Metastatic Neoplasms in the Optic Disc: The 1999 Bjerrum Lecture: Part 2

Background  Little information is available on metastatic tumors to the optic disc.

Objective  To determine the clinical features and prognosis of patients with optic disc metastasis.

Design  Retrospective chart review.

Results  Of 660 consecutively evaluated patients with intraocular metastasis, 30 (4.5%) (31 eyes) had metastatic cancer to the optic disc; 24 (80%) were women and 6 (20%) were men. Mean age at the time of ocular diagnosis was 55 years. The primary neoplasm was in the breast in 13 patients (43%), in the lung in 8 (27%), in the intestine in 1 (3%), in the kidney in 1 (3%), and in the prostate in 1 (3%); the primary neoplasm was never determined in 6 patients (20%). The optic disc metastasis was unilateral in 29 patients (97%) and bilateral in 1 (3%). Ophthalmoscopically, the disc metastasis appeared as a diffuse enlargement of the optic disc in 26 eyes (84%) and as a distinct nodule in 5 (16%). There was an adjacent juxtapapillary choroidal component to the metastatic disc lesion in 23 eyes (74%), and the optic disc was involved without a retinal or choroidal component in 8 (26%). Other associated findings included some degree of secondary disc edema in all eyes, buried disc blood vessels in 23 (74%), and splinter hemorrhages in 13 (42%). Fine needle aspiration biopsy was useful in establishing the diagnosis in all 5 eyes in which it was performed. Mean survival was 13 months after diagnosis of the disc metastasis.

Conclusions  Metastasis to the optic disc accounts for 5% of all intraocular metastases. It can occur as invasion from a juxtapapillary choroidal metastasis or as isolated optic disc metastasis. Breast and lung cancers are the most common primary neoplasms that account for metastasis to the optic disc. The primary site is never determined in 20% of patients. The characteristic clinical features of optic disc metastasis should help differentiate it from other causes of swollen optic disc. Patient prognosis is poor.

MOST METASTATIC cancers to the intraocular structures occur in the uveal tract, and metastases to the retina and optic disc are rare.^1-5 Isolated cases of metastasis involving the optic disc have been published,^6-19 but there have been no reports of large series of cases. Criteria are needed to differentiate metastatic cancer to the optic disc from papilledema, papillitis, and other causes of swollen optic disc.

The purpose of this article is to elucidate information on metastatic neoplasms that involve the optic disc, with emphasis on primary cancers, associated ocular findings, and ophthalmoscopic features. Based on our observations, we discuss the features that help differentiate optic disc metastasis from other lesions of the disc.

We searched the computerized database of the oncology Service at Wills Eye Hospital, Philadelphia, Pa, for cases of metastatic cancer to the optic disc. We reviewed the charts and tabulated patient age, sex, eye(s) affected, visual acuity, intraocular pressure, results of slitlamp biomicroscopy, and ophthalmoscopic findings. To determine the specific ophthalmoscopic features of optic disc metastasis we reviewed clinical notes, drawings, and fundus photographs and recorded the tumor location in the disc, the color of the mass, and the tumor configuration (diffuse or nodular). We studied the disc blood vessels and the presence and type of hemorrhage (splinter or round). In cases in which such ancillary studies were performed, we reviewed the results of fluorescein angiography, ultrasonography, computed tomography, and magnetic resonance imaging. We studied clinical records and made follow-up telephone calls to determine the site of the primary cancer and noted whether the ocular metastasis involved only the optic disc or also the adjacent choroid and other ocular tissues. Finally, we tabulated ocular and systemic follow-up information for affected patients.

Because our goal was to elucidate the ophthalmoscopic features of optic disc metastasis, we excluded from this study patients with retrobulbar metastasis without disc involvement, meningeal carcinomatosis, lymphoma, and leukemia.

Of 820 affected eyes in 660 consecutive patients with metastatic cancer to the intraocular structures evaluated by the Oncology Service at Wills Eye Hospital from February 1, 1974, through July 31, 1998, 31 eyes in 30 patients (4.5%) had metastatic cancer that involved the optic disc.

The most important clinical data for the 30 patients are shown in Table 1. At the time of diagnosis of optic disc metastasis, the mean and median patient age was 55 years (range, 39-78 years). There were 24 women (80%) and 6 men (20%). The right eye was affected in 15 patients (50%), the left eye was affected in 14 (47%), and both eyes were affected in 1 (3%).

