Eur Radiol DOI 10.1007/s00330-015-3762-8

CHEST

C-Arm Cone-Beam CT Virtual Navigation-Guided Percutaneous Mediastinal Mass Biopsy: Diagnostic Accuracy and Complications Hyungjin Kim 1,2,3 & Chang Min Park 1,2,4 & Sang Min Lee 1,2 & Jin Mo Goo 1,2,4

Received: 19 December 2014 / Revised: 7 March 2015 / Accepted: 1 April 2015 # European Society of Radiology 2015

Abstract Objectives To assess the usefulness of C-arm cone-beam computed tomography (CBCT) virtual navigation-guided percutaneous mediastinal mass biopsy in terms of diagnostic accuracy and complication rates. Methods Seventy-eight CBCT virtual navigation-guided percutaneous mediastinal mass biopsies were performed in 75 patients (M:F, 38:37; mean age, 48.55±18.76 years). The procedural details, diagnostic sensitivity, specificity, accuracy and complication rate were investigated. Results Mean lesion size was 6.80±3.08 cm, skin-to-target distance was 3.67±1.80 cm, core needle biopsy rate was 96.2 % (75/78), needle indwelling time was 9.29±4.34 min, total procedure time was 13.26±5.29 min, number of biopsy specimens obtained was 3.13±1.02, number of CBCTs performed was 3.03±0.68, rate of lesion border discrimination from abutting mediastinal structures on CBCT was 26.9 % (21/78), technical success rate was 100 % (78/78), estimated effective dose was 5.33±4.99 mSv, and the dose area product was 12,723.68±10,665.74 mGy⋅cm2. Among the 78 biopsies, 69 were malignant, 7 were benign and 2 were indeterminate. Diagnostic sensitivity, specificity and accuracy for the

diagnosis of malignancies were 97.1 % (67/69), 100 % (7/7) and 97.4 % (74/76), respectively, with a complication rate of 3.85 % (3/78), all of which were small pneumothoraces. Conclusions CBCT virtual navigation-guided biopsy is a highly accurate and safe procedure for the evaluation of mediastinal lesions. Key Points • CBCT virtual navigation-guided percutaneous mediastinal biopsy is highly accurate • CBCT virtual navigation-guided percutaneous mediastinal biopsy is a safe procedure • Mediastinal vascular injury can be avoided under CBCT virtual navigation guidance Keywords Cone-beam computed tomography . Image-guided biopsy . Mediastinal mass . Respiratory system . Sensitivity and specificity

Abbreviations CBCT Cone-beam computed tomography 3D Three-dimensional DLBL Diffuse large B-cell lymphoma DAP Dose area product

* Chang Min Park [email protected]

Introduction 1

Department of Radiology, Seoul National University College of Medicine, 101, Daehangno, Jongno-gu, Seoul 110-744, Korea

2

Institute of Radiation Medicine, Seoul National University Medical Research Center, 101, Daehangno, Jongno-gu, Seoul 110-744, Korea

3

Aerospace Medical Group, Air Force Education and Training Command, Jinju, Korea

4

Cancer Research Institute, Seoul National University, 101, Daehangno, Jongno-gu, Seoul 110-744, Korea

Patients with mediastinal masses often require biopsy procedures, as treatment strategies can differ significantly according to their pathology—e.g., tumours such as thymoma are often treated surgically, whereas lymphoma or metastatic lesions would be treated with chemotherapeutic agents. For imaging-guided biopsy of mediastinal lesions, computed tomography (CT) is widely used, since it allows detailed

Eur Radiol

visualization of the lesion and neighbouring mediastinal structures as well as precise localization and documentation of the biopsy needle through cross-sectional imaging with good inplane resolution [1, 2]. However, there are several disadvantages to using CTguidance, including its inability to provide real-time imaging guidance; the small gantry bore size of the CT system, leading to operator discomfort, which can lower operator performance; and limitations in image plane orientation owing to the limited degree of gantry tilt allowed with the CT system, resulting in difficulties in selecting the optimal route to the target [3–5]. Another modality that has been utilized for imaging-guided biopsy is the CT fluoroscopy system, which allows for real-time imaging guidance, but the other inherent problems of CTguided biopsy, as well as radiation exposure to the operator, still remain [6]. Within the past decade, the C-arm cone-beam CT (CBCT) system has been introduced into the field of interventional radiology [7], allowing radiologists to perform interventional procedures in a comfortable open gantry system under real-time fluoroscopic guidance, resulting in accurate planning with three-dimensional (3D) CT data [8]. In addition, the CBCT virtual navigation software program provides a virtual needle pathway for targeting lesions, helping operators more easily navigate the needle into the target after initially determining the skin entry site and destination target based on preprocedural CBCT data [8]. Indeed, owing to this system’s ability to facilitate more accurate and safer needle placement, it has already been shown to provide excellent diagnostic accuracy for biopsy of the lung [2, 4, 8–16] (Table 1) and bone lesions [17, 18]. In this study, we hypothesized that the mediastinum would also be a suitable candidate, as it consists of vital organs and complex vascular structures surrounded by the lungs and bones, and its anatomic window for a safe percutaneous needle approach is limited. To the best of our knowledge, however, CBCT-guided mediastinal mass biopsy has not been investigated in the literature to date. Thus, the purpose of our study was to assess the usefulness of the C-arm CBCT virtual navigation system for percutaneous biopsy of mediastinal masses with regard to diagnostic accuracy and complication rates in 75 patients with mediastinal masses.

