British Journal of Ophthalmology, 1984, 68, 272-275

Uveal melanoma: case report of extension through theoptic nerve to the surgical margin in the orbital apexJEREMY CHESS,' DANIEL M. ALBERT,' A. ROBERT BELLOWS, ANDRICHARD DALLOW3

From 'Harvard Medical School, Massachusetts Eye and Ear Infirmary, Boston; 2Ophthalmic Consultantsof Boston, Inc., Boston; and 38 Hawthorne Place, Boston, Massachusetts, USA

SUMMARY An unsuspected uveal melanoma arising in a phthisical eye extended through the opticnerve to the surgical margin in the orbital apex without involvement of other orbital structures.Step sections showed the continuous extension through the nerve. Factors likely to predispose tooptic nerve extension include epithelioid cell type, large tumour size, peripapillary involvement,and glaucoma. Optic nerve extension carries the risks noted with orbital extension of intraoculartumour, but may in addition lead more directly to intracranial invasion, including a chiasmalsyndrome. The role of exenteration and radiotherapy in these cases remains uncertain.

Optic nerve extension of uveal melanomas is currentlya relatively uncommon occurrence. It was noted, forexample, in 3-6% of the reported cases summarisedby Reese' and in less than 0-6% of the extensiveseries reported by Wilder and Paul.' In an earlierseries optic nerve extension was seen in approximately20% of cases.We report a case ofan unsuspected uveal melanoma

in a phthisical eye which we believe is unique in itsdocumentation by stepped sections of optic nerveextension to the orbital apex without invasion ofother orbital tissues. Several previous reports3-7have made note of this form of tumour spread buthave not to our knowledge eliminated the possibilityof orbital or haematogenous spread of tumour to theoptic nerve in the orbital apex. This case also bringsto light problems encountered in the diagnosis ofunsuspected intraocular malignant melanomas andthe management ofextraocular spread of this tumour.

Case report

A 59-year-old white man presented in the EmergencyRoom at the Massachusetts Eye and Ear Infirmary on21 November 1977. His previous ophthalmic historywas reported to include the development of aninferotemporal quadrantanopsia in 1969 following aretinal vascular occlusion. Six months later he becameCorrcspondencc to Danicl M. Albert, MD. Massachusetts Eye andEar Infirmary, 243 Charics Strect. Boston, MA 02114, USA.

totally blind in the eye. He had been followed up by aprivate ophthalmologist, now deceased, for increasedintraocular pressure and intermittent pain associatedwith iris neovascularisation.

His previous medical history included medicallycontrolled seizures dating back to childhood andsystemic hypertension. Previous reported surgeryhad been limited to mastoidectomies.At the time of his first examination by one of us

(A.R.B.) on 21 November 1977 visual acuity was nolight perception in the right eye and 20/25 in the leftwith best correction. Pertinent findings were limitedto the right eye. There was 3+ conjunctival injection.The cornea was clear. A large blood clot filled theanterior chamber, and no other ocular structurescould be visualised. Intraocular pressure was 29mmHg OD and 16 mmHg OS. Miotics were discon-tinued, and treatment with topical atropine andsteroids was instituted.The haemorrhage had not entirely cleared until 17

February 1978, at which time fluffy anterior chamberdebris and a mature cataract were noted along withcontinued glaucoma. Inability to visualise the fundusprompted a referral for ultrasound examination on 2March 1978.

Ultrasound examination revealed a globe"somewhat smaller than average' with normal con-figuration. Debris in the anterior chamber andcataract were noted. The vitreous cavity was 'almostentirely filled with confluent moderate and low

272

on June 4, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.68.4.272 on 1 April 1984. D

ownloaded from

Uveal melanoma: extension through the optic nerve

Fig. 1 External appearaince of phtllisicalglobeprior toenuacleation.

density membranes.' It was felt that the findings'probably represent very marked intraocular debrisfrom hemorrhage or inflammation; however, retinaldetachment or tumor cannot be excluded.'

In February 1979 the eye was found to be dis-coloured and deformed, and the intraocular pressurewas too low to be measured accurately (Fig. 1). Thepatient was advised to undergo enucleation, but herefused. However, after continued bouts of dis-comfort and inflammation he consented to anenucleation. An additional ultrasound examinationwas performed and revealed findings interpreted as'disorganized intraocular contents with the prob-ability of organized retinal detachment. No intra-ocular tumor was identified.' Enucleation wasperformed on 21 May 198 1.

