Figure 1. A, Gross photograph of Case 1. The posterior segment contains a chronic hemorrhage and necrotic tumor with calcific foci. B, Histopathology of case 1. A totally necrotic tumor rests on inner surface of the atrophic retina (hematoxylin-eosin [H-E], original magnification ×10). C, Gross photograph of Case 3. The posteriorly detached vitreous contains seeds of viable tumor. Extensive chorioretinal atrophy is present. D, Histopathology of case 3. Seeds of poorly differentiated retinoblastoma are seen in the vitreous overly ciliary body (H-E, original magnification ×25). E, Gross photograph of Case 6. A residual tumor fills half of the posterior segment. F, Histopathology of case 6. Most of the residual tumor is necrotic (H-E, original magnification ×5). A few sleeves and cuffs of viable cells persist within necrotic retinoblastoma. G, Gross photograph of Case 5. A viable retinoblastoma fills most of the posterior segment. H, Histopathology of case 5. A viable retinoblastoma invades the optic nerve, reaching the interior of the lamina cribrosa (H-E, original magnification ×50).
Figure 2. A, Fundus photograph of case 7. Severe chorioretinal atrophy involves the posterior pole. A few foci of viable retinoblastoma are evident in the photograph. B, Histopathology of case 7. The severely atrophic posterior retina shows total loss of the outer nuclear layer and photoreceptors. The retinal pigment epithelium is absent, and the choroid is markedly atrophic (hematoxylin-eosin [H-E], original magnification ×100). C, Histopathology of case 1. The severely atrophic posterior retina rests on fibrotic choroid that shows pigment dispersion indicative of prior necrosis. The retinal pigment epithelium is absent. D, Histopathology of case 1. The orbital vessel contains calcified thrombus (H-E, original magnification ×100). E, Histopathology of case 6. Calcified thrombus fills the lumen of the orbital vessel (H-E, original magnification ×25). F, Histopathology of case 6. Intraluminal calcium is seen at higher magnification (H-E, original magnification ×60). G, Histopathology of case 2. The arrow points to a foreign body in the lumen of the central retinal artery (H-E, original magnification ×50). H, Histopathology of case 2. A foreign body giant cell abuts the annular foreign body, consistent with cellulose fiber (H-E, original magnification ×250). I, Histopathology of case 2. The ciliary artery contains foreign material of unknown composition and foreign body giant cells (H-E, original magnification ×100). J, Histopathology of case 2. The foreign material is birefringent (H-E with crossed polarizers, original magnification ×100). K, Histopathology of case 2. A giant cell envelops the annular cellulose fiber next to the choroidal vessels (H-E, original magnification ×400). L, Histopathology of case 2. The cellulose fiber is birefringent. M, Histopathology of case 6. Thrombus within the orbital artery contains cross-sectioned birefringent fragments of synthetic fiber that incited granulomatous inflammation. N, Histopathology of case 6. A longitudinal segment of birefringent synthetic fiber is seen within lumen.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Eagle RC, Shields CL, Bianciotto C, Jabbour P, Shields JA. Histopathologic Observations After Intra-arterial Chemotherapy for Retinoblastoma. Arch Ophthalmol. 2011;129(11):1416–1421. doi:10.1001/archophthalmol.2011.223
Author Affiliations: Ophthalmic Pathology Laboratory (Dr Eagle) and Ocular Oncology Service (Drs C. L. Shields and J. A. Shields), Wills Eye Institute, and the Department of Endovascular Neurosurgery (Dr Jabbour), Thomas Jefferson University, Philadelphia, Pennsylvania.
Objective To describe histopathologic observations in eyes enucleated after intra-arterial chemotherapy (IAC) for retinoblastoma (Rb).
Methods Retrospective histopathologic analysis of 8 eyes.
Results The eyes were enucleated for tumor viability (n = 4), neovascular glaucoma (n = 2), anaphylactic reaction from IAC (n = 1), and persistent retinal detachment with poor visualization of the tumor (n = 1). Of the 2 eyes judged clinically with complete tumor regression and the 5 with viable tumor, the findings were confirmed on histopathology. The Rb response ranged from minimal (n = 1) to moderate (n = 1) to extensive (n = 4) to complete regression (n = 2). Viable vitreous seeds (n = 4 eyes), invasion into the optic nerve (n = 3), reaching the lamina cribrosa in 2 cases, and invasion into the choroid (n = 1) were observed. Histopathologic evidence of ischemic atrophy involving the outer retina and choroid was found in 4 eyes. One eye treated at another center with IAC and enucleated by our team for recurrence was observed to have extensive choroidal and outer retinal atrophy. This case showed orbital vascular occlusion and subendothelial smooth muscle hyperplasia. Intravascular birefringent foreign material was observed in 5 cases within occluded vessels, stimulating a granulomatous inflammatory response. The foreign material comprised cellulose fibers (n = 3), synthetic fabric fibers (n = 1), or unknown composition (n = 2). Thrombosed blood vessels were identified in 5 eyes and involved ciliary arteries in the retrobulbar orbit (n = 5), scleral emissarial canals (n = 1), small choroidal vessels (n = 1), and central retinal artery (n = 1).
