Mechanical Thrombectomy for Acute Vertebrobasilar Occlusion Associated with Mycoplasma pneumoniae Pneumonia in a Child: A Case Report

Abstract

Objective

Due to its rarity, high-quality evidence regarding mechanical thrombectomy for pediatric acute ischemic stroke remains limited. Mycoplasma pneumoniae is a common cause of community-acquired pneumonia in children, and a recognized but uncommon cause of pediatric ischemic stroke. Reports describing large-vessel occlusion associated with M. pneumoniae infection are rare, and the role of mechanical thrombectomy in this setting remains unclear.

Case Presentation

A 5-year-old boy hospitalized for pneumonia developed a sudden onset of impaired consciousness and quadriplegia. Brain MRI demonstrated acute infarctions involving the brainstem and left cerebellar hemisphere, and MRA revealed vertebrobasilar artery occlusion. Mechanical thrombectomy using contact aspiration and a combined technique resulted in successful reperfusion (modified Thrombolysis in Cerebral Infarction score 2b) with retrieval of red thrombi. Subsequent investigations confirmed M. pneumoniae pneumonia, and an extensive evaluation failed to identify an alternative embolic source. These findings, together with the coagulation abnormalities observed during the acute phase, suggested that the ischemic stroke was most likely caused by thromboembolism associated with M. pneumoniae infection. Despite requiring decompressive craniectomy and intensive care, the patient achieved a favorable neurological outcome, with the modified Rankin Scale score improving from 3 at discharge to 1 at the 3-month follow-up.

Conclusion

This case highlights the feasibility and potential efficacy of mechanical thrombectomy in selected pediatric patients with acute ischemic stroke, even in the setting of rare etiologies such as M. pneumoniae infection.

Introduction

Pediatric arterial ischemic stroke is a rare condition, with an incidence of approximately 1–2 per 100000 children per year.¹⁾ Large-vessel occlusion (LVO), including basilar artery occlusion (BAO), represents an even less common subset of pediatric stroke, but is associated with substantial morbidity and mortality.²⁾ Due to its rarity, high-quality evidence regarding mechanical thrombectomy for LVO in pediatric patients remains limited. Nevertheless, recent observational studies and international registries, including the Save ChildS, Save ChildS Pro, and KidClot studies, have demonstrated the feasibility and potential benefit of mechanical thrombectomy in selected pediatric patients.^3–5)

Mycoplasma pneumoniae (M. pneumoniae) is a common cause of community-acquired pneumonia in pediatric patients, and a recognized but uncommon cause of pediatric ischemic stroke.⁶⁾ Reports describing vertebrobasilar artery occlusion associated with M. pneumoniae infection are rare, and the role of mechanical thrombectomy in this setting remains unclear. Here, we report a pediatric patient with vertebrobasilar artery occlusion associated with M. pneumoniae pneumonia who was successfully treated with mechanical thrombectomy.

Case Presentation

A 5-year-old boy who had contracted influenza B infection 2 weeks earlier was admitted to our hospital with persistent fever, cough, and loss of appetite lasting for 1 week. Laboratory examinations revealed a white blood cell count of 8760/μL (neutrophils 82.3%) and a C-reactive protein level of 3.29 mg/dL. Chest radiography demonstrated decreased radiolucency in the left lung field, based on which he was diagnosed with acute pneumonia and intravenous ampicillin therapy was initiated. On hospital day 2, he developed impaired consciousness, with a Japan Coma Scale score of III-100 and a Glasgow Coma Scale score of 7 (E1V1M5), together with quadriplegia and sensory disturbances in all four extremities (PedNIHSS score, 37). Brain MRI demonstrated acute ischemic infarctions involving the pons, medulla oblongata, and left cerebellar hemisphere (PC-ASPECTS 7) (Fig. 1A and 1B). MRA revealed occlusion of the vertebral arteries bilaterally and the proximal basilar artery (Fig. 1C). The distal basilar artery was visualized via collateral flow from the right posterior communicating artery.

