Advertisement

Indications for Fine Needle Aspiration Biopsy of Posterior Segment Intraocular Tumors

  • Zélia Maria Corrêa
    Correspondence
    Inquiries to Zélia Maria Corrêa, Ocular Oncology Service, Retina Division, Department of Ophthalmology, Wilmer Eye Institute, Johns Hopkins Medicine, 600 North Wolfe St, Maumenee 711, Baltimore, MD 21287 USA
    Affiliations
    Ocular Oncology Service, Department of Ophthalmology, University of Cincinnati College of Medicine, Cincinnati, Ohio, USA

    Wilmer Eye Institute, Johns Hopkins School of Medicine, Baltimore, Maryland, USA
    Search for articles by this author
  • James Jay Augsburger
    Affiliations
    Ocular Oncology Service, Department of Ophthalmology, University of Cincinnati College of Medicine, Cincinnati, Ohio, USA
    Search for articles by this author

      Purpose

      To define 4 indication categories of fine needle aspiration biopsies (FNABs) of solid intraocular tumors and describe the differences among the patients, tumors, and results of biopsy in these 4 indication categories.

      Methods

      Retrospective descriptive analysis of 880 FNABs of a solid intraocular tumor of the posterior ocular segment performed by in the authors' ocular oncology practice during the period July 1980 through July 2014.

      Results

      FNABs were performed as a separate procedure in 372 cases (42.3%), at plaque implantation in 279 (31.7%), post-enucleation in 225 (25.6%), and post-resection/pre-laser in 4 (0.4%). FNABs were categorized as diagnostic in 292 (33.2%), confirmatory in 121 (13.8%), investigational in 187 (21.3%), and prognostic in 280 (31.8%). Prior to the biopsy, all patients who underwent diagnostic FNAB had a tumor of uncertain pathologic type, while all patients who underwent confirmatory FNAB had a clinically diagnosed malignant intraocular neoplasm. In contrast, all patients who underwent a prognostic FNAB had an unequivocal primary posterior uveal melanoma clinically, while patients who underwent an investigational biopsy all had an unequivocal or probable malignant intraocular tumor of a specific type. Most diagnostic FNABs were performed transvitreously (n = 255, 87.3%) compared to prognostic FNABs (n = 143, 51.1%) (P < .001). Overall, 733 FNABs (83.3%) yielded a sufficient specimen for cytologic diagnosis. Diagnostic and confirmatory FNABs were the most likely to yield an insufficient aspirate (28.4% and 20.7%, respectively), while investigational and prognostic FNABs were most likely to yield sufficient specimens (94.7% and 90.0%, respectively) (P < .001). Mean post-FNAB follow-up of the entire group was 62.7 months (standard deviation [SD] = 59.8, 95% confidence interval [CI] = 58.7–66.6). Mean posttreatment follow-up or total follow-up of untreated patients was 53.7 months (SD = 58.7, 95% CI = 49.8–57.6).

      Conclusions

      Because of the substantial differences shown among FNAB cases performed for different indications, it seems appropriate to report results of FNAB or other biopsy methods according to the category (indication) of the biopsy. NOTE: Publication of this article is sponsored by the American Ophthalmological Society.
      A biopsy may be defined as a surgical procedure intended to obtain a representative (and sufficient) specimen of cells or tissue for pathologic (or prognostic) assessment.
      • Frable W.J.
      Fine-needle aspiration biopsy: a review.
      There are at least 4 different general types of biopsy that can be performed for a clinically identified tumor. An excisional biopsy is a surgical procedure in which the surgeon attempts to remove the intact tumor entirely. Examples in ophthalmic surgery include enucleation and transscleral en bloc tumor resection. In contrast, an incisional biopsy is a surgical procedure in which the surgeon cuts away a portion of the tumor using instruments such as scissors, a scalpel, a curette, a suction-cutting tissue extractor, or a large-caliber cutting needle (core needle biopsy). Examples in ophthalmic surgery include transscleral incisional biopsy with scissors, a transvitreal-transretinal vitrector biopsy, and transvitreal endoresection.
      • Peyman G.A.
      • Fishman G.A.
      • Sanders D.R.
      • Apple D.J.
      • Vlchek J.K.
      Biopsy of human scleral-chorioretinal tissue.
      An aspirational biopsy is a surgical procedure in which a small-caliber hollow lumen needle and suction force without cutting are employed to aspirate a small specimen of the tumor.
      • Foulds W.S.
      The uses and limitations of intraocular biopsy.
      This type of biopsy is generally referred to as fine needle aspiration biopsy (FNAB). Finally, an exfoliative (desquamatory) biopsy is a surgical procedure in which superficial or exposed tumor cells are scraped off the tumor's surface or lifted off the tumor using an impression method. This type of biopsy is usually employed for superficial tumors of the skin or mucous membranes and is not employed currently for intraocular tumors. The particular type of biopsy employed in a given patient depends on factors such as the anatomic site of the tumor, the diagnosis or differential diagnosis of the tumor, the planned analysis of the obtained specimen, and the anticipated benefits and risks of the procedure.
      • Midena E.
      • Segato T.
      • Piermarocchi S.
      • Boccato P.
      Fine needle aspiration biopsy in ophthalmology.

       Historical Perspective

      From a historical perspective, the beginning of the 20th century was marked by significant indifference towards aspiration biopsy. Despite a series of articles published in the 1930s by Martin and Ellis
      • Martin H.E.
      • Ellis E.B.
      Biopsy by needle puncture and aspiration.
      and Stewart
      • Stewart F.W.
      The diagnosis of tumors by aspiration.
      describing procedures in which they used needles to aspirate tumor tissue from several bodily organs for pathologic analysis, aspiration biopsy did not become a popular procedure in the United States.
      • Frable W.J.
      Fine-needle aspiration biopsy: a review.
      The interest of the medical community in aspiration biopsy only returned in the 1960s owing to extensive work done by cytologists including Franzen and Zajicek
      • Franzen S.
      • Zajicek J.
      Aspiration biopsy in diagnosis of palpable lesions of the breast. Critical review of 3479 consecutive biopsies.
      and Söderström
      • Söderström N.
      Fine Needle Aspiration Biopsy.
      from the Karolinska Institute in Sweden. In ophthalmology, publications on the use of FNAB for solid intraocular tumors started in the late 1970s.
      • Jakobiec F.A.
      • Chattock A.
      The role of cytology and needle biopsies in the diagnosis of ophthalmic tumors and simulating conditions.
      • Jakobiec F.A.
      • Coleman D.J.
      • Chattock A.
      • Smith M.
      Ultrasonically guided needle biopsy and cytologic diagnosis of solid intraocular tumors.
      FNAB evolved technically thanks to the work of Augsburger and associates
      • Augsburger J.J.
      • Shields J.A.
      Fine needle aspiration biopsy of solid intraocular tumors.
      • Augsburger J.J.
      Fine·needle aspiration biopsy in the diagnosis of suspected intraocular cancer.
      • Augsburger J.J.
      • Shields J.A.
      Fine needle aspiration biopsy of solid intraocular tumors: indications, instrumentation and techniques.
      but it did not become more widely used until the recent adoption of prognostic testing for posterior uveal melanoma.
      • Onken M.D.
      • Worley L.A.
      • Char D.H.
      • et al.
      Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma.
      • Schopper V.J.
      • Correa Z.M.
      Clinical application of genetic testing for posterior uveal melanoma.
      At the present time, FNAB is the most frequently employed biopsy method for intraocular tumors in ocular oncology centers today.
      • Eide N.
      • Walaas L.
      Fine-needle aspiration biopsy and other biopsies in suspected intraocular malignant disease: a review.
      • Augsburger J.J.
      Fine needle aspiration biopsy of suspected metastatic cancers to the posterior uvea.
      This method is less invasive than incisional and excisional biopsy methods and seems to have a good safety profile in multiple reported series.
      • Augsburger J.J.
      • Correa Z.M.
      • Schneider S.
      • et al.
      Diagnostic transvitreal fine-needle aspiration biopsy of small melanocytic choroidal tumors in nevus versus melanoma category.
      • Eide N.
      • Syrdalen P.
      • Walaas L.
      • Hagmar B.
      Fine needle aspiration biopsy in selecting treatment for inconclusive intraocular disease.
      However, this method is also associated with a number of pertinent limitations, including the small number of tumor cells obtained from many tumors
      • Eide N.
      • Walaas L.
      Fine-needle aspiration biopsy and other biopsies in suspected intraocular malignant disease: a review.
      • Correa Z.M.
      • Augsburger J.J.
      Sufficiency of FNAB aspirates of posterior uveal melanoma for cytologic versus GEP classification in 159 patients, and relative prognostic significance of these classifications.
      • Augsburger J.J.
      • Correa Z.M.
      • Trichopoulos N.
      Prognostic implications of cytopathologic classification of melanocytic uveal tumors evaluated by fine-needle aspiration biopsy.
      and the lack of preservation of histologic architecture of the tumor.
      • Foulds W.S.
      The uses and limitations of intraocular biopsy.
      It has been more than 36 years (J.J.A.) and 20 years (Z.M.C.) since initial introduction of FNAB of selected intraocular tumors into the authors' practices. Since then, periodic assessment of findings and outcomes were performed with the purpose of improving surgical technique, sample procurement and processing, clinical application of results, and reporting. As we expanded the use of FNAB in our clinical practice, it gradually became apparent that these FNAB cases fell into 4 discrete indication categories—diagnostic, confirmatory, investigational, and prognostic—and that FNAB indication could be correlated to certain clinical and pathologic features of these tumors. These realizations led the authors to pursue this project to create a nomenclature and standard reporting scheme for FNAB of intraocular tumors.

       Hypothesis

      We hypothesize that we will be able to retrospectively classify a group of FNABs for tumors of the posterior ocular segment according to indication for the biopsy and identify 4 distinct patient groups, tumor features, biopsy yield, and patient outcomes. These findings could provide important new insights into a more proper nomenclature for categorizing FNAB indication and FNAB reporting in the literature. This fact alone can potentially make studies more reproducible and understandable to our colleagues. We also hypothesize that FNABs carry a low rate of complications if appropriately indicated and performed, making for a safe tool to evaluate solid intraocular tumors. Finally, we hypothesize that diagnostic FNABs have a significant impact in changing patient management.
      To address the hypotheses above, we reviewed our combined experience using FNAB to evaluate solid intraocular tumors. The purposes of this study are (1) to define these distinct indication categories and (2) to evaluate the differences between the patients, tumors, and results of biopsy in these 4 FNAB indication categories.

      Methods

      This is a retrospective study approved by the Institutional Review Board of the University of Cincinnati. Patient information was accessed in compliance with the Health Insurance Portability Act. The FNAB logbooks of the Ocular Oncology Service were reviewed and all patients evaluated by an FNAB of an intraocular tumor that had been performed by 1 of the authors during the period July 1980 through July 2017 (n = 1161) were identified. Detailed chart review of each of these patients during academic year 2017 (ie, July 1, 2016 through June 30, 2017) was performed to ensure adequate follow-up documentation. Some of the reported patients (n = 117) were part of the Collaborative Ocular Oncology Group 1.
      • Onken M.D.
      • Worley L.A.
      • Char D.H.
      • et al.
      Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma.

