(Radiology. 1999;213:568-572.)
© RSNA, 1999
Vascular and Interventional Radiology |
Orbital Mass Lesions: US-guided Fine-Needle Aspiration Biopsy-Experience in 37 Patients1
Sanjay Gupta, MD,
Bimal Sood, MD,
Madhu Gulati, MD,
Deepak Takhtani, MD,
Rajiv Bapuraj, MD,
Niranjan Khandelwal, MD,
Usha Singh, MD,
Arvind Rajwanshi, MD,
Subhash Gupta, MD and
Sudha Suri, MD
1 From the Departments of Radiodiagnosis (Sanjay G., B.S., M.G., D.T., R.B., N.K., S.S.), Ophthalmology (U.S.), and Cytology (A.R., Subhash G.), Post Graduate Institute of Medical Education and Research, Chandigarh 160012, India. Received November 25, 1998; revision requested December 28; revision received January 29, 1999; accepted April 9. Address reprint requests to Sanjay G. (e-mail: dsgupta@yahoo.com).
 |
Abstract
|
|---|
PURPOSE: To evaluate the safety and effectiveness of using ultrasonographic (US) guidance for performing fine-needle aspiration biopsies of orbital mass lesions.
MATERIALS AND METHODS: Thirty-seven patients with mass lesions in the orbit underwent US-guided fine-needle aspiration biopsy. Computed tomographic scans were available in all patients. In 19 patients, the lesions were located in the posterior orbit, whereas in 18 patients the lesions were located in (n = 3) or extended up to (n = 15) the anterior compartment. Fine-needle aspiration biopsy was performed with 22–25-gauge needles and use of the freehand technique.
RESULTS: Needle biopsies were performed safely and easily, and real-time US monitoring of the needle position was used to avoid injury to the eyeball. No major complications were encountered. Diagnostic specimens were obtained in 29 (78%) of the 37 patients, and 19 benign and 10 malignant disease processes were diagnosed. In eight patients (22%), an appropriate diagnosis could not be made, as aspiration samples yielded insufficient tissue.
CONCLUSION: US provides safe and effective guidance for performing fine-needle aspiration biopsy in orbital mass lesions and is especially useful in deep-seated nonpalpable retrobulbar lesions.
Index terms: Biopsies, 22.126, 22.12985 • Orbit, CT, 22.12112 • Orbit, neoplasms, 22.343, 22.362, 22.363, 22.365 • Orbit, US, 22.12981, 22.12985 • Ultrasound (US), guidance, 22.12985
|
Introduction
|
|---|
Surgical biopsy of space-occupying lesions of the orbit, particularly of those retrobulbar in location, presents substantial technical difficulties. In recent years, results of several studies (1–9) have shown the safety and reliability of fine-needle aspiration biopsy in the evaluation of orbital tumors. To be effective, this technique requires precise localization of the target lesion, particularly so in cases of deep retrobulbar lesions. Most of the previous studies have been limited to patients with palpable orbital and eyelid lesions. Review of the literature revealed only a few articles (10-13) describing the role of imaging guidance (computed tomography [CT] or ultrasonography [US]) in performing orbital fine-needle aspiration biopsies. In this study, we evaluated our experience with US-guided fine-needle aspiration biopsies of orbital lesions in 37 patients to determine its safety and effectiveness. Herein, we discuss the technique, advantages, and limitations of the procedure.
|
MATERIALS AND METHODS
|
|---|
During a 4-year period from August 1994 through August 1998, 37 patients (27 male patients, 10 female patients; age range, 4 months to 70 years; average age, 33 years) underwent US-guided fine-needle aspiration biopsy of orbital lesions. CT scans were available in all patients prior to the biopsy procedure. Nineteen lesions were limited to the posterior part of the orbit (ie, behind the posterior aspect of the globe) either in the retrobulbar area (n = 14) or involving the posterior part of an extraocular muscle (n = 5), three lesions were located in the anterior orbit, and 15 lesions were located predominantly posteriorly but also extended anteriorly along the side of the eyeball. The maximum dimensions of the lesions in which biopsy was performed was 1.2–3.9 cm.
All biopsies were performed with use of an RT 3600 or RT 4600 US system (GE Medical Systems, Milwaukee, Wis) and a 5-MHz sector transducer with an 18-mm footprint. The study was performed in accordance with institutional guidelines, and written informed consent was obtained from all patients. The coagulation parameters were checked and corrected, if necessary, to our acceptable range for performance of biopsy procedures: prothrombin time of less than 3 seconds over control (11–13 seconds), partial thromboplastin time of less than 7 seconds over control (30–45 seconds), and a platelet count of more than 100 x 109/L. The patients were placed in the supine position. By using an aseptic technique and a sterile gel as an acoustic coupling agent, the probe tip was placed on the patient's closed eyelid between the globe and orbital rim; the orbit was scanned in axial, sagittal, and various oblique planes; and the site from which the lesion was best visualized was selected. The eyeball was fixed in position by applying firm pressure with the probe itself; also, the patient was asked not to move the eyeball during the procedure. The needle was inserted along the side of the transducer for the paraocular approach (Fig 1a) with use of continuous real-time US visualization. A freehand technique without any attachable needle guides was used in all patients.
View larger version (25K):
[in this window]
[in a new window]
|
Figure 1a. (a) Diagram shows the paraocular approach for lesion visualization and biopsy. The probe (p) is placed on the closed eyelid between the orbital rim and the globe, and the needle is passed into the lesion (m) either from above the probe (a) for sagittal scanning or from its side (b) for transverse scanning. (b) Diagram shows the transocular approach for lesion visualization and biopsy. The probe (p) is placed on the closed eyelid, the ultrasound beam is directed through the globe to localize the lesion (m), and the needle (a) is introduced from the opposite meridian. In a few small lesions, a curved needle (b) was used to avoid injury to the globe.
|
|
View larger version (26K):
[in this window]
[in a new window]
|
Figure 1b. (a) Diagram shows the paraocular approach for lesion visualization and biopsy. The probe (p) is placed on the closed eyelid between the orbital rim and the globe, and the needle is passed into the lesion (m) either from above the probe (a) for sagittal scanning or from its side (b) for transverse scanning. (b) Diagram shows the transocular approach for lesion visualization and biopsy. The probe (p) is placed on the closed eyelid, the ultrasound beam is directed through the globe to localize the lesion (m), and the needle (a) is introduced from the opposite meridian. In a few small lesions, a curved needle (b) was used to avoid injury to the globe.
|
|
More posterior lesions that could not be imaged adequately by means of the paraocular approach or in which the space between the globe and the orbital rim was inadequate for simultaneous placement of the probe and the needle were located by means of the transocular approach, with the probe tip placed on the closed eyelid at a meridian opposite to that of the lesion, and the ultrasound beam was directed through the eyeball. The needle was introduced from the side opposite to that of the probe and advanced along the bony orbital wall (Fig 1b) into the lesion. The path of the needle was monitored by means of continuous US visualization; this approach was used in five patients. In three of these five patients with small deeply situated posterior lesions, a gentle curve (150°–160°) was given to the needle shaft to conform approximately to the inner contour of the orbital wall (Fig 1b); a straight needle would have either missed the lesion or passed dangerously close to the globe.
In the majority of patients, the biopsies were performed without use of local anesthesia or sedation; however, a few anxious patients (n = 4) required intravenous sedation with midazolam hydrochloride (Fulsed; Ranbaxy Labs, Dewas, Maharashtra, India): an initial 1–2-mg bolus was administered and increments of 1–2 mg were added but not to exceed 5 mg/h. In children, the procedures were performed with use of intravenous ketamine hydrochloride (Ketamax 50 [0.5–1.0 mg per kilogram of body weight]; Parenterals, Thol, Gujarat, India) sedation that was administered by an anesthesiologist. We used a 3.5-cm-long 22-gauge needle for anterior lesions (n = 18) and a 9.8-cm-long 22-gauge spinal needle for posterior lesions (n = 19). Once the needle tip was seen within the target lesion, suction was applied, and the needle was moved to and fro within the lesion with use of real-time US visualization to ensure that the needle tip remained within the lesion. The suction was released before the needle was removed. If considerable blood was present in the aspirate, another sample was obtained either without applying suction or by using a thinner (23–25 gauge) needle; this was done in seven patients.