Materials and methods This retrospective study was approved by the institutional review board of the Seoul National University Hospital, with waiver of patient informed consent.

Study population From April 2010 to October 2014, 81 consecutive CBCT virtual navigation-guided percutaneous mediastinal mass biopsies were performed in 78 patients in our hospital. From among this group, we excluded three patients, for the following reasons: both ultrasound and CBCT were used for needle guidance (n=2), and the biopsy needle targeted the sternuminvading portion of the lesion and was not advanced into the mediastinum (n=1). For the latter case, the main bulk of the lesion was located in the mediastinum, and it was deemed easier to biopsy a superficial sternal portion of the lesion than a deep-body lesion such as that located in the mediastinum. Thus, ultimately, 75 patients who underwent 78 biopsy procedures comprised our study population (mean age, 48.55± 18.76 years; range, 1–83 years). In our study, three patients underwent repeat biopsy for the same mediastinal lesion, and these repeat biopsies were considered as unique initial cases. Among the 75 patients, 38 were male (mean age, 49.42± 21.15 years; range, 1–83 years) and 37 were female (mean age, 47.65±16.19 years; range, 20–77 years). Thirty-eight patients had a single mediastinal lesion, and the other 37 had more than two lesions, with or without conglomeration. Biopsy procedure Mediastinal mass biopsies were performed using two different CBCT systems: the AXIOM Artis dTA/VB30 flat-panel detector with a 2048 × 1538 element (Siemens, Erlangen, Germany) (n=12); and the Allura Xper FD20 flat-panel detector with a 2480×1920 element (Philips Healthcare, Best, the Netherlands) (n =66), with virtual guidance (iGuide, Siemens Medical Solutions; XperGuide, Philips Healthcare). All procedures were performed using the coaxial technique by or under the supervision of one experienced chest radiologist (C.M.P., with 8 years of experience in image-guided procedures). Prior to the procedure, previously acquired diagnostic contrast-enhanced chest CTs were carefully reviewed to determine the most appropriate needle route (Fig. 1a). The mean length of time from acquisition of the diagnostic CT image to the biopsy procedure was 12.73 ± 12.33 days (range, 0– 61 days). Patients in the intervention suite first underwent pre-procedural CBCT (Fig. 1b), and these CT images were then transferred to dedicated workstations (Leonardo with DynaCT, Siemens Medical Solutions, Erlangen, Germany; XtraVision, Philips Healthcare). The three-dimensional CT images were reconstructed using multiplanar reformations in axial, coronal and sagittal planes. Thereafter, using the virtual navigation software program, operators determined the safest and most effective route to obtain an appropriate sample and to avoid vessels and other mediastinal structures (Fig. 1c). Automatic vertical alignment from the skin entry site to the target lesion (bull’s-eye

Retrospective Lung nodule

Retrospective Lung nodule Retrospective Lung nodule

Retrospective Lung nodule Retrospective Lung nodule

2010

Cheung et al.a [10] 2011 Choi et al.c [11] 2012

2012 2012

2012

2013

2014

2014

Hwang et al.b [12]

Lee et al.d [15] Choi et al.b [4]

Braak et al.e [9]

Choo et al.f [8]

Jiao et al.g [2]

Lee et al.h [14]

Lung nodule

Biopsies were targeted for lung nodules ≤2 cm

h

g

f

e

d

Mean lesion size was 2.7±1.7 cm (range, 0.5–13.0 cm)

Mean lesion size was 4.6±3.0 cm (range, 0.6–15.0 cm)

Biopsy was performed for lung nodules ≤1 cm

Mean lesion size was 32.5 mm (range, 3–93 mm)

