Preoperative CT-guided Fiducial Marker Placement for Surgical Localization of Pulmonary Nodules

Abstract

Purpose

To assess the technical success and complication rates of CT-guided fiducial marker placement for the localization of pulmonary nodules and to assess the surgical localization failure rate.

Materials and Methods

This was a single-center, retrospective analysis of consecutive patients who underwent CT-guided fiducial marker placement procedures between 2014 and 2020. End points included the technical success of the fiducial marker placement, procedural complications, and the surgical localization failure rate. A two-sample t test and a Fisher exact test were used to compare continuous and categorical variables, respectively. Multivariate logistic regression was used to identify independent risk factors for complications.

Results

A total of 198 preoperative CT-guided fiducial marker placement procedures were performed in 190 patients (mean age, 64 years ± 12 [standard deviation]; 121 women) to localize 205 nodules (mean size, 10 mm ± 4; mean distance to the pleura, 10 mm ± 9). The technical success rate was 98.5% (195 of 198). There were no major complications. A total of 202 nodules were resected during 193 procedures performed 5 days ± 13 after the fiducial marker placement (range, 0–123 days). Surgical localization failure occurred in one patient (0.5%). Of the resected nodules, 146 were lung cancers, 26 nodules were metastases, two were carcinoid tumors, and 28 were benign.

Conclusion

The CT-guided fiducial marker placement of pulmonary nodules was safe, effective, and resulted in a low surgical localization failure rate.

Keywords: CT, Percutaneous, Thorax, Lung

Summary

CT-guided fiducial marker placement was a safe and effective technique for localizing pulmonary nodules for preoperative planning.

Key Points

• ■ The technical success rate of CT-guided fiducial marker placement was 98.5% (195 of 198).

• ■ The rate of complications was low (6% of patients had a pneumothorax that required chest tube placement).

• ■ The surgical localization failure rate after CT-guided fiducial marker placement was low (n = 1 patient, 0.5%).

Introduction

Advances in CT technology, the greater use of CT in emergency departments, the implementation of lung cancer screening programs, and the more frequent depiction of nodules on any given scan (1) have resulted in an increased number of detected small pulmonary nodules that require surgery for diagnostic resection and treatment. Minimally invasive techniques for surgical resection, including video-assisted thoracoscopic surgery (VATS) and robot-assisted thoracoscopic surgery (RATS), are associated with decreased morbidity and reduced recovery time and hospital stay without compromising patient care when compared with open thoracotomy (2–4). However, the surgical access ports used for VATS and RATS impede bimanual palpation of the lung parenchyma that facilitates nodule identification. Failed resection or conversion to open thoracotomy are the consequences of nodule localization failure (5). The need for the accurate localization of nodules has created an opportunity for interventional radiologists to aid in minimally invasive surgical resection. Preoperative localization of small, subsolid, or deep pulmonary nodules is performed through the image-guided placement of liquid or metal markers in or adjacent to the lung nodule (6).

The percutaneous insertion of rodlike, gold fiducial markers is an easily implementable technique that is well tolerated by patients and can be performed on a separate day to surgical excision. The aim of this study was to evaluate the success rate of CT-guided percutaneous fiducial marker placement for nodule localization prior to surgical resection. The secondary aims were to evaluate the complication rates of fiducial marker placement and the rate of successful fiducial marker–guided, minimally invasive resection.

Materials and Methods

This Health Insurance Portability and Accountability Act–compliant, retrospective cohort study was performed at an academic medical center. Institutional review board approval was obtained for chart review, and the need for written informed consent was waived.

Study Sample

Over a 6-year period (September 18, 2014, to September 17, 2020), a search of the thoracic interventional radiology database identified 198 CT-guided fiducial marker placement procedures performed in 190 patients to localize 205 nodules prior to surgical resection, all of which were included in our study. The decision to request preoperative nodule localization was at the discretion of the referring thoracic surgeon. The thoracic surgeons typically referred patients with suspicious or growing nodules that were potentially difficult to identify at surgery because of nodule size, attenuation, or location (with both central nodules and peripheral nodules being amenable to localization). A prior study reported on the feasibility and safety of CT-guided fiducial marker placement in the initial 19 patients, who are included in this current cohort (7).

