Mechanisms of Intraocuular Pressure Elevation after Pars Plana Vitrectomy

Mechanisms of Intraocular Pressure Elevation after Pars Plana Vitrectomy DENNIS P. HAN, MD, HILEL LEWIS, MD, FRED H. LAMBROU, Jr., MD, WILLIAM F. MIELER, MD, ARTHUR HARTZ, MD, PhD

Abstract: A prospective study of 222 consecutive patients undergoing pars plana vitrectomy was done to determine the incidence and mechanisms of postoperative intraocular pressure (lOP) elevation. Within 48 hours of surgery, postoperative lOP increased by at least 5 to 22 mmHg in 136 eyes (61.3%) and to 30 mmHg or more in 79 eyes (35.6%). Presumed mechanisms of openangle glaucoma included intraocular gas expansion (28.4 %), inflammatory trabecular meshwork obstruction (4.5%), silicone oil-related glaucoma (3.6%), and erythroclastic glaucoma (2.2%) . Closed-angle mechanisms included pupillary block glaucoma (6.8%) and ciliary body edema (3.6%). Factors which were associated with postvitrectomy pressure elevation included placement of a scleral buckle, either intraoperatively (P = 0.003) or before vitrectomy (P = 0 .001), intraoperative scatter endophotocoagulation (P = 0.041), intraoperative lensectomy (P = 0.024) , and development of postoperative fibrin membranes (P = 0.038) . Surgery was required to lower lOP or relieve pupillary block in 25 eyes (11.3%). Ophthalmology 96:1357-1362, 1989

Transient or sustained elevation of intraocular pressure (lOP) is a frequent occurrence after pars plana vitrectomy.':" Severe pressure elevations may result in visual loss due to central retinal artery occlusion or optic nerve ischemia." There are many mechanisms for acute postvitrectomy pressure elevation including erythroclastic glaucorna.l-" ? inflammatory glaucoma , and neovascular glaucoma.Yln addition , specifictherapeutic interventions commonly used in conjunction with vitrectomy surgery, such as scleral buckling, endophotocoagulation, or use of expansile gases, may be associated with lOP elevation."" Preexisting glaucoma or postoperative complications such as fibrin format ion" and choroidal hemorrhage or detachment I may also increase the risk of lOP

Originally received: September 30 , 1988. Revision accepted: March 7, 1989. From the Departm ent of Ophthalmology, Eye Institute, Medical College of Wisconsin, Milwaukee. Supported in part by an unrestricted grant from Research to Prevent Blindness, Inc, and by Ophthalmic Research core grant EY01931. Reprint reque sts to Dennis P. Han, MD, Medical College of Wiscons in, 8700 West Wiscons in Ave, Milwaukee, WI 53226.

elevation. However, the relative contribution of each of these factors is unclear. We attempted to determine prospectively, the severity, incidence, and mechanisms of postoperative lOP elevation in 222 consecutive patients undergoing pars plana vitrectomy in conjunction with various adjunctive techniques . Preoperative, intraoperative, and postoperative factors were also assessed to determine their influence on the risk of significant postvitrectomy lOP elevation.

MATERIALS AND METHODS The study was done at the Medical Collegeof Wisconsin between July 1986 and June 1987. The main indications for surgery are listed in Table 1. All eyes underwent three-port pars plana vitrectomies with sclerotomies in the region between 2.5 and 4.0 mm posterior to the limbus . In cases requiring lensectomy, phacofragmentation and aspiration of the lens was done with a 19-or 20-gauge ultrasonic fragmentation-aspiration needle inserted through the pars plana. A posterior vitrectomy was then done with an aspiration-cutting vitrectom y instrument and separate endoilluminator. In 1357

OPHTHALMOLOGY

•

SEPTEMBER 1989

Table 1. Indications for Pars Plana Vitrectomy in 222 Eyes Preoperative Diagnoses Rhegmatogenous RD RD with PVR ~ grade C Diabetic vitreous hemorrhage Diabetic traction RD Idiopathic macular pucker Posterior segment trauma Miscellaneous Endophthalmitis Posterior uveitis Nondiabetic vitreous hemorrhage Posterior chamber lens dislocation Nondiabetic traction RD Aphakic and pseudophakic cystoid macular edema Removal of massive subretinal hemorrhage Dislocated crystalline lens Impending macular hole formation Total no. of patients

