Coagulation Functions Are Associated With Hemorrhagic Transformation in Non-Atrial Fibrillation Patients: A Case Control Study

Background: Hemorrhagic transformation (HT) is a serious neurological complication of acute ischemic stroke (AIS) after revascularization. The majority of AIS patients do not have atrial brillation (AF) which could also develop to HT. In this study, we aimed to explore whether coagulation parameters are risk factors of HT in non-AF patients. Methods: We consecutively enrolled 285 AIS patients with HT. Meanwhile, age- and sex-matched 285 AIS patients without HT were included. The diagnosis of HT was determined by brain CT or MRI during hospitalization. All patients were divided into two subgroups based on the presence of AF and explore the differences between the two subgroups. Blood samples were obtained within 24 hours of admission, and all patients were evenly classied into three tertiles according to platelet counts (PLT) levels. Results: In this study, we found the rst PLT tertile (OR = 3.509, 95%CI = 1.268-9.711, P = 0.016) was independently associated with HT in non-AF patients, taking the third tertile as a reference. Meanwhile, we also found mean platelet volume (MPV) (OR = 0.605, 95%CI = 0.455-0.805, P = 0.001) and brinogen (FIB) (OR = 1.928, 95%CI = 1.346-2.760, P <0.001) were signicantly associated with HT in non-AF patients. But in AF patients, coagulation parameters showed no signicant difference. Meanwhile, we found the MPV (OR = 1.314, 95%CI = 1.032-1.675, P = 0.027) and FIB (OR = 1.298, 95%CI = 1.047-1.610, P = 0.018) were signicantly associated with long-term outcomes in non-AF HT patients. Conclusions: Additionaly, and FIB outcomes in non-AF HT patients. Our study showed that coagulation functions at admission may be benecial for clinicians to recognize patients with high risk of HT at early stage and improve unfavorable long-term outcomes in non-AF patients.

It has been reported that AF and some medication used for its treatment such as Warfarin could cause HT [9][10][11][12]. In 2019, Jiao et al. found that AF independently correlated with HT and was a risk factor of HT [8]. In a prospective trial of 101 AIS patients, Tu et al. found that AF patients had signi cant hypoperfusion that may damage vascular integrity leading to a more frequent HT [11].
Meanwhile, Altavilla et al. found that patients with AF who received low-molecular-weight heparin have a higher risk of HT than nonbridged patients [13]. Besides, an animal experiment found that the incidence of HT in rats treated with Warfarin was increased [12].
However, HT has been underexplored in non-AF patients; the majority of AIS patients are non-AF patients [14,15] and they could also develop HT. Meanwhile, AF and medication for AF would in uence coagulation functions [16,17].Thus, the risk factors of non-AF patients to develop HT after AIS needs to be further explored, especially coagulation functions.
In this research, we mainly explored the association between coagulation functions and HT in patients with AF and without AF and aimed to identify coagulation parameters as risk factors in non-AF HT patients.

Subjects
This was a retrospective study of patients with and without hemorrhagic transformation after stroke. The study protocol obtained the approval of the Ethics Committee of the First A liated Hospital of Wenzhou Medical University. We didn't have informed consent for it was a retrospective study and the patient pro le was anonymous. The cohort was made up of the First A liated Hospital of Wenzhou Medical University's clinical database of HT; the same number of stroke patients without HT from our stroke center was matched by age and sex using the same inclusion and exclusion criteria. All patients who were diagnosed with HT in the First A liated Hospital of Wenzhou Medical University were included in this study from December 2013 to December 2015. The inclusion criteria were as follows: (1) age between 18 and 99 years; (2) patient was included within 7 days of stroke; (3) the diagnosis of stroke was con rmed by computerized tomography (CT) or magnetic resonance imaging (MRI) at the time of admission. The exclusion criteria were as follows: (1) hemorrhagic stroke or transient ischemic attack (TIA); (2) patient received intravenous thrombolytic therapy; (3) patients with any severe liver or kidney dysfunction; (4)patients failed to receive a repeat CT/MRI scan; (5) patients' medical record was incomplete.

Diagnosis of HT
All patients received a brain CT scan or MRI including diffusion-weighted imaging (DWI) and T2-weighted gradient-echo within 24 h after stroke onset. HT was diagnosed in a subsequent CT/MRI performed 7 ± 2days after stroke onset or whenever a worsening clinical condition. Two neurologists evaluated the CT/MRI scans independently and diagnosed HT, who were blinded to the results of clinical and laboratory measurements.

