Effectiveness and Safety of Image-Directed Biopsies: Coaxial Technique Versus Conventional Fine-Needle Aspiration.

ABSTRACT

Background. We compared the effectiveness and safety of image-guided biopsies done with coaxial guides versus fine-needle aspiration done without coaxial guides.

Methods. With the use of hospital computer records and chart reviews, all image-guided biopsies done during a 4-year period at our institution were assessed for adequacy and complications. For each biopsy, the use of a coaxial guide, the site, and the imaging modality were recorded. Adequacy of the biopsy and complications were compiled. Success rates were calculated for conventional and coaxial biopsies and by modality and site.

Results. Coaxial technique reduced the number of unsatisfactory biopsies compared with conventional technique in extrathoracic sites. The decrease was statistically significant. No major complications occurred from extrathoracic biopsies with either technique. No difference was found in success rates or complication rates between ultrasound-guided and CT-guided biopsies using coaxial technique.

Conclusion. Coaxial technique reduces the number of inadequate biopsies in extrathoracic sites, without a detectable increase in complications.

The purpose of coaxial technique is to maintain access to a lesion after successful needle placement. This allows multiple specimens to be obtained without additional passes through overlying tissue. An increase in the proportion of successful biopsies would be expected when using coaxial technique because it allows sampling of a larger tissue volume and special tissue studies (eg, flow cytometry). Further, coaxial guides facilitate the use of cutting needle biopsy guns, which have previously been shown to provide a higher diagnostic yield than standard biopsy needles. (1-4) Safety of the technique is harder to predict. Although the number of passes through overlying tissue should be reduced, the number of passes into the lesion may be increased and larger gauge needles are used. A previous single institution study (5) found a 100% adequacy rate for core biopsies and a 96% adequacy rate for fine-needle aspiration when coaxial technique was used. The major complication rate was 3%. However, the authors did not provide a definition of an adequate histology specimen, and their criteria for an adequate cytology specimen were liberal (only acellular specimens and specimens consisting solely of blood were considered inadequate). In addition, the study contained no control group to see whether these high success rates represented a change from conventional technique. (5) We decided to study the safety and diagnostic yield of image guided biopsies done with coaxial guides compared with fine-needle aspiration done without coaxial guides at our institution.

METHODS

All image-guided biopsies done at our institution over a 4-year period were assessed for adequacy and complications using hospital computer records and chart review. If a specific pathologic diagnosis was made, the specimen was classified as adequate. If the biopsy was read as unsatisfactory or insufficient for diagnosis, it was classified as inadequate. Biopsies read as normal or with nonspecific pathologic diagnoses were classified as adequate if they were used to determine management. Normal findings or nonspecific diagnoses were considered inadequate if repeat biopsy was done or if biopsy results were ignored. Biopsy material was reviewed by one of us (R.H.A.) if the initial reading was normal, nonspecific, or unsatisfactory.

Radiology residents under the direct supervision of staff radiologists did the biopsies. The coaxial biopsies were done using 17 or 19 gauge Temno (Bauer Medical, Clearwater, Fla) or Monopty (Bard, Livingston, Ga) guides with their corresponding 18 and 20 gauge biopsy guns and 20 or 22 gauge Chiba (Cook, Bloomington, Ind) needles. The conventional biopsies were done using Chiba needles ranging from 18 to 22 gauge. A Picker International 1200 (Cleveland, Ohio) or an Elscint CT Twin (Haifa, Israel) was used for CT-guided procedures. An American Technology Laboratory Ultramarak 9 HDI (Bothell, Wash) was used for ultrasound-guided procedures. Cases were assigned randomly among a group of eight staff radiologists, according to the coverage schedule and the time of submission of the request. The choice of imaging modality, needle type, and size was made by the supervising radiologist. The procedures were routinely attended by a cytopathologist for immediate evaluation of the tissue sampled, by fine-needle aspirati on in most cases. This was accomplished by preparing 1 or 2 slides with a portion of the specimen, air-drying, staining with Diff-Quik stain, and viewing for adequacy. If cytologic evaluation showed an adequate specimen, core biopsies were then obtained when coaxial technique was used. In some cases, the cytology specimen was obtained from the core biopsy tissue by either dragging the core over a slide (touch preparation), scraping the core with a slide for application to a second slide (scrape preparation), or compressing the specimen between two slides (crush preparation). All needle biopsies and cytology specimens were read jointly.

