Beyeler, Morin (2025). Use of biomarkers in ischemic strokes : the particular case of cancer-related strokes. (Thesis). Universität Bern, Bern

Preview

Text 25beyeler_m.pdf - Thesis Available under License Creative Commons: Attribution (CC-BY 4.0). Download (42MB) | Preview

Abstract

Background: Around 5–10% of all patients with acute ischemic stroke (AIS) have underlying active cancer at the time of their stroke and constitute the group of “cancer-related strokes”. Cancer patients are known to have a higher risk of first and recurrent AIS, more severe stroke, and increased mortality after AIS. Although cancer-related strokes are thought to be caused by paraneoplastic coagulopathy, the complex relationship between cancer and AIS remains difficult to understand due to the multitude of cancer types and differing stages of cancer progression at the time of AIS. Consequently, it is essential to improve our knowledge of cancer-related strokes and the management of the affected patients. The general aim of this PhD thesis was to elucidate the description of the incidence, characteristics and outcomes of cancer-related strokes. Methods: The analyses that make up this PhD thesis are grouped into three subchapters corresponding to its three specific aims: The first was to better understand the characteristics of known biomarkers for cancer-related strokes and to identify new and, if possible, specific biomarkers of cancer-related strokes. Such biomarkers are essential for assessing the severity of paraneoplastic coagulation, identifying patients at risk of underlying undetected cancer at the time of AIS (occult cancer) and assessing the prognosis of AIS patients with underlying active cancer. The second aim was the investigation of AIS due to occult cancer. By proposing a clinical score to predict the presence of underlying occult cancer and investigating the outcomes for AIS patients with occult cancer, we sought to support the future implementation of a cancer-screening procedure in high-risk AIS patients. The third aim was to investigate factors influencing the outcomes of cancer-related strokes. In addition to assessing the influence of biomarkers associated with cancer-related strokes on long-term outcomes after AIS, we investigated which type of antithrombotic drug was associated with the best outcomes in patients with cancer-related stroke. To meet the aims of this PhD thesis, we created the Bernese Malignancy-in-Stroke (BMS) Database, which included cancer-specific data from all consecutive AIS patients treated between 2015 and 2020 at the Inselspital, University Hospital of Bern, Switzerland (N=5012). In parallel, we created the Bernese Transient Ischemic Attack (TIA) Database, which included data on all consecutive TIA patients treated at the Inselspital between 2015 and 2020 (N=1436). We used TIA patients without persisting brain damage as a comparison group in some of the studies presented in this thesis. Results: In this PhD thesis, we demonstrated that common cancer-associated biomarkers in AIS patients were also present in TIA patients with cancer. Compared to TIA patients without cancer, those with cancer were more likely to have a history of smoking (adjusted odds ratio [aOR] 2.77, 95% confidence interval [CI] 1.34–5.7), elevated D-dimer (aOR 1.77, 95% CI 1.26–2.49), elevated lactate dehydrogenase (aOR 1.003, 95% CI 1.00–1.005), lower hemoglobin (aOR 1.02, 95% CI 1.00–1.04), and lower leukocyte count (aOR 1.20, 95% CI 1.04–1.38). In AIS patients (independently of the presence of cancer), we demonstrated a time-dependent fluctuation in D-dimer levels. An early increase in D-dimer levels occurred within the first 6 hours after symptom onset (standardized beta coefficient [β] 0.728, 95% CI 0.324–1.121). Following this initial increase, the levels decreased (β −13.022, 95% CI −20.401 to −5.643) and a second increase was seen after 35 hours from symptom onset (β 11.750, 95% CI 4.71–18.791). No time-dependent fluctuation in D-dimer levels was observed in the control group of patients with TIA. In our study on the identification of new biomarkers for cancer-related strokes, we demonstrated an association between active cancer and the susceptibility vessel sign (SVS), a non-invasive, in situ representation of an occlusive thrombus, reflecting its microscopic composition. The absence of the SVS was associated with the presence of cancer in AIS patients (aOR 3.14, 95% CI 1.45–6.80). In another study, we demonstrated an association between the absence of a right-to-left cardiac shunt (encompassing patent foramen ovale and atrial septal defect) and active cancer (aOR 2.29, 95% CI 1.14–4.58). We were unable to demonstrate any association between increased left atrial volume index (≥35 mL/m2 on echocardiography), representing atrial cardiopathy, and the presence of