Data collection
Patients' demographic data and personal hobbies included gender, age, current smoking, and drinking. Past medical history included the previous history of stroke, hypertension, diabetes mellitus, coronary artery disease (CAD), atrial brillation, and hyperlipidemia.
Meanwhile, the National Institutes of Health Stroke Scale (NIHSS) was used to evaluate the severity of stroke within 24 hours after admission. All patients identi ed the stroke subtypes according to the TOAST criteria [18]. Blood samples and blood pressure were obtained within 24 h of hospital admission and sent blood samples for tests as soon as possible. Laboratory tests included leukocyte counts, erythrocyte counts, platelet counts (PLT), mean platelet volume (MPV), prothrombin time (PT), international normalized ratio (INR) and brinogen (FIB). Also, drug applications in the hospital including anticoagulant, antiplatelet, and lipidlowering drugs were collected and documented. All patients were evenly classi ed into three tertiles according to the PLT levels (tertile 1, < 175; tertile 2, 175-222; and tertile 3, > 222).
In December 2019, HT patients were followed up by telephone to collect their up-to-date functional conditions. We used the modi ed Rankin Scale (mRS) to assess long-term functional outcomes, and an unfavorable functional outcome was de ned as an mRS score of ≥ 2 [19][20][21].

Statistical analyses
Statistical descriptions of continuous variables were mean ± SD or medians (quartiles), while categorical variables were percentages.
The comparison of continuous variables used the Student's t-test or Mann-Whitney test, while categorical variables used the Chisquare test or Fisher's exact test. The comparison of the differences between three PLT tertiles in continuous variables used the Kruskal-Wallis test or one-way analysis of variance (ANOVA), while categorical variables used the Pearson's chi-square or Fisher's exact tests. Multiple logistic regression analysis was performed to identify signi cant independent related factors of HT and adjust potential confounding factors. Cox proportional hazards model was performed to identify the signi cant independent related factors of unfavorable long-term outcomes in non-AF HT patients and adjust potential confounding factors. All statistical analyses used IBM SPSS Statistics for Windows, Version: 19.0.0 (Chicago, IL).Two-tailed P-values < 0.05 were considered statistically signi cant.

Baseline characteristics of patients strati ed by HT
In this study, 285 AIS patients with HT and age-and sex-matched 285 AIS patients without HT were included. The baseline demographic, clinical, and laboratory characteristics of the study patients were presented in Table 1. The mean age of the enrolled patients was 68.9 ± 12.4 years. 388 (68.1%) patients were male and 182 (31.9%) were female. Also, 137 (24.0%) patients had a history of AF. As shown in Table 1, patients with HT had lower baseline systolic blood pressure (SBP), PLT, and MPV; higher leukocyte counts, PT, INR, and FIB compared with patients without HT. They also had higher NIHSS scores at admission. AF was more frequently found in patients with HT and the proportion of AF among HT and non-HT patients was 108 (37.9%) and 29 (10.2%) respectively. Less HT patients received anticoagulant therapy and more received antiplatelet therapy. Besides, patients with HT were more likely to smoke and drink. Demographic and laboratory variables according to PLT tertiles were presented in Table 2. The incidence of HT was signi cantly higher in the rst PLT tertile than the second and third MLR tertiles (63.7% versus 42.8% and 43.2%, respectively; P < 0.001). As shown in Table 2, patients with lower PLT levels were more likely to be drinker and CAD patients; had lower leukocyte counts and FIB levels, and had higher MPV, PT and INR levels; and were more likely to receive anticoagulant therapy, less likely to receive antiplatele therapy.

Characteristics of patients with HT in subcategorized groups of AF
The incidence of HT was higher in the AF subgroup than non-AF (78.8% vs. 40.9%, P < 0.001). In the subgroup of non-AF patients, the signi cant parameters between HT and non-HT (Supplemental Table) were generally consistent with parameters in Table 1. But in AF patients, patients with HT had higher levels of leukocyte counts, lower SBP; higher NIHSS scores at admission; and more smoker and drinker. In AF patients, there were no signi cant differences in coagulation parameters among HT and non-HT patients.

Association between coagulation parameters and HT
The occurrence of HT was used as a dependent variable and the third PLT tertile was used as a reference in the multivariate regression in all patients. After adjusting for confounding and risk factors, multivariate regression analysis showed that the rst PLT tertile (OR = 3.517, 95%CI = 1.526-8.106, P = 0.003; Table 3) was independently associated with HT in all patients. Meanwhile, MPV (OR = 0.698, 95%CI = 0.557-0.875, P = 0.002; Table 3) and FIB (OR = 1.613, 95%CI = 1.199-2.169, P = 0.002; Table 3) were also signi cantly associated with HT in all patients. Considering the potential relationship between AF and coagulation function, we divided all patients into AF and non-AF subgroups.