For each biopsy, the use of a coaxial guide, the modality, the largest needle gauge, and the total number of needle passes were recorded. The site of the biopsy was recorded and classified as abdominopelvic, chest, superficial soft tissue, or bone. Adequacy of the biopsy and complications were compiled. If the use of a coaxial needle could not be determined, the case was excluded from the study. Success rates were calculated for conventional and coaxial biopsies and by site. The data were stratified by site and analyzed by the MantelHaenszel common odds ratio estimate. Calculations were repeated with chest data removed because so few chest biopsies in our study were done with coaxial technique. Success rates were calculated by imaging modality and analyzed by Fisher's Exact Test. The median number of needle passes for coaxial and conventional biopsies was calculated and analyzed by the Wilcoxon rank sum test. The mean number of needle passes was also calculated.

RESULTS

Of 32 abdominal biopsies done by conventional technique, 25 (78%) were successful. Of 35 abdominal biopsies done using coaxial technique, 32 (91%) were successful. Of 56 chest biopsies done by conventional technique, 50 (89%) were successful. All 3 of the chest biopsies done by coaxial technique were successful (100%). Four of 5 biopsies of superficial soft tissues done by conventional technique were successful (80%), all 4 of those done by coaxial technique were successful (100%). Of 3 bone biopsies done by conventional technique, 2 (67%) were successful, and all 4 bone biopsies done by coaxial technique were successful (100%). Forty extrathoracic biopsies were done using conventional technique; 31 (78%) were successful. Forty-two extrathoracic biopsies were done using coaxial technique; 39 (93%) were successful.

A total of 96 biopsies were done using conventional technique; 81 (84%) were successful. A total of 48 biopsies were done by coaxial technique; 45 (94%) were successful (Table). The differences were statistically significant for the overall data (P < .05) and the extrathoracic data (P < .05). Of 16 biopsies done by coaxial technique with ultrasound guidance, 15 (93%) were successful. Of 33 biopsies done by coaxial technique with CT guidance, 31 (94%) were successful. There was no statistical difference between ultrasound and CT success rates (P = 1.0). No biopsy was reclassified on review.

Seven major complications (5 large pneumothoraces and i case each of hemoptysis and chest pain) occurred during conventional chest biopsies (12.5%). Nine minor complications (9 small pneumothoraces) occurred during conventional chest biopsy (16%). A pneumothorax was considered large if a chest tube was placed and small if no treatment was required. One minor complication (small pneumothorax) occurred during coaxial chest biopsy (33%). One minor complication (pain requiring cessation of the procedure) occurred during coaxial abdominal biopsy (3%). No other complications were found.

The total number of needle passes was available for 90 conventional biopsies and 33 coaxial biopsies. In coaxial biopsies, the records usually did not specify the number of passes made with the coaxial guide, biopsy gun, or aspiration needle but gave only the total number of passes. A mean of 2.1 [+ or -] 0.10 passes and a median of 2.0 passes were made in conventional biopsies. A mean of 3.4 [+ or -] 0.25 passes and a median of 3.0 passes were made in coaxial biopsies. The difference in the medians is statistically significant (P< .0001).

The maximum needle gauge used was available for 79 conventional biopsies and 20 coaxial biopsies. One 18 gauge, nineteen 20 gauge, eight 21 gauge, and fifty-one 22 gauge needles were used in conventional biopsies. Eleven 17 gauge and nine 19 gauge coaxial guides were used.

DISCUSSION

Relative Safety and Effectiveness

In this study, all but one of the conventional biopsies was done with smaller gauge needles than the smallest gauge coaxial guide used. In addition, the number of needle passes was significantly higher when coaxial technique was used. This increased aggressiveness resulted in a significantly higher diagnostic yield for coaxial technique. In extrathoracic biopsies, the diagnostic yield was increased by 15%. However, there was no appreciable difference in the low number of complications between conventional and coaxial methods in extrathoracic biopsies. These results are similar to those from previous studies, which showed low complication rates in abdominal biopsies.(1-6) Although not addressed by our study, it has been our experience that coaxial technique decreases patient discomfort and the amount of time to complete a biopsy because repeated needle positioning is not necessary. We have also noticed that coaxial guides are easier to visualize than conventional needles by both CT and ultrasound because of t heir larger gauges. On the basis of the increased diagnostic yield without detectable loss of safety, we recommend the use of coaxial technique for all image-guided extrathoracic biopsies, with possible rare exceptions. Because of the small number of thoracic biopsies done with coaxial technique in our study, we cannot assess its relative safety and effectiveness in the thorax. Use of coaxial technique was recommended for almost all thoracic biopsies in a recent article.(7)

In our study, no difference was found in the diagnostic yield between CT-guided and ultrasound-guided biopsies when a coaxial guide was used. (No ultrasound-guided procedures were done without a coaxial guide [Fig 1]). This contrasts with the results of a previous study that showed a higher diagnostic yield for ultrasound-guided procedures.(8)

Choice of Pathologic Specimen

Placement of a coaxial guide allows both core biopsy and tine-needle aspiration specimens to be obtained. Cytologic specimens not only provide a method for immediate confirmation of needle placement within a lesion, but also provide cytologic detail that is sometimes superior to that seen in core biopsies. Hence, cytologic specimens provide information complementary to that provided by core biopsy and frequently are diagnostic alone. Fine-needle aspiration is also useful when flow cytometric studies are needed because aspiration helps to separate and disengage the cells from surrounding stroma. Core biopsies show portions of the architecture, matrix, and border of a lesion and may provide a more definitive pathologic diagnosis when cytology is inconclusive.