Smears were made, and the slides were stained routinely with May-Grünwald and Giemsa stains. A few slides were fixed in 95% alcohol for hematoxylin-eosin staining. Other special stains (periodic acid-Schiff or Ziehl-Neelsen stains) were used whenever necessary. Syringes with the aspirated material were sent for culture in patients with high clinical suspicion of infection.
|
RESULTS
|
|---|
Twenty-nine (78%) of the 37 fine-needle aspiration biopsies yielded adequate specimens for cytologic analysis (Table); in all these patients, cytologic findings along with the clinical presentation and radiologic findings provided sufficient information to treat the patients without further tissue analysis. Ten fine-needle aspiration biopsies revealed a malignant process: metastatic carcinoma (n = 2); non-Hodgkin lymphoma (n = 2); adenoid cystic carcinoma of the lacrimal gland (n = 2) (Fig 2); embryonal rhabdomyosarcoma (n = 2); malignant round cell tumor, possibly primitive neuroectodermal tumor (n = 1); and alveolar soft-tissue sarcoma (n = 1).
View larger version (180K):
[in this window]
[in a new window]
|
Figure 2a. Adenoid cystic carcinoma of the lacrimal gland in a 56-year-old man. (a) Contrast material-enhanced axial CT scan shows a homogeneous retrobulbar mass (m) extending into the anterior orbit along the lateral aspect of the globe. (b) Transverse oblique sonogram shows the needle (arrow) within the mass, which is displacing the globe (g) medially.
|
|
View larger version (179K):
[in this window]
[in a new window]
|
Figure 2b. Adenoid cystic carcinoma of the lacrimal gland in a 56-year-old man. (a) Contrast material-enhanced axial CT scan shows a homogeneous retrobulbar mass (m) extending into the anterior orbit along the lateral aspect of the globe. (b) Transverse oblique sonogram shows the needle (arrow) within the mass, which is displacing the globe (g) medially.
|
|
Benign lesions were diagnosed at fine-needle aspiration biopsy in 19 patients: meningioma (n = 3) (Fig 3), pleomorphic adenoma of the lacrimal gland (n = 3), pseudotumor (n = 2) (Fig 4), lymphangioma (n = 1), hemangioma (n = 1), epidermal inclusion cyst (n = 1), sinus histiocytosis (n = 1), and inflammatory lesion (n = 7). In one of these seven patients, aspirate yielded frank pus; culture, however, was negative, and the patient improved with administration of broad spectrum antibiotics. Fine-needle aspiration biopsy in three other patients revealed acute inflammatory reactions; one of these patients had associated disease in adjacent sinuses and a positive blood culture for Aspergillus fumigatus and improved with antifungal treatment, whereas the other two patients improved with use of antibiotics. The other three patients had a nonspecific chronic inflammatory reaction.
View larger version (150K):
[in this window]
[in a new window]
|
Figure 3a. Orbital meningioma in a 24-year-old woman. (a) Contrast-enhanced axial CT scan shows a well-defined homogeneous mass (m) surrounding the right optic nerve (arrow). (b) Transverse sonogram shows the needle (arrows) within the hyperechoic mass (m) visualized through the globe by means of the transocular approach.
|
|
View larger version (165K):
[in this window]
[in a new window]
|
Figure 3b. Orbital meningioma in a 24-year-old woman. (a) Contrast-enhanced axial CT scan shows a well-defined homogeneous mass (m) surrounding the right optic nerve (arrow). (b) Transverse sonogram shows the needle (arrows) within the hyperechoic mass (m) visualized through the globe by means of the transocular approach.
|
|
View larger version (168K):
[in this window]
[in a new window]
|
Figure 4a. Inflammatory pseudotumor in a 55-year-old woman. (a) Contrast-enhanced coronal CT scan shows an enlarged right superior rectus muscle (m). (b) Sagittal sonogram shows the needle (arrow) passing above the globe (g) toward the enlarged hypoechoic superior rectus muscle (m).
|
|
View larger version (166K):
[in this window]
[in a new window]
|
Figure 4b. Inflammatory pseudotumor in a 55-year-old woman. (a) Contrast-enhanced coronal CT scan shows an enlarged right superior rectus muscle (m). (b) Sagittal sonogram shows the needle (arrow) passing above the globe (g) toward the enlarged hypoechoic superior rectus muscle (m).
|
|
In eight (22%) of the 37 patients, we obtained either no tissue or simply blood and/or nonspecific fibrous tissue, and cytologic diagnosis was not possible. Chronic pseudotumors were subsequently diagnosed in three of these patients on the basis of clinical and radiologic findings and response to steroid therapy or radiation therapy. In one patient, histopathologic examination revealed fibrocollagenous tissue separated by fibrous septae. Two patients were lost to further follow-up, and in two other patients the final diagnosis remains to be established as they are receiving empiric steroid therapy on the basis of the presumptive diagnoses of pseudotumors.
