Peripheral Cryoablation for Treatment of Active Pars Planitis: Long-Term Outcomes of a Retrospective Study

ELLIOTT H. SOHN, BENJAMIN C. CHAON, DOUGLAS A. JABS, AND JAMES C. FOLK

PURPOSE: To compare the long-term outcomes of pe- CONCLUSIONS: These data suggest that peripheral ripheral retinal cryoablation to conventional treatment retinal cryoablation therapy is an effective treatment for active pars planitis. for active pars planitis and may be better than DESIGN: Retrospective, interventional, comparative conventional regional injections and oral case series. corticosteroid therapy for induction of remission. (Am METHODS: Review at a single institution was conduct- J Ophthalmol 2016;162:35–42. Ó 2016 by Elsevier Inc. ed to compare the effect of cryotherapy to eyes with pars All rights reserved.) planitis to those receiving conventional therapy (topical, regionally injected, or oral corticosteroid therapy). Best- corrected visual acuity (VA), complications, resolution ARS PLANITIS IS A SUBSET OF INTERMEDIATE of cystoid (CME), and anterior chamber characterized by chronic vitreous inflammation and and vitreous inflammation were assessed. P the accumulation of fibroinflammatory debris over RESULTS: One hundred thirty-six eyes were treated the , occurring in the absence of any associated 1 conventionally, 50 eyes were treated with cryotherapy. infection or systemic disease. Clinically, these collections Median follow-up was 60.8 months (range 8.1– of inflammatory material resemble yellow-gray or white ag- 223.1 months) in the cryotherapy group and 45.0 months gregates visible in the mid- and posterior vitreous and col- (range 3.1–339.0 months) in the controls. There were no lections of inflammatory debris on the pars plana significant differences in baseline VA, anterior chamber (particularly inferiorly), known as ‘‘snowballs’’ and ‘‘snow- and vitreous inflammation, presence of CME, and prior banks,’’ respectively. The most frequent structural compli- use of regional corticosteroid injections. VA improved cation of pars planitis is macular edema, and most over time in the cryotherapy group (slope of L0.0018 treatment regimens are directed at controlling macular 2,3 logMAR units per month; P [ .023) but declined in edema. Other, less common structural complications the controls (slope of D0.0011 logMAR units per include traction retinal detachments and vitreous month; P [ .023). Kaplan-Meier survival estimates hemorrhage from neovascularization of the snowbank. demonstrated faster times to resolution of anterior cham- Although a subset of patients with pars planitis do not ber cell, vitreous cell, and CME in the cryotherapy- have macular edema or and may be treated eyes. Hazard ratios of remission (adjusted for observed, the majority of patients need treatment. Con- confounding factors) for vitreous cell and CME for those ventional treatment regimens typically have used a step- treated with cryotherapy compared to controls were 4.73 ped approach starting with regional corticosteroid (95% confidence interval 1.63, 13.63; P [ .004) and injections (either periocular or intravitreal), followed by 6.85 (95% confidence interval 1.06, 44.78; P [ .044), oral , supplemented with immunosuppres- 2,3 respectively. No ocular complications were identified in sion as needed for more difficult or bilateral cases. the cryotherapy group. Regional corticosteroid injections have the potential for elevated and formation/ progression and typically have a duration of effect of 3 months, thereby requiring repetitive injections. Accepted for publication Nov 5, 2015. Although when used properly, oral corticosteroids and From the Department of and Visual Sciences, Stephen immunosuppression can be administered with minimal A. Wynn Institute for Vision Research, University of Iowa Hospitals and Clinics, Iowa City, Iowa (E.H.S., B.C.C., J.C.F.); Departments of to no increase in side effects in adults, in children long- Ophthalmology and Medicine, Icahn School of Medicine at Mount term use of oral corticosteroids is avoided owing to the Sinai, New York, New York (D.A.J.); and Department of Epidemiology, risk of growth retardation.4–6 Furthermore, the chronic The Johns Hopkins University Bloomberg School of Public Health, Baltimore, Maryland (D.A.J.). nature of the disease often requires long-term therapy; Benjamin C. Chaon is now at the University of Minnesota Department there is a low sustained, drug-free remission rate; and of Ophthalmology, Minneapolis, Minnesota. some patients’ uveitis will not be controlled by this Inquiries to Elliott H. Sohn, Department of Ophthalmology, University 7 of Iowa Hospitals and Clinics, 200 Hawkins Dr, Iowa City, IA 52242; approach. Although some clinicians have advocated e-mail: [email protected] therapeutic vitrectomy for pars planitis, a review

