American Heart Association

Monthly Archives: June 2013

Low DHEAS associated with increased risk of ischemic stroke in women

Waimei Tai, MD



  • A large observational nested case-control study by Jimenez et. al. using collected data from the Nurses Health Study demonstrated decreased dehydroepiadrosterone sulfate (DHEAS) in women who have an ischemic stroke during the >10 year follow up period. 

    DHEAS is the major precursor for 50% of androgens in men, 75% of estrogens in premenopausal women, and almost 100% of estogrens in postmenopausal women. Low DHEAS level has been previously associated with increased incidence of ischemic heart disease.  

    In this study, the lowest quartile of DHEAS was associated with being older, less educated, higher prevalence of hypertension, high cholesterol, history of heart disease, and higher proportion of hormone therapy use. However, after adjusting for matching risk factors, women in the lowest quartile of DHEAS versus the highest quartile of DHEAS had a non-significant increased risk for ischemic stroke.

    The authors propose that a novel association with a stable hormonal biomarker with increased ischemic stroke risk. Whether this is merely an association and a risk marker of underlying vascular disease, or a causative factor in ischemic stroke, be it directly or through its downstream androgenic/estrogenic activity, is unclear.

    I think this is an interesting study, as it potentially proposes new markers of disease that may not have been previously studied. As the authors’ point out, this initial study may suggest other new research questions. Much work in the neuro-hormonal arena has been focused on how hormones act as neuro-transmitters, or interact in the brain’s chemical milieu. Now potentially new directions of research can look at how hormones act on cerebral vessels, or neuronal activity and recovery.

    HgbA1c Predicts Symptomatic Hemorrhage after Thrombolysis in Ischemic Stroke

    Waimei Tai, MD

    Rocco A, Heuschmann PU, Schellinger PD, Köhrmann M, Diedler J, Sykora M. Glycosylated Hemoglobin A1 Predicts Risk for Symptomatic Hemorrhage After Thrombolysis for Acute Stroke. Stroke. 2013


    Rocco, et al. published an interesting article recently looking at the relationship between glycosylated hemoglobin A1c and the risk of bleeding after an acute ischemic stroke. This was a registry study at a single, large stroke center of acute ischemic stroke patients treated with iv-tpa. Patients were evaluated for symptomatic intracranial hemorrhage (sICH) and other clinical factors including HgbA1c. Without surprise, higher NIHSS, history of diabetes, entry serum glucose and higher HgbA1c were associated with a higher sICH rate. In fact, the c-statistic for HgbAc was 0.8 to predict sICH which is quite high for a clinical parameter.

    In addition, after multivariate anaylsis, age, NIHSS and HgbA1c were independent predictors of dependent status (mRS 3-5). Similarly age, history of diabetes sICH, and HgbA1c predicted risk of death at 90 days.

    Previous studies have suggested that diabetes and hyperglycemia at stroke presentation increases sICH and subsequent poor outcomes for ischemic stroke patients. This has been the premise for such randomized controlled trials such as SHINE (Stroke Hyperglycemia Insulin Network Effort Trial, NCT01369069) which is evaluating strict glycemic control with intravenous insulin versus more traditional subcutaneous insulin regimens in the first 72 hours after ischemic stroke.

    This study bolsters current clinical knowledge about the relationship between diabetes and outcomes from acute ischemic stroke. HgbA1c, a quantifiable measure, is a better independent predictor for sICH and functional status after stroke, than merely the history of diabetes alone.

    I don’t think this would really change current clinical practice. We already routinely check HgbA1c as a marker for diabetes in our acute ischemic stroke patients. We also routinely get repeat imaging 1-2 days after acute iv thrombolysis treatment and monitor patients carefully to evaluate for sICH. This may perhaps persuade me to be more cautious or delay in starting antithrombotic therapy if poorly controlled diabetic patients had any hint of micro hemorrhages on imaging. How do you see this affecting your clinical practice?

    Dabigatran May Be Safer Than Warfarin

    Claude Nguyen, MD

  • Hori M
  • Connolly SJ
  • Zhu J, Li
  • u LS
  • Lau C-P
  • Pais P. 
  • Dabigatran Versus WarfarinEffects on Ischemic and Hemorrhagic Strokes and Bleeding in Asians and Non-Asians With Atrial Fibrillation. Stroke. 2013


    For Asians with atrial fibrillation, dabigatran may be safer than warfarin.The RE-LY trial provided evidence that dabigatran is superior to warfarin in the secondary prevention of stroke in those with atrial fibrillation, with similar rates of hemorrhagic complications at the 150mg BID dose. Despite this promise, however, it may be years before we understand the true pharmacogenomics of any new medication. Hori et al., sponsored by Boehringer Ingelheim, represented a consortium of Asian and Southeast Asian countries in addition to the US and Canada, which studied the bleeding rates of Asians versus non-Asians as a subgroup analysis of the RE-LY trial. They were motivated by prior evidence that Asians taking warfarin are at higher risk of developing hemorrhage. 