Mean lesion size was 37±23 mm (range, 8–120 mm)

Mean lesion size was 30±16 mm (range, 8–86 mm)

b

c

Procedures were performed for lung nodules ≤3 cm

a

Retrospective Lung nodule

Retrospective Lung nodule

Retrospective Lung nodule

Prospective

Retrospective Lung nodule

2010

Jin et al.a [13]

99.6 (1148/1153)

98.2 (108/110)

100 (107/107)

N/A

N/A 100 (173/173)

N/A N/A

100 (27/27)

100 (71/71)

95.7 (733/766)

96.7 (88/91)

96.7 (58/60)

90.0 (63/70)

93.1 (54/58) 96.8 (91/94)

90.6 (48/53) 95.8 (69/72)

94.1 (16/17)

97 (35/36)

100 (323/323)

100 (17/17)

100 (38/38)

100 (14/14)

100 (36/36) 100 (69/69)

100 (21/21) 100 (27/27)

88.9 (8/9)

100 (25/25)

91.7 (77/84)

Virtual guidance software

iGuide, XperGuide

iGuide

XperGuide

XperGuide

Pneumothorax: 25.4 (18/71) Not used Hemoptysis: 14.1 (10/71) Chest pain: 1.4 (1/71) Pneumothorax: 14.8 (4/27) N/A Hemoptysis: 3 (1/27) Pneumothorax: 25.7 (19/74) Not used Pneumothorax: 16.2 (16/99) Not used Hemoptysis: 2.0 (2/99) Subcutaneous emphysema: 1.0 (1/99) Pneumothorax: 25.5 (24/94) Not used Pneumothorax: 31.8 (55/173) N/A Hemoptysis: 14.5 (25/173)

Complication rate (%)

Chest pain: 1.2 (2/173) Pneumothorax: 20.2 (17/84) Hemoptysis: 1.2 (1/84) 98.0 (96/98) Pneumothorax: 6.5 (7/107) Hemoptysis: 5.6 (6/107) 97.2 (105/108) Pneumothorax: 12.7 (14/110) Hemoptysis: 6.4 (7/110) 97.0 (1056/1089) 23.9 (276/1153)

95.7 (90/94) 98.2 (160/163)

93.2 (69/74) 97.0 (96/99)

92.3 (24/26)

98.4 (60/61)

Biopsy target Technical success Sensitivity (%) Specificity (%) Accuracy (%) rate (%)

Year Study design published

Recent studies on CBCT-guided biopsy procedures in the field of chest intervention

First author

Table 1

Eur Radiol

Eur Radiol Fig. 1 C-arm cone-beam (CB) CT virtual navigation-guided percutaneous mediastinal mass biopsy in a 71-year-old man. (a) On contrast-enhanced diagnostic CT taken prior to the procedure, a paraaortic dumbbell-shaped mass (arrow) is noted. (b) Preprocedural CBCT image shows the paraaortic mass (arrow). (c) The needle entry site, approach technique and needle advancement length are determined on pre-procedural CBCT. (d) A needle (not shown) is introduced into the bull's-eye view of the fluoroscopic image, in which the skin entry site (pink circle) is superimposed on the target (green circle). (e) CBCT image shows a coaxial introducer needle targeting the lesion using a parasternal approach

view) was established using virtual colour spots on the fluoroscopic image (Fig. 1d). A 17-gauge coaxial introducer was then inserted into the target lesion in the bull’s-eye view, and procedural cone-beam CT was performed to check the exact needle location (Fig. 1e). If the needle tip was correctly located within the target lesion, 18-gauge cutting needle biopsy (Stericut, TSK Laboratory, Japan) or aspiration biopsy (Westcott, Angiotech USA) was conducted through a coaxial introducer at the operator’s discretion. After the specimens were obtained, postprocedural CBCT was performed to identify procedurerelated complications. A more detailed description of the process of CBCT virtual navigation-guided biopsy can be found in Choo et al.’s study [8].