0002-9394/$36.00 Ó 2016 BY ELSEVIER INC.ALL RIGHTS RESERVED. 35 http://dx.doi.org/10.1016/j.ajo.2015.11.009 concluded that methodological problems and study con- for this project, which adhered to the principles of the founders resulted in insufficient evidence to support or Declaration of Helsinki. refute its use.8 In 1973, Aaberg and associates reported Data were extracted from review of patient charts at the in 23 eyes in 14 patients affected by pars planitis that pe- time of pars planitis diagnosis (the ‘‘baseline’’ visit) and at ripheral retinal ablation directed at the inflammatory ex- consecutive follow-up visits. Follow-up visits were grouped udates overlying the pars plana with transconjunctival as sequential 3-month ‘‘contiguous visit windows’’ extend- cryopexy (cryotherapy) was effective at decreasing vitritis ing until the last follow-up visit (eg, the 3-month follow-up and improving visual acuity.9 These observations subse- window extended from 1.5 to 4.5 months post-baseline, the quently were supported by data from other small case se- 6-month follow-up visit window from 4.5 to 7.5 months ries; however, these studies evaluated the effect of post-baseline, etc). Data collected at presentation, base- cryotherapy on relatively few eyes and were limited by a line, and follow-up visits included (1) best-corrected relatively short duration of follow-up.10–13 Snellen visual acuity; (2) presence or absence of cystoid There are limited data on the long-term effectiveness of macular edema identified by stereoscopic funduscopic ex- peripheral retinal cryoablation therapy in controlling amination and/or fluorescein angiography, which was intraocular inflammation, improving visual acuity out- confirmed by optical coherence tomography imaging comes, and reducing structural complications of pars plani- when this modality became available (2004); (3) grading tis. Our retrospective study compared the long-term of anterior chamber cell; (4) grading of vitreous chamber clinical outcomes of eyes receiving cryotherapy to those cell; (5) lenticular status (phakic, aphakic, pseudophakic); receiving conventional therapy for active pars planitis. (6) use of topical corticosteroid medications; (7) use of oral We sought to determine if cryotherapy resulted in better vi- corticosteroid medications; (8) use of periocular corticoste- sual outcomes and faster time to resolution of cystoid mac- roid injections within the previous 120 days; (9) treatment ular edema, vitreous, and anterior chamber cell compared of with topical medication; (10) interval glau- to conventional therapy. coma surgery; (11) interval ; and (12) pres- ence of . Also noted in treated eyes was the date of the first cryotherapy and the total number of cryotherapy treatments per eye. Visual acuity data were converted to logMAR (loga- METHODS rithm of the minimal angle of resolution) values (logMAR ¼ log [1/Snellen visual acuity]) to allow for sta- A RETROSPECTIVE, INTERVENTIONAL, COMPARATIVE tistical analysis. The following logMAR values were study of all known patients with active pars planitis treated assigned: for hand motion, 3.3 logMAR; light perception, in the Department of Ophthalmology and Visual Sciences 4.3 logMAR; and no light perception, 5.3 logMAR. Each at the University of Iowa from January 1, 1983 (near the difference of þ0.3 logMAR between the baseline and time cryotherapy for pars planitis was first used at the Uni- follow-up visual acuity corresponds to a loss of 3 lines on versity of Iowa) to January 1, 2013 was conducted. The re- an Early Treatment Diabetic Study (ETDRS) cords of patients with ICD-9 diagnosis code 363.21 eye chart and a doubling of the visual angle.10,11 corresponding to a diagnosis of pars planitis were reviewed Much of the grading of anterior chamber cell and vitre- for inclusion in this study. Included patients met the ous chamber cell was performed prior to publication of the following criteria: (1) presence of intraocular inflammation Standardization of Uveitis Nomenclature’s (SUN) publica- in the vitreous in at least 1 eye; (2) presence of inflamma- tion of their standardized grading schema.1 However, the tory vitreous debris with snowbanks on the pars plana of the grading used appears to correspond reasonably well with inflamed eye; and (3) a follow-up ophthalmic evaluation the SUN schema. The early grading used a 1 3 1mm performed at least 3 months after the diagnosis of pars beam on the brightest illumination on slit-lamp bio- planitis. Exclusion criteria included (1) serologic evidence microscopy and had the following grades: 0 for no cells, of an infectious etiology for the ocular inflammation or a 0.5þ for 1–2 cells, 1þ for 5–10 cells, 2þ for 10–20 cells, known associated systemic disease (eg, or multi- 3þ for 20–30 cells, and 4þ for over 30 cells. Anterior ple sclerosis); (2) clinical history or physical examination chamber and vitreous inflammatory activity outcomes finding(s) to suggest a diagnosis other than pars planitis were dichotomized based on a threshold of anterior cham- (eg, oral lesions in patients with Behc¸et disease); (3) inter- ber cell or vitreous chamber cell ranking less than or equal mediate uveitis without evidence of snowbanks; (4) to 0.5þ, corresponding to ‘‘trace’’ inflammatory activity or isolated anterior or posterior uveitis; and (5) significant dia- less. The proportion of patients with anterior chamber cell betic retinopathy, macula-off retinal detachment, uncon- or vitreous cell ranking less than or equal to 0.5þ at base- trolled glaucoma, or a disease other than pars planitis line and the rate at which patients’ anterior chamber and that could account for a loss of visual acuity. Only de- vitreous inflammation fell below this threshold was identified data were used in the analyses. Institutional Re- compared between eyes treated with cryotherapy and view Board approval at the University of Iowa was obtained eyes treated with conventional therapy alone.