Fourteen (47%) of 30 patients had a known primary tumor when the optic disc metastasis was found. Mean time from diagnosis of the primary cancer to diagnosis of optic disc metastasis was 19 months (range, 1-120 months). In 16 patients (53%), there was no known primary tumor when the patient was seen initially with visual symptoms secondary to the optic disc metastasis. However, in 14 of these 16 patients, ophthalmoscopy revealed an adjacent choroidal mass in addition to the optic disc lesion, thus suggesting the diagnosis and prompting a systemic evaluation for a primary neoplasm or other signs of metastatic disease. Subsequent systemic evaluation revealed the primary neoplasm in 10 of the 16 patients. In 6 patients (20%), the primary neoplasm was never specifically identified.

The primary neoplasm was in the breast in 13 patients (43%), in the lung in 8 (27%), in the intestine in 1 (3%), in the kidney in 1 (3%), and in the prostate in 1 (3%). The 6 patients (20%) in whom the primary neoplasm was never determined all eventually died of widespread metastatic carcinoma (Table 1 and Table 2).

With regard to sex, among the 24 women, the primary tumor site was the breast in 13 (54%), the lung in 6 (25%), the intestine in 1 (4%), and the kidney in 1 (4%). In 3 women (12%) the primary site was never determined. Among the 6 men, the primary site was the lung in 2 (33%), the prostate in 1 (17%), and undetermined in 3 (50%).

Twenty-seven patients (90%) had no ocular pain and 3 (10%) experienced severe pain due to elevated intraocular pressure. All 3 had a total retinal detachment secondary to choroidal metastasis in the affected eye and underwent enucleation of the affected eye to relieve the severe pain. When first seen by us with the optic disc metastasis, visual acuity in the affected eye (N = 31) ranged from 6/6 to no light perception. It was measured at 6/6 in 2 eyes, 6/9 in 3, 6/12 in 2, 6/15 in 2, 6/21 in 1, 6/24 in 2, 6/30 in 2, 6/60 in 2, 6/120 in 1, counting fingers in 5, hand movements in 6, light perception in 1, and no light perception in 2. Patients with better visual acuity had smaller tumors with a more eccentric location on the optic disc. We did not observe any relation between the type of primary tumor and visual acuity. Technicians usually checked the pupils, which were often dilated before examination by a physician. However, an afferent pupillary defect was recorded in 23 of 30 patients, mainly in those with poor vision in the affected eye.

The ophthalmoscopic features of optic disc metastasis are shown in Table 3. In 25 (81%) of the 31 metastases to the optic disc in 30 patients the tumor was centrally located, affecting the entire disc (Figure 1 and Figure 2). It assumed an eccentric location over a margin of the optic disc in 6 eyes (19%) (Figure 3, Figure 4, and Figure 5). There was minimal extension into the nerve fiber layer of the retina in all eyes, but none had a separate metastasis to the retina. Some degree of venous stasis was present in 20 eyes (64%) (Figure 1 and Figure 2, and Figure 6), but none had a complete retinal vein obstruction. The color of the tumor tissue in the optic disc was white in 16 eyes (52%) (Figure 4 and Figure 5), yellow in 10 (32%) (Figure 2), and fleshy pink in 5 (16%) (Figure 7). In 5 eyes (16%) the disc tumor was seen as a nodule (Figure 3 and Figure 4), and in 26 (84%) it was diffuse, without a distinct nodule (Figure 1, Figure 2, Figure 6, and Figure 7). The blood vessels were visualized crossing the disc in 8 eyes (26%) (Figure 3, Figure 4, and Figure 5) and were obscured by tumor tissue and edema in 23 (74%) (Figure 1, Figure 2, and Figure 6). Hemorrhages were present in or adjacent to the disc in 13 eyes (42%), and in all instances they were flame shaped and in the nerve fiber layer (Figure 1 and Figure 7). No deep dot or blot hemorrhages were observed.

Overall, 23 (74%) of the 31 eyes had juxtapapillary choroidal metastasis adjacent to the optic disc metastasis. In 17 of these the optic disc metastasis and the adjacent choroidal metastasis were present at the initial visit to us (Figure 6 and Figure 7). In 6 eyes we initially observed a juxtapapillary choroidal metastasis only, and the infiltration of the optic disc developed during follow-up examinations (Figure 8). In 8 eyes (26%) the metastasis was only to the optic disc, with minimal juxtapapillary retinal component and no choroidal involvement.

