Echocardiographically occult nonbacterial thrombotic endocarditis with tumor cell-containing valvular vegetations in metastatic breast cancer Authors: Kristina Charaya, Irina Sokolova, Elnara Abdullaeva, Aleksei Volovchenko, Tamerlan Erdniev, Nurudin Nurutdinov, Nana Gogiberidze, Dmitry Shchekochikhin, Petr Chomakhidze, Alexandra Bogdanova, Irina Ingisonis, Yulia Lerner, Alexander Zanozin, Yulia Andreeva, Denis Andreev, Maria Poltavskaya Published: 2026 ABSTRACT BackgroundNonbacterial thrombotic endocarditis (NBTE) is an uncommon but important cause of systemic arterial embolism, most often occurring in advanced malignancy. Diagnosis is frequently difficult because vegetations may be small, transient, or not visualized on echocardiography, which may delay recognition.Case presentationA 37-year-old woman presented with multifocal ischemic strokes and visceral infarctions. Initial transesophageal echocardiography (TEE) showed a small isoechoic structure on the mitral valve, interpreted as a possible vegetation. Despite empiric broad-spectrum antibiotics and therapeutic anticoagulation, she developed progressive thrombocytopenia, hemolytic anemia, disseminated intravascular coagulation, and multiorgan failure. Repeated blood cultures were negative, and repeat TEE showed no valvular masses. The patient died despite plasma exchange, plasma infusion, corticosteroids, and transfusion support.Postmortem examination revealed previously undiagnosed bilateral metastatic invasive lobular breast carcinoma. Histopathology demonstrated tumor cell–containing fibrin–platelet vegetations on the mitral and tricuspid valves, with fibrin-rich microthrombi in renal arterioles consistent with secondary paraneoplastic thrombotic microangiopathy.ConclusionThis autopsy-confirmed case underscores that malignancy-associated NBTE may remain echocardiographically occult despite extensive arterial embolization. It also demonstrates that valvular thrombi may incorporate tumor cells, providing a direct substrate for embolic events.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40959-026-00498-8. Background Nonbacterial thrombotic endocarditis (NBTE) is defined by sterile fibrin–platelet thrombi on cardiac valves and is most often associated with malignancy, autoimmune disease, or severe systemic inflammation. In contrast to infective endocarditis (IE), NBTE typically occurs without persistent bacteremia, fever, or destructive valvular lesions; however, it carries a high risk of systemic arterial embolization [1]. NBTE is frequently underdiagnosed, particularly when vegetations are small, transient, or not visualized on echocardiography [2]. Autopsy studies suggest that clinically unrecognized valvular thrombi may represent an important embolic source in cancer-related hypercoagulability [3]. Case presentation A 37-year-old woman without known cardiovascular or oncologic disease was admitted to our tertiary cardiology center on October 29, 2025, for evaluation of recurrent ischemic strokes with systemic arterial thromboembolism. The patient had been previously diagnosed with polycystic ovary syndrome and had used combined oral contraceptives for several years. Obstetric history included two pregnancies: a miscarriage in 2014 and a full-term pregnancy in 2015, during which prophylactic low-molecular-weight heparin was administered. According to the patient, she had been evaluated by a hematologist before pregnancy and was informed of a possible “thrombophilia,” but no documentation was available. Initial presentation Symptoms began in early September 2025 with episodic burning and constricting pain involving the neck, shoulders, and chest, with gradual escalation in frequency and intensity. She was initially treated as an outpatient with nonsteroidal anti-inflammatory drugs, pregabalin, and duloxetine with no improvement. On October 1, she developed acute left-sided weakness and was admitted to a tertiary neurological center. At initial neurological evaluation, the National Institutes of Health Stroke Scale (NIHSS) score was 12. Magnetic resonance imaging (MRI) of the brain demonstrated multiple acute and subacute ischemic lesions