    Asians in the RE-LY trial were compared to non-Asians in terms of the 3 treatment groups in the trial (dabigatran 110mg BID vs. 150mg BID vs. warfarin). Asians were younger, had lower body weight, and higher rate of prior stroke, while non-Asians had higher proportion of previous MI. The group also performed a geographic analysis of the data, finding that it correlated with stroke and bleeding rates by ethnic origin. In both Asians and non-Asians, rates of stroke or systemic embolism was lower with the higher dabigatran dose compared to warfarin. There was no evidence of a treatment effect by region. Given that Asians were younger than non-Asians in the study, there was a greater benefit for dabigatran amongst Asians, which was attenuated after age adjustment. 
    The group found that Asians treated with warfarin had higher incidences of bleeding, despite being younger and having lower therapeutic INR ranges than non-Asians. They also found that bleeding rates were significantly lower with dabigatran compared to warfarin amongst the Asian group; however, an interaction between treatment and region was found. Major bleeding was higher in Asians, thought to be not only due to genetic factors, but to a higher proportion of previous stroke compared to non-Asians.

    As with any subgroup analysis, this study is not without limitations, particularly as the proportion of Asians was small compared to the non-Asian group. Even though there are similarities between Asian subgroups, it is important to note that Asians are a heterogeneous group, as evidenced by the regional differences between Asians in the study and its effect on treatment. Therefore, these results may not necessarily be applicable to Americans with Asian heritage. Nevertheless, this is the best evidence we have that dabigatran may be a safer alternative than warfarin in Asians with atrial fibrillation. Before we accept these findings, however, they must be replicated in other studies.

    Life’s Simple 7

    Jose Gutierrez, MD, MPH

  • Kulshreshtha A
  • Vaccarino V
  • Judd SE
  • Howard VJ
  • McClellan WM
  • Muntner P. 
  • Life’s Simple 7 and Risk of Incident StrokeThe Reasons for Geographicand Racial Differences in Stroke Study. Stroke. 2013

    Life simple seven (LS7) have been promoted by the American Heart Association as a tool to promote cardiovascular health in the population and track it over time. Some have advocated that this tool also represent an opportunity to focus in modifiable risk factors that can reduce the risk of cardiovascular disease. Evidence supporting the validity of the LS7 to identify those at risk of cardiovascular disease, particularly stroke, has been piling up. In this issue of Stroke, Cushman et al report that individuals from the REGARDS study who had the highest score (representing the healthier group) had the smallest risk of incident stroke while those in the lower end of the LS7 had the highest risk of stroke. The author identified a reduction of 8% in stroke risk per score point in the LS7. The risk of stroke was similar for White and Blacks in equal categories of LS7, although Blacks had a lower prevalence of cardiovascular health factors. Only minority had five or more ideal LS7: 3% of Whites and 8% of Blacks. 

    The results by Kulshreshtha et al. and those of other groups that have demonstrated the usefulness of the LS7 to measure cardiovascular health and to identify those at higher risk of vascular events shed a dire picture about the current status of ideal cardiovascular health in the US. At the same time, these results may bring a glimpse of hope to curb the vascular-related mortality and morbidity epidemic by focusing in promoting changes at a population levels. It is encouraging to see that a single point change in the score is associated with reduce stroke risk, so even a modest gain should be encouraged. From a public health perspective, investing resources in accomplishing a healthier lifestyle seems the smartest way to go in preventing vascular disease, particularly stroke.

    Wake-Up Stroke t-PA

    Aaron Tansy, MD
    Manawadu D, Bodla S, Jarosz J, Keep J, and Kalra L. A case-controlled comparison of thrombolysis outcomes between wake-up and known time of onset ischemic stroke patients. Stroke. 2013

    Unknown symptom duration, often a problem in patients with wake-up ischemic stroke (WUIS), is a major barrier to the use of t-PA. Consequently, there has been a strong interest within the stroke community to evaluate the feasibility and safety of ways to identify ways to expand thrombolytic eligibility to this cohort. Dulka Manawadu and colleagues show, in a forthcoming paper in stroke, that thrombolysis is a potential treatment option in some patients who wake up with a stroke. Analyzing clinical data culled from a comprehensive stroke center registry, they performed a small, retrospective comparison of post-thrombolytic clinical outcomes (90 day mRS, ICH, and mortality) in patients with symptom duration between 4.5 – 12 hours (WUIS) and <4.5 hours (reference group, RG). WUIS major screening criteria for t-PA eligibility included NIHSS >/5 and no significant evidence of early ischemic change (EIC) on head CT by ASPECTS scoring. Data were analyzed from 68 WUIS and 326 RG cases treated with t-PA. Results revealed that there were no significant differences between the demographically well-matched treatment groups in any of the prescribed clinical outcome metrics.

    Although the group’s study has some significant limitations including its retrospective nature, small case size and high potential for selection bias in the WUIS cohort, it nevertheless holds significance: the findings support the idea that certain cases of WUIS may be amenable to thrombolytic therapy. Not only does this study foster hope that t-PA may be a future treatment option in WUIS but it also provides data for planning a prospective, randomized controlled studies, like the SPOTRIAS-sponsored MR WITNESS trial currently underway. Only then we will truly know if t-PA for wake-up strokes is more than just a dream.

    Canabis Lifestyle


    Modifiable risk factor modification is obviously an important component of stroke prevention. Although marijuana use has been linked to adverse cardiovascular effects and vascular conditions such as reversible cerebral vasoconstriction syndrome, its effect on stroke risk is unclear. Barber et al. looked at patients who received urine drug screens and compared them to controls who were matched for age, sex, and ethnicity. Of the patients who were positive for cannabis, they found a higher proportion of men and smokers. Using a logistic regression model adjusting for age, sex, and ethnicity, there was an association between cannabis use and increased risk of ischemic stroke or TIA; however, this association was weakened when adjusting for tobacco use. 

    A major limitation of the paper was in its selection of the control group, as socioeconomic status and use of other substances could not be adequately matched. As a result, it remains unclear whether the association between marijuana and stroke/TIA is confounded by these other factors. Nevertheless, it suggests that marijuana use is not benign. As many states are at a crossroads in the consideration for marijuana to be legalized for medicinal purposes, it is important to consider the growing evidence that cannabis may have more harmful effects than previously thought.

    Clinical outcomes associated with degree of reperfusion achieved in target mismatch: Pooled data from the DEFUSE studies

    Vasileios-Arsenios Lioutas, MD


    The goal of all our interventions in acute ischemic stroke is salvaging as much hypoperfused tissue as possible, with the hope that this will improve clinical outcomes. Although clinically intuitive, the idea that targeting the penumbra would lead to better outcomes was not rigorously tested until the last few years when several studies, including DEFUSE 1 (published in 2006) and DEFUSE 2 (published in 2012) studied the use of MRI diffusion and perfusion imaging to select patients for treatment. The underlying idea is that following an ischemic insult, a “core” infarct of irreparably damaged tissue ensues, corresponding to increased intensity signal in the Diffusion Weighted MRI sequences. This is surrounded by hypoperfused, “stunned” tissue that is potentially salvageable, widely known as ischemic penumbra. Its imaging correlate is thought to be altered signal in perfusion weighted images that exceed the extent of the core infarction in DWI sequences; that is the penumbra is represented by the “mismatch” between PWI and DWI sequences.

    Starting with broad definitions, the investigators identified imaging mismatch patterns that were thought to predict better outcome if reperfusion was achieved. This mismatch pattern was termed “target mismatch” and its main feature was defined as a ratio of critically hypoperfused tissue and ischemic core of ≥1.8. The leading hypothesis was that achieving reperfusion of the ischemic penumbra would translate into meaningful clinical outcome in patients with target mismatch profile (TMM), whereas no such relation exists in patients with no target mismatch (no TMM).

    As already reported in each of the 2 studies individually, the findings support the primary hypothesis: Reperfusion correlates with favorable clinical outcome in TMM patients, whereas no TMM patients do not seem to benefit. Furthermore, increasing degree of reperfusion was associated with improved clinical outcome in TMM patients, with those within the highest quartile (94-100% reperfusion ) benefitting significantly more compared to those in the lower three quartiles. After adjusting for other independent outcome predictors (DWI lesion volume and age), each 10% increase in the degree of reperfusion had an OR of 1.3 (95% CI: 1.13-1.49) for favorable clinical response and 1.31 for good functional outcome (95% CI: 1.14-1.50).

    The data is interesting, but it should be kept in mind that the total number of patients is relatively small (121 patients in total analyzed in this substudy) and especially the number of no TMM patients was limited (23 patients only), not allowing for detailed analysis in this subgroup. Also, all patients included in the two studies received either intravenous tPA or endovascular treatment (in the extended time window of >3hrs from symptom onset), essentially not including a control group.

    From one point of view, the two trials could be viewed as proof-of-principle studies: they provide convincing evidence that perfusion is a reliable marker of underlying cerebral dysfunction/ischemia and that reperfusion correlates well with clinical improvement and functional outcome in a patient subgroup with a certain neuroimaging profile. Whether perfusion imaging should become routine and affect our management options (essentially classifying patients as “responders” vs “non-responders”) needs to be more extensively investigated, perhaps including patients within the 3 hour treatment window.