Data collection For each procedure, operators documented the lesion’s characteristics as well as the procedural records in our radiology database. With regard to lesion characteristics, the lesion size (largest diameter), location in the mediastinum (anterior, middle or posterior mediastinum) [19], and lesion feature (solid, cystic or mixed solid and cystic) were described. With respect to procedural records, skin-totarget distance, biopsy method (core needle or aspiration biopsy), patient position (supine or prone), coaxial needle indwelling time, total procedure duration (defined as the length of time from local anaesthesia injection to the end of

Eur Radiol

post-procedure CBCT), number of biopsies obtained, number of CBCT data acquisitions, technical success or failure, and total dose area product (fluoroscopy dose plus CBCT dose) were documented. Technical success was defined as appropriate location of the coaxial needle within the target lesion on procedural CBCT images and adequate tissue sampling on visual inspection. The dose area product in milligrays was converted into effective dose using a conversion factor obtained in a separate phantom study. As previously described by Choi et al. [4] and Choo et al. [8], the conversion factors used to convert dose area product to effective dose were 0.17 mSv/Gy⋅cm 2 with the AXIOM Artis system and 0.45 mSv/Gy⋅cm 2 with the Allura Xper FD20 system. In addition, the mediastinal needle approach (parasternal, transsternal, transpulmonary, subxiphoid, suprasternal or paravertebral) (Figs. 2 and 3) and the availability for discriminating the lesion border from abutting mediastinal structures such as great vessels on pre-procedural CBCT images (available or unavailable) were determined by one of the authors (H.K with 5 years of experience in radiology) after reviewing each case. Fig. 2 Transpulmonary approach in a 77-year-old man with a middle mediastinal mass. (a) Contrast-enhanced CT revealed a lobulated enhancing mass in the aortopulmonary window. (b) The patient underwent pre-procedural cone-beam (CB) CT for biopsy planning. (c) Needle trajectory using a transpulmonary approach was determined. (d) Inserted needle tip did not reach the mass on procedural CBCT. Therefore, needle was advanced an additional 1 cm to the target (not shown), and the biopsy was performed successfully

Assessment of diagnostic accuracy The final diagnosis of a mediastinal lesion was determined according to the following criteria: (a) If the patient underwent surgical resection, the surgical pathology report decided the final diagnosis; (b) if the result of the biopsy revealed a specific malignant or benign pathology such as thymoma or sarcoidosis, it was accepted as the final diagnosis; (c) in cases of non-specific benign pathology (e.g. negative for malignancy, chronic inflammation, etc.), follow-up CTs and medical records assisted the diagnosis, with lesions considered benign when they showed a 20 % or greater decrease in diameter on follow-up CT [12]; (d) if the nonspecific benign lesions did not meet ‘criterion (c)’, the final diagnosis was listed as indeterminate. Indeterminate lesions were not included in the calculation of diagnostic accuracy. In our study, thymoma was classified as a malignancy, as it has been considered clinically malignant in several studies based on its invasive and metastasizing properties [20, 21]. All biopsy-related complications such as pneumothorax or hematoma were also recorded. Pneumothorax was evaluated with post-procedural CBCT

Eur Radiol Fig. 3 Paravertebral approach in a 62-year-old woman with a posterior mediastinal mass. (a) Contrast-enhanced diagnostic CT shows a homogeneous right paravertebral mass (arrow). (b) Pre-procedural cone-beam CT taken in prone position. (c) Paravertebral approach is planned and needle insertion route is estimated. (d) Biopsy needle is advanced along the estimated needle pathway to the centre of the mass

and follow-up chest radiographs during the patient’s hospitalization. All data were analyzed using Excel 2010 (Microsoft Corp., Redmond, WA, USA) and SPSS software (version 19.0; SPSS, Chicago, IL).

Results Lesion characteristics The mean size of the 75 mediastinal lesions among the 75 patients was 6.80±3.08 cm (range, 1.00–14.00 cm). Sixtythree of the lesions were located in the anterior mediastinum, five in the middle mediastinum, and seven in the posterior mediastinum. Fifty-eight lesions were solid, one was cystic, and 16 showed mixed solid and cystic features. A detailed description of lesion characteristics is provided in Table 2. Procedural records The mean distance from the skin to the target was 3.67 ± 1.80 cm (range, 1.30–9.60 cm). With regard to the mediastinal needle approach used, the parasternal approach was used for 65 lesions, the transpulmonary approach for 6, the

subxiphoid approach for 2, and the paravertebral approach for 5 lesions. Core needle biopsy was performed in 75 cases and aspiration biopsy in 3 cases. The patients underwent biopsy in the supine position in 70 cases and in the prone position in 8 cases. With respect to procedure duration, the coaxial introducer indwelling time was 9.29 ± 4.34 min (range, 1–29 min) and the total procedure time was 13.26± 5.29 min (range, 3–31 min). The number of biopsy specimens obtained per biopsy session was 3.13 ± 1.02 (range, 1–7). The number of CBCTs performed per biopsy was 3.03±0.68 (range, 1–6). The boundary of the lesion was discriminable from the abutting mediastinal vascular structures on pre-procedural CBCT in 26.9 % of procedures (21/78) (Fig. 4). The technical success rate was 100 %. Finally, the mean estimated effective radiation dose and dose area product were 5.33±4.99 mSv (range, 0.06–33.94 mSv) and 12, 723.68 ± 10,665.74 mGy⋅cm2 (range, 379.60–75, 416.00 mGy⋅cm2), respectively. A detailed description of the procedural records is summarized in Table 3. Diagnostic accuracy and complications Among the 78 biopsy results, 69 were malignant, 7 were benign, and 2 were indeterminate (Table 4). The final

