Impact of Traditional and Non-Traditional Lipid Parameters on Outcomes after Intravenous Thrombolysis in Acute Ischemic Stroke

Contradicting evidence exists regarding the role of lipids in outcomes following intravenous (IV) thrombolysis with tissue plasminogen activator (tPA). Restricted cubic spline curves and adjusted logistic regression were used to evaluate associations of low-density lipoprotein cholesterol (LDL-C), non-high-density lipoprotein cholesterol (non-HDL-C), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C) and LDL-C/HDL-C ratio with poor functional outcome, symptomatic intracranial hemorrhage (SICH) and 90-day mortality, among 1004 acute ischemic stroke (AIS) patients who received IV tPA in a comprehensive stroke center. Quartile (Q) 1, Q2 and Q3 of HDL-C were associated with increased odds of poor functional outcome (adjusted odds ratio (adjOR) 1.66, 95% CI 1.06–2.60, p = 0.028, adjOR 1.63, 95% CI 1.05–2.53, p = 0.027, adjOR 1.56, 95% CI 1.01–2.44, p = 0.048) compared to Q4. Q2 and Q4 of non-HDL-C were associated with increased odds of SICH (adjOR 4.28, 95% CI 1.36–18.90, p = 0.025, adjOR 5.17, 95% CI 1.64–22.81, p = 0.011) compared to Q3. Q1 and Q2 of LDL-C was associated with increased odds of mortality (adjOR 2.57, 95% CI 1.27–5.57, p = 0.011 and adjOR 2.28, 95% CI 1.10–5.02, p = 0.032) compared to Q3. In AIS patients who received IV tPA, low LDL-C was associated with increased odds of mortality while HDL-C may be protective against poor functional outcome.


Introduction
Stroke incidence and stroke-related mortality increased substantially from 1990 to 2019, and the stroke burden will continue to increase globally, especially in underdeveloped countries [1]. In Singapore, a study reported an increased prevalence of stroke risk factors and the increased crude incidence rate of stroke among those younger than 65 [2]. There has been conflicting evidence on how lipid parameters affect post-thrombolysis outcomes [3,4], with non-traditional lipid parameters like non-HDL-C suggested to have similar functions as LDL-C in predicting hemorrhagic transformation (HT) [5]. Pre-stroke statin use was also shown to influence ischemic stroke outcomes [6].Hence, we aimed to explore the general associations of selected traditional and non-traditional lipid parameters with postthrombolysis outcomes, to compare their clinical utility in prognostication. One significance of this study is to facilitate determination of cholesterol target levels in achieving optimal post-stroke recovery. Lastly, there are varied results on the relationship between lipid parameters and large artery atherosclerosis (LAA) stroke subtype. Bang et al. found that higher non-HDL-C and TG, but not LDL-C, was associated with LAA. Several studies have also shown that LDL-C may not best predict atherosclerotic vascular risk [7]. However, Hindy et al. suggested LDL-C lowering likely reduces LAA risk [8]. These should be further assessed to evaluate the utility of targeting lipid parameters for LAA stroke prevention.

