UIA growth score obtained from the present score included partly same factors as treatment score based on risk factors of UIA rupture [21].
The treatment score (the AUC for the ROC 0.755) predicted slightly better aneurysm rupture than did present growth score for the growth of UIA. Both these scores included cigarette smoking, patient age <40 years, and female gender, the latter being of borderline statistical significance in the treatment score. Aneurysm size and some locations were included in the treatment score but not in the growth score.
This is not surprising since e.g., aneurysms in the anterior communicating artery are more prone to rupture with a smaller size than those in middle cerebral artery which grow more before rupture [15].
Previously published results from this cohort [15,17,18] have been reported to have the best data quality and the lowest bias of all UIA growth studies to date, in spite of reporting the lowest approximate annual incidence of growth [13,14]. ELAPSS score study [16] with a relatively short follow-up (median 2.5 years) has some contradictory results with PHASES score study which was planned to predict aneurysm rupture risk (e.g., for previous SAH and some aneurysm locations) [12].
Although ELAPSS score study is a pooled individual patient data meta-analysis its results also conflict with those of two other recent meta-analyses for sex, age, previous SAH, populations, and some aneurysm locations [13,14]. Predictors for aneurysm growth were Earlier subarachnoid hemorrhage (decrease risk!), Location of the aneurysm, Age > 60 years, Population, Size of the aneurysm, and Shape of the aneurysm (ELAPSS).
Of the total 1507 patients with UIAs only 18 patients (1.2%) had an aneurysm rupture, and of 257 growing UIAs only 8 (3.1%) ruptured. Cigarette smoking which has been most significant risk factor for UIA growth and SAH [24,25] was not analyzed in ELAPSS score study. UIA growth in long-term follow-up also correlated better with its rupture [14,15].
Mean follow-up among patients of working age between diagnosis and rupture in the present study was >10 years. Routine follow-up angiography during 1–3 years after the diagnosis seems also to be unreliable since aneurysm may be stable >20 years, then grow 1 mm and rupture [15,18].
Selection criteria for aneurysm occlusion treatment and conservative follow-up with/without control angiograms in participating centers in the ELAPSS score study were heterogeneous. In some countries, small UIAs in elderly are routinely followed radiologically. In Finland only those patients with uncertainty for treatment are followed up on clinical-based criteria.
This was seen also in the ELAPSS score study where patients with angiographic follow-up in Finland were significantly younger than elsewhere and these patients are known to be more likely smokers and have higher UIA growth risk [13,15,17,18]. This has leaded arbitrarily to a high ELAPSS score among Finns although in one meta-analysis, the growth rate of their UIAs was even lower [13].
ELAPSS score describes preferably the growth of low rupture risk UIAs in patients left to conservative treatment. This was also seen in the low aneurysm rupture rate in the ELAPSS score study. These results are difficult to be generalized to all patients with UIAs.
The strengths of this study lie in the complete and almost lifelong follow-up of patient of working age and the very limited treatment selection bias [9,15,18]. Correspondingly, a previous aneurysm growth study based on this cohort [15,17,18] was considered to be of high study quality and have low sources of bias relative to other UIA growth studies [13,14].
Although Finnish people have been considered to be subject to a higher risk of aneurysm rupture, the incidence of SAH is no higher in Finland than elsewhere when standardized for the study design with its inclusion and exclusion criteria, the accuracy of diagnosis, and the sex and age distributions of the population [8].
The Nordic countries seem to have similar incidences of SAH. A recent twin study has shown that the contribution of genetic factors as the cause of SAH appears to be slight, and the main causal factor seems to be cigarette smoking [26].
A limitation of this study is the relatively small sample size, despite a long total follow-up time as compared with other large prospective study populations, which on the other hand, had shorter follow-ups and a high treatment selection bias. Patients with a prior history of SAH have not been shown to have a significantly higher risk of aneurysm rupture or growth than others when confounding factors are taken into account [12].
In this study, all initial UIA measurements were obtained from two-dimensional conventional angiograms, which may yield less accurate measures than three-dimensional angiograms. For UIA growth measurement the same projections were done as was done in initial angiograms.
The present simple, rapid scoring system for UIA growth is significantly better and easier to use than the ELAPSS score. This new growth score contains also the same factors than the treatment score obtained from this cohort. The latter score is directed to evaluate rupture risk of UIAs and its value for prediction is somewhat better than that for growth score. The treatment score is preferable in clinical practice for treatment decision and it likely reduces need of follow-up angiographies.
reference link: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7603243/
More information: Stroke (2021). DOI: 10.1161/STROKEAHA.120.031622
