Pre-operative CT localization for patients with subsolid opacities expecting video-assisted thoracoscopic surgery—single center experience of fluorescent iodized emulsion and hook-wire localization technique

Abstract

Objective:

To describe our clinical experience of using combination of fluorescent iodized emulsion, indocyanine green & lipiodol, and hook-wire in pre-operative CT localization of patients with subsolid lung lesions prior to video-assisted thoracoscopic surgery (VATS).

Methods:

A retrospective review between June 2018 and July 2019 of consecutive Chinese patients whom underwent VATS for subsolid lung lesions with pre-operative CT localization done with combination of fluorescent iodized emulsion and hook-wire technique in a tertiary hospital (Tuen Mun Hospital, Hong Kong SAR). The duration and complications related to the localization procedure were recorded The clinical records, operative findings and pathology reports were retrieved from the hospital electronic clinical management system.

Results:

Combination fluorescent iodized emulsion with hook-wire enabled accurate localization and resection of all subsolid lung lesions in VATS. No major complications were reported.

Conclusion:

Combination of fluorescent iodized emulsion and hook-wire placement under CT guidance is a simple, safe and cost- effective procedure that enabled accurate localization and resection of subsolid nodule in VATS.

Advances in knowledge:

VATS has been the mainstay for indeterminate pulmonary nodules for diagnostic and/or curative purpose. The main problem that surgeons may encounter during operation is the difficulty in locating the target lesion particularly for subsolid lesions. Many pre-operative localization methods have been developed in this regard. With the novel technique that we described, we were able to overcome disadvantages of most described methods.

Objectives

Video-assisted thoracoscopic surgery (VATS) has been increasingly utilized for indeterminate subsolid nodules, that include pure ground glass opacity (GGO) and part solid lesions, for both diagnostic and/ or curative purposes. These subsolid nodules are either non-palpable or difficult to detect without pre-operative localization via thoracoscopy. In this report, we describe and evaluate our experience using combination of fluorescent iodized emulsion, indocyanine green (ICG) & lipiodol, with hook-wire placement for localizing subsolid nodules prior to VATS.

Background

With the widespread use of CT, incidental solid and subsolid pulmonary nodules has been increasingly detected, this pose considerable diagnostic and management challenge for clinicians and radiologists. The Fleischner Society and British Thoracic Society published guidelines and management recommendations for these lesions that include interval CT scans, additional imaging such as positron emission tomography (PET).^1,2 When nature of these nodules remains indeterminate or if there is evidence of enlargement upon imaging follow-up, excisional biopsy or surgical resection may become necessary.

VATS has been the mainstay for indeterminate pulmonary nodules for diagnostic and/or curative purpose. Detection of a lesion during VATS depends on visual confirmation on the television monitor and/or palpation by the cardiothoracic surgeon. The main problem related to thoracoscopic resection of these subsolid nodules is the difficulty in locating the target nodule, especially if the lesion is a small, deeply seated, or pure ground-glass in nature. Pre-operative localization facilitate intraoperative localization of the lesion, minimizes the volume of lung parenchyma that needs to be sacrificed with the target lesion and also facilitates the pathologists to find the target lesions on frozen sections. Many methods have been developed in this regard, including the use of hookwires, microcoils, radiotracers, radiocontrasts, and dyes.^3–6 Each method has its own advantages and disadvantages. For instance, dislodgement of hookwires prior to or during operation are not uncommon; there is radiation exposure issues and unclear resection margin regarding the use of radiocontrast agents, etc. Before using the ICG–lipiodol mixture, we have tried conventional dyes with methylene blue as pre-operative localization materials with non-satisfactory results as dark coloration of the lung surface due to anthracosis render the methylene blue, i.e. site of the lesion difficult to detect during VATS. Every radiologist/surgeon chooses the method of pre-operative localization most suitable for him, and there is no consensus of which is the most perfect technique. To overcome the disadvantages of each individual method, our center adopted a combination of fluorescent iodized emulsion, prepared by mixing ICG with lipiodol, together with hook-wire placement since June 2018 with satisfying results.