Fig. 1. Preoperative MRI and MRA. Diffusion-weighted imaging demonstrated acute ischemic infarctions involving the medulla oblongata (A, arrow), pons (B, arrows), and left cerebellar hemisphere (A and B, arrowheads). MRA demonstrated occlusion of the vertebral arteries bilaterally and the proximal basilar artery (C). The distal basilar artery was visualized via collateral flow from the right posterior communicating artery (C, arrows).

After obtaining informed consent from his parents, mechanical thrombectomy was subsequently performed under general anesthesia. A 5-Fr FUBUKI Dilator Kit (Asahi Intecc, Aichi, Japan) was advanced into the left vertebral artery via a right transfemoral approach. Intraoperative left vertebral angiography demonstrated abrupt occlusion of the intracranial segment of the left vertebral artery (Fig. 2A). Two passes of contact aspiration using a 5-Fr SOFIA catheter (Terumo Neuro, Aliso Viejo, CA, USA) restored flow in the proximal basilar artery (Fig. 2B and 2C). Subsequently, a combined technique using a Solitaire X 3 × 20-mm stent retriever (Medtronic, Irvine, CA, USA) and a 5-Fr SOFIA catheter achieved recanalization of the distal basilar artery and bilateral posterior cerebral arteries, although the left posterior inferior cerebellar artery remained occluded (Fig. 2D and 2E). Additional angiography of the right vertebral artery confirmed recanalization of the occluded intracranial segment of the right vertebral artery, resulting in successful reperfusion with a modified Thrombolysis in Cerebral Infarction score of 2b (Fig. 2F). No residual stenosis, vessel wall irregularity, or other angiographic abnormalities were observed in the vertebral arteries bilaterally or the basilar artery. Red thrombi were retrieved both from within the stent retriever and through the aspiration system (Fig. 2G).

Fig. 2. Mechanical thrombectomy for vertebrobasilar occlusion. Left vertebral angiography demonstrated occlusion of the intracranial segment of the left vertebral artery (A, arrow). Contact aspiration using a 5-Fr SOFIA catheter (B, arrow) achieved recanalization of the proximal basilar artery (C). A combined technique using a Solitaire X 3 × 20-mm stent retriever and a 5-Fr SOFIA catheter achieved recanalization of the distal basilar artery and posterior cerebral arteries bilaterally (D, E). In panel (D), the arrowhead indicates the distal marker of the Solitaire X stent retriever, and the arrow indicates the SOFIA catheter. Although successful reperfusion was achieved, the left posterior inferior cerebellar artery remained occluded (E). Right vertebral angiography confirmed recanalization of the intracranial segment of the right vertebral artery (F). Red thrombi were retrieved during mechanical thrombectomy (G).

SOFIA, Terumo Neuro, Aliso Viejo, CA, USA; Solitaire X, Medtronic, Irvine, CA, USA

Immediate postoperative brain CT demonstrated no hemorrhagic complications, but revealed worsening cerebellar edema associated with left cerebellar infarction, for which an emergency decompressive craniectomy was performed (Fig. 3A). A chest CT performed at this time demonstrated bilateral diffuse ground-glass opacities and consolidative infiltrates (Fig. 3B). A BioFire FilmArray Pneumonia Panel (bioMérieux Japan, Tokyo, Japan) detected Haemophilus influenzae (H. influenzae) and M. pneumoniae, and the antibody titers for M. pneumoniae, measured by the particle agglutination assay on the day after stroke onset, were markedly elevated at 1:5120. Although a fourfold rise in antibody titer could not be demonstrated by paired serology, this markedly elevated titer strongly supported an acute M. pneumoniae infection. Based on these findings, the patient was diagnosed with pneumonia due to coinfection with M. pneumoniae and H. influenzae. Coagulation studies demonstrated elevated D-dimer levels (8.9 μg/mL) and reduced fibrinogen levels (114 mg/dL). His prothrombin time–international normalized ratio (PT-INR) was mildly prolonged at 1.28, whereas the activated partial thromboplastin time remained within the normal range (28.5 seconds). Further thrombophilia testing, performed before the initiation of intravenous heparin therapy, revealed a positive lupus anticoagulant by the diluted Russell viper venom time assay, although no other major thrombophilic abnormalities, including anticardiolipin antibodies and anti-β2-glycoprotein I antibodies, were identified. Extensive investigations, including transthoracic echocardiography, lower-extremity venous ultrasonography, and contrast-enhanced CT, revealed no identifiable embolic source, such as an intracardiac thrombus, right-to-left shunt, or deep venous thrombosis, and no evidence of systemic embolism, including pulmonary embolism or splenic infarction. These findings further support the hypothesis that transient hypercoagulability associated with M. pneumoniae infection was a plausible mechanism of cerebral embolism.