No. of Eyes

(%)

(11.2) (24.8) (16.7) (11.2) (9.5) (12.2) (14.4)

25 55

37

25 21

27

32 8

6 4 3

3 3

2 2 1 222

(100.0)

RD = retinal detachment; PVR = proliferative vitreoretinopathy. Table 2. Protocol for Management of Postvitrectomy Intraocular Pressure Elevation lOP (mmHg)

,,;25

26-30 (inclusive) 31-40 (inclusive) If unresponsive to above medications and lOP (mmHg) is:

Intervention Observation Timoptic or Betoptic one drop two times daily Above regimen plus carbonic anhydrase inhibitor systemically Ocular paracentesis

~40

>35 if corneal edema or pain present >25 if patient has lost light perception lOP

=

intraocular pressure.

cases requiring scleral buckling, hard silicone episcleral implants were used, and 5-0 nylon horizontal mattress sutures were placed to imbricate the sclera. In most cases, two or three spots of peripheral retinal cryopexy were placed prophylactically just posterior to each of the sclerotomy sites. Fluid-gas exchanges, if necessary, were done with a continuous fluid-air exchange unit (CFAE-20, Trek Medical Products, Mukwonago, WI), and insufflation of the vitreous cavity was completed with varying mixtures of room air, sulfur hexafluoride gas (20-25% in room air), or perfluoropropane gas (15-20% in room air). Preoperative evaluation consisted of recording the patient's age, sex, race, ocular laterality, preoperative diagnosis, prior surgical procedures, and preexisting history of glaucoma. Preoperative examination consisted ofi (1) 1358

•

VOLUME 96

•

NUMBER 9

measurement of the Snellen visual acuity, with and without a pinhole, with the patient wearing the most recent spectacle correction, (2) Goldmann applanation tonometry, (3) gonioscopy with grading of anterior chamber angle width,"? (4) slit-lamp examination with recording of the status of the lens (whether phakic, aphakic, or pseudophakic), and (5) determination of preoperative rubeosis of the angle or iris. Indirect ophthalmoscopy was done, and the presence or absence of a preexisting scleral buckle was recorded. Intraoperative factors were recorded, including whether scleral buckling (encircling type), scatter endophotocoagulation, pars plana lensectomy, fluid-gas exchange, silicone oil injection, iridectomy, or sodium hyaluronate injection were done. In eyes receiving fluid-gas exchange, the type and concentration of gas (perfluoropropane, sulfur hexafluoride, or room air) was recorded. In eyes receiving sodium hyaluronate, the amount and location (anterior vs. posterior segment) of the injection was recorded. Occurrence of intraoperative and early postoperative complications of vitreous hemorrhage, hyphema, and choroidal detachment were also noted. Postoperative evaluation consisted ofIOP measurement by applanation pneumotonometry at 2 to 4,6 to 8, 12 to 24, and 36 to 48 hours after surgery. (Although we followed a number of eyes for 72 hours or more, most study eyes had maximum measured pressures in the 12-to-24 hour interval, resulting in most hospital discharges occurring by the end of the 48-hour interval.) In eyes with postoperative lOP elevation, gonioscopy and slit-lamp examination were done to determine whether the angle was open, closed, or narrowed. Elevation of lOP was treated medically according to a standard protocol (Table 2) and surgically according to clinical judgment of the mechanism of glaucoma involved. Patients with preexisting glaucoma were treated according to the same protocol, except that their current glaucoma medications, excluding miotics, were continued pre- and postoperatively. Late follow-up examination was obtained at least 6 weeks after surgery and consisted ofmeasurement of the Snellen visual acuity, Goldmann applanation tonometry, and indirect ophthalmoscopy. CLASSIFICATION OF GLAUCOMA MECHANISMS

We classified the various mechanisms ofIOP elevation based on previous clinical descriptions of postvitrectomy glaucoma 1-3,5.6 and our observations on the anatomic configurations of eyes in which acute postvitrectomy lOP increases develop and their response to therapeutic intervention. We defined "gas expansion" as elevated lOP in the presence of an expansile intraocular gas bubble, with an open angle, absence of iris bombe with free access of intravitreal fluid into the anterior chamber, minimal inflammation (,,; +2 cell flare), and absence of inflammatory or cellular debris on the trabecular meshwork. We also included in this category eyes with fibrin formation but without evidence of pupillary block. We defined "pupillary block" as elevated lOP in the presence of a closed or narrowed anterior chamber angle, with access to the an-