Ancillary studies were not performed in all eyes because of diagnostic certainty or patient preference. Transocular fine needle aspiration biopsy²⁰ of the optic disc mass was performed in 5 eyes, usually in patients who did not have a known primary neoplasm or who had no other evidence of metastatic disease. In all 5 eyes, malignant cells compatible with metastatic carcinoma were demonstrated (Figure 9, A and E), and there were no false-negative fine needle aspiration biopsy results. Fluorescein angiography was performed in 21 of 30 patients. Although there was some variability, most tumors showed relative hypofluorescence in the arterial phase, with gradual hyperfluorescence beginning in the late venous phase and moderate staining in the late frames (Figure 9, B, C, and D). Ultrasonograms, obtained in 20 eyes, generally showed a mild nonspecific elevation of the optic disc, with high internal reflectivity. In the 2 patients who underwent orbital magnetic resonance imaging, there was no convincing evidence of optic nerve involvement posterior to the lamina cribrosa.

Management of the affected eye varied with the clinical circumstances. Although a few patients had combinations of treatment, the main treatments for the 31 affected eyes were external beam irradiation in 14, systemic chemotherapy in 10 (given for systemic extraocular metastasis), enucleation in 3 (for severe ocular pain), and observation in 4. Because most patients were treated at a different institution, it was not usually possible to perform a detailed evaluation of treatment results and visual outcome. However, it was our general impression that patients who underwent chemotherapy or irradiation generally showed stability or regression of the optic disc tumor, but visual acuity usually did not improve after such treatment.

Follow-up ranged from 3 to 48 months after our diagnosis of optic disc metastasis. Twenty-five patients (83%) died of metastatic carcinoma, and 3 patients were still alive and undergoing treatment for systemic metastasis with less than 1-year follow-up after the diagnosis of optic disc metastasis. Two patients were lost to follow-up and are presumed dead because of systemic metastasis at the time of diagnosis of the optic disc metastasis. Of the 25 patients known to have died of metastatic disease, survival ranged from 1 to 72 months (mean, 13 months; median, 12 months).

Metastatic neoplasms to the optic nerve can occur as direct hematogenous metastasis to the neural tissue or to the overlying meninges (meningeal carcinomatosis). Most metastatic neoplasms that involve the optic disc are carcinomas,^1-18 but cutaneous melanoma can rarely occur in the optic disc.¹⁹ Hematogenous metastasis to the optic nerve can settle in the retrobulbar portion of the nerve, where it can simulate retrobulbar neuritis, or can occur anteriorly in the optic disc, where it manifests as an elevated optic nerve head lesion. Metastatic tumors that affect the optic disc can develop as an optic disc lesion without adjacent choroidal involvement or as optic disc extension from a juxtapapillary choroidal metastasis.

In our study and literature review, we included only cases in which there was an ophthalmoscopically visible swelling of the optic disc in which clinical circumstances strongly indicated metastasis from a distant organ. Because we wanted to focus on the ophthalmoscopic features, we excluded from our review patients with (1) optic nerve metastasis without disc involvement,^21-24 (2) meningeal carcinomatosis,^25-27 and (3) insufficient information as to what part of the optic nerve was involved.²⁸

In our review of the English-language literature on optic disc metastasis we identified 16 previously reported cases^6-19 that we believed were adequately documented (Table 4). Thus, our contribution of 30 patients adds substantially to the published cases and provides considerable information on optic disc metastasis.

Based on our results, metastasis to the optic disc occurs almost exclusively in adulthood at a median age of 55 years. The predilection for breast cancer accounting for a large portion of optic disc metastasis in our series is in keeping with previously reported cases of optic disc metastasis (Table 4) as well as choroidal,¹ iris,²⁹ and orbital³⁰ metastases. Of our 8 patients with lung cancer metastatic to the optic disc, 6 were women and 2 were men. Of the 6 men in our series, only 2 (33%) had lung cancer as a proven primary malignancy. Of the 4 previously described patients with lung cancer metastatic to the optic disc, 2 were men and 2 were women (Table 4). Thus, of 12 patients now described in the literature of lung cancer metastatic to the optic disc, 8 (67%) are women.

Most patients with metastatic cancer to the optic disc have a characteristic contiguous choroidal metastasis, which facilitates the diagnosis. Of our 31 eyes, 23 (74%) had a juxtapapillary choroidal component (Table 1). Of the 16 patients described in the literature, 8 (53%) had a juxtapapillary component. This difference probably can be explained by the fact that tumors confined to the optic disc are more likely to be reported than are those with a typical choroidal metastasis adjacent to the optic nerve.

