evaluation of a difficult and potentially life-threateningproblem.

ANDREW G. LEE, MD

Houston, TexasPAUL W. BRAZIS, MD

Jacksonville, Florida

REFERENCE

1. Lee AG, Onan H, Brazis PW, Prager TC. An imaging guide tothe evaluation of third cranial nerve palsies. Strabismus1999;7:153–168.

AUTHOR REPLY

THE IMAGING GUIDELINES DRS LEE AND BRAZIS REFER TO1

were not published by the time of our original submissionor response. Their four queries address more general issuesof the clinical evaluation of patients with third cranialnerve palsy, instead of specific ones related to the role ofmagnetic resonance angiography, the focus of our Perspec-tive.

1. In deriving our guidelines, we considered the clinicalevaluation of the pupil separately from that of theextraocular muscles because the relative function ofeach is such an important tool for estimating thelikelihood that the third nerve injury is the result ofcompression or infarction.2 By complete extraocularmuscle impairment, we imply that all extraocularmuscles are totally paralyzed. By incomplete impair-ment, we imply that those extraocular muscles af-fected are not totally paralyzed, or that not all of theextraocular muscles are affected. Our rationale issimilar to that articulated by Drs Lee and Brazis, andreflected in our imaging guidelines, which acknowl-edge that a patient with an incomplete third cranialnerve palsy, regardless of the status of the pupil,might be harboring an aneurysm.

2. The issue of initially performing computed tomogra-phy (CT) or magnetic resonance imaging (MRI) inthis setting is often debated. Our imaging guidelineswere designed for patients with neurologically isolatedthird cranial nerve palsy. We doubt that a patientwith aneurysmal third cranial nerve compression andsubarachnoid hemorrhage would present with oph-thalmoplegia alone.3 Moreover, although we ac-knowledge that obtaining CT may be more practicalthan obtaining MRI in some emergent situations, wecaution that CT is often normal in patients withsubarachnoid hemorrhage, especially after the firstday of symptoms.4

3. Obviously, we agree with an individualized approachto these patients.

4. We also instruct the patient or the relatives of thepatient to check the pupil and report immediately if

it has enlarged. The reliability of this request, how-ever, remains untested and is no substitute for serialexaminations looking for delayed pupil dilation,which may occur with aneurysmal compression of theoculomotor nerve.5

DANIEL M. JACOBSON, MD

Marshfield, WisconsinJONATHAN D. TROBE, MD

Ann Arbor, Michigan

REFERENCES

1. Lee AG, Onan H, Brazis PW, Prager TC. An imaging guide tothe evaluation of third cranial nerve palsies. Strabismus1999;7:153–168.

2. Trobe JD. Third nerve palsy and the pupil. Footnotes to therule. Arch Ophthalmol 1988;106:601–602.

3. Jefferson G. Isolated oculomotor palsy caused by intracranialaneurysm. Proc R Soc Med 1947;40:419–432.

4. Schievink WI. Intracranial aneurysms. New Engl J Med1997;336:28–40.

5. Kissel JT, Burde RM, Klingele TG, Zeiger HE. Pupil-sparingoculomotor palsies with internal carotid-posterior communi-cating artery aneurysms. Ann Neurol 1983;13:149–154.

Modified Sutureless Sclerotomies inPars Plana Vitrectomy

EDITOR:

AFTER INITIAL REPORTS OF SUTURELESS SCLEROTOMIES

for vitrectomy,1,2,3 Kwok and associates4 (Am J Ophthal-mol 127:731–733, June 1999) recently advocated a modi-fied circumferential approach. We have also performed aconsecutive series using a circumferential approach andpresent a prospective, independent assessment of surgicalcomplications.

We performed 90 self-sealing sclerotomies on 30 con-secutive patients using the technique originally describedby Chen,1 but with a modified circumferential approach.The surgeon had performed approximately 60 self-sealingsclerotomies before commencing this prospective study.

A surgical proforma was completed by the surgeon at theend of each operation. Patients were assessed for thefollowing complications; intraoperative or end-operativewound leak (gas, fluid, or oil), wound extension or hem-orrhage, flap dehiscence, retinal dialysis or tear, andvitreous or retinal incarceration. Because of the high rateof wound leakage reported by Milibak and Suveges,2 allsclerotomies were closed at the end of the vitrectomy, witha single 7-0 absorbable suture, and the proforma was usedto record all complications occurring up to the point wherethe sclerotomies were sutured.