Discussion
To the best of our knowledge, this is the rst study to explore the association of coagulation parameters and HT in non-AF patients; and to explore the effect of coagulation parameters on long-term outcomes after HT in non-AF patients. Our present study showed that three coagulation parameters including low PLT, low MPV, and high FIB were signi cant and dependent risk factors with HT in non-AF patients, instead of AF patients. Furthermore, we found that MPV and FIB levels were independently and signi cantly associated with unfavorable long-term functional outcomes in non-AF HT patients.
Consistent with many studies, we identi ed that AF was a reliable risk factor of HT, and the incidence of HT among AF patients was higher than non-AF patients [7,8,22,23]. Considering the complex interaction between AF and HT [24], we did a subgroup analysis of our patients to explore the association between coagulation functions and HT. Interestingly, we found none of the coagulation parameters was associated with HT among AF patients. This may be that most AF patients received anticoagulant therapy before AIS which may in uence coagulation parameters.
However, our study suggested that PLT, MPV, and FIB were associated with HT signi cantly in the non-AF patients. PLT is a key factor in the coagulation process. According to clinical guidelines from the American Heart Association/American Stroke Association published in 2018, low PLT levels (< 100,000 counts/µl) would increase the risk of HT [25] and do not recommend these patients to receive reperfusion therapies. Several studies also pointed out that PLT was associated with the occurrence of HT [8,26]. Furthermore, a previous study identi ed that lower coated-platelets counts increased the likelihood of early HT in patients with nonlacunar ischemic stroke [27]. Platelet-endothelial interactions could maintain the structural integrity of blood vessels when stroke occours [28]. The patients with low PLT levels are weak in maintaining the blood vessels' integrity and are more likely to develop HT after AIS. Salas-Perdomo et al. found in the experiment that using an anti-platelet serum in mice would lead to larger intraparenchymal hematomas after stroke [29], thus stressed the vital hemostatic function of platelets in AIS. This may because that platelets can physically cover the damaged vascular endothelium or release protective factors to protect the endothelial barrier function. These all indicated the importance of PLT in HT.
MPV describes the sizes of PLT, which is a marker of PLT activity and in uences bleeding [30,31]. In our study, MPV among HT patients was lower than non-HT. Besides, logistic regression analysis showed low MPV was a risk factor of HT. Early study found that larger PLT had more granules and would produce more vascular activity as well as pre-clotting factors, which may raise the hemostatic e ciency [31,32]. A retrospective study suggested that baseline MPV was associated with unfavorable stroke outcomes but its relationship with HT was still uncertain [33].
Previous studies brie y mentioned the relationship between FIB and HT [26,34,35]. And we found that higher FIB may be associated with a higher incidence of HT among AIS patients, which is congruent with some reports [34,35]. However, Wang et al. found that FIB < 1.50 g/L was a risk factor for HT [26]. We hypothesize that high levels of FIB may be associated with HT by participating in the in ammatory process. When HT occurs, endothelial dysfunction of capillaries would further lead to abnormal blood-brain barrier permeability within the infarcted area [36]. The dysfunction of blood-brain barrier in the infarcted area was proved to be the main casue of HT. Meanwhile, a review proposed that FIB was a ligand of cell surface receptors and this could promote the intercellular adhesion between in ammatory cells and endothelial [37]. In another study, an anti-in ammatory thrombolytic drug called SMTP-7 could decrease the incidence of HT, which suggested the relationship between HT and in ammation [38]. Thus, the relationship between FIB and HT remains unknown and this needs to be further investigated.
In our study, we found that coagulation functions were associated with unfavorable long-term outcomes in non-AF HT patients. Its exact mechanism remains unknown; however, the mechanism by which coagulation function is related to unfavorable outcomes in intracerebral hemorrhage (ICH) is unclear. Unfavorable outcomes after hemorrhage were reported to be conducted by the in ammatory effects of intraparenchymal blood [39].Recently, Krenzlin H et al. found that the activated cerebral thrombin system was related to poor outcome after ICH in mice. They suspected that the activated cerebral thrombin system may contribute to secondary brain damage [40]. Meanwhile, a previous study suggested that coagulation function may in uence neurovascular injury and neuroprotection [41].
Our study has some limitations. First, this was a single-center and retrospective study, so it is necessary to conduct multi-center, prospective studies to establish causality and provide long-term prognostic information. Second, our study did not discuss the association between different subtypes of HT (hemorrhagic infarction or parenchymal hematoma) and coagulation functions. In the further study, we could explore the association between the severity of HT and coagulation functions. Third, owing to the infarction size were not documented, the associations between infarction size and HT was not described in detail.

Conclusions
In summary, we found that three coagulation parameters including low PLT, low MPV, and high FIB were risk factors among the non-AF patients, while these parameters were not associated with HT in AF patients. Meanwhile, MPV and FIB levels were independently and signi cantly associated with unfavorable long-term outcomes in non-AF HT patients. Our study demonstrated that the coagulation functions may be useful hematological markers for clinicians to recognize patients with high risk of HT at early stage and improve unfavorable long-term outcomes in non-AF patients.

Declarations
Ethical standard This study was approved by the Institutional Review Board (IRB), the First A liated Hospital of Wenzhou Medical University, and was following the Declaration of Helsinki promulgated by the National Institute of Health. We didn't have informed consent for it was a retrospective study and the patient pro le was anonymous.