HAN et al •

POSTVITRECTOMY PRESSURE ELEVATION

terior chamber of intravitreal fluid limited by a fibrin membrane, intraocular lens, or gas or silicone bubble blocking the pupillary space . We distinguished these cases from those defined as " iridocorneal apposition." In the latter, deforming forces from an intraocular gas bubble, rather than an isolated pupillary block mechanism, push the iris into complete apposition with the cornea in an aphakic eye.' ? Here the bubble could be seen immediately to displace this iris against the cornea whenever in contact with it. Although initially reversible , iridocorneal apposition eventually becomes permanent, causing secondary angle-closure glaucoma. "Ciliary bod y edem a" was characterized by a deep or slightly shallowed central anterior chamber without iris bombe and a closed or narrowed chamber angle due to peripheral, anterior rotation of the ciliary body. An "inflammatory" mechanism was characterized by lOP elevation with inflammatory debris or marked cell and flare (+ 3 or worse) with an open-chamber angle. " Erythroclastic glaucoma" was defined as the presence of degenerated erythrocytes blocking the trabecular meshwork with an open-chamber angle. "Corticosteroidinduced" lOP elevation was diagnosed when the patient had received long-standing preoperative ocular administration of corticosteroid eyedrops and had continuous postoperative lOP elevation which was reversed by discontinuing the drops, often days or weeks later. " Posttraumatic angle injury" was diagnosed when angle recession, iridodialysis, or disruption ofthe iris near the anterior chamber angle was noted pre- or intraoperatively. " Silicone oil-related" glaucoma was defined by elevated lOP due to silicone oil-induced pupillary block or elevated lOP in the presence of a silicone oil-filled vitreous cavity without an y of the other previously mentioned mechanisms. STATISTICAL ANALYSIS

The incidence, severity , and mechanisms of lOP elevation were determined for all 222 eyes within each diagnostic category. We applied the term "postvitrectomy lOP elevation" to eyes whose maximum-measured pressure in the first 48 hours after surgery exceeded the preoperative measurement by 5 mmHg or more. We categorized the severity of postvitrectomy lOP elevation as: (1) mild, if the maximum-measured pressure was 22 to 29 mmHg; (2) moderate, if the maximum-measured pressure was 30 to 39 mmHg, and (3) severe, if the maximum-measured pressure was 40 mmHg or more. For analysis, we further defined postvitrectomy lOP elevation as " significant" if the maximum-measured pressure fell into the moderate or severe groups. We attempted to determine the mechanism of lOP elevation in all eyes with mild elevation or more. Chi-square tests were done to determine if any of the recorded pre-, intra-, or postoperative factors were associated with a statistically significant increased risk of postvitrectomy lOP elevation. Factors which appeared to be statistically significant were an alyzed by multiple stepwise logistic regression to determine whether they influenced the risk of significant postvitrectomy lOP elevation independently. Fisher's exact test was

Table 3. Preoperative, and Early and Late Postoperative Intraocular Pressure Measurements in Vitrectomized Eyes lOP (mmHg)

Preoperative lOP Early postoperative lOP Late postoperative lOP

No. of Eyes

Mean/SO

Range

P*

198 222 132

26.9/8.8

16.3/7.4

3-70 10-60 0-31

0.0001

15.4/5.0

NS

lOP = intraocular pressure; SO = standard deviation; NS = not significant. * Paired t test for difference between pre- and postoperative lOPs.

used to determine the significance ofapparent associations between certain mechanisms ofIOP elevation and specific perioperative factors when the expected values were too small for chi-square testing. Correlations among the preoperative, maximum-measured early postoperative (within 48 hours of surgery), and late postoperative lOPs were made using Spearman's rank correlation test. Determination as to whether the last follow-up visual acuity (as measured by the change in the minimal angle of resolution [log MAR]) or final lOP were influenced by postvitrectomy pressure elevation was done using analysis of variance and the paired l test. Comparisons between the preoperative, early postoperative, and late postoperative lOPs for the group as a whole were made using the independent l test. P values ofless than 0.05 were considered significant.

