ORBITAL RECURRENCE OF CHOROIDAL MELANOMA 20 YEARSAFTER ENUCLEATION
JERRY A. SHIELDS, M. D., JAMES J. AUGSBURGER, M. D.,
AND MALVIN J. DOUGHERTY, M.D.
A 54-year-old woman developed an orbital mass 20 years afterenucleation of the left eye for a choroidal melanoma, which proved tobe of spindle-A type. Computed tomography of the orbit suggested thediagnosis of a cavernous hemangioma, but orbital biopsy confirmed thediagnosis of recurrent malignant melanoma and the patient was man-aged by orbital exenteration followed by radiotherapy. The orbitalrecurrence was of the mixed cell type, suggesting a gradual 20-yeartransition of the spindle-A cells to spindle-B and epithelioid cells. Werecommend that the selected surgical management of extrascleralextension of uveal melanomas be determined by the degree of extra-scleral involvement and by whether the extrascleral extension isdiscovered clinically, at surgery, or in the pathology laboratory. Werecommend that orbital computed tomography be performed periodi-cally after enucleation in all patients who have orbital extension of uvealmelanomas.
Orbital recurrence of malignant mela-noma of the choroid is uncommon, occur-ring in about 3% of patients who undergoenucleation for this disease and in 18% ofpatients who have gross or microscopicevidence of extrascleral extension of thetumor at the time of enucleation.' Beforethe advent of computed tomography,there were no reliable methods for theearly detection of orbital recurrence insuch cases. We report herein the comput-ed tomographic characteristics and histo-
Accepted for publication April 4, 1984.From the Oncology Service, Wills Eye Hospital,
Jefferson Medical College, Thomas Jefferson Univer-sity, Philadelphia, Pennsylvania. This study was sup-ported in part by the Pennsylvania Lions SightConservation and Eye Research Foundation, Inc.,and the Ocular Oncology Fund, Wills Eye Hospital,Philadelphia.
Reprint requests to Oncology Service, Wills EyeHospital, Ninth and Walnut Streets, Philadelphia, PA19107 (Dr. Shields).
pathologic findings of an orbital melano-ma which recurred 20 years afterenucleation for choroidal melanoma. Themelanoma cells showed an interestingtransition from predominantly spindle-Acells in the original tumor to mixed celltype in the recurrence 20 years later.
A 54-year-old woman had undergone enucleationof her left eye at age 34 because of a malignantmelanoma of the choroid. She was examined periodi-cally for the next 18 years and on each visit theprosthesis was removed, and examination of thesocket showed no evidence of recurrence. One yearbefore our examination, however, she noticed thegradual onset of progressive forward displacement ofthe prosthesis in the anophthalmic left orbit. She wasreferred here for diagnosis and treatment.
Her family history and medical history were non-contributory. Results of physical and ocular examina-tion were normal except for the findings in the leftorbit. The left orbital tissues were displaced anterior-ly to such a degree that her prosthesis could not beconsistently retained (Fig. 1). A ball implant waseasily visible through the thin, stretched overlyingconjunctiva. No orbital mass was palpable around the
AMERlCAN JOURNAL OF OPHTHALMOLOGY 97:767-770, 1984 767
768 AMERICAN JOURNAL OF OPHTHALMOLOGY JUNE, 1984
Fig. 1 (Shields, Augsburger, and Dougherty). Ex-ternal photograph showing loosely fitting prosthesisover the ball implant in the left eye.
ball implant, which resisted retropulsion. Our clini-cal diagnosis was orbital recurrence of choroidalmelanoma.
A- and B-scan ultrasonography of the orbit wereinconclusive because the ball implant caused markedacoustic shadowing of the posterior orbital tissues.Computed tomography of the orbit, however, dis-closed a large, well-circumscribed mass that occu-pied most of the left orbit posterior to the ballimplant (Fig. 2). There was pressure erosion of themedial wall of the orbit by the mass. The computedtomographic findings were interpreted by the neuro-radiologist as highly suggestive of a cavernous he-mangioma, in disagreement with our clinical impres-sion. On the basis of the computed tomographyreport, we elected to perform an excisional biopsy ofthe mass. Results of preoperative laboratory studies,including a complete blood cell count, serum gamma
Fig. 2 (Shields, Augsburger, and Dougherty).Contrast computed tomography showing large orbitalmass posterior to the ball implant. Note the erosionof bone on the nasal wall of the orbit.
glutamyl transpeptidase, lactate dehydrogenase, anda chest x-ray were normal.