Fig. 3. Postoperative cranial CT after decompressive craniectomy and chest CT findings. Cranial CT obtained after decompressive craniectomy demonstrated persistent edema associated with left cerebellar infarction despite posterior fossa decompression (A). Chest CT demonstrated bilateral multifocal ground-glass opacities and nodular infiltrates in both lung fields, consistent with severe pneumonia (B).

The patient underwent intensive care management, including mechanical ventilation, antibiotic therapy (cefotaxime plus azithromycin), and continuous intravenous heparin infusion. Subsequently, the pneumonia gradually improved, and mechanical ventilation was discontinued on hospital day 7. No recurrent ischemic stroke occurred during hospitalization, and antithrombotic therapy was transitioned from intravenous heparin to oral aspirin (100 mg/day), as transient hypercoagulability associated with M. pneumoniae infection was considered the most likely mechanism of cerebral embolism and coagulation abnormalities had resolved following improvement of the pneumonia. The patient’s neurological deficits gradually improved, and he was discharged home on hospital day 34 with a modified Rankin Scale (mRS) score of 3, although limb and truncal ataxia persisted at discharge. Coagulation abnormalities, including prolonged PT-INR, elevated D-dimer, and reduced fibrinogen levels, had normalized by the time of discharge, although the lupus anticoagulant remained weakly positive at the 3-month follow-up. At the 3-month follow-up, his neurological function had further improved, and his mRS score had improved to 1.

Discussion

We herein report a rare case of acute vertebrobasilar occlusion associated with M. pneumoniae pneumonia in a 5-year-old boy who was successfully treated with mechanical thrombectomy, resulting in a favorable neurological outcome despite severe initial neurological deficits. This case provides additional evidence supporting the feasibility of endovascular treatment in carefully selected pediatric patients with LVO.

In the present case, neurological symptoms developed during the acute phase of pneumonia caused by M. pneumoniae and H. influenzae, whereas the preceding influenza B infection had clinically improved approximately 2 weeks earlier. Therefore, influenza B was considered less likely to have been the principal trigger of the ischemic stroke. Previous studies have associated M. pneumoniae infection with pediatric arterial ischemic stroke through mechanisms including transient hypercoagulability, antiphospholipid antibody production, immune-mediated vascular injury, and cerebral vasculitis.^7–11) In contrast, evidence directly linking respiratory H. influenzae infection to arterial ischemic stroke is limited. Moreover, transient antiphospholipid antibodies, including lupus anticoagulant, have been reported following M. pneumoniae infection,^9,10) whereas similar findings have not been well documented in respiratory H. influenzae infection. Accordingly, the transient hypercoagulable state and lupus anticoagulant positivity observed in our patient were considered more compatible with previously reported M. pneumoniae–associated thrombotic mechanisms. We therefore considered M. pneumoniae to be the leading suspected contributor to the cerebral embolism in the present case. Although the lupus anticoagulant remained weakly positive approximately 3 months after stroke onset, anticardiolipin antibodies and anti-β2-glycoprotein I antibodies were negative, and coagulation abnormalities had resolved. These findings were considered more consistent with transient infection-associated lupus anticoagulant than with antiphospholipid syndrome.^9,10) Previous pediatric reports have shown that infection-associated lupus anticoagulant is usually transient and may resolve spontaneously within several weeks to approximately 6 months after resolution of the infection.^12,13) Therefore, continued follow-up is warranted to determine whether lupus anticoagulant resolves spontaneously over time.