RESULTS The median patient age was 55 years (range, 4-84 years), and 126 patients were male and 96 female . There were 197 white patients, 19 black, and six of other racial origin. Of the 222 study eyes, 147 were phakic, 36 were aphakic, and 39 were pseudophakic. In 24 eyes, we were unable to measure the preoperative lOP due to an open corneoscleral wound. Follow-up of 6 weeks or more was obtained in 132 patients (median, 14.6 weeks; range , 685 weeks). EFFECT OF VITRECrOMY ON lOP

The means, standard deviations, and rang es of the preoperative, maximum-measured, and late postoperative lOPs for the entire group are shown in Table 3. For the entire group, a significant difference between the preoperati ve and maximum-measured early postoperative lOPs was found (P = 0.0001, paired l test), with an average lOP elevation of 10.6 mmHg. No significant difference was found between the preoperative and late pressure measurements for the group as a whole . However, there was a significant correlation between earl y postoperative lOP elevation from baseline and late postoperative lOP elevation from baseline values (r = 0.492 , P = 0.000 I, Spearman's rank correlation test), indicating a tendency 1359

OPHTHALMOLOGY

•

SEPTEMBER 1989

•

VOLUME 96

•

NUMBER 9

Table 4. Factors Associated with Significant Acute Postvitrectomy Intraocular Pressure Elevation " Frequency of Glaucoma (%) Factor Not Present

Factor Present

• Maximum-measured lOP within 48 hours postoperatively

~30

mmHg and

Table 5. Mechanisms of Acute Postvitrectomy Intraocular Pressure Elevation in 222 Eyes Frequency (no. of eyes) Open-angle mechanisms Gas expansion Inflammatory Silicone oil-related (without pupillary block) Corticosteroid response Erythroclastic Closed-angle mechanisms Pupillary block Gas-pupil block Fibrin-pupil block IOL-induced pupil block Silicone pupil block Subtotal (pupillary block) Ciliary body edema Iridocorneal apposition Preexisting mechanisms Undetermined Combined Total IOL

=

(%)

63 10

(28.4) (4.5)

8 6 5

(3.6) (2.7) (2.2)

4

7

2

2 15

(6.8)

8 3 18 5

-5

(3.6) (1.4) (8.1) (3.2) (-3.2)

136

(61.3)

intraocular lens.

for eyes with greater early postoperative lOP elevation to sustain a greater lOP elevation from baseline 6 weeks or more after surgery. INCIDENCE AND RISK FACfORS FOR POSTVITRECfOMY PRESSURE ELEVAnON

The incidence of significant and severe postvitrectomy lOP elevation within each diagnostic category are indicated respectively as follows: proliferative vitreoretinopathy, 54.6 and 16.4%; rhegmatogenous retinal detachment, 40 and 0%; posterior segment trauma, 33.3 and 11.1%; diabetic vitreous hemorrhage, 32.4 and 2.7%; diabetic traction detachment, 20 and 4%; and idiopathic macular pucker, 9.5 and 0%. Perioperative factors predicting a significantly increased risk of acute postvitrec1360

49/167 47/163 53/175 50/172 26/110

30/55 (54.6) 32/59 (54.2) 26/47 (55.3) 29/49 (59.2) 53/112 (47.3)

Previously placed scleral buckle Intraoperative scleral buckle Intraoperative lensectomy Postoperative fibrin formation Intraoperative scatter photocoagulation

~5

(29.3) (28.8) (30.3) (29.1) (23.6)

Multivariate Logistic Regression (P)

0.001 0.003 0.024 0.038 0.041

mmHg increase from preoperative measurement.

tomy lOP elevation by both chi-square testing and multivariate logistic regression include intraoperative or previous placement of a scleral buckle, intraoperative scatter endophotocoagulation, intraoperative pars plana lensectomy, and postoperative fibrin formation . The incidence of postvitrectomy pressure elevation associated with each of these factors and the corresponding P values are shown in Table 4. The presence of any of these factors increased the risk of significant lOP elevation approximately twofold. MECHANISMS OF ACUTE POSTVITRECfOMY lOP ELEV AnON

