Introduction

Mechanical aortic valve replacement remains a definitive therapeutic option for patients with severe aortic valve disease who require long‑term durability and structural reliability. Despite substantial advances in prosthetic valve engineering and perioperative management, mechanical valves are intrinsically chromogenic and necessitate lifelong anticoagulation therapy. Prosthetic valve thrombosis (PVT), although less frequent in the aortic position compared with the mitral valve, represents a serious and potentially catastrophic complication that may lead to acute valve obstruction, systemic embolization, heart failure, or sudden death. The clinical burden of PVT underscores the importance of accurately identifying patients at increased thrombotic risk, yet effective prediction remains a persistent challenge in contemporary cardiovascular practice. (1)

Long‑term anticoagulation with vitamin K antagonists, most commonly warfarin, constitutes the cornerstone of thrombosis prevention in patients with mechanical heart valves. Anticoagulation intensity is routinely monitored using the international normalized ratio (INR), with guideline‑recommended target ranges tailored to valve position and patient‑specific risk factors.

For most mechanical aortic valves, a target INR between 2.0 and 3.0 is considered sufficient. However, real‑world anticoagulation control is often suboptimal due to dietary variability, drug interactions, genetic polymorphisms affecting warfarin metabolism, and inconsistent monitoring. Importantly, prosthetic valve thrombosis has been documented even among patients with apparently therapeutic INR levels, highlighting the limitations of INR as a sole surrogate for thrombotic risk. (2)

The pathophysiology of thrombosis in mechanical the aortic position compared with the mitral valve, represents a serious and potentially catastrophic complication that may lead to acute valve obstruction, systemic embolization, heart failure, or sudden death. The clinical burden of PVT underscores the importance of accurately identifying patients at increased thrombotic risk, yet effective prediction remains a persistent challenge in contemporary cardiovascular practice. Long‑term anticoagulation with vitamin K antagonists, most commonly warfarin, constitutes the cornerstone of thrombosis prevention in patients with mechanical heart valves. Anticoagulation intensity is routinely monitored using the international normalized ratio (INR), with guideline‑recommended target ranges tailored to valve position and patient‑specific risk factors. For most mechanical aortic valves, a target INR between 2.0 and 3.0 is considered sufficient. However, real‑world anticoagulation control is often suboptimal due to dietary variability, drug interactions, genetic polymorphisms affecting warfarin metabolism, and inconsistent monitoring. Importantly, prosthetic valve thrombosis has been documented even among patients with apparently therapeutic INR levels, highlighting the limitations of INR as a sole surrogate for thrombotic risk. The pathophysiology of thrombosis in mechanical aortic valves is fundamentally distinct from that observed in native cardiac structures or atrial fibrillation. Thrombus formation in this setting is driven by direct blood contact with non‑endothelial zed artificial surfaces, abnormal shear stress, flow turbulence, and localized areas of flow stagnation, particularly near hinge mechanisms. These factors promote platelet activation, fibrin deposition, and local coagulation cascade activation independent of systemic hypercoagulable states. Consequently, valve‑specific mechanical and hemodynamic determinants may play a more dominant role in PVT development than traditional demographic or clinical risk factors. (3)

In contrast, thromboembolic risk prediction in atrial fibrillation has been successfully standardized through the use of the CHA₂DS₂‑VASc score. This widely validated clinical tool incorporates congestive heart failure, hypertension, age, diabetes mellitus, prior stroke or transient ischemic attack, vascular disease, and female sex to estimate the risk of systemic embolism. Its simplicity and strong prognostic performance have led to widespread adoption in routine clinical practice. Given its success in atrial fibrillation, the CHA₂DS₂‑VASc score has been increasingly explored in other thrombotic conditions, including patients with prosthetic heart valves, despite limited pathophysiological justification for such extrapolation. (4)

