Tumor expectoration following drug-eluting bead bronchial artery chemoembolization for advanced non-small cell lung cancer

Abstract

Drug-eluting bead bronchial artery chemoembolization (DEB-BACE) has been investigated as a locoregional salvage treatment for advanced non-small cell lung cancer (NSCLC). It may be considered for treatment-refractory or symptomatic central tumors. We retrospectively describe case of three patients with centrally located, treatment-refractory NSCLC who underwent DEB-BACE. All three subsequently expectorated soft-tissue material. Clinical features, procedures, imaging, pathology, and outcomes were reviewed. Potential bronchial and non-bronchial systemic tumor-feeding arteries were assessed using contrast-enhanced computed tomography and selective digital subtraction angiography (DSA). Lobaplatin was infused intra-arterially. Embolization was then performed with 300 to 500 μm epirubicin-loaded Bead microspheres. The procedural endpoint was marked reduction of tumor blush with stasis or near-stasis. One patient had SMARCA4-deficient NSCLC. Two had squamous cell carcinoma. Soft-tissue expectoration occurred 3 days to 2 weeks after DEB-BACE. Histopathology confirmed carcinoma with extensive necrosis in Cases 1 and 2. Pathological confirmation was unavailable in Case 3. Case 1 achieved a durable partial response during repeated DEB-BACE and continued sintilimab treatment. The tumor decreased from 94 mm × 55 mm to 22 mm × 14 mm at 13 months and remained stable at 29 months. Case 2 developed tumor expectoration 3 days after treatment. Bronchial patency was subsequently restored. The patient died 10 months after diagnosis. Case 3 developed recurrent hemoptysis and underwent emergency hemostatic bronchial artery embolization. The patient subsequently expectorated a soft-tissue mass and died from progressive pulmonary infection 14 months after diagnosis. Tumor expectoration may represent treatment-associated necrosis after DEB-BACE. Careful assessment of vascular anatomy, bleeding risk, infection, and non-target embolization risk is essential.

Introduction

Systemic therapy, including platinum-based chemotherapy and immune checkpoint inhibitors, remains the cornerstone of treatment for advanced non-small cell lung cancer (NSCLC). However, treatment options become limited after progression on first-line or subsequent systemic regimens (1,2). This problem is particularly relevant in patients with large central tumors. Such tumors may cause airway obstruction, post-obstructive pneumonia, superior vena cava syndrome (SVCS), or refractory hemoptysis. Systemic therapy may not provide sufficiently rapid locoregional control in these settings (3-5).

Bronchial artery chemoembolization (BACE) delivers chemotherapy and embolic material directly to tumor-feeding arteries and has been investigated as a locoregional treatment in selected patients (6,7). Drug-eluting bead bronchial artery chemoembolization (DEB-BACE) uses drug-eluting microspheres for calibrated embolization and local delivery of chemotherapy. Whether these features improve clinical outcomes compared with conventional embolization remains uncertain (8,9). DEB-BACE has been reported in patients with NSCLC complicated by obstructive pneumonia, hemoptysis, and treatment-refractory central tumors (8,10-12).

Our previous retrospective study reported the feasibility, local treatment activity, and safety profile of DEB-BACE combined with bronchial arterial infusion chemotherapy in patients with advanced treatment-refractory NSCLC (13). The observational design did not establish comparative efficacy. Clinical experience with DEB-BACE remains limited. Its clinical effects and complications, including tumor necrosis and expectoration, require further characterization.

We report a case series of three patients with advanced, treatment-refractory NSCLC who underwent DEB-BACE and subsequently experienced expectoration of soft-tissue material. Histopathological confirmation of necrotic carcinoma was obtained in two patients. This report focuses on the technical aspects of DEB-BACE, the possible relationship between treatment-associated tumor necrosis and expectoration, and relevant procedural safety considerations. All three patients included in this study were identified and treated at Guangdong Provincial People’s Hospital.

General DEB-BACE technique

Potential bronchial and non-bronchial systemic tumor-feeding arteries were evaluated on preprocedural contrast-enhanced chest computed tomography (CT) and selective DSA. A 5-Fr diagnostic catheter was used to catheterize suspected feeding arteries. A 2.7-Fr microcatheter (Progreat, Terumo, Japan) was advanced superselectively into each confirmed tumor-feeding artery as close to the tumor bed as safely feasible (13). Selective DSA from the intended treatment position was reviewed for tumor blush and spinal arterial branches. The images were also assessed for other non-target branches and for bronchial or systemic arterial communications with the pulmonary arterial or venous circulation.

