| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vascular and Interventional Radiology |
2-cm) Subpleural Pulmonary Lesions: Short- versus Long-Needle-Path CT-guided Biopsy—Comparison of Diagnostic Yields and Complications1
1 From the Departments of Diagnostic Radiology (S.G., F.A.M., M.J.W., K.A., D.C.M., R.M., M.E.H.), Pathology (S.K.), and Biostatistics (L.D.B.), University of Texas M.D. Anderson Cancer Center, 1515 Holcombe Blvd, Box 325, Houston, TX 77030. From the 2003 RSNA Annual Meeting. Received September 3, 2003; revision requested November 20; final revision received April 5, 2004; accepted April 28. Address correspondence to S.G. (e-mail: sgupta@di.mdacc.tmc.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: The study was approved by the institutional review board, and the requirement for informed patient consent was waived. The medical and imaging records of patients who underwent CT-guided biopsy of subpleural pulmonary nodules measuring up to 2 cm in diameter were reviewed. The study included 176 patients (79 men, 97 women; age range, 18–84 years) who were divided into two groups: In group A, a direct approach in which the needle traversed a short lung segment was used. In group B, an indirect approach involving the use of a longer needle path was used. Diagnostic yield, accuracy, and pneumothorax and chest tube placement rates were compared between the two groups. Two-tailed t tests and Pearson 
2 tests were used to analyze continuous and categorized variables, respectively.
RESULTS: Group A comprised 48 patients; and group B, 128 patients. The mean needle path length was 0.4 cm in group A and 5.6 cm in group B. The short-path approach necessitated more needle punctures (mean, 2.9 vs 1.8 with long-path approach, P < .001) through the pleura. The diagnostic yield in group A was significantly lower than that in group B (71% vs 94%, P < .001), particularly in patients with small (0–1-cm) nodules (40% in group A vs 94% in group B, P < .001). The frequency of postbiopsy pneumothorax was identical (69%) in the two groups. However, more group B than group A patients required chest tube placement for treatment of pneumothorax (38% vs 17%, P = .006).
CONCLUSION: Use of long-needle-path biopsy of subpleural lesions resulted in a higher diagnostic yield, especially for small nodules. However, compared with the short-needle-path technique, this approach was associated with a higher frequency of chest tube placement for pneumothorax.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Subpleural lesions can be sampled by using a direct approach involving a short intrapulmonary needle path or an indirect approach with which the needle traverses a longer segment of aerated lung. Tanaka et al (8) observed that using an oblique needle path, on which the needle was inclined 45 or more degrees within the same plane as the transverse CT section showing the nodule and advanced almost parallel to the pleura over a long distance, rather than the standard direct needle path resulted in a diagnostic rate increase from 43% to 81%.
Although a few other investigators (9,12) have advocated the use of a tangential approach to biopsy of subpleural lesions, data from comparisons of the diagnostic accuracy of the two techniques are limited. Also, the complication rates associated with the two approaches have not been extensively studied. The purpose of our study was to retrospectively compare the diagnostic yield and complications associated with using short versus long needle paths for CT-guided biopsy of small subpleural lung lesions.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Biopsy Procedures
The biopsies were performed by one of seven staff interventional radiologists (S.G., F.A.M., M.J.W., K.A., D.C.M., R.M., or M.E.H.), who had 3–10 years experience performing thoracic biopsy, or by a radiology trainee (resident or fellow) with direct supervision by one of the staff interventional radiologists. Written informed consent was obtained from each patient before each biopsy. All biopsies were performed with CT guidance by using a CTi Smart View scanner (GE Medical Systems, Milwaukee, Wis) with the patient in a prone, supine, or lateral decubitus position, depending on the lesion location. The CT images were obtained through the region of interest by using 3–5-mm collimation and were viewed by using lung window settings.