The incidence of the various mechanisms of acute postvitrectomy lOP elevation are shown in Table 5. Of 222 study eyes, 136 eyes (61.3%) had at least mild postvitrectomy lOP elevation, 79 eyes (35.6%) had significant elevation , and 17 eyes (7.7%) had severe elevation. Of 136 eyes with at least mild lOP elevation , 92 eyes (67.6%) had an open-angle configuration, 27 eyes (19.8%)had a closedor narrow-angle configuration, and two eyes (1.3%) had a combined appearance from preexisting peripheral anterior synechiae. In the remaining 15 eyes, the angle configuration could not be determined because of poor visualization of the angle structures with gonioscopy or inability of the patient to cooperate during gonioscopic examination. Certain perioperative factors appeared to be associated with characteristic mechanisms of glaucoma. Of five eyes with erythroclastic glaucoma , four had had nonclearing diabetic vitreous hemorrhage. The association between erythroclastic glaucoma and diabetic vitreous hemorrhage was statistically significant (P = 0.003, Fisher's exact test). The remaining eye had undergone vitrectomy to remove subretinal hemorrhage due to age-related macular degeneration." We also found a statistically significant association (P < 0.00 I, chi-square test) between the presence of postoperative fibrin formation and pupillary block glaucoma (all types combined). Four of eight eyes (50%) undergoing vitrectomy for endophthalmitis had inflammatory glaucoma compared with only 6 of 214 eyes(2.8%) without endophthalmitis. This association was statistically significant. (P < 0.001, chi-square test, Yates corrected). We also found a higher incidence of angle-closure glaucoma due to ciliary body edema in eyes with previous or

HAN et at

•

POSTVITRECTOMY PRESSURE ELEVATION

Table 6. Surgical Management of Postvitrectomy Intraocular Pressure Elevation-Indications and Results Mechanism of lOP Elevation Gas expansion Fibrin-induced pupillary block Silicone oil-, 10L-, or gasinduced pupillary block Ciliary body edema

Mean lOP before Treatment (mmHg)

Mean lOP within 24 hrs after Treatment (mmHg)

Treatment

No. of Eyes

Anterior chamber paracentesis Argon laser memoranotomv"

10

5

38.1 31.4

29.8 22.0

5 5

32.0 33.4

22.2 21.0

Argon laser iridotomy Argon laser iridoplasty17

10L = intraocular lens.

intraoperative placement of a scleral buckle (6 of 104 eyes, 5.8%) compared with eyes without a scleral buckle (2 of 118 eyes, 1.7%). The difference between these two groups was not statistically significant (P = 0.151 , Fisher's exact test). Of 30 eyes with a history of glaucoma , 11 had at least mild postvitrectomy lOP elevation due to the mechanism responsible for their glaucoma history. Nine eyes (30%) had significant postvitrectomy lOP elevation compared with 70 of 192 eyes (36.5%) without a history of glaucoma. Therefore, although preexisting mechanisms appeared to be a significant cause of postvitrectomy lOP elevation , a history of glaucoma did not increase the overall rate of postvitrectomy lOP elevation. SURGICAL MANAGEMENT OF ACUTE POSTVITRECTOMY PRESSURE ELEVAnON

Surgical intervention was required in 25 eyes (11.3 %) with postoperative lOP elevations uncontrolled by medical therapy alone. The indications and results of treatment are described in Table 6. Adequate treatment required recognition of the mechanism of lOP elevation and use of a treatment modality for that mechanism. In all cases, ocular paracentesis or argon laser treatment lowered lOP to acceptable levels. One eye required repeated, but successful,laser membranotomy 3 days after initial treatment due to recurrence of a fibrin membrane. The occurrence of significant acute postvitrectomy lOP elevation did not influence the final visual acuity 6 weeks or more after the surgery (n = 127). Eyes without acute postvitrectomy lOP elevation gained a mean of 8.2 lines of visual acuity, compared with eyes with mild, moderate, and severe elevation, (7.6, 11.1, and 4.9 lines of visual acuity, respectively). These differences were not statistically significant (by analysis of variance).

DISCUSSION We found that significant, acute lOP elevation

(;dO mmHg) was a frequent complication after pars plana vitrectomy, occurring in 35.6% of the study population.