36 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 2016 FIGURE 1. Technique for an eye with active pars planitis undergoing external cryoablation. (Left) Cryotherapy application with probe to snowbanks. (Right) En face view of the during treatment; note additional single-freeze thaw applications given 1 cryo- therapy spot posterior to the snowbanks.

CRYOABLATION PROCEDURE AND CONVENTIONAL Patients receiving conventional therapy (ie, controls) THERAPY: With visualization using the indirect ophthal- received either topical acetate 1% eye drops moscope and condensing , single freeze-thaw, confluent, or a posterior superior sub-Tenon or an intravitreal triam- but not overlapping applications of cryotherapy were applied cinolone acetonide injection for unilateral active pars to all areas of snowbanking along involved areas of the pars planitis. Oral prednisone was used for recalcitrant unilat- plana (Figure 1, Left). Procedures were performed by 1 of 2 eral cases, eyes with a history of a corticosteroid-related surgeons (E.H.S. or J.C.F.). A single row of cryotherapy intraocular pressure rise, or bilateral active disease; patients was also applied just posterior to the most posterior aspect who could not tolerate prednisone or had recalcitrant dis- of snowbanking (Figure 1, Right). Depending on the amount ease despite prednisone 80 mg daily were started on of snowbanking, the average eye may receive between 10 corticosteroid-sparing immunomodulatory therapy. Rela- and 20 spots of cryotherapy. If the snowbanking involved tively few eyes were treated with pars plana vitrectomy more than 6 clock hours of the periphery, treatment was with or without adjunctive cryotherapy; any eyes receiving planned in 2 sessions because of the concern that more cryotherapy were placed in the cryotherapy group. Some extensive cryopexy at a single session could cause exacerba- eyes with bilateral active disease received cryotherapy in tion of the inflammation and/or peripheral retinal traction. only 1 eye; in these instances the eye with more severe dis- A second session of cryotherapy was also performed if the ease was always selected for cryotherapy. inflammation remained active. All patients received prepro- cedure topical anesthesia with proparacaine eye drops and, if STATISTICAL ANALYSIS: Kaplan-Meier curves and the cryotherapy was performed in the clinic, subconjunctival random-effects Cox proportional hazards models were fit 2% lidocaine without epinephrine; those intolerant of this for time-to-event data, such as resolution of macular approach were given a peribulbar block with 2% lidocaine edema, vitreous cell, and anterior chamber cell. Visual acu- without epinephrine. Children under the age of 16 years ity (logMAR after conversion from Snellen) trajectories or those with substantial anxiety or discomfort were brought were obtained using linear model with random effects (ie, to the operating room for monitored anesthesia care or mixed model). Resolution was defined as going from a vit- laryngeal mask anesthesia for the procedure. No postopera- reous (or anterior) chamber cell rating greater than 0.5þ at tive antibiotic drops were given to patients. To blunt any in- baseline (ie, time of cryotherapy or time or presentation in flammatory response and the potential for retinal the control group) to the first occurrence of vitreous (or detachment from cryotherapy, some patients not already anterior) cell rating less than 0.5þ at least 3 months after on oral prednisone received either (1) a posterior superior baseline, with additional qualifiers of having no periocular sub-Tenon injection of 20 mg triamcinolone acetonide or or intravitreal corticosteroid injections within the previous intravitreal 2 mg triamcinolone acetonide injection at the 120 days and no oral corticosteroids or immunosuppressive time of cryotherapy (n ¼ 21) or (2) oral prednisone at a agents. The 120-day time lag was used to account for the dose of 20–60 mg daily for 3 days prior to treatment, then duration of the effect of regional corticosteroid injections a rapid taper and discontinuation over 9 days after cryo- in both groups. Baseline for the controls was defined as therapy (n ¼ 6). the time of presentation; for the cryotherapy group it was