All patients tolerated the procedure well. Two patients, one with pleomorphic adenoma of the lacrimal gland and the other with orbital pseudotumor, developed mild diffuse orbital hemorrhage following biopsy. Both patients were admitted to the hospital, and clinical and CT follow-up showed resorption of the hemorrhage; there were no permanent sequelae. One patient developed a localized lid hematoma at the puncture site, and it also subsided spontaneously. One patient with malignant embryonal rhabdomyosarcoma with extension into the nasal cavity developed bleeding through the nose immediately after the procedure; it was successfully managed with nasal packing. No other complications occurred.
|
DISCUSSION
|
|---|
Prior to the introduction of US and CT, aspiration biopsies of orbital lesions, guided with ophthalmologic examination and conventional radiologic techniques, were difficult to perform and were limited to palpable orbital and eyelid tumors (4,5,8). Without accurate tumor visualization, there was no way to ensure that the sample was being obtained from the lesion and not from the adjacent orbital soft tissue.
To our knowledge, articles (10–13) describing the role of imaging guidance for orbital biopsies are scanty. Czerniak et al (10) successfully performed needle biopsies of 13 orbital lesions (nine of these lesions were purely retrobulbar in location) with use of CT guidance. However, CT precludes real-time visualization during needle placement and biopsy, which often results in multiple attempts before the needle can be positioned accurately. Multiple axial and coronal sections are often required to visualize the needle tip, and lack of visualization during biopsy in small lesions may result in sampling of adjacent normal tissue.
US recently has been gaining recognition as a useful and versatile guidance technique. It has many advantages over CT guidance, including real-time imaging, decreased procedure time and cost, portability, and lack of ionizing radiation. Also, with advances in technology, sector probes with small footprints are available, and these allow adequate visualization of the orbital structures through a closed eyelid. However, review of the literature revealed only one article (13) that described the use of US guidance for performing orbital biopsies in three patients with retrobulbar tumors.
The results of our study suggest that US is efficacious as a guidance technique for biopsy of most orbital lesions; we were able to clearly visualize and effectively perform biopsy in lesions that were located in various parts of the orbit and that were as small as 1.5 cm. During this study, only three patients were refused biopsy because of inadequate lesion visualization or lack of availability of a safe path for biopsy; all three patients had lesions smaller than 1 cm in the posterior part of the orbit.
We were able to perform biopsy in the majority of the lesions in the present study by means of the paraocular approach; this was useful both for anterior orbital lesions and for deep retrobulbar masses. The presence of large orbital masses with associated proptosis usually resulted in appreciable widening of the paraocular space (Fig 2), thus facilitating the biopsy procedure. Even in patients with smaller lesions (Fig 4), biopsy could be performed easily by means of the paraocular approach simply by pushing the eyeball with the probe itself away from the needle path.
However, not all lesions permit paraocular biopsy; in some patients with posterior orbital lesions, the space between the globe and the orbital rim proved inadequate for simultaneous probe placement and needle puncture. This problem was solved by using the transocular approach (Fig 3) for lesion visualization.
Giving a gentle curve to the needle to conform to the inner orbital contour was useful in three patients with small deeply situated posterior orbital lesions; this results in the needle passing along the orbital wall, thus decreasing the risk of injury to vital structures. To our knowledge, this maneuver has not previously been reported for orbital biopsies, although Jasinki (14) successfully used a curved needle to perform biopsy in an anterior superior mediastinal mass located behind the sternum by means of the suprasternal approach.
In keeping with most of the previous reports (3,5,8), we also did not use local anesthesia in the majority of the adult patients, as it leads to distortion and swelling of orbital tissues and also requires an additional needle puncture. In our experience, injection of local anesthetic sometimes changes the echo characteristics of superficial tissues, probably secondary to microscopic air bubbles being introduced, and hampers subsequent US visualization. As pointed out by Kennerdell et al (5), because of the loose anatomic composition of the orbital tissues, needle insertion causes only brief discomfort; all of the patients in our study tolerated the procedure well.
