Purpose
To report the results and complications of an aspiration cutter-assisted anterior segment tumor biopsy performed under viscoelastic.
Design
Retrospective, consecutive, interventional case series.
Methods
Fifty-five patients with anterior segment tumors underwent aspiration cutter-assisted biopsy at a single center. Fifty-six biopsies were performed, because 1 eye underwent biopsy twice. Indications for biopsy included: suspected anterior segment malignancy, patient’s desire for pathologic confirmation before treatment, atypical tumor, and genetic tumor analysis. The main evaluated outcomes were the perioperative visual acuity, the biopsy technique, and related complications.
Results
The initial sample comprised 55 consecutively biopsied patients, from which a minimum 1-month follow-up existed for 52 biopsy samples. The median total follow-up was 30.4 months (range, 1 to 190 months). The mean initial visual acuity was 20/50 before and 20/50 1 month after biopsy. No patient lost vision. Most cases were suspected iris melanoma (n = 39/56; 69.6%), followed by suspicious iris nevi (n = 4/56; 7.1%) and melanocytoma (n = 4/56; 7.1%). Seven (n = 7/52; 13.4%) wounds required a single 10-0 nylon suture to achieve negative Seidel test results at the corneal entry site. Postoperative surgical findings included transiently increased intraocular pressure (n = 6/52; 11.5%), 1 hyphema, 1 flare, and 1 persistent pupillary defect. All but the pupillary defect resolved within 4 weeks of the biopsy procedure. There were no secondary infections or cataracts.
Conclusions
This study suggests that small-incision, aspiration cutter-assisted anterior segment biopsy seems to be a safe and effective procedure. No short- or long-term complications that would prevent its use were noted.
Biopsy techniques are used in all fields of medicine. Nearly all parts of the body are accessible through core biopsy or fine-needle aspiration biopsy (FNAB) as well as open biopsy. For example, in 2007, the Swedish Cancer Registry reported that nearly 99% of primary malignant extraocular tumors were diagnosed by a cytologist or histopathologist. In contrast, ophthalmologists have been reluctant to biopsy intraocular tumors because of concerns about tumor seeding and vision preservation. Therefore, the diagnosis of most iris tumors has been based on clinical and ultrasound characteristics as well as evidence of tumor growth.
In cases where clinical diagnostic techniques are inconclusive (particularly when there is documented tumor growth), anterior segment biopsy can be performed. In general, surgical iridectomy, iridocyclectomy, and choroidal biopsy will provide relatively large histopathologic specimens. However, these invasive techniques require relatively large incisions, tissue rearrangement, and loss of tissue function. Large incisions also increase risk of infection and intraocular complications.
In contrast, minimally invasive techniques used for anterior segment biopsy include FNAB and aspiration cutter-assisted techniques. In general, FNAB is used to obtain cytologic specimens, and aspiration cutter-assisted techniques can obtain both cells and tissue. Both FNAB and aspiration cutter techniques share the advantage of smaller entry sites and less normal tissue damage and risk to vision. In 2005, we first reported the Finger iridectomy technique. In that study, a 25-gauge aspiration cutter was used to obtain biopsy samples from 10 patients. Herein, we present a much larger series with longer-term follow-up and compare our results with those published on FNAB.
Methods
A retrospective, interventional, noncomparative case series study was performed. All patients were evaluated by history, clinical examination, and pathologic analysis of retrieved tumor tissue. In total, 55 patients underwent an aspiration cutter-assisted biopsy for a tumor in their anterior segment. One tumor was biopsied twice because of persistent growth of an iridociliary nevus. Therefore, we report on 56 biopsies. Four of these biopsies were immediately followed by an additional procedure during the same operative session and are excluded from our analysis of postoperative complications. This leaves 52 biopsies in the complication study group.
Entry criteria included patients who were clinically diagnosed with an iris or iridociliary tumor and who were determined to require biopsy. Our general indications for biopsy included: atypical tumor, invasive conjunctival tumor to determine intraocular extension, and metastatic tumor with no known primary, the patient requiring pathologic proof of malignancy to proceed with treatment, or genetic tissue analysis and documented growth. Atypical tumors were defined as those without significant clinical or ultrasonographic characteristics pathognomonic of the diagnosis.