This incidence is higher than that of previous large series ofvitrectomized eyes3,7 in which acute postvitrectomy lOP elevation occurred in 20 to 28%, due primarily to ery-

throclastic glaucoma or hyphema. In a separate series of

206 eyes, late postvitrectomy glaucoma occurred in 26%

of eyes, due to neovascular and open-angle glaucoma . 1 In these studies, the indication for surgery in appro ximately half of each of the study populations was diabetic retinopathy. The higher incidence of acute lOP elevation in our study probably reflects a different spectrum of diseases and interventions within our study population. In addition more frequent , scheduled lOP measurements were made which may have detected otherwise unrecognized lOP elevations. The lower proportion of diabetic eyes in our study may also explain why we saw a comparatively mild effect of vitrectomy on late lOP. The incidence of sustained lOP elevation from neovascular glaucoma is likely to be lower in a population with fewer diabetic eyes. Previous or intraoperative scleral buckling or intraoperative scatter endophotocoagulation appeared to increase the risk of postvitrectomy lOP elevation significantly and independently. In many cases, this is probably due to the inherent ability of each of these modalities to cause elevated lOP by ciliary body congestion or choroidal detachment with secondary angle closure. Angle-closure glaucoma is estimated to occur in 6 to 22% of eyes undergoing single-session panretinal fundus photocoagulation,13,20 and 3.6% of eyes undergoing scleral buckling with episcleral implants." Although a scleral buckle was present in only 46.8 % (104 of 222) of the eyes in this study, we had a similar incidence of ciliary body edema (3.6%) as did Perez et al? in a series of buckled, nonvitrectomized eyes. Our study suggeststhat vitrectomized eyesmay develop lOP elevation equally, if not more likely, in response to scleral buckling or fundus photocoagulation , as do nonvitrectomized eyes undergoing these procedures. Operative lensectomy also appeared to predispose the eye to glaucoma, through mechanisms not entirely clear to us. Such a phenomenon could be explained by increased inflammation, uveal congestion from increased manipulation of the eye by the surgical sclerotomies, or obstruction of trabecular meshwork outflow by lens particles or uveal pigment dispersion from ultrasonic phacofragmentation. Although we did not observe that our patients accumulated significant amounts of debris on the trabecular meshwork, we cannot exclude physiologic obstruction by material which was invisible or cleared before our gonioscopic examination. 1361

OPHTHALMOLOGY

•

SEPTEMBER 1989

Our study confirms the findings of a previous study by Abrams et al." They showed a significant association between postvitrectomy fibrin formation and postoperative lOP elevation. We have characterized the fibrin response further as directly causing postvitrectomy lOP elevation in some cases through pupillary block. The frequent occurrence of this complication has not been reported in previous large series of vitrectomized eyes. Isolated cases of postvitrectomy fibrin-pupillary block glaucoma have been managed with surgical disruption of the pupillary membrane with a needle, argon laser membranotomy," and more recently, intraocular injection of tissue plasminogen activator." We found a strong correlation between early and late postoperative lOP elevation in our study population, suggesting that eyes which sustain acute lOP elevation are more likely to have a persistent increase in lOP lasting 6 weeks or more (median, 14.6 weeks) after vitrectomy. Nevertheless, we found no significant differences between initial and final pressures for the group as a whole, suggesting that the magnitude of this late effect is relatively small in comparison with the magnitude ofIOP elevation seen acutely after vitrectomy. We attribute this primaril y to the transient and reversible mechanisms involved (ciliary body edema, gas expansion, corticosteroid response, erythroclastic glaucoma, etc.) and the early intervention used to prevent the occurrence of synechial angle closure in eyes with closed-angle glaucomas (e.g., pupillary block, ciliary body edema). Although we did not find that the occurrence of acute postvitrectomy lOP elevation had an effect on the final visual acuity, the relatively short median follow-up duration of 14.6 weeks was not sufficient to assess the long-term effects on visual function by either acute or chronic lOP elevations. Such an assessment would also require more sophisticated methods of measuring visual function (e.g., visual fields) which were not investigated in this study. In addition, the significance of a moderate lOP elevation in nonglaucomatous eyes is uncertain. Our classification of postvitrectomy lOP elevation should be considered syndromic rather than etiologic in nature, because: (1) it assumes an underlying etiology based on our present understanding of pressure-elevation mechanisms in vitrectomized eyes; and (2) it is based primarily on clinical findings which are consistent with, but do not prove, the presumed etiology. This consideration is especially valid in the classification of open-angle mechanisms, in which the anatomic configuration of the anterior chamber angle may not discriminate between individual mechanisms or superimposed mechanisms. Nevertheless, we found our classification to be useful in determining, by clinical examination, the most likely cause of lOP elevation and the best course of treatment in a given patient. Acute lOP elevation is a frequent postoperative complication of pars plana vitrectomy and is associated with several surgical modalities used in conjunction with vitreous surgery, including scleral buckling, endophotoco-

1362

•

VOLUME 96

•

NUMBER 9

agulation , and pars plana lensectomy. Postoperative fibrin formation also appears to increase the likelihood of postvitrectomy lOP elevation . Recognition of the various mechanisms responsible, including those involving angle closure and potential synechial formation, is important to reduce the likelihood of long-term lOP elevation and prevent the visual complications which may result from a severe postoperative lOP rise.

