Stroke is a leading cause of death and disability in all developed countries. Non-traumatic intracerebral hemorrhage (ICH) constitutes ~10–15% of acute strokes and has the highest case fatality among all strokes, with ~40% of patients dying within the first month (1, 2). Despite the significant advances in acute ischemic stroke care, pharmacological treatment in ICH is still limited, and therefore, it remains the most devastating cerebral vascular disease (1). This is at least partly due to an existing knowledge gap in the pathophysiology of intracerebral bleeding and the mechanisms driving hematoma progression. Coagulation disorders and impaired hemostasis have been shown to increase the risk of ICH (3, 4). However, studies on associations between hemostasis abnormalities and the outcome of ICH remain limited. In a prospective series of patients with ICH, significant hemorrhage growth occurred in 38% of patients, which was associated with outcomes (5). Besides a handful of studies indicating that increased admission D-dimer levels predict mortality (6–8), the impact of the hemostasis system on the evolution of hematoma and the outcome of acute non-traumatic ICH has not been fully investigated.

Thrombin generation assay (TGA) is a global hemostasis test providing information about the speed and amount of generated thrombin in plasma (9). The analysis of the thrombin generation curve (thrombogram) reveals the concerted action of plasmatic pro- and anticoagulant factors; therefore, the assay has been suggested to be a promising candidate to demonstrate hypo-or hypercoagulability in patients. TGA has been shown to correlate with the clinical phenotype in hemophilia and with the response to treatment (10–12). TGA seems also potentially useful in the field of thrombotic disorders, to predict the risk of recurrent venous thromboembolism and to help tailor antithrombotic treatments (10, 12). The aim of this study was to evaluate whether the TGA correlates with the evolution of hematoma in patients with non-traumatic, spontaneous ICH and to see whether this test might serve as a useful tool to predict outcomes.

In this study, 87 patients with non-traumatic, spontaneous ICH and 164 healthy controls were enrolled. Two patients with COVID-19 were not included in the study: in case of one patient SARS-CoV-2 infection was detected on admission, another patient acquired SARS-CoV-2 infection on day 25 after the event, and thus, this patient was also excluded from the study. The assumed cause of ICH was hypertension in all patients, as based on the exclusion criteria, other major causes of ICH, including cerebral aneurysm, arteriovenous malformation, malignancy, severe liver insufficiency, hemorrhagic diathesis, amyloidosis, or vasculitis were excluded. Baseline characteristics of the study population are shown in Table 1. The mean age of patients with ICH was 67 (± 12) years, and 63% of patients were men. Mean age of healthy controls was 53 (± 9) years, and 34% were men. The most frequent cerebrovascular risk factor was hypertension (97%) in patients. CRP and fibrinogen levels, WBC, and platelet count were significantly higher in patients with ICH as compared to controls. PT and APTT were significantly shorter in patients, whereas TT was marginally, but significantly longer in patients as compared to controls. D-dimer levels and FVIII activity were significantly higher in patients as compared to healthy controls. Stroke characteristics and outcomes of enrolled patients are shown in Table 2. Median NIHSS on admission was 13 (IQR: 6–19). Mortality was 23% by day 14 after the event and 41% by day 90. In terms of long-term functional outcome, only 20% of the patients had a favorable outcome (mRS 0–1) by the end of the 3rd month. Median volume of hemorrhage was 12.0 (IQR: 3.5–29.8) cm³ on admission. The median ICH score was 0 (IQR: 0–2). After the event of stroke, most patients received antihypertensive (89%) and antacid therapy (94%). Low molecular weight heparin (LMWH) prophylaxis was administered in case of 26 high-risk patients (30%); all of these patients had high degree of disability or died (mRS 2–6) by the end of the 3rd month. Antiplatelet drugs were not used during the follow-up. Four cases of thrombotic events (1 patient with deep vein thrombosis and 3 patients with pulmonary embolism) were registered during the follow-up.

The result of the TGA measurements in patients and in controls is shown in Table 3. Peak thrombin was significantly higher in patients as compared to controls (397.2 ± 93.9 nM vs. 306.3 ± 85.3 nM, respectively, p < 0.0001). Time to peak parameter was significantly shorter in patients with ICH [median: 5.5 (IQR: 5.0–6.0) min in patients vs. 6.5 (IQR: 5.6–7.8) min in controls, p < 0.0001]. Lag time and ETP did not differ between groups. Lag time, ETP, and time to peak parameters showed a significant positive correlation with CRP in patients (Figure 1) and in controls (Supplementary Tables 1, 2). Lag time, ETP, and time to peak parameters showed a similar significant positive correlation with fibrinogen in patients, but not in controls (Supplementary Tables 1, 2). In patients, the peak thrombin parameter showed a significant positive correlation with white blood cell count (r = 0.2385, 95% CI: 0.0293–0.4277, p = 0.0261) and a significant negative correlation with PT (r = −0.3080, 95% CI: −0.4871– −0.1041, p = 0.0037) and APTT (r = −0.4277, 95% CI:−0.5856– −0.2385, p < 0.0001) (Supplementary Table 1). D-dimer and FVIII activity did not show significant correlation with any of the TGA parameters in patients and in controls (Supplementary Tables 1, 2).

