Transthoracic Computed Tomography–Guided Lung Nodule Biopsy: Comparison of Core Needle and Fine Needle Aspiration Techniques

Transthoracic Computed Tomography–Guided Lung Nodule Biopsy: Comparison of Core Needle and Fine Needle Aspiration Techniques

Canadian Association of Radiologists Journal xx (2016) 1e6 www.carjonline.org Thoracic and Cardiac Imaging / Imagerie cardiaque et imagerie thoraciqu...

218KB Sizes 0 Downloads 43 Views

Canadian Association of Radiologists Journal xx (2016) 1e6 www.carjonline.org

Thoracic and Cardiac Imaging / Imagerie cardiaque et imagerie thoracique

Transthoracic Computed TomographyeGuided Lung Nodule Biopsy: Comparison of Core Needle and Fine Needle Aspiration Techniques Bippan S. Sangha, MDa,*, Cameron J. Hague, MDa, Jennifer Jessup, MDa, Robert O’Connor, MBb, John R. Mayo, MDa a

Department of Radiology, University of British Columbia, Vancouver, British Columbia, Canada Department of Pathology, University of British Columbia, Vancouver, British Columbia, Canada

b

Abstract Purpose: To determine if there is a statistically significant difference in the computed tomography (CT)eguided trans-thoracic needle biopsy diagnostic rate, complication rate, and degree of pathologist confidence in diagnosis between core needle biopsy (CNB) and fine needle aspiration biopsy (FNAB). Methods: A retrospective cohort design was used to compare the diagnostic biopsy rate, diagnostic confidence, and biopsy-related complications of pneumothorax, chest tube placement, pulmonary hemorrhage, hemoptysis, admission to hospital, and length of stay between 251 transthoracic needle biopsies obtained via CNB (126) or FNAB (125). Complication rates were assessed using imaging and clinical followup. Final diagnosis was confirmed via surgical pathology or clinical follow-up over a period of up to 10 years. Results: CNB provided diagnostic samples in 91% and FNA in 80% of biopsies, which was statistically significant (P < .05). The sensitivities for CNB and FNAB were 89% (85 of 95) and 95% (84 of 88), respectively. The specificity of CNB was 100% (21 of 21) and for FNAB was 81% (2 of 11) with 2 false positives in the FNAB group. The differences in complication rate was not statistically significant for pneumothorax (50% vs 46%; determined by routine postbiopsy CT), chest tube (2% vs 4%), hemoptysis (4% vs 6%), and pulmonary hemorrhage (38% vs 47%) between FNAB and CNB, respectively. Seven patients requiring chest tube were admitted to hospital, 2 in the FNAB cohort for an average of 2.5 days and 5 in the CNB cohort for an average of 4.6 days. Conclusions: CNB provided more diagnostic samples with no statistical difference in complication rate. Resume Objet : Determiner s’il existe un ecart statistiquement significatif entre la biopsie au trocart et la cytoponction (aspiration a l’aiguille) en ce qui a trait aux taux de diagnostic par biopsie transthoracique a l’aiguille guidee par tomodensitometrie (TDM), aux taux de complication et au degre de confiance diagnostique du pathologiste. Methodes : Un examen de cohorte retrospectif a permis de comparer le taux de diagnostic, la confiance diagnostique et les complications (pneumothorax, insertion d’un drain thoracique, hemorragie pulmonaire, hemoptysie, admission et duree du sejour a l’h^ opital) associes a 251 biopsies transthoraciques a l’aiguille, dont 126 biopsies au trocart et 125 cytoponctions. Les taux de complication ont ete determines gr^ace au suivi en imagerie et au suivi clinique. Le diagnostic definitif a pour sa part ete confirme au moyen de resultats pathologiques chirurgicaux ou d’un suivi clinique pouvant s’echelonner sur 10 ans au plus. Resultats : 91 % des echantillons preleves par biopsie au trocart ont ete utilises a des fins diagnostiques, contre 80 % des echantillons preleves par cytoponction (ecart statistiquement significatif; P < 0,05). Le degre de sensibilite des biopsies au trocart a ete etabli a 89 % (85 sur 95), alors que celui des cytoponctions a ete etabli a 95 % (84 sur 88). Les biopsies au trocart ont affiche une specificite de 100 % (21 sur 21) et les cytoponctions, une specificite de 81 % (2 sur 11) avec deux resultats faussement positifs au sein du groupe. En ce qui concerne les taux de complication, les ecarts entre les cytoponctions et les biopsies au trocart ne se sont pas averes significatifs sur le plan statistique au chapitre des pneumothorax (50 % contre 46 %; determines par la TDM postbiopsie d’usage), de l’insertion d’un drain thoracique (2 % contre 4 %), de l’hemoptysie (4 % contre 6 %) et de l’hemorragie pulmonaire (38 % contre 47 %). Enfin, des 7 patients

* Address for correspondence: Bippan S. Sangha, MD, Diagnostic Radiology Residency Program, Department of Radiology, Vancouver General

Hospital, University of British Columbia, 3350 - 950 West 10th Avenue, Vancouver, British Columbia V5Z 4E3, Canada. E-mail address: [email protected] (B. S. Sangha).

0846-5371/$ - see front matter Ó 2016 Canadian Association of Radiologists. All rights reserved. http://dx.doi.org/10.1016/j.carj.2015.10.005

2

B. S. Sangha et al. / Canadian Association of Radiologists Journal xx (2016) 1e6

qui ont ete hospitalises en raison de l’insertion d’un drain thoracique, 2 avaient subi une cytoponction et ont passe en moyenne 2,5 jours a l’h^ opital, alors que 5 avaient subi une biopsie au trocart et ont passe en moyenne 4,6 jours a l’h^ opital. Conclusions : La biopsie au trocart preleve davantage d’echantillons diagnostics que la cytoponction. Toutefois, les deux interventions n’affichent aucun ecart statistiquement significatif au chapitre des taux de complication. Ó 2016 Canadian Association of Radiologists. All rights reserved. Key Words: Core needle biopsy; Fine needle aspiration biopsy; Lung nodule; Transthoracic lung nodule biopsy

Computed tomography (CT)eguided percutaneous biopsy is commonly used to provide tissue diagnosis in indeterminate lung nodules and masses and is an essential element of the diagnostic algorithm [1]. Recent data from the National Lung Cancer Screening Trial showed a significant reduction in lung cancer mortality in high risk lung cancer subjects receiving screening low-dose chest CT examinations [2]. The results of this landmark randomized trial are expected to increase the demand for CT-guided percutaneous biopsy. The most commonly used CT-guided biopsy techniques are core needle (CN) and fine needle aspiration (FNA) biopsy. Despite numerous previous studies [3e11] debate remains around the optimal technique, with CN advocates focusing on the larger volume of tissue provided, while FNA advocates note the less traumatic aspect of a smaller needle gauge biopsy technique. In addition to differences in the volume of tissue provided, there are substantial operational differences. At the majority of sites, FNA requires either a cytopathology technologist or cytopathologist presence in the CT scanner suite during the biopsy procedure to confirm an adequate tissue sample is obtained. Provision of this manpower is not possible in some institutions, and if not provided it has been documented that the rate of insufficient tissue may reach 20% [12]. In comparison, CN biopsy can be successfully performed in the absence of onsite pathology support since it only requires placing the sample in a labeled formalin container. However, it has been suggested that CN leads to increased complications including pneumothorax, hemorrhage, and hemoptysis compared to FNA [3,13]. Patient-specific factors have been associated with complications including lesion size, degree of emphysema and needle path length [14e16] suggesting the optimal technique may be patient specific. Finally, although previous studies have reported lower accuracy of FNA for both benign [1,4,8] and malignant [11] lesions compared to CN, a recent meta-analysis [9] did not show significant (P < .05) differences in diagnostic yield or complication rates. Thus, clinical equipoise exists at this time. In 2004, due to retirement of the senior cytopathologist and operational restructuring, the preferred CT-guided biopsy technique changed from FNA to CN at our institution. This change provided an opportunity to compare the 2 techniques to an external gold standard consisting of surgical resection and/or a minimum of 6 years of clinical follow-up. We hypothesized that a larger volume of tissue from CN biopsy would lead to a decrease in the number of

nondiagnostic samples, increase in pathologist diagnostic confidence, and no change in the rate of procedural complications. Methods CT-guided biopsies of 251 undiagnosed lung lesions were performed in 243 patients (126 male, 117 female) using a single 16-detector-row CT scanner without fluoroscopic capability (Siemens Healthcare, Forcheim, Germany). Biopsies were obtained by either a staff radiologist (J.R.M.) with 20 years of experience or a thoracic imaging fellow supervised by the same staff radiologist. This retrospective study was approved by the Institutional Review Board of the University of British Columbia (Vancouver, Canada). We retrospectively reviewed 2 cohorts of biopsied lesions: 124 consecutive FNA biopsies (70 male, 54 female) between January 1, 2001, and July 31, 2003; 128 CN biopsies (60 male, 68 female) composed of 119 consecutive biopsies between July 1, 2006, and June 30, 2008; and 9 consecutive CN biopsies after unsuccessful FNA between January 2001 and July 31, 2003. Procedural informed consent was obtained from all patients prior to biopsy. All patients had normal coagulation parameters (prothrombin time, partial thromboplastin time, platelets, international normalized ratio) within 3 weeks prior to biopsy. Pre biopsy imaging was reviewed prior to booking to evaluate for the presence and extent of emphysema, proximity of the lesion to central pulmonary vessels, needle path planning, and patient positioning. Patients with emphysema classified by visual estimation as either severe (50%-75% of lung volume) or very severe (>75% lung volume) received prebiopsy lung function testing. Biopsy was not attempted if the forced expiratory lung volume at 1 second (FEV 1.0) was less than 1 L, or if the patient was on home oxygen. Immediately prior to FNA and CN biopsy, a planning CT was performed with the patient in the biopsy position to confirm feasibility of the planned needle path. The needle path was chosen to avoid crossing of interlobar fissures and large vascular structures. Once the needle path was confirmed, 3-7 mL of 1% Lidocaine (Alveda Pharma, Toronto, ON) was injected subcutaneously using CT guidance. FNA biopsy was performed in the presence of a cytopathology technologist using a tandem needle approach with identical 10 or 15 cm long 22-gauge Westcott cutting needles (Cook Medical, Bloomington, IN). The initial needle was placed into the lesion and position confirmed using CT. The

Comparison of core and fine needle lung nodule biopsy / Canadian Association of Radiologists Journal xx (2016) 1e6

second identical needle was placed alongside the first. Two aspiration biopsies were then performed identically through the 2 needles. A cytopathology technologist was present to ensure adequate samples and prepare slides. Samples were reviewed by a cytopathologist with 30 years of experience. CN biopsies were performed using a thin-walled 19-gauge guide needle through which a spring-loaded 20-gauge core biopsy needle (Quick-Core; Cook Medical) was placed. Depending on the size of the lesion a 10 or 20 mm throw was chosen. The guide needle was placed within the lesion using CT guidance. The stylet of the guide needle was removed and replaced by the 20 gauge core biopsy needle. Six cores were routinely obtained. Five core biopsy samples were placed in formalin and sent to the pathology laboratory. The sixth core was placed on a gauze pad soaked in saline and sent to the microbiology laboratory in a sterile vial. Immediately postbiopsy, a limited coverage chest CT was used to assess for pneumothorax or parenchymal hemorrhage. Patients with these complications received immediate post biopsy baseline chest radiographs. The decision to place a chest tube was based on the clinical status of the patient and the size of the pneumothorax. Asymptomatic patients with a pneumothorax involving less than 25% of lung volume were observed. Larger or symptomatic pneumothoraces were drained using a percutaneously placed pigtail catheter with a Heimlich valve (Cook Critical Care). All patients were observed for 4 hours in a day bed unit then returned to radiology for a pre discharge follow-up chest radiograph. If a pneumothorax increased in size on follow-up chest radiograph, the referring clinical service (thoracic surgery or respirology) was consulted, with the decision to place a chest tube dependent on their opinion. All patients receiving a chest tube were admitted and their in-hospital course tracked through hospital medical records. Biopsy was aborted in 3 patients (2 CN and 1 FNA) before a sample was obtained due to pneumothorax (1 FNA, 1 CN) and substantial hemoptysis (1 CN). These cases were categorized as inadequate samples. Data Extraction For each biopsy a single fellowship trained thoracic radiologist with 6 years of experience (C.H.) reviewed the prebiopsy CT scan and hospital medical record for: age, sex, biopsy technique, lesion mean diameter, lesion morphology (solid, semisolid, or ground glass), emphysema extent (Table 1), and chest tube placement. The same radiologist also assessed postbiopsy CT scans and reports for complications including: pneumothorax severity (Table 2), pulmonary hemorrhage, hemoptysis, chest tube placement, and fatality. The pathology reports were scored using a 7-point Likert scale (Table 3) by a single observer (B.S.S.), blinded to the follow-up reference standard. In all cases receiving complete surgical excision, the follow-up reference standard was the pathologic diagnosis on the specimen. All biopsies not receiving surgical excision were cross-referenced with the British Columbia Lung

3

Table 1 Comparison of patient parameters between the CN and FNA biopsy groups Biopsy type

CN

FNA

Mean patient age, y Mean lesion diameter, cm

65.8  11.9 3.2  2.1 (range 0.7-10.4) 1.4  1.4 (range 0-6.2) 103 25

67.3  11.0 3.0  1.8 (range 0.8-9.2) 1.5  1.4 (range 0-6.5) 101 23

41 48 23 12 4

58 38 19 7 2

Mean pleural distance, cm Malignant lesions Benign lesions Degree of emphysema < 5% of lung volume 6%-25% of lung volume 26%-49% of lung volume 50%-74% of lung volume > 75% of lung volume

Grade 1 2 3 4 5

CN ¼ core needle; FNA ¼ fine needle aspiration.

Cancer Registry. As lung cancer is a reportable disease in our province, this identified all subjects receiving treatment for lung cancer who did not receive surgical resection. All other biopsies had a follow-up reference diagnosis made by telephone contact with the referring or family physician. Followup was complete to February 1, 2012 (mean follow-up CN vs. FNA: 1041 vs 1738 days). No subjects were lost to follow-up. Statistical Analysis Continuous variables were described with mean and standard deviation. Categorical variables were described with counts or percentages. Complications, pathologic diagnosis, and comparison between the biopsy method was tested with a chi-square test. Logistic regression was used to test if the biopsy type impacted the lesion characteristic to pathology diagnosis relationship as well as the lesion characteristic to complication relationship. Statistical software R version 3.10 (Vienna, Austria) was used. Sensitivity and specificity of the lung biopsy pathologic diagnosis was made by comparison to the follow-up reference standard. A true negative biopsy was defined as a lesion assessed via pathology to be benign (Likert scale scores 5, 6, or 7) that did not demonstrate malignancy on subsequent follow-up. True positive results were defined as lesions assessed by pathology to be malignant (Likert scale scores 1, 2, or 3) that were confirmed as malignant at follow-up. False positives were defined as biopsies assessed via pathology to be malignant (Likert scale scores 1, 2, or 3) Table 2 Grading the severity of pneumothorax on immediate postbiopsy CT Degree of emphysema on CT

Score

None <5% of lung volume 6%-25% of lung volume 26%-49% of lung volume >50% of lung volume

None Trivial Mild Moderate Severe

CT ¼ computed tomography.

4

B. S. Sangha et al. / Canadian Association of Radiologists Journal xx (2016) 1e6

Table 3 Criteria and scoring of pathology report confidence in diagnosis Likert score

Classification

Criteria assessed

1 2 3 4 5 6 7

Definitely malignant Highly suspicious for malignancy Mildly suspicious for malignancy Nondiagnostic Definitely benign Benign features No malignancy, adequate sample

Meeting pathologic requirement for malignancy >2 malignant features, but not meeting all requirements for pathologic diagnosis Single malignant feature Nondiagnostic assessment by pathologist Specific benign diagnosis provided Benign features described, benign diagnosis favored without definite diagnosis Adequate lung tissue, no malignant features seen

that were subsequently confirmed to be nonmalignant on follow-up. False negatives were defined as biopsies assessed as benign by pathologic evaluation (Likert scale scores 5, 6, or 7) that were subsequently shown to be malignant at follow-up.

Results No significant difference was found (P > .05) between FNA and CN for mean patient age, lesion diameter, distance from the pleura surface, rate of malignancy, or degree of emphysema (Table 1). Comparing FNA to CN, FNA had significantly (P < .05) more inadequate samples (26 of 124 vs 12 of 128, respectively). Between FNA and CN, there was no significant difference (P > .05) in the rate of definitely malignant (67 vs 74) or definitely benign (4 vs 6) biopsies (Table 4). Similarly, when the biopsy was called pathologically highly suspicious or suspicious for malignancy, there was no significant (P > .05) difference in the rate of malignancy on reference follow-up for FNA (15 of 19) vs CN (11 of 12). For those with no malignant features or some benign features, there was no difference in the rate of benignity on reference follow-up for FNA (4 of 8) vs CN (13 of 23). The odds ratio for obtaining a diagnostic sample in a patient with underlying emphysema was significantly (P < .05) higher using FNA (OR 1.56) compared to CN (OR 0.71). Inadequate samples were significantly (P < .05) more frequent as lesion size decreased for both CN and FNA. There was no significant difference in the rate of adequate sampling between FNA and CN with respect to lesion Table 4 Distribution of pathology report scores with results of subsequent follow-up Biopsy type

CN

morphology (Table 5). No significant difference (P > .05) was seen between FNA and CN for any complication (Table 6) and there were no fatalities. Increasing extent of emphysema was related to a higher pneumothorax rate with CN compared to FN (Table 7). There was a significant (P < .005) relationship between chest tube placement and increasing degree of emphysema in the CN, but not the FN cohort. Small lesion size was demonstrated to be a significant driver of hemorrhage and pneumothorax in both study arms, but not hemoptysis or chest tube placement.

Discussion Diagnostic Performance The transition from FNA to CN at our institution resulted in a significant decrease in the rate of inadequate samples without a significant increase in biopsy related complications. Compared to the external gold standard we found no difference between the diagnostic accuracy of FNA and CN, with 100% accuracy for definitively malignant (67 of 67 vs 74 of 74, respectively), or definitively benign (4 of 4 vs 6 of 6, respectively) lesions. This data is strengthened by the 100% follow-up provided by the composite gold standard of either surgical resection pathology, cross reference with the provincial lung cancer registry or long-term patient follow-up. We graded pathologic diagnostic confidence of biopsy specimens using a 7-point Likert scale. We hypothesized that pathologists interpreting samples obtained via CN would be more confident in their diagnosis, reflected in the wording of the pathology report. While performance of CN or FNA biopsy varied between Likert scores, other than definitively malignant, inadequate

FNA

Definitively malignant 74 (74 MFU) 67 (67 MFU) Highly suspicious for 7 (7 MFU) 8 (6 MFU, 2 BFU) malignancy Suspicious for malignancy 5 (4 MFU, 1 BFU) 11 (9 MFU, 2 BFU) Inadequate sample 12 (6 MFU, 6 BFU) 26 (14 MFU, 12 BFU) Definitively benign 6 (6 BFU) 4 (4 BFU) Some benign features 4 (2 MFU, 2 BFU) 2 (1 MFU, 1 BFU) No malignant features, 19 (8 MFU, 11 BFU) 6 (3 MFU, 3 BFU) adequate sample BFU ¼ benign on follow-up; CN ¼ core needle; FNA ¼ fine needle aspiration; MFU ¼ malignant on follow-up.

Table 5 Odds of obtaining an adequate sample with respect to lesions size, morphology, pleural distance, and degree of emphysema CN

FNA

Biopsy type

Odds ratio (95% CI)

P

Odds ratio (95% CI)

P

Lesion size Morphology Pleural distance Emphysema

1.87 2.44 0.84 0.71

.01 .21 .25 .11

1.52 0.95 0.82 1.56

.05 .93 .25 .19

(1.16e3.00) (0.61e9.82) (0.61e1.14) (0.47e1.08)

(1.00e2.32) (0.31e2.90) (0.59e1.15) (0.80e3.06)

CI ¼ confidence interval; CN ¼ core needle; FNA ¼ fine needle aspiration.

Comparison of core and fine needle lung nodule biopsy / Canadian Association of Radiologists Journal xx (2016) 1e6 Table 6 Rates of complications using CN and FNA techniques Complications Hemorrhage None Mild Moderate Severe Hemoptysis Pneumothorax None Trivial Mild Moderate Severe Chest tube placement

CN

FNA

68 50 8 2 8

(53.1%) (39.1%) (6.3%) (1.6%) (6.3%)

60 53 11 1 5

(48%) (42.4%) (8.8%) (0.8%) (4%)

58 43 10 5 0 5

(45.3%) (34.1%) (7.9%) (4%)

63 46 14 2 0 2

(50%) (36.8%) (11.2%) (1.6%)

(3.9%)

(1.6%)

CN ¼ core needle; FNA ¼ fine needle aspiration.

sample and definitively benign (Likert scores 1, 4, and 5 respectively), no statistical difference was identified between the 2 techniques with regard to indeterminate biopsies (Likert scores of 2, 3, 6, and 7). We found that in about half of cases (CN, 10 of 21; and FN, 4 of 8) biopsy samples with no malignant features or some benign features, but lacking a definitively benign diagnosis, were malignant on follow-up. This is similar to our rate of malignancy for cases with inadequate biopsy samples (CN, 6 of 12; and FN, 14 of 26). Although we had hoped the more elaborate Likert scale may have teased out subtle performance differences in the 2 study arms, the results ultimately highlight the necessity to aggressively pursue a final diagnosis for all lung lesions not determined to be definitively malignant or benign on percutaneous biopsy. The main factor behind the overall Table 7 Odds of having complications with respect to prebiopsy patient parameters using CN and FNA CN Complication Hemorrhage Lesion size Morphology Pleural distance Emphysema Pneumothorax Lesion size Morphology Pleural distance Emphysema Hemoptysis Lesion size Morphology Pleural distance Emphysema Chest tube Lesion size Morphology Pleural distance Emphysema

FNA

Odds ratio

P

Odds ratio

P

0.58 1.91 2.35 0.90

< .005 .07 < .005 .50

0.61 7.18 2.28 0.82

< .005 < .005 < .005 .33

0.51 1.25 1.70 1.61

< .005 .52 < .005 .01

0.79 1.42 1.74 0.94

.03 .38 < .005 .73

0.56 0.43 1.57 0.82 0.57 0.85 0.74 3.44

.07 .39 .02 .56 .21 .87 .46 < .005

CN ¼ core needle; FNA ¼ fine needle aspiration. Bold indicates statistical significance.

0.64 1.72 1.58 0.47

.37 .57 .17 .42

0.45 1.38 1.91 1.23

.33 .79 .10 .76

5

difference in diagnostic performance between the 2 biopsy techniques was the avoidance of obtaining an inadequate sample. Overall this was significantly better for the biopsies performed with CN as compared to FNA (P < .01). Lesion size (maximal transaxial diameter) was a driver of the ability to obtain an adequate sample in both groups, with no significant difference (P > .05) found between the techniques. Diagnostic accuracy in small lesions has been the subject of previous research [17] and is likely to be important in the future if CT lung cancer screening programs are instituted. For these small lesions the improved accuracy of excision biopsy using guided video assisted thoracic surgery may be useful [18]. We found similar diagnostic accuracy between FNA and CN for benign diagnoses. This contrasts with previous studies showing improved CN accuracy for benign lesions [4,5,11] possibly secondary to the increased biopsy volume. The similar performance of FNA and CN in this study is likely due to the lower rate of benign diagnoses (10 of 243) in our population compared to Staroselsky et al [5] (41 of 182), Laurent et al [11] (25 of 221), and Arakawa et al [4] (35 of 122). Complication Rates Complication rates did not differ significantly between the 2 biopsy techniques. No overall difference in pneumothorax rate was seen between CN and FNA in our study, with the pneumothorax rate in either cohort similar to prior studies ranging from 4%61% [19,20]. When controlling for patient parameters, a greater degree of emphysema was associated with an increased risk of pneumothorax in the CN group, but not the FNA group. However, there was not a statistically increased risk of chest tube placement in the CN group. Our result differs from Cox et al [16] who found an increased risk of pneumothorax with greater emphysema severity using a FNA technique. In our study, FNA used a tandem needle technique, minimizing needle gauge at the cost of increased number of pleural passes compared to a coaxial needle technique. Multiple previous studies have not shown a relationship between emphysema and pneumothorax risk [21e23]. Interestingly, multivariate analysis demonstrated that with increasing emphysema, FNA had a greater diagnostic yield. This finding has not been previously described and may reflect the greater relationship between emphysema degree and pneumothorax rate in the CN vs FNA cohorts. Our rate of chest tube placement for either CN or FN was in the range reported in the literature between 0.5% [24] and 14.2% [25]. FNA or CN did not have an effect on rate of chest tube placement. Similar to previous studies [9,11,17] we found no significant difference between FNA and CN on pulmonary hemorrhage or hemoptysis. We note there has been a wide range of reported rates of hemorrhage associated with CT-guided biopsy which may be due to differing methods of grading. Some papers have defined hemorrhage

6

B. S. Sangha et al. / Canadian Association of Radiologists Journal xx (2016) 1e6

using hemoptysis [14] while others have used hemoptysis and ground glass opacity on postbiopsy CT [26]. Consistent with previous studies [4e6,8,9,27] we found no difference between FNA and CN on hemoptysis and our hemoptysis rate was concordant with previous data [5,14]. In this study, pneumothorax and pulmonary hemorrhage were more common in small lesions. This finding again may support the use of guided VATS excision biopsy for small peripheral lung nodules. Limitations of this study include the relatively small sample sizes within some Likert categories. This may have limited our ability to use a Likert scale to examine the effect the 2 biopsy techniques might have on interpreting pathologist confidence. In addition, the Likert scale was applied to the clinical pathology report, as opposed rereviewing the pathology itself. FNA biopsies used a tandem needle technique limiting the applicability to centres using a coaxial FNA technique. The low incidence of major complications (chest tube placement, fatality) limits conclusions on patient safety issues. Conclusion In conclusion, we found compared to FNA, CN has a greater diagnostic yield with no difference in procedural complications. Improved diagnostic yield is secondary to a decrease in the rate of inadequate samples. It is important to note that lesion size was the primary driver of diagnostic yield, with smaller lesions resulting in fewer definitive diagnoses in either group. Therefore, surgical excision biopsy techniques should be considered in patients with small lesions. References [1] Klein J, Salomon G, Stewart E. Transthoracic needle biopsy with a coaxially placed 20-gauge automated cutting needle: results in 122 patients. Radiology 1996;198:715e20. [2] National Lung Screening Trial Team, Aberle DR, Adams AM, et al. Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl Med 2011;365:395e409. [3] Lourenc¸o R, Camacho R, Barata MJ, Canario D, Gaspar A, Cyrne C. CT-guided percutaneous transthoracic biopsy in the evaluation of undetermined pulmonary lesions. Rev Port Pneumol 2006;12:503e24. [4] Arakawa H, Nakajima Y, Kurihara Y, Niimi H, Ishikawa T. CT-guided transthoracic needle biopsy: a comparison between automated biopsy gun and fine needle aspiration. Clin Radiol 1996;51:503e6. [5] Staroselsky A, Schwarz Y, Man A, Marmur S, Greif J. Additional information from percutaneous cutting needle biopsy following fineneedle aspiration in the diagnosis of chest lesions. Chest 1998;113: 1522e5. [6] Ohno Y, Hatabu H, Takenaka D, Imai M, Ohbayashi C, Sugimura K. Transthoracic CT-guided biopsy with multiplanar reconstruction image improves diagnostic accuracy of solitary pulmonary nodules. Eur J Radiol 2004;51:160e8. [7] Yamagami T, Iida S, Kato T, Tanaka O, Nishimura T. Combining fineneedle aspiration and core biopsy under CT fluoroscopy guidance: a better way to treat patients with lung nodules? AJR Am J Roentgenol 2003;180:811e5.

[8] Moulton J, Moore P. Coaxial percutaneous biopsy technique with automated biopsy devices: value in improving accuracy and negative predictive value. Radiology 1993;186:515e22. [9] Yao X, Gomes MM, Tsao MS, Allen CJ, Geddie W, Sekhon H. Fine needle aspiration biopsy vs core needle biopsy in diagnosing lung cancer: a systematic review. Curr Oncol 2012;19:e16e27. [10] Szle˛zak P, Srutek E, Gorycki T, Kowalewski J, Studniarek M. Needle biopsy under CT fluoroscopy guidance and fine-needle aspiration cytology: comparison of diagnostic yield in the diagnosis of lung and mediastinum. Pol J Rad 2014;79:175e80. [11] Laurent F, Latrabe V, Vergier B, Montaudon M, Vernejoux JM, Dubrez J. CT-guided transthoracic needle biopsy of pulmonary nodules smaller than 20 mm: results with an automated 20-gauge coaxial cutting needle. Clin Radiol 2000;55:281e7. [12] Austin J, Cohen M. Value of having a cytopathologist present during percutaneous fine-needle aspiration biopsy of lung: report of 55 cancer patients and meta-analysis of the literature. AJR Am J Roentgenol 1993;160:175e7. [13] Geraghty PR, Kee ST, McFarlane G, Razavi MK, Sze DY, Dake MD. CT-guided transthoracic needle aspiration biopsy of pulmonary nodules: needle size and pneumothorax rate. Radiology 2003;229:475e81. [14] Yeow K. Risk factors of pneumothorax and bleeding multivariate analysis of 660 CT-guided coaxial cutting needle lung biopsies. Chest 2004;126:748. [15] Yildirim E, Kirbas I, Harman A, et al. CT-guided cutting needle lung biopsy using modified coaxial technique: factors effecting risk of complications. Eur J Radiol 2009;70:57e60. [16] Cox J, Chiles C, McManus C, Aquino SL, Choplin RH. Transthoracic needle aspiration biopsy: variables that affect the risk of pneumothorax. Radiology 1999;212:165e8. [17] Choi S, Chae E, Kim J, et al. Percutaneous CT-guided aspiration and core biopsy of pulmonary nodules smaller than 1 cm: analysis of outcomes of 305 procedures from a tertiary referral center. AJR Am J Roentgenol 2013;201:964e70. [18] Mayo J, Clifton J, Powell T, et al. Lung nodules: CT-guided placement of microcoils to direct video-assisted thoracoscopic surgical resection 1. Radiology 2009;250:576e85. [19] Laspas F, Roussakis A, Efthimiadou R, Papaioannou D, Papadopoulos S, Andreou J. Percutaneous CT-guided fine-needle aspiration of pulmonary lesions: results and complications in 409 patients. J Med Imaging Radiat Oncol 2008;52:458e62. [20] Wallace M, Krishnamurthy S, Broemeling L, et al. CT-guided percutaneous fine-needle aspiration biopsy of small (<1cm) pulmonary lesions. Radiology 2002;225:823e8. [21] Rizzo S, Preda L, Raimondi S, et al. Risk factors for complications of CT-guided lung biopsies. Radiol Med 2011;116:548e63. [22] Laurent F, Michel P, Latrabe V, Tunon de Lara M, Marthan R. Pneumothoraces and chest tube placement after CT-guided transthoracic lung biopsy using a coaxial technique: incidence and risk factors. AJR Am J Roentgenol 1999;172:1049e53. [23] Heyer C, Reichelt S, Peters S, Walther JW, M€uller KM, Nicolas V. Computed tomography-navigated transthoracic core biopsy of pulmonary lesions. Acad Radiol 2008;15:1017e26. [24] Montaudon M, Latrabe V, Pariente A, Corneloup O, Begueret H, Laurent F. Factors influencing accuracy of CT-guided percutaneous biopsies of pulmonary lesions. Eur Radiol 2004;14:1234e40. [25] Saji H. The incidence and the risk of pneumothorax and chest tube placement after percutaneous CT-guided lung biopsy: the angle of the needle trajectory is a novel predictor. Chest 2002;121:1521e6. [26] Khan M, Straub R, Moghaddam S, et al. Variables affecting the risk of pneumothorax and intrapulmonal hemorrhage in CT-guided transthoracic biopsy. Eur Radiol 2008;18:1356e63. [27] Anderson J, Murchison J, Patel D. CT-guided lung biopsy: factors influencing diagnostic yield and complication rate. Clin Radiol 2003; 58:791e7.