Findings & procedure details

This is a retrospective review of consecutive patients between June 2018 and July 2019 referred for pre-operative CT localization of subsolid nodules in a single tertiary center in Tuen Mun Hospital, Tuen Mun, HKSAR. There was a total of 19 patients with 19 subsolid nodules during the study period. A total of 13 patients had combination of fluorescent iodized emulsion with hook-wire placement. The clinical, radiological findings, intraoperative records and pathological findings were reviewed from the electronic patient record system of the Hospital Authority, HKSAR. All patients were referred from a team of cardiothoracic surgeons.

There were a total of 8 males and 11 females, with age ranging from 45 to 70 years old (mean 60.2). The average size of the lung lesions ranged from 7.7 to 37.5 mm (mean 16.0 mm, located from subpleural to a distance of 17.8 mm from the nearest pleural surface (mean 7.0 mm).

Localization technique

We made use of a spiral end hookwire (SOMATEXR Lung Marker System^®, Somatex Medical Technologies GmbH, Germany), which is a wiring system using coaxial loading cannula technique. It consists of a puncture cannula (15 cm length and 1.20 mm diameter) and a marking wire which is pre-loaded in a loading cannula. The distal end of the marking wire is of spiral shape which would anchor in the lung parenchyma. The target of the tip of the marking wire was usually close to or inside the pulmonary nodule to be operated on, which is determined based on location of the lesion and discussion between the operating surgeons and the radiologists.

Under local anesthesia, with aseptic technique and CT guidance, the radiologist inserted the puncture cannula through the thoracic wall until it reached the target lesion. After removing the introducer needle from the puncture cannula, a mixture of 0.2 to 0.4 ml of fluorescent iodized emulsion (mixture of ICG with lipiodol in 1 to 9 or 2 to 8 ratio), was injected via the puncture cannula. After CT confirmation of the adequacy of the amount of dye, the loading cannula containing the spiral wire was introduced through the puncture cannula for final deployment of the localizing hookwire. The patient was then sent back to the ward and wait for the availability of the operation theatre usually around 5 h after the localization procedure. During VATS wedge resection, a NIR thoracoscope was used to localize the nodules and to determine with the surgical margin for resection (Figure 1). The resected specimen was examined under the NIR thoracoscope to further confirm the adequacy of resection. The surgical specimens were all sent for pathological evaluation.

Figure 1.

(a) Different patient with intra operative findings using NIR thoracoscope during uniportal VATS with fluorescence merge image showing the target lesion. (b) The specimen was illuminated ex-vivo under NIR thoracoscope, this also guide the adequacy of resection margin for the cardiothoracic surgeons. NIR,near infrared; VATS, video-assistedthoracoscopic surgery.

Results

The localization procedure was technically successfully in all patients. In six cases, fluorescent iodized emulsion was injected alone without combined hook-wire placement, data were also included and arrange in chronological order on Table 1 in order to allow readers to have a better overview of the pathological diagnosis of the subsolid lesions. In 13 of the 19 procedures (68.4%), the localizing material (fluorescent iodized emulsion alone or with hookwire) was inserted through the lesion. In all the remaining six cases (31.6%), the emulsion and hookwire were placed in the adjacent lung tissue within 5 mm distance from the lesion. There are no major complications reported. All patients developed trace to mild <5% pneumothorax that did not require chest tube placement. Another minor complication encountered in some patients was minimal needle tract parenchymal hemorrhage with new patchy ground glass opacities developed, but none of the patients developed hemoptysis. Some patients also complained of wound pain and mild cough. The average time of the localization procedure was ~42.6 min. After the localization procedure, all patients were sent back to the ward for observation while waiting for the availability of the operation theatre.

Table 1.

Patients’ characteristics and details of lesions. # CT coronary angiogram. *GGO type: Pure type, P1 (regular border), P2 (irregular border); Mixed type, M1 (<50% solid component), M2 (>50% solid component)