Conversely, in 8 (26%) of our 31 eyes and 7 (47%) of the 15 eyes described in the literature, there was no clinically evident choroidal metastasis. In such cases it would be helpful to have criteria for making the diagnosis of optic disc metastasis. This can be facilitated by obtaining a good history, performing a systemic evaluation, and having knowledge of the ophthalmoscopic features of optic disc metastasis. For example, if the affected patient has a history of cancer, particularly if there is systemic metastasis, the diagnosis of optic disc metastasis becomes more likely. If the patient is found on systemic evaluation to have a primary cancer, this also makes the diagnosis of optic disc metastasis more likely.

If there is no history of cancer and no detectable systemic cancer, then the ophthalmoscopic features must be taken into consideration and the optic disc swelling must be differentiated from papilledema and other causes of the swollen optic disc. Optic disc metastasis is usually unilateral and characterized by central swelling of the disc. Some optic disc metastases were characterized by an intensely white, chalky infiltrate that appears with ophthalmoscopy (Figure 3 and Figure 4) and fluorescein angiography to be relatively avascular. In other instances, the metastasis is creamy yellow (Figure 2). The margins are often sharply circumscribed with scallops or nodules (Figure 1 and Figure 2). Hemorrhages are observed in the swollen tissue in 42% of the eyes, and they are primarily flame shaped. Deeper, round hemorrhages were not observed in any of the eyes we studied. Papilledema, due to increased intracranial pressure from brain tumor or pseudotumor cerebri, is usually bilateral and rather symmetrical and produces peripheral visual field loss with little loss of central vision, and associated hemorrhage is rare. Avascular, chalky white infiltrates are unlikely to occur with papilledema. In addition, papilledema tends to cause earlier hyperfluorescence with fluorescein angiography. Optic papillitis, as seen with demyelination disease, is more likely to be unilateral and less extensive and lacks a solid nodule. Optic disc granuloma, due to sarcoidosis and other granulomatous disease, is usually unilateral and appears as a distinct nodule, sometimes with overlying vitreous cells and often a history or clinical findings of systemic inflammatory disease. It is more likely to be eccentrically located on the disc, with localized swelling. Drusen of the optic disc are more likely to be bilateral and to appear as distinct yellow bodies. They tend to autofluoresce and are prone to produce deeper round or oval hemorrhages, often secondary to choroidal neovascularization. Capillary hemangioma of the optic disc is generally found in younger patients who have von Hippel-Lindau syndrome, and it often produces yellow intraretinal exudation and shows early hyperfluorescence with fluorescein angiography.

With attention to the previously mentioned criteria, the diagnosis of optic disc metastasis should be established in most patients. Although most of our patients did not have overt evidence of orbital or brain metastasis, the patient should also undergo periodic magnetic resonance imaging or computed tomography of the orbit and brain to detect metastasis in those locations. Transvitreal fine needle aspiration biopsy can be used to make the diagnosis in uncertain cases if there is significant visual loss.

With regard to treatment, our patients had various cancers managed by different approaches at various centers. Therefore, we do not have enough data to make specific recommendations. However, based on our observations, metastasis to the optic disc can lead to severe visual loss if not treated early. The patient should be treated with appropriate chemotherapy for any systemic cancer and undergo external ocular irradiation to the posterior segment and anterior orbit of the affected eye. About 35 to 40 Gy is generally recommended and usually should be given shortly after the diagnosis of optic disc metastasis. The affected patient should be followed up closely for other systemic metastasis.

The systemic prognosis for patients with optic disc metastasis is generally poor. In our series and in the literature, mean survival is about 13 months.

In summary, we described the demographics, clinical features, diagnostic approaches, treatment, and prognosis of patients with cancer metastatic to the optic disc. It is a condition of adulthood that generally occurs unilaterally as enlargement of the optic disc due to tumor infiltration. More than 50% of patients have no history of cancer. Breast and lung cancers are the most common primary tumors to account for optic disc metastasis. Fine needle biopsy can be used in difficult cases. Management includes appropriate chemotherapy and/or ocular radiotherapy. The systemic prognosis is poor.

Accepted for publication September 4, 1999.

This study was supported by the Eye Tumor Research Foundation, Philadelphia, Pa; the Award of Merit in Retina Research, Houston, Tex (Dr J. Shields); and the Macula Foundation, New York, NY (Dr C. Shields).

Presented as part of the 1999 Bjerrum Lecture, Danish Ophthalmological Society, Copenhagen, Denmark, December 4, 1999.

Reprints: Jerry A. Shields, MD, Oncology Service, Wills Eye Hospital, 900 Walnut St, Philadelphia, PA 19107.