Conclusion Retinoblastoma can be controlled with IAC, but histopathology of enucleated eyes reveals that ocular complications including thromboembolic events can occur.
Intra-arterial chemotherapy (IAC) is a new and somewhat controversial therapy for retinoblastoma (Rb). This therapy involves intravascular cannulation of the ophthalmic artery with direct delivery of chemotherapy to the eye and periocular region. The main goal of this approach is to provide sufficient chemotherapy to eradicate the Rb and avoid the toxicities of systemic chemotherapy. Initially developed by Kaneko and colleagues1-3 in Japan, the treatment has been studied by Abramson and colleagues4-6 in New York City and Shields and colleagues7-11 in Philadelphia, as well as several other centers in the United States and abroad.12,13
There have been relatively few reports on the histopathologic findings in eyes with Rb enucleated following IAC. In a preliminary study, histopathologic examination was performed on 2 eyes enucleated for chronic retinal detachment after IAC for Rb.4 The pathology discussion briefly remarks only that “pathological analysis demonstrated no viable tumor.” There were no comments on the remaining findings within the globes. In 2010, Vajzovic and colleagues13 reported the clinicopathologic findings of 3 eyes of 3 patients diagnosed with advanced Rb whose treatment had included IAC in addition to other therapies including systemic and periocular chemotherapy. All 3 eyes contained viable Rb. Two of the three eyes had optic nerve invasion by viable tumor that reached the lamina cribrosa in 1 eye, and 1 eye had nonmassive choroidal invasion. The authors of these articles made no mention of vascular occlusion or the presence of intravascular birefringent foreign material. Herein, we describe the histopathologic observations of 8 eyes that had been treated with IAC for Rb.
All specimens were submitted to the Wills Eye Institute Ophthalmic Pathology Laboratory following enucleation at Wills Eye Institute. Seven of the eight eyes had been treated by the Ocular Oncology Service at the Wills Eye Institute in conjunction with the Department of Endovascular Neurosurgery at Thomas Jefferson University Hospital under a protocol approved by the Wills Eye Institute institutional review board. The remaining eye was enucleated at Wills following treatment at a center in New York. Patient characteristics, treatment parameters, and other pertinent data are shown in Table 1.
The enucleated eyes were fixed in neutral buffered formalin and processed routinely for light microscopy using the same sectioning protocol as the Children's Oncology Group14 clinical trial investigating chemotherapy in eyes with high-risk histopathologic features (ARET 0332). Microscopic sections examined included 2 levels cut from the main pupil-optic nerve segment through the optic nerve head, multiple sections from both calottes, and a transverse section of the optic nerve from the surgical margin. Additional sections of orbital tissue were prepared and examined in a few instances. Sections were stained with hematoxylin-eosin and periodic acid–Schiff. Polarization microscopy was used to detect birefringent foreign material. All eyes were dissected and examined histopathologically by an ophthalmic pathologist (R.C.E.) with extensive experience in the histopathologic evaluation and diagnosis of Rb.
Indications for enucleation included poor tumor response to therapy (n = 2), recurrent vitreous seeding (n = 2), termination of intra-arterial therapy necessitated by an intraoperative anaphylactic reaction (n = 1), neovascular glaucoma (n = 2), and persistent retinal detachment that precluded view of the fundus (n = 1). Most eyes in this study initially harbored advanced Rb for which the Oncology Service at Wills Eye Institute had advised enucleation, but the parents insisted on IAC. In 2 instances, clinical examination prior to enucleation revealed evidence of vascular occlusion and ischemic damage. One patient was found to have ophthalmic artery occlusion at the time of an attempted second treatment. The second patient, who was advised initially by our group to have enucleation, had 6 cycles of IAC and plaque brachytherapy elsewhere and was found to have extensive atrophy and nonperfusion of the posterior retina and choroid as well as massive recurrent tumor.
Histopathologic findings are summarized in Table 2 and Table 3 and illustrated in Figure 1 and Figure 2. Microscopic examination disclosed varying degrees of tumor response to therapy (Figure 1). No viable tumor or seeding was found in 2 eyes that had been enucleated for neovascular glaucoma (Figure 1, A and B). Both of these eyes contained regression scars comprising gliosis and characteristic foci of calcified tumor cells consistent with a complete therapeutic response. Four other eyes enucleated for residual or recurrent tumor also contained tumor regression scars as well as viable tumor, which typically manifested as vitreous seeds (Figure 1, C and D). In one instance, the Rb appeared totally viable with little or no evidence of treatment response (Figure 1G). That tumor was well differentiated and contained numerous rosettes and areas of photoreceptor differentiation. Another eye, which was enucleated because the tumor did not appear to respond clinically, contained a large, extensively necrotic Rb with relatively small foci of residual viable tumor consistent with a partial treatment response (Figure 1, E and F).