Patient assessment included age, gender, and medical history of ocular or systemic malignancy. Tumor-specific initial examination parameters included: best-corrected visual acuity; pupillary and ocular motor evaluation; and slit-lamp examination of the eyelids, conjunctival surfaces, and anterior segment, noting the presence of tumor, enlarged episcleral vessels, pigment dispersion, corectopia, ectropion uveae, hemorrhage, and glaucoma. Tumor assessment included: morphologic features, location, intrinsic vascularity, height, and maximal basal diameter (eg, high-frequency ultrasound).
In this study, the aspiration cutter biopsy procedure was evaluated for diagnostic yield, postoperative visual acuity, and ophthalmic complications. Each eye was assessed at the time of the biopsy and at the 1-week and 1-month follow-up. More long-term follow-up was performed; however, outcome data typically were complicated by the necessity for subsequent treatment.
Aspiration Cutter Technique
The technique was described previously. Briefly, preoperative pilocarpine 1% and acetazolamide 500 mg were given 1 hour before surgery. Then, the procedures were performed in the operating theater and under a operating microscope. A 0.3 forceps stabilized the eye, and an inked 25-gauge microvitreoretinal blade or trocar was used to create a stab incision through clear juxtalimbal cornea, into the anterior chamber. The incision site typically is on the side of the tumor to avoid the pupil and risking damaging the natural lens. Inking the blade or trocar allowed for staining and subsequent recognition of the clear corneal incision site. When necessary, acetylcholine chloride 10 mg/mL (Miochol-E; CibaVision-Novartis, Basel, Switzerland) was introduced into the anterior chamber to stabilize the iris, to induce miosis and thus to protect the natural lens. Then, sodium hyaluronate 1% (Healon; Advanced Medical Optics, Santa Ana, California, USA) was introduced into the anterior chamber to maintain anterior chamber depth during biopsy and (on occasion) to position the iris for biopsy (away from the natural lens). In contrast to aqueous infusions, viscoelastic stabilized the anterior segment and did not wash out potential specimen.
All specimens were obtained using a 25-gauge aspiration cutter inserted into the anterior chamber through the marked clear corneal opening. As needed, the aspiration port could be rotated as to be partially or completely occluded by the tumor or iris. Complete occlusion increases the yield, but risks full-thickness iridotomy. Our typical initial settings (Accurus; Alcon, Fort Worth, Texas, USA) were 300 mm Hg of suction and a cutting rate of 600 cuts per minute. These settings then were adjusted to maximize the efficiency of the process under direct visualization, while trying to keep the cut rate as low as possible. Two to 3 biopsy passes usually were required to obtain a diagnostic specimen. Each time, the aspiration cutter was removed from the eye and its contents aspirated into a 3-mL syringe. The barrel of the syringe was examined for tissue. Fresh specimens were analyzed by our ophthalmic pathologists for adequacy. The diagnosis was made by cytologic analysis and later was confirmed by histopathologic and immunohistochemistry analysis (where possible).
Most corneal entry points proved to be self-sealing. If Seidel test results were positive, a 10-0 nylon suture was placed and the wound was retested. At the end of surgery, an antibiotic steroid solution was injected beneath the conjunctiva, 1 drop of timolol maleate 0.5% and antibiotic steroid ointment was placed on the eye, and then it was patched and shielded. Patients were discharged (on the same day) with topical steroid, antibiotic, and ocular hypotensive agents. All eye pressures were checked the following day.
Results
Herein, we report on 55 patients (27 females) with a median follow-up of 30.4 months (range, 1 to 190 months). The mean age was 60 years (median, 61 years; range, 33 to 88 years). Although most patients had no previous history of cancer (n = 45/55; 81.8%), 4 (7.3%) had a history of breast cancer and 1 (1.8%) (each) had a history of cutaneous melanoma, prostate cancer, endometrial cancer, kidney cancer, lung cancer, and lymphoma.
