Unexpected Complications 25 Years after Coil Embolization for Pulmonary Arteriovenous Fistula

Abstract

An 87-year-old woman who had undergone coil embolization 25 years ago for pulmonary arteriovenous fistula, which was diagnosed following repeated cerebral infarction, presented with massive hemoptysis. The coils migrated and were excreted in stool following hemoptysis during long-term follow-up. Although the technical success rate of coil embolization for pulmonary arteriovenous malformations is extremely high, and coil embolization-related complications are rare, little is known about the long-term complications. We herein report the clinical course of our case, review previous reports related to coil migration as a long-term complication, and discuss the associated mechanism.

Introduction

Owing to technological advances, coil embolization for pulmonary arteriovenous malformations (PAVMs) can be safely performed with few complications. The technical success rate of coil embolization for PAVM is extremely high (≥99%), and coil embolization-related complications are rare (≤0.65%) (1). However, little is known about its long-term complications.

We herein report a case of coil excretion through the stool following hemoptysis during long-term follow-up, with a review of the pertinent literature.

Case Report

An 87-year-old woman presented to the emergency department with severe respiratory failure due to massive hemoptysis. She had a history of chronic right paresis and dysphasia owing to repeated cerebral infarction caused by bilateral pulmonary arteriovenous fistulae, for which she had undergone coil embolization 25 years previously. She had age-associated changes in the skeletal structure, including kyphosis and scoliosis, which were the pathological background disorders that predisposed her to aspiration. She had no other comorbidities and had never smoked. Six months before admission, she was intubated because of massive hemoptysis. Intensive care, including red blood cell transfusion, hemostatic agents, and rest, enabled extubation within three days, and she was treated for aspiration pneumonia.

On admission, the patient's height and body weight were 141.3 cm and 33.1 kg, respectively. Her blood pressure, heart rate, and oxygen saturation level were 149/103 mmHg, 104 beats/min, and 94% on a 15-L/min reservoir mask, respectively. Multiplex polymerase chain reaction testing of a nasopharyngeal swab for severe acute respiratory syndrome coronavirus 2 and other common respiratory viruses was negative. She was diagnosed with aspiration pneumonia caused by Pseudomonas aeruginosa and urinary tract infection caused by extended-spectrum beta-lactamase-producing Escherichia coli. Table 1 presents the patient's data on admission. The patient's hemoglobin level was 7.6 mg/dL, which was significantly lower than her usual level of approximately 11.0 mg/dL.

Table 1.

Laboratory Findings on Admission.

Hematology			Serology
RBC	23×104	/μL	Antinuclear antibodies	Negative
Hemoglobin	7.6	g/dL	p-ANCA	Negative
Hematocrit	22.3	%	c-ANCA	Negative
WBC	6,630	/μL	Anti-GBM antibodies	Negative
Neutrophils	81.8	%	BNP	60.5	pg/mL
Lymphocytes	12.1	%	QFT test	Negative
Platelets	20.3×104	/μL	Aspergillus GM antigen	Negative
			Anti-MAC antibodies	Negative
Biochemistry
Albumin	3.1	g/dL	Coagulation
Total bilirubin	0.5	mg/dL	PT	12.8	s
AST	22	U/L	APTT	31	s
ALT	11	U/L	Fib	433	mg/dL
γ-GTP	15	U/L	D-dimer	2.3	μg/mL
LDH	235	U/L
ALP	105	U/L	Culture
BUN	12.4	mg/dL	Sputum	Pseudomonas aeruginosa
Creatinine	0.40	mg/dL	Urine	ESBL-producing Escherichiacoli
CRP	4.58	mg/dL

ALP: alkaline phosphatase, ALT: alanine aminotransferase, ANCA: antineutrophil cytoplasmic antibodies, APTT: activated partial thromboplastin time, AST: aspartate aminotransferase, BNP: brain natriuretic peptide, BUN: blood urea nitrogen, CRP: C-reactive protein, ESBL: extended-spectrum beta-lactamase, FDP: fibrinogen degradation products, Fib: fibrinogen, GBM: glomerular basement membrane antibodies, γ-GTP: γ-glutamyl transpeptidase, INR: international normalized ratio, LDH: lactate dehydrogenase, PT: prothrombin time, QFT: quantiferon, RBC: red blood cell, WBC: white blood cell

