By internally draining subretinal fluid at the time of the scleral thinning procedure, we were able to reattach the retina immediately and prevent addition­al photoreceptor damage by slowly resorbing subreti­nal fluid. External drainage of subretinal fluid repre­sents an alternative approach to vitreous surgery in such cases.3 Problems with an external approach include a significant risk of subretinal hemorrhage while penetrating thickened, congested choroid; pro­found intraoperative hypotony; and difficulty access­ing the posteriorly shifting subretinal fluid. Although it is more invasive, we prefer the internal approach, which decreases the risk of choroidal hemorrhage, affords excellent control over the intraocular pressure, and allows complete drainage of subretinal fluid.

REFERENCES

1. Gass JDM. Uveal effusion syndrome: a new hypothesis con­cerning pathogenesis and technique of surgical treatment. Trans Am Ophthalmol Soc 1983;81:247-260.

2. Johnson MW, Gass JDM. Surgical management of the idio-pathic uveal effusion syndrome. Ophthalmology 1990;97: 778-785.

3. Brockhurst RJ. Ciliochoroidal (uveal) effusion. In: Ryan SJ, Schachat AP, Murphy RB, editors. Retina, 2nd ed. Chicago: Mosby, 1994:1745-1752.

Vitreous Cells as an Indicator of Retinal Tears in Asymptomatic or Not Recently Symptomatic Eyes Edwin E. Boldrey, MD

PURPOSE: To define the relationship between vitreous cells and retinal tears in eyes without recent flashes or floaters. METHOD: Five hundred eighty-five consecutive patients who complained of light flashes, floaters, or both in one eye had their asymptomatic or not recently symptomatic fellow eye examined pro-spectively for vitreous cells and retinal tears. RESULTS: Of those fellow eyes with 2 + or more vitreous cells (10 or more cells per l-mm slit-lamp field), 31.6% (6/19) had one or more tears vs 1.6% (9/566) of those with 1+ or fewer cells

(less than nine cells per l-mm slit-lamp field) (P = .0001). CONCLUSIONS: A substantial number of vitreous cells, even in an eye with no recent symptoms, is correlated with the presence of a retinal tear. Additionally, a retinal tear may exist in an eye with 1 + or fewer cells.

THE ASSOCIATION BETWEEN VITREOUS CELLS AND retinal tears in eyes with floaters of recent onset or

with light flashes has been well documented.1"3 To my knowledge, this association has not been studied previously in a large number of asymptomatic or not recently symptomatic eyes.

I prospectively studied 589 consecutive patients, aged 16 to 92 years, referred to a retinal practice for light flashes, floaters, or both. The correlation be­tween vitreous cells and retinal tears in these symp­tomatic eyes has been reported previously, as have the patient demographics.2 In 585 patients, the asymp­tomatic (with no vitreous floaters or with floaters that had not changed for at least 6 months) fellow eye was also examined. Eyes with light flashes or with a history of eye surgery, trauma, diabetic retinopathy, or uveitis were not included. In four fellow eyes, ade­quate peripheral examination was not possible.

After dilation, slit-lamp examination of the ante­rior to midvitreous was performed in all patients, and the density of vitreous cells was graded. These measurements were taken using a 1 X 9-mm vertical slit-lamp beam to examine the vitreous from immedi­ately behind the lens to the midvitreous. Cell density ranged from 0 to 4 + , with 0 indicating no cells; trace, occasional cells; 1 + , one to nine cells per field; 2+ , 10 to 30 cells; 3 + , 31 to 100 (estimated) cells; and 4 + , innumerable cells. For simplification, cells were not differentiated by cell type. Peripheral retinal examination was performed, using indirect ophthal-moscopy and scleral depression in all eyes and three-mirror contact lens biomicroscopy in some. Fifteen (2.6%) of 585 asymptomatic eyes had at least one flap

Accepted for publication Oct 4, 1996. Department of Ophthalmology, Stanford University. Inquiries to Edwin E. Boldrey, MD, Retina-Vitreous Associates, Inc,

2512 Samaritan Ct, Ste A, San Jose, CA 95124; fax: (408) 356-8849.

VOL.123, No. 2 BRIEF REPORTS 263

Table. Relation Between Vitreous Cells and Retinal Tears in Asymptomatic or Not Recently Symptomatic Eyes

Vitreous Celt Density Percent of Eyes With Tears

None 0.9% (4/440) Trace 4.5% (4/88) 1+ 2.6% (1/38) 2+ 25.0% (3/12) 3+ 42.9% (3/7) 4+ 0 Total 2.6% (15/585)

or operculated retinal tear. Twenty-one eyes (3.6%) also had at least one atrophic round retinal hole without traction or local retinal detachment. These nontractional holes were not included in this analysis.