Case	Age	Sex	Smoking history (Ex/Y/N)	Presentation	Tumor location and size (mm)	GGO type *	Depth from pleural surface (mm)	Technique	ICG fluroescence	Pathological diagnosis
1	57	M	N	Cough	RLL (7 × 4.5)	P1	7.4	Icg/lipoidol	Clear	Interstitial fibrosis
2	70	M	Ex	Incidental CT findings due to chest injury	RUL (23 × 15)	M2	1	Icg/lipoidol + hookwire	Clear	Adenocarcinoma EGFR mutation not available
3	47	M	Y	Incidental finding on CTCA#	RML (10 × 10)	P1	17.8	Icg/lipoidol	Clear	Adenocarcinoma EGFR positive
4	58	M	N	Incidental finding on CTCA#	LLL (10.4 × 8.8)	P	3	Icg/lipoidol	Clear	Adenocarcinoma EGFR positive
5	57	F	N	Incidental CXR findings due to chest injury	RUL (8.4 × 7.3)	M1	8.5	Icg/lipoidol	Clear	Fibrosis with no malignancy
6	64	F	N	Known CA lung with tumor excision. New lung lesion detected on surveillance CT	RUL (40 × 35)	M2	2	Icg/lipoidol + hookwire	Clear	Adenocarcinoma EGFR positive
7	65	F	N	Known CA sigmoid in remission. New lung lesion on surveillance CT	RLL (17 × 13)	P	subpleural	Icg/lipoidol + hookwire	Clear	Atypical adenomatous hyperplasia
8	69	M	Y	CXR shadow	RUL (17 × 13)	P	14	Icg/lipoidol + hookwire	Clear	Scar with no malignancy
9	56	F	N	CXR shadow	RLL (18.2 × 17)	M1	11	Icg/lipoidol + hookwire	Clear	Adenocarcinoma
10	67	F	N	Known CA lung with tumor excision. New lung lesion on PET-CT follow-up	RUL (10.6 × 9.4)	P2	3	Icg/lipoidol	Clear	Adenocarcinoma EGFR negative
11	64	F	N	Known CA breast in remission. Lung lesion on survillence CT	LUL (16 × 13)	M2	Subpleural	Icg/lipoidol	Clear	Adenocarcinoma EGFR negative
12	61	F	N	Incidental CT findings due to CXR suspected opacity	RUL (18 × 13.5)	M2	21	Icg/lipoidol + hookwire	Clear	Adenocarcinoma EGFR negative
13	61	M	N	Incidental CT findings due to CXR suspected opacity	RUL (20 × 8.4)	M2	5.6	Icg/lipoidol + hookwire	Clear	Adenocarcinoma EGFR negative
14	45	F	N	Incidental CT findings due to chest pain	RML (8.5 × 7)	M1	11.5	Icg/lipoidol + hookwire	Clear	Adenocarcinoma EGFR negative
15	67	F	N	Known CA breast in remission. Incidental CT findings due to CXR suspected opacity	RLL (30 × 28.3)	P2	4	Icg/lipoidol + hookwire	Clear	Adenocarcinoma EGFR negative
16	66	M	N	Lung shadow on CXR upon health check-up	LUL (31.5 × 20.1)	M2	Subpleural	Icg/lipoidol + hookwire	Clear	Minimal invasive Adenocarcinoma
17	53	F	N	Incidental CT findings due to unrelated pleural-based lesion	LUL (14.6 × 10.8)	P1	20	Icg/lipoidol + hookwire	Clear	Adenocarcinoma in-situ
18	58	F	N	Incidental CT findings done for investigation of hematuria	RLL (19 × 16.2)	M1	2	Icg/lipoidol	Clear	Adenocarcinoma EGFR positive
19	58	M	N	Incidental CT findings due to CXR suspected opacity	RUL (18.9 × 16)	P2	2	Icg/lipoidol + hookwire	Clear	Adenocarcinoma EGFR status not available

CXR, chest radiograph; EGFR, epidermal growth factor receptor; GGO, ground glass opacity; ICG, indocyanine green;LLL, left lower lobe; LUL, left upper lobe; RLL, right lower lobe; RML, right middle lobe; RUL, Right upper lobe.

According to the intraoperative records, all lesions were successfully localized during operation with clear visualization by the NIR scope. No dislodgement of the spiral end hookwire was reported.

All patients underwent successful VATS, with an average operation time of 73.4 min (excluding two cases, one with subsequent completion lobectomy and the other with thoracotomy for concomitant pleural lesion). All patients were discharged from the hospital without major complications with an average of hospital stay of ~3.69 days (range from 1 to 7 days)

The pathological diagnosis for the 19 patients (include cases with pre-operative localization with ICG & lipiodol emulsion alone) were primary lung adenocarcinomas (68.4%, n = 13); minimal invasive adenocarcinoma (5.3%, n = 1); atypical adenomatous hyperplasia (5.3%; n = 1); adenocarcinoma in-situ (5.3%, n = 1); interstitial fibrosis/scar (15.7%, n = 3).