       Data Collection

      During this records review, the following information was abstracted on every case: (1) demographic variables: date of birth, age of patient on date of FNAB, sex of patient; (2) general ophthalmic variables: best-corrected distance visual acuity [Snellen] in eye subjected to FNAB at examination immediately preceding FNAB, absence vs presence and duration of ocular-visual symptoms attributable to the intraocular tumor; (3) tumor variables: location of tumor (choroid, choroid and ciliary body, retina, retina and ciliary body); largest basal diameter (LBD) of tumor (mm); smallest basal diameter of tumor (mm); thickness of tumor (mm); topographic location of tumor relative to optic disc, fovea, ocular equator, and ora serrata; percentage of retina detached; prebiopsy clinical diagnosis or differential diagnosis of tumor; (4) biopsy variables: date of FNAB, principal indication category of FNAB (diagnostic, confirmatory, investigational, prognostic; discussed in Definitions below), timing of FNAB relative to any intervention (separate procedure, immediately prior to radioactive plaque implantation or other treatment, immediately post-enucleation, immediately post-resection), caliber of needles employed, shape of biopsy needles employed (straight tip vs bent tip), route of the biopsy needle(s) (transvitreal via lamellar scleral incision in pars plana region, transscleral via lamellar scleral flap or pouch, translimbal-transaqueous into iridociliochoroidal tumor, direct transscleral via full-thickness sclera post-enucleation, direct biopsy of exposed tumor via postenucleation eye-wall window, direct biopsy of tumor post-resection), number of needle passes and tumor sites sampled during the biopsy, intraoperative complications, number of FNABs performed on each patient's tumor; (5) postbiopsy variables: sufficiency of aspirate(s) for cytopathologic classification, cytopathologic diagnosis in sufficient cases, cytopathologic melanoma cell type in sufficient cases categorized as uveal melanoma, working diagnosis or differential diagnosis of tumor following FNAB, agreement vs disagreement between prebiopsy clinical diagnosis or differential diagnosis and postbiopsy working diagnosis, delayed complications of biopsy evident within 3 months following clinical FNAB; (6) management variables: initial management of tumor following FNAB, date of initial tumor treatment (or most recent follow-up evaluation if tumor was left untreated), whether management provided following FNAB was the same or different than that planned prior to FNAB; (7) final diagnosis: clinical diagnosis of tumor at most recent follow-up encounter based on all available evidence; (8) outcome variables: agreement vs disagreement between post-FNAB working diagnosis and final diagnosis, survival status of patient through available follow-up, date of death of deceased patients or most recent follow-up encounter date for surviving patients; (9) interval variables: interval between date of FNAB and date of initial treatment or date of most recent follow-up encounter if no treatment was provided, interval between FNAB and date of death of deceased patients or most recent follow-up encounter of surviving patients, interval between date of initial treatment and date of death of deceased patients or most recent follow-up encounter of surviving patients, interval between FNABs in patients who underwent more than 1 FNAB.

       Definitions of Fine Needle Aspiration Biopsy Indication Categories

      Although FNAB indication categories have been briefly shared by our group,
      • Correa Z.M.
      • Augsburger J.J.
      Sufficiency of FNAB aspirates of posterior uveal melanoma for cytologic versus GEP classification in 159 patients, and relative prognostic significance of these classifications.
      • Augsburger J.J.
      • Correa Z.M.
      • Trichopoulos N.
      Prognostic implications of cytopathologic classification of melanocytic uveal tumors evaluated by fine-needle aspiration biopsy.
      the proposed definitions below are a product of thoughtful revision and analysis of patients' pre- and post-FNAB diagnosis, and clinical presentation.
      Diagnostic: A biopsy performed primarily to establish a pathologic diagnosis in cases with an uncertain clinical diagnosis; by definition, the clinical diagnosis for the lesion prompting biopsy must be a differential diagnosis that includes at least 1 malignant neoplasm or a microbial intraocular tumor at a reasonably strong level of probability. A biopsy of this type is usually performed as a separate surgical procedure and not generally in conjunction with therapeutic intervention for the tumor; some exceptions to this timing rule occur, including (1) when cytopathologic slides are prepared and reviewed in the operating room by a pathologist and a decision to provide treatment at that time is based on the pathologist's verbal report, and (2) when the differential diagnosis is between 2 malignant intraocular neoplasms (eg, amelanotic choroidal melanoma vs nonophthalmic primary cancer metastatic to choroid) for which treatment by plaque radiotherapy is planned regardless of which tumor type is identified cytopathologically. Patients with a “probable diagnosis” prior to the biopsy were included in the next group.
      Confirmatory: A biopsy performed primarily (1) to convince a skeptical patient about the accuracy of a clinical diagnosis and appropriateness of recommended treatment, or (2) to justify patient management that may be complex, expensive, or potentially complicated (eg, intravenous chemotherapy, external beam radiation therapy) to professional colleagues who will have to provide that therapy; by definition, the prebiopsy clinical diagnosis for a confirmatory biopsy must be a single tumor type about which there is no clinically relevant doubt; in addition, a biopsy of this type is almost always clinical (ie, performed in the in vivo setting prior to any therapeutic intervention as a separate procedure) rather than post-enucleation or post-resection.
      Investigational: A biopsy performed to evaluate some aspect of performance of a particular method or set of instruments in a specific setting. Typical reasons for performing an investigational biopsy include (1) for the surgeon to gain experience with a particular technique or set of instruments,
      • Larsen A.C.
      • Holst L.
      • Kaczkowski B.
      • et al.
      MicroRNA expression analysis and Multiplex ligation-dependent probe amplification in metastatic and non-metastatic uveal melanoma.
      (2) for a pathologist to gain experience handling and processing the specimen and interpreting findings, (3) for evaluation of yield of tumor cells (sufficiency vs insufficiency) for the intended test, (4) for evaluation of clinical diagnostic accuracy (sensitivity, specificity, predictive values) using the cytopathologic diagnosis obtained from the biopsy as the correct diagnosis, (5) for evaluation of cytologic-histologic correlation in biopsied eyes treated by enucleation or en bloc tumor resection,
      • Staby K.M.
      • Gravdal K.
      • Mork S.J.
      • et al.
      Prognostic impact of chromosomal aberrations and GNAQ, GNA11 and BAP1 mutations in uveal melanoma.
      (6) for evaluation of the safety (frequency of various complications) of the employed biopsy method, and (7) for assessment of the predictive value of suspected potential prognostic factors determined by conventional or investigational testing of the obtained tumor specimens in a laboratory; a biopsy of this type may be clinical (in vivo) or performed post-enucleation or post-resection (in vitro) and may be performed for tumors of any clinical diagnosis or differential diagnosis
      Prognostic: A biopsy performed to obtain a representative specimen of a tumor of a particular clinical type for prognostic classification using a validated method of specimen analysis performed in a CLIA-certified laboratory; a biopsy of this type may be clinical or performed post-enucleation or post-resection. At the present time, the only category of intraocular tumors for which validated prognostic testing is available is melanocytic uveal tumors, and the only validated prognostic test for such tumors at the present time is the DecisionDx-UM test (Castle Biosciences, Inc, Phoenix, Arizona, USA).
      • Onken M.D.
      • Worley L.A.
      • Char D.H.
      • et al.
      Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma.

       Biopsy Techniques

      The instruments and techniques used to perform FNABs of solid intraocular tumors in the authors' ocular oncology practice have been reported in detail previously
      • Augsburger J.J.
      • Shields J.A.
      Fine needle aspiration biopsy of solid intraocular tumors.
      • Augsburger J.J.
      Fine·needle aspiration biopsy in the diagnosis of suspected intraocular cancer.
      • Augsburger J.J.
      • Shields J.A.
      Fine needle aspiration biopsy of solid intraocular tumors: indications, instrumentation and techniques.
      and have changed minimally through the years. Briefly, almost all the cases were performed using 25 gauge, sharp, disposable, hollow-lumen needles without an obturator connected to a disposable 20-inch plastic tubing and a 10 mL plastic syringe to obtain biopsy specimens. The needle length depended on tumor location and route. For clinical (in vivo) FNAB of most posterior segment intraocular tumors whose anterior margin was located at or anterior to the ocular equator, the tumor was first localized intraoperatively by its ocular transillumination shadow. The margins of the tumor shadow were marked on the sclera with a sterile marking pen. Over the marked tumor site, a partial-thickness scleral flap was raised (early years of our experience) or a lamellar scleral pouch was created (more recent years of our experience) to serve as the site of transscleral introduction of the tip of the biopsy needle. The tip of a short (5/8-inch-long) 25 gauge or 27 gauge biopsy needle was passed through the inner scleral lamellae at the site of the flap or pouch and into the substance of the underlying tumor by the ophthalmic surgeon. The surgical assistant then performed several cycles of suction–release of suction using the 10 mL aspirating syringe. In most cases, approximately 10 such cycles were performed during each needle pass. Once all suction force had been discontinued, the surgeon withdrew the tip of the needle and applied light pressure at the biopsy site using the tip of a cotton-tipped applicator stick. In cases sampled in several tumor sites, this process was repeated using a new needle for each needle pass. Following the last needle pass, the lamellar scleral flap or lamellar scleral incision for the pouch was closed with interrupted sutures, usually ones of 8-0 nylon. Cryotherapy to the transscleral biopsy site was not performed in any of these cases.
      For clinical (in vivo) FNAB of most posterior segment intraocular tumors whose anterior margin was located posterior to the ocular equator, the authors employed a transvitreal approach via a lamellar scleral incision parallel to the limbus in the pars plana region (usually about 3.5 mm from the limbus) in the meridian of the tumor and indirect ophthalmoscopy visualization of the passage of the tip of the biopsy needle through the vitreous humor and into the visible intraocular tumor. A 25 gauge or 27 gauge long (1.5-inch-long) biopsy needle was used for such biopsies.
      • Augsburger J.J.
      • Correa Z.M.
      • Schneider S.
      • et al.
      Diagnostic transvitreal fine-needle aspiration biopsy of small melanocytic choroidal tumors in nevus versus melanoma category.
      For tumors ≤2 mm thick, the authors employed straight biopsy needles; however, for most tumors <2 mm thick, the authors bent the tip of the biopsy needles to an angle of 45–60 degrees relative to the needle shaft and directed the tip of the needle into the tumor parallel to the sclera underlying the tumor. Once the tip of the needle had been introduced into the tumor, the surgical assistant performed multiple cycles of suction–release of suction as described in the preceding paragraph. The surgeon then withdrew the needle from the eye and applied light pressure at the eye-wall puncture site in the pars plana region using the tip of a cotton-tipped applicator stick. If the eye became hypotonic when the needle was withdrawn (a not-infrequent occurrence in eyes with predominantly liquid posterior vitreous), an intravitreal injection of balanced salt solution was generally performed using a 30 gauge injection needle to restore the ocular pressure before any additional needle passes or suture closure of the lamellar scleral site. Following the last needle pass, the lamellar scleral incision in the pars plana region was closed with a mattress suture, generally one of 7-0 polyglactin 910. Cryotherapy to the transscleral needle site in the pars plana region was not performed in any of these cases. Laser photocoagulation to or around the site(s) of tumor sampling was not performed in any of these cases.
      Postenucleation and post–transscleral resection (in vitro) biopsies of selected tumors performed for investigational or prognostic purposes were performed immediately following surgical excision. Most of the postenucleation FNABs were performed by direct puncture of full-thickness sclera over the tumor (as localized by postenucleation transillumination) in the intact eye. Some were performed by the in vivo techniques described above. A few were sampled via a full-thickness eye-wall window created to obtain fresh tumor tissue for various investigational studies. Posttransscleral resection biopsies of intraocular tumors obtained via an en bloc excision technique were sampled directly by the surgical assistant while the procedure was continuing. The eye or tumor specimen was then immersed in formalin and transferred to the pathology laboratory for histopathologic processing and analysis.
      During the early years of this study period, the aspiration biopsy specimen contained within the lumen of each biopsy needle was flushed into a sterile disposable syringe using 1–2 mL of balanced salt solution followed by an equivalent volume of absolute alcohol or a cytofixative solution. The syringe was then delivered to the pathology laboratory as soon as possible for further processing by Millipore filter or cell block techniques. Standard cytopreparatory techniques were employed and multiple slides were prepared for evaluation. During more recent years, specimen handling has been simplified by priming the aspirating needle with about 0.05 mL of balanced salt solution (sufficient to be visible in the needle hub but not extending into the connector tubing between the biopsy needle and aspirating syringe) prior to FNAB. Immediately following the biopsy, the aspirating syringe was separated from the tubing and filled with 5–10 mL of air. The syringe was reattached to the tubing and used to spray the biopsy specimen and small amount of balanced salt solution onto prelabeled glass slides. The specimens were fixed immediately using a spray cytofixative, and the slides prepared in this way were submitted to the pathology laboratory for histopathologic and (if indicated) immunocytochemical staining. Aspirates obtained from melanocytic uveal tumors for gene expression profile (GEP) testing (investigational between 2007 and 2009 and prognostic from 2009 through the present) were flushed into a sterile vial using a small aliquot of extraction medium supplied by the laboratory where the test was performed, snap frozen by placing the vial in a Styrofoam box containing dry ice, and shipped to the testing laboratory.