1. Aaberg TM, Van Hom DL. Late complications of pars plana vitreous surgery. Ophthalmology 1978; 85:126-40. 2. Wilensky JT, Goldberg MF, Alward P. Glaucoma after pars plana vitrectom y. Trans Am Acad Ophthal Otolaryng 1977; 83:0P114-21 . 3. Weinberg RS, Peyman GA, Huamonte FU. Elevation of intraocular pressure after pars plana vitrectomy. Albrecht von Graefes Arch Klin Exp Ophth alrnol 1976; 200:157-61 . 4. Abram s GW, Swanson DE, Sabates WI, Goldman AI. The results of sulfur hexafluoride gas in vitreous surgery. Am J Ophthalmol 1982; 94:165-71 . 5. Campbell DG, Simmons RJ, Grant WM. Ghost cells as a cause of glaucoma. Am J Ophthalmol 1976; 81:441-50. 6. Campbell 00. Simmons RJ, Tolentino FI, McMeel JW. Glaucoma occurring after closed vitrectomy. Am J Ophth almol 1977; 83:63-9. 7. Ghartey KN, Tolentino FI, Freeman HM, et al. Closed vitreous surg ery. XVII. Results and complicat ions of pars plana vitrectomy. Arch Ophthalmol1980; 98:1248-52. 8. Blankenship G. Preoperative iris rubeosis and diabetic vitrectomy results. Ophthalmology 1980; 87:176-82. 9. Perez RN, Phelps CD, Burton TC. Angle-closure glaucoma following scleral buc kling operations . Trans Am Acad Ophthal Otolaryng 1976; 81:247-52. 10. Phelps CD, Burton TC. Glaucoma and retinal detachment. Arch OphthalmoI1977; 95:418-22. 11. Lewis H, Han D, Williams GA. Management of fibrin pupillary-bloc k glauc oma after pars plana vitrectomy with intravitreal gas injection. Am J Ophthalmol1987; 103:180-2. 12. Han DP, Lewis H, and Williams GA. Management of complete iridocomeal apposition after vitrectomy [Letter]. Am J Ophtha/moI1987; 103:108-9. 13. Doft BH, Blankenship Gw. Single versus multiple treatment sessions of argon laser pametinal phot ocoagulatio n for proliferative diabet ic retinopathy . Ophthalmology 1982; 89:772-9. 14. Faulbom J. Conway BP, Machemer R. Surgical complications of pars plana vitreous surgery. Ophthalmology 1978; 85:116-25. 15. Killey FP, Edelhauser HF, Aaberg TM. Intraocular sulfur hexafluoride and octofluorocyclobutane . Arch Opthalmol 1978; 96:511-5. 16. Sebestyen JG. Fibrinoid syndrome: a severe complication of vitrectomy surgery in diabetics. Ann Ophthalmol1982; 14:853-6. 17. Kolker AE, Hetherington J Jr. Becker-Shaffer's Diagnosis and Therapy of the Glaucomas, 5th ed. St. Louis: CV Mosby, 1983; 42. 18. Hanscom TA, Diddie KR. Early surgical drainage of macular subretinal hemorrhage . Arch Ophthalmol1987; 105:1722-3. 19. Burton TC, Folk JC. Laser iris retraction for angle closure glaucoma after retinal detachment surgery. Ophthalmology 1988; 95:742-8. 20. Huamonte FU, Peyman GA. Goldbe rg MF, Lock etz A. Immediate fundus compli cations after retinal scatter photoc oagulation. I. Clinical picture and pathogene sis. Ophthalmic Surg 1976; 7(1):88-99. 21. Williams GA, Lambrou FH, Jaffe GA, et aI. Treatment of postvitrectomy fibrin formation with intraocular tissue plasminog en activator. Arch Ophthalmol 1988; 106:1055-8.