Eur Radiol Table 2 Characteristics of the 75 lesions in 75 patients

Lesion characteristics

Value

Size of target lesion (cm)a Location in the mediastinum Anterior Middle Posterior

6.80±3.08

Lesion feature Solid Cystic Mixed

63 5 7 58 1 16

Note: Unless otherwise indicated, data are number of lesions a

Data are means±standard deviations

Fig. 4 C-arm cone-beam (CB) CT virtual navigation-guided biopsy in a 77-year-old woman with an anterior mediastinal mass. (a) Contrast-enhanced diagnostic CT shows a mediastinal mass (arrow) closely abutting the ascending aorta and left main pulmonary artery (arrowheads). (b) Pre-procedural CBCT reveals that the boundary of the mass (arrow) cannot be distinguished from the surrounding vascular structures. (c) Needle insertion route and depth are estimated while taking into account the lesion contour and vessel locations. (d) The biopsy needle is safely advanced without vascular injury. (e) No bleeding is detected around the biopsy site (arrow) on the post-procedural CBCT

diagnosis of malignancy was made on the basis of surgical pathology (n=15) and specific malignant biopsy results (n=54). As for the diagnosis of benignity, confirmation was made through surgical pathology (n=3), specific benign biopsy results (n=3) and clinical data (n=1). Two of the 69 malignant cases in our study were revealed to be false negatives on biopsy, and were surgically excised and diagnosed as thymomas. Of the seven benign cases, one was diagnosed clinically, as this lesion showed a decrease in diameter greater than 20 % on the follow-up chest CT (interval, 267 days; pathology, chronic granulomatous inflammation with necrosis). With regard to the two indeterminate cases, one was suspected of being residual DLBL after chemotherapy, and its pathologic diagnosis was not established, since the specimen consisted mainly

Eur Radiol Table 3 Procedural records of 78 CBCT virtual navigation-guided mediastinal biopsies

Discussion

Biopsy records

Value

Skin to the target distance (cm)a Approach technique Parasternal Transpulmonary

3.67±1.80

To the best of our knowledge, this is the first study to evaluate the diagnostic accuracy of a CBCT virtual navigation system for percutaneous mediastinal mass biopsy, revealing high accuracy (97.4 %) for the diagnosis of malignant mediastinal lesions, with a low rate of complications (3.85 %). Previous studies have reported diagnostic accuracy of CTguided mediastinal mass biopsy ranging from 77.5 to 96.4 % [22–27]. However, considering that inadequate samples were excluded from the calculation of accuracy in those studies, the true diagnostic biopsy rate may have been only as high as 91 % [22–25]. In our study, the technical success rate of the CBCT virtual navigation system was 100 %, and diagnostic accuracy was 97.4 %. Even if the two indeterminate cases were actually malignancies and we had regarded them as false negatives, the diagnostic accuracy would still have been noteworthy (94.7%). The exceptionally high diagnostic accuracy achieved in our study may be for several reasons, including the biopsy method, operator skill, and the patient population. Nevertheless, the CBCT system along with the use of virtual navigation provided clear advantages to the operator, helping to target the lesion more precisely. Moreover, the larger spatial room and the more flexible imaging plane enabled by the open gantry C-arm system would have also increased the operator’s comfort and confidence. In addition, optimal imaging projection through the rotations and angulations of the C-arm is made possible with real-time visualization of the needle position and progression to the target. Virtual guidance also offers the automatic alignment of the detector, skin entry site, and target in a bull’s-eye view, enabling operators to easily approach the target along the planned path while avoiding bones and lungs. In a recent phantom study performed by Busser et al. [28], the authors demonstrated that CBCT was the preferred modality over CT, irrespective of the level of operator experience, in terms of accurate needle placement in difficult guidance procedures requiring double-angulated needle paths. In other recent clinical studies dealing with lung nodules, CBCT-guided biopsy procedures also showed overall excellent technical success rates and diagnostic accuracy (Table 1). As the mediastinum is a complex body part comprising the heart, trachea, oesophagus and multiple vascular structures, elaborate needle placement is essential. As such, CBCT virtual navigation-guided biopsy may be an exceptionally wellsuited procedure for lesions in the mediastinum. With respect to the number of complications, previous studies on CT-guided mediastinal biopsy have reported a wide range of complication rates, from 0 to 45.5 % [22–27, 29–31]. Hagberg et al. [23] reported that no complications had occurred in their study, albeit with a small study population (n=21) using only the transsternal approach. As for the types of complications that can occur, the most frequently documented complication of mediastinal biopsy has been