involving the terminal branches of both middle cerebral arteries, the right posterior cerebral artery, and the right posterior inferior cerebellar artery. Computed tomography (CT) angiography showed occlusion of the right internal carotid artery involving both extracranial and intracranial segments. Intravenous thrombolysis was not administered because the patient was outside the established 4.5-hour time window from symptom onset and therefore did not meet eligibility criteria for systemic thrombolysis [4]. Initial laboratory tests showed leukocytosis, moderate thrombocytopenia, and markedly elevated inflammatory markers. Body temperature and blood pressure remained normal throughout hospitalization. Transthoracic echocardiography revealed preserved left ventricular systolic function and mild mitral and tricuspid regurgitation. Transesophageal echocardiography (TEE) identified a small (4.7-mm) isoechoic structure on the atrial side of the posterior mitral leaflet. Given its subtle appearance and lack of independent motion, this finding was interpreted as a possible vegetation, although imaging artifact or a transient thrombotic deposit could not be excluded. Continuous electrocardiographic monitoring showed persistent sinus rhythm. Despite repeated microbiological blood cultures obtained, no bacteremia was detected. Contrast-enhanced CT of the chest and abdomen did not reveal a primary malignancy but demonstrated embolic complications, including splenic infarction and bilateral renal infarctions. Given the combination of multifocal cerebral ischemia, visceral infarctions, and a valvular echocardiographic finding, a working diagnosis of embolic stroke of presumed cardiac origin was made, and IE was considered, although fever was absent and blood cultures remained negative. Empiric broad-spectrum antibiotics and therapeutic anticoagulation with low-molecular-weight heparin (LMWH) were started given the presumed embolic mechanism and the possibility of malignancy-associated thrombosis. During this period, inflammatory markers partially decreased, but progressive cytopenias developed, including worsening anemia and thrombocytopenia with the appearance of immature myeloid forms and nucleated red blood cells in the peripheral smear. Admission to cardiology center After discharge, the patient was readmitted on October 29 for further cardiology workup. On admission, she complained of persistent generalized burning pain and profound weakness. Physical examination revealed pallor and petechial rash on the lower extremities and feet. A firm, painless 1-cm breast mass was palpated. Vital signs were stable, and no cardiac murmurs were noted. Repeat transthoracic and TEE demonstrated no valvular masses, no evidence of intracardiac shunting, and intact mitral and aortic valves (Fig. 1). The repeat TEE provided diagnostic image quality for evaluation of the cardiac valves, without major technical limitations (poor acoustic windows, significant shadowing, or non-diagnostic motion artifact). The TEE video is available in the Supplementary Material. Fig. 1Transesophageal echocardiography showing no visible mitral valve vegetation on repeat examination. Mid-esophageal TEE view demonstrates preserved mitral valve morphology without discrete vegetations or thrombotic masses. Figure 1. Repeat transesophageal echocardiography demonstrating preserved mitral valve morphology without discrete vegetations Transesophageal echocardiography showing no visible mitral valve vegetation on repeat examination. Mid-esophageal TEE view demonstrates preserved mitral valve morphology without discrete vegetations or thrombotic masses. Figure 1. Repeat transesophageal echocardiography demonstrating preserved mitral valve morphology without discrete vegetations By the time of readmission to our cardiology center, she had already completed approximately 4 weeks of antibiotic treatment. Given persistently negative blood cultures, absence of fever, and no valvular masses on repeat TEE, antibiotics were discontinued. On October 30, the patient developed new neurological symptoms, including diplopia and impaired visual focusing. Brain MRI revealed