Lobaplatin was administered first through the microcatheter. When several tumor-feeding arteries were treated, the total dose was divided according to the degree of tumor staining in each artery (13). Embolization was then performed with 300 to 500 μm DC Bead^TM microspheres (Boston Scientific, Marlborough, MA, USA). The microspheres are biocompatible, hydrophilic, non-resorbable, size-calibrated polyvinyl alcohol-based hydrogel particles. They were loaded with 50 mg epirubicin. The microspheres were delivered through the microcatheter under fluoroscopic monitoring. The target endpoint was marked reduction or disappearance of tumor blush with stasis or near-stasis of the treated vessel (13). Embolization was completed when this endpoint was reached.

Preprocedural contrast-enhanced CT assessment

Preprocedural contrast-enhanced chest CT was performed using a dual-source multidetector CT scanner (SOMATOM Definition Flash, Siemens Healthineers, Forchheim, Germany). The scan range extended from the lung apices to the level of the diaphragmatic domes. Ioversol at 350 mg I/mL was administered through an antecubital vein using a power injector. The total contrast volume was 80 mL and the injection rate was 4.0 mL/s. This was followed by 50 mL of saline. Arterial-phase imaging was performed using bolus tracking. The region of interest was placed in the aortic arch. Image acquisition began 5 seconds after attenuation reached 100 Hounsfield units (HU). Images were reconstructed at a slice thickness of 1.0 mm. Multiplanar and vascular reconstructions were generated.

The images were reviewed for the origin and course of orthotopic and ectopic bronchial arteries. Potential non-bronchial systemic tumor-feeding arteries were also assessed. These included intercostal, internal thoracic, subclavian-derived, and inferior phrenic arteries when applicable. The relationship between the tumor and the pulmonary arteries and veins was evaluated. Tumor enhancement, necrosis, vascular invasion, vascular occlusion, and other vascular findings relevant to treatment planning were also assessed.

Case 1: SMARCA4-deficient NSCLC with SVCS

Clinical presentation

A 34-year-old man with a 30-pack-year smoking history presented with chest tightness and facial swelling of 2 weeks’ duration. He had been diagnosed 5 months earlier with a right hilar/paraspinal mass (66 mm × 43 mm) causing right lower lobe atelectasis. Biopsy revealed an undifferentiated carcinoma with loss of SMARCA4 expression [programmed death-ligand 1 (PD-L1) tumor proportion score <1%], consistent with SMARCA4-deficient NSCLC (clinical stage cT4N2M0, IIIB). SMARCA4-deficient NSCLC is a rare and aggressive subtype with frequent disease progression and poor prognosis (14).

First-line therapy with paclitaxel, cisplatin, and the anti-programmed cell death protein 1 (PD-1) antibody sintilimab was administered for four cycles. The patient did not receive second-line chemotherapy primarily because he declined further systemic treatment after completing first-line therapy. Restaging demonstrated disease progression with tumor enlargement to 94 mm × 55 mm, persistent right middle and lower lobe atelectasis, and bronchial occlusion. Contrast-enhanced CT confirmed SVCS, with thrombus in the mid-to-distal superior vena cava and extensive chest wall collaterals (Figure 1A). His performance status declined to 2. Representative follow-up CT images demonstrating subsequent tumor regression and durable local control are shown in Figure 1B,1C, respectively.

Figure 1.

Case 1. Imaging in SMARCA4-deficient NSCLC with superior vena cava syndrome. (A) Contrast-enhanced CT after progression on first-line chemoimmunotherapy, showing a right hilar/paraspinal mass (94 mm × 55 mm, red arrows) with SVC thrombus (yellow arrows) and chest wall collaterals (white arrows). (B) CT 4 months after the third DEB-BACE, demonstrating marked tumor regression (36 mm × 25 mm, red arrows), resolution of atelectasis and collaterals, with the SVC stent and contained thrombus in place (yellow arrows). (C) CT at 29-month follow-up, confirming durable local control with the primary tumor stable at 22 mm × 14 mm (red arrows) and a patent SVC stent (yellow arrows). (D) DSA during the second DEB-BACE demonstrating persistent hypervascular tumor supply from multiple bronchial and non-bronchial systemic arterial branches, including intercostal and internal thoracic arterial feeders, which were treated superselectively. (E) Composite gross photographs of necrotic tissue fragments obtained over approximately 12 expectoration episodes about 2 weeks after the second DEB-BACE. (F) Hematoxylin and eosin stain of expectorated material, revealing necrotic debris with scattered nests of atypical epithelial cells (poorly differentiated carcinoma) (magnification, 400×). CT, computed tomography; DEB-BACE, drug-eluting bead bronchial artery chemoembolization; DSA, digital subtraction angiography; NSCLC, non-small cell lung cancer.