Preprocedure Image Analysis

The last diagnostic CT scan obtained prior to CT-guided placement was reviewed by two radiologists, an attending interventional thoracic radiologist with 5 years of experience and a thoracic radiology fellow (S.M. and N.E.F., respectively), blinded to clinical details and outcomes. The nodule location, attenuation (solid, ground glass, part solid, or part cystic), size (mean diameter), and distance from the nodule margin to the nearest pleural surface were determined in consensus.

CT-guided Fiducial Localization

The fiducial marker placement procedure was performed in the radiology department under CT guidance, either by or under direct supervision of one of five attending interventional thoracic radiologists with 2–28 years of interventional experience (S.M., F.J.F., M.C.P., J.O.S., and A.S.). The technique used has been described previously (6,7) and is similar to that of our percutaneous nodule biopsy technique (8). All but one procedure was performed under conscious sedation with a combination of midazolam and fentanyl. One patient required monitored anesthesia care because of the inability to tolerate the procedure under conscious sedation. Briefly, 3 × 0.8–mm gold fiducial markers (Best Medical International) were deployed through the lumen of a 19-gauge coaxial needle to lie adjacent to or within the targeted nodule (Fig 1). The positioning of the needle tip away from a vessel was confirmed prior to deployment. Ideally, fiducial markers were either placed next to the deep margin of the nodule, thereby sandwiching the nodule between the free edge of the pleura and the fiducial markers, or the fiducials were placed on either side of the nodule, bracketing the nodule. The mean number of fiducial markers placed were two (range, one to four) per nodule. The relative location of the fiducial marker to the nodule was directly communicated to the surgeon following placement. If requested by the surgeon, a percutaneous, image-guided needle biopsy was performed during the same session as the placement.

Figure 1:

Images in a 50-year-old woman with an incidentally detected, enlarging, 10-mm, left upper lobe, part-solid nodule proven to be adenocarcinoma (100% lepidic). (A) Preliminary axial CT image shows a left upper lobe nodule (arrow). (B) The coaxial introducer needle was advanced through the subcutaneous tissue to the pleura. (C) The needle was advanced deeply to the nodule. (D) The first 3-mm gold fiducial marker was deployed, and the needle was retracted. (E) The second fiducial marker was deployed with the nodule sandwiched between fiducial markers and the pleura. (F) Postprocedure radiograph demonstrates the two fiducial markers in the left upper lung (arrow). (G) Radiograph of the specimen with two fiducial markers in the specimen.

The patient was immediately turned puncture side down on removal of the coaxial needle. For patients undergoing same-day surgery, one bedside chest radiograph was obtained in the recovery room to rule out a large pneumothorax and to help confirm the presence of fiducial markers prior to transport to the preoperative area for surgical resection. All other patients underwent routine postprocedure supine or prone bedside chest radiography at 1 hour and upright radiography at 3 hours prior to discharge (7).

Postprocedure Image Analysis

The final images of the procedure were analyzed by two radiologists, an attending interventional thoracic radiologist with 5 years of experience and a thoracic radiology fellow (S.M. and N.E.F., respectively), and the length of the needle tract through the lung and the distance between the fiducial markers and the targeted nodule were measured. The presence of a pneumothorax was demonstrated on both the procedural CT scan and the postprocedure radiographs.

In cases in which additional imaging was performed between the day of the placement procedure and surgery, these were reviewed to assess for migration of the fiducial markers.

Surgical Resection

The decision to perform the VATS or RATS resection of the targeted nodule the same day as the placement procedure or at a later date was at the discretion of the thoracic surgeon. Intraoperative on-table fluoroscopy was used to localize the fiducial markers, and radiographs were obtained of the resection specimens to help confirm the excision of the fiducial markers. The option of intraoperative frozen section examination was available. The decision to proceed with same-day segmentectomy or lobectomy following wedge resection was made on the basis of the findings of the intraoperative frozen-section and/or nodal sampling.

Outcomes

The primary outcome was technical success, defined as the successful deployment of at least one fiducial marker in or within 10 mm of the targeted nodule measured on the final images of the procedure. Complications were assessed retrospectively and classified according to the Society of Interventional Radiology guidelines (9). The secondary outcome was the surgical localization failure rate, which was defined as the conversion of a planned VATS or RATS to an open thoracotomy because of the difficulty in localizing the nodule despite fiducial markers. The decision was made to report the technical success rate and complication rate of fiducial marker placement and the surgical localization rate on the basis of the number of procedures rather than the number of nodules, as it more accurately reflects patient outcomes.

Statistical Analysis

Continuous variables, such as age, lesion size, and distance to pleura, are presented as means ± standard deviations, with ranges in parentheses. A two-sample t test was used to compare continuous variables. Categorical variables, such as sex, ethnicity (taken from electronic health records), history of malignancy, nodule location, and occurrence of a complication, are presented as counts or counts with percentages in parentheses. A Fisher exact test was used to compare categorical variables. The factors yielding a P value less than .2 in the univariate analysis were used as variables in multivariate logistic regression to identify independent risk factors for complications. Statistical analysis was performed by using statistical software (Stata, version 13.1), with statistical significance set at a P value less than .05.

Results

Demographics

CT-guided fiducial marker placement was performed in 190 patients (mean age, 64 years ± 12; 121 women) for the preoperative localization of 205 nodules (Table 1).

Table 1:

Patient and Nodule Characteristics

Technical Success

Three of 198 fiducial marker placement procedures were unsuccessful, and the technical success rate was 98.5% (Table 2). One procedure was aborted prior to placing a needle in the lung because of the difficulty in distinguishing an enlarging 13-mm ground-glass nodule from the adjacent atelectasis. Subsequent successful fiducial marker placement was followed by the surgical excision of focal fibrosis. Another patient developed self-limited hemoptysis resulting in the abortion of the procedure before fiducial marker deployment. The patient was discharged home but was subsequently lost to follow-up. In a third patient, two fiducial markers were inadvertently deployed into the pleural space, and the procedure had to be interrupted to place a chest tube for a pneumothorax. The result of a concurrently performed needle biopsy was suspicious for malignancy, so the decision was made to perform a middle lobectomy rather than a wedge resection. There was a confirmed finding of adenocarcinoma at lobectomy.

Table 2:

Procedure Outcomes

Concurrent Percutaneous Needle Biopsy

In 15 procedures, a needle biopsy was performed at the same time as the fiducial marker insertion. In 80% of these combined procedures (12 of 15), the biopsy result was concordant with the final diagnosis (Table 3). In the remaining three procedures, the biopsy result was nondiagnostic: Two nodules were malignant at surgical resection, and one nodule was yet to be resected by the end of the study period.

Table 3:

Comparison of Biopsy Results and Final Diagnoses

Complications

In four procedures, at least one of the fiducial markers migrated into the pleural space immediately after deployment (including the procedure referenced to above; Table 4). There was no significant difference in the length of the needle tract through the lung between procedures in which a fiducial marker fell into the pleural space (33 mm ± 16) and those in which they did not (43 mm ± 20; P = .3). In two of the four procedures, the introducer needle crossed a fissure, and in the other two, the needle caused unrecognized pleural tenting, resulting in fiducial marker deployment into the pleural space. Three of the four procedures had one remaining fiducial marker in place that allowed for the fluoroscopic identification and the successful surgical excision of the nodule. Of the 140 patients who underwent surgery on a later day to the placement procedure, 66 (47%) underwent additional imaging between the day of the placement procedure and surgery, none of which identified the migration of the fiducial markers (Fig 2).

Table 4:

Procedure Complications

Figure 2:

Images in a 61-year-old woman with a history of pancreatic cancer and a slowly growing right lower lobe nodule. (A) Preprocedural axial CT image shows a 7-mm solid nodule in the right lower lobe (arrow). (B) Axial CT image obtained at the conclusion of the procedure shows two fiducial markers bracketing the nodule (arrow) with minor adjacent parenchymal hemorrhage (arrowhead). (C) Axial CT image obtained 15 months later demonstrates the two fiducial markers to be unchanged in location.

One procedure was complicated by the fiducial marker embolization into a small branch of the right inferior pulmonary vein at the time of placement; the fiducial marker eventually lodged in the left ventricle, where it stayed until the patient’s death 588 days later, without causing dysrhythmia or other symptoms.