Tumor Characteristics and Staging before Biopsy
The great majority of tumors were nodular and in the iris (n = 29/55; 52.7%) or iridociliary are (n = 23/55; 41.8%). Tumors were centered in the supranasal quadrant (n = 7: 12.7%), inferonasal quadrant (n = 23; 41.8%), inferotemporal quadrant (n = 16; 29.1%), or supertemporal quadrant (n = 9; 16.4%). Of note, 71% were in the inferior quadrant. There was 1 multifocal (1.8%) and 2 diffuse (3.6%) iris lesions. There were only 2 cases of pigment dispersion (n = 2/55; 3.6%). Intrinsic tumor vascularity (n = 15/55; 27.3%) and corectopia (n = 15/55; 27.3%) were noted in slightly less than one third of cases. Mean tumor thickness (before biopsy) was 1.8 mm (median, 1.5 mm; range, 0.7 to 6.0 mm), and mean tumor base was 5.2 mm (median, 4.5 mm; range, 0.7 to 19.0 mm). Twenty of the 39 tumors biopsy proven to be uveal melanomas were staged TI, with the rest being staged TIIA (according to the seventh edition of the AJCC classification).
Visual Acuity and Biopsy Results
Although the mean initial visual acuity was 20/50, the majority could see better (median, 20/20; range, 20/20 to 5/200). One month after biopsy, all visual acuities returned to their preoperative levels. A cytopathologic diagnosis was obtained in 98.2% (n = 55/56) of cases, including malignant melanoma of the iris (n = 39/56;69.6%), iris nevus (n = 4/56; 7.1%), melanocytoma (n = 4/56; 7.1%), lymphoma (n = 2/56; 3.6%), and iris epithelial cyst (n = 2/56; 3.6%). Biopsy samples showed adenocarcinoma of the ciliary body (n = 1/56; 1.8%), an invasive squamous cell carcinoma (n = 1/56; 1.8%), a metastatic endometrial carcinoma (n = 1/56; 1.8%), and normal iris stroma (n =1/56; 1.8%). In 1 case, the biopsy sample was inconclusive (showing nonspecific chronic inflammation), prompting a second open biopsy at the same session. However, a retrospective evaluation of the aspiration cutter biopsy specimen found it consistent with the established diagnosis of sarcoid granuloma.
Postoperative Course and Complications
Rates of complications were studied in 52 biopsies. Excluded were 4 cases in which the biopsy concurred with other surgical procedures. For example, there were 2 cases of enucleation directly after pathologic confirmation of malignant melanoma, 1 case in which surgical iridectomy was performed only to verify a melanocytoma, and 1 case of an open biopsy that revealed a sarcoid granuloma after the primary aspiration cutter biopsy had yielded nonspecific inflammatory cells.
Thus, complications related to aspiration cutter biopsy included 5 (n = 5/52; 9.6%) cases of transient increased intraocular pressure (thought to be related to retained viscoelastic) lasting no more than 1 week ( Table ). One patient was found to have an iridociliary melanoma in whom neovascular glaucoma developed on the first day after the aspiration cutter iris biopsy. The glaucoma was controlled successfully with intravitreal bevacizumab (Genentech/Roche; South San Francisco, California, USA), followed by palladium-103 brachytherapy with adjuvant cryotherapy. In 1 case, the intraocular pressure remained increased for 1 month, related to a postoperative hyphema. Biopsy of 1 iris melanocytoma left a small and asymptomatic pupillary defect. Minimal (n = 5/52; 9.6%) and mild (n = 1/52; 1.9%) flare, defined as 1 and 2 crosses in the Tyndall scale, respectively, were noted in 6 asymptomatic patients. The presence of red blood cells in the anterior chamber was detected in 8 cases (n = 8/52; 15.3%), only seen during the first week of biopsy (without secondary glaucoma). There were no cases of secondary tumor formation at the wound or within the anterior segment. Finally, although not considered a complication, there were 7 intraoperative Seidel-positive corneal wounds (n = 7/52; 13.4%), each treated with a single 10-0 nylon suture ( Table ).