Applying the CHA₂DS₂‑VASc score to patients with mechanical aortic valves assumes that systemic cardiovascular comorbidities exert a similar influence on thrombus formation across fundamentally different biological environments. However, bus formation in this setting is driven by direct blood contact with non‑endothelial zed artificial surfaces, abnormal shear stress, flow turbulence, and localized areas of flow stagnation, particularly near hinge mechanisms. These factors promote platelet activation, fibrin deposition, and local coagulation cascade activation independent of systemic hypercoagulable states. Consequently, valve‑specific mechanical and hemodynamic determinants may play a more dominant role in PVT development than traditional demographic or clinical risk factors. In contrast, thromboembolic risk prediction in atrial fibrillation has been successfully standardized through the use of the CHA₂DS₂‑VASc score. This widely validated clinical tool incorporates congestive heart failure, hypertension, age, diabetes mellitus, prior stroke or transient ischemic attack, vascular disease, and female sex to estimate the risk of systemic embolism. Its simplicity and strong prognostic performance have led to widespread adoption in routine clinical practice. Given its success in atrial fibrillation, the CHA₂DS₂‑VASc score has been increasingly explored in other thrombotic conditions, including patients with prosthetic heart valves, despite limited pathophysiological justification for such extrapolation. Applying the CHA₂DS₂‑VASc score to patients with mechanical aortic valves assumes that systemic cardiovascular comorbidities exert a similar influence on thrombus formation across fundamentally different biological environments. However, while atrial fibrillation–related thrombosis is primarily mediated by blood stasis and endothelial dysfunction within the left atrium, prosthetic valve thrombosis arises from localized interactions between circulating blood elements and an artificial surface under non‑physiological flow conditions. As a result, the relative contribution ofbus formation in this setting is driven by direct blood contact with non‑endothelial zed artificial surfaces, abnormal shear stress, flow turbulence, and localized areas of flow stagnation, particularly near hinge mechanisms. These factors promote platelet activation, fibrin deposition, and local coagulation cascade activation independent of systemic hypercoagulable states. Consequently, valve‑specific mechanical and hemodynamic determinants may play a more dominant role in PVT development than traditional demographic or clinical risk factors. (5)

In contrast, thromboembolic risk prediction in atrial fibrillation has been successfully standardized through the use of the CHA₂DS₂‑VASc score. This widely validated clinical tool incorporates congestive heart failure, hypertension, age, diabetes mellitus, prior stroke or transient ischemic attack, vascular disease, and female sex to estimate the risk of systemic embolism. Its simplicity and strong prognostic performance have led to widespread adoption in routine clinical practice. Given its success in atrial fibrillation, the CHA₂DS₂‑VASc score has been increasingly explored in other thrombotic conditions, including patients with prosthetic heart valves, despite limited pathophysiological justification for such extrapolation. (6)

Applying the CHA₂DS₂‑VASc score to patients with mechanical aortic valves assumes that systemic cardiovascular comorbidities exert a similar influence on thrombus formation across fundamentally different biological environments. However, while atrial fibrillation–related thrombosis is primarily mediated by blood stasis and endothelial dysfunction within the left atrium, prosthetic valve thrombosis arises from localized interactions between circulating blood elements and an artificial surface under non‑physiological flow conditions. As a result, the relative contribution of age, hypertension, diabetes, and sex to thrombosis risk may be attenuated or obscured by the overwhelming pro‑thrombotic stimulus of the mechanical prosthesis itself. (7)

The aortic valve position introduces additional complexity due to its unique hemodynamic environment. High transvalvular flow velocities reduce global blood stasis but generate substantial shear stress, which can paradoxically enhance platelet activation and von Willebrand factor degradation. Furthermore, micro‑regions of low shear and recirculation near valve hinges may serve as niduses for thrombus formation despite overall favorable flow conditions. These nuanced flow characteristics are not captured by conventional clinical risk scores, further questioning the relevance of CHA₂DS₂‑VASc in predicting aortic mechanical valve thrombosis. (7)

Existing evidence evaluating the association between CHA₂DS₂‑VASc score and prosthetic valve thrombosis is limited and inconsistent. Most available studies have focused on composite thromboembolic outcomes rather than confirmed valve thrombosis and frequently include heterogeneous valve types and positions. In several analyses, any apparent association between higher CHA₂DS₂‑VASc scores and thromboembolic events diminishes after adjustment for anticoagulation intensity or the presence of atrial fibrillation. These findings suggest that CHA₂DS₂‑VASc may reflect general vascular risk rather than valve‑specific chromogenic potential. (9)

From a clinical perspective, predicting prosthetic valve thrombosis is further complicated by the dynamic nature of anticoagulation control. Transient periods of sub therapeutic INR, even if brief, may substantially increase thrombotic risk, while inflammatory states, infections, renal dysfunction, or hormonal changes can transiently alter coagulation age, hypertension, diabetes, and sex to thrombosis risk may be attenuated or obscured by the overwhelming pro‑thrombotic stimulus of the mechanical prosthesis itself. (10)

The aortic valve position introduces additional complexity due to its unique hemodynamic environment. High transvalvular flow velocities reduce global blood stasis but generate substantial shear stress, which can paradoxically enhance platelet activation and von Willebrand factor degradation. Furthermore, micro‑regions of low shear and recirculation near valve hinges may serve as niduses for thrombus formation despite overall favorable flow conditions. These nuanced flow characteristics are not captured by conventional clinical risk scores, further questioning the relevance of CHA₂DS₂‑VASc in predicting aortic mechanical valve thrombosis. (11)

Existing evidence evaluating the association between CHA₂DS₂‑VASc score and prosthetic valve thrombosis is limited and inconsistent. Most available studies have focused on composite thromboembolic outcomes rather than confirmed valve thrombosis and frequently include heterogeneous valve types and positions. In several analyses, any apparent association between higher CHA₂DS₂‑VASc scores and thromboembolic events diminishes after adjustment for anticoagulation intensity or the presence of atrial fibrillation. These findings suggest that CHA₂DS₂‑VASc may reflect general vascular risk rather than valve‑specific chromogenic potential. (12)

From a clinical perspective, predicting prosthetic valve thrombosis is further complicated by the dynamic nature of anticoagulation control. Transient periods of sub therapeutic INR, even if brief, may substantially increase thrombotic risk, while inflammatory states, infections, renal dysfunction, or hormonal changes can transiently alter coagulation balance. Static clinical scores fail to incorporate these temporal fluctuations, limiting their predictive accuracy. Consequently, reliance on simplified scoring systems may lead to false reassurance or unnecessary intensification of therapy in mechanical valve recipients. (13)

Despite these limitations, the CHA₂DS₂‑VASc score continues to attract interest because of its familiarity and ease of use. Clinicians often seek practical tools that can be readily applied in busy clinical settings, and the concept of a universal thromboembolic risk score is appealing. However, emerging data increasingly suggest that mechanical valve thrombosis represents a distinct clinical entity requiring disease‑specific risk assessment strategies rather than adaptation of atrial fibrillation–based models. (14)

Patients undergoing mechanical aortic valve replacement represent a unique population characterized by relatively younger age, fewer systemic comorbidities, and prolonged exposure to anticoagulation therapy. Their long‑term outcomes depend not only on achieving target INR values but also on maintaining stable anticoagulation over time. Variability in INR control, rather than absolute INR values, has been increasingly recognized as a critical determinant of thrombotic and bleeding events. Risk stratification tools that fail to account for this longitudinal dimension provide an incomplete assessment of true thrombotic vulnerability. (15)

Clarifying the role of CHA₂DS₂‑VASc in this context carries important implications for the future of personalized valve care. Demonstrating limited or absent predictive value would support the development of valve‑specific risk models incorporating prosthesis design, flow dynamics, anticoagulation quality metrics such as time in therapeutic range, and emerging biomarkers of coagulation or inflammation. Such approaches align with the broader movement toward precision medicine in cardiovascular disease. (16)

The present study was therefore designed to evaluate the association between the CHA₂DS₂‑VASc score and prosthetic valve thrombosis in patients with mechanical aortic valve replacement within a real‑world clinical cohort. By examining clinical characteristics alongside anticoagulation parameters, this investigation aims to determine whether CHA₂DS₂‑VASc provides independent prognostic information beyond INR control. We hypothesize that the score does not independently predict aortic mechanical valve thrombosis, reinforcing the need for valve‑specific predictive frameworks. Ultimately, improving risk prediction for prosthetic valve thrombosis has the potential to reduce morbidity, optimize anticoagulation strategies, and guide more individualized patient management. As mechanical valve implantation continues to play a central role in the treatment of valvular heart disease worldwide, refining our understanding of thrombosis risk and the tools used to assess it remains a clinical and scientific priority.

Material and methods

Study Design and Setting: The present study was designed as a case–control investigation conducted at Shahid Madani Heart Center, a tertiary referral hospital affiliated with Tabriz University of Medical Sciences, Tabriz, Iran. Patient recruitment and data collection were performed over a five‑year period, from the beginning of 2017 to the end of 2021. The study population consisted of patients who had previously undergone mechanical mitral valve replacement and were followed regularly at the study center during the specified time frame.

Sampling

A census sampling strategy was employed, whereby all eligible patients meeting the predefined inclusion criteria during the study period were consecutively enrolled. Based on this approach, a total of 100 patients were included in the final analysis. Patients were categorized into two groups according to the presence or absence of confirmed prosthetic valve thrombosis, forming the case and control groups, respectively.

Inclusion and Exclusion Criteria: Eligible participants were adult patients aged 18 years or older who had undergone mechanical mitral valve replacement and case–control investigation conducted at Shahid Madani Heart Center, a tertiary referral hospital affiliated with Tabriz University of Medical Sciences, Tabriz, Iran. Patient recruitment and data collection were performed over a five‑year period, from the beginning of 2017 to the end of 2021. The study population consisted of patients who had previously undergone mechanical mitral valve replacement and were followed regularly at the study center during the specified time frame.

Study Procedures: Baseline demographic characteristics, including age and sex, were extracted from medical records. Clinical variables such as hypertension, diabetes mellitus, heart failure, vascular disease, history of stroke or transient ischemic attack, and atrial fibrillation were systematically recorded. These variables were used to calculate the CHA₂DS₂‑VASc score for each participant according to standard definitions. Echocardiographic parameters, including left ventricular ejection fraction and prosthetic valve function, were obtained from the most recent examinations prior to the diagnosis of thrombosis or the corresponding follow‑up date in controls.

Anticoagulation‑related data were carefully reviewed, including prescribed warfarin dose, recorded INR values, and INR status at the time of thrombosis diagnosis or last follow‑up. For patients in the case group, INR values closest to the time of confirmed prosthetic valve thrombosis were analyzed. In the control group, INR measurements obtained during routine follow‑up visits were used for comparison. All data were collected using a standardized data extraction form to minimize information bias.

The primary objective of the study was to evaluate the association between the CHA₂DS₂‑, heart failure, vascular disease, history of stroke or transient ischemic attack, and atrial fibrillation were systematically recorded. These variables were used to calculate the CHA₂DS₂‑VASc score for each participant according to standard definitions. Echocardiographic parameters, including left ventricular ejection fraction and prosthetic valve function, were obtained from the most recent examinations prior to the diagnosis of thrombosis or the corresponding follow‑up date in controls.

Statistical Analysis: Statistical analyses were performed using standard statistical software. Continuous variables were expressed as mean ± standard deviation or median with interquartile range, as appropriate, and were compared using the independent‑samples t‑test or Mann–Whitney U test. Categorical variables were presented as frequencies and percentages and were compared using the chi‑square test or Fisher’s exact test. Logistic regression analysis was used to identify independent predictors of prosthetic valve thrombosis, with results reported as odds ratios and 95% confidence intervals. A two‑sided p‑value of less than 0.05 was considered statistically significant.

Ethical Considerations: The study protocol was reviewed and approved by the Ethics Committee of Tabriz University of Medical Sciences (Ethics code: IR. TBZMED.1400.1256). This research was conducted in accordance with the principles of the Declaration of Helsinki. Patient confidentiality was strictly maintained, and all data were anonymized prior to analysis. The present study represents the results derived from Specific Objective No. 2 of the medical thesis, registered under Thesis Number: 67370.

Results

During a four‑year observational period, a total of 104 patients were diagnosed with prosthetic valve thrombosis. Detailed review of anticoagulation status revealed that 82 of these patients (78%) had international normalized ratio (INR) values below or outside the therapeutic range, reflecting inadequate adherence to regular warfarin therapy. Because sub therapeutic anticoagulation represents a well‑established confounding During a four‑year observational period, a total of 104 patients were diagnosed with prosthetic valve thrombosis. Detailed review of anticoagulation status revealed that 82 of these patients (78%) had international normalized ratio (INR) values below or outside the therapeutic range, reflecting inadequate adherence to regular warfarin therapy. Because sub therapeutic anticoagulation represents a well‑established confounding factor for prosthetic valve thrombosis, these patients were excluded from further analysis. Consequently, the study population consisted of 22 patients with confirmed prosthetic valve thrombosis who maintained INR values within the therapeutic range, along with 78 patients with mechanical prosthetic valves who had no clinical or imaging evidence of thrombosis and served as the control group.

The mean age of the overall study cohort was 54.93 ± 11.52 years. Patients in the thrombosis group had a mean age of 51.5 ± 14.68 years, while those without thrombosis had a mean age of 56 ± 10.31 years; this difference was not statistically significant (p=0.148) (Table 1‑4). Regarding sex distribution, 62 patients (62%) were female and 38 patients (38%) were male. Comparison between the thrombosis and non‑thrombosis groups demonstrated no significant difference in sex composition, indicating that gender was not associated with the occurrence of prosthetic valve thrombosis in this cohort (p=0.944).

Among the patients with mechanical prosthetic valves, 94% had a mechanical leaflet valve, and 6% had a mechanical monoleaflet valve. All patients experiencing thrombotic events had leaflet valves, but this difference was not statistically significant (p = 0.405). The left ventricular ejection fraction (EF) was similar between the two groups, with a mean EF of 44.5% in the thrombosis group and 45.18% in the non-thrombosis group, showing no significant difference (p=0.723) (table 1).

Regarding the time elapsed since valve surgery, patients with thrombosis had a longer mean postoperative duration (127 months) compared to those without thrombosis (100 months), but this was not statistically significant (p=0.149). The average warfarin dose was slightly higher in the thrombotic group (4.47 mg) versus the control group (4.33 mg), yet this difference was not statistically significant (p=0.447). Similarly, the mean INR values over the last three measurements were comparable between groups, with no significant variation (p = 0.620), indicating consistent anticoagulation management across the study population (table 1).

 Table 1. Clinical Characteristics of Patients with Prosthetic Valves

Analysis of thrombosis location demonstrated that the mitral valve was the most frequently affected site, accounting for half of all thrombotic events. Isolated tricuspid valve thrombosis represented more than one‑quarter of cases, while aortic valve involvement was less common. Concomitant thrombosis involving both the mitral and tricuspid valves was rare, observed in only a small minority of patients. When comparing left‑sided and right‑sided valve involvement, left‑sided prosthetic valve thrombosis (mitral with or without aortic involvement) was more prevalent overall than isolated right‑sided events. However, the relatively high proportion of tricuspid valve thrombosis highlights that right‑sided mechanical valves are also susceptible to thrombotic complications, emphasizing the need for vigilant anticoagulation and follow‑up regardless of valve position (figure 1).

 The overall mean CHA₂DS₂‑VASc score in the study population was 2.58 ± 1.65. Patients with prosthetic valve thrombosis had a higher mean score (2.68 ± 1.63) compared with those without thrombosis (2.23 ± 1.72); however, this difference did not reach statistical significance (p = 0.23) (Figure 2). These findings indicate that, despite a numerically higher CHA₂DS₂‑VASc score among patients who developed thrombosis, the score did not meaningfully discriminate between patients with and without thrombotic events in this cohort (figure 2).

The mean CHA₂DS₂‑VASc scores, analyzed according to prosthetic valve type, showed no significant differences between thrombotic and non‑thrombotic groups. Among patients with aortic mechanical valves, the mean score was 3 ± 2 in the non‑thrombosis group and 2 ± 2 in those with thrombosis (p=0.841). For mitral mechanical valves, both groups demonstrated similar mean scores of 3 ± 2 in non‑thrombotic patients and 3 ± 2 in thrombotic patients (p=0.204) (figure 3).

When the CHA₂DS₂‑VASc score was examined separately for aortic and mitral prostheses, no meaningful difference emerged between patients with and without thrombosis. In the aortic valve subgroup, the thrombotic cohort actually had a slightly lower mean score than those without thrombosis, but the difference was statistically no significant. This indicates that the score did not distinguish patients at higher thrombotic risk among those carrying an aortic mechanical valve. Similarly, in the mitral valve subgroup, mean CHA₂DS₂‑VASc scores were almost identical between individuals with thrombotic events and those without, again with no statistically significant separation. Collectively, these findings reinforce that the CHA₂DS₂‑VASc score traditionally validated for predicting embolic risk in atrial fibrillation—offers limited discriminatory value for identifying prosthetic valve thrombosis risk, regardless of whether the prosthesis is located in the aortic or mitral position.