       Exclusion Criteria

      Because the target group for this study consisted of patients who underwent FNAB of a solid intraocular tumor of the posterior ocular segment, cases were excluded if the aspirated material was intraocular fluid and not a solid intraocular tumor and cases undergoing FNAB of a solid tumor of the anterior ocular segment (ie., a tumor of the iris, iris and ciliary body, or ciliary body that did not involve the choroid or retina). In addition, all cases performed using a biopsy needle larger than 25 gauge (a few cases in the early years of this experience) were excluded. Other cases excluded were FNABs performed after July 2014 (to be certain that all cases given the final diagnosis of a benign neoplasm or nonneoplastic ocular lesion simulating a malignant intraocular neoplasm had at least 2 years of potential follow-up after the biopsy until the date of medical records review), and those of patients returned to the referring ophthalmologist for management and follow-up after the biopsy in which no follow-up information had been received from that ophthalmologist. Following these exclusions, the final case series subjected to data analysis numbered 880 cases, including 557 FNABs performed by Z.M.C. and 323 performed by J.J.A.
      During the medical records review, the author recognized that some of the patients in this series underwent FNAB for more than 1 indication (eg, for both diagnostic and prognostic purposes or for both investigational and prognostic purposes). If FNAB was performed for diagnostic or confirmatory purposes and also for investigational or prognostic purposes, the diagnostic or confirmatory purpose was regarded as the principal indication.

       Statistical Analysis

      The statistical analysis was performed using IBM SPSS software (Statistical Package for the Social Sciences; Version 23.0, IBM Corp, Armonk, New York, USA). Conventional descriptive statistics (including mean, standard deviation, minimal, maximal, median) was used for all continuous numerical variables that included age of patient, LBD of the tumor, and maximal tumor thickness. Frequency distributions (including number and percentage of cases) were computed for all study variables. Patients were divided accordingly for intrinsically dichotomous variables such as sex and intraocular tumor location. For intrinsically numerical variables, conventional values for those variables as cut points between trichotomous subgroups were selected based on our series distribution in an attempt to create comparable-size groups and included descriptive statistics, mean, median, and standard deviation. The distributions of LBD and thickness for the tumors in the 4 groups according to the FNAB indication were computed and compared using 1-way analysis of variance (ANOVA), and the distribution of intraocular tumor location was computed and compared using the χ2 statistic.
      To evaluate the relationships between FNAB indication and other evaluated variables in this study, we performed cross-table analysis with computation of χ2 values. An alpha equal to or less than 0.001 was selected as the nominal significance level for all statistical tests performed in this study. Kaplan-Meier cumulative probability of survival was computed and plotted for each of the 4 FNAB indication groups. Additional melanoma specific survival by FNAB indication category was computed to evaluate this subset of patients.

      Results

       Whole-Group Results

      As indicted above, the complete dataset for this study consisted of 880 cases of FNAB performed on a solid intraocular tumor of the posterior ocular segment during the period July 1980 through July 2014. The patients in the entire group ranged in age from <1 month to 99 years (mean age 56.6 years, median age 59.0 years) at the time of biopsy. Four hundred and forty-six patients (50.7%) were female and 436 (49.3%) were male. The tumor was located in the choroid in 625 (71.0%), in both the choroid and ciliary body in 208 (23.7%), and in the retina in 47 (5.3%).
      The principal indication category for FNAB in the 880 cases was diagnostic in 292 (33.2%), confirmatory in 121 (13.8%), investigational in 187 (21.3%), and prognostic in 280 (31.8%).
      The tumors in this series ranged in size from 2.0 mm to 24.0 mm in LBD (mean 12.0 mm, median 12.0 mm) and from 1.0 mm to 21.3 mm in maximal tumor thickness (mean 5.5 mm, median 4.4 mm). The FNABs were performed as a separate procedure prior to any therapeutic intervention for the intraocular tumor in 372 (42.3%), at the time of radioactive plaque implantation in 279 (31.7%), immediately post-enucleation in 225 (25.6%), immediately following transscleral en bloc tumor resection in 2 (0.2%), and immediately prior to laser therapy for the tumor in 2 (0.2%).

       Differential Distributions of Patient and Tumor Variables and Relevant Outcomes in Fine Needle Aspiration Biopsy Categories

       Age of Patients

      Table 1 depicts the age distribution of patients on the date of FNAB for the total group and for FNAB indication category subgroups. Although the mean age of the patients was quite similar in 3 of the FNAB category groups (55.4 years in the diagnostic FNAB group, 54.3 years in the confirmatory FNAB group, and 50.8 years in the investigational group), the patients in the prognostic FNAB group were significantly older (62.6 years) than those in the other groups (ANOVA F statistic = 9.5, P < .0001). Furthermore, analysis of age distributions by 20-year-wide categories revealed several noteworthy differences among the FNAB category groups. For example, only 0.7% of the patients (n = 2) in the prognostic FNAB category vs 14.4% of the patients (n = 27) in the investigational FNAB group were under the age of 20 years (n = 46). Also, 9.6% of the patients (n = 27) in the prognostic FNAB group were over the age of 80 years (n = 49) while no other FNAB group had over 0.6% of its patients in that age category (P < .001).
      Table 1Age Distribution of Patients on the Date of Fine Needle Aspiration Biopsy for the Total Group and for Indication Category Subgroups
      Age Category of Patient on Date of FNABFNAB Indication Category
      Indication for FNAB.
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      ≤20 years14 (4.8)3 (2.5)27 (14.4)2 (0.7)46 (5.2)
      >20 but ≤40 years37 (12.7)28 (23.1)25 (13.4)13 (4.6)103 (11.7)
      >40 but ≤60 years113 (38.7)39 (32.2)54 (28.9)107 (38.2)313 (35.6)
      >60 but ≤80 years117 (40.1)46 (38.0)75 (40.1)131 (46.8)369 (41.9)
      >80 years11 (3.8)5 (4.1)6 (3.2)27 (9.6)49 (5.6)
      Totals292 (100.0)121 (100.0)187 (100.0)280 (100.0)880 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of patients evaluated; % = relative and cumulative percentage; Total = total number of patients in each category.
      Patient age difference among the indication category subgroups was statistically significant, P < .0001.
      a Indication for FNAB.

       General Ophthalmic Variables

      The majority (74.9%) of the evaluated patients had vision symptoms attributed to the biopsied tumor. The respective percentages of symptomatic patients was 64.4% (188 of 292 patients) in the diagnostic FNAB group, 77.7% (94 of 121 patients) in the confirmatory FNAB group, 83.4% (156 of 187 patients) in the investigational FNAB group, and 78.9% (221 of 280 patients) in the prognostic FNAB group.
      Table 2 shows a trichotomous breakdown of visual acuity in the different indication categories. Most of patients (62.0%) in the diagnostic FNAB group had good baseline visual acuity compared to the other groups. Although there is a trend of better vision among patients evaluated by diagnostic biopsy, the difference was not statistically significant.
      Table 2Best-Corrected Visual Acuity at Baseline in the 4 Defined Fine Needle Aspiration Biopsy Indication Categories (N = 880)
      Visual Acuity at Baseline (Trichotomous)Indication Category of FNAB
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      Poor (20/200 or worse)61 (20.9)40 (33.1)92 (49.2)67 (23.9)260 (29.5)
      Intermediate (worse than 20/50 but better than 20/200)50 (17.1)16 (13.2)35 (18.7)65 (23.2)166 (18.9)
      Good (better than 20/50)181 (62.0)65 (53.7)60 (32.1)148 (52.9)454 (51.6)
      Column totals292 (100.0)121 (100.0)187 (100.0)280 (100.0)880 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of patients evaluated; % = relative and cumulative percentage; Total = total number of patients in each category.
      Difference did not reach statistical significance (P = .005).

       Tumor Size

      Table 3 shows summary descriptive data on LBD (mm) and Table 4 shows the data on maximal thickness (mm) of the tumor in the 4 FNAB indication categories. As shown by these tables, the tumors in the diagnostic FNAB category were smaller on average (in both LBD and thickness) than those in the other FNAB categories (ANOVA F statistic 49.0, P < .0001). Tumors in the confirmatory FNAB group were intermediate in size between those in the investigational and prognostic FNAB groups.
      Table 3Largest Basal Diameter of Biopsied Tumor for the Total Group and for Fine Needle Aspiration Biopsy Indication Category Subgroups (ANOVA F Statistic 49.0, P < .0001)
      Comparative Means, Standard Deviations of Means, Extreme Values, and Medians of Largest Basal Diameter (mm) of Biopsied Tumor
      FNAB Indication CategoryMeanSDMinimumMaximumMedian
      Diagnostic10.13.72.023.09.0
      Confirmatory11.63.64.522.011.5
      Investigational13.93.76.022.513.0
      Prognostic12.83.55.024.013.1
      Total
      Categorical Summarization of Results
      Categorical LBDFNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      Small (LBD ≤10 mm)177 (60.6)51 (42.1)34 (18.2)79 (28.2)341 (38.8)
      Medium (LBD >10 mm but ≤15 mm)89 (30.5)52 (43.0)89 (47.6)137 (48.9)367 (41.7)
      Large (LBD >15 mm)26 (8.9)18 (14/9)64 (34.2)64 (22.9)172 (19.5)
      Column totals292 (100.0)121 (100.0)187 (100.0)280 (100.0)880 (100.0)
      FNAB = fine needle aspiration biopsy; LBD = largest basal diameter; N = number of patients evaluated; % = relative and cumulative percentage.
      Table 4Maximal Thickness of Biopsied Tumor for the Total Group and for Fine Needle Aspiration Biopsy Indication Category Subgroups (ANOVA F Statistic 66.0, P < .0001)
      Comparative Means, Standard Deviations of Means, Extreme Values, and Medians of Tumor Thickness (mm) of Biopsied Tumor
      FNAB Indication CategoryMeanSDMinimumMaximumMedian
      Diagnostic3.72.51.018.02.7
      Confirmatory4.62.61.315.93.9
      Investigational7.93.72.021.36.0
      Prognostic6.23.21.318.05.6
      Total
      Categorical Summarization of Results
      Categorical Tumor ThicknessFNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      Thin (TH ≤3.5 mm)200 (68.5)57 (47.1)25 (13.4)70 (25.0)352 (40.0)
      Intermediate (TH> 3.5 mm but ≤7 mm)61 (20.9)43 (35.3)68 (36.4)114 (40.7)286 (32.5)
      Thick (TH >7 mm)31 (10.6)21 (17.4)94 (50.3)96 (34.3)242 (27.5)
      Column totals292 (100.0)121 (100.0)187 (100.0)280 (100.0)880 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of patients evaluated; TH = thickness; % = relative and cumulative percentage.