Among patients with subsolid nodules whom subsequently confirmed to be malignancy, 13 (92.9%) were non-smoker (female n = 9; male n = 4), and 1 patient (male) was reported to be an ex-smoker whom quitted smoking for more than 10 years. Characteristics of subsolid nodules subsequently proven to the malignancy; average size is 17.6 mm (ranging from 9.6 to 37.5 mm), irregular margin n = 6 (40%), presence of solid components n = 10 (66.6%).

Discussion

Small pulmonary nodules are being increasingly encountered with the widespread use of multidetector CT scans. Pulmonary nodules are further classified into solid, GGO and part solid nodule. GGO, are generally defined as nodules with increased attenuation in the lung on CT with preserved bronchial and vascular markings (Figure 2a); and subsolid nodules are regarded as presence of additional solid nodule components other than normal vascular or bronchial structures.

Figure 2.

(a) CT scan taken with patient in prone position showing a ground glass opacity (black arrow) measuring about 2 cm with slight irregular outline located at left lower lobe about 2 cm deep to the pleural surface. Of note, the opacity does not obscure a normal pulmonary vessel (white arrow) that is seen running through the lesion. (b) CT scan taken with patient in supine position after fluorescent iodized emulsion (black arrow), and hook-wire (arrowhead) placement in pre-operative localization of ground glass opacity. Paper-thin layer of pneumothorax noted after procedure (white arrow). Wedge excision was subsequently performed in the operation theatre and frozen section came back to be adenocarcinoma.

Different professional bodies such as Fleischner Society and British Thoracic Society published guidelines regarding follow-up and management of incidental lung nodules detected on CT images based on nodule size, density and patient risk factors.^1,2 Details are beyond the scope of discussion of this article. Persistent subsolid nodules (including both GGO and part solid nodule) pose additional diagnostic and therapeutic challenge for respiratory physicians, radiologists and cardiothoracic surgeons because of they have a much slower growth rate but carry a much higher risk of malignancy. Kakinuma et al reported that 10% of solitary ground glass nodules 5 mm or smaller, will grow and 1% will develop into invasive adenocarcinomas or minimally invasive adenocarcinomas. They suggested solitary GGO 5 mm or smaller should be rescanned 3.5 years later to look for development of a solid component.⁷ Flesichner society 2017 consensus guideline suggests annual follow-up CT up to 5 years for solitary GGO more than 6 mm in size. Incidental CT findings of a pulmonary nodule create a lot of patient anxiety and patients are often uncomfortable with the prospect of waiting for follow-up CT scan. For instance, for solitary part-solid nodules 6 mm or more in size, the Fleischner guideline recommends first CT at 3–6 months to confirm persistence of the lesion and then annual CT follow-up for 5 years.¹ With increasing use of CT scans, the “snow-ball” effects of following up these incidental lung lesions, will create an additional burden to the limited resources of healthcare system particularly the public sector. We recommend multidisciplinary discussion of the management of these nodules, with expertise input from respiratory physicians, cardiothoracic surgeons, radiologists and oncologists, to derive a practical and sustainable module to cater the foreseeable increasing number of follow-up cases. Stakeholders in charge of healthcare budget from the government should allocate additional resources and have long-term planning ahead of anticipated problems in the future.

To our knowledge, there is no consensus regarding whether to have pre-operative biopsy of subsolid nodules before surgery. Some surgeons recommend pre-operative biopsy in order to have better treatment planning for patients, others prefer direct surgical resection because of relative low invasiveness of VATS in removing a subsolid nodule. Pre-operative fine needle aspiration biopsy by either CT-guided, endobronchial ultrasound and virtual bronchoscopy is an effective approach in experienced hands, but it has important limitations for very small nodules and ground-glass lesions due to potential problems with inadequate sampling and false-negative results.^8–11 Currently, minimally invasive surgery with lung sparing technique enables both accurate pathological diagnosis and therapeutic treatment in selected patients with little morbidity.^12,13 However, it is difficult to visually identify and palpable subsolid nodules even if they are peripherally located. Furthermore, accurate pre-operative localization of indiscernible lesions is important to save operative time to find the obscure targets, and to avoid thoracotomy conversion in VATS surgery.