A pneumothorax occurred in 21% of procedures (42 of 198); 6% of patients (11 of 198) required chest tube placement and admission for less than 48 hours (minor adverse event). The risk of a pneumothorax was independent of the size of the nodule (10.2 vs 10 mm; P = .9) and the distance from the nearest pleura (9.4 vs 10.6 mm; P = .4). There was no increased risk of pneumothorax when more than one nodule was localized during the same procedure (12.5% vs 21.6%; P > .99). There was also no increased risk of pneumothorax (21.2% vs 13.3%; P = .7) or the need for a chest tube (5.5% vs 6.7%; P = .6) when fiducial marker placement alone or biopsy and fiducial marker placement was performed. Self-limited hemoptysis occurred in four patients, which was managed conservatively (minor adverse event).

Surgical Procedure

A total of 202 localized nodules were removed during 193 surgical procedures, 5 days ± 13 after the fiducial marker placement (range, 0–123 days) (Table 5). In 183 procedures, one nodule was resected. In eight procedures, two nodules were resected during the same procedure, and in one procedure, three nodules were resected. In one patient with fibrotic nonspecific interstitial pneumonia, at the time of surgery, the nodule was found to be in a location that would necessitate a basilar segmentectomy or lobectomy that the patient could not tolerate, so the resection was aborted. Three patients underwent resection of two localized nodules on separate occasions. One patient underwent resection of three localized nodules on three separate occasions. One patient underwent two surgeries to resect four localized nodules.

Table 5:

Surgical Approach

Forty-nine patients (25.4%) underwent surgery the same day as the placement procedure. In the remaining 140 patients, surgery was performed on a later date (n = 139) or was yet to be performed by the end of the study period (n = 1); one patient underwent successful placement; however, resection has been delayed because of the SARS-CoV-2 pandemic and the stability of the nodule.

The surgical approach was VATS in 65.3% (126 of 193), RATS in 30.6% (59 of 193), planned thoracotomy in 1% (two of 193), and initial VATS converted to thoracotomy in 3.1% (six of 193) because of dense adhesions present in three procedures, desaturation in one procedure, multiple concurrent wedge resections in one procedure, and the failure to localize the nodule in one procedure. Therefore, the surgical localization failure rate was 0.5% (one of 193).

In 80.7% of procedures (163 of 202), a wedge resection was performed to resect the nodule; in 2% (four of 202), a segmentectomy was performed; and in 11.4% (23 of 202), the wedge resection was followed by completion lobectomy. Eleven patients, with a total of 12 nodules (5.4%), underwent lobectomy without a prior wedge because the nodule location was unsuitable for wedge resection (n = 7), because the biopsy performed concurrently with fiducial marker placement was already positive for malignancy (n = 2), because two lesions were resected (n = 1), or because of the intraoperative discovery of malignant lymphadenopathy (n = 1).

Final Diagnoses

The final diagnosis was lung cancer in 69.8% (146 of 202), metastases in 12.9% (26 of 202), carcinoid tumor in 1% (two of 202), and benign in 13.9% (28 of 202; Table 6).

Table 6:

Final Diagnosis of Lung Nodule

Discussion

The use of minimally invasive, parenchymal-sparing surgery for the resection of small nodules has caused a paradigm shift, with the emergence of both image-guided and bronchoscopic techniques to aid in localization. To our knowledge, this is the largest study of preoperative CT-guided gold fiducial marker placement, a technique reported by only a couple of other institutions (10,11). Other studies have reported the use of percutaneously placed gold fiducial markers for nodule localization prior to radiation (12–16); however, the precise marker placement in or near the targeted nodule may not be as essential for radiation guidance as it is for surgical resection.

Our study confirms that CT-guided fiducial marker placement is a safe and easily performed method of preoperative nodule localization. The primary end point of successful deployment was achieved in 98.5% of procedures when performed by five radiologists with wide-ranging experience in interventional thoracic procedures. Similar success rates are reported for hookwire (98%), microcoil (98%–100%), and lipiodol localization (99%) (17). However, these techniques often have a longer learning curve. Dendo et al (18) reported initial successful hookwire localization at years 1–3 in 92% cases, improving to 100% at years 4–6. Technically challenging localization methods have required modifications to improve operator success. For example, initial descriptions of microcoil placement involved placing one end of the microcoil in the pleural space and the other within a nodule. However, this could be complicated by microcoil retraction into the pleural space or chest wall (19). A modified version, in which all of the coil is deployed into the lung parenchyma without pleural marking, is easier to perform and similar to that of our approach (6,20).