Histopathological examination revealed an intra-ocular melanoma with tumour present at the site oftransection of the optic nerve. The patient wasreadmitted to the hospital and examined formetastases. A physical examination, chest x-ray,skeletal survey, bone scan, liver and spleen scan,serum chemistry profile, and orbital computerisedtomography scan failed to reveal evidence of meta-static disease. An orbital exenteration was performed.Postoperative external beam radiotherapy wasadministered. He received 5000 rads to the right orbitand intracranial optic nerve during a 5-week period.The left eye retained normal visuail acuity and anormal visual field. However, the patient presented17 months after enucleation with lung and livermetastases and died with widespread metastases. Nopost-mortem examination was performed.

GROSS PATHOLOGICAL. I)ESCRIPTIONThe deformed right eye measured 17 x 19 x 2() mm.The opacified cornea measured 9x 10 mm. The pupilwas not visible. Minimal conjunctival and orbitaltissue was attached. The optic nerve was 1 mm inlength and the cut end appeared black. The eye wassectioned horizontally.The globe was filled with a homogeneous, black,

poorly cohesive material. The anterior chamber wasobliterated, no angle structures were visible, and theiris and ciliary body could not be identified. Theretina, choroid, and optic disc could not be dis-tinguished grossly. The sclera was intact.

MICROSCOPICA1 D)ESCRIPTIONA large diffuse uveal melanoma of mixed cell typeextended from the optic nerve to the ciliary body andoccupied approximately one half of the vitreouscavity. There was a predominance of spindle B cellsbut an average of 5 epithelioid cells per high-powerfield were present (Fig. 2).The cornea showed stromal scarring; Descemet's

membrane was thrown into folds. Disorganisedremnants of necrotic iris were present, and the anglestructures were not identifiable. The anteriorchamber was filled with a thick proteinaceousmaterial containing cholesterol crystals. Thecataractous lens was dislocated anteriorly.Cholesterol crystals and red blood cells were found inthe anterior vitreous. The retina was disorganisedand detached. Tumour cells extended into the innerscleral lamellae and also invaded the optic nerve,extending to the site of surgical transection. Sections

Fig. 2 Epitheloidnmelanoma, intraocilar portion.HaeimatoxYlin? mitl eosin (X 1(00).

273

on June 4, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.68.4.272 on 1 April 1984. D

ownloaded from

Jeremy Chess, Daniel M. Albert, A. Robert Bellows, and Richard Dallow

Fig. 3 Melanoma invading atrophicoptic nerve. Haematoxylin and eosin(x 100).

of tumour invading the optic nerve papilla were notavailable because of inadvertent thick sectioning ofthe paraffin block through this area prior to discoveryof the intraocular tumour.

MICROSCOPIC FINDINGS OF THE EXENTERATIONSPECIMENInvasion of the optic nerve by melanoma cellsextended from the proximal to distal end of thisspecimen. Stepped sections revealed the continuityof the tumour through the nerve from the globe to thesurgical margin at the level of the annulus (Fig. 3). Notumour was present in the other orbital tissues.

Discussion

The present case is of interest as an example of anuncommon form of spread of a uveal melanoma inthe absence of extraocular extension through thescleral emissary canals. We showed with steppedsections of the exenterated orbital tissue that spreadof tumour to the orbital apex was entirely within theoptic nerve and that there has been no invasion ofother orbital structures.The possibility of intracranial invasion through the

optic nerve was first suggested by Knapp3 in 1864 andillustrated in a report by Landsberg.4 Two welldocumented cases were reported by Terry' from ourinstitution in 1935 in which intracranial extensionalong the optic nerve led to a hemianopia in thefellow eye from chiasmal involvement and in whichthe eventual deaths resulted from the intracranialinvolvement. A similar case encountered by Blodiwas reported by Spencer.7