Kennerdell et al (5), in their study of 156 orbital biopsies performed without imaging guidance, emphasized that the needle length should not exceed 3.75 cm for an adult orbit and 2.5 cm for infants and patients with smaller orbits to avoid puncture of the superior orbital fissure. However, we encountered no problems with the use of needles 9.8 cm long; this is because the entire biopsy procedure, including the sampling, was performed with use of continuous US monitoring, which ensures that the needle tip remains within the lesion at all times. However, since a possibility of inadvertent puncture of the superior orbital fissure exists, it would be safer to use a shorter needle for orbital biopsies, and there is no added advantage in using a longer needle.
Although several authors (15,16) used biopsy guides for US-guided biopsies, we believe that the freehand technique offers considerable flexibility to the radiologist and allows for fine adjustments while maneuvering the needle. This is especially useful in orbital biopsies in which the window available for visualization and needle insertion is often small and precludes the use of biopsy guides. Also, with biopsy guides, it is not always possible to follow the prescribed path owing to intervening vital structures (eg, the globe); the transocular approach that we used in five patients exemplifies this point well and would not have been possible with the biopsy guides.
We found preprocedural CT useful as a guide to performing US-guided biopsies. CT, because of its large field of view, allows adequate visualization of the lesion, its extent, and its relationship with adjacent structures. This reduces the time required for US localization and also helps in selecting the best approach for biopsy.
Since the application by Schyberg (17) of fine-needle aspiration biopsy to the diagnosis of orbital tumors, several studies (1–9) have reported the use of this technique in the evaluation of deep and superficial orbital tumors but with varying success. Krohel et al (18), in their series of 34 patients, found orbital fine-needle aspiration biopsy to have an accuracy of only 50%. However, other authors (1–9) have reported much higher success rates ranging from 83% to 92%; we achieved a somewhat lower success rate of 78%. The wide variety of disease processes encountered in the orbit and the different selection criteria used by different authors probably explains this wide variation in the results of orbital fine-needle aspiration biopsy (1–9,18). As pointed out by Kennerdell et al (4,5) and other authors (1–3,6–9,18), the best use of orbital fine-needle aspiration biopsy is in establishing a diagnosis of malignant (primary or secondary) nonresectable neoplasms, thus eliminating the need for surgical intervention; cytologic diagnosis of malignancy obviated the need for a diagnostic orbitotomy in 10 patients in our study. Fine-needle aspiration biopsy is also useful in establishing a correct diagnosis in patients suspected of having acute orbital inflammation, as was illustrated in four of the patients in our study.
The role of fine-needle aspiration biopsy in the evaluation of patients suspected of having orbital pseudotumors remains controversial (3,5,8,18,19). It is generally held that it is often difficult to separate patients with lymphoid hyperplasia in the pseudotumor group from those with low-grade malignant lymphomas; three of the fine-needle aspiration biopsies with negative results in our study were from chronic pseudotumors. Recent studies (9,19), however, suggest that use of cell surface markers, immunocytochemistry, and electron microscopy can substantially improve the accuracy of fine-needle aspiration biopsy in the diagnosis of pseudotumors. In keeping with the previous reports (4,5,8,9,18), our results also show that negative fine-needle aspiration biopsy results mostly come from fibrous lesions with sparse or exceedingly cohesive cell population or inadequate samples from lymphoid lesions; of the eight patients with negative fine-needle aspiration biopsy results, three had chronic pseudotumors and one had fibrocollagenous tissue at histopathologic examination. Hence, negative results should not preclude the use of other methods to confirm a clinical suspicion.
The complications of orbital fine-needle aspiration biopsy are of great concern, and extreme care should be taken to avoid damage to the globe or the optic nerve (3,5,8,9,20). Real-time US monitoring allows simultaneous visualization of the globe and the needle as it is advanced toward the lesion; this is very important to prevent inadvertent grazing or puncture of the globe. Although retrobulbar hemorrhage is an expected complication, it is uncommon and usually does not cause permanent ocular disability (5,20). We encountered this complication in only two of the patients in our study; in both, it resorbed spontaneously without any sequelae.