VOL. 162 CRYOABLATION FOR PARS PLANITIS 37 TABLE. Cryoablation for Treatment of Pars Planitis: Baseline Characteristics of Cryotherapy-Treated Versus Control Eyes

Cryotherapy Control P Characteristic Group Group Value

Number of patientsa 40 86 Number of eyes 50 136 Median age (y) 17.4 23.0 <.001 Age range 12–70 12–74 Sex (%) Men 55 46 .37 Women 45 54 Race/ethnicity (%) White, non-Hispanic 95.0 98.8 .19 African-American 2.5 1.2 .59 Asian 2.5 0 .14 FIGURE 2. Visual acuity (logMAR) trajectory over time Bilateral uveitis (%) 92.5 76.7 .03 showing a significant improvement after cryotherapy and wors- Treatment at baseline ening in the control group of eyes with active pars planitis. % who received regional 0 1.5 .44 corticosteroids within 3 months of baseline Current prednisone use (%) 28.2 2.9 .001 6 cryotherapy patients with unilateral pars planitis, and 20 Immunosuppression (%) 0 0 patients with bilateral uveitis where 1 eye was treated with Ocular characteristics cryotherapy and the second eye was in the control group). Presenting visual acuity Characteristics of the study population are shown in the Median logMAR 0.4 0.3 .23 Table. Cryotherapy-treated patients were followed for a Range, logMAR 0.12 to 3.3 0.12 to 3.3 median of 60.8 months (range 8.1–223.1 months) and Snellen equivalent 20/50 20/40 controls were followed for a median of 45.0 months (range Anterior chamber median 0.5 0.5 .68 3.1–339.0 months). Ten eyes in the cryotherapy group sus- grade cell tained more than 1 treatment. The 2 groups were relatively Range 0–3 0–3 well balanced at baseline with no significant differences Vitreous cell median grade 2 2 .20 between the groups in baseline visual acuity, anterior Range 0–3 0–3 Lens status (%) chamber and vitreous inflammation, presence of macular Phakic 100.0 99.3 .54 edema, and prior use of regional corticosteroid injections. Pseudophakic 0 0.07 .56 There were significant differences in the ages of the 2 Macular edema (%) 47.9 45.2 .71 groups, with the cryotherapy group being somewhat younger (median 17.4 years vs 23.0 years) and in the higher a Twenty bilateral patients were treated with cryotherapy in 1 use of oral corticosteroids in the cryotherapy group at base- eye and not in the second eye and are included in both groups. line (28.2% vs 2.9%, P ¼ .001). Regarding other potential confounders occurring during follow-up, 13 eyes (9.6%) in the control group underwent the date the treatment was performed or most recent visit pars plana vitrectomy, compared to 11 eyes (22%) in the before cryotherapy. Time-varying covariate analysis was cryotherapy group (P ¼ .038). Of the 11 cryotherapy- performed to account for confounding factors found be- treated eyes that underwent vitrectomy, 4 underwent it tween the groups at baseline and during follow-up, at the time of cryotherapy and 7 underwent it subsequently. including bilaterality, vitrectomy, and regional corticoste- The mean number of regional (ie, sub-Tenon or intravi- roid injections given after baseline. treal) corticosteroid injections after baseline in the cryotherapy-treated group was 2.13/eye and in the control group was 0.85/eye (P < .001). Potential confounders at RESULTS baseline and follow-up between the 2 groups were accounted for using time-varying covariate analyses. OF THE CHARTS REVIEWED FOR INCLUSION IN THIS STUDY, 58 patients were excluded because they did not meet the VISUAL ACUITY OUTCOMES: Over time, visual acuity on diagnostic criteria for pars planitis and 3 patients were average improved in the cryotherapy-treated eyes and excluded owing to follow-up of less than 3 months. The worsened slightly in the control eyes (Figure 2). Specif- study population consists of 186 eyes (136 control eyes; ically, in the cryotherapy-treated eyes, visual acuity values 50 cryotherapy eyes) from 95 patients (69 control patients, tended to improve (decreasing logMAR) during follow-up

38 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 2016 by an average of 0.0018 logMAR units per month (P ¼ .023). In the control eyes, the visual acuity values tended to worsen (increasing logMAR) during follow-up by 0.0011 logMAR units per month (P ¼ .023). A formal test of interaction to see if the slopes were different was sig- nificant (P ¼ .024). After adjusting for bilateral status, vit- rectomy, and regional corticosteroid injections during follow-up, there was still a suggestion that the slopes were different (P ¼ .069).