ICG is a NIR fluorescent dye with fluorescence absorption of 800 nm. Although ICG has amphiphilic properties (both hydrophilic and lipophilic), it must be dissolved in water to yield an aqueous solution for injection in patients.^14–16 In addition, ICG is the only NIR fluorescent dye currently approved for clinical use in patients to detect the sentinel lymph node, tumor margin, cardiovascular imaging, or tissue perfusion¹⁷; it can also be used to localize pulmonary nodules pre-operatively premixed with lipiodol.¹⁸ A note of caution is that ICG contains sodium iodide and is contraindicated in patients with known iodine hypersensitivity such as iodinated contrast medium commonly used for CT scan. And allergy and anaphylactic reaction may occur as other drugs administrate to the body.

NIR fluorescence imaging using ICG during VATS has several advantages compared with other conventional dyes such as methylene blue. First, NIR light is only visible under thoracoscopy with fluorescence imaging system; thus, ICG does not obscure the surgical field. Second, ICG has higher tissue penetration than that of methylene blue and hence, ICG is more useful to localize deep-seated pulmonary nodules. Our experience concurs with that of Rho et al¹⁸ that low concentration and volume of ICG should be used for localization of pulmonary nodules as at high concentration, quenching effect could occur that obscures the resection margin.

Lipiodol, an oil-based radiopaque contrast agent, enables direct CT visualization of the amount of fluorescent iodized emulsion injected at/around the lesion; and also enables the operator to detect presence of any leakage of the emulsion along the needle tract into the pleural cavity; with multiplanar reformed images constructed on CT, the lipiodol also enhances the three-dimensional orientation of the exact location of the lesion during thoracoscopy. In addition, lipiodol also reduces the diffusion of ICG solution into the adjacent tissue, and thereby providing a more precise resection margin.

The ICG lipiodol forms an emulsion that would enable retaining of ICG around the lesion for localization, thus the procedure can be done in the CT suite instead of inside the operation theatre. In return, the valuable OT session could be utilized more efficiently in our institute.

We also utilized a spiral end hookwire in our pre-operative localization technique. The hookwire other than its primary function for localization, can also double-up as a bulk forming agent, so that the surgeons could have a better assessment on the resection margin of the lesion. In addition, this avoids failed localization by using ICG fluorescence alone as this has been reported due to either insufficient or leakage of ICG into the thoracic cavity.¹⁴

Some centers utilize electromagnetic navigation bronchoscopy for pre-operative localization of lung nodules. However, this requires expertise training, utilization of valuable operation theater time and is equipment-dependent. The consumable cost of ICG/lipiodol with hookwire for localization in our institute was estimated to be around $5000 HK dollars, this is much lower than electromagnetic navigation bronchoscopy for localization (estimated around $20,000 HK dollars) that include steering probe, forceps/brush and fiducial marker etc.

Lung cancer is the leading cause of cancer death in the world. Almost all our patients subsequently diagnosis with adenocarcinoma (including the case with minimal invasive adenocarcinoma) were never smokers (92.9%). There were more female patients with subsolid pulmonary nodules subsequently confirmed to be malignant diseases than male patients. Among patients subsequently diagnosed to have malignancy, according to the clinical notes, none of them has presenting chest symptoms such as cough or hemoptysis; the subsolid lung lesions were incidental found on CT done for other clinical reasons, such as chest pain with CT coronary angiogram detected lung base lesions in five patients (35.7%); chest radiograph detected shadow in six patients (42.9%), in which the chest radiograph shadows were often unrelated to the subsolid lesions detected on subsequent CT thorax; detected on surveillance imaging CT or PET-CT for tumors in remission in three patients (21.4%). Typical CT morphological features associated with malignancy include margin speculation (40%) and internal solid components (66.6%). Our findings concur with the findings published in other studies.^19–21

Pre-operative CT localization with combination of fluorescent iodized (ICG & lipiodol) emulsion and hookwire was a safe and cost-effective procedure. None of the reported cases developed adverse reactions or major complications.

Conclusion

Our preliminary results have shown that the use of fluorescent iodized emulsion and hookwire under CT guidance is a simple, safe and cost-effective procedure that enabled accurate localization and resection of subsolid nodule in VATS.