Stroke severity and short-term outcomes showed no association with TGA parameters (data not shown). Poor long-term outcome was significantly associated with a higher BMI, higher D-dimer levels, higher NIHSS on admission, larger estimated hemorrhage volume at any measured time-point, higher ICH score, higher peak thrombin parameter, and being a non-smoker (Table 4). In patients with worse long-term functional outcomes, peak thrombin was significantly higher as compared to those with favorable outcomes [mRS 2–6 median: 402.5 (IQR: 344.8–473.8) vs. mRS 0–1: 326.4 (294.2–416.1) nM, p = 0.0096] (Table 4). ROC analysis was performed to investigate the diagnostic performance and the optimal threshold of the peak thrombin parameter for predicting long-term outcomes (Figure 2). Based on the statistically optimal threshold of 339.1 nM peak thrombin, the sensitivity and specificity of peak thrombin parameter to determine mRS 2–6 as outcome were 80.8 and 64.7%, respectively. Using this threshold, poor outcome can be ruled out with a high negative value (NPV: 0.9016, 95% CI: 0.8016–0.9541; PPV: 0.4583, 95% CI: 0.2789–0.6493). In a binary logistic regression model including age, sex, BMI, smoking status, NIHSS on admission, D-dimer (>0.5 mg/L), and peak thrombin (>339.1 nM), only NIHSS and the peak thrombin parameters remained in the model as significant, independent predictors of poor outcome (OR: 1.458, 95% CI: 1.180–1.801, p < 0.001 per unit change in NIHSS and OR: 48.292, 95% CI: 4.667–499.720, p = 0.001, respectively). Lag time and time to peak showed a modest, significant negative correlation with intracerebral bleeding volume on admission (r = −0.2603, 95% CI: −0.4642– −0.0303, p = 0.0231 and r = −0.3698, 95% CI: −0.5542– −0.1509, p = 0.0010, respectively) (Table 5). During the follow-up of patients, hemorrhage volumes on day 90 showed significant positive association with the ETP and peak thrombin parameters (r = 0.3838, 95% CI: 0.0161–0.6600, p = 0.0363 and r = 0.5383, 95% CI: 0.2103–0.7574, p = 0.0021, respectively). FVIII activity showed no correlation with hemorrhage volumes (data not shown), whereas D-dimer levels correlated with day 14 estimated hemorrhage volumes only (r = 0.3504, 95% CI: 0.0860–0.5689, p = 0.0087). Mortality by days 14 and 90 was not associated with TG parameters, FVIII activity, or D-dimer levels in this cohort (data not shown). The peak thrombin parameter on admission was above the 339.1 nM threshold in all patients with VTE events during the follow-up (data not shown).

To our knowledge, this study shows for the first time that the TGA is associated with the evolution of hematoma in patients and it might be a promising tool to predict the outcome of ICH. Despite the clear benefit of diagnostic tests with acceptable predictive value regarding outcomes in patients with acute ICH stroke, surprisingly, few studies are available on this topic. Here, we show that the most relevant TGA parameters (e.g., peak thrombin and ETP) of patients on admission correlate with the estimated size of hematoma at 90 days post-event, which is an important parameter of outcome. According to our results, an elevated peak thrombin is associated with worse long-term functional outcome in patients, and this parameter proved to be a significant, independent predictor of long-term outcomes. Although one might expect a lower extent of thrombin generation in patients with ICH with hematoma growth and poor outcomes, these results are in line with previous studies and our current study revealing that an elevated D-dimer, indicating more extensive coagulation and breakdown of clots, is associated with poor outcomes in patients with ICH (6–8). Our study indicates that the TGA might be even more relevant in the prognosis of ICH as D-dimer, as in the backward logistic regression analysis including all relevant variables, only peak thrombin and NIHSS remained in the model as independent predictors of poor outcome (mRS 2–6). Our results clearly show a potential distortion of hemostatic balance in patients with ICH toward a hypercoagulable state. When comparing thrombin generation in patients and healthy controls, peak thrombin parameter was significantly higher in patients. Correlation between thrombin generation parameters and baseline laboratory characteristics revealed that the difference might be explained by the presence of higher inflammatory parameters in patients. As the major cause of ICH was assumed to be hypertension in all patients in this cohort, it can be surmised that a low-grade, chronic inflammation exerted by an enhanced atherosclerotic process might induce a higher extent of thrombin generation in this cohort. Although fibrinogen, an acute phase protein, was positively associated with most thrombin generation parameters in patients, fibrinogen levels were within the normal range in this cohort and showed no significant difference between groups of different outcomes, implying that fibrinogen itself might not be directly involved in this process. Interestingly, PT and APTT showed a moderately strong negative correlation with peak thrombin, implying that factors resulting in shortened PT and APTT might be associated with a higher peak of thrombin formation. Although FVIII was significantly higher in patients with ICH as compared to healthy controls, FVIII levels did not show a correlation with any TGA parameters. Further studies are required to reveal mechanisms contributing to the observed higher thrombin generation in patients with ICH.