Conscious sedation was induced in all patients. The skin was aseptically prepared and draped, and 1% lidocaine (Xylocaine; Astra, Westborough, Mass) was administered subcutaneously to induce local anesthesia. Coaxial biopsy was performed by using an 18-gauge guide needle and an inner 22-gauge Chiba needle (Cook, Bloomington, Ind). With use of intermittent CT scans to evaluate the needle trajectory, the guide needle was inserted through the pleura and advanced to a position close to the target lesion. However, in some patients with peripheral lesions, in whom a short needle path was used, the guide needle was placed in the chest wall just outside the pleura. Subsequently, the inner 22-gauge needle was advanced coaxially through the guide needle into the lesion. After the needle tip’s position was confirmed on CT scans, samples were obtained by using an aspiration or capillary (nonaspiration) technique.
An on-site cytotechnologist prepared the biopsy sample slides, which were immediately evaluated for specimen adequacy by a cytopathologist. If there was any clinical or cytopathologic suggestion of infection, additional fine-needle aspirates were obtained for microbiologic evaluation. Patients for whom the initial cytologic evaluation was inconclusive underwent additional fine-needle aspiration biopsy, core-needle biopsy with a 20-gauge biopsy needle (Quick-Core; Cook), or both, if they were deemed safe and feasible by the radiologist.
After undergoing biopsy, the patients were monitored by the nursing staff in the radiology department holding area. Immediately or within 30 minutes after the biopsy, an expiratory posteroanterior chest radiograph was obtained with the patient in an upright position. If pneumothorax was absent, this examination was followed by another chest radiographic examination 3 hours after the biopsy. If the initial chest radiograph showed a pneumothorax, however, a follow-up radiograph was obtained after 1 hour. A chest tube was inserted if the pneumothorax was moderate or large (ie, >25% free air in the pleural cavity, as estimated at visual assessment of the chest radiograph); if the size of the pneumothorax seen on follow-up chest radiographs continued to increase; or if the patient experienced substantial pain, shortness of breath, or decreased oxygen saturation in the presence of a small pneumothorax. Chest tubes were placed by the interventional radiologist who performed the biopsy or by another interventional radiologist on clinical service.
Data Collection and Statistical Analyses
For this study, one radiologist (S.G., with 8 years experience performing chest biopsy) reviewed the pre- and intrabiopsy CT scans and the postbiopsy chest radiographs stored digitally on a Web-based image distribution system (Stentor, San Francisco, Calif) for the following parameters: The size of the lesion and the length of the needle path (ie, the length of the aerated lung segment traversed by the needle from the surface of the pleura to the proximal margin of the target lesion) were measured by using electronic calipers on a personal computer–based workstation (Windows; Microsoft, Redmond, Wash). The lesion location, lesion contact or noncontact with the pleura, patient position, biopsy approach (anterior, posterior, or lateral), crossing of the fissure by the biopsy needles, number of visceral pleural layers punctured, total number of pleural punctures, biopsy sample type (fine-needle aspiration, core-needle biopsy, or both), and development of pneumothorax, hemorrhage, or both during the biopsy also were noted. The patients’ computerized records were reviewed (S.G., M.J.W.) for the following data: presence or absence of pneumothorax on the postbiopsy chest radiographs, placement or nonplacement of chest tube, and results of cytopathologic, histopathologic, and microbiologic examinations of the biopsy specimens.
We divided the patients into two groups according to the length of the transpulmonary needle path: In group A, a direct approach in which the needle traversed a short (
1-cm) segment of aerated lung was used. In group B, an indirect approach involving the use of a longer (>1-cm) transpulmonary needle path was used. The choice of approach used in any given patient was based on the discretion of the interventional radiologist performing the biopsy.
The cytopathologic and surgical-pathologic findings were reviewed (by S.K.) and categorized as malignant, suggestive of malignancy, benign, or nondiagnostic. A definite malignant diagnosis or one suggestive of malignancy was considered a positive result. Results were considered negative for malignancy when the specimen contained no malignant cells, showed specific benign findings (eg, granuloma, hamartoma, cores of fibrosis), or showed nonspecific benign changes (eg, giant cells, leukocytes, histiocytes, fragments of fibrosis) or when a microorganism was identified at culture or histopathologic examination.