       Intraocular Tumor Location

      Table 5 describes the intraocular tissue location of biopsied tumor for the total group and for FNAB indication category subgroups. Tumor location in the eye also differed substantially among the FNAB groups. This difference was statistically significant (P < .001) For example, 8% of the tumors in the confirmatory FNAB group vs 0.9% of the tumors in the prognostic FNAB group were located exclusively in the ciliary body. Because the prognostic FNAB group (n = 280) consisted only of patients with clinically diagnosed primary posterior uveal melanoma, no tumors in the prognostic FNAB group were located in the retina. In contrast, 13.9% of the tumors in the investigational FNAB group (n = 26/187), 6.3% of tumors in the diagnostic FNAB group (n = 15/292), and 5.3% of tumors in the confirmatory FNAB group (n = 6/121) were located in the retina.
      Table 5Intraocular Tissue Location of Biopsied Tumor for the Total Group and for Fine Needle Aspiration Biopsy Indication Category Subgroups
      Intraocular Tissue Location of TumorFNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      Choroid alone235 (80.5)91 (75.2)111 (59.4)188 (67.1)625 (71.0)
      Ciliary body and choroid35 (12.0)23 (19.0)45 (24.1)78 (27.9)181 (20.6)
      Iris, ciliary body, and choroid7 (2.4)1 (0.8)5 (2.7)14 (5.0)27 (3.1)
      Retina15 (5.1)6 (5.0)26 (13.9)0 (0.0)47 (5.3)
      Column totals292 (100.0)121 (100.0)187 (100.0)280 (100.0)880 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of patients evaluated; % = percentage of the overall cohort.
      (P < .001.)

       Diagnosis or Differential Diagnosis

      The pre-FNAB diagnoses and differential diagnoses in the 880 study cases by FNAB category are presented in Table 6. Consistent with our definitions of the 4 FNAB categories, each diagnostic FNAB was associated with a differential diagnosis that included at least 1 malignant intraocular neoplasm while each confirmatory FNAB was associated with an unequivocal prebiopsy clinical diagnosis. The most frequent differential diagnoses in our diagnostic FNAB subgroup (n = 292) were “uveal nevus vs uveal melanoma” in 133 (45.5%), “uveal melanoma vs metastatic carcinoma to uvea” in 57 (19.7%), “probable uveal melanoma but atypical” in 52 (17.8%), and “intraocular lymphoma vs inflammatory tumor” in 11 (3.8%). The most frequent prebiopsy clinical diagnoses in the confirmatory FNAB subgroup (n = 121) were “uveal melanoma” in 70 (57.9%) and “metastatic carcinoma to uvea” in 14 (11.6%). In contrast, the clinical diagnosis in the investigational FNAB subgroup (n = 187) was “uveal melanoma” in 161 (86.1%) and “retinoblastoma” in 26 (13.9%) cases. In contrast, the clinical diagnosis in all prognostic FNAB subgroup (n = 280) was melanocytic uveal tumor (suspected or unequivocal uveal melanoma in 279 [99.6%] and probable small uveal melanoma in 1 case [0.4%]).
      Table 6Prebiopsy Diagnosis or Differential Diagnosis for the Total Group and for Fine Needle Aspiration Biopsy Indication Category Subgroups
      FNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      SPECIFIC TUMOR DIAGNOSIS
       Uveal melanoma070 (8.0)161 (18.3)279 (31.7)510 (58.0)
       Probable uveal melanoma but atypical030 (3.4)0030 (3.4)
       Retinoblastoma02 (0.2)26 (2.9)-28 (3.1)
       Metastatic tumor014 (1.6)0-14 (1.6)
       Primary intraocular lymphoma04 (0.5)0-4 (0.5)
       Other specified tumor type
      The tumor was a clinically diagnosed intraocular plasmacytoma (confirmed cytopathologically by FNAB) in a patient with well-documented systemic multiple myeloma.
      01 (0.1)0-1 (0.1)
      DIFFERENTIAL DIAGNOSIS
       Uveal nevus vs uveal melanoma (tumor ≤3.5 mm thick)133 (15.1)001
      This patient had a very small (thickness = 1/0 mm) melanocytic macular choroidal tumor associated with serous subretinal fluid, surface clumps of orange pigment, and visual symptoms; the tumor aspirates obtained by FNAB were submitted only for gene expression profiling in this case. The aspirates were sufficient for gene expression profile testing and showed the tumor to be a class 1B tumor.
      (0.1)
      134 (15.2)
       Dormant melanoma >3.5 mm thick vs large uveal nevus52 (5.9)00-52 (5.9)
       Uveal melanoma vs metastasis58 (6.6)00-58 (6.6)
       Uveal melanoma vs inflammatory tumor
       Intraocular lymphoma vs inflammatory tumor11 (1.3)00-11 (1.3)
       Other differential diagnosis (<5 cases of any other differential)38 (4.3)00-38 (4.3)
      Column totals292 (33.2)121 (13.8)187 (212)280 (31.8)880 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of patients evaluated; % = percentage of the overall cohort.
      a The tumor was a clinically diagnosed intraocular plasmacytoma (confirmed cytopathologically by FNAB) in a patient with well-documented systemic multiple myeloma.
      b This patient had a very small (thickness = 1/0 mm) melanocytic macular choroidal tumor associated with serous subretinal fluid, surface clumps of orange pigment, and visual symptoms; the tumor aspirates obtained by FNAB were submitted only for gene expression profiling in this case. The aspirates were sufficient for gene expression profile testing and showed the tumor to be a class 1B tumor.

       Timing of the Biopsy

      Timing of the FNAB relative to initial tumor treatment in the 4 FNAB indication categories is presented in Table 7. Consistent with the previously described 4 FNAB categories, both diagnostic and confirmatory FNABs were (by definition) always performed as a separate procedure prior to any therapeutic intervention and were intended to establish a pathologic diagnosis for the tumor that could be used to justify postbiopsy therapeutic recommendations. In contrast, many of the FNAB procedures in the investigational category (158 cases, 84.5%) and some of the cases in the prognostic FNAB category (69 cases, 24.6%) were performed post-enucleation or immediately post-resection of the tumor. Two hundred three biopsies in the prognostic FNAB indication group (72.5%) were performed immediately prior to radioactive plaque implantation; however, this timing of the biopsy relative to treatment occurred in <20% of cases in all 3 other FNAB indication categories. The timing of the biopsy correlated strongly with the FNAB indication (P < 0.001).
      Table 7Timing of Fine Needle Aspiration Biopsy Relative to Initial Tumor Treatment in the 880 Study Cases, by Indication Category (P < .001)
      Timing of FNABFNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      Separate procedure266 (91.1)98 (81.1)1 (0.5)7 (2.5)372 (42.3)
      Immediately before radioactive plaque implantation25 (8.6)23 (19.0)28 (15.0)203 (79.5)279 (31.7)
      Immediately post-enucleation0 (0.0)0 (0.0)158 (84.5)67 (23.9)225 (25.6)
      Immediately post-resection of tumor0 (0.0)0 (0.0)0 (0.0)2 (0.7)2 (0.2)
      Immediately before laser therapy1 (0.3)0 (0.0)0 (0.0)1 (0.4)2 (0.2)
      Column totals292 (100.0)121 (100.0)187 (100.0)280 (100.0)880 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of patients evaluated; % = percentage of the overall cohort.

       Planned Testing of Tumor Aspirates

      All cases prior to September 2007 were sampled only for cytopathologic analysis and pathologic classification. In the early years of our series, only standard chemical stains (hematoxylin-eosin, periodic acid–Schiff) and occasional special chemical stains (eg, Fontana) were employed as aids to cytopathologic analysis. Immunocytochemical stains gradually became available and were employed in some cases to supplement the chemical stains (provided that enough cellular material was obtained to prepare several slides). In contrast, virtually all melanocytic uveal tumors biopsied between September 2007 and November 2009 were also sampled for GEP testing and classification
      • Onken M.D.
      • Worley L.A.
      • Char D.H.
      • et al.
      Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma.
      • Harbour J.W.
      A prognostic test to predict the risk of metastasis in uveal melanoma based on a 15-gene expression profile.
      in a Collaborative Ocular Oncology Group prospective validation study of the GEP test's prognostic value,
      • Onken M.D.
      • Worley L.A.
      • Char D.H.
      • et al.
      Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma.
      and almost all of the posterior uveal melanocytic tumors managed in our center since the conclusion of that validation study had been tested for both cytopathologic classification and GEP testing using the DecisionDx-UM test (Castle Biosciences, Inc, Friendwood, Texas, USA).
      • Harbour J.W.
      • Chen R.
      The DecisionDx-UM gene expression profile test provides risk stratification and individualized patient care in uveal melanoma.

       Biopsy Needles Employed in Fine Needle Aspiration Biopsy

      Table 8 summarizes study data on the biopsy needles employed in the 4 FNAB indication subgroups. The type of needle used and any additional adjustments depend on tumor location and size. Over 95% of FNABs in all subgroups were performed using 25 gauge needles during the interval under study; however, during 2016, we switched to 27 gauge needles for most biopsies and continue to use these smaller-caliber needles for most biopsies at present. Long (1.5-inch) biopsy needles were used in the majority of all FNABs in all 4 FNAB indication categories. The majority of the FNABs in the confirmatory, investigational, and prognostic subgroups were performed using straight biopsy needles; however, the majority of biopsies in the diagnostic FNAB category (n = 166, 56.8%) were performed using our bent-needle technique, contrasting with the prognostic groups, in which a bent needle was used in only 49 cases (17.5%) (P < .001).
      Table 8Biopsy Needles Used to Perform Fine Needle Aspiration Biopsy for the Total Group and for Indication Category Subgroups
      FNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      Needle caliber
       25 gauge288 (97.9)117 (96.7)183 (97.9)279 (99.6)865 (98.3)
       27 gauge4 (1.4)4 (3.3)3 (1.6)1 (0.4)12 (1.4)
       30 gauge2 (0.7)0 (0.0)1 (0.5)0 (0.0)3 (0.3)
      Needle length
       5/8 inch long31 (10.6)7 (5.8)46 (24.6)104 (37.1)188 (21.4)
       1.5 inches long261 (89.4)114 (94.8)141 (75.4)176 (62.9)692 (78.6)
      Needle shape
      Indicates if the needle shaft was bent to allow complete imbedding of the needle hub during the biopsy.
       Straight needles126 (43.2)68 (56.2)163 (87.2)231 (82.5)588 (66.8)
       Bent needles166 (56.8)53 (43.8)23 (12.8)49 (17.5)292 (33.2)
      FNAB = fine needle aspiration biopsy; N = number of cases; (%) = relative percentages.
      a Indicates if the needle shaft was bent to allow complete imbedding of the needle hub during the biopsy.