a new lacunar infarct in the head of the right caudate nucleus, in addition to multiple chronic ischemic lesions. Laboratory tests showed rapidly worsening thrombocytopenia, severe normocytic anemia, elevated lactate dehydrogenase, low haptoglobin, and schistocytes on peripheral blood smear, consistent with intravascular hemolysis (Table 1). The direct antiglobulin test (DAT) was negative, and complement levels (C3/C4) were within the reference range. Coagulation studies demonstrated hypofibrinogenemia, prolonged clotting times, and markedly elevated D-dimer levels, consistent with evolving disseminated intravascular coagulation (DIC). ADAMTS13 activity was normal, arguing against thrombotic thrombocytopenic purpura. Testing for antiphospholipid syndrome, autoimmune hemolytic anemia, atypical hemolytic uremic syndrome, and inherited thrombophilias was negative. Antithrombin activity and coagulation factors V and VIII were within the reference range at the time of testing. Table 1Longitudinal changes in laboratory parameters01 Oct 202521 Oct 202529 Oct 202505 Nov 202507 Nov 2025Сomplete blood count Hemoglobin, g/L138112857558 Erythrocytes, ×10¹²/L5.344.193.342.822.25 Platelets, ×10⁹/L123566758** Leukocytes, ×10⁹/L22.810.715.649.9518.51 Neutrophils, %70.835.535.440.938.2 Absolute neutrophil count, ×10⁹/L16.63.985.544.067.06 Absolute lymphocyte count, ×10⁹/L4.514.437.634.879.83 Lymphocytes, %19.84348.848.953.1 Myelocytes, %35NA53 Blast cells, %01000 Normoblasts, ×10⁹/L2.0416NANANABiochemical blood analysis Procalcitonin, ng/mL0.1460.2730.050.20NA CRP, mg/L964366.646.239.9 ALT, U/L80109180198204 AST, U/L75106206221373 Total protein, g/L68706667NA Albumin, g/L36.834.934.435.933.1 Total bilirubin, µmol/L12.211.816.646.1126.9 Direct bilirubin, µmol/L4.84.95.719.563.8 Creatinine, µmol/L110108119108167 LDH, U/LNANA130713472451 Haptoglobin, mg/dLNANA6NANAСoagulation profile INR1.311.121.611.91.85 PT, s14.312.217.821.221.5 aPTT, s24.3NA31.96> 180*> 180* Fibrinogen, g/LNA2.941.31.351.06 D-dimer, ng/mLNANA331024,18089,530ALT Alanine aminotransferase, aPTT Activated partial thromboplastin time, AST Aspartate aminotransferase, CRP C-reactive protein, INR International normalized ratio, LDH lactate dehydrogenase, PT Prothrombin time* Upper limit of the assay (reportable range)** Platelet count on 07 Nov 2025 was measured after platelet transfusion Table 1Longitudinal changes in laboratory parameters01 Oct 202521 Oct 202529 Oct 202505 Nov 202507 Nov 2025Сomplete blood count Hemoglobin, g/L138112857558 Erythrocytes, ×10¹²/L5.344.193.342.822.25 Platelets, ×10⁹/L123566758** Leukocytes, ×10⁹/L22.810.715.649.9518.51 Neutrophils, %70.835.535.440.938.2 Absolute neutrophil count, ×10⁹/L16.63.985.544.067.06 Absolute lymphocyte count, ×10⁹/L4.514.437.634.879.83 Lymphocytes, %19.84348.848.953.1 Myelocytes, %35NA53 Blast cells, %01000 Normoblasts, ×10⁹/L2.0416NANANABiochemical blood analysis Procalcitonin, ng/mL0.1460.2730.050.20NA CRP, mg/L964366.646.239.9 ALT, U/L80109180198204 AST, U/L75106206221373 Total protein, g/L68706667NA Albumin, g/L36.834.934.435.933.1 Total bilirubin, µmol/L12.211.816.646.1126.9 Direct bilirubin, µmol/L4.84.95.719.563.8 Creatinine, µmol/L110108119108167 LDH, U/LNANA130713472451 Haptoglobin, mg/dLNANA6NANAСoagulation profile INR1.311.121.611.91.85 PT, s14.312.217.821.221.5 aPTT, s24.3NA31.96> 180*> 180* Fibrinogen, g/LNA2.941.31.351.06 D-dimer, ng/mLNANA331024,18089,530ALT Alanine aminotransferase, aPTT Activated partial thromboplastin time, AST Aspartate aminotransferase, CRP C-reactive protein, INR International normalized ratio, LDH lactate dehydrogenase, PT Prothrombin time* Upper limit of the assay (reportable range)** Platelet count on 07 Nov 2025 was measured after platelet transfusion Longitudinal changes in laboratory parameters ALT Alanine aminotransferase, aPTT Activated partial thromboplastin time, AST Aspartate aminotransferase, CRP C-reactive protein, INR International normalized ratio, LDH lactate dehydrogenase, PT Prothrombin time * Upper limit of the assay (reportable range) ** Platelet count on 07 Nov 2025 was measured after platelet transfusion A formal 4Ts score was not documented; however, heparin-induced thrombocytopenia was considered unlikely based on the timing of thrombocytopenia relative to LMWH exposure. Peripheral blood flow cytometry did not detect a paroxysmal nocturnal hemoglobinuria clone. Bone marrow aspiration was attempted twice to exclude hematologic malignancy, given the presence of leukoerythroblastosis and cytopenias. Both procedures yielded markedly hypocellular samples without diagnostic evidence of leukemia or lymphoma. Trephine biopsy and breast tissue sampling were deferred due to critical thrombocytopenia, active bleeding, and DIC. Breast imaging revealed a small lesion (approximately 1 cm, BI-RADS 3) and axillary lymphadenopathy; the lesion was initially considered probably benign. Biopsy was deferred due to critical thrombocytopenia and bleeding risk. At this stage, the presence of progressive cytopenias, intravascular hemolysis, and widespread arterial thrombosis raised concern for a thrombotic microangiopathy (TMA) with predominant arterial involvement. Plasma exchange, plasma infusion, corticosteroids, and extensive transfusion support were initiated in accordance with treatment protocols for TMA. Despite these measures, the patient’s condition deteriorated rapidly, with progressive hemorrhagic complications, metabolic acidosis, renal failure, and refractory shock. She died on November 7, 2025. Postmortem findings The autopsy revealed a previously undiagnosed multicentric bilateral synchronous invasive lobular carcinoma in both breasts, classified as grade 3 and triple-negative, with a proliferation index exceeding 70% (Fig. 2). Given that a triple-negative profile is uncommon for classic invasive lobular carcinoma [5], lobular differentiation was supported by loss of E-cadherin expression (with GATA3 positivity), consistent with breast origin. Fig. 2Histopathology and immunophenotype of metastatic breast carcinoma (lobular phenotype) with a triple-negative profile. Representative autopsy sections show a carcinoma composed of discohesive tumor cells with an infiltrative growth pattern, consistent with lobular differentiation (A-В; H&E, ×200; scale bar 100μm.). Immunohistochemistry demonstrates loss of E-cadherin expression in tumor cells (С), supporting lobular phenotype. Tumor cells are negative for estrogen receptor (ER) and progesterone receptor (PR) (D-E) and show no HER2 overexpression/amplification (F), consistent with a triple-negative immunoprofile. The proliferation index is high (Ki-67 >70%) (G). GATA3 highlights tumor cell nuclei, while mammaglobin is detected in the cytoplasm and cell membranes. This immunophenotype supports breast lineage (H-I). Histopathology and immunophenotype of metastatic breast carcinoma (lobular phenotype) with a triple-negative profile. Representative autopsy sections show a carcinoma composed of discohesive tumor cells with an infiltrative growth pattern, consistent with lobular differentiation (A-В; H&E, ×200; scale bar 100μm.). Immunohistochemistry demonstrates loss of E-cadherin expression in tumor cells (С), supporting lobular phenotype. Tumor cells are negative for estrogen receptor (ER) and progesterone receptor (PR) (D-E) and show no HER2 overexpression/amplification (F), consistent with a triple-negative immunoprofile. The proliferation index is high (Ki-67 >70%) (G). GATA3 highlights tumor cell nuclei, while mammaglobin is detected in the cytoplasm and cell membranes. This immunophenotype supports breast lineage (H-I). The cancer had spread extensively, affecting the lymph nodes, ovaries, bone marrow, liver, lungs, spleen, brain, thyroid gland, and epicardial tissue. Gross examination of the heart demonstrated friable thrombotic deposits on the mitral valve apparatus (Fig. 3). Histopathology confirmed NBTE, showing fibrin–platelet vegetations on the mitral and tricuspid valves containing clusters of malignant cells (Fig. 4). The underlying valve tissue was preserved without ulceration or destructive inflammatory infiltrate, arguing against infective endocarditis and supporting NBTE. Immune-complex staining was not performed, as histological findings were characteristic of NBTE, with no evidence of inflammatory or immune-mediated valvular injury. Histopathologic examination of the aortic valve revealed no significant abnormalities and no evidence of NBTE vegetations (Fig. 5). Fibrin-rich microthrombi were identified in renal arterioles, consistent with secondary paraneoplastic TMA (Fig. 6). Additional findings included systemic arterial thromboembolism and severe DIC. Fig. 3Gross pathological examination of the heart: thrombotic deposits on the mitral valve apparatus. The mitral valve cusps are thin and translucent, with loose pink finely verrucous vegetations located along the perimeter. Histopathology of nonbacterial thrombotic endocarditis with tumor cell–containing valvular vegetations Gross pathological examination of the heart: thrombotic deposits on the mitral valve apparatus. The mitral valve cusps are thin and translucent, with loose pink finely verrucous vegetations located along the perimeter. Histopathology of nonbacterial thrombotic endocarditis with tumor cell–containing valvular vegetations Fig. 4Histopathology of nonbacterial thrombotic endocarditis with tumor cell–containing valvular vegetations. Representative autopsy sections of the mitral (A; H&E stain. Magnification: ×100. Scale bar: 200 μm.) and tricuspid (B; H&E stain. Magnification: ×200. Scale bar: 100 μm) valves show fibrin–platelet vegetations consistent with NBTE, with intralesional clusters of malignant cells embedded within the thrombus, which are extensively positive in IHC reactions with antibodies against PAN-CK (С; Magnification: ×200. Scale bar: 100 μm. D Magnification: ×400. Scale bar: 50 μm). The underlying valve tissue is preserved without ulceration or destructive inflammatory infiltration, arguing against infective endocarditis Histopathology of nonbacterial thrombotic endocarditis with tumor cell–containing valvular vegetations. Representative autopsy sections of the mitral (A; H&E stain. Magnification: ×100. Scale bar: 200 μm.) and tricuspid (B; H&E stain. Magnification: ×200. Scale bar: 100 μm) valves show fibrin–platelet vegetations consistent with NBTE, with intralesional clusters of malignant cells embedded within the thrombus, which are extensively positive in IHC reactions with antibodies against PAN-CK (С; Magnification: ×200. Scale bar: 100 μm. D Magnification: ×400. Scale bar: 50 μm). The underlying valve tissue is preserved without ulceration or destructive inflammatory infiltration, arguing against infective endocarditis Fig. 5Clinical timeline using relative time points (Day 0 = first presentation). LDH = lactate dehydrogenase; MRI = magnetic resonance imaging Clinical timeline using relative time points (Day 0 = first presentation). LDH = lactate dehydrogenase; MRI = magnetic resonance imaging Fig. 6Fibrin-rich microthrombi were identified in renal arterioles, consistent with secondary paraneoplastic TMA. Arteriolar wall thickening was observed, indicating residual effects of previous vasculitis, which is a manifestation of secondary TMA. Histologic images reveals fibrin thrombi in the lumina of glomerular capillary loops and afferent arterioles (А-В; H&E, ×400; scale bar 50-20μm), thickening of the arteriolar wall (С; H&E, ×400; scale bar 50μm) Fibrin-rich microthrombi were identified in renal arterioles, consistent with secondary paraneoplastic TMA. Arteriolar wall thickening was observed, indicating residual effects of previous vasculitis, which is a manifestation of secondary TMA. Histologic images reveals fibrin thrombi in the lumina of glomerular capillary loops and afferent arterioles (А-В; H&E, ×400; scale bar 50-20μm), thickening of the arteriolar wall (С; H&E, ×400; scale bar 50μm) Discussion This case provides autopsy-confirmed evidence that malignancy-associated NBTE may remain echocardiographically occult despite recurrent systemic arterial embolization. It also offers clinicopathological confirmation of tumor cell–containing valvular vegetations, supporting a direct link between advanced malignancy and embolic events. In the present case, the initial presentation closely mimicked cardioembolic stroke due to IE, a reasonable diagnostic assumption given the embolic pattern of infarctions and the initial echocardiographic findings. Clinically, the patient fulfilled several criteria that commonly trigger evaluation for IE: multifocal embolic strokes, visceral infarctions, and an echocardiographic valvular mass [6]. However, IE was considered unlikely given persistently negative blood cultures, absence of sustained fever during most of the disease course, and lack of progressive valvular destruction on serial imaging [7]. Instead, the overall course was more consistent with a malignancy-driven hypercoagulable state. Alternative explanations, including imaging artifact or a transient thrombotic deposit, were considered. Accordingly, upon admission to the cardiology department, antibiotics were discontinued after an approximately 4-week course. The absence of detectable vegetations on repeat imaging highlights the limited sensitivity of echocardiography in NBTE. In this case, image quality was sufficient for reliable assessment of the cardiac valves, suggesting that the negative findings were not due to technical limitations. This likely reflects the intrinsic properties of NBTE vegetations, which may be small, friable, transient, or may embolize between examinations [3, 5]. Autopsy series have demonstrated that clinically unrecognized valvular thrombi are common in patients with malignancy and may nonetheless serve as clinically significant embolic sources [3]. Autopsy studies have shown that valvular thrombi may be clinically unrecognized in patients with malignancy and still serve as a source of systemic emboli. These observations underscore that negative echocardiographic findings do not exclude clinically relevant valvular pathology in the appropriate clinical context. Interpretation of small or equivocal echocardiographic findings remains challenging. Overinterpretation of minor structures may lead to false-positive diagnoses, whereas excessive caution may delay recognition of clinically important lesions. In the present case, the initial finding of a small isoechoic structure on the mitral valve was interpreted cautiously, with consideration given to both vegetation and artifact. This approach reflects the need to integrate imaging findings with the overall clinical and laboratory picture rather than relying on echocardiography in isolation. Although TEE remains the primary imaging modality for suspected valvular pathology, additional imaging techniques are unlikely to substantially improve detection of NBTE in similar settings [8]. Limitations in spatial and temporal resolution, particularly for small or highly mobile lesions, reduce the diagnostic yield of modalities such as cardiac CT or MRI [5]. Moreover, their use may be restricted in critically ill patients [9]. As a result, the diagnosis of NBTE often remains clinical and is confirmed only at autopsy. Advanced techniques such as 3D TEE or shorter-interval repeat imaging might theoretically improve detection, but were not feasible in this case. Consistent with NBTE rather than IE, histology demonstrated sterile fibrin-platelet vegetations without valvular destruction or inflammatory infiltrate [1]. In contrast to most published clinical reports [3], our case provides clinicopathological confirmation of valvular vegetations incorporating tumor cell clusters, identified only at autopsy. It remains uncertain whether tumor cells actively contribute to valvular thrombogenesis or are passively incorporated into fibrin-platelet aggregates. However, their consistent intrathrombotic presence across multiple valves argues against a purely incidental finding. An additional feature of this case was the coexistence of widespread arterial thrombosis with rapidly progressive hemolytic anemia, thrombocytopenia, and consumptive coagulopathy requiring plasma exchange and transfusion support, consistent with malignancy-associated TMA. Unlike primary TMAs, cancer-related TMA is typically driven by uncontrolled prothrombotic and proinflammatory signaling and is frequently refractory to plasma exchange-based therapy [10]. In this context, the parallel development of microvascular thrombosis (TMA/DIC) and macroscopic valvular thrombi (NBTE) suggests