The evidence supporting mechanical thrombectomy for pediatric acute ischemic stroke remains limited because randomized controlled trials have not been conducted in pediatric patients. Current evidence is derived mainly from retrospective cohort studies, registry data, and meta-analyses. The Save ChildS study, the largest international multicenter retrospective cohort of pediatric patients undergoing thrombectomy, demonstrated high rates of successful reperfusion and favorable functional outcomes with a low rate of procedure-related complications.³⁾ More recently, the Save ChildS Pro registry, currently the only large prospective multicenter study in this field, showed that endovascular treatment was associated with better functional outcomes than best medical treatment alone, despite greater baseline stroke severity in patients who underwent thrombectomy.⁴⁾ These findings are consistent with those of the KidClot study, which also supported the feasibility and safety of recanalization therapies in pediatric acute ischemic stroke.⁵⁾ In addition, systematic reviews and meta-analyses have consistently reported high rates of successful recanalization and favorable neurological outcomes after thrombectomy in children.^14,15) Collectively, these findings suggest that mechanical thrombectomy might be a feasible and effective treatment option in selected children, although the current evidence remains largely based on non-randomized data. Reflecting on these accumulating data, the 2026 American Heart Association/American Stroke Association guideline noted that endovascular treatment may be considered in carefully selected pediatric patients with LVO.¹⁶⁾ Although outcomes of pediatric BAO have been reported to be more favorable than those of adult BAO,^2,17) it remains a potentially devastating condition, with reports of severe disability even under conservative medical treatment.¹⁸⁾ Therefore, in carefully selected pediatric patients with BAO, attempting mechanical thrombectomy may be reasonable when performed by experienced neurointerventionalists after consideration of stroke severity, the extent of salvageable brain tissue, and technical feasibility. However, because thrombectomy devices have been developed primarily for adults, and the safety and efficacy of endovascular treatment in pediatric patients remain incompletely established, informed consent should be obtained after careful discussion of the limited evidence available in pediatric patients. Compared with BAO in adult patients, BAO in pediatric patients is characterized by greater etiological heterogeneity. In addition to cardioembolism, a wide range of underlying conditions, including vertebral artery dissection, arteriopathies, thrombophilic disorders, cardiac diseases, and inflammatory vasculopathies, may contribute to the occurrence of stroke.^2,17) This etiological heterogeneity has important implications for treatment planning, because the underlying mechanism of vessel occlusion may influence both the feasibility and efficacy of endovascular treatment. For example, vertebral artery dissection, a common etiology of pediatric BAO, may require a different endovascular strategy from that used for embolic occlusions, since the primary pathology involves vessel wall injury. In such cases, minimizing device manipulation and avoiding unnecessary passes may be important for reducing the risk of further vascular injury.

