Thoracoscopic retrieval of a rare migrated fish bone in the thorax: a case report with sequential CT evidence and literature review

Abstract

Background

Esophageal foreign bodies represent a common clinical emergency. Sharp foreign bodies, such as fish bones, carry a risk of esophageal perforation and subsequent severe complications. Extraluminal migration into the thoracic cavity is an uncommon but clinically significant occurrence.

Case presentation

A 58-year-old male presented with a 6-hour history of fish bone ingestion. Initial chest computed tomography (CT) demonstrated an esophageal foreign body with associated perforation; however, rigid esophagoscopy was unsuccessful in localizing the object. Follow-up contrast-enhanced CT revealed migration of the foreign body into the thoracic cavity posterior to the left subclavian artery, accompanied by left-sided pneumothorax and 80% lung compression. Single-port thoracoscopic surgery performed 28 h post-admission successfully retrieved a 2.5-cm fish bone and addressed the pulmonary injury. The patient recovered without complications, and follow-up imaging confirmed no delayed sequelae.

Conclusions

Sequential CT imaging facilitates accurate localization of migrated foreign bodies, while minimally invasive thoracoscopic techniques provide an effective therapeutic option. Multidisciplinary collaboration and imaging-guided clinical decision-making play a critical role in minimizing complications and optimizing outcomes.

Background

Esophageal foreign bodies (FBs) represent a common emergency in otolaryngology and gastroenterology. Sharp objects such as fish bones account for the majority of cases, with reported esophageal perforation rates of 1.4% and an overall mortality rate of 0.85% in adults [1]. Migrated FBs typically lodge in the neck or mediastinum, leading to complications such as neck abscesses, mediastinitis, or injury to adjacent organs (e.g., trachea or major blood vessels) [2]. However, extraluminal migration beyond the mediastinum into the thoracic cavity is exceptionally rare. Since the first case reported by So et al. in 1982—a 42-year-old man with a 3-cm chicken bone embedded in the right upper lung lobe following a 5-month history of hemoptysis [3]—only 10 similar cases have been documented worldwide. Advances in high-resolution computed tomography (CT) and minimally invasive thoracoscopic techniques now enable early detection and intervention for such rare migrations. Here, we present a case of a fish bone migrating into the thoracic cavity, a dynamic process visualized by sequential CT imaging, and successfully retrieved via video-assisted thoracic surgery (VATS) within 1.5 days of ingestion. This study highlights the critical role of imaging-guided decision-making and rapid multidisciplinary management in averting life-threatening complications.

Case presentation

A 58-year-old male patient was admitted to the hospital with a “6-hour history of accidental fish bone ingestion.” His medical history included chronic gastritis and prior cholecystectomy. Six hours prior, the patient experienced sudden retrosternal pain and dysphagia without dyspnea after swallowing a fish bone during a meal. He initially visited a local hospital, where a chest CT scan revealed “esophageal FB with suspected esophageal wall penetration (Fig. 1A).” Subsequently, the patient was transferred to the Department of Otolaryngology, the First Affiliated Hospital of Zhejiang University School of Medicine. An urgent thoracic aortic computed tomography angiography (CTA) further indicated “a FB in the upper thoracic esophageal lumen with esophageal perforation (Fig. 1B, F and H).”

Fig. 1.

Temporal evolution of intrathoracic foreign body migration on sequential CT imaging. A (1-hour post-ingestion): Initial axial CT reveals a transversely oriented FB (red arrow) embedded in the esophagus, with partial mural penetration (tip projecting into the mediastinum). B (6-hour post-ingestion): Follow-up axial imaging shows complete transmural penetration, with the sharp tip (red arrow) embedding in the left upper lobe parenchyma. C & D (22-hour and 32-hour post-ingestion): Subsequent axial images document the FB (red arrow) positioned posterior to the left subclavian artery (yellow triangle) within the left thoracic cavity, accompanied by pneumothorax. The interval between C and D demonstrates positional stability relative to the vessel despite ongoing pneumothorax. E (52-hour post-ingestion): Postoperative axial imaging confirms complete FB removal, with resolution of pneumothorax and preserved vascular integrity of the left subclavian artery (yellow triangle). F & G (6-hour and 22-hour post-ingestion): Coronal reformats corresponding to B and C, respectively, clearly depict the craniocaudal relationship of the FB (red arrow) to the left subclavian artery (yellow triangle) and the evolving pneumothorax. H & I (6-hour and 22-hour post-ingestion): Sagittal reformats corresponding to B and C provide lateral views, further illustrating the anterior-posterior trajectory of the migrating FB (red arrow) and its consistent proximity to the left subclavian artery (yellow triangle)

Under general anesthesia, a rigid esophagoscopy was performed 3 hours after admission as the first intervention. Intraoperatively, no FB was identified; however, significant edema was noted in the esophageal lumen approximately 18 cm from the incisors, with an ulcerative mucosal wound observed on its left lateral wall. Postoperative management included nasogastric feeding diet, intravenous fluid support, and close monitoring. The following day, the patient reported persistent chest pain. The first non-contrast CT scan after the procedure showed further migration of the FB, “a linear high-density shadow along the posterior margin of the left subclavian artery and left-sided pneumothorax (30% lung compression, Fig. 1C, G, and I).” The Department of Thoracic Surgery was initially consulted, recommending nil per os, bed rest, and a CTA to assess possible injury to the left subclavian artery. That afternoon, the patient developed new chest tightness, although oxygen saturation remained stable. By evening, a contrast-enhanced CT scan revealed the FB in a similar location, with progression of the left pneumothorax (80% lung compression, Fig. 1D).”

The Department of Thoracic Surgery was re-consulted and confirmed surgical indications. Due to the proximity of the FB to the artery, the vascular surgery team was involved to prepare for potential arterial hemorrhage during the procedure. At 28 h post-admission, a left-sided thoracocentesis was performed prior to anesthesia induction to partially decompress the pneumothorax. The patient then underwent single-port VATS under general anesthesia. Intraoperatively, a 2.5-cm white FB (consistent with a fish bone based on history) was identified posterior to the left subclavian artery, approximately 5 cm above the aortic arch (Fig. 2). The FB was mostly embedded in the thoracic cavity, accompanied by mild pleural adhesions and a small amount of yellowish pleural effusion. The parietal pleura surrounding the FB was incised with an electrohook posterior to the left subclavian artery. After confirming that the end of the FB was not hooked into the artery, the FB was removed with double-jointed dissecting forceps. Needle-like marks on the adjacent lung surface were sutured and repaired. The procedure lasted 1 h, with intraoperative placement of a thoracic drainage tube. Postoperatively, the patient reported significant relief of chest pain and tightness. Secondary infection and leakage were closely monitored.

Fig. 2.

Intraoperative photograph of the FB. A: Thoracoscopic location of the FB (blue arrow) was posterior to the left subclavian artery (yellow triangle); B: The FB was a sharp fish bone at a length of approximately 2.5 cm

A repeat CTA performed one day after the VATS (Fig. 1D) revealed no evidence of pseudoaneurysms formation. A subsequent esophagogram performed five days postoperatively ruled out contrast extravasation. The patient was therefore allowed to resume oral intake. After four days of oral feeding without recurrence of chest pain, fever, or increased drainage, the chest tube was removed. The patient was discharged the following day following an uneventful recovery. A one-month postoperative follow-up CTA ruled out delayed pseudoaneurysms formation. One year later, gastroscopy performed for his chronic gastritis revealed no esophageal scarring or strictures.

Discussion

In this case, sequential CT imaging confirmed thoracic migration of the esophageal FB, which was successfully retrieved via VATS 28 h post-admission. A review of the literature (Table 1) based on the Pubmed database identified a median interval of 20 days (range: 1.5 days to 5 years) between ingestion and FB retrieval in migrated intrathoracic esophageal cases, with the current case representing the shortest documented interval. Among 11 reviewed cases, 7 originated from Asian countries, with fish bones constituting the predominant FB type.

Table 1.

A review of past literature on ingested FBs migrating into the thoracic cavity

Year	Authors	Country	Patient (Sex/Age)	Special conditions	Main symptoms	FBs location	Interval from impaction to retrieval	Surgical approach	Lung management	Object type	Perioperative complications	Outcome
1982[3]	So et al.	USA	M/42	None	Hemoptysis	Right upper lobe	5 months	Thoracotomy	Pneumotomy	Chicken bone	Lung abscess	Recovered
1988[11]	Radford & Wells	UK	M/51	None	Chest pain, weight loss, fever	Right upper lobe	4 months	Thoracotomy	Partial lung resection	Pork bone	Mediastinitis, pulmonary inflammatory pseudotumor	Recovered
2011[25]	Ozkan et al.	Turkey	F/20	None	Asymptomatic	Left lower lobe	10 days	Thoracotomy	Pneumotomy	Sewing needle	Asymptomatic colon perforation	Recovered
2012[7]	Pan et al.	China	M/57	None	Recurrent pneumonia, hemoptysis	Right upper lobe	5 years	Thoracotomy	Right upper lobectomy	Fish bone	Lung abscess	Recovered
2012[18]	Fujino et al.	Japan	F/55	None	Chest pain, fever	Right lower lobe	31 days	VATS	Ligament division	Fish bone	None	Recovered
2013[20]	Matsuda et al.	Japan	F/60	None	Fever, odynophagia	Right upper lobe	3 days	Thoracotomy	Right upper lobectomy	Fish bone	Lung abscess	Recovered
2014[14]	Mogi et al.	Japan	F/85	Dementia	Cough, anorexia	Right upper lobe	At least 1 week	Thoracotomy	Right upper lobectomy	Denture	Esophagopulmonary fistula, lung abscess	Recovered
2015[16]	Lee et al.	South Korea	M/50	None	Chest pain, hemoptysis	Right upper lobe (near subclavian artery)	4 weeks	VATS + Stent	Wedge resection	Fish bone	Pseudoaneurysm, pneumothorax	Recovered
2015[12]	Tan et al.	China	M/47	None	Dry cough, fever	Right upper lobe	20 days	VATS	Abscess debridement	Fish bone	Lung abscess	Recovered
2016[15]	Ekeke et al.	USA	M/33	Psychiatric history (recurrent ingestion)	Pleuritic chest pain	Left pleural apex	8 days (post-laparotomy)	VATS	Decortication	Ballpoint pen	Pleural effusion, empyema	Recovered
2025	Ours	China	M/58	None	Chest pain, dysphagia	Left upper lobe (near subclavian artery)	1.5 days	VATS	Suture	Fish bone	Pneumothorax	Recovered

This study provides the first documentation of intrathoracic FB migration following esophageal perforation, visualizing the dynamic process through sequential CT imaging. The object demonstrated sequential progression: initial esophageal wall penetration within 1 h post-ingestion (Fig. 1A), complete transmural passage confirmed by rigid esophagoscopy at 9 h, and subsequent migration into the thoracic cavity within 28 h (Fig. 1C). The failure of esophagoscopic visualization, despite the FB’s presence on initial CT, can be explained by its rapid and complete transmural migration. The linear configuration of the fish bone likely promoted swift penetration, resulting in its complete embedding within the esophageal submucosa or muscular layers by the time of endoscopy. Although preoperative CT indicated migration by showing the FB tip abutting the pleura, it could not definitively discern whether the FB still partially resided within the lumen or had already fully penetrated the esophageal wall. This detailed visualization of the migratory pathway, akin to the process observed in thyroid gland migration [4], underscores the utility of sequential CT imaging in tracking FB movement. Potential contributing factors may involve horizontal orientation [5], esophageal peristaltic forces [6], and mediastinal displacement during respiratory cycles [7]. Additionally, the potential influence of intraoperative manipulation cannot be ruled out.

The scenario where rigid esophagoscopy fails to visualize the FB, as encountered in this case, presents a critical management dilemma, indicating probable complete transmural migration. Preoperative CT becomes indispensable for confirming migration and anatomical localization. If the migrated FB is not immediately accessible upon surgical exploration, adjunctive real-time techniques may be considered. The utility of these techniques depends on the FB’s nature and location. Endoscopic ultrasound (EUS) is highly sensitive for detecting both radiolucent and radiopaque FBs, but its efficacy is limited to lesions adjacent to the esophagus within the mediastinum; it cannot reach FBs that have migrated distally into the pleural cavity or diaphragm. In contrast, C-arm fluoroscopy is effective for intraoperative localization of radiopaque objects (e.g., metallic fragments [8]), regardless of their specific intrathoracic position, provided they are sufficiently dense. These observations suggest that when transmural migration is suspected, cautious decision-making is required. After thorough vascular risk assessment, options include EUS localization for mediastinal FBs [9], mucosa-incising retrieval techniques [10], or direct thoracic surgical consultation for VATS, which may offer viable management alternatives. Thoracic surgical collaboration remains essential for complex cases requiring multidisciplinary intervention.

The present case manifested solely as pneumothorax, contrasting with delayed presentations (>2 weeks) frequently complicated by mediastinitis [11] or pulmonary abscess formation [3, 7, 12]. The development of pneumothorax in this case is likely attributable to pleural injury caused by the migrating FB. The fish bone, upon traversing the esophageal wall and entering the thoracic cavity, presumably created a pleural defect. While this defect may have been initially subtle, positive-pressure ventilation during rigid esophagoscopy could have facilitated air entry into the pleural space, analogous to cases where ventilation exacerbates an undetected pleural breach [13]. Subsequent persistent air leakage from this defect then led to the progressive chest tightness and increased lung collapse observed on sequential CT imaging after the procedure. Timely diagnosis proves critical for complication mitigation, though clinical challenges persist. Linear sharp objects (e.g., fish bones, needles) often induce minimal symptoms during migration, potentially delaying medical consultation. For instance, in Tan et al.‘s report [12], the patient alleviated FB sensation by swallowing rice, delaying diagnosis until cough and fever prompted CT detection of a right upper lobe abscess two weeks later. Furthermore, cognitive impairments in specific populations (e.g., dementia [14], psychiatric disorders [15]) may further exacerbate diagnostic delays.