The biopsy was deemed nondiagnostic if the specimen contained few cells and showed only skeletal muscle cells, blood or normal lung cells, or a few atypical cells and a definite opinion regarding the benignancy or malignancy of the lesion could not be rendered or if an adequate sample could not be obtained because the presence of pneumothorax or hemorrhage complicated the procedure.
Follow-up clinical, imaging, and pathologic records also were reviewed (by S.K. and S.G.) so that a final decision regarding the benign or malignant nature of the sampled lesion could be reached. A positive biopsy result was considered to be true-positive when there was surgical confirmation, when the histologic findings were compatible with the patient’s known primary malignancy, when the results of biopsy of another site revealed cancer with the same histologic characteristics, when the lesion increased in size and other proved metastases were found, or when the patient was treated for malignancy and the subsequent clinical course and response to therapy were deemed appropriate. A positive biopsy result was considered to be false-positive when there was no evidence of malignancy at surgical resection without preoperative chemotherapy or when there was nodule regression at follow-up imaging in the absence of therapy. A negative biopsy result was considered to be true-negative when no tumor was identified at histopathologic examination of the surgical specimen, when the lesion subsequently disappeared or decreased in size with or without antibiotic administration, or when the lesion remained stable at follow-up CT for at least 12 months. The identification of organisms at culture or histopathologic examination was considered to be a true-negative result. Nodule growth or a malignant lesion confirmed at surgery was considered to be a false-negative finding.
We compared the following results between the two biopsy groups: percentage of samples considered adequate for histopathologic analysis, diagnostic accuracy, sensitivity, and specificity. Diagnostic accuracy was calculated by comparing the histopathologic diagnosis with the final diagnosis based on surgical, culture, or clinical and/or radiologic follow-up findings. Patients in whom the tissue specimens were deemed nondiagnostic and in whom a definite diagnosis of lesion benignancy or malignancy could not be established—either because they were lost to follow-up or because the radiologic follow-up period for a lesion suspected to be benign was less than 12 months—were excluded from the statistical analysis used to estimate diagnostic accuracy, sensitivity, and specificity.
Data analyses were performed by a biostatistician (L.D.B.) with use of computer software (SPSS, version 11.0; SPSS, Chicago, Ill). The two biopsy groups were compared with respect to patient demographics, lesion characteristics, biopsy technique–related variables, complications, diagnostic yield, and accuracy. We also evaluated the effects that lesion size and lesion contact or noncontact with the pleura had on the diagnostic yield and the pneumothorax and chest tube placement rates associated with the two biopsy techniques. The significance calculated from the mean of a continuous variable was analyzed by using two-tailed t tests, and the categorized variables were analyzed by using Pearson 2 tests. P < .05 was considered to indicate statistical significance.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
Of the 128 patients in group B, 44 (34%) required more than one pleural puncture, and in only eight of these patients was this related to the needle slipping out of the pleura. In the remaining 36 of these 44 patients, multiple pleural punctures were secondary to traversal of the fissure (n = 11) or inadvertent puncture of the contralateral pleural surface that was abutting the lung nodule during sampling (n = 25).
Diagnostic Yield
The histopathologic findings in the two patient groups are summarized in Table 3. In group A, 71% of the biopsy procedures yielded sufficient material for cytologic and/or histologic analysis. In contrast, 94% of the biopsy procedures in group B yielded diagnostic samples. This between-group difference was statistically significant (P < .001). After excluding the nondiagnostic biopsy procedures and the patients in whom the final diagnosis could not be established, we found that the diagnostic accuracy, sensitivity, and specificity were similar between the two groups.
|
|
|
|
|
Alveolar hemorrhage along the needle path and/or around the target lesion was seen at 17 (35%) of the 48 biopsies performed in group A and in 97 (76%) of the 128 biopsies performed in group B. In nine (7%) group B patients, pulmonary hemorrhage totally obscured the lesion and precluded any further biopsy attempts; this problem was not encountered in any group A patient.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Moreover, repeat access into the lung is difficult in the presence of a pneumothorax because the lung is no longer fixed in position. In our study, biopsies involving short intrapulmonary paths were associated with the development of pneumothorax during the procedure in 73% of the patients, and the pneumothorax prevented further sampling or necessitated additional pleural punctures in 31% of the patients.