Treatment course

Initial management and first DEB-BACE

At presentation, urgent endovascular treatment was performed for symptomatic SVCS. Two self-expanding Wallstent endoprostheses measuring 90 mm × 14 mm and 60 mm × 14 mm were deployed in the SVC. Both stents were manufactured by Boston Scientific, USA. Facial swelling and chest tightness improved immediately after stent placement. This early symptomatic improvement was attributed primarily to restoration of venous outflow. DEB-BACE was performed during the same session to address the underlying tumor burden.

Diagnostic angiography identified hypervascular tumor supply from the right bronchial and adjacent intercostal arteries. Preprocedural imaging showed no pulmonary arterial or pulmonary venous tumor invasion. After superselective microcatheterization of the tumor-feeding artery, 60 mg lobaplatin was first infused intra-arterially through the microcatheter. This was followed by embolization using 300 to 500 μm epirubicin-loaded DC Bead^TM microspheres. Embolization was continued until near-stasis was achieved. Near-stasis was defined as persistence of contrast material in the target vessel for approximately 3 to 4 cardiac cycles. Because the patient developed substantial back discomfort after prolonged recumbency and declined continuation of the procedure, additional planned arterial catheterization was not performed during this session. Systemic therapy with sintilimab was maintained, and therapeutic anticoagulation with rivaroxaban was initiated.

Second DEB-BACE and tumor expectoration

Follow-up imaging 2 months after the first DEB-BACE demonstrated stable disease by RECIST 1.1 criteria (15), with reduced central enhancement and slight improvement in atelectasis. A second DEB-BACE was performed 3 months after the first. Angiography demonstrated persistent tumor hypervascularity with additional collateral tumor-feeding branches. Multiple bronchial and non-bronchial systemic feeders contributed to the tumor supply. These included intercostal and internal thoracic arterial branches, which were treated superselectively (Figure 1D). Lobaplatin at 60 mg was administered before embolization with 300 to 500 μm epirubicin-loaded DC Bead^TM microspheres. One distal branch was additionally embolized with a 5 mm × 5.5 mm microcoil. Embolization was completed after marked reduction of tumor blush with stasis or near-stasis of the treated vessels. CT 1 month after the second procedure showed tumor reduction to 63 mm × 55 mm. Obstructive atelectasis and bronchial narrowing also improved substantially. The tumor response met RECIST 1.1 criteria for partial response (15).

Approximately 2 weeks after the second DEB-BACE, the patient expectorated large fragments of necrotic tissue over approximately 12 episodes (Figure 1E). Pathology confirmed poorly differentiated carcinoma with extensive necrotic debris, hemorrhagic exudate, and scattered aggregates of atypical epithelial cells (Figure 1F).

Third DEB-BACE and follow-up

A third DEB-BACE was performed 2 months after the second using the same sequence of intra-arterial lobaplatin infusion followed by epirubicin-loaded microsphere embolization. Angiography demonstrated markedly diminished residual tumor vascularity, and embolization was performed to the predefined angiographic endpoint. Serial imaging demonstrated continued regression. By 4 months after the third procedure, the tumor measured 36 mm × 25 mm, maintaining partial response. Post-obstructive atelectasis had nearly resolved, with restoration of bronchial patency. CT at 13 months showed further reduction to 22 × 14 mm. At 29 months after initiating interventional therapy (32 months from diagnosis), the tumor remained stable at 22 mm × 14 mm without locoregional progression. The patient remained alive with performance status 1, reported subjective improvement in daily functioning and respiratory symptoms, and showed no signs of local progression.