We compared the images taken at another hospital six months before admission (Fig. 1a: actual size; Fig. 1b: magnified view) with those taken at our hospital during admission (Fig. 1c: actual size; Fig. 1d: magnified view). In the magnified view, the coils over the right diaphragm appeared to have lost their volume, with two consecutive coils reduced to one. Coil fragments were observed in the small intestine (Fig. 1e). The right medial branches were the predicted source of hemorrhaging, as seen on the computed tomography scan (Fig. 1f, g).

Figure 1.

(a-d) Images six months before (a: actual size, b: magnified view) and during admission (c: actual size, d: magnified view). The coils over the left diaphragm are unchanged, whereas the coils over the right diaphragm appear to have lost their volume. (e) Coil fragments are visible in the small intestine (red circle). Tracheal deviation and silhouetting of the right heart border and right hemidiaphragm suggest atelectasis in the right middle lobe. Coronal computed tomography images on admission (f: mediastinal window, g: lung window) revealing atelectasis and consolidation around the coils, particularly in right segment 5b.

As initial treatment after hemoptysis, the patient was managed in the right-side down-decubitus position to protect the nonbleeding lung, as spillage of blood into the nonbleeding lung may impair gas exchange by blocking the airway with clot or filling the alveoli with blood. As a result, blood components were deposited into the right upper lobe during the course of the case, temporarily enhancing the shadow in the right upper lung field. Hemoptysis subsided with two units of red blood cell transfusion, hemostatic agents, and rest; oxygen administration was terminated on day 5. With positional drainage, the shadows showed steady improvement, and antibiotics administered for aspiration pneumonia and urinary tract infection were effective.

Bronchoscopy performed on day 13 to determine the source of hemorrhaging and confirm the protrusion of the coils into the bronchus showed no obvious abnormal findings (Fig. 2). Two coils were excreted in the stool on the same day (Fig. 3).

Figure 2.

Bronchoscopic imaging and chest radiography during the procedure. (a) Bronchoscopy showing no localized active hemorrhaging or protrusion of the coil. (b) Fiberscope positioned near the site of coil embolization.

Figure 3.

(a, b) Coils excreted through the patient's stool 25 years after coil embolization.

Chest radiograph showed improvement of atelectasis and disappearance of the coils from the intestine, and the patient was discharged on day 23. The patient has been followed up in the outpatient clinic, and to date, the hemoptysis has not recurred.

Discussion

We encountered a case in which coils were excreted via the stool following hemoptysis. It had been 25 years since the patient had undergone coil embolization for pulmonary arteriovenous fistula. Her symptoms improved with conservative treatment. Based on the radiological findings, we speculated that endobronchial coil migration was the cause of the hemoptysis and coil excretion through the stool.

PAVMs are abnormal direct connections between a pulmonary artery and a pulmonary vein. Since the first report of PAVM in 1897 (2), therapeutics for the disease have progressed. The first successful surgical treatment was a pneumonectomy performed in 1942 (3). With further advancements in thoracic surgery, local resection became the preferred treatment by 1959 (4). Although surgical resection has proven to be effective in the treatment of PAVMs, the invasiveness of the procedure, along with the loss of viable lung tissue, prompted the development of transcatheter embolization therapy. Embolization has been considered the standard treatment for PAVMs since 1983 (5,6). Because of ethical concerns, there have been no randomized controlled trials specifically examining the efficacy of embolization therapy for PAVMs. However, observational studies indicate that embolization reduces morbidity (7).

Since the introduction of embolization therapy, a variety of devices have been utilized, including silicone balloons, coils, and vascular plugs. Long-term complications mostly include reperfusion or recanalization of PAVMs. Wu et al. reported that PAVMs had relapsed in 15 of 43 patients by 12 months post-surgery (8).

Endobronchial coil migration as a long-term complication is rare, with only 13 reported cases (Table 2) (1,9-18). In 7 of the 13 patients, coil migration occurred more than a year after coil embolization, with a median period of 2.3 (1.2-12) years. Most patients presented with hemoptysis and cough, with two patients expectorating the coils. In the migrated endovascular coil cases, common management strategies are observation, coil retrieval with loop cutters or bronchoscopy, and surgical intervention, including lobectomy and segmentectomy.

Table 2.

Previous Cases of Migrated Coils of Pulmonary Arteriovenous Malformations.

References	Age/sex	Primary indication for	Time from coiling	Treatment
		coil embolization	to detection of migration
(11)	18/M	Saccular aneurysm of the	6 weeks	Lobectomy
		pulmonary artery
(12)	59/M	Racemose hemangioma	2.3 years	BAE and right S2 segmentectomy
				for complete resection of a
				racemose hemangioma
(13)	40/M	Aortopulmonary collaterals	Unknown	Unknown
		due to tetralogy of Fallot
(14)	38/F	Pulmonary artery aneurysm	6 months	Lobectomy
(10)	56/M	Arteriovenous	10 years	Lobectomy
		malformation
(15)	32/F	Saccular aneurysm of the	2 weeks	Conservative
		pulmonary artery
(15)	50/M	Tumor invasion to	3 weeks	Conservative
		pulmonary artery
(16)	33/M	Cryptogenic hemoptysis	4.5 years	Additional super-selective BAE
(16)	67/F	Cryptogenic hemoptysis	2 years	Forceps extraction after split by a
				loop cutter
(1)	44/M	Hereditary hemorrhagic	12 years	Thoracoscopic right basal
		telangiectasia		segmentectomy
(17)	32/M	Bilateral mycotic aneurysms	1.2 years	Forceps extraction under rigid
		due to endocarditis		bronchoscopy
		complicated by pulmonary
		and septic emboli
(18)	54/M	Pulmonary artery	1 month	Pneumonectomy
		pseudoaneurysm
(9)	45/F	Hereditary hemorrhagic	8 years	Bronchoscopy-guided removal
		telangiectasia

BAE: bronchial artery embolization

Complications caused by coil migration after pulmonary embolization include fistulization, erosion into the bronchus, migration into the contralateral pulmonary vasculature, pulmonary hypertension, pneumothorax, and migration to the heart (9). The exact mechanism underlying coil migration remains unclear. Arterial wall weakness, adjacent bronchus erosion, and failure of the coils to maintain their spiral shape may contribute to migration (10). Considering the histology, we speculated that the effects of the coils on the structural changes in the blood vessels and bronchi are similar to the microvasculopathy observed in chronic thromboembolic pulmonary hypertension involving the pulmonary arterioles, venules, and capillaries. Although its pathophysiology is complex, and much remains to be learned, chronic thromboembolic pulmonary hypertension is known to be caused by not only chronic mechanical obstruction from residual thrombotic material but also severe remodeling and microvasculopathy of the distal pulmonary arteries (0.1-0.5 mm) (19).

In our case, local mechanical irritation associated with repeated inflammation, including repeated aspiration pneumonia and a history of respiratory failure due to hemoptysis requiring intubation six months before admission, might have contributed to the weakening of the tissue around the coil and aggravation of the coil migration. Furthermore, the 25 years of occlusion of the PAVM caused remodeling of the vessel, and the discrepancy between the coil and vessel diameters resulted in the protrusion of the coil outside the vessel. Since the patient was at a high risk of aspiration, the coils were extracted not via her airway but rather via the stool.

To our knowledge, this is the first case report of migrated coils being excreted through the stool following hemoptysis during long-term follow-up. Careful observation of coil migration is warranted, as complications can occur unexpectedly, as in the present case.

The authors state that they have no Conflict of Interest (COI).