Interestingly, the evolution of the intracerebral hematoma was associated with the extent of thrombin generation in this cohort. This information might serve as important knowledge in the future, as data on the factors that drive the enlargement or the dissolution of the bleeding are potentially important when designing future pharmacological therapies. As a first step, it is crucial to understand the underlying pathomechanism leading to poor outcomes in patients. Second, adequate diagnostic tools could serve as a prerequisite in such approaches.

Thrombin generation assay is a method that has been shown to be potentially useful to predict the outcomes in a wide range of pathologies where the hemostasis balance has been tilted (16–18). The assay has a list of benefits and limitations, and most importantly, optimal conditions of the test have not been clearly defined, as yet. An undisputable advantage of the assay is being a global test of clot formation, thus saving efforts to determine the levels of individual factors of hemostasis using laborious and time-consuming methods. TGA as a global test incorporates the multiple procoagulant and anticoagulant pathways in its final result, and therefore, it more likely represents the situation occurring in the in vivo environment. TGA is highly variable among individuals, and studies suggest that this may allow individualized treatment in patients with bleeding disorders or on anticoagulant therapy (21). TGA has been used to predict bleeding tendency in various clinical conditions. In hemophiliac patients and in hemophiliacs with inhibitors, its usefulness has been proved, and it has been a matter of debate whether the TGA might describe the risk of bleeding better than traditional tests (22–24). TGA, when performed preoperatively, was found to be useful in providing information predictive of blood loss after cardiac surgery (25). On the other hand, in patients with thrombotic disorders, the test is potentially useful to identify those at higher risk for recurrent thrombosis (21). Patients with stroke, particularly those with ICH, are at significant risk of venous thromboembolic events (VTEs) (21). In a recent prospective observational study, it has been shown that after adjusting for confounders, patients with hemorrhagic stroke have significantly higher risk of in-hospital VTE than patients with acute ischemic stroke (1). In this studied cohort, 4 VTE events were registered, despite the use of prophylactic anticoagulation in high-risk patients. As the exact underlying pathophysiology of the increased VTE risk in patients with ICH has not been clarified, our results might provide additional data to understand more about the hypercoagulable state in this patient population. Additional studies are warranted to find out whether the TGA or other hemostasis assays might provide help in the VTE risk assessment and the antithrombotic management of patients with ICH.

It must be noted that a major factor limiting the use of TGA in the clinical practice currently is the fact that despite the attempts to standardize the test, analytical challenges remain to be elaborated. In future studies, assay conditions might be further improved and standardized, allowing direct comparison between laboratories. The ultimate goal in the development and standardization of the TGA will be to generate standardized assay conditions that lead to highly sensitive and specific tests that aid clinical decision-making, while allowing interlaboratory comparison of larger datasets. Clinical studies testing the utility of various hemostasis tests in predicting the outcome of intracerebral hemorrhage are scarcely found in the literature as yet. Our study, similarly to a few previous studies testing hemostasis abnormalities in patient cohorts, is a hypothesis-generating study to obtain insights into whether TGA is associated with poor outcomes in patients with ICH (26–28). Further long-term follow-up studies are warranted to verify our results and optimize assay conditions.

In patients with ICH, TG was increased as compared to healthy controls, which might be explained by the presence of higher inflammatory parameters in patients. Peak thrombin measured on admission might be a useful tool to predict outcomes in patients with ICH.

The results of this study should be interpreted in the context of its limitations and strengths. The sample size is limited; however, as compared to other published prospective studies of consecutive patients with non-traumatic, spontaneous ICH, involving the measurement of hemostasis biomarkers from admission samples, it is among the largest studies as yet. Nevertheless, the results presented here must be confirmed and validated by larger studies. The study was single-centered, which contributed to the limited sample size, but it had the advantages of uniform sample handling and patient care, and the major benefit that only one patient was lost to follow-up. As the TGA has no reference ranges provided by the distributor as yet, TGA results of healthy controls were included in this study. Despite our best efforts, based on the criteria of healthy control group, the age and sex of patients and controls were significantly different, which limits the direct comparison of results between groups. Assessment of TGA with different reagents (e.g., with low tissue factor concentration) was out of the scope of this study. The study was not designed to find out whether the TGA or other hemostasis assays might provide help in the assessment of VTE risk and the personalized antithrombotic management of patients with ICH; therefore, additional studies are warranted to focus on this aspect.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Institutional Ethics Committee of the University of Debrecen and the Ethics Committee of the National Medical Research Council. The patients/participants provided their written informed consent to participate in this study.

LL and RO-K collected clinical samples, performed experiments, analyzed, and interpreted the data. TÁ, IF, KF, MH, and JT collected clinical data and analyzed and interpreted the data. LC designed the research and interpreted the data. ZB designed the research, analyzed and interpreted the data, and wrote the manuscript. All authors have read and approved the final manuscript.

This work was supported by grants from the National Research, Development and Innovation Fund (K120042 and FK128582), by GINOP-2.3.2-15-2016-00043, the Lóránd Eötvös Research Network (ELKH-DE Cerebrovascular Research Group), and by ÚNKP 20-3-I-DE-220.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.