       Route of Needle Passage During Fine Needle Aspiration Biopsy

      The route of needle passage (ie, transsclerally into a peripheral choroidal or ciliochoroidal tumor [direct route] vs transvitreally into a posterior choroidal or retinal tumor [indirect route]) into the tumor during FNAB is relevant with regard to potential complications of the biopsy (see below) and investigations of potential needle tract seeding by tumor cells (Table 9). While we studied the needle tracts of a number of early cases in which FNAB was performed post-enucleation or as a clinical biopsy shortly prior to enucleation, most of our later cases coming to enucleation did not undergo such analysis.
      Table 9Route of Needle Passage Relative to Indication for Fine Needle Aspiration Biopsy
      Route of Needle Passage During FNABFNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      Transvitreous into tumor via pars plana entry site255 (87.3)109 (90.1)119 (63.6)143 (51.1)626 (71.1)
      Transaqueous-iris-vitreous via limbus4 (1.4)4 (3.3)13 (7.0)2 (0.7)23 (2.6)
      Transaqueous into tumor via limbus
      This route was used to biopsy 2 children with clinically diagnosed retinoblastoma whose parents refused enucleation without pathologic confirmation of the diagnosis, 6 other children with suspected retinoblastoma in a blind eye, and most of the eyes enucleated for retinoblastoma and investigational biopsies in this series.
      3 (1.0)1 (0.8)7 (3.7)2 (0.7)13 (1.5)
      Transscleral into tumor via scleral flap29 (9.9)7 (5.8)11 (5.9)77 (27.5)124 (14.1)
      Transscleral into tumor post-enucleation1
      This case involved a blind eye with an intraocular tumor of uncertain nature pre-enucleation. The tumor proved to be a predominantly necrotic uveal melanoma pathologically.
      (0.3)
      37 (19.8)54 (19.2)92 (10.5)
      Direct into tumor post-resection2 (0.7)2 (0.7)
      Total292 (100.0)121 (100.0)187 (100.0)280 (100.0)880 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of patients evaluated; % = percentage of the overall cohort.
      a This route was used to biopsy 2 children with clinically diagnosed retinoblastoma whose parents refused enucleation without pathologic confirmation of the diagnosis, 6 other children with suspected retinoblastoma in a blind eye, and most of the eyes enucleated for retinoblastoma and investigational biopsies in this series.
      b This case involved a blind eye with an intraocular tumor of uncertain nature pre-enucleation. The tumor proved to be a predominantly necrotic uveal melanoma pathologically.

       Number of Tumor Sites Sampled During Fine Needle Aspiration Biopsy

      Table 10 summarizes the number of tumor sites sampled during FNAB in the 4 FNAB indication categories. The number of tumor sites sampled during FNAB is potentially relevant with regard to sufficiency of the biopsy for cytopathologic classification and diagnosis and (in clinical biopsies) the frequency of complications such as major intravitreal bleeding during the biopsy and postbiopsy rhegmatogenous retinal detachment. The number of tumor sites sampled in our entire series of 880 cases was 1 in 255 (29.0%), 2 in 169 (19.2%), 3 in 307 (34.9%), 4 in 140 (15.9%), and 5 in 9 (1.0%). The median number of tumor sites sampled in the entire group was 2. Inspection of the table reveals that 238 of the 280 biopsies (85.0%) in the prognostic FNAB group entailed sampling of 3 or more tumor sites while <50% of the biopsies in the other 3 FNAB groups entailed sampling this many tumor sites. The table also reveals that a large proportion of the investigational FNABs (130 of 187 biopsies, 69.5%) entailed sampling of only 1 tumor site. Most of these biopsies were performed during the first several years of our study, prior to the decision to sample 2 or more tumor sites in most FNABs to reduce the likelihood of an insufficient tumor aspirate for cytopathologic classification and diagnosis.
      Table 10Number of Sampled Sites Relative to Indication for Fine Needle Aspiration Biopsy in 880 Patients
      Number of Needle Passes (Sampled Sites)Indication Category of FNAB
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      One73 (25.0)49 (40.5)130 (69.5)3 (1.1)255 (29.0)
      Two76 (26.0)34 (28.1)20 (10.7)39 (13.9)169 (19.2)
      Three111 (38.0)28 (23.1)11 (5.9)157 (56.1)307 (34.9)
      Four30 (10.3)10 (8.3)26 (13.9)74 (26.4)140 (15.9)
      Five2 (0.7)7 (2.5)9 (1.0)
      Column total292 (100.0)121 (100.0)187 (100.0)280 (100.0)880 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of cases; (%) = relative percentages.

       Frequency of Insufficiently Cellular Specimen for Planned Test

      In the entire series, 146 of the 880 FNABs (16.6%) were associated with an insufficient tumor sample for cytopathologic classification. As shown in Table 11, the highest proportion of insufficient aspirates for cytopathologic classification of tumor type occurred in the diagnostic FNAB subgroup (28.4%), while the lowest proportion occurred in the investigational FNAB subgroup (5.3%). As indicated previously, the tumors in the investigational FNAB group were substantially larger than those in the other 3 groups and were frequently biopsied post-enucleation for cytohistopathologic correlation. In contrast, tumors in the diagnostic FNAB indication category tended to be smaller than the tumors in the other groups and were frequently biopsied to determine whether a small melanocytic choroidal tumor was a melanoma or an elevated, activated uveal nevus. Consequently, these results for insufficiency of the aspirates for cytologic classification and pathologic diagnosis are not really surprising.
      Table 11Sufficiency of Fine Needle Aspiration Biopsy Aspirates for Cytopathologic Classification for the Total Group and for Indication Category Subgroups
      Sufficiency of Aspirates for Cytopathologic ClassificationFNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      Sufficient208 (71.2)96 (79.3)177 (94.7)252 (90.0)733 (83.3)
      Insufficient83 (28.4)25 (20.7)10 (5.3)28 (10.0)146 (16.6)
      Not submitted for cytology
      Because of small tumor dimension and posterior location, 1 patient did not have a specimen submitted to cytology; only submitted for gene expression profile testing.
      1 (0.3)0 (0.0)0 (0.0)0 (0.0)1 (0.1)
      Column totals292 (100.0)121 (100.0)187 (100.0)280 (100.0)880 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of cases; (%) = relative percentages.
      a Because of small tumor dimension and posterior location, 1 patient did not have a specimen submitted to cytology; only submitted for gene expression profile testing.
      Table 12 shows the frequencies of insufficiently cellular specimens for GEP classification in the 4 FNAB indication subgroups in this series. As shown by the table, over 98% of the melanocytic uveal tumors whose aspirates were submitted for GEP testing in this series yielded a sufficiently cellular specimen of the tumor for GEP testing in all 4 of the FNAB indication categories. One of these patients (1.1%) was in the diagnostic FNAB group and 4 (1.5%) were in the prognostic FNAB group. The very low frequency of an insufficient aspirate of a melanocytic posterior uveal tumor for GEP testing is consistent with the observation that only a few tumor cells are generally required for successful GEP testing.
      • Correa Z.M.
      • Augsburger J.J.
      Sufficiency of FNAB aspirates of posterior uveal melanoma for cytologic versus GEP classification in 159 patients, and relative prognostic significance of these classifications.
      Table 12Sufficiency of Fine Needle Aspiration Biopsy Aspirates of Clinical Diagnosed Melanocytic Uveal Tumors for Gene Expression Profile Testing and Classification for the Subgroup With a Prebiopsy Diagnosis of Melanocytic Uveal Tumor
      Uveal melanoma, probable melanoma but atypical, dormant uveal melanoma vs large uveal nevus, uveal nevus vs uveal melanoma.
      and for Indication Category Subgroups
      Sufficiency of FNAB Aspirates for GEPFNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      Sufficient93 (98.9)19 (100.0)40 (100.0)265 (98.5)417 (98.8)
      Insufficient1 (1.1)0 (0.0)0 (0.0)4 (1.5)5 (1.2)
      Column totals94 (100.0)19 (100.0)40 (100.0)269 (100.0)422
      This number does not include several cases in which a laboratory or technician error resulted in the specimen not arriving at and/or not being processed in the testing laboratory.
      (100.0)
      FNAB = fine needle aspiration biopsy; GEP = gene expression profile test; N = number of cases; (%) = relative percentages.
      a Uveal melanoma, probable melanoma but atypical, dormant uveal melanoma vs large uveal nevus, uveal nevus vs uveal melanoma.
      b This number does not include several cases in which a laboratory or technician error resulted in the specimen not arriving at and/or not being processed in the testing laboratory.

       Intraoperative Complications

      The frequencies of different intraoperative complications of biopsy encountered in the 4 FNAB indication categories are shown in Table 13. Intraoperative complications can only be assessed in cases undergoing a clinical biopsy (ie, an in vivo biopsy, not a postenucleation or postresection tumor biopsy). In the 653 clinical FNABs (74.2%) in this series, 40.7% (n = 358) had no complications from the FNAB procedure, 30.9% (n = 272) had minor intraoperative bleeding from 1 or more of the tumor puncture sites, and only 2.5% (n = 22) had major intraoperative bleeding during the procedure. The only other intraoperative complication was a traumatic cataract (in 1 early case) that was successfully managed by cataract extraction and intraocular lens implantation.
      Table 13Intraoperative Complications for the Subgroup Undergoing Clinical Biopsy and for Fine Needle Aspiration Biopsy Indication Category Subgroups
      Intraoperative ComplicationFNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      N (%)N (%)N (%)N (%)N (%)
      None134 (45.9)48 (39.7)18 (62.1)158 (74.9)358 (54.8)
      Minor intraocular bleeding154 (52.7)71 (58.7)7 (24.1)40 (19.0)272 (41.7)
      Major intraocular bleeding3 (1.0)2 (1.7)4 (13.8)14 (6.2)22 (3.4)
      Traumatic cataract1 (0.3)0 (0.0)0 (0.0)0 (0.0)1 (0.2)
      Column totals292 (100.0)121 (100.0)29 (100.0)211 (100.0)653 (100.0)
      FNAB = fine needle aspiration biopsy; N = number of cases; (%) = relative percentages.