The inert nature of fiducial markers allows surgery to be performed at a different date, and even in a different hospital, providing flexibility between radiology and operating schedules, which is one of the advantages of this localization technique. In our study, the procedure was performed on a different date than the surgery in just less than 75% of the procedures. Same-day, often immediate, resection can cause scheduling constraints but is a necessity with other techniques, particularly with methylene blue dye, radioisotope localization, and hookwires or coils that extend into the pleura or chest wall.

Our technique and equipment for lung biopsy and the placement of fiducial markers are very similar and allow the option for both procedures to be performed during the same session without an additional increase in complications or a decrease in technical success. A study of patients undergoing percutaneous fiducial marker placement prior to stereotactic radiation therapy also found no additional risk of a major complication when biopsy was concurrently performed (16). In addition, the needle entry point, angle, and trajectory for the placement of the fiducial marker are independent of the planned surgical approach (8). In contrast, other localization methods for surgery with a visible tract from the pleura to the nodule, as with methylene blue injection, hookwire, or microcoil with pleural marking, require the radiologist to use the shortest perpendicular path to a nodule for localization. This limitation in approach may be challenging and can cause failure in the proper placement of the wire or coil (6,19,21,22). In contrast, fiducial markers and microcoils without pleural marking may be placed from any safe pathway through the lung.

Our complication rates were low and comparable with those of other localization methods (17). Our pneumothorax rate of 21.2% was similar to that of biopsies performed at our institution of 24%, and our chest tube rate was slightly higher (5.6% vs 2.2%). Our hemoptysis rate was also very similar (2% vs 2.2%) (23). We had one case of embolization of a fiducial marker into the left ventricle via a peripheral pulmonary vein, without any long-term adverse effects. Although systemic embolization presents risks and should be prevented by avoiding the transgression of pulmonary venous structures, four other cases of embolization of fiducial markers have been described at other centers, without clinical consequences, including cases of migration to the heart and a coronary artery (10,11,24). We had four cases of inadvertent deployment of one or more fiducial marker(s) into the pleural space at the time of placement. In three of these cases, one fiducial marker was successfully placed in the lung, enabling the successful localization and resection of the nodule. Therefore, as long as this complication is identified at the time of the procedure, the procedure can be salvaged by placing additional fiducial markers within the lung parenchyma. This is far lower than the failure rate of the hookwire technique because of dislodgment of the wire, reaching up to 47% (25). A recent study on the microcoil localization of small peripheral nodules reported that, at the time of surgery, the microcoil had dislodged in 38 patients (7.6%), including 10 microcoils that completely retracted into the lungs and 28 microcoils that were completely withdrawn from the lungs (26). There were no cases of late fiducial marker movement after successful placement in our study, which has been reported in up to 30% of fiducial markers placed bronchoscopically, likely because of expectoration through the tract to the airway (27).

The surgical localization failure rate with fiducial markers occurred in one case (0.5%), resulting in the conversion of a planned VATS or RATS to an open thoracotomy. Our results are similar to those of other techniques. Successful VATS rates range from 92% to 96% for hookwire localization (17,21), 97%–100% for microcoil (17,19,21), 100% for dye localization (22), 97%–100% for radiotracer localization (28–30), and 99% for lipiodol localization (17). In a recent study on electromagnetic transthoracic nodule localization, 7% required conversion to a thoracotomy, although the sample size was small (31).

In our study, nearly 14% of nodules resected were benign. More than one nodule was resected in 11 patients in our study, with even one patient undergoing resection of four nodules, again underlining the importance of parenchymal-sparing surgery. Limiting the size of resection is important in cases of benign cause, multiple lesions, and in lung cancer with poor pulmonary reserve. Minimizing the size of pulmonary resection while maintaining adequate surgical margins can be facilitated by preoperative localization.

A limitation of this study included its retrospective design. This study was also a single-institute study, with the procedure performed or supervised by one of five interventional thoracic radiologists (S.M., F.J.F., M.C.P., J.O.S., A.S.). There was no control arm, limiting our ability to make comparisons between the use of CT-guided fiducial marker placement and no placement or for comparison of success rates and complications of localization through different techniques. A randomized trial that compares CT-guided fiducial marker placement with alternative methods of placement and includes economic considerations may be warranted in the future.

In conclusion, the CT-guided fiducial marker placement of pulmonary nodules is a safe, effective procedure that results in a low surgical localization failure rate.