Purpose
To describe and analyze results from the fine-needle aspiration biopsy (FNAB) technique, used as a diagnostic tool, in patients with orbital lesions.
Design
Retrospective case series.
Methods
setting : Institutional (Karolinska Hospital, Stockholm). study population : 207 patients with 210 orbital lesions. intervention : 225 FNABs of the orbit. main outcome measures : Successful diagnosis from FNAB.
Results
Of the 210 orbital lesions evaluated with FNAB, a successful cytologic diagnosis was achieved in 176 (84%). In more than half of the orbital lesions (54%), the FNAB diagnosis in addition to imaging appearance, clinical appearance, and clinical history provided sufficient information for treatment, and the patient did not require an incisional or excisional biopsy. Ninety-seven patients underwent additional excisional or incisional biopsy; FNAB diagnoses and the histopathologic diagnoses corresponded in 87% of the cases with an intraclass correlation coefficient of 0.84 (95% confidence interval [CI] 0.76–0.90). The difference was statistically significant between the ability to make a successful cytologic FNAB diagnosis in palpable lesions vs nonpalpable lesions (successful diagnosis in 90% [CI = 85%–95%] vs 75% [CI = 66%–84%]; P < .01). Neither the orbital quadrant location, nor the radiologic appearance (diffuse vs encapsulated), nor size of the lesion affected the success of FNAB diagnoses (all P > .7). There was a complication in 6 cases (3%). All complications were temporary and none led to permanent damage.
Conclusions
FNAB proved effective and exceedingly safe. With the current healthcare climate of minimally invasive surgery and cost control, FNAB should be considered as a valid alternative to open surgery in the evaluation and management of orbital lesions.
Orbital lesions are rare and can be a diagnostic challenge for the ophthalmologist. In some situations, clinical in combination with radiologic findings are sufficient to arrive at a correct diagnosis. In most patients, histologic evaluation is needed. Morphology can be acquired by incisional/excisional biopsy and by fine-needle aspiration biopsy (FNAB). An incisional biopsy in the orbital area carries inherent risks and complications of surgery, but provides more material for a higher diagnostic yield. While FNAB represents a less invasive alternative, it gives less material to examine and is therefore dependent on an experienced cytopathologist. If FNAB does not provide adequate tissue for a diagnosis, a secondary incisional biopsy may still be required.
Since the early 1960s FNAB has been used in the diagnosis of tumors, and since the 1970s it has been used in the diagnosis of orbital tumors in Stockholm, Sweden. With time, the technique and procedure have been refined, and currently FNAB cytology is the mainstay of morphologic diagnosis of orbital tumors in Stockholm. In some instances FNAB is replaced by incisional biopsy as the primary procedure. Earlier studies report a diagnostic accuracy of 75%–88% for orbital FNAB. More recent data show a diagnostic accuracy up to 99% (with the use of immunocytochemistry and other techniques). This high diagnostic accuracy makes FNAB an attractive alternative to an orbitotomy.
In many countries, surgeons prefer incisional biopsy or direct excision over FNAB for lesions within the orbit. This cautious attitude toward orbital FNAB may be based on studies from the 1980s where the results from FNAB were reported to be poor and potentially risky. Since the 1980s, there have been advances in sampling techniques, immunocytochemistry, the skills of the cytopathologist, and the knowledge of which lesions are suited for FNAB.
The purpose of this study is to describe and analyze the results from the FNAB technique, used as a diagnostic tool, in all patients who underwent FNAB in the orbital area between the years 2005 and 2013 at the Division of Clinical Cytology, Karolinska Hospital, Solna.
Methods
This is a retrospective chart review of all patients who underwent FNAB in the orbital area between January 1, 2005 and December 31, 2013 at the Division of Clinical Cytology, Karolinska Hospital, Solna. Institutional review board approval was applied for and granted by the Regional Ethical Review Board in Stockholm. The described research adhered to the tenets of the Declaration of Helsinki. Patients with orbital lesions were referred from the Oculoplastic and Orbital Service and the Oncology Service at St Erik Eye Hospital in Stockholm to the Division of Clinical Cytology, Karolinska Hospital, Solna. The criteria for selection for FNAB or excisional/incisional biopsy was the following: if a lesion was thought to be inflammatory or lymphoproliferative and anterior enough to obtain a representative sample with the patient awake, then an incisional biopsy was usually performed; excisional biopsies were performed if the lesion radiographically was unambiguously a lesion best treated with total excision (eg, cavernous hemangioma). If neither of those 2 situations was present, then the patient was referred to the cytopathologist for FNAB.