Case 2: squamous cell carcinoma with tumor expectoration

Clinical presentation

A 60-year-old man with a 30-pack-year smoking history (quit 5 years prior) and hypertension presented with cough and sputum of 6 months’ duration. Chest CT at an outside hospital showed right lower lobe atelectasis and bronchial obstruction (Figure 2A). The corresponding postprocedural CT demonstrating restored right lower bronchial patency after DEB-BACE and tumor expectoration is shown in Figure 2B. Biopsy confirmed moderately differentiated squamous cell carcinoma [cT2N3M0 (IIIB)]. The tumor caused distal obstructive atelectasis and post-obstructive pneumonia. Molecular profiling was negative for driver mutations. PD-L1 expression was 10%. The patient had an Eastern Cooperative Oncology Group (ECOG) performance status of 1.

Figure 2.

Case 2. Imaging in squamous cell carcinoma. (A) Preprocedural contrast-enhanced chest CT demonstrating a centrally located tumor causing right lower lobe atelectasis and bronchial obstruction. Red arrows indicate the centrally located lesion and associated bronchial obstruction. (B) Postprocedural CT after DEB-BACE and tumor expectoration, showing resolution of atelectasis with restored patency of the right lower bronchus. Red arrows indicate the restored patency of the right lower lobe bronchus. (C) Digital subtraction angiography during DEB-BACE demonstrating hypervascular tumor supply from the bronchial artery before superselective intra-arterial infusion of lobaplatin followed by embolization with 300 to 500 μm epirubicin-loaded DC BeadTM microspheres. Red arrows indicate the bronchial arterial branches supplying the tumor. (D) Gross specimen of necrotic tumor tissue expectorated 3 days after the procedure. (E) Hematoxylin and eosin stain of the expectorated material, confirming moderately differentiated squamous cell carcinoma with extensive necrosis (magnification, 200×). CT, computed tomography; DEB-BACE, drug-eluting bead bronchial artery chemoembolization.

Treatment course

The patient received two cycles of first-line chemoimmunotherapy with paclitaxel liposome, cisplatin, and tislelizumab. Reassessment CT showed stable disease. Treatment was subsequently interrupted because of Haemophilus influenzae pneumonia and coronavirus disease 2019 (COVID-19). Antimicrobial treatment was administered. After recovery, the patient underwent DEB-BACE combined with tislelizumab at 200 mg. Selective DSA demonstrated tumor supply from the bronchial artery. After superselective catheterization, lobaplatin at 60 mg was administered through the microcatheter. Embolization was then performed with 300 to 500 μm DC Bead^TM microspheres loaded with 50 mg epirubicin. The microspheres were delivered through the microcatheter under fluoroscopic monitoring. Embolization was completed after marked reduction of tumor blush with stasis or near-stasis of the target vessel. Only the bronchial arterial supply was embolized during this treatment session. No non-bronchial systemic artery was treated (Figure 2C).

Tumor expectoration

Three days after the first DEB-BACE, the patient developed dyspnea and severe cough. He then expectorated tissue that was histopathologically confirmed as moderately differentiated squamous cell carcinoma with extensive necrosis (Figure 2D,2E). Following expectoration, repeat evaluation demonstrated restoration of right lower bronchial patency with resolution of dyspnea.

Follow-up

The patient subsequently received additional DEB-BACE treatment and systemic therapy. He died 10 months after the initial diagnosis.

Case 3: locally advanced squamous cell carcinoma with tumor expectoration

Clinical presentation

A 67-year-old man with a 40-pack-year smoking history presented with cough and hemoptysis of 3 months’ duration. Evaluation at an outside hospital, including positron emission tomography (PET)-CT and biopsy, established a diagnosis of non-keratinizing squamous cell carcinoma of the left lower lobe, staged as cT4N1M0 (IIIA). The tumor was centrally located, causing distal obstructive atelectasis and abutting the descending aorta.

He received several cycles of chemotherapy and immunotherapy at another institution. Restaging CT demonstrated disease progression with tumor enlargement, invasion of the left inferior pulmonary vein, and persistent atelectasis. Due to persistent hemoptysis and refractory disease, the patient was referred to our center. Baseline contrast-enhanced CT showed a large left lower lobe mass measuring 83 mm × 74 mm. The mass encased the descending aorta and occluded the left inferior pulmonary vein. Hilar lymph node metastasis invaded the left main bronchus and was associated with post-obstructive pneumonia (Figure 3A). The follow-up contrast-enhanced CT obtained approximately 10 days after the second DEB-BACE is shown in Figure 3B.

Figure 3.