Some authors have advocated the use of a near-tangential or oblique approach in which the needle is advanced parallel to the pleura to circumvent the problems associated with a direct approach (8). However, biopsy performed by using a tangential approach also may be unsuccessful if a short transpulmonary path is chosen because the needle tends to either slip over the pleura during entry or fall out of the lung during breathing. On the basis of our experience, we believe that it is not the obliquity of the needle path in relation to the pleura but rather the length of the transpulmonary path that is more important in preventing these problems during CT-guided biopsy of subpleural nodules. Sometimes, if a tangential approach will enable the traversal of only a short segment of the lung, it will be more convenient to use an indirect approach from the opposite side of the lung.
A long transpulmonary needle path, by enabling the use of a posterior or anterior approach to avoid overlying bone structures, offers more access options than the short path (8,9). Use of the long path facilitates secure anchoring of the guide needle, making it less prone to dislodgment from the lung during respiration and thus substantially reducing the likelihood of intraprocedural pneumothorax, which was seen in 30% of the group B patients as compared with 73% of the group A patients in the current study. Even if a pneumothorax does occur during biopsy performed with this approach, the guide needle usually is not displaced from the lung, so further sampling can be performed.
Another advantage to using the long transpulmonary path is that the distance between the pleural puncture site and the lesion leaves considerable room for course adjustment. The needle can be partially withdrawn and readvanced in the correct direction without having to be pulled back across the pleural surface. In addition, slightly imperfect needle trajectories can be corrected by pulling the guide needle back 1 or 2 cm proximal to the lesion and applying torque on the hub of the guide needle to direct the passes of the inner biopsy needle in the desired direction. Also, the inner 22-gauge biopsy needle tip may be given a small curve to correct its course at the time of sampling. These techniques enable most needle malpositions to be corrected without placing additional needles or repuncturing the pleura. They also allow a larger area of the lesion to be targeted for sampling from different parts of the lesion and thus potentially facilitate increases in the diagnostic yield.
The higher prevalence of intrapulmonary hemorrhage along the needle track seen with use of the long needle path was not unexpected because use of a longer transpulmonary path increases the probability of blood vessels along the course to the lesion being injured. Although this is usually no cause for concern, alveolar bleeding around the lesion has the potential to obscure the lesion, precluding further attempts at sampling. This occurred in 7% of the group B patients in our study. In most cases, however, the lesion can be seen through the hemorrhage because of differences in lesion density, so more attempts at sampling are possible.
In their series of 61 peripheral lung nodules smaller than 2.5 cm that were analyzed with CT-guided biopsy, Tanaka et al (8) reported that an oblique approach was significantly more likely to be effective (in 81.2% of cases) for obtaining a diagnostic sample than a direct right-angled approach (in 43.3% of cases). In another study involving 30 subpleural nodules measuring up to 1 cm, Wallace et al (9) observed a significant difference in accuracy between tangential (100%) and direct (64%) approaches. In our study also, the diagnostic yield with the long-needle-path approach (94%) was significantly higher than that with the short-needle-path approach (71%).
Various factors, most of which have been mentioned earlier and which include needle dislodgment due to breathing, pneumothorax development during biopsy, the overlying rib preventing needle angulation, and the short distance between the pleural entry site and the lesion leaving no room for course correction, contribute to the increased technical difficulty of obtaining an adequate sample—and thus the lower diagnostic yield—with the direct approach (8,9,11,13). The smaller the lung nodule, the more difficult it becomes to "hit" the nodule (ie, insert the needle into the nodule) by using the direct approach. In the present study, the diagnostic yield with the short approach was significantly lower for 0–1-cm lesions (40%) than for lesions larger than 1 cm to 2 cm in diameter (79%).