The FNAB technique evaluated was used during the study period 2005–2013. Both superficial and deeper lesions were biopsied in a strictly outpatient basis. A 27 or 25 gauge (0.4–0.5 mm) needle was used for both palpable and nonpalpable lesions. Eight different cytopathologists carried out the FNABs, with confirmation of diagnosis by the 2 senior cytopathologists. In more posterior tumors the biopsy was performed under CT guidance by 3 different radiologists. No local anesthesia was required, but with children, FNAB was performed under general anesthesia. When the FNAB was performed by a radiologist, the cytopathologist immediately performed staining to determine if the material was adequate; if not, a new attempt was performed. Rarely more than 2 FNAB attempts were performed in the same session. If the patient was taking any type of anticoagulant medication, this was stopped for the appropriate amount of time prior to the procedure (depending on the type of medication and whether termination was approved by the patient’s prescribing physician). A more detailed description of the FNAB technique used and staining methods has been previously reported by Seregard and Tani.
An FNAB was considered successful if it led to a cytologic diagnosis. If the result was “inconclusive” it was regarded as unsuccessful. The FNABs were pooled into 5 groups of diagnoses: inconclusive, malformation/cyst, normal findings, tumor, and inflammation. When the FNAB showed “benign cytology” the label “normal findings” was applied in order to be able to compare the cytologic and incisional/excisional biopsies with each other. To determine if the location of the lesion affected the result of the FNAB, the patients were divided into 4 groups depending on in which quadrant the lesion resided: superotemporal, superonasal, inferotemporal, or inferonasal. The study group was also divided depending on how posterior the lesion was positioned: anterior (anterior to the posterior part of the globe), posterior intraconal, or posterior extraconal. The lesions were also divided into palpable or nonpalpable and into focal/encapsulated or diffuse, depending on imaging appearance of the mass, to see if this affected the success of the procedure. The size of the lesion was measured in millimeters (mm) in vertical, horizontal, and anterior-posterior dimensions using computed tomography (CT) and/or magnetic resonance imaging (MRI) scans.
If a second FNAB was performed in the same patient, this was referred to as a re-FNAB if made within 6 months from the first one. A re-FNAB was performed if the first one was inconclusive or if the cytologic result did not correlate with the clinical suspicion. If the first FNAB and the re-FNAB diagnosis did not correspond, the re-FNAB diagnosis was chosen as the “FNAB diagnosis.”
Data were analyzed for the group as a whole and for subgroups. The proportion (percent) was calculated for the group as a whole and for each subgroup with 95% confidence intervals. The potential differences in the proportion of successful FNABs between subgroups were calculated with the χ 2 test along with the Yates correction. The potential influence of the size of the lesion was analyzed using the t test. P values ≤.05 were considered statistically significant. When comparing the cytologic diagnosis from the successful FNABs with the histologic diagnosis from incisional biopsy or resection of the lesion, the intraclass correlation coefficient was calculated with 95% confidence intervals. All statistical analyses were performed with MedCalc (version 15.2.2; MedCalc Software, Ostend, Belgium).
Results
Between 2005 and 2013, 225 FNAB samples were recorded in 210 orbits at the Division of Clinical Cytology, Karolinska Hospital, Solna. The material was collected from 207 orbits (87 [42%] men and 120 [58%] women). The mean age was 54 years (median age 59 years; range, 2 months to 95 years). In 117 individuals (56%) the age ranged between 20 and 65 years (ie, the productive working years in Sweden). Three of the patients had bilateral biopsies; thus 210 orbits were investigated. Fifteen patients had a re-FNAB.
Of the 225 FNABs performed in the 210 orbits, there was a complication documented in 6 cases (3%). The complication noted in 5 of the 6 cases was a small hematoma in the aspiration area or eyelid, which was observed for 1–2 hours after the procedure. In 1 case there was a subconjunctival hemorrhage. This patient was observed overnight without any sequelae. All complications were temporary and none led to permanent damage.