cancer in patients with AIS (aOR 0.91, 95% CI 0.60–1.37). Based on our studies on occult cancers, we proposed the OCCULT-5 Score, which comprises five variables: age ≥ 77 years, embolic stroke of undetermined source, multi-territorial brain infarcts, D-dimer levels ≥ 820 μg/L, and female sex. A score of ≥ 3 predicted an underlying occult cancer in AIS patients, with a sensitivity of 64% and a specificity of 73%. Outcomes in patients with new cancer diagnosed immediately after the index AIS (during hospitalization) were no better than those with a cancer diagnosis obtained after discharge and within one year after the index AIS. In particular, there was no difference in long-term mortality between patient groups (adjusted hazard ratio [aHR] 1.16, 95% CI 0.53–2.52). Regarding factors influencing outcomes, in AIS patients treated with mechanical thrombectomy, we demonstrated an association between the absence of the SVS and long-term mortality (aHR 2.11, 95% CI 1.35–3.29) and poor functional outcome in the long term (aOR 2.90, 95% CI 1.29–6.55). Interaction analyses did not reveal any substantial influence of the presence of active cancer on these associations (p for interaction=0.79 and 0.71, respectively). Finally, there was no difference in outcomes between the two types of antithrombotic drug used for secondary prevention in AIS patients with active cancer. Anticoagulant and antiplatelet therapy were associated with similar risks of 1-year mortality (aHR 0.76, 95% CI 0.36–1.63) and long-term recurrent AIS (aHR 0.49, 95% CI 0.08–2.83). Discussion: The projects comprising this PhD thesis covered many aspects of the “cancer-related stroke” field. The absence of the SVS, reflecting the predominance of platelets and fibrin in the microscopic composition of the occluding thrombus, was associated with cancer-related strokes (and also independently associated with platelet- and fibrin-rich thrombi). Since the presence of the SVS, by contrast, was mainly associated with erythrocyte-rich thrombi and cardioembolic AIS, the assessment of SVS status (giving us a direct insight into the thrombus composition in situ) is helpful to assess the most likely underlying stroke etiology. An association between active cancer and arterial causes of AIS is presumed due to the negative association with right-to-left cardiac shunts (as a surrogate marker for paradoxical embolism). Furthermore, our studies showed a trend towards non-cardioembolic causes of AIS in patients with active cancer due to the association with absence of the SVS and the lack of association with atrial cardiopathy. As only 1.4% of AIS patients are diagnosed with a new cancer in the year following an AIS (and up to 6.2% following cryptogenic strokes), it is important to identify patients at risk of occult cancer to speed up the time to cancer diagnosis and potentially positively influence patient outcomes. To our knowledge, our OCCULT-5 Score is the only model that provides a clinical score that can be easily used at the bedside to evaluate the probability of underlying occult cancer in AIS patients. Despite the large size of our BMS, only 59 patients with occult cancer were included in the study, which failed to demonstrate a difference in outcomes between patients diagnosed with a new cancer during hospitalization versus after discharge. This underscores the need for multicenter studies that include larger numbers of AIS patients with active cancer, and particularly occult cancers, to address unanswered research questions. Finally, regarding secondary prevention, there are no conclusive guideline-based recommendations for the optimal antithrombotic strategy for patients with cancer-related strokes. Although our study (not published yet) on whether anticoagulants or antiplatelet drugs offered better secondary prevention has the advantage of including only patients with active cancer, it failed to demonstrate an advantage for either medication emphasizing the need for randomized clinical trials on this subject in the future. Outlook: The establishment of a multicenter registry of cancer-related strokes with detailed information on individual cancers will enable investigators to perform studies analyzing the different types of cancer on a case-by-case basis, and potentially to propose individualized treatments. In addition, prospective studies are needed, both to assess the risk of occult cancer in AIS patients and to evaluate new biomarkers for cancer-related strokes.

Item Type:	Thesis
Dissertation Type:	Cumulative
Date of Defense:	3 March 2025
Subjects:	600 Technology > 610 Medicine & health
Institute / Center:	04 Faculty of Medicine
Depositing User:	Sarah Stalder
Date Deposited:	23 Oct 2025 09:47
Last Modified:	23 Oct 2025 09:47
URI:	https://boristheses.unibe.ch/id/eprint/6822

Actions (login required)

View Item