Subxiphoid Paravertebral Biopsy method Core needle biopsy Aspiration biopsy Patient position Supine Prone Procedure time (min)a

65 6 2 5 75 3 70 8

Needle dwelling time 9.29±4.34 Total procedure time 13.26±5.29 Number of biopsies obtaineda 3.13±1.02 Number of CBCT data acquisitionsa 3.03±0.68 Lesion border discrimination from its abutting mediastinal vessels on pre-procedural CBCT Available 21 Unavailable 57 Technical success rate (%) 100 Radiation dosea Absorbed radiation dose (DAP, mGy∙cm2) 12,723.68±10,665.74 Effective dose (mSv) 5.33±4.99 Complication Pneumothorax 3 Note: Unless otherwise indicated, data are number of biopsy procedures a

Data are mean±standard deviation

CBCT C-arm cone-beam computed tomography, DAP dose area product

of necrotic material. The patient underwent radiotherapy without additional biopsy, and the lesion decreased in size over follow-up (interval, 781 days). The other case was suspected of being ovarian cancer metastasis; however, the biopsy specimen was negative for malignant cells. This patient was not followed up after biopsy in our hospital. Therefore, the overall sensitivity, specificity and accuracy for the diagnosis of malignancy for CBCT virtual navigation-guided percutaneous mediastinal mass biopsies were 97.1 % (67/69), 100 % (7/7) and 97.4 % (74/76), respectively. Complications occurred in 3 of 78 procedures (3.85 %), all of which were pneumothoraces. However, no action other than oxygen inhalation was required in these patients, as the pneumothoraces were small and asymptomatic.

Eur Radiol Table 4 Final diagnoses of the 78 mediastinal biopsies Malignant lesions (n=69)

Benign lesions (n=7) Indeterminate lesions (n=2)

Final diagnosis

Number of cases

Thymic carcinoma Thymoma Metastasis Diffuse large B-cell lymphoma Small cell carcinoma, Hodgkin lymphoma, lymphoblastic lymphoma Squamous cell carcinoma Yolk sac tumor, poorly differentiated carcinoma, marginal zone B-cell lymphoma Adenocarcinoma, carcinosarcoma, combined small cell carcinoma and squamous cell carcinoma, lymphoepithelial-like carcinoma, malignant mesenchymal tumor, seminoma, ganglioneuroblastoma, undifferentiated pleomorphic sarcoma Schwannoma Amyloidosis, mature teratoma, chronic granulomatous inflammation with necrosis Diffuse large B-cell lymphoma, metastasis (presumed diagnosis based on clinical suspicion)

12 11 8 6 5 for each entity

pneumothorax, as was the case in our study [26]. However, as the pneumothoraces in our study were small and asymptomatic, none of the patients required chest tube insertions or prolonged hospital stays. Interestingly, no hematomas owing to vascular injury developed among patients in our study, although lesion border discrimination from abutting mediastinal vessels on pre-procedural CBCT was available in only 26.9 % of cases. Previous studies have reported the occurrence of hematoma or haemorrhage in up to 10 % of CT-guided mediastinal biopsies [22, 24–26, 29–31]. Considering the intrinsic disadvantage of CBCT, in which the large cone angle increases scattered radiation as well as noise, which in turn can disturb the discrimination of low-contrast objects [32], vascular injury during CBCT-guided biopsy is of significant concern. However, after scrupulous review of the diagnostic contrast-enhanced CT data taken prior to the procedure, estimation of the lesion boundary and the safe range for needle advancement was feasible. Thus, we believe that the low contrast resolution of CBCT may not be a critical obstacle and that the biopsy procedure can be performed safely without contrast media injection. In terms of radiation exposure, the mean effective dose in the present study was 5.33±4.99 mSv. Although the radiation dose for CT-guided or CT fluoroscopy-guided mediastinal biopsy has not been reported in the literature thus far, recent studies on CBCT-guided percutaneous transthoracic lung biopsy revealed an average effective dose ranging 4.6 to 8.6 mSv [4, 8, 12, 14]. The radiation dose in our study falls within that range, and is thought to be acceptable, considering that the mean effective dose of a regular-dose chest CT has been reported as approximately 11.05 mSv [33]. To further reduce radiation exposure, there may be a few strategies that we can explore, including the use of low-quality 3D rotational