FNAB resulted in a cytologic diagnosis in 176 of 210 orbits (84%) ( Table 1 ). The main reason for an “inconclusive” FNAB was that the aspiration only gave “sparse material” with limited cellularity; this occurred in 18 out of 34 samples (53%). Tumor subgroups are listed in Table 1 .
| Fine-Needle Aspiration Biopsy Diagnosis | Number | (%) |
|---|---|---|
| Inconclusive | 34 | (16%) |
| Malformation/cysts | 18 | (9%) |
| Normal findings | 25 | (12%) |
| Tumor | 93 | (44%) |
| Lymphoproliferative tumor | 48 | (23%) |
| Metastasis | 16 | (8%) |
| Lacrimal gland tumor | 13 | (6%) |
| Mesenchymal tumor | 8 | (4%) |
| Secondary tumor | 5 | (2%) |
| Neurogenic/meningeal tumor | 3 | (1%) |
| Inflammation | 40 | (19%) |
| Total | 210 | (100%) |
Regarding lesion location, there were 42 posterior lesions that included 26 intraconal and 16 extraconal lesions. More than half of the lesions (115 of 210) were located superotemporal and anterior. The percentage of successful FNABs in the 4 quadrants ranged from 78% (CI = 65%–91%) for superonasal to 86% (CI = 72%–100%) for inferotemporal lesions ( Table 2 ). The likelihood of making a cytologic diagnosis was not dependent on which quadrant the lesion was positioned in ( P > .70). Both palpable (vs nonpalpable) and anterior (vs posterior) lesions showed a statistically significant difference in the percentage of successful cytologic FNAB diagnoses ( Table 2 ). Of the 89 “nonpalpable” lesions, 14 were performed without CT guidance (5 of these 14 were “inconclusive” and none of them had a re-FNAB; 2 of these 5 “inconclusive” cases had a subsequent incisional biopsy, which in 1 case showed orbital inflammation and in the other the incisional biopsy was also inconclusive).
| FNAB Investigated | FNAB Not Successful | FNAB Successful | % Successful | P Value | |
|---|---|---|---|---|---|
| Quadrant | >.70 | ||||
| Inferonasal | 23 | 4 | 19 | 83% (CI = 68%–98%) | |
| Inferotemporal | 22 | 3 | 19 | 86% (CI = 72%–100%) | |
| Superonasal | 37 | 8 | 29 | 78% (CI = 65%–91%) | |
| Superotemporal | 128 | 19 | 109 | 85% (CI = 79%–91%) | |
| Anterior | 168 | 22 | 146 | 87% (CI = 82%–92%) | <.02 |
| Posterior | 42 | 12 | 30 | 71% (CI = 57%–85%) | |
| Palpable | 121 | 12 | 109 | 90% (CI = 85%–95%) | <.01 |
| Nonpalpable | 89 | 22 | 67 | 75% (CI = 66%–84%) | |
| Focal | 77 | 13 | 64 | 83% (CI = 75%–91%) | >.80 |
| Diffuse | 133 | 21 | 112 | 84% (CI = 78%–90%) | |
| CT-guided | 92 | 19 | 73 | 79% (CI = 71%–87%) | >.10 |
| Not CT-guided | 118 | 15 | 103 | 87% (CI = 81%–93%) |
Neither imaging appearance (diffuse vs focal) nor CT guidance (CT guided vs not CT guided) showed a statistically significant difference in the percentage of successful cytologic FNAB diagnoses ( Table 2 ).
In 167 of the lesions it was possible to measure the size of the lesion on previously performed CT or MRI scans. The vertical size was 14 mm (standard deviation [SD] 4 mm) in the inconclusive lesions and 16 mm (SD 6 mm) in the successful ones. The horizontal size was 12 mm (SD 5 mm) in the inconclusive lesions compared to 13 mm (SD 6 mm) in the successful lesions. Finally, the anterior-posterior size was 18 mm (SD 6 mm) in the inconclusive FNABs and 20 mm (SD 8 mm) in the successful FNABs. The difference in the vertical-, horizontal-, and anterior-posterior mean sizes of the lesions did not statistically differ between inconclusive and successful cytologic FNAB diagnoses (all P > .7).
In 56 patients, additional material was needed for diagnosis or for classification, and an incisional biopsy was performed. An additional 41 patients underwent a complete surgical excision or surgical debulking of the lesion after FNAB diagnosis. Thus, there was a histopathologic diagnosis from subsequent open biopsy on 97 patients, which could be compared with the FNAB cytologic diagnosis. The correspondence between the histopathologic biopsy diagnosis and the cytologic FNAB diagnosis is shown in Table 3 . Of the 97 patients there were 21 with an “inconclusive” FNAB diagnosis, which were thereby regarded as unsuccessful. Of the remaining 76 patients with a successful cytologic diagnosis, the cytologic and histopathologic diagnosis corresponded in 66 patients (87%) ( Table 3 ). The intraclass correlation coefficient for the corresponding diagnosis was 0.84 (CI = 0.76–0.90). The most common FNAB diagnosis that did not correspond to the histopathologic diagnosis was “normal findings”—that is, benign cells (6 out of 8). In 4 of these 6 an incisional biopsy showed “inflammation.”