The present case is also instructive in relation tothe clinical problems encountered in the managementof phthisical eyes. The risk of a 'hidden' melanoma ina phthisical eye has been well documented, and 12%of histologically proved uveal melanomas were un-suspected preoperatively in the series reported byYanoff and Zimmerman.8 There has been increasedreliance on ultrasound examination to detect thesetumours, but this may be difficult even in experthands.A number of factors noted by others to increase the

likelihood of optic nerve extension were prominentfeatures in our patient. Reese,' Samuels,' and Wilderand Paul2 emphasised that this form of invasion wasprominent in advanced cases, often with juxta-papillary location. Shammas and Blodi'" reviewed 26cases of peripapillary melanoma with optic nerveextension and found that most were necrotic tumourswith mixed or epithelioid cell type. Further, theyfound invasion of the retina, blindness, and secondaryglaucoma in all their cases. Spencer7 commented onthe possible mechanism by which glaucoma may leadto optic nerve extension. He indicated that the pos-sible development of cavernous atrophy secondary toincreased intraocular pressure might predispose thenerve to tumour invasion. A rubeotic glaucoma,presumably secondary to the reported vascularocclusion occurring some 10 years prior to enuclea-tion, was a noted aspect of our patient's clinicalcourse.The prognosis for patients with optic nerve

extension was noted by Shammas and Blodi"' to be'mainly dependent on the presence of orbitalextension.' Orbital exenteration has been the most

274

on June 4, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.68.4.272 on 1 April 1984. D

ownloaded from

Uveal melanoma: extension through the optic nerve

frequent surgical approach to this problem, butits effectiveness is controversial. Starr andZimmerman'" described the poor outcome afterexenteration for orbital extension. Fourteen of their15 patients were dead of metastases within an averageperiod of 13 months. Subsequently Shammas andBlodi'2 found that exenteration within 8 weeks ofenucleation led to a significant improvement in 5-year survival rates in their 13 patients undergoingorbital exenteration. However, Affeldt et al. '" didnot observe an improved survival rate in a series of 7patients treated with early exenteration. Similarly,the therapeutic role of radiotherapy'4 or

neurosurgery in cases such as ours remains uncertain.

References

I Rccsc AB. 1hotnors ofth(1 Evye. Ncw York: Ilarpcr. 1951.2 Wildcr [IC, Paul EV. Mialignant miclanoria of thc choroid and

ciliary body: a study of 2535 cases. Milit Stirg 195 1; 109: 370-8.3 Knaipp I1. A treatise 0ot intraocul(ir tlitnor.s Jrotii origiaiol clinical

observation,s (1t1d ainaionicwal iniv'estigaitions. Trians Colc S. New

York: Wood. 1869.

4 Landsherg M. Saircom der Choroidcea schr langsamer,cigenthumlicher Vcrhaiuf. Arch Op/tdialtnol 1869; 15: 21(1-4.

5 Tcrry TL, Johns JP. Uvcal sarcoma-malignant mclainoma:statisticail study of 94 cases. Ain J Oplithalmnol 1935; 18: 90)3-13.

6 Tcrry TL. Malignant so-calilcd saircomia of uvcea. Arch Oph-t oillnol 1940; 24: 20)6-14.

7 Spcnccr WH. Optic ncrvc cxtension of intraocular ncoplaisms.A]nJ Op/ithalmnol 1975; 80: 465-71.

8 Yainoff M, Zimmcrmian LE. Histogcncsis of malignaintmclainomais of the uvci. 11. Relationship of uvcal ncvi tomalignant melanomas. Cancer 1967; 20: 493-50)7.

9 Samucls B. Study of ainaitomic and clinicail mainifestiations ofnecrosis in 84 cases of choroidall sarcomas. Arc/i Ophthalmnol1934; I1: 998-10)27.

1() Sharmmas HF. Blodi FC. Pcripapillliry choroidili melanomas.Arch Ophtalanol 1978; 96: 44(0-5.

11 Stairr HJ, Zimmcrmain LE. Extrasclcral cxtcnsion of choroidiland ciliary body melanomas. Ihit Opitl/ialmnol Cliii 1962; 2:369-85.

12 Shammas -IF. Blodi FC. Orbitatl cxtension of choroidal aindciliairy body melanomas. Arch Ophialhanol 1977; 95: 2(X)2-5.

13 Affcldt JC, Mincklcr DS, Azen SP, Ych L. Prognosis in uvceilmclainoma with cxtraisclcrail extcnsion. Arch Ophtlialmnol 1981);98: 1975-9.

14 Char DII, Phillips RD. The potential for adjuvant raidiotherapyin choroidal mclanoma. Arch Op/it/ia/nol 1982; 100: 247-8.

275

on June 4, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.68.4.272 on 1 April 1984. D

ownloaded from