VITREOUS CELL: Resolution of vitreous cell occurred with greater frequency in cryotherapy-treated eyes. At 1 year, 49.8% of eyes in the cryotherapy group had experi- enced resolution of vitreous cell and were not taking either prednisone or immunosuppressive agents, compared to FIGURE 3. Kaplan-Meier survival estimate showing more 22.5% in the control group (P ¼ .0071). At 1 year, rapid resolution of vitreous cell in the cryotherapy group 43.7% of eyes in the cryotherapy group had resolution of compared to control eyes with active pars planitis. the vitreous cell, were not receiving oral corticosteroids or immunosuppressive agents, and had not received a regional corticosteroid injection within the previous 120 days, compared to 18.5% in the control group (P ¼ .0115). Kaplan-Meier survival estimates depicting resolu- tion of vitreous cell are presented in Figure 3. The hazard ratio for ‘‘remission’’ (inactive disease without treatment) for the cryotherapy group compared to the control group was 3.79 (95% confidence interval [CI] 1.47, 9.76, P ¼ .002). After adjusting for bilaterality, vitrectomy, and regional corticosteroid use, the hazard ratio for remission with cryoablation vs none was 4.73 (95% CI 1.65, 13.63, P ¼ .004).

CYSTOID MACULAR EDEMA: Cryotherapy-treated eyes appeared more likely to obtain resolution of macular edema compared to control eyes. The Kaplan-Meier survival esti- FIGURE 4. Kaplan-Meier survival estimate showing more mates of the resolution of macular edema are shown in rapid resolution of cystoid macular edema in the cryotherapy Figure 4; differences were significant by 6 months (P ¼ group compared to control eyes with active pars planitis. .007). At 1 year resolution of macular edema without recur- rence and without use of systemic medications or use of regional corticosteroid injections within the previous estimates of resolution of anterior chamber cell were signif- 120 days occurred in 57.7% of eyes treated with cryo- icantly different between the 2 groups by 6 months therapy compared to 41.0% of control eyes. At 1 year, res- (P ¼ .005). The crude hazard ratio for remission (inactive olution of macular edema without systemic medications disease off all therapy) was 12.62 (95% CI 0.96, 165.25; P ¼ but allowing regional corticosteroid injections within the .053), and after adjustment for bilaterality, vitrectomy, and previous 120 days occurred in 68.9% of cryotherapy- regional corticosteroid use during follow-up it was 21.74 treated eyes compared to 42.8% of control-treated eyes. (95% CI 1.11, 425.68; P ¼ .043). These percentages are Kaplan-Meier estimates based on 18 cryotherapy-treated eyes and 58 control eyes. The crude COMPLICATIONS AND OTHER TREATMENTS: Other hazard ratio for remission of the macular edema was 4.12 than transient erythema of the over the (95% CI 0.87, 19.44; P ¼ .074) in favor of the treated areas, there were no complications resulting from cryotherapy-treated eyes, and after adjustment for bilater- cryotherapy. No eyes developed retinal detachment as a ality, vitrectomy, and regional corticosteroid injections result of cryotherapy. In fact, 1 eye had an inferior retinal the hazard ratio was 6.85 (95% CI 1.06, 44.78; P ¼ .044). detachment before treatment that resolved 3 months after cryotherapy. Six eyes in the cryotherapy group had poste- ANTERIOR CHAMBER CELL: Cryotherapy appeared to be rior subcapsular cataract noted on examination (5 had associated with resolution of anterior chamber cell when cataract surgery); 18 of the control eyes had posterior sub- compared to control eyes (Figure 5). Kaplan-Meier survival capsular cataract (16 had cataract surgery)—the number

VOL. 162 CRYOABLATION FOR PARS PLANITIS 39 our series no retinal detachments were identified despite relatively long follow-up. Nevertheless, it is possible, or perhaps even likely, that peripheral retinal ablation by either means (cryotherapy or laser photoablation) may be similarly effective for the treatment of pars planitis, as the key issue is not the mode of ablation but that peripheral retinal ablation was performed. Histopathologic studies of human eyes with pars plani- tis have demonstrated fibroglial proliferation and lympho- cytic infiltration of the snowbank as well as , cyclitis, and choroiditis.19,20 Though B cells have been identified posterior to the junction of the and , there was a predominance of T cells in the pars plana of these donor eyes along with a high FIGURE 5. Kaplan-Meier survival estimate showing more ratio of helper T cells to suppressor cells.20 Elevated rapid resolution of anterior chamber cell in the cryotherapy levels of specific T cell subtypes in the peripheral blood group compared to control eyes with active pars planitis. of patients with pars planitis further implicate the role of T cell activation in this disorder.21,22 Thus a mechanism of action for cryotherapy is unknown; that had was not significantly different between hypotheses have included altering the blood-ocular ¼ the 2 groups (P .83). barrier (allowing better ingress of drugs and egress of Four eyes in the control group and 4 eyes in the cryo- inflammatory mediators) and destruction or reduction of therapy group needed glaucoma surgery, and all had multi- an antigenic stimulus in the pars plana. The retinal S- ple periocular corticosteroid injections preceding the antigen that has been used to cause experimental autoim- glaucoma procedure. Five patients started corticosteroid- mune uveitis21,22 has been suggested as a potential sparing immunosuppressive agents during follow-up, 2 in candidate,23 but antibodies to the S-antigen were not the control group and 3 in the cryotherapy group. All 3 found in a donor eye with pars planitis, suggesting that of these cryotherapy group patients had bilateral disease it may not be important in the pathogenesis.20 There is with cryotherapy given to only 1 eye. no doubt that further research is needed to elucidate the pathogenetic mechanisms involved. There are several potential limitations to this study, DISCUSSION many inherent in its retrospective nature. Although we could not identify any evident selection bias in the eyes PARS PLANITIS IS THE THIRD MOST COMMON UVEITIC DIS- chosen for cryotherapy (aside from choosing the ‘‘worse’’ ease affecting the pediatric population1,9,12 and has an eye for unilateral treatment in a patient with bilateral dis- overall annual incidence of 2 per 100,000 persons.3,13–16 ease), unidentified selection bias could have influenced the Standard approaches to treatment have advantages and results. Observers were not masked as to treatment, allow- disadvantages, including cataracts and ocular ing for potential observer bias to influence the results. As hypertension/glaucoma from regional corticosteroid this was a retrospective study, it is difficult to know for injections.17,18 Nevertheless, because of the chronic certain which type of patients would benefit most from nature of this disease, these treatments often require cryotherapy. These potential biases can only be addressed long-term use. Our data suggest that many eyes treated through a randomized clinical trial with masked assessment with peripheral retinal cryoablation can achieve ‘‘inac- of outcomes. Variable concomitant therapy is a potential tive’’ disease without attendant systemic medications confounder. There was a baseline imbalance in oral corti- and that this rate may be greater than that in eyes treated costeroid use, with a higher percentage of patients in the conventionally. Furthermore, cryotherapy-treated eyes cryotherapy group using oral corticosteroids at baseline. appeared to have faster resolution of macular edema and Given our stepwise treatment algorithm, this greater use better visual outcomes. In addition, we did not identify could reflect more severe disease or the greater frequency any greater rate of ocular complications in the of bilateral disease in the cryotherapy group. Given the cryotherapy-treated eyes, suggesting that this treatment similarity of vision, macular edema percentage, and inflam- can be administered safely. mation median grades, it is likely that the difference reflects Some clinicians have employed peripheral laser photoa- the difference in bilateral disease. As some of these patients blation instead of cryoablation for pars planitis, with good were treated before the advent of optical coherence tomog- results reported in uncontrolled case series. These authors raphy, it is possible that cystoid macular edema could have have expressed concern about the complications of cryo- been present despite not being seen ophthalmoscopically; therapy, particularly retinal detachment.13 However, in however, both groups would have been subjected to this

40 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 2016 and thus the effect on the results of this study should be bias. As such, our data should be viewed as highly sugges- minimal. tive rather than conclusive. Any differential treatment during follow-up also could There were no standard follow-up protocols, resulting in have influenced the outcomes, including resolution of variable follow-up intervals. We attempted to mitigate this anterior chamber and vitreous cells, resolution of macular variable follow-up by recording only visits that corre- edema, and improvement in vision. For instance, on sponded to 3-month intervals after baseline and using average there were fewer periocular injections given to contiguous visit windows. Heterogeneous follow-up also the control group during the follow-up period, and pars was addressed by using statistical tools such as Kaplan- plana vitrectomy was used in a greater percentage of pa- Meier product-limit estimation and random-effects tients (albeit still a minority) in the cryotherapy group. modeling. The sample size of eyes treated with cryotherapy We sought to mitigate any potential confounding factors was small, so that estimates of side effects also are impre- in 2 ways. The first was analyzing outcomes that included cise. For example, the upper 95% CI on the detachment being off systemic and regional treatments as part of the rate is approximately 6% of eyes treated with peripheral definition of remission, and the second was adjusting for po- retinal cryoablation. tential confounders statistically (including postbaseline In conclusion, for eyes with active pars planitis, our data periocular injections and vitrectomy) using time-varying suggest that peripheral retinal cryoablation appears to be covariate analysis. In this regard, our data suggest greater an effective treatment that can potentially result in a rates of resolution of anterior chamber cells, vitreous cells, ‘‘remission,’’ a result that appears to be seen less commonly and macular edema despite discontinuing concomitant oral with conventional corticosteroid treatments. This treat- therapy in the cryotherapy group. Nevertheless, despite the ment appears to have a low risk of side effects and to be a apparent success of cryotherapy and the significant adjusted reasonable addition to the armamentarium of therapies hazard ratios (favoring cryotherapy), the small sample size for pars planitis. Because of the limitations inherent in a led to wide confidence intervals, so that the magnitude of retrospective study of this type, its role vis-a`-vis conven- the benefit is less certain. Finally, observers were not tional treatment will need to be evaluated in a randomized masked as to treatment, potentially introducing observer clinical trial.

FUNDING/SUPPORT: NO FUNDING OR GRANT SUPPORT. FINANCIAL DISCLOSURES: ELLIOTT H. SOHN: OXFORD BIOMEDICA, Oxford, United Kingdom (research support); GlaxoSmithKline, Middlesex, United Kingdom (research support); Douglas A. Jabs: Applied Genetic Technologies, Inc, Alachua, Florida (Data and Safety Monitoring Committee); James C. Folk: IDx LLC, Iowa City, Iowa (Shareholder, Officer, and Board Member). The following author has no financial disclosures: Benjamin C. Chaon. All authors attest that they meet the current ICMJE criteria for authorship. The authors thank Jeffrey Dawson, ScD (University of Iowa) for his assistance with statistical analysis and Alton Szeto (altonszeto.com) for creation of the eye schematic figure.

REFERENCES a prospective longitudinal assessment in the childhood Asthma Management Program (CAMP) study. Pediatrics 1. Jabs DA, Nussenblatt RB, Rosenbaum JT. Standardization of 2008;122(1):e53–61. Uveitis Nomenclature (SUN) Working Group. Standardiza- 7. Multicenter Uveitis Steroid Treatment (MUST) Trial tion of uveitis nomenclature for reporting clinical data. Re- Research Group, Kempen JH, Altaweel MM, et al. Ran- sults of the First International Workshop. Am J Ophthalmol domized comparison of systemic anti-inflammatory ther- 2005;140(3):509–516. apy versus fluocinolone acetonide implant for 2. Raja SC, Jabs DA, Dunn JP, et al. Pars planitis: clinical fea- intermediate, posterior, and panuveitis: the multicenter tures and class II HLA associations. Ophthalmology 1999; uveitis steroid treatment trial. Ophthalmology 2011; 106(3):594–599. 118(10):1916–1926. 3. Donaldson MJ, Pulido JS, Herman DC, Diehl N, Hodge D. 8. Becker M, Davis J. Vitrectomy in the treatment of uveitis. Am Pars planitis: a 20-year study of incidence, clinical fea- J Ophthalmol 2005;140(6):1096–1105. tures, and outcomes. Am J Ophthalmol 2007;144(6): 9. Aaberg TM, Cesarz TJ, Flickinger RR. Treatment of periph- 812–817. eral uveoretinitis by cryotherapy. Am J Ophthalmol 1973; 4. Zhang L, Prietsch SOM, Ducharme FM. Inhaled corticoste- 75(4):685–688. roids in children with persistent asthma: effects on growth. 10. Lehpamer B, Moshier E, Goldberg N, Ackert J, Godbold J, Cochrane Database Syst Rev 2014;7:CD009471. Jabs DA. Subretinal fluid in uveitic macular edema: effect 5. MacGregor RR, Sheagren JN, Lipsett MB, Wolff SM. Alter- on vision and response to therapy. Am J Ophthalmol 2013; nate-day prednisone therapy. Evaluation of delayed hypersen- 155(1):143–149. sitivity responses, control of disease and steroid side effects. N 11. Klein R, Klein BE, Lee KE, Cruickshanks KJ, Chappell RJ. Engl J Med 1969;280(26):1427–1431. Changes in visual acuity in a population over a 10-year 6. Kelly HW, Van Natta ML, Covar RA, et al. Effect of long- period: the Beaver Dam Eye Study. Ophthalmology 2001; term corticosteroid use on bone mineral density in children: 108(10):1757–1766.

VOL. 162 CRYOABLATION FOR PARS PLANITIS 41 12. Smith JA, Mackensen F, Sen HN, et al. Epidemiology and edema complicating noninfectious uveitis. Am J Ophthalmol course of disease in childhood uveitis. Ophthalmology 2009; 2011;152(3):441–448.e442. 116(8):1544–1551. 1551.e1. 18. Sallam A, Taylor SRJ, Habot-Wilner Z, et al. Repeat intravi- 13. Okinami S, Sunakawa M, Arai I, Iwaki M, Nihira M, treal triamcinolone acetonide injections in uveitic macular Ogino N. Treatment of pars planitis with cryotherapy. oedema. Acta Ophthalmol 2012;90(4):e323–325. Ophthalmologica 1991;202(4):180–186. 19. Pederson JE, Kenyon KR, Green WR, Maumenee AE. Pathol- 14. Verma L, Kumar A, Garg S, Khosla PK, Tewari HK. Cryo- ogy of pars planitis. Am J Ophthalmol 1978;86(6):762–774. pexy in pars planitis. Can J Ophthalmol 1991;26(6): 20. Wetzig RP, Chan CC, Nussenblatt RB, Palestine AG, 313–315. Mazur DO, Mittal KK. Clinical and immunopathological 15. Josephberg RG, Kanter ED, Jaffee RM. A fluorescein angio- studies of pars planitis in a family. Br J Ophthalmol 1988; graphic study of patients with pars planitis and peripheral 72(1):5–10. exudation (snowbanking) before and after cryopexy. Ophthal- 21. Boyd SR, Young S, Lightman S. Immunopathology of the mology 1994;101(7):1262–1266. noninfectious posterior and intermediate uveitides. Surv 16. Prieto JF, Dios E, Gutierrez JM, Mayo A, Calonge M, Ophthalmol 2001;46(3):209–233. Herreras JM. Pars planitis: epidemiology, treatment, and asso- 22. Caspi RR, Silver PB, Luger D, et al. Mouse models of experimental ciation with . Ocul Immunol Inflamm 2001; autoimmune uveitis. Ophthalmic Res 2008;40(3-4):169–174. 9(2):93–102. 23. Nussenblatt RB, Gery I, Ballintine EJ, Wacker WB. Cellular 17. Leder HA, Jabs DA, Galor A, Dunn JP, Thorne JE. Periocular immune responsiveness of uveitis patients to retinal S-anti- triamcinolone acetonide injections for cystoid macular gen. Am J Ophthalmol 1980;89(2):173–179.

42 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 2016 Biosketch

Dr Elliott H. Sohn is vitreoretinal surgeon-scientist at the Wynn Institute for Vision Research/University of Iowa with special expertise in inherited retinal degenerations, AMD, and and is surgeon/PI/co-investigator for several gene therapy trials. His laboratory focuses on efficacy and immunologic effects of stem cell and gene therapy and the pathophysiology and novel therapies for AMD. He serves on ARVO, AAO, and ABO committees and is editorial board member for Scientific Reports (Nature Publishing).

VOL. 162 CRYOABLATION FOR PARS PLANITIS 42.e1 Biosketch

Dr James C. Folk is the Judith (Gardner) and Donald Beisner, Professor of Vitreoretinal Disease and Surgery in the Department of Ophthalmology and Visual Sciences and Wynn Institute of Vitreoretinal Research at the University of Iowa. He is an officer of the Macula Society and member of the Retina Society and American Society of Retinal Specialists. Although he treats a wide variety of vitreoretinal disease he has a particular interest in ocular inflammatory diseases. He is active in both clinical and translational research.

42.e2 AMERICAN JOURNAL OF OPHTHALMOLOGY FEBRUARY 2016