3 2 for each entity 1 for each entity

4 1 for each entity 1 for each entity

angiography-mode CBCT with strictly limited scanning area and number of CBCT acquisitions. In addition, iterative image reconstruction algorithms, which have not been established in CBCT systems to date, may also reduce radiation exposure in the near future by lowering image noise [32]. There are several limitations to our study. First, our retrospective study included a relatively small number of patients. Therefore, a detailed analysis revealing the factors influencing diagnostic accuracy or complication rate was not conducted. Future studies examining a larger population using this system are warranted in this regard. Second, we did not compare the diagnostic capability of CBCT virtual navigation-guided mediastinal mass biopsy with other procedures such as CTguided or CT fluoroscopy-guided biopsy. Thus, it is still unclear whether the CBCT virtual navigation-guided biopsy can provide a direct benefit over other biopsy methods. A prospective randomized controlled clinical trial dealing with these systems would help resolve this question. In conclusion, CBCT virtual navigation-guided percutaneous mediastinal mass biopsy is an accurate procedure that allows safe access to lesions in various locations within the mediastinum. As such, this method is highly recommended for mediastinal lesions, which typically require complex needle paths. Acknowledgments The scientific guarantor of this publication is Chang Min Park, MD, Ph.D. The authors of this manuscript declare no relationships with any companies whose products or services may be related to the subject matter of the article. This work was supported by the Industrial Strategic Technology Development Program (grant number: 10041618) funded by the Ministry of Trade Industry and Energy (MI, Korea). No complex statistical methods were necessary for this paper. Institutional review board approval was obtained. Written informed consent was waived by the institutional review board. Study subjects have not been previously reported in other journals. Methodology: retrospective, diagnostic study (case series), performed at one institution.

Eur Radiol

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

Gupta S, Seaberg K, Wallace MJ et al (2005) Imaging-guided percutaneous biopsy of mediastinal lesions: different approaches and anatomic considerations. Radiographics 25:763–786 de Jiao C, Li TF, Han XW et al (2014) Clinical applications of the C-arm cone-beam CT-based 3D needle guidance system in performing percutaneous transthoracic needle biopsy of pulmonary lesions. Diagn Interv Radiol 20:470–474 Carlson SK, Felmlee JP, Bender CE et al (2005) CT fluoroscopyguided biopsy of the lung or upper abdomen with a breath-hold monitoring and feedback system: a prospective randomized controlled clinical trial. Radiology 237:701–708 Choi JW, Park CM, Goo JM et al (2012) C-arm cone-beam CTguided percutaneous transthoracic needle biopsy of small (≤20 mm) lung nodules: diagnostic accuracy and complications in 161 patients. AJR Am J Roentgenol 199:W322–W330 Yankelevitz DF, Vazquez M, Henschke CI (2000) Special techniques in transthoracic needle biopsy of pulmonary nodules. Radiol Clin N Am 38:267–279 Nawfel RD, Judy PF, Silverman SG, Hooton S, Tuncali K, Adams DF (2000) Patient and personnel exposure during CT fluoroscopyguided interventional procedures. Radiology 216:180–184 Wallace MJ, Kuo MD, Glaiberman C et al (2009) Threedimensional C-arm cone-beam CT: applications in the interventional suite. J Vasc Interv Radiol 20:S523–S537 Choo JY, Park CM, Lee NK, Lee SM, Lee HJ, Goo JM (2013) Percutaneous transthoracic needle biopsy of small (≤1 cm) lung nodules under C-arm cone-beam CT virtual navigation guidance. Eur Radiol 23:712–719 Braak SJ, Herder GJ, van Heesewijk JP, van Strijen MJ (2012) Pulmonary masses: initial results of cone-beam CT guidance with needle planning software for percutaneous lung biopsy. Cardiovasc Intervent Radiol 35:1414–1421 Cheung JY, Kim Y, Shim SS, Lim SM (2011) Combined fluoroscopy- and CT-guided transthoracic needle biopsy using a C-arm cone-beam CT system: comparison with fluoroscopyguided biopsy. Korean J Radiol 12:89–96 Choi MJ, Kim Y, Hong YS, Shim SS, Lim SM, Lee JK (2012) Transthoracic needle biopsy using a C-arm cone-beam CT system: diagnostic accuracy and safety. Br J Radiol 85:e182–e187 Hwang HS, Chung MJ, Lee JW, Shin SW, Lee KS (2010) C-arm cone-beam CT-guided percutaneous transthoracic lung biopsy: usefulness in evaluation of small pulmonary nodules. AJR Am J Roentgenol 195:W400–W407 Jin KN, Park CM, Goo JM et al (2010) Initial experience of percutaneous transthoracic needle biopsy of lung nodules using C-arm cone-beam CT systems. Eur Radiol 20:2108–2115 Lee SM, Park CM, Lee KH, Bahn YE, Kim JI, Goo JM (2014) Carm cone-beam CT-guided percutaneous transthoracic needle biopsy of lung nodules: clinical experience in 1108 patients. Radiology 271:291–300 Lee WJ, Chong S, Seo JS, Shim HJ (2012) Transthoracic fineneedle aspiration biopsy of the lungs using a C-arm cone-beam CT system: diagnostic accuracy and post-procedural complications. Br J Radiol 85:e217–e222

16.

Kim JI, Park CM, Lee SM, Goo JM (2015) Rapid needle-out patient-rollover approach after cone beam CT-guided lung biopsy: effect on pneumothorax rate in 1,191 consecutive patients. Eur Radiol. doi:10.1007/s00330-015-3601-y 17. Carrafiello G, Fontana F, Mangini M et al (2012) Initial experience with percutaneous biopsies of bone lesions using XperGuide conebeam CT (CBCT): technical note. Radiol Med 117:1386–1397 18. Leschka SC, Babic D, El Shikh S, Wossmann C, Schumacher M, Taschner CA (2012) C-arm cone beam computed tomography needle path overlay for image-guided procedures of the spine and pelvis. Neuroradiology 54:215–223 19. Whitten CR, Khan S, Munneke GJ, Grubnic S (2007) A diagnostic approach to mediastinal abnormalities. Radiographics 27:657–671 20. Riedel RF, Burfeind WR Jr (2006) Thymoma: benign appearance, malignant potential. Oncologist 11:887–894 21. Suster S, Moran CA (2006) Thymoma classification: current status and future trends. Am J Clin Pathol 125:542–554 22. Gupta S, Wallace MJ, Morello FA Jr, Ahrar K, Hicks ME (2002) CT-guided percutaneous needle biopsy of intrathoracic lesions by using the transsternal approach: experience in 37 patients. Radiology 222:57–62 23. Hagberg H, Ahlstrom HK, Magnusson A, Sundstrom C, Astrom GK (2000) Value of transsternal core biopsy in patients with a newly diagnosed mediastinal mass. Acta Oncol 39:195–198 24. Kulkarni S, Kulkarni A, Roy D, Thakur MH (2008) Percutaneous computed tomography-guided core biopsy for the diagnosis of mediastinal masses. Ann Thorac Med 3:13–17 25. Lin ZY, Li YG (2009) Artificial pneumothorax with position adjustment for computed tomography-guided percutaneous core biopsy of mediastinum lesions. Ann Thorac Surg 87:920–924 26. Priola AM, Priola SM, Cataldi A et al (2008) CT-guided percutaneous transthoracic biopsy in the diagnosis of mediastinal masses: evaluation of 73 procedures. Radiol Med 113:3–15 27. Zwischenberger JB, Savage C, Alpard SK, Anderson CM, Marroquin S, Goodacre BW (2002) Mediastinal transthoracic needle and core lymph node biopsy: should it replace mediastinoscopy? Chest 121:1165–1170 28. Busser WM, Braak SJ, Futterer JJ et al (2013) Cone beam CT guidance provides superior accuracy for complex needle paths compared with CT guidance. Br J Radiol 86:20130310 29. Greif J, Staroselsky AN, Gernjac M et al (1999) Percutaneous core needle biopsy in the diagnosis of mediastinal tumors. Lung Cancer 25:169–173 30. Sklair-Levy M, Polliack A, Shaham D et al (2000) CT-guided coreneedle biopsy in the diagnosis of mediastinal lymphoma. Eur Radiol 10:714–718 31. Zinzani PL, Corneli G, Cancellieri A et al (1999) Core needle biopsy is effective in the initial diagnosis of mediastinal lymphoma. Haematologica 84:600–603 32. Schegerer AA, Lechel U, Ritter M, Weisser G, Fink C, Brix G (2014) Dose and image quality of cone-beam computed tomography as compared with conventional multislice computed tomography in abdominal imaging. Investig Radiol 49:675–684 33. Lee CH, Goo JM, Ye HJ et al (2008) Radiation dose modulation techniques in the multidetector CT era: from basics to practice. Radiographics 28:1451–1459