Study Design
In this study, we included consecutive patients who received IV tPA from September 2006 to June 2018. All these patients had no contraindications to IV tPA use. The study obtained ethics approval from the National Healthcare Group-Domain Specific Review Board (NHG DSRB Ref: 2010/00509). Patients were assessed by a neurologist for eligibility to receive intravenous thrombolysis according to institutional protocol and American Heart Association/American Stroke Association guidelines at a standard dose of 0.9 mg/kg body weight [9]. All thrombolysed stroke patients underwent standard non-contrast head computed tomography and computed tomography brain and neck angiography. Patients who were deemed as unsuitable for IV-tPA, or who underwent endovascular thrombectomy were excluded from the study. Patients that were included had valid lipid parameters of TC, HDL-C and TG collected in fasting conditions, in mmol/litre, that were taken within 24 h of their AIS admission as per our institution protocol ( Figure 1). outcomes [3,4], with non-traditional lipid parameters like non-HDL-C suggested to have similar functions as LDL-C in predicting hemorrhagic transformation (HT) [5]. Pre-stroke statin use was also shown to influence ischemic stroke outcomes [6].Hence, we aimed to explore the general associations of selected traditional and non-traditional lipid parameters with post-thrombolysis outcomes, to compare their clinical utility in prognostication. One significance of this study is to facilitate determination of cholesterol target levels in achieving optimal post-stroke recovery. Lastly, there are varied results on the relationship between lipid parameters and large artery atherosclerosis (LAA) stroke subtype. Bang et al. found that higher non-HDL-C and TG, but not LDL-C, was associated with LAA. Several studies have also shown that LDL-C may not best predict atherosclerotic vascular risk [7]. However, Hindy et al. suggested LDL-C lowering likely reduces LAA risk [8]. These should be further assessed to evaluate the utility of targeting lipid parameters for LAA stroke prevention.

Study Design
In this study, we included consecutive patients who received IV tPA from September 2006 to June 2018. All these patients had no contraindications to IV tPA use. The study obtained ethics approval from the National Healthcare Group-Domain Specific Review Board (NHG DSRB Ref: 2010/00509). Patients were assessed by a neurologist for eligibility to receive intravenous thrombolysis according to institutional protocol and American Heart Association/American Stroke Association guidelines at a standard dose of 0.9 mg/kg body weight [9]. All thrombolysed stroke patients underwent standard non-contrast head computed tomography and computed tomography brain and neck angiography. Patients who were deemed as unsuitable for IV-tPA, or who underwent endovascular thrombectomy were excluded from the study. Patients that were included had valid lipid parameters of TC, HDL-C and TG collected in fasting conditions, in mmol/litre, that were taken within 24 h of their AIS admission as per our institution protocol. ( Figure  1) LDL-C was calculated by the Friedewald equation (TC-HDL-C-TG/5) [10], while some had LDL-C collected directly. Non-HDL-C levels were calculated by subtracting HDL-C levels from TC. LDL-C/HDL-C ratio was calculated by dividing LDL-C by HDL-C. Other baseline demographics, clinical parameters and ischemic stroke characteristics were collected and tabulated within 24 h of AIS admission. Diabetes mellitus (DM) was defined as pre-existing diagnosis of diabetes mellitus or an admitting fasting blood glucose level greater than or equal to 7.0 mmol/L or an HbA1c greater than or equal to 6.5% [11]. The severity of stroke at presentation was assessed using the National Institute of LDL-C was calculated by the Friedewald equation (TC-HDL-C-TG/5) [10], while some had LDL-C collected directly. Non-HDL-C levels were calculated by subtracting HDL-C levels from TC. LDL-C/HDL-C ratio was calculated by dividing LDL-C by HDL-C. Other baseline demographics, clinical parameters and ischemic stroke characteristics were collected and tabulated within 24 h of AIS admission. Diabetes mellitus (DM) was defined as pre-existing diagnosis of diabetes mellitus or an admitting fasting blood glucose level greater than or equal to 7.0 mmol/L or an HbA1c greater than or equal to 6.5% [11]. The severity of stroke at presentation was assessed using the National Institute of Health Stroke Scale (NIHSS) [12]. This assessment was made by credentialed nurses as part of the acute stroke response team. The presence of Large Vessel Occlusion (LVO) was defined as occlusions of the first and second segment of the middle cerebral artery (MCA), M1 and M2, the Internal Carotid Artery (ICA) and as well as its terminus, tandem occlusions involving ICA-MCA, or occlusion of the basilar artery. The Trial of ORG 10172 in Acute Stroke Treatment (TOAST) criteria was used by the treating stroke neurologist to classify stroke subtypes [13]. Investigations to evaluate the TOAST mechanism included: vessel imaging with either computed tomography angiography (CTA), magnetic resonance angiography (MRA), transcranial doppler (TCD), carotid duplex or a combination of these, as well as Holter monitoring and transthoracic echocardiography.
The primary outcome measured was poor functional outcome (90-day modified Rankin Scale (mRS) of 3-6). Secondary outcomes measured were symptomatic intracranial hemorrhage (SICH) and 90-day all-cause mortality. SICH was based on the European Cooperative Acute Stroke Study (ECASS) II definition [14]. The 90-day mRS was evaluated during follow-up visit to the stroke clinic. If not, mRS was evaluated via telephone call instead.

Data Analysis
Analyses were performed using SPSS for Windows version 27.0 (SPSS Inc, Chicago, IL, USA) and R version 4.0.5. Restricted cubic splines with 5 knots were plotted to visually assess the univariate associations between the lipid parameters and the three outcomes. These served qualitative and descriptive purposes only. Since non-linear (U-shaped/inverse U-shaped/J-shaped/reverse tick) associations were found, lipid parameters were divided into quartiles for analysis. Descriptive statistics for continuous variables were presented as mean (SD) when normality and homogeneity assumptions were satisfied, otherwise as median (interquartile range) (IQR), and n (%) for categorical variables. Differences in continuous variables were assessed using 2 sample t-test when normality and homogeneity assumptions were satisfied; otherwise, Mann-Whitney U test was used where data was not distributed normally. Chi-square or Fisher's exact test was used for categorical variables. Covariates that were selected a priori for variable adjustment were gender, age, hypertension, atrial fibrillation, large vessel occlusion, diabetes mellitus, admitting NIHSS and admitting SBP. Logistic regression assessed the associations between lipid levels and outcomes, and with LAA. Results were presented as adjusted odds ratio (adjOR) with 95% confidence interval (CI). Statistical significance was set at two-sided p < 0.05.

Results
This section may be divided by subheadings. It should provide a concise and precise description of the experimental results, their interpretation, as well as the experimental conclusions that can be drawn. Comparing LAA and non-LAA groups, median LDL-C, non-HDL-C, LDL-C/HDL-C ratio, as well as white blood cell (WBC) count, neutrophils, platelets, admitting NIHSS and presence of LVO were significantly higher in the LAA compared to non-LAA group (Table S1). Regarding lipid parameters and associations with LAA, after adjustment for gender, age, hypertension, atrial fibrillation, large vessel occlusion, diabetes mellitus, admitting NIHSS and admitting SBP, the following results were obtained. Q4 of LDL-C was associated with increased odds of LAA (adjusted odds ratio (adjOR) 1.69, 95% CI 1.07-2.69, p = 0.024) compared to Q1. Q2 and Q4 of non-HDL-C was significantly associated with increased odds of LAA (adjOR 1.64, 95% CI 1.04-2.60, p = 0.035 and adjOR 1.75, 95% CI 1.10-2.80, p = 0.018 respectively) compared to Q1. Q3 of HDL-C was associated with increased odds of LAA (adjOR 1.84, 95% CI 1.16-2.95, p = 0.010) compared to Q4. Lastly, Q3 and Q4 of LDL-C/HDL-C were associated with increased odds of LAA (adjOR 1.88, 95% CI 1.18-3.00, p = 0.008 and adjOR 1.71, 95% CI 1.07-2.77, p = 0.027 respectively) compared to Q1 (Table 2).

Associations of Lipid Parameters with Poor Functional Outcome, SICH and Mortality
When evaluating the following relationships between lipid parameters and outcomes measured, variables adjusted for in the multivariate model include gender, age, hypertension, atrial fibrillation, large vessel occlusion, diabetes mellitus, admitting NIHSS and admitting SBP.

Poor Functional Outcome
Restricted cubic spline curves showed a 'reverse tick' relationship between LDL-C, non-HDL-C and TC with poor functional outcome (Figure 2).

Associations of Lipid Parameters with Poor Functional Outcome, SICH and Mortality
When evaluating the following relationships between lipid parameters and outcomes measured, variables adjusted for in the multivariate model include gender, age, hypertension, atrial fibrillation, large vessel occlusion, diabetes mellitus, admitting NIHSS and admitting SBP.

90-Day Mortality
Restricted cubic spline curves showed a 'reverse tick' relationship of LDL-C, non-HDL-C and TC with mortality. (Figure 2) In multivariate analysis, Q1 and Q2 of LDL-C were significantly associated with increased odds of mortality (adjOR 2.57, 95% CI 1.27-5.57, p = 0.011 and adjOR 2.28, 95% CI 1.10-5.02, p = 0.032) when compared to Q3. Q2 of LDL-C/HDL-C was associated with increased odds of mortality (adjOR 2.51, 95% CI 1.30-5.05, p = 0.008) when compared to Q3. No significant associations were found to relate non-HDL-C, TC and HDL-C with mortality on multivariate analysis ( Table 4). Results of logistic regressions that adjusted for age and gender only, and age, gender, admitting NIHSS and LVO, are illustrated in the Supplementary Material (Tables S4 and S5).

Discussion
It was found that firstly, low HDL-C was associated with poor functional outcome, and secondly, a U-shaped relationship was found between non-HDL-C and SICH. Finally, low LDL-C was also found to be associated with increased mortality.

Lipid Parameters and Functional Neurological Outcome
We found that high HDL-C tended to be protective against poor functional outcome. This corroborates with a past Japanese study that found higher HDL-C and increased odds of favourable functional outcome after tPA thrombolysis [15]. This can be explained by how HDL reduces neuronal injury after ischemic stroke, possibly through anti-oxidative or anti-inflammatory pathways [16]. By suppressing inflammatory responses, HDL-C could promote neurological recovery because inflammatory cells and responses like neutrophils and neutrophil accumulation have been shown to cause poorer neurological outcome [17]. Next, our study found that the relationship between LDL-C/HDL-C and functional outcome may be non-linear, expanding on a previous Chinese study of 763 AIS patients treated with tPA, that demonstrated a LDL-C/HDL-C ratio cut-off of <2.71 was associated with higher risk of poor outcome [18]. Higher LDL-C/HDL-C was also found to be protective against mRS >2 at 3 months [19]. However, the relationship between LDL-C/HDL-C and functional outcome after ischemic stroke thrombolysis continues to be inconclusive.

Lipid Parameters and SICH
When investigating relationships with SICH, we found a U-shaped and non-linear relationship of non-HDL-C and TC with SICH respectively. Previously, a Chinese study found that lower non-HDL-C resulted in greater risk of hemorrhagic transformation [5], but other studies were unable to validate this association [3,20]. Conclusions on associations between TC and SICH were also varied [21,22]. Our finding of a U-shaped relationship between non-HDL-C and SICH where increased risk of SICH occurs at either moderately low or high non-HDL concentrations could explain the discrepancy in prior studies' results. Postulated mechanisms include how low cholesterol reflects poor general health and undernourishment that predisposes individuals to hemorrhagic transformation unrelated to cholesterol pathways [23]. High non-HDL-C could increase risk of SICH through the development of arterial stiffness, which independently increases the risk of hemorrhagic transformation in thrombolysis-treated stroke patients [24]. Next, some studies have suggested a correlation between lower LDL-C and increased SICH risk [25,26]. In our study, non-HDL-C, but not LDL-C, was significantly associated with SICH. This is a surprising finding because LDL-C is a major component of non-HDL-C. However, other lipoproteins which contribute to non-HDL-C, such as very-low-density lipoproteins, intermediate-density lipoproteins and lipoprotein(a) could affect SICH occurrence. For instance, lipoprotein(a) was found to reduce bleeding risk in the brain due to its hemostatic properties [27]. Past studies have also demonstrated significant differences in the predictive ability of non-HDL-C and LDL-C on major adverse cardiac events (MACE) [28], reinstating the different roles of non-HDL-C and LDL-C in cardiovascular and cerebrovascular events.