In all series comparing core needle biopsy and fine-needle aspiration, cases were reported in which cytologic diagnosis was correct when histologic diagnosis was incorrect and vice versa. (5) Therefore, we chose to routinely obtain both types of specimen. However, if immediate cytologic analysis is done and the attending cytopathologist can make a definitive diagnosis from an aspirate, a core biopsy may be unnecessary. One may chose to do core biopsies first, since cytologic preparations may be prepared by touch, crush, or scrape preparations of core biopsy tissue, thus saving the expense of an aspiration needle (Figs 2 and 3) Touch preparations have previously been shown to adequately evaluate needle placement. (9) Crush and scrape preparations frequently yield more cells per slide than touch preparations but may destroy a core biopsy. Fortunately, this is usually acceptable when using a coaxial guide, since additional cores. can be easily obtained in most cases. Frozen sections of core biopsies are also po ssible but are particularly difficult on core biopsies obtained with 18 or 20 gauge guns. Some institutions obtain multiple core biopsies and/or perform fine-needle aspiration without cytologic review of adequacy, which decreases the time and manpower necessary for the procedure (N. Evans, MD, oral communication, July 1999; M. Gavant, MD, oral communication, November 1999).

Use in Special Situations

Bone Biopsy. All four coaxial and three of four conventional biopsies of bone tumors were successful (Fig 4). The cortex was eroded in all of these cases. Thus, bone biopsy can be done without bone biopsy needles in selected cases. If a coaxial guide is used and attempts at bone biopsy with standard needles fail, a trephine needle biopsy can be done without additional passes through overlying tissue. Passage of a trephine needle through a coaxial guide to perform bone biopsy has been described previously. (10)

Biopsy of Suspected Lymphoma. Coaxial technique is particularly useful in the diagnosis of lymphoma because multiple specimens are essential for histologic, immunophenotypic, molecular, and cytogenetic studies. However, the use of percutaneous biopsy in the diagnosis of lymphoma is controversial and may not be adequate in every case. Subclassification of lymphomas depends, in part, on architectural arrangement of neoplastic cells, which may be more difficult to assess on core biopsy than on a surgically resected specimen and is impossible to assess on aspirates. Also, more than one cell type may be present (eg, T-cell-rich B-cell lymphoma), which could go undetected or be misinterpreted because of less extensive sampling. Despite these concerns, the diagnostic accuracy ranges from 68% to 94% for percutaneous biopsy of suspected lymphoma. (11) In addition, some patients can be treated without exact sub-typing because of a definite previous diagnosis or poor prognostic indicators. (11) In two large series of ly mphoma diagnoses by core biopsy, 72% (12) and 86% (11) of patients were treated on the basis of core biopsy results alone. In our series, 3 of 4 coaxial-guided biopsies of lymphoma were used to determine treatment.

Limitations of the Study

Nonspecific and normal pathologic diagnoses are difficult problems in clinical practice and for this study. A normal pathologic diagnosis may represent the benign histology of a lesion seen on an imaging study (eg, adrenal adenoma) or may represent sampling error (eg, normal adrenal adjacent to a metastasis). Likewise, a nonspecific histologic diagnosis may represent the true histology of a lesion (eg, chronic inflammation and fibrosis may be the only findings in a chronic abscess) or may represent sampling error (chronic inflammation and fibrosis may be seen at the edge of a malignancy). We divided nonspecific and normal findings into adequate and inadequate categories according to whether the pathologic diagnoses were used to determine management. Since consultation between clinician, radiologist, and pathologist usually occurred when the biopsy was reported as nonspecific or normal and the various explanations were discussed, this is a reasonable way to categorize these cases as adequate or inadequate. Alt hough we recognize that this method is subject to error, it is more accurate than placing all these cases into one category.