Use of the real-time capability of CT fluoroscopy has been advocated for needle biopsies of pulmonary nodules. Katada et al (14) reported that the diagnostic accuracy of CT fluoroscopy–guided biopsy was 100% for lesions larger than 10 mm in diameter but only 67% for those 10 mm in diameter or smaller. Irie et al (15) found that although CT fluoroscopy involved a reduced procedure time, its use did not lead to improved diagnostic accuracy in patients with lesions 1 cm in diameter or smaller. These results show that obtaining diagnostic samples from small lesions is difficult, even with CT fluoroscopy. Moreover, CT fluoroscopy–guided procedures expose the radiologist to more radiation.
In our study, the two biopsy techniques were associated with similar rates of pneumothorax. Cox et al (16) observed that although the rate of pneumothorax dramatically increased when any amount of aerated lung was traversed (50% vs 15% when no segment of aerated lung was traversed to sample pleura-based lesions), this rate did not increase with increasing lesion depth. Other investigators, however, have reported that the risk of pneumothorax increases with increasing lesion depth (17–21). In our study, the rate of postbiopsy pneumothorax in group A was as high as that in group B, despite the significantly shorter transpulmonary path used in group A. Similar findings have been reported by other investigators (8,12,22). Tanaka et al (8) noted a pneumothorax rate of 20% for biopsies of subpleural nodules that were performed with a near-90° short approach, as compared with a 12.5% pneumothorax rate with a longer oblique approach. Moulton and Moore (22) observed an increased risk of pneumothorax with subpleural lesions, especially those abutting the pleural surface. Yeow et al (12) reported that the risk of pneumothorax associated with needle biopsy of a subpleural lesion within 2 cm of the pleural surface was seven times higher than that associated with needle biopsy of deeper lesions.
As mentioned earlier, using a short needle path frequently results in the inadvertent dislodgment of the guide needle into the pleural space, which allows air ingress. In our study, the significantly larger number of pleural punctures required in the group A patients also may have contributed to the comparable rates of pneumothorax in the two groups. In a recent study, the pneumothorax rate in patients with single punctures (18%) was significantly lower than that in patients with three punctures (73%) (21).
The relatively high overall rate of pneumothorax in our study patients, as compared with the rates reported in the literature, could be related to various factors. One possible explanation could be that we studied only the cases of patients with small lesions (up to 2 cm). The results of several studies (16,18,20) have shown that the pneumothorax rate increases with decreasing lesion size. In the study by Cox et al (16), the pneumothorax rate was 64% for lesions 2 cm in diameter or smaller and 41% for lesions larger than 2 cm in diameter. Some investigators have also suggested that CT-guided lung biopsy may be associated with a higher prevalence of pneumothorax than fluoroscopically guided lung biopsy because of the increased duration of the needle placement (16,23). All of the biopsies in our series were performed by using CT guidance. The large number of pleural punctures (more than one puncture in 69% of group A patients and 34% of group B patients) in our study may have been another contributory factor. The number of pleural punctures can affect the pneumothorax rate: Ohno et al (21) found that the pneumothorax rates associated with one, two, and three pleural punctures were 18%, 53%, and 73%, respectively.