LDL-C and Mortality
Our study found low LDL-C was significantly associated with increased odds of 90-day all-cause mortality after multivariate adjustment. This association has also been previously suggested outside of stroke-in a systematic review involving 68094 elderly patients, an inverse association between LDL-C and all-cause mortality was found, hypothesized by how low LDL-C increased vulnerability to fatal illnesses [29]. In our study, patients with low LDL-C had more comorbidities, (Table S3) which likely increased mortality predisposition. Hence, it was proposed that low LDL-C is an indirect indicator of severe illness rather than the cause of increased mortality [30]. Another explanation, although speculative, is the interaction between low LDL-C with dysbiosis and changes to bile acid metabolism that eventually leads to mortality [31]. The finding of low LDL-C increasing mortality may also be attributed to AIS patients without prior statin use, supported by Cheng et al. who found that low LDL-C was associated with higher mortality rates in statin-naïve acute ischemic stroke [32]. This finding is therefore generalized without comparison between statin, non-statin or proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor use.

Analysis of Lipid Parameters with LAA Subtype
This study also found that LDL-C, non-HDL-C, HDL-C and LDL-C/HDL-C were significantly associated with LAA after multivariate adjustment. Previous studies found elevated LDL-C a risk factor for atherothrombotic infarct, and LDL-C higher in LAA than other stroke subtypes [33,34]. Our finding of high LDL-C and increased odds of LAA can be explained by LDL-C's association with intracranial and extracranial stenosis [35]. Next, our study found that higher HDL-C may be more desirable than moderate HDL-C levels in protection against LAA, reasoned by how HDL-C increases LDL-C reverse transport, delivers antioxidants to LDL-C and decreases susceptibility of LDL-C to oxidation in endothelium, slowing the process of atherosclerosis [36]. Lastly, our finding of high LDL-C/HDL-C ratio and association with increased odds of LAA can be explained by the atherogenicity of LDL-C and atheroprotection by HDL-C, supporting the previous finding of LDL-C/HDL-C ratio and its association with increased intima-media thickness, a measure of subclinical atherosclerosis [37]. Thus, high LDL-C/HDL-C ratio may be a useful indicator of atherosclerosis to help identify patients at higher risk of LAA stroke.

Strengths and Limitations
Our study is a comprehensive report detailing associations of selected lipid parameters with post-thrombolysis outcomes, representing a relatively large cohort size (1004 patients analysed) compared to previous thrombolysis studies. However, this study was a single institution retrospective cohort study that solely evaluated intravenous thrombolysis patients. This may limit the generalisability of results to other cohorts, warranting more prospective studies on lipid parameters and ischemic stroke outcomes. We would like to acknowledge the possibility of Type I error in the multivariate analyses in which significant associations may no longer hold true after Bonferroni correction. Excessive correction of statistical level of significance may also increase the likelihood of Type II error as a trade-off to reduce Type I error, which may increase false negatives. The multivariate relationship between HDL-C and poor functional outcome is to be interpreted with caution as no significant relationship was found on univariate analysis, in which the possible reasons can be explained statistically by Lo et al. and Wang et al. [38,39]. Hence further studies in this area are required to confirm this association. Our study did not distinguish statin and non-statin users or evaluated use of other lipid-lowering agents, nor compare antithrombotic drug use which could theoretically affect SICH risk. Comorbidities like chronic kidney disease could be a confounder that should be explored in future studies. We would suggest future studies compare ischemic stroke outcomes between intravenous thrombolysis, intra-arterial thrombolysis and mechanical thrombectomy cohorts, and evaluate the role of other non-traditional lipid measures like TG/HDL-C and TC/HDL-C ratio in these treatment options.

Conclusions
In AIS patients who received IV tPA, low LDL-C was associated with increased odds of mortality while HDL-C may be protective against poor functional outcome.   Informed Consent Statement: Due to the retrospective conduct of this study, patient informed consent was waived, and this waiver was approved by NHG DSRB.