Mycoplasma pneumoniae is a major cause of community-acquired pneumonia in pediatric patients and is associated with a variety of extrapulmonary manifestations.⁷⁾ In severe cases, thrombotic complications involving both the arterial and venous systems, including cerebral infarction, deep venous thrombosis, pulmonary embolism, and splenic infarction, have been reported.⁷⁾ Several pathogenic mechanisms for thrombosis associated with M. pneumoniae have been proposed, including direct vascular invasion, immune-mediated vasculopathy,⁸⁾ and infection-associated hypercoagulability.^7,9,10) Infection-associated hypercoagulability may be mediated by inflammatory cytokine release, complement activation, and endothelial dysfunction.⁶⁾ In addition, transient induction of antiphospholipid antibodies, including lupus anticoagulant, has been reported during acute M. pneumoniae infection, which may further contribute to a prothrombotic state.^7,9,10) In the present case, post-thrombectomy angiography demonstrated no residual stenosis or vessel wall abnormalities in the vertebral arteries bilaterally or the basilar artery, making underlying vasculitis or vasculopathy less likely. Additionally, retrieval of a red thrombus during mechanical thrombectomy supported an embolic mechanism rather than in situ thrombosis, while extensive investigations failed to identify an alternative embolic source. Moreover, coagulation testing revealed a positive lupus anticoagulant, markedly elevated D-dimer levels, and reduced fibrinogen concentrations, indicating a prothrombotic state with activation and consumption of the coagulation system. Taken together, these findings suggest that a hypercoagulable state associated with M. pneumoniae infection was the most plausible mechanism underlying vertebrobasilar artery occlusion in this patient.

To the best of our knowledge, only 1 previous report has described mechanical thrombectomy for ischemic stroke associated with M. pneumoniae infection in a child, which involved an M2 occlusion.¹¹⁾ The present case may therefore represent the first reported case of mechanical thrombectomy for posterior circulation stroke in a child associated with M. pneumoniae infection. Although effective reperfusion was achieved in both cases, the efficacy of thrombectomy in ischemic stroke associated with M. pneumoniae may depend on the underlying stroke etiology. In patients in whom arterial occlusion results primarily from thromboembolism related to infection-associated hypercoagulability, as in the present case, the underlying pathophysiology may resemble that of cardioembolic LVO in adults. Consequently, mechanical thrombectomy may be more likely to achieve successful reperfusion in such patients. In contrast, recent reports have demonstrated vessel wall enhancement on MRI and arterial stenosis in patients with Mycoplasma-associated central nervous system vasculitis.⁸⁾ In such cases, vasculopathic changes may extend over long arterial segments,⁸⁾ and underlying stenotic lesions may persist even after thrombus removal, potentially limiting the efficacy of mechanical thrombectomy. In line with this concept, a recent multicenter study comparing mechanical thrombectomy outcomes between pediatric ischemic strokes with focal cerebral arteriopathy and those with cardiogenic embolism found lower rates of successful reperfusion and less favorable functional outcomes in the focal cerebral arteriopathy group.¹⁹⁾ Furthermore, inflammatory arteriopathies may be associated with increased vascular fragility, which could increase the risk of hemorrhagic complications during endovascular procedures.³⁾ These factors should be considered when planning mechanical thrombectomy in patients with suspected vasculopathy. However, distinguishing thromboembolic occlusion from vasculopathy-driven occlusion before mechanical thrombectomy is often challenging in clinical practice. In such cases, angiographic findings may provide useful clues to the underlying mechanism. For example, the clot meniscus sign has been reported as an angiographic marker associated with embolic vessel occlusion,²⁰⁾ whereas arterial stenosis may raise the suspicion of an underlying vasculopathy.⁸⁾ Nevertheless, the diagnostic value of these findings in ischemic stroke associated with M. pneumoniae infection remains uncertain because of the extremely limited number of reported cases. Further accumulation of well-characterized cases is needed to clarify how the underlying stroke etiology might influence patient selection and treatment strategies for mechanical thrombectomy in ischemic stroke associated with M. pneumoniae infection.

Conclusion

We report a pediatric case of vertebrobasilar artery occlusion associated with M. pneumoniae infection that was successfully treated with mechanical thrombectomy. This case highlights the feasibility and potential efficacy of mechanical thrombectomy in selected pediatric patients with acute ischemic stroke, even in the setting of rare etiologies such as M. pneumoniae infection. Given the heterogeneous etiologies underlying pediatric stroke, further accumulation of well-characterized cases is needed to refine patient selection and optimize endovascular treatment strategies for ischemic stroke in pediatric patients.

Ethics Statement

Written informed consent for publication of this report was obtained from the patient’s parents.

Disclosure Statement

The authors declare no conflicts of interest.