Increased numbers of vitreous cells were associated with an increased likelihood that a retinal tear was present (Table). Of patients with 2+ or more cells (10 or more cells per 1-mm slit beam), 31.6% (6/19) had at least one retinal tear vs those with 1 + or fewer cells (one to nine cells per 1-mm slit-lamp field), of which 1.6% (9/566) had a tear (P = .0001, chi-square).

This study implies that routinely performing slit-lamp examination of the anterior vitreous in search of cells may help to identify asymptomatic eyes harbor­ing retinal tears, especially if 2+ or more vitreous cells are present. However, only 40% of eyes (6/15) with tears had 2+ cells or more, and four tears (26.7%) had no associated cells. Tears without cells were generally small and did not expose much retinal pigment epithelium to the vitreous cavity, whereas larger tears tended to have more cells associated with them.

It is possible that these eyes are not representative of all asymptomatic eyes because all were selected by referral, and all had a recently symptomatic fellow eye. However, the 2.6% of asymptomatic eyes that had a tear is similar to the 1.9% and 2.1% found in two large unselected series,4,5 implying that patients in this study may be representative of the general population.

Examination of the anterior vitreous for cells can be performed quickly and easily. Because vitreous cells may indicate the presence of a retinal tear, slit-lamp examination of the anterior vitreous should

be performed in all patients undergoing a routine eye examination regardless of whether or not that patient has reported recent light flashes or floaters. Addition­ally, because nine eyes with 1 + or fewer cells had a retinal tear, the absence of vitreous cells does not rule out the presence of a tear.

REFERENCES

1. Hamilton AM, Taylor W. Significance of pigment granules in the vitreous. Br J Ophthalmol 1972;56:700-702.

2. Boldrey EE. Risk of retinal tears in patients with vitreous floaters. Am J Ophthalmol 1983;96:783-787.

3. Brod RD, Lightman DA, Packer AJ, Saras HP. Correlation between vitreous pigment granules and retinal breaks in eyes with acute posterior vitreous detachment. Ophthalmology 1991;98:1366-1369.

4. Byer NE. Clinical study of retinal breaks. Trans Am Acad Ophthalmol Otolaryngol 1967;71:461-472.

5. Foos RY. Postural peripheral retinal tears. A n n Ophthalmol 1974;6:679-687.

Conjunctival Biopsy in Infantile Neuroaxonal Dystrophy Rosane C. Ferreira, MD, Gary W. Mierau, PhD, and J. Bronwyn Bateman, MD

PURPOSE: To describe a case of infantile neuroax­onal dystrophy with optic nerve atrophy and to discuss the diagnostic role of conjunctival biopsy. METHODS: We performed a complete ophthalmo-logic examination and a diagnostic conjunctival biopsy on a girl with a neurodegenerative disease. RESULTS: On the basis of "spheroid" inclusions in the unmyelinated axons, we diagnosed infantile neuoroaxonal dystrophy. CONCLUSIONS: Optic atrophy is an important finding in infantile neuroaxonal dystrophy, and conjunctival biopsy is a reliable and very conve­nient diagnostic test.

Accepted for publication Sept 20, 1996. Departments of Ophthalmology (R.C.F., J.B.B.) and Pathology

(G.W.M.), The Children's Hospital, University of Colorado; and Depart­ment of Ophthalmology, Federal University of Sao Paulo (R.C.F.).

Inquiries to ]. Bronwyn Bateman, MD, Department of Ophthal­mology, University of Colorado Health Sciences Center, Box B204, 4200 E Ninth Ave, Denver, CO 80262; fax: (303) 315-5644; e-mail: bronwyn.bateman@uchsc.edu

264 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 1997