Optic nerve invasion by viable tumor was present in 3 eyes. The optic nerve invasion was prelaminar in 1 eye. In the remaining 2 eyes, viable tumor was observed within the lamina cribrosa (Figure 1H). There was no histopathologic evidence of retrolaminar invasion. A microscopic focus of choroidal invasion by viable tumor was found in 1 eye. There was histopathologic evidence of severe outer retinal and choroidal atrophy secondary to ischemia in 4 eyes (Figure 2, A-C). Dilated retrobulbar arteries that contained large calcified thrombi and orbital fat necrosis were found in 1 case that was noted to have an ophthalmic artery occlusion clinically (Figure 2D). In a second case in which 6 cycles of IAC was administered at a center in New York and that had severe posterior retinal and choroidal atrophy documented fluorangiographically by our team, a long segment of posterior choroid and corresponding outer retinal layers were markedly atrophic (Figure 2, A and B). In that case, orbital arteries showed luminal compromise secondary to subendothelial smooth muscle proliferation. That eye also contained a small intraocular focus of birefringent foreign material surrounded by granulomatous inflammation in an area of regressed tumor. Two cases had large calcified thrombi within dilated orbital arteries (Figure 2, D, E, and F). Both cases also had orbital fat necrosis.
An unusual unexpected finding was the presence of birefringent foreign material within the thrombosed lumina of vessels in 5 of the 8 eyes examined, including the single case in which chemotherapy was delivered elsewhere (Figure 2, G-N). Although the particles of foreign material were found most often within ciliary arteries in the retrobulbar orbit or intrascleral emissarial canals (Figure 2, I, J, M, and N), identical particles of foreign material also were detected within small intrachoroidal vessels (Figure 2, K and L), central retinal vessels in the optic nerve head (Figure 2, G and H), or within an area of regressed intraocular tumor. In 3 cases, the size and annular configuration of the intraluminal birefringent foreign material was consistent with cellulose or cotton fibers (Figure 2, G, H, K, and L). In another case, intraluminal foreign bodies in orbital vessels appeared to be synthetic fabric fibers (Figure 2, M and N). They were intensely birefringent solid cylinders that contained characteristic particles of delustering agent and knife chatter marks on their sectioned ends. The synthetic fibers were situated in the thrombosed lumen of an orbital arteriole proximal to a dilated area filled with a large calcific thrombus directly behind the eye. In 2 instances, the identity of prominent deposits of birefringent material in posterior ciliary arteries was uncertain (Figures 2, I and J)
Our histopathologic observations of eyes with Rb enucleated after IAC show a variable response of the tumor to therapy including total regression in some cases and varying degrees of response in others. In eyes in which there was complete clinical response but enucleation was performed for neovascular glaucoma or vitreous hemorrhage, histopathology confirmed complete tumor response with no viable tumor. In such cases, IAC was successful in eradication of Rb. In cases in which enucleation was performed for incomplete tumor response clinically, histopathology confirmed viable Rb. Viable vitreous seeds were the most common indication for enucleation and were found despite complete regression of the primary tumor.
The most striking finding on histopathology was evidence of retinal and choroidal atrophy induced by vascular compromise. Histopathology disclosed areas of choroidal and outer retinal atrophy consistent with treatment-induced ischemia in 4 eyes. In 1 eye that was treated at another center and referred to our center for our opinion, we noted massive tumor recurrence but additionally were impressed with extensive choroidal atrophy clinically. Histopathology of that case showed profound choroidal vascular compromise and smooth muscle proliferation in the walls of orbital arterioles. Munier and colleagues recently have recognized similar findings of sectoral choroidal occlusive vasculopathy leading to chorioretinal atrophy following IAC for Rb in 2 of 13 eyes and additional retinal arteriolar emboli in 1 eye.15
The most disturbing finding was the presence of intravascular foreign material within thrombi in 5 of the 8 cases. The intraluminal birefringent foreign material, which had stimulated a granulomatous inflammatory response, presumably gained access to the eye or orbital tissue during administration of the IAC. Symptomatic and asymptomatic thromboembolic complications caused by the inadvertent administration of foreign substances during angiography or chemotherapy have been reported in the nonophthalmic literature.16-22 In some instances, the material was derived from catheters used during angiography or their coatings. In other instances, cotton fibers were identified. Particulates suspended within room air are thought to be a likely source of such contaminants, since cellulose fibers are a major constituent of dust. In our protocol administration of IAC, there is no cotton fiber material in the operating room and all surfaces are covered in plastic wrap. The chemotherapy is filtered to remove particulates. The chemotherapy is diluted and delivered in standard pulsatile fashion over 30 minutes. The source of the foreign body material in our cases and in the one case performed elsewhere remains unknown. After the birefringent material was initially identified, we modified our technique by eliminating all linen towels from the table and the patient's drapings during the procedure and by flushing the microcatheter with saline at the end of the chemotherapeutic agent delivery to avoid any crystallization of the agent in the ophthalmic artery.