       Postoperative Complications Within or By 3 Months Post-biopsy

      The frequencies of different postoperative (delayed) complications of biopsy encountered in the 4 FNAB indication categories within 3 months of the biopsy are shown in Table 14. Postoperative complications of FNAB can be addressed only in patients who underwent a clinical FNAB and were managed by an eye-preserving treatment (including plaque radiotherapy or proton beam irradiation) following the FNAB. Patients who were managed by enucleation or transscleral tumor resection shortly after the biopsy would not be appropriate cases for assessment of postoperative complications of the biopsy. In the 573 study cases managed conservatively following FNAB, minor vitreous hemorrhage (evident on ophthalmoscopy but not sufficient to cause visual blurring or impairment of our fundus view) was evident at 3 months in 277 cases (48.3%), moderate vitreous hemorrhage (sufficient to cause some visual impairment and partial impairment of our fundus view) was evident in 48 cases (8.4%), and major vitreous hemorrhage (precluding any fundus view and prompting discussion of vitrectomy for blood washout) was evident in 21 cases (3.7%). A rhegmatogenous retinal detachment developed post-FNAB in 4 patients in this series (0.7%) and endophthalmitis developed in 2 patients (0.3%). Fortunately, both of these latter patients were treated promptly and regained visual acuity to their prebiopsy level. Overall, 22 patients (3.8%) required an additional intervention to manage their FNAB complication(s).
      Table 14Complications of Fine Needle Aspiration Biopsy Evident Within 3 Months Following Biopsy for the 573 Patients Who Underwent Clinical Biopsy and Were Managed by an Eye-Preserving Method After the Biopsy, and for Indication Category Subgroups
      Complications of FNAB Evident Within 3 MonthsFNAB Indication Category
      DiagnosticConfirmatoryInvestigationalPrognosticTotal
      Column percentages add to >100.0% because some patients experienced more than 1 complication of the biopsy.
      N (%)N (%)N (%)N (%)N (%)
      None92 (36.8)45 (51.7)7 (24.1)83 (40.1)227 (39.6)
      Minor vitreous hemorrhage, clearing spontaneously145 (58.0)37 (42.5)15 (51.7)80 (38.6)277 (48.3)
      Moderate vitreous hemorrhage, clearing spontaneously9 (3.6)3 (3.4)6 (20.7)30 (14.5)48 (8.4)
      Major vitreous hemorrhage, not clearing spontaneously &/or prompting PPV4 (1.6)2 (2.3)1 (3.4)14 (6.8)21 (3.7)
      Rhegmatogenous retinal detachment2 (0.8)1 (1.1)0 (0.0)1 (0.5)4 (0.7)
      Endophthalmitis1 (0.4)1 (1.1)0 (0.0)0 (0.0)2 (0.4)
      Traumatic cataract1 (0.4)0 (0.0)0 (0.0)0 (0.0)1 (0.2)
      FNAB = fine needle aspiration biopsy; N = number of cases; PPV = pars plana vitrectomy; (%)= relative percentages.
      a Column percentages add to >100.0% because some patients experienced more than 1 complication of the biopsy.

       Frequency With Which Biopsy Changed Planned Treatment

      The frequency with which treatment provided for the intraocular tumor following FNAB differed from treatment anticipated for that tumor prior to the biopsy can be assessed only in patients who underwent biopsy as a separate procedure and not in conjunction with or immediately following tumor excision by enucleation or transscleral resection. In the 653 applicable cases of clinical FNABs in our series, 242 patients (37.1%) underwent a different treatment following the biopsy than had been anticipated prior to the biopsy. The biopsy was not the sole cause for this change in all these patients but was the principal determinant of altered posttreatment patient management in most of them.

       Diagnostic Accuracy of Post–Fine Needle Aspiration Biopsy Working Diagnosis

      Diagnostic accuracy of fine needle aspiration biopsy as a standalone test can be assessed only in patients whose biopsy yielded a sufficient aspirate for cytopathologic classification and diagnosis and subsequently underwent enucleation or surgical resection of the intraocular tumor or had sufficient postbiopsy clinical follow-up to confirm the nature of the intraocular tumor beyond reasonable doubt.
      • Folberg R.
      • Augsburger J.J.
      • Gamel J.W.
      • Shields J.A.
      • Lang W.R.
      Fine-needle aspirates of uveal melanomas and prognosis.
      As mentioned previously, a substantial number of our biopsies (mostly ones in which a relatively small tumor was biopsied) were associated with an insufficiently cellular specimen for cytopathologic classification. Rather than evaluating the cytopathologic classification of the aspirates of each case as a standalone diagnostic marker, the authors formulated a postbiopsy working diagnosis that not only took into account the FNAB cytopathology report (and GEP report on appropriately submitted tumor aspirates of melanocytic uveal tumors) but also considered our level of diagnostic certainty or uncertainty regarding tumor type pre-biopsy, the size of the tumor, and the early postbiopsy behavior of the tumor. We then compared this early postbiopsy working diagnosis to the final diagnosis of each tumor in this series (definition of final diagnosis is described in Methods). This comparison is presented for the entire study group of 880 cases in Table 15. Inspection of this table shows that our post-FNAB working diagnosis was concordant with the final diagnosis in 820 of the 880 cases (93.0%). Assuming that one regards cases having both a working diagnosis and final diagnosis of “uveal nevus vs uveal melanoma” as true positive cases, our results indicate a sensitivity of 0.95, a specificity of 0.72, a positive predictive value of 0.98, and a negative predictive value of 0.51 in this series (Table 16).
      Table 15Cross-tabulation of Working Diagnosis After Fine Needle Aspiration Biopsy
      Clinical diagnosis taking into account FNAB results and clinical features of the case.
      vs Final Diagnosis for the Total Group
      Final Diagnosis
      Malignant Intraocular NeoplasmBorderline Intraocular NeoplasmBenign Intraocular TumorTotal
      Uveal MelanomaMetastatic CancerPrimary Intraocular LymphomaRetino-blastomaOther Malignant Intraocular NeoplasmUveal Nevus vs. MelanomaUveal NevusNon-neoplastic Tumor
      Working diagnosis after FNAB
      Malignant intraocular neoplasm
      Uveal melanoma6810001521690
      Metastatic cancer23000100033
      Primary intraocular lymphoma008000008
      Retinoblastoma00028000028
      Other malignant intraocular neoplasm300030006
      Borderline intraocular neoplasm
      Uveal nevus vs melanoma320000274063
      Benign intraocular tumor
      Uveal nevus50000018023
      Nonneoplastic tumor10201002529
      Column totals7243010286322426880
      FNAB = fine needle aspiration biopsy.
      Proportion of cases with correct working diagnosis (ie, working diagnosis = final diagnosis) = 820 of 880 = 0.93.
      a Clinical diagnosis taking into account FNAB results and clinical features of the case.
      Table 16Statistical Analysis of Diagnostic Accuracy in the Entire Study Group, Assuming That Cases That Had a Working Diagnosis and Final Diagnosis of Uveal Nevus vs Melanoma Were True Positives With Regard to Diagnosis
      Final Diagnosis
      Malignant NeoplasmBenign Neoplasm
      Working diagnosis after FNAB
       Malignant neoplasm78412
       Benign neoplasm4143
      Sensitivity: 784/823 = 0.95
      Specificity: 43/60 = 0.72
      Positive predictive value: 784/796 = 0.98
      Negative predictive value: 43/84 = 0.51
      Positive likelihood ratio: (784/823)/(12/60) = 4.75
      Negative likelihood ratio: (41/823)/(43/60) = 0.07
      Risk ratio: (784/796)/(41/84) = 1.2
      FNAB = fine needle aspiration biopsy.

       Rebiopsy of Selected Tumors

      Fifty-two of the 880 tumors in this series (5.9%) were biopsied more than once. Two of these 52 patients actually underwent a third FNAB of their tumor. The indication for the initial biopsy in this subset of patients was diagnostic in 37 (71.2%), confirmatory in 6 (11.5%), investigational in 3 (5.8%), and prognostic in 6 (11.5%). The indications for the second biopsy were diagnostic in 12 (23.1%) (all cases in which the initial FNAB had been inconclusive), confirmatory in 7 (13.5%) (most of which were melanocytic uveal tumors that had enlarged following an inconclusive initial FNAB and were rediagnosed as uveal melanomas), investigational in 3 (5.8%), and prognostic in 30 (57.7%). The 2 cases biopsied a third time included 1 very slowly enlarging amelanotic but presumably melanocytic choroidal tumor sampled 3 times in 10+ years (insufficiently cellular aspirates for cytopathologic classification obtained at each biopsy) and 1 small melanocytic choroidal tumor that was biopsied initially for diagnosis, was confirmed to be a uveal melanoma, was treated by plaque radiotherapy, relapsed locally following plaque therapy, was once again biopsied to see if the relapsed portion of the tumor had changed in cell type or GEP from the initial biopsy, was re-treated by a second radioactive plaque, eventually resulted in blindness and pain secondary to radiation-induced neovascular glaucoma and ischemic radiation retinopathy, was enucleated, and was biopsied again for investigational purposes.

       Survival and Follow-up Interval

      Overall survival (cumulative probability of survival over time) of patients following FNAB of an intraocular tumor is a function of the patient's age and general health prior to the biopsy, the type of tumor that prompted the biopsy, and the aggressiveness and metastatic potential of the tumor that prompted the biopsy. Mean post-FNAB follow-up of the entire group was 62.7 months (standard deviation [SD] = 59.8, 95% confidence interval [CI] = 58.7–66.6). Mean posttreatment follow-up or total follow-up of untreated patients was 53.7 months (SD = 58.7, 95% CI = 49.8–57.6). The mean survival time among patients in the diagnostic FNAB group was 268.8 months (standard error = 12.2, 95% CI = 244.7–293.0). The mean survival time among patients in the confirmatory FNAB group was 256.6 months (standard error = 20.5, 95% CI = 216.4–296.9). The mean survival time among patients in the investigational FNAB group was 206.9 months (standard error = 18.5, 95% CI = 170.6–243.2). The mean survival time among patients in the prognostic FNAB group was 131.0 months (standard error = 6.6, 95% CI = 118.0–144.0). The more relevant survival statistic in patients undergoing FNAB for a suspected malignant intraocular tumor is tumor-specific cumulative actuarial mortality. Obviously, this differs substantially with the tumor type. The only tumor type for which cytopathologic features of the tumor cells and GEP classification of those cells have been studied as potential prognostic factors is primary posterior uveal melanoma. Our results show that in 784 patients evaluated by FNAB for a posterior uveal tumor, the survival of patients submitted to a diagnostic FNAB was substantially better than that of the other indication category groups, especially the investigational FNAB category. This finding also reached statistical significance (P = .0001).

      Discussion

      FNAB of solid intraocular tumors is performed for many different reasons. In the past 3 decades several groups have published on their experience, technique, yield, and preparation techniques for the specimens.
      • Midena E.
      • Segato T.
      • Piermarocchi S.
      • Boccato P.
      Fine needle aspiration biopsy in ophthalmology.
      • Augsburger J.J.
      Fine·needle aspiration biopsy in the diagnosis of suspected intraocular cancer.
      • Onken M.D.
      • Worley L.A.
      • Char D.H.
      • et al.
      Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma.
      • Augsburger J.J.
      Fine needle aspiration biopsy of suspected metastatic cancers to the posterior uvea.
      • Cohen V.M.
      • Dinakaran S.
      • Parsons M.A.
      • Rennie I.G.
      Transvitreal fine needle aspiration biopsy: the influence of intraocular lesion size on diagnostic biopsy result.
      • Shields J.A.
      • Shields C.L.
      • Ehya H.
      • Eagle Jr., R.C.
      • De Potter P.
      Fine-needle aspiration biopsy of suspected intraocular tumors. The 1992 Urwick Lecture.
      More recently, FNAB has gained new interest with the use of uveal melanoma tumor tissue for prognostic testing.
      • Onken M.D.
      • Worley L.A.
      • Char D.H.
      • et al.
      Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma.
      • Correa Z.M.
      • Augsburger J.J.
      Independent prognostic significance of gene expression profile class and largest basal diameter of posterior uveal melanomas.
      However, there has been limited reporting of different indications for FNAB.
      • Shields J.A.
      • Shields C.L.
      • Ehya H.
      • Eagle Jr., R.C.
      • De Potter P.
      Fine-needle aspiration biopsy of suspected intraocular tumors. The 1992 Urwick Lecture.
      • Singh A.D.
      • Biscotti C.V.
      Fine needle aspiration biopsy of ophthalmic tumors.
      • Augsburger J.J.
      • Correa Z.M.
      Biopsy of iris tumors.
      • Singh A.D.
      • Medina C.A.
      • Singh N.
      • et al.
      Fine-needle aspiration biopsy of uveal melanoma: outcomes and complications.
      • Klufas M.A.
      • Itty S.
      • McCannel C.A.
      • et al.
      Variable results for uveal melanoma-specific gene expression profile prognostic test in choroidal metastasis.

       Why Biopsy Intraocular Tumors

      Performing FNAB of an intraocular tumor simply “because it is there” seems to be a futile exercise. It is the authors' belief that intraocular tumors should be biopsied only if there is a substantial likelihood of their being malignant. As shown by our results, clinical FNAB of nonmelanocytic intraocular tumors is rare unless patient management is likely to be influenced by this invasive testing.
      It has been routine to identify a principal indication category (and a secondary indication category as well, if one exists) for every FNAB performed since the start of our FNAB experience. It recently became apparent that evaluated tumors in the different FNAB indication categories differ substantially with regard to clinical diagnosis or differential diagnosis, size of the tumor, expected cellular yield, and testing intended to be performed on the obtained specimens, and it was determined that these and other factors were pertinent to the expected results of the biopsy. It has been also routine to perform multisite sampling of most tumors to lessen the likelihood of an insufficiently cellular aspirate for cytopathologic classification based on technical problems with the biopsy procedure; to increase the likelihood of obtaining sufficiently cellular specimens for both conventional chemical staining and immunocytochemical testing, if indicated; and to provide a specimen of melanocytic uveal tumors for prognostic GEP profile testing and classification.
      The available literature leads us to believe that some ocular tumor specialists currently do not submit specimens of melanocytic tumors they biopsy for both cytology and a prognostic cytogenetic test (GEP or other genetic/chromosomal test).
      • Shields C.L.
      • Ganguly A.
      • Bianciotto C.G.
      • Turaka K.
      • Tavallali A.
      • Shields J.A.
      Prognosis of uveal melanoma in 500 cases using genetic testing of fine-needle aspiration biopsy specimens.
      • Murray T.G.
      Uveal Melanoma Fine Needle Aspiration Biopsy (FNAB) for Molecular Genomic Classification: Evaluation of Transscleral Versus Transvitreal Biopsy.
      , Such personal practices assume that the clinical diagnosis of every intraocular tumor is accurate (or only infrequently erroneous) and rely on the results of cytogenetic testing alone. Also, this approach has led to circumstances such as performing GEP classification of metastatic cancers and interpreting class 1 GEP as evidence of the melanocytic nature of the sampled tumor.
      • Klufas M.A.
      • Itty S.
      • McCannel C.A.
      • et al.
      Variable results for uveal melanoma-specific gene expression profile prognostic test in choroidal metastasis.
      Many ocular tumor specialists also biopsy mostly larger tumors, and that may lead readers to believe it unsafe to perform FNAB on small melanocytic posterior uveal tumors suspected of being small uveal melanomas.
      • Cohen V.M.
      • Dinakaran S.
      • Parsons M.A.
      • Rennie I.G.
      Transvitreal fine needle aspiration biopsy: the influence of intraocular lesion size on diagnostic biopsy result.
      Others still do not employ FNAB or any other method of biopsy for any melanocytic tumors they treat. While there is no direct criticism to any personal practice pattern, our approach to patient management based on the FNAB results shown herein supports earlier treatment as well as delaying treatment for those patients that do not need it.
      Nevertheless, it is important to recognize that formulating a clinical diagnosis is always an exercise in probability and that the correct clinical diagnosis in any given case can never be regarded as completely certain; however, we regard only biopsies performed on lesions or tumors assigned a differential diagnosis clinically (ie, 2 or more diagnoses, both of which are regarded as reasonably likely) by an experienced clinician to be true diagnostic biopsies.
      • Augsburger J.J.
      • Correa Z.M.
      Biopsy of iris tumors.
      It is important to understand that diagnostic and confirmatory biopsies can potentially change the management of the patient. This study clearly shows that patient management changed in more than one-third of all clinical FNABs (Table 15). This result is clinically important for patient management and once again reinforces the importance of uniform use of FNAB as a diagnostic and confirmatory tool.

       The Issue of Insufficient Aspirate for Cytology Diagnosis

      After several years of experience with FNAB we realized 1 of the significant issues was the nondiagnostic (or insufficient) specimen.
      • Foulds W.S.
      The uses and limitations of intraocular biopsy.
      We also realized (as others did) that increasing the needle caliber had insignificant impact and increased the rate of complications and bleeding.
      • Midena E.
      • Segato T.
      • Piermarocchi S.
      • Boccato P.
      Fine needle aspiration biopsy in ophthalmology.
      • Augsburger J.J.
      • Shields J.A.
      Fine needle aspiration biopsy of solid intraocular tumors: indications, instrumentation and techniques.
      • Shields J.A.
      • Shields C.L.
      • Ehya H.
      • Eagle Jr., R.C.
      • De Potter P.
      Fine-needle aspiration biopsy of suspected intraocular tumors. The 1992 Urwick Lecture.
      • Singh A.D.
      • Medina C.A.
      • Singh N.
      • et al.
      Fine-needle aspiration biopsy of uveal melanoma: outcomes and complications.
      As we began to perform at least 2-site biopsies, it became clear that most tumors (but certainly not all) that yielded an insufficient specimen at 1 site also did so at other sampled sites.
      • Augsburger J.J.
      • Correa Z.M.
      • Trichopoulos N.
      Prognostic implications of cytopathologic classification of melanocytic uveal tumors evaluated by fine-needle aspiration biopsy.
      We subsequently hypothesized that an insufficiently cellular aspirate for cytopathologic classification of the tumor cells (a “quantity not sufficient” [QNS] result) from 2 or more sites was indicative of cohesiveness of the tumor cells and probable relative if not absolute benignity of that tumor.
      • Augsburger J.J.
      • Correa Z.M.
      • Trichopoulos N.
      Prognostic implications of cytopathologic classification of melanocytic uveal tumors evaluated by fine-needle aspiration biopsy.
      We evaluated the cumulative actuarial mortality owing to metastasis in patients with melanocytic uveal tumors evaluated by FNAB in 2 or more sites retrospectively and confirmed that a QNS result was almost always associated with long-term metastasis-free survival.
      • Augsburger J.J.
      • Correa Z.M.
      • Trichopoulos N.
      Prognostic implications of cytopathologic classification of melanocytic uveal tumors evaluated by fine-needle aspiration biopsy.
      Based on this evidence, we currently regard a QNS result at 2 or more tumor sites sampled by FNAB as an indicator of probable tumor benignity (nevus or “spindle A melanoma”). In our experience, diagnostic FNABs usually include smaller tumors and, for that reason, are the ones with highest insufficient aspirate for diagnosis.
      • Augsburger J.J.
      • Correa Z.M.
      • Schneider S.
      • et al.
      Diagnostic transvitreal fine-needle aspiration biopsy of small melanocytic choroidal tumors in nevus versus melanoma category.
      • Cohen V.M.
      • Dinakaran S.
      • Parsons M.A.
      • Rennie I.G.
      Transvitreal fine needle aspiration biopsy: the influence of intraocular lesion size on diagnostic biopsy result.
      An open discussion with patients prior to performing a biopsy about FNAB results and how they are applied to management decisions is an important part of clinical care.
      It is our belief that a biopsy performed on a small melanocytic uveal tumor that some experienced ocular oncologists would classify clinically as a “large melanocytic uveal nevus vs small uveal melanoma” (also termed “suspicious uveal nevus,” “indeterminate pigmented uveal tumor,” “nevoma,” or “borderline melanocytic uveal tumor” by some ocular tumor specialists) needs to first confirm (if possible) the melanocytic nature of the tumor and not just its GEP and chromosomal status.
      • Correa Z.M.
      • Augsburger J.J.
      Sufficiency of FNAB aspirates of posterior uveal melanoma for cytologic versus GEP classification in 159 patients, and relative prognostic significance of these classifications.
      • Shields C.L.
      • Ganguly A.
      • Bianciotto C.G.
      • Turaka K.
      • Tavallali A.
      • Shields J.A.
      Prognosis of uveal melanoma in 500 cases using genetic testing of fine-needle aspiration biopsy specimens.
      • Murray T.G.
      Uveal Melanoma Fine Needle Aspiration Biopsy (FNAB) for Molecular Genomic Classification: Evaluation of Transscleral Versus Transvitreal Biopsy.
      • Harbour J.W.
      • Chao D.L.
      A molecular revolution in uveal melanoma: implications for patient care and targeted therapy.
      Failure to do so may lead to mismanagement of patient and misuse of resources, since prognostic tests have a higher cost that may not be covered by insurance.
      • Klufas M.A.
      • Itty S.
      • McCannel C.A.
      • et al.
      Variable results for uveal melanoma-specific gene expression profile prognostic test in choroidal metastasis.
      Similarly to diagnostic biopsies, confirmatory FNABs are also performed in vivo prior to tumor or lesion treatment. The principal factor distinguishing confirmatory biopsies from diagnostic biopsies is that the treating physician previously assigns an unequivocal clinical diagnosis to the tumor or lesion in this category.

       Investigational Fine Needle Aspiration Biopsy

      Investigational FNABs warrant a special comment. Biopsies in this category can be performed for a variety of reasons, including providing experience to the ophthalmic surgeon with the selected instrumentation, routes, and techniques of biopsy and methods of handling and processing the obtained aspirates for pathologic study and permitting studies of cellular yield (ie, sufficiency or insufficiency of the aspirates for the intended testing), diagnostic accuracy of the cytopathologic classification (usually reported as sensitivity, specificity, and positive and negative predictive values), percentage of clinical biopsies that change patient management, frequency of various complications of clinical biopsies, histopathologic findings along the biopsy needle tract in enucleated eyes, and potential prognostic value of nonvalidated laboratory tests performed in a research (non-CLIA-approved) laboratory.
      • Augsburger J.J.
      • Shields J.A.
      Fine needle aspiration biopsy of solid intraocular tumors: indications, instrumentation and techniques.
      • Correa Z.M.
      • Augsburger J.J.
      Sufficiency of FNAB aspirates of posterior uveal melanoma for cytologic versus GEP classification in 159 patients, and relative prognostic significance of these classifications.
      In our opinion, all such studies require review and approval by an institutional review board and investigational patient consent.