To conclude, US-guided fine-needle aspiration biopsy of orbital lesions appears feasible and safe, is especially useful for deeply situated retrobulbar lesions, and may yield valuable diagnostic information, thus obviating major surgical procedures in a substantial number of patients.
|
Footnotes
|
|---|
Author contributions: Guarantor of integrity of entire study, Sanjay G.; study concepts, S.S., Sanjay G.; study design, Sanjay G., D.T.; definition of intellectual content, Sanjay G., B.S.; literature research, Sanjay G., M.G.; clinical studies, Subhash G., N.K., U.S., A.R.; data acquisition, D.T., Sanjay G.; data analysis, Sanjay G., Subhash G.; manuscript preparation, R.B., Sanjay G., Subhash G.; manuscript editing, Sanjay G.; manuscript review, S.S., Sanjay G., Subhash G.
|
References
|
|---|
-
Arora R, Rewari R, Betheria SM. Fine needle aspiration cytology of eyelid tumours. Acta Cytol 1990; 34:227-239.[Medline]
-
Das DK, Das J, Bhatt NC, Chachra KL, Natarajan R. Orbital lesions: diagnosis by fine needle aspiration cytology. Acta Cytol 1994; 38:158-164.[Medline]
-
Dey P, Radhika S, Rahwanshi A, Ray R, Nijhawan R, Das A. Fine needle aspiration biopsy of orbital and eyelid lesions. Acta Cytol 1993; 37:903-907.[Medline]
-
Kennerdell JS, Dekker A, Johnson BL, Dubois PJ. Fine needle aspiration biopsy: its use in orbital tumours. Arch Ophthalmol 1979; 97:1315-1317.[Abstract]
-
Kennerdell JS, Slamovits TL, Dekker A, Johnson BL. Orbital fine needle aspiration biopsy. Am J Ophthalmol 1985; 99:547-551.[Medline]
-
Tijil JWM, Koornneef L. Fine needle aspiration biopsy in orbital tumours. Br J Ophthalmol 1991; 75:491-492.[Abstract/Free Full Text]
-
Westman-Naeser S, Naeser P. Tumours of the orbit diagnosed by fine needle biopsy. Acta Ophthalmol 1978; 56:969-976.[Medline]
-
Zajdela A, Vielh P, Schlienger P, Haye C. Fine needle cytology of 292 palpable orbital and eyelid tumours. Am J Clin Pathol 1990; 93:100-104.[Medline]
-
Zeppa P, Tranfa F, Errico ME, et al. Fine needle aspiration (FNA) biopsy of orbital masses: a critical review of 51 cases. Cytopathology 1997; 8:366-372.[Medline]
-
Czerniak B, Woyke S, Daniel B, Krzysztolic Z, Koss LG. Diagnosis of orbital tumours by aspiration biopsy guided by computerised tomography. Cancer 1984; 54:2385-2389.[Medline]
-
Dubois PJ, Kennerdell JS, Rosenbaum AE, Dekker A, Johnson BR, Swink CA. Computed tomography localization for fine needle aspiration biopsy of orbital tumors. Radiology 1979; 131:149-152.[Abstract]
-
Kennerdell JS, Dubois PJ, Dekker A, Johnson BL. CT guided fine needle aspiration biopsy of orbital optic nerve tumours. Ophthalmology 1980; 87:491-496.[Medline]
-
Spoor TC, Kennerdell JS, Dekker A, Johnson BL, Rehkopf P. Orbital fine needle aspiration biopsy with B-scan guidance. Am J Ophthalmol 1980; 89:274-277.[Medline]
-
Jasinki RW. CT guided suprasternal approach for mediastinal biopsy. J Comput Assist Tomogr 1992; 16:669-670.[Medline]
-
Momel DS, Dodd GD, III, Esola CC. Efficacy of sonography as a guidance technique for biopsy of abdominal, pelvic and retroperitoneal lymph nodes. AJR 1996; 167:957-962.[Abstract/Free Full Text]
-
Fisher AJ, Paulson EK, Sheafor DH, Simmon CM, Nelson RC. Small lymph nodes of the abdomen, pelvis, and retroperitoneum: usefulness of sonographically guided biopsy. Radiology 1997; 205:185-190.[Abstract/Free Full Text]
-
Schyberg E. Fine needle aspiration biopsy of orbital tumours. Acta Ophthalmol 1975; 125(suppl):11-18.
-
Krohel GB, Tobin DR, Chavis RM. Inaccuracy of fine needle aspiration biopsy. Ophthalmology 1985; 92:666-670.[Medline]
-
Char DH, Miller T. Orbital pseudotumour: fine needle aspiration biopsy and response to therapy. Ophthalmology 1993; 100:1702-1710.[Medline]
-
Liu D. Complications of fine needle aspiration biopsy of the orbit. Ophthalmology 1985; 92:1768-1771.[Medline]
TOP
Abstract
Introduction