An attempt to remove the tumor intact failed whenthe portion attached to the ethmoid bone developeda rent in the capsule, liberating black tumor tissueinto the surgical field. The tumor was thereforeremoved piecemeal and the specimens submitted forpermanent sections. After confirmation of the diag-nosis of a mixed cell type malignant melanoma 24hours later, an eyelid-sparing exenteration of theorbit was performed. No skin graft was applied andthe orbit was allowed to heal by granulation. Thepatient was treated postoperatively with 4,000 rads ofexternal beam irradiation to the left orbit, in divideddoses over a six-week period. Approximately 30months after surgery, the patient developed a pulmo-nary tumor, which was excised and diagnosed histo-logically as a bronchogenic carcinoma. The patient isstill alive with no clinical evidence of metastaticmelanoma 48 months after exenteration.
A review of the histologic sections ofthe enucleated eye demonstrated a cho-roidal melanoma with a serous detach-ment of the retina. Multiple representa-tive sections showed the tumor to becomposed predominantly of minimallypigmented spindle-A melanoma cells(Fig. 3). Mitotic figures were extremelyrare. A small focus of extrascleral exten-sion of the tumor was evident (Fig. 4).Sections through this area showed flat-tened, benign-appearing spindle-A mela-noma cells but no distinct nodules. Areview of the gross examination made 20
Fig. 3 (Shields, Augsburger, and Dougherty).Structure of the intraocular tumor, showing minimal-ly pigmented, spindle-A melanoma cells (hematoxy-lin and eosin, x 1(0).
VOL. 97, NO.6 CHOROIDAL MELANOMA 769
Fig. 4 (Shields, Augsburger, and Dougherty).Low-power photomicrograph showing extrascleralnodule of the tumor (arrow). The intraocular tumor isabove, the sclera in the center, and the extrascleralnodule is below (hematoxylin and eosin, X 10).
years earlier indicated no note of extra-scleral involvement. Sections of both theorbital tumor and the exenteration speci-men disclosed residual spindle-B and epi-thelioid melanoma cells (Fig. 5). Mitoticactivity within the tumor cells was rela-tively low with only one to two mitoticfigures per 40 high-power fields.
Orbital recurrence of choroidal and cil-iary body melanomas can occur severalyears after enucleation. 1 There have beenreports of orbital recurrence at 13 years,"
Fig. 5 (Shields, Augsburger, and Dougherty).S~ructure of ~rbital recurrence, showing minimallypigmented spmdle-B and epithelioid melanoma cells(hematoxylin and eosin, x 100).
24 years." 28 years," and 30 years" afterenucleation. A recent report described anorbital recurrence that did not becomeclinically apparent for 15 years after evis-ceration of the ocular contents in an eyewith an unsuspected intraocular melano-rna."
An interesting aspect of the presentcase is the difference in the cytologicfindings between the original choroidaltumor and the orbital recurrence 20 yearslater. The original tumor was predomi-nantly of the spindle-A cell type and therecurrence was predominantly of themixed cell type. This suggests that theremay be a gradual evolution of uveal mela-nomas from more benign to more malig-nant cell types, a possibility we havesuspected but have not been able to doc-ument previously.
The prognosis for life becomes worsewhen a patient with a uveal melanomadevelops a recurrence in the orbit. 1 Thismay occur in part because the recurrencedoes not usually become clinically appar-ent until it is quite large, as illustrated inour case. In view of the poor prognosis,some investigators have recommendedearly exenteration as the treatment ofchoice for choroidal melanoma with orbit-al extension," although others have disa-greed with this approach.P The recentcontroversy regarding the role of enucle-ation in the management of uveal melano-mas 9,lO could also raise questions as to theadvisability oforbital exenteration for thistumor.
The melanoma in the orbit of our pa-tient had probably been growing slowlyfor many years, perhaps since the enucle-ation 20 years before, but it had not beenclinically evident. When it did becomeclinically recognizable, it was far ad-vanced, with compression on the orbitalbones, making the definitive surgerymore difficult. Based on our experiencewith a number of affected patients, wenow recommend the following approach
770 AMERICAN JOURNAL OF OPHTHALMOLOGY JUNE, 1984
to the management of choroidal melano-mas with extrascleral extension providedthe patient has no evidence of systemicdissemination of the tumor. 11,12
If an extrascleral nodule is detected atthe time of enucleation, the managementdepends on the extent of orbital involve-ment. In the case of a small, grosslyencapsulated nodule, we recommendsurgically covering the extrascleral tu-mors with Tenon's fascia and adjacentorbital tissues and suturing them to thesuperficial scleral fibers around the nod-ule. A freezing, "no-touch" enucleation isthen perforrned.P If massive extraocularextension is evident at surgery, then aprimary orbital exenteration should beperformed.
In cases where extrascleral extension isnot detected until histopathologic exami-nation of the enucleated eye, the bestmanagement is also not yet established.However, it seems reasonable that if thenodule is poorly encapsulated and ap-pears to be capable of future recurrence,the patient should be managed by teno-nectomy with removal of the orbital im-plant and adjacent orbital tissues." Inretrospect, this procedure may have beenadvisable when our patient had enuclea-tion 20 years previously. All patients withextrascleral extension, after initial surgi-cal management as outlined, should haveperiodic clinical examinations, and com-puted tomography should be performed ifthere is suspicion of orbital recurrence.The role of radiation therapy as supple-mental treatment in patients with extra-ocular extension of uveal melanoma is notestablished, but it seems advisable be-cause of the poor prognosis for thesepatients.
In our case, computed tomographysuggested a cavernous hemangioma ofthe orbit rather than a melanoma. Inthe case reported by Levine, Putterman,and Korey" computed tomography alsoshowed a large orbital tumor with en-hancement with contrast media, similar
to that seen in our patient. In their casethe diagnosis of recurrent melanoma wasnot suspected until surgical exploration.We believe that all patients with extra-scleral extension of choroidal melanomashould have periodic computed tomogra-phy and any orbital tumor detected insuch patients be considered to be recur-rent malignant melanoma until provenotherwise.
1. Starr, H. J., and Zimmerman, L. E.: Extra-scleral extension and orbital recurrence of malignantmelanoma of the choroid and ciliary body. Int.Ophthalmol. Clin. 2:369, 1962.
2. Newton, F. H.: Local recurrence of melanomaof the choroid 13 years after enucleation. Am. J.Ophthalmol. 21:668, 1938.
3. Saunders, D. H., Rodrigues, M. M., andShannon, G. M.: Orbital recurrence of malignantmelanoma of the choroid 24 years after enucleation.Ophthalmic Surg. 8:31, 1977.
4. Allen, J. c.. and Jaeschle, W. H.: Recurrenceof malignant melanoma in the orbit after 28 years.Arch. Ophthalmol. 76:79, 1966.
5. Reese, A. B.: Tumors of the Eye, 3rd ed.Hagerstown, Harper & Row, 1976, p. 210.
6. Levine, R. A., Putterman, A. M., and Korey,M. S.: Recurrent orbital malignant melanoma afterthe evisceration ofan unsuspected choroidal melano-ma. Am. J. Ophthalmol. 89:571, 1980.
7. Shammas, H. F., and Blodi, F. D.: Orbitalextension of choroidal and ciliary body melanomas.Arch. Ophthalmol. 95:2002, 1977.
8. Affeldt, J. C., Minckler, D. S., Azen, S. P.,and Yeh, L.: Prognosis in uveal melanoma withextrascleral extension. Arch. Ophthalmol. 98:1975,1980.
9. Zimmerman, L. E., McLean, I. W., and Fos-ter, W. D.: Does enucleation of the eye containing amalignant melanoma prevent or accelerate the dis-semination of tumor cells? Br. J. Ophthalmol. 62:420,1978.
10. Shields, J. A., and Augsburger, J. J.: Themanagement of choroidal melanomas. Am. J. Oph-thalmol. 90:266, 1980.
11. Shields, J. A.: Diagnosis and Management ofIntraocular Tumors. Chapt. 10, The Management ofPosterior Uveal Melanomas. St. Louis, C. V. Mosby,1983.
12. Shields, J. A., Cruess, A. F., Augsburger,J. J., and Corwin, C.: The management of extraocu-lar extension of posterior uveal melanomas. Orbit1:209, 1982.
13. Fraunfelder, F. T., Boozman, F. W., III, Wil-son, R. S., and Thomas, A. H.: No-touch techniquefor intraocular malignant melanomas. Arch. Ophthal-mol. 95:1616, 1977.
14. Wolter, J. R.: Epibulbar extension of a choroi-dal melanoma treated with tenonectomy. Ophthal-mic Surg. 5:48, 1974.