As expected, the presence or absence of pleural contact did not affect the pneumothorax rate in the patients who underwent long-needle-path biopsy because the lesion abutting or not abutting the pleura had no effect on the length of the aerated lung segment traversed by the needle with this approach. However, contrary to the findings of some previous studies (11,16), which suggest that direct biopsy of pleura-based lesions is associated with a relatively low risk of pneumothorax because no part of the aerated lung is traversed, we observed no significant difference in pneumothorax rate based on pleural contact in the patients who underwent short-needle-path biopsy. This minimal difference in pneumothorax rate is probably because the problems associated with using a short needle path—namely, needle dislodgment with breathing; failure to hit the lesion during the first attempt, with resulting inadvertent traversal of the adjacent normal aerated lung region; and the frequent need for multiple pleural punctures—which increase the risk of pneumothorax with this approach, can happen irrespective of whether or not the lesion is abutting the pleura. These problems are more likely to be encountered with small lesions, as was the case in our study. In the group of patients who had nodules that were in contact with the pleural surface and underwent short-needle-path biopsy, pneumothorax was more common in those patients who had 0–1-cm lesions (100%) than in those who had lesions larger than 1 cm to 2 cm in diameter (62%). These findings are consistent with Moore’s (11) observation that a direct approach for biopsy of a pleural-based lesion should be used only when the lesion is large enough to ensure a direct hit during the first attempt.
Although the rates of pneumothorax were similar between the two biopsy techniques in this study, the chest tube placement rate was significantly higher in the patients who underwent biopsy involving the long transpulmonary needle path than in those who underwent biopsy involving the short path. This was not surprising considering the fact that a biopsy approach involving the traversal of a very short segment of aerated lung is unlikely to result in substantial air leakage. Also, a lung that is peripheral to a subpleural lesion is likely to be collaterally ventilated because of the obstruction of the main bronchus by the lesion itself, and there is unlikely to be substantial air leakage from an area of collaterally ventilated lung (11).
One potential limitation of our study was that its retrospective nature may have led to bias in the comparisons of the two patient-biopsy groups because the choice of biopsy approach could have been influenced by the size and location of the lesion or by the presence or absence of pleural contact. However, the fact that the two groups had similar patient demographics and lesion characteristics (Table 1) suggests that such bias was avoided, and, thus, a statistically meaningful comparison of the two biopsy techniques was possible. Another limitation of the study was the fact that, due to the retrospective nature of the data collection, we had no control over the sample sizes and a power analysis was not performed to calculate the sample sizes. However, it appears that the sample sizes for the two groups were sufficient to enable the detection of differences in the major end points of this study.
In conclusion, the results of this study suggest that the use of a long transpulmonary needle path for biopsy of subpleural lesions results in a higher diagnostic yield, especially for small nodules; however, this approach is associated with a higher frequency of chest tube placement compared with the short-needle-path approach.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, S.G.; study concepts, S.G., S.K.; study design, S.G.; literature research, S.G.; clinical studies, S.G., M.J.W., F.A.M., K.A., D.C.M., R.M., M.E.H.; data acquisition, S.G., M.J.W.; data analysis/interpretation, S.G., F.A.M.; statistical analysis, L.D.B., S.G.; manuscript preparation, S.G., D.C.M.; manuscript definition of intellectual content, S.G., K.A., R.M.; manuscript editing, S.G., K.A.; manuscript revision/review and final version approval, S.G.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
1-cm) pulmonary lesions. Radiology 2002; 225:823-828.This article has been cited by other articles:
|
|
 |
|
  H. G. Welch, S. Woloshin, L. M. Schwartz, L. Gordis, P. C. Gotzsche, R. Harris, B. S. Kramer, and D. F. Ransohoff Overstating the Evidence for Lung Cancer Screening: The International Early Lung Cancer Action Program (I-ELCAP) Study Arch Intern Med, November 26, 2007; 167(21): 2289 - 2295. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Eberhardt, D. Anantham, F. Herth, D. Feller-Kopman, and A. Ernst Electromagnetic Navigation Diagnostic Bronchoscopy in Peripheral Lung Lesions Chest, June 1, 2007; 131(6): 1800 - 1805. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. R. Gildea, P. J. Mazzone, D. Karnak, M. Meziane, and A. C. Mehta Electromagnetic Navigation Diagnostic Bronchoscopy: A Prospective Study Am. J. Respir. Crit. Care Med., November 1, 2006; 174(9): 982 - 989. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