Five patients underwent pars plana vitrectomy for vit-reous hemorrhage, four had diabetic tractional retinaldetachments, seven had rhegmatogenous retinal detach-

AMERICAN JOURNAL OF OPHTHALMOLOGY116 JANUARY 2000

ments, five had rhegmatogenous retinal detachments withproliferative vitreoretinopathy, and nine had other in-dications, including giant retinal tear, removal of softlens matter, epiretinal membrane peel, removal of sub-retinal neovascular membrane, and infective or idio-pathic vitreitis.

Ten of the 30 patients (13 of 90 sclerotomies) had awound leak either during or at the end of the vitrectomy.One patient had a wound extension, and one had a flapdehiscence. One patient developed a wound hemorrhageat two sites. Two patients had vitreous incarceration in thewound, and in one of them, an associated retinal incarcer-ation was successfully disengaged. No retinal tears ordialyses occurred (Table 1). Nineteen patients had uncom-plicated self-sealing sclerotomies.

The risk of wound extension and flap dehiscence may bereduced by using a needle to puncture the internal scle-rotomy,4 although vitreous and retinal incarceration arepotentially problematic as the incarceration may occurwholly within the scleral tunnel and not appear at theexternal opening. In both of our patients, the vitreousincarceration was easily recognized, but a risk exists thatthe surgeon may fail to recognize vitreous or retinalincarceration if the internal opening is not carefullyinspected at the end of the vitrectomy.

We agree with Kwok and associates4 that a circumfer-ential approach makes access easier, and we observed asimilar rate of wound leakage. Our experience differs inthat we experienced complications other than wound leak.

TIM JACKSON, FRCOPHTH

London, United Kingdom

REFERENCES

1. Chen JC. Sutureless pars plana vitrectomy through self-sealingsclerotomies. Arch Ophthalmol 1996;114:1273–1275.

2. Milibak T, Suveges I. Complications of sutureless pars planavitrectomy through self-sealing sclerotomies [letter]. ArchOphthalmol 1998;116:119.

3. Chen JC. Reply to “Complications of sutureless pars plana

vitrectomy through self-sealing sclerotomies.” Arch Ophthal-mol 1998;116:119.

4. Kwok AKH, Tham CCY, Lam DSC, Li ML, Chen JC.Modified sutureless sclerotomies in pars plana vitrectomy.Am J Ophthalmol 1999;127:731–733.

AUTHOR REPLY

UNLIKE IN OUR SERIES, DR JACKSON AND ASSOCIATES

encountered complications, including wound extension,flap dehiscence, and vitreous and retinal incarceration. Asthey pointed out, these complications could be related totheir use of a myringo-vitreal-retinal blade, as described byDr Chen,1 to perforate the internal sclerotomy instead ofour use of a 20-gauge round-body hypodermic needle. Thismodification may be a major factor that helps to reducerisk of scleral flap laceration made by the sharp edges of themyringo-vitreal-retinal blade.

We agree with Dr Jackson and associates that vitreous oreven retinal incarceration can happen. In order to mini-mize the chance of incarceration, we routinely pinch offthe infusion cannula before instruments are withdrawnfrom the sclerotomy sites. Moreover, the assistant surgeonwould press the scleral flap tightly against the scleraltunnel to avoid wound leakage before infusion is resumed.

Dr Jackson and associates sutured all sclerotomy sites atthe end of vitrectomy. In our opinion, this method is a safeprophylactic measure, but seems unnecessary. We onlysuture those sites with leakage.

ALVIN K. H. KWOK, FRCS

CLEMENT C. Y. THAM, FRCS

DENNIS S. C. LAM, FRCS, FRCOPHTH

Shatin, Hong Kong

REFERENCE

1. Chen JC. Sutureless pars plana vitrectomy through self-sealingsclerotomies. Arch Ophthalmol 1996;114:1273–1275.

TABLE 1. Complication Rates

Patient

Complications

Complication

Rate (%)

Sclerotomy

Complications

Sclerotomy Complication

Rate (%)

Wound leakage 10 33 13 14

Wound extension 1 3 1 1

Wound dehiscence 1 3 1 1

Wound hemorrhage 1 3 2 2

Vitreous incarceration 2 7 2 2

Retinal incarceration 1 3 1 1

Retinal tears 0 0 0 0

Retinal dialysis 0 0 0 0

CORRESPONDENCEVOL. 129, NO. 1 117