This study was done retrospectively and is subject to some of the usual limitations of this method. There were problems with incomplete data collection. Information on needle gauge and number of needle passes was available in 97 and 90 cases, respectively, of a total of 144. When a coaxial guide was used, the records rarely specified the number of passes made with a coaxial guide versus the number made through the guide. Although we believe that it is safe to assume that the increased number of needle passes found in coaxial biopsies resulted from an increased number of passes through the guide, we do not have data to support this. Fortunately, this is a peripheral issue in our study. Since individual attending physicians chose the biopsy technique that was used, selection bias could have been introduced. Almost all chest biopsies were done with conventional technique. We addressed this problem by calculating success rates with and without including the chest biopsies. The difference was minimal. Attending ph ysicians may have chosen coaxial technique for more difficult cases. This would decrease the apparent diagnostic yield of coaxial technique. If such a bias existed in our study, an increase in diagnostic yield was shown nonetheless.

CONCLUSION

Our study showed a statistically significant increase in diagnostic yield for image-guided extrathoracic biopsies when a coaxial guide was used. There was no detectable loss of safety. Successful biopsies of lymphoma and certain bone lesions were demonstrated. Coaxial technique appears to decrease the time and patient discomfort of an image-directed biopsy. Coaxial guides are more easily visualized than conventional needles by both CT and ultrasound. Because of these multiple advantages, we use coaxial technique almost exclusively when performing image-guided biopsies and recommend that others do the same.

Acknowledgment. We thank William H. Frawley, PhD, for assistance with statistical analysis.

From the Departments of Radiology and Pathology, University of Texas Southwestern Medical Center at Dallas.

Reprint requests to Alan H. Appelbaum, MD, University of Tennessee, Department of Radiology, 800 Madison Ave, Memphis, TN 38163.

References

(1.) Hopper KD, Abendroth CS, Sturtz KW, et al: Blinded comparison of biopsy needles and automated devices in vitro: biopsy of diffuse hepatic disease. AJR 1993; 161:1293-1297

(2.) Hopper KD, Abendroth CS, Sturtz KW, et al: Blinded comparison of biopsy needles and automated devices in vitro: biopsy of medical renal disease. AJR 1993; 161:1299-1301

(3.) Parker SH, Hopper KD, Yakes WF, et al: Image-directed percutaneous biopsies with a biopsy gun. Radiology 1989; 171:663-666

(4.) Nyman RS, Cappelen-Smith J, Brismar J, et al: Yield and complications in ultrasound-guided biopsy of abdominal lesions. Acta Radiologica 1995; 36:485-490

(5.) Mouton JS, More PT: Coaxial percutaneous biopsy technique with automated biopsy devices: value in improving accuracy and negative predictive value. Radiology 1993; 186:515-522

(6.) Smith EH: Complications of percutaneous abdominal fine needle biopsy. Radiology 1991; 178:253-258

(7.) Moore EH: Technical aspects of needle aspiration lung biopsy: a personal perspective. Radiology 1998; 208:303-308

(8.) Sheafor DH, Paulson EK, Simmons CM, et al: Abdominal percutaneous interventional procedures: comparison of CT and US guidance. Radiology 1998; 207:705-710

(9.) Hahn PF, Eisenberg PJ, Pitman MB, et al: Cytopathologic touch preparations (imprints) from core needle biopsies: accuracy compared with that of fine-needle aspirates. AJR 1995; 165:1277-1279

(10.) White LM, Schweitzer ME, Deely DM: Coaxial percutaneous needle biopsy of osteolytic lesions with intact cortical bone. AJR 1996; 166:143-144

(11.) Ben-Yehuda D, Polliach A, Okon E, et al: Image-guided core needle biopsy in malignant lymphoma: experience with 100 patients that suggests the technique is reliable. J Clin Oncol 1996; 198:371-375

(12.) Silverman SG, Lee BY, Mueller PR, et al: Impact of positive findings at image-guided biopsy of lymphoma on patient care: evaluation of clinical history, needle size, and pathologic findings on biopsy performance. Radiology 1994; 190:759-764

TABLE

Adequacy by Site and Biopsy Type

                Inadequate  Adequate  Success Rate

Chest
  Conventional      6          50         89%
  Coaxial           0           3        100%
Abdominal
  Conventional      7          25         78%
  Coaxial           3          32         91%
Soft Tissue
  Conventional      1           4         80%
  Coaxial           0           4        100%
Bone
  Conventional      1           2         67%
  Coaxial           0           3        100%
Extrathoracic
  Conventional      9          31         78%
  Coaxial           3          39         93%
Total
  Conventional     15          81         84%
  Coaxial           3          45         94%

RELATED ARTICLE: KEY POINTS

* This study showed a 15% increase in success rates of image-guided biopsies in extrathoracic, sites when coaxial technique was used. This increase was statistically significant.

* There was no detectable change in complication rates when coaxial technique was used.

* Coaxial technique appears to decrease the time and patient discomfort of image-directed biopsies.