       The Growing Role of Prognostic Fine Needle Aspiration Biopsy in Melanocytic Choroidal Tumors

      While prognostic FNABs should not change patient management, results should be applied for personalized clinical decision making, surveillance testing, and, it is hoped, more targeted therapeutic options for metastatic disease. Prognostic FNABs are mainly used for posterior uveal melanoma.
      • Harbour J.W.
      • Chao D.L.
      A molecular revolution in uveal melanoma: implications for patient care and targeted therapy.
      Novel techniques such as GEP and multiplex ligation-dependent probe amplification have improved upon the initial methods for genetic testing like karyotype analysis, fluorescence in situ hybridization, and comparative genomic hybridization.
      • Correa Z.M.
      Assessing prognosis in uveal melanoma.
      However, only GEP testing has been prospectively validated in a clinical setting and some of the reported FNABs in this study were reported among the validation cohort.
      • Onken M.D.
      • Worley L.A.
      • Char D.H.
      • et al.
      Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma.
      We specifically analyzed the cumulative probability of survival of patients with posterior uveal melanoma by FNAB indication category. The main reason to further investigate this subset of patients was to explore the potential impact of insufficient aspirates for cytologic diagnosis. Interestingly, the diagnostic FNAB group showed the best survival outcome while yielding higher rates of insufficient aspirate of cytology diagnosis. This finding serves as evidence of our previous report that an insufficient aspirate is an indicator of a favorable survival outcome
      • Augsburger J.J.
      • Correa Z.M.
      • Trichopoulos N.
      Prognostic implications of cytopathologic classification of melanocytic uveal tumors evaluated by fine-needle aspiration biopsy.
      and that diagnostic FNAB does not worsen patient prognosis. This curve also illustrates that prognostic and confirmatory biopsies have similar survival, showing again that there is no downside to confirming a diagnosis prior to deciding management.

       Limitations and Further Work

      One limitation of this work is its retrospective nature, and it can be argued that through the years our indication criterion has changed. Such assumption is indeed appropriate, as we began to perform prognostic biopsies in 2007, and such category has been created since genomic testing has become available. At the same time, we (and other groups) did not biopsy posterior uveal melanomas routinely prior to eye-sparing treatments. To address this limitation, the authors chose to personally review and discuss any controversial cases prior to including them in the study to ensure that the information was complete and there was agreement regarding indication for FNAB. Further, we agreed that in any cases with multiple indications, such as diagnostic/prognostic and confirmatory/prognostic, the primary indication would be the main reason for the biopsy to be performed at a particular time (ie, as a separate procedure or prior to treatment).
      Another limitation is the potential bias in the selection of study patients associated with a retrospective study. However, because all tumors meeting our inclusion criteria were entered in this study we expect that the full spectrum of FNABs in posterior segment tumors was shown herein.

       Conclusions

      Because of the substantial differences observed among FNAB cases performed for different indications, it appears that reported results of FNABs or other biopsy methods should explicitly identify the category of the biopsy in all of their cases and thus improve the readership's understanding of the procedure and its applications.
      Funding/Support: This work was supported in part by the James J. Augsburger Ocular Oncology Fund , University of Cincinnati College of Medicine , Cincinnati, Ohio, USA. Financial Disclosures: Zélia Maria Corrêa: Consultant, Castle Biosciences, Inc. James Jay Augsburger has no financial disclosures. Both authors attest that they meet the current ICMJE criteria for authorship.

      Supplemental Data

      References

        • Frable W.J.
        Fine-needle aspiration biopsy: a review.
        Hum Pathol. 1983; 14: 9-28
        • Peyman G.A.
        • Fishman G.A.
        • Sanders D.R.
        • Apple D.J.
        • Vlchek J.K.
        Biopsy of human scleral-chorioretinal tissue.
        Invest Ophthalmol. 1975; 14: 707-710
        • Foulds W.S.
        The uses and limitations of intraocular biopsy.
        Eye (Lond). 1992; 6: 11-27
        • Midena E.
        • Segato T.
        • Piermarocchi S.
        • Boccato P.
        Fine needle aspiration biopsy in ophthalmology.
        Surv Ophthalmol. 1985; 29: 410-422
        • Martin H.E.
        • Ellis E.B.
        Biopsy by needle puncture and aspiration.
        Ann Surg. 1930; 92: 169-181
        • Stewart F.W.
        The diagnosis of tumors by aspiration.
        Am J Pathol. 1933; 9: 801-812
        • Franzen S.
        • Zajicek J.
        Aspiration biopsy in diagnosis of palpable lesions of the breast. Critical review of 3479 consecutive biopsies.
        Acta Radiol Ther Phys Biol. 1968; 7: 241-262
        • Söderström N.
        Fine Needle Aspiration Biopsy.
        Almqvist and Wiksell/Gebers, Stockholm1966
        • Jakobiec F.A.
        • Chattock A.
        The role of cytology and needle biopsies in the diagnosis of ophthalmic tumors and simulating conditions.
        in: Jakobiec F.A. Ocular and Adnexal Tumors. Aesculapius Publishing Co, Birmingham1978: 341-358
        • Jakobiec F.A.
        • Coleman D.J.
        • Chattock A.
        • Smith M.
        Ultrasonically guided needle biopsy and cytologic diagnosis of solid intraocular tumors.
        Ophthalmology. 1979; 86: 1662-1681
        • Augsburger J.J.
        • Shields J.A.
        Fine needle aspiration biopsy of solid intraocular tumors.
        Trans Pa Acad Ophthalmol Otolaryngol. 1983; 36: 169-172
        • Augsburger J.J.
        Fine·needle aspiration biopsy in the diagnosis of suspected intraocular cancer.
        in: Jakobiec F.A. Advanced Techniques in Ocular Surgery. WB Saunders, Philadelphia1984: 491-521
        • Augsburger J.J.
        • Shields J.A.
        Fine needle aspiration biopsy of solid intraocular tumors: indications, instrumentation and techniques.
        Ophthalmic Surg. 1984; 15: 34-40
        • Onken M.D.
        • Worley L.A.
        • Char D.H.
        • et al.
        Collaborative Ocular Oncology Group report number 1: prospective validation of a multi-gene prognostic assay in uveal melanoma.
        Ophthalmology. 2012; 119: 1596-1603
        • Schopper V.J.
        • Correa Z.M.
        Clinical application of genetic testing for posterior uveal melanoma.
        Int J Retina Vitreous. 2016; 2: 4
        • Eide N.
        • Walaas L.
        Fine-needle aspiration biopsy and other biopsies in suspected intraocular malignant disease: a review.
        Acta Ophthalmol. 2009; 87: 588-601
        • Augsburger J.J.
        Fine needle aspiration biopsy of suspected metastatic cancers to the posterior uvea.
        Trans Am Ophthalmol Soc. 1988; 86: 499-560
        • Augsburger J.J.
        • Correa Z.M.
        • Schneider S.
        • et al.
        Diagnostic transvitreal fine-needle aspiration biopsy of small melanocytic choroidal tumors in nevus versus melanoma category.
        Trans Am Ophthalmol Soc. 2002; 100 (discussion 232–234): 225-232
        • Eide N.
        • Syrdalen P.
        • Walaas L.
        • Hagmar B.
        Fine needle aspiration biopsy in selecting treatment for inconclusive intraocular disease.
        Acta Ophthalmol Scand. 1999; 77: 448-452
        • Correa Z.M.
        • Augsburger J.J.
        Sufficiency of FNAB aspirates of posterior uveal melanoma for cytologic versus GEP classification in 159 patients, and relative prognostic significance of these classifications.
        Graefes Arch Clin Exp Ophthalmol. 2014; 252: 131-135
        • Augsburger J.J.
        • Correa Z.M.
        • Trichopoulos N.
        Prognostic implications of cytopathologic classification of melanocytic uveal tumors evaluated by fine-needle aspiration biopsy.
        Arq Bras Oftalmol. 2013; 76: 72-79
        • Larsen A.C.
        • Holst L.
        • Kaczkowski B.
        • et al.
        MicroRNA expression analysis and Multiplex ligation-dependent probe amplification in metastatic and non-metastatic uveal melanoma.
        Acta Ophthalmol. 2014; 92: 541-549
        • Staby K.M.
        • Gravdal K.
        • Mork S.J.
        • et al.
        Prognostic impact of chromosomal aberrations and GNAQ, GNA11 and BAP1 mutations in uveal melanoma.
        Acta Ophthalmol. 2018; 96: 31-38
        • Harbour J.W.
        A prognostic test to predict the risk of metastasis in uveal melanoma based on a 15-gene expression profile.
        Methods Mol Biol. 2014; 1102: 427-440
        • Harbour J.W.
        • Chen R.
        The DecisionDx-UM gene expression profile test provides risk stratification and individualized patient care in uveal melanoma.
        PLoS Curr. 2013; 5
        • Folberg R.
        • Augsburger J.J.
        • Gamel J.W.
        • Shields J.A.
        • Lang W.R.
        Fine-needle aspirates of uveal melanomas and prognosis.
        Am J Ophthalmol. 1985; 100: 654-657
        • Cohen V.M.
        • Dinakaran S.
        • Parsons M.A.
        • Rennie I.G.
        Transvitreal fine needle aspiration biopsy: the influence of intraocular lesion size on diagnostic biopsy result.
        Eye (Lond). 2001; 15: 143-147
        • Shields J.A.
        • Shields C.L.
        • Ehya H.
        • Eagle Jr., R.C.
        • De Potter P.
        Fine-needle aspiration biopsy of suspected intraocular tumors. The 1992 Urwick Lecture.
        Ophthalmology. 1993; 100: 1677-1684
        • Correa Z.M.
        • Augsburger J.J.
        Independent prognostic significance of gene expression profile class and largest basal diameter of posterior uveal melanomas.
        Am J Ophthalmol. 2016; 162: 20-27.e21
        • Singh A.D.
        • Biscotti C.V.
        Fine needle aspiration biopsy of ophthalmic tumors.
        Saudi J Ophthalmol. 2012; 26: 117-123
        • Augsburger J.J.
        • Correa Z.M.
        Biopsy of iris tumors.
        Am J Ophthalmol. 2011; 152: 720-722.e722
        • Singh A.D.
        • Medina C.A.
        • Singh N.
        • et al.
        Fine-needle aspiration biopsy of uveal melanoma: outcomes and complications.
        Br J Ophthalmol. 2016; 100: 456-462
        • Klufas M.A.
        • Itty S.
        • McCannel C.A.
        • et al.
        Variable results for uveal melanoma-specific gene expression profile prognostic test in choroidal metastasis.
        JAMA Ophthalmol. 2015; 133: 1073-1076
        • Shields C.L.
        • Ganguly A.
        • Bianciotto C.G.
        • Turaka K.
        • Tavallali A.
        • Shields J.A.
        Prognosis of uveal melanoma in 500 cases using genetic testing of fine-needle aspiration biopsy specimens.
        Ophthalmology. 2011; 118: 396-401
        • Murray T.G.
        Uveal Melanoma Fine Needle Aspiration Biopsy (FNAB) for Molecular Genomic Classification: Evaluation of Transscleral Versus Transvitreal Biopsy.
        Journal of VitreoRetinal Diseases. 2018; 2: 208-212
        • Harbour J.W.
        • Chao D.L.
        A molecular revolution in uveal melanoma: implications for patient care and targeted therapy.
        Ophthalmology. 2014; 121: 1281-1288
        • Correa Z.M.
        Assessing prognosis in uveal melanoma.
        Cancer Control. 2016; 23: 93-98