a shared upstream mechanism rooted in extreme malignancy-driven hypercoagulability, rather than sequential or unrelated processes. LMWH was initiated during the initial hospitalization in accordance with recommendations for malignancy-associated thrombosis, where it is generally preferred over vitamin K antagonists [11]. However, further anticoagulation was limited by progressive thrombocytopenia, coagulopathy, and bleeding complications, which significantly narrowed the therapeutic window. Although NBTE is most commonly associated with mucin-producing adenocarcinomas, breast cancer is a less frequent underlying cause [13]. In this patient, extent of thrombotic involvement observed in this case illustrates the capacity of advanced malignancy to drive both systemic embolization and microvascular injury. DIC and NBTE often coexist in patients with malignancies [12]. In our case, they appear to represent different manifestations of the same underlying condition. The simultaneous involvement of large vessels, cardiac valves, and the microcirculation supports the presence of severe malignancy-associated hypercoagulability as a unifying mechanism. Although NBTE most commonly affects the mitral and aortic valves [1], involvement of both atrioventricular valves in this patient likely reflects the severity of the underlying hypercoagulable state. These findings support the possibility that cardiac valves can serve as sites of tumor-associated thrombogenesis rather than simply serving as passive recipients of circulating thrombi. The presence of tumor cells embedded within fibrin-platelet vegetations provides a plausible mechanistic link between systemic hypercoagulability, valvular thrombus formation, and recurrent arterial embolization, even in the absence of persistent echocardiographic abnormalities. Conclusion In summary, this case emphasizes that NBTE should be considered in patients with unexplained systemic arterial emboli, particularly in the absence of microbiological evidence of infection. It also highlights that negative echocardiographic findings do not exclude clinically significant valvular thrombi. The identification of tumor cell–containing vegetations further contributes to the understanding of malignancy-associated thrombosis and its potential mechanisms. Patient perspective Not available due to the patient’s death. Learning points Multifocal arterial emboli with persistently negative blood cultures should prompt early consideration of NBTE and evaluation for occult malignancy;Repeated negative TEE does not exclude NBTE in patients with ongoing embolic phenomena;Tumor cell–containing valvular vegetations can serve as a direct embolic source;Overlapping macrovascular (NBTE, arterial emboli) and microvascular (TMA, DIC) thrombotic manifestations may reflect a shared upstream oncologic driver and are often refractory to standard plasma exchange-based therapies. Multifocal arterial emboli with persistently negative blood cultures should prompt early consideration of NBTE and evaluation for occult malignancy; Multifocal arterial emboli with persistently negative blood cultures should prompt early consideration of NBTE and evaluation for occult malignancy; Repeated negative TEE does not exclude NBTE in patients with ongoing embolic phenomena; Repeated negative TEE does not exclude NBTE in patients with ongoing embolic phenomena; Tumor cell–containing valvular vegetations can serve as a direct embolic source; Tumor cell–containing valvular vegetations can serve as a direct embolic source; Overlapping macrovascular (NBTE, arterial emboli) and microvascular (TMA, DIC) thrombotic manifestations may reflect a shared upstream oncologic driver and are often refractory to standard plasma exchange-based therapies. Overlapping macrovascular (NBTE, arterial emboli) and microvascular (TMA, DIC) thrombotic manifestations may reflect a shared upstream oncologic driver and are often refractory to standard plasma exchange-based therapies. Supplementary Information Supplementary Material 1. The transesophageal echocardiography (TEE) video. Supplementary Material 1. The transesophageal echocardiography (TEE) video.