The selection of imaging investigations profoundly impacts early detection efficacy. Current literature demonstrates a high level of underdiagnosis when relying solely on endoscopy or X-ray [3, 7, 11, 12, 16]. For radiolucent objects, such as fish bones, the false-negative rate for X-ray alone can reach up to 85%, thus CT has already been recommended as the first-line investigation for suspected perforation or high-risk cases [17]. Sequential CT monitoring may further enhance complication prediction and optimize therapeutic strategies in migration cases.

Migration trajectory critically determines complication severity. In the case reported by Fujino et al. [18], a fish bone migrated through the pulmonary ligament into the lung without involving major vessels or pleura, thereby avoiding the formation of pseudoaneurysms or pneumothorax. Conversely, Lee et al. [16] reported a case of a fish bone migrating to the right upper lobe, resulting in a subclavian artery pseudoaneurysm that required stent repair, while Huang et al. [19] described fatal sequential vertebral and subclavian pseudoaneurysms despite multiple interventions. In the present case, the trajectory of the FB in close proximity to the left subclavian artery and the visceral pleura placed the patient at potential risk for both immediate and delayed injuries. The immediate concerns included vascular injury or pneumothorax [16] from direct perforation, as manifested in this case. Had the FB remained in situ, delayed complications such as mediastinitis [11], abscess formation [3, 12, 20], or the development of an esophagopulmonary fistula [14] could have ensued, underscoring the critical need for timely intervention. The necessity of preoperative collaboration with the vascular surgery team and meticulous intraoperative dissection in such high-risk trajectories is therefore emphasized. Postoperative CTA surveillance ensured no delayed vascular injury, underscoring the importance of multidisciplinary planning.

Future research directions should prioritize FB migration prediction and trajectory analysis. Although preliminary attempts have been made to use artificial intelligence technology to assist endoscopy [21] or X-ray imaging [22, 23] in identifying FB impactions or assessing risks, there is still a lack of in-depth exploration of CT imaging in this regard. For instance, the development of predictive models to determine whether a FB remains partially within the esophageal lumen based on CT imaging features, or the assessment of critical structure risk stratification along migration paths, could offer more precise decision-making support for surgical approach selection, complication prevention, and perioperative management.

Among the cases we reviewed, VATS has gradually become the mainstream approach for FB removal since the report by Fujino et al. in 2012 [18], which may be closely related to its minimally invasive nature and clear surgical field. A comprehensive review of 71 VATS-treated mediastinal/thoracic FB cases demonstrated 100% retrieval success with favorable prognoses [24], supporting VATS as the preferred option for hemodynamically stable patients as an alternative to traditional thoracotomy. The decision to proceed with VATS in this case, despite the vascular proximity, was based on the following considerations. Preoperative CTA showed no definitive evidence of pseudoaneurysm formation. During the initial VATS exploration, the absence of significant perivascular hematoma or arterial bulging supported the decision to continue with the minimally invasive approach. Nevertheless, the feasibility of conversion to thoracotomy remained an essential precaution against potential vascular injury.

In the present case, the FB was retrieved without deliberate esophageal exposure. This was based on the short ingestion-to-surgery interval (28 h), the linear configuration of the fish bone, and its preoperative CT localization—almost entirely within the thoracic cavity—which collectively suggested a small esophageal perforation with minimal surrounding inflammation. This approach aimed to minimize surgical trauma and preserve the local blood supply of the esophagus, potentially reducing the risk of impaired healing at the perforation site.

Conclusions

This case demonstrates early VATS intervention for thoracically migrated esophageal FBs, after rigid esophagoscopic visualization fails. Sequential CT comparisons provided a clear visualization of migration dynamics and recovery. Rapid multidisciplinary management played a vital role in facing underlying life-threatening complications, and resulted in the shortest reported interval in literature. Future research might prioritize predictive models for migration risks preoperatively, ensuring timely and minimally invasive management of migrated esophageal FBs.

Abbreviations

FB

Foreign body

CT

Computed tomography

VATS

Video-assisted thoracic surgery

CTA

Computed tomography angiography

Author contributions

HC C contributed to the study design and drafted. HH M provided intraoperative imaging data. Z C conducted the literature search. ZZ C and NC X assisted in investigation. SH Z acquired the data and wrote the article. ZT W performed data analysis and gave the final approval of the version to be submitted. All authors read and approved the manuscript.

Funding

Not applicable.

Data availability

The findings of this study are supported by data included in the article. Additional datasets and materials can be obtained from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Helsinki Declaration II and was approved by the Clinical Research Ethics Committee of the First Affiliated Hospital, Zhejiang University School of Medicine (approval number: [2025B] IIT Ethics Approval No. 0296).

Consent for publication

The patient provided written informed consent for the publication of his clinical details and images in this study by signing the consent form.

Competing interests

The authors declare no competing interests.