Case 3. Imaging in locally advanced squamous cell carcinoma. (A) Baseline contrast-enhanced chest CT showing a large left lower lobe mass (83 mm × 74 mm) encasing the descending aorta, occluding the left inferior pulmonary vein, with hilar lymphadenopathy invading the left main bronchus and resulting in post-obstructive pneumonia. Red arrows indicate the baseline left lower lobe tumor. (B) Contrast-enhanced CT obtained approximately 10 days after the second DEB-BACE, demonstrating the left lower lobe tumor measuring 82 mm × 70 mm with markedly heterogeneous enhancement and intratumoral poorly enhancing necrotic areas. Red arrows indicate the left lower lobe tumor after DEB-BACE. (C) DSA during the second DEB-BACE showing hypervascular tumor supply from the bronchial and internal thoracic arteries before superselective treatment. (D) Gross photograph of the expectorated soft-tissue mass. The specimen was not available for histopathological examination. Its tumor origin was therefore not pathologically confirmed. CT, computed tomography; DEB-BACE, drug-eluting bead bronchial artery chemoembolization.

Treatment course

The first DEB-BACE was performed on October 25, 2024, together with tislelizumab at 200 mg. The second DEB-BACE was performed on November 18, 2024. During both procedures, lobaplatin at 60 mg was administered first through a superselectively positioned microcatheter. Embolization was then performed with 300 to 500 μm DC Bead^TM microspheres loaded with 50 mg epirubicin. Selective DSA showed tumor supply from the bronchial artery and the internal thoracic artery (Figure 3C). Both angiographically confirmed tumor-feeding arteries were treated superselectively during the DEB-BACE procedures. The target endpoint was marked reduction of tumor blush with stasis or near-stasis of the treated vessels. Embolization was completed after this endpoint was reached. No immediate intraprocedural complication was documented.

Contrast-enhanced CT was obtained on November 28, approximately 10 days after the second DEB-BACE. The left lower lobe mass measured 82 mm × 70 mm compared with 83 mm × 74 mm at baseline. The lesion remained markedly heterogeneous and contained poorly enhancing necrotic areas. The change in tumor size did not meet RECIST 1.1 criteria for partial response. Radiological necrotic change was present.

Tumor expectoration

Approximately 2 weeks after the second DEB-BACE, the patient presented with recurrent hemoptysis. Emergency transcatheter bronchial artery embolization was performed for hemostasis. Shortly after this procedure, the patient expectorated a soft-tissue mass (Figure 3D). The specimen was not preserved for histopathological examination. The contemporaneous medical record did not document a more specific reason. The nature of the expectorated material was therefore not histologically confirmed. CT before the hemostatic procedure had already demonstrated intratumoral necrotic change after the second DEB-BACE. The expectorated material was considered clinically compatible with treatment-associated necrotic tumor tissue. The relative contributions of the preceding DEB-BACE and the subsequent hemostatic embolization cannot be determined with certainty.

Follow-up

A further contrast-enhanced chest CT was obtained on December 2 after the emergency hemostatic embolization. The scan continued to show heterogeneous enhancement and intratumoral necrotic areas. Formal size-based tumor response assessment was limited by concurrent left lung atelectasis, a large left pleural effusion, and bilateral pulmonary infection. The subsequent course was complicated by progressive pulmonary infection requiring intensive care unit admission. The patient died 14 months after the initial diagnosis. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Guangdong Provincial People’s Hospital (KY2024-118-01). Written informed consent was obtained from all participants.

Discussion

This article describes three patients with advanced, treatment-refractory NSCLC who experienced expectoration of soft-tissue material after DEB-BACE. All three had centrally located tumors with angiographically evident systemic arterial supply and had previously received systemic anticancer treatment. The expectoration events occurred 3 days to 2 weeks after DEB-BACE. Histopathological examination confirmed carcinoma with extensive necrosis in Cases 1 and 2, whereas pathological confirmation was unavailable in Case 3. These observations support a substantial local treatment effect in selected patients but should not be interpreted as evidence of comparative superiority of DEB-BACE. They also highlight the need for careful assessment of hemorrhagic and infectious complications as well as the risk of non-target embolization.

Tumor expectoration and local treatment effect

Expectoration of necrotic tumor tissue may provide a visible manifestation of local treatment-associated necrosis when pathological confirmation is available. The three patients had centrally located tumors with bronchial involvement and angiographically evident tumor vascularity. In Case 1, facial swelling and chest tightness improved immediately after SVC stent placement. This improvement was primarily attributed to restoration of venous outflow. Tumor shrinkage and improvement in bronchial obstruction developed later during repeated DEB-BACE and continued sintilimab treatment. The individual contribution of DEB-BACE to long-term survival cannot be determined. This case also cannot establish synergy with immunotherapy. Case 2 expectorated tissue 3 days after DEB-BACE and subsequently showed restoration of bronchial patency. Case 3 developed recurrent hemoptysis after the second DEB-BACE and required emergency hemostatic embolization before expectorating soft-tissue material. The specimen was not examined histologically, so its nature and causal relationship to DEB-BACE remain less certain. These observations are consistent with published reports of local treatment activity after DEB-BACE (4,9,16,17). They do not establish comparative efficacy.

The mechanistic rationale of DEB-BACE combines arterial devascularization with local delivery of chemotherapy from the loaded microspheres (8,13). Angiographically evident tumor vascularity provides the anatomical basis for selective transarterial treatment. Pharmacokinetic observations from other tumors or vascular territories should not be assumed to represent drug exposure within lung tumors. The interval between DEB-BACE and expectoration should therefore be regarded as a temporal association rather than proof of a specific drug-release mechanism. Treatment-associated ischemia and chemotherapy may contribute to tumor necrosis. Necrotic material that communicates with the airway may then be expelled by coughing. This proposed mechanism remains hypothetical.

Potential role of DEB-BACE and current evidence

Drug-eluting microspheres provide a platform for calibrated particulate embolization and local delivery of a loaded chemotherapeutic agent. A retrospective comparative study in NSCLC reported favorable outcomes with DEB-BACE compared with BACE using gelatin sponge particles (9). A systematic review and meta-analysis also reported favorable pooled outcomes with drug-eluting embolic BACE (17). However, the available evidence is largely based on nonrandomized and heterogeneous studies. Selection bias and residual confounding remain important limitations. The current lung cancer literature therefore does not establish the clinical superiority of DEB-BACE.

International experience also supports the feasibility of transarterial treatment beyond the bronchial arteries. Hori et al. retrospectively reported transarterial treatment in 98 patients with advanced or recurrent lung cancer. The treated vessels included bronchial arteries and, when appropriate, mediastinal branches arising from the subclavian circulation. Selective arterial infusion was followed by embolization (18). This study further emphasizes the importance of identifying the complete systemic arterial supply to the tumor rather than limiting angiographic assessment to the bronchial arteries.

Evidence from other vascular territories also shows that theoretical advantages of drug-eluting microspheres do not always improve clinical outcomes. In a recent multicenter randomized trial in hepatocellular carcinoma, transarterial embolization and drug-eluting microsphere chemoembolization produced similar time to progression, overall survival (OS), tumor response, and adverse event outcomes (19). These findings cannot be directly extrapolated to lung cancer because disease biology and vascular anatomy differ. They nevertheless support the need for randomized lung-cancer-specific evidence before superiority of DEB-BACE can be concluded.

Accordingly, DEB-BACE should currently be regarded as a promising locoregional treatment platform for selected patients rather than an established superior alternative to conventional BACE.

Safety considerations for hemorrhage and non-target embolization

A major safety concern after tumor-directed embolization is hemorrhage from vessels affected by tumor invasion, necrosis, or cavitation. Pulmonary arterial invasion is an important source of life-threatening bleeding risk in advanced lung cancer (5). Bronchial arterial tumor vascularity may represent a separate source of bleeding.

Treatment-associated necrosis may further weaken tumor tissue and adjacent vessels. Case 3 illustrates this concern. Recurrent hemoptysis occurred after the second DEB-BACE and required emergency hemostatic bronchial artery embolization. The precise vascular source could not be established with certainty. Necrotic material within an obstructed airway may also predispose to secondary infection, as illustrated by the subsequent pulmonary infection in Case 3.

Bronchial or non-bronchial systemic artery-to-pulmonary venous shunting is another serious concern during particulate embolization. Embolic particles can pass through such a communication into the pulmonary veins. They may then reach the left heart and systemic arterial circulation. This creates a risk of cerebral, coronary, or other systemic embolization. Selective DSA should be reviewed for early pulmonary venous opacification and other signs of high-flow shunting before particulate embolization. When a large pulmonary arterial or venous shunt is present, larger particles or coil embolization may reduce the risk of non-target embolization (20).

Inadvertent embolization of spinal arterial branches is rare but potentially catastrophic. Radiculomedullary or anterior spinal arterial supply may arise from intercostobronchial or other systemic arteries. Inadvertent particulate embolization of these branches may cause spinal cord ischemia. Selective and superselective angiograms should be reviewed for vessels coursing toward the spinal canal and for the characteristic hairpin appearance of a radiculomedullary artery. Particulate embolization should not be performed from a position where a branch supplying the spinal cord, brain, or heart cannot be safely excluded or bypassed (20).

During DSA, early and delayed images were assessed for tumor blush and abnormal enlargement or tortuosity of the target artery. The images were also reviewed for pulmonary arterial shunting, early pulmonary venous filling, and hazardous non-target branches. Before particulate embolization, superselective angiography was repeated from the intended microcatheter position. Embolization was not performed from a position that showed a critical spinal or other non-target branch. The catheter position or embolic strategy was modified when a safe injection position could not be established (20).

Patient selection and preprocedural risk stratification

DEB-BACE should be considered a selective locoregional salvage option rather than a substitute for established systemic or radiation-based treatments. Potential candidates include adults with histologically confirmed unresectable NSCLC who have local progression or cannot tolerate standard therapy (13). A clear locoregional treatment goal should be present. The target tumor should have accessible systemic arterial supply. Patients should also have adequate performance status, organ function, and coagulation status for an arterial intervention. In our previous study, ECOG performance status above 2 and untreatable coagulation disorders were exclusion criteria (13).

A separate retrospective comparative study excluded patients with severe coagulation dysfunction, severe liver, kidney, or cardiopulmonary dysfunction, active infection, and other contraindications to interventional treatment (9). These findings support deferral or reconsideration of DEB-BACE in patients with severe uncorrectable coagulopathy, severe organ dysfunction, or uncontrolled infection. The same applies when local treatment is unlikely to provide meaningful clinical benefit. A vessel that supplies the spinal cord, brain, or heart should not be embolized with particles if the critical branch cannot be safely bypassed (20). A substantial artery-to-pulmonary venous shunt should also prompt modification or avoidance of particulate embolization when it cannot be safely managed (20).

Pulmonary arterial invasion, extensive tumor necrosis or cavitation, and active hemoptysis warrant particular caution (5). Severe post-obstructive infection should also be considered during procedural risk assessment. Additional treatment-associated necrosis may increase hemorrhagic or infectious risk. Patient selection should therefore consider the treatment goal, systemic disease burden, performance status, bleeding status, infection status, vascular anatomy, and the feasibility of safe superselective catheterization. Prospective studies are needed before formal DEB-BACE-specific contraindications can be established.

Integration with systemic therapy and future directions

All three patients received immune checkpoint inhibitors. Combination strategies involving BACE or DEB-BACE and immunotherapy have been investigated in prospective single-arm and retrospective studies (6,16,21,22). A phase II study of BACE plus tislelizumab reported a median progression-free survival of 10.5 months and a median OS of 15.0 months (6). Retrospective comparative data have also reported favorable outcomes with DEB-BACE plus PD-1 blockade (21). These study designs cannot establish biological synergy or determine the independent contribution of each treatment component. The prolonged disease control observed in Case 1 likewise cannot establish synergy between DEB-BACE and sintilimab.

Superselective BACE or DEB-BACE is technically repeatable and may provide a locoregional treatment option in selected patients (11,13). The present study does not compare DEB-BACE with external beam radiation therapy and does not establish superiority over other locoregional treatments. This report is limited by its retrospective design, small sample size, and heterogeneity of prior treatments. Prospective studies with standardized protocols are needed to validate these findings and define optimal strategies for procedural risk reduction.

Conclusions

Tumor expectoration may represent a clinically observable manifestation of treatment-associated necrosis after DEB-BACE in selected patients with centrally located NSCLC. In this series, the expectorated material was histologically confirmed as necrotic carcinoma in two of three cases. Preprocedural assessment should address pulmonary vascular invasion, non-bronchial systemic tumor supply, artery-to-pulmonary venous shunting, and critical non-target arterial branches. DEB-BACE may have a role as a locoregional salvage option in carefully selected patients. Its comparative benefit over conventional embolization remains unproven and requires prospective evaluation.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was approved by the Institutional Review Board of Guangdong Provincial People’s Hospital (KY2024-118-01). Written informed consent was obtained from all participants.