Foreign substances such as cellulose fibers may be introduced during fluoroangiography, IAC catheter insertion, or delivery of chemotherapy more often than has been previously appreciated. Theoretically, small emboli might cause clinically inapparent areas of ischemia in large organs. The foreign material might not be disclosed by limited tissue sampling or might be overlooked when larger organs are examined histopathologically. Theoretically, foreign bodies might be detected more often in eyes enucleated after IAC for Rb because a relatively small volume of tissue with delicate vascularity is subject to intense histopathologic scrutiny. In addition, compared with other organs, eyes containing Rb are more likely to be removed and examined histopathologically after treatment failure.
In 3 of the eyes examined, viable Rb had invaded the optic nerve head, invading the lamina cribrosa in 1. Optic nerve invasion, reaching the lamina in 1 instance, was also observed in 2 of 3 eyes described by Vajzovic and colleagues.14 We also observed a microscopic focus of choroidal invasion in 1 eye. Retrolaminar optic nerve invasion and massive choroidal invasion are high-risk histopathologic features that are an indication for adjuvant chemotherapy in many institutions. Current treatment regimens for Rb including systemic chemotherapy administered prior to enucleation (chemoreduction) or on an adjuvant basis after high-risk histopathogic findings are detected in enucleated eyes and have markedly decreased the mortality due to primary Rb in developed countries in recent years. In an upcoming article, we describe 100% long-term survival in 51 children with high-risk Rb using our protocol of systemic vincristine, etoposide, and carboplatin for 6 cycles. In addition, systemic chemotherapy has greatly reduced the incidence of pineal tumors in patients with germline mutations.23 At major institutions in the United States and Europe, second nonocular tumors that develop in patients with germline mutations in the Rb1 gene are now the major cause of death in patients with Rb.24
In a large series of enucleated eyes with Rb that were accessioned by the Ophthalmic Pathology Laboratory at the Wills Eye Institute between 1986 and 2008, 18.5% contained high-risk features that would serve as an indication for adjuvant systemic chemotherapy.25 Other centers have reported an even higher incidence of high-risk features. Adjuvant systemic chemotherapy is presumed to eradicate tumor cells that have metastasized from the primary intraocular tumor to extraocular tissues. Selective IAC that is delivered solely to the eye might not eradicate extraocular tumor cells at remote sites if they are present in a patient with Rb and, furthermore, IAC might not effectively control malignant pineal tumors that could arise in patients with germline mutation. Theoretically, one might expect to find a higher incidence of metastatic disease after IAC because high-risk features, which would serve as an indication for adjuvant systemic intravenous chemotherapy, might not be detected. The latter does not appear to be entirely theoretical, because there are known cases of patients treated with IAC for Rb elsewhere in the United States and abroad who have developed metastatic disease.
In summary, histopathology of eyes with Rb following IAC showed evidence of complete tumor regression in eyes in which there was clinical tumor regression and also confirmed viable tumor in those in which tumor was suspected clinically. An equally important finding is that histopathology reveals evidence of ocular vascular occlusive events, some with intraluminal foreign bodies. This evidence confirms previously published observations.12 We suggest that IAC for Rb be used with caution.
Correspondence: Ralph C. Eagle, Jr, MD, Ophthalmic Pathology Laboratory, Wills Eye Institute, 840 Walnut St, 14th Floor, Philadelphia, PA 19107 (email@example.com).
Accepted for Publication: April 26, 2011.
Published Online: July 11, 2011. doi:10.1001/archophthalmol.2011.223
Author Contributions: Dr Eagle had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Financial Disclosure: None reported.
Funding/Support: This study was supported by the Noel T. and Sara L. Simmonds Endowment for Ophthalmic Pathology, Wills Eye Institute, and the Eye Tumor Research Foundation, Philadelphia, Pennsylvania (C. L. Shields and J. A. Shields).
Role of the Sponsors: The funders had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; and in the preparation, review, or approval of the manuscript.
Previous Presentations: Paper presented at the Wills Eye Conference; March 11, 2011; Philadelphia, Pennsylvania.
Create a personal account or sign in to: