Retrospective cohort study of 163 pediatric glaucoma patients Jean-François Bussières,* BPharm, MSc, FCSHP; Roxane Therrien,* BPharm, MSc; Patrick Hamel,† MD, FRCSC; Pierre Barret,* DPharm Candidate; Sonia Prot-Labarthe,* DPharm, PhD ABSTRACT •RÉSUMÉ Objective: The aim of our study was to describe a cohort of pediatric glaucoma patients in Quebec. Design: This study was a retrospective medical record review. Participants: The study included patients younger than 18 years who were diagnosed with glaucoma between 1980 and 2000 and monitored at the Ophthalmology Clinic of the Sainte-Justine University Hospital Centre (Centre Hospitalier Universitaire Sainte-Justine) and had ocular hypertension or glaucoma in at least 1 eye. Methods: The data gathered concerned patients’ information, the surgical procedures performed post diagnosis associated with the glaucoma diagnosis, and the drugs prescribed. Results: The study included 163 patients (254 eyes), a total of 374 surgical procedures, and the use of 2885 anti­ glaucoma drug therapies. For the 4 most frequent pathologies, patients were monitored for 8.4 (SD 4.2) years for aphakic glaucoma/pseudophakic glaucoma, 10.0 (SD 5.5) years for congenital glaucoma, 9.0 (SD 5.2) years for Axenfeld-Rieger syndrome, and 7.5 (SD 3.4) years for uveitic glaucoma. In total, 113 patients had at least 1 surgical procedure (69.3%). Before 1985, only timolol, pilocarpine, epinephrine, acetazolamide, and dipivefrin were used. Other beta blockers then appeared (betaxolol, levobunolol between 1985 and 1990, and the timolol-pilocarpine association between 1990 and 1995). After 1995, we saw the arrival of a new class of prostaglandin F2 alpha analogues, with latanoprost and other carbonic acid anhydrase inhibitors such as dorzolamide and brinzolamide. Conclusions: This study illustrates the great variety of glaucoma diagnostic subgroups and the use of surgery and drug therapies. Objet : Description d’une cohorte de patients atteints de glaucome pédiatrique au Québec. Nature : Examen rétrospectif de dossiers médicaux. Participants : Patients de moins de 18 ans ayant eu un diagnostic de glaucome entre 1980 et 2000 à la Clinique d’Ophtalmologie du Centre Hospitalier Universitaire Saint-Justine et qui souffraient d’hypertension oculaire ou de glaucome dans au moins 1 œil. Méthodes : Les données relevées portaient sur les patients, les procédures chirurgicales associées et postérieures au diagnostic de glaucome et les médicaments prescrits. Résultats : L’étude a porté sur 163 patients (254 yeux), 374 procédures chirurgicales et le recours à 2885 pharmacothérapies anti-glaucome. Pour les 4 pathologies les plus fréquentes, les patients ont été suivis de près pendant 8,4 (ÉT 4,2) ans pour glaucome aphaque ou pseudophaque, 10,0 (ÉT 5,5) ans pour glaucome congénital, 9,0 (ÉT 5,2) ans pour le syndrome d’Axenfeld-Rieger et 7,5 (ÉT 3,4) ans pour le glaucome uvéitique. En tout, 113 patients ont subi au moins 1 procédure chirurgicale (69,3 %).   Avant 1985, seuls le timolol, la pilocarpine, l’épinéphrine, l’acétazolamide et la dipivéfrine étaient utilisés. D’autres bêta-bloquants sont ensuite apparus (le bétaxolol, le lévobunolol entre 1985 et 1990, et l’association timolol-pilocarpine entre 1990 et 1995).   Après 1995, nous avons vu l’arrivée d’une nouvelle classe d’analogues de la prostaglandine F2-alpha, avec le latanoprost et d’autres inhibiteurs de l’anhydrase carbonique comme le dorzolamide et le brinzolamide. Conclusions : Cette étude illustre la grande variété des sous-groupes de diagnostic du glaucome et du recours à la chirurgie et à la pharmacothérapie.

C

hildhood glaucoma is rare. Glaucoma management has been evolving rapidly over the last few years, especially when it comes to surgery, but also regarding the drug therapies offered, which often play a supporting role.1–3 Because of the low incidence of the disease in the population and the varied differential diagnoses of pediatric glaucoma, few pediatric cohort studies have been published.4-6 The aim of our study is to describe a cohort of pediatric

glaucoma patients in Quebec who were monitored at the Sainte-Justine University Hospital Centre (Centre Hospitalier Universitaire Sainte-Justine [CHUSJ]). Methods

The study included patients younger than 18 years who were diagnosed with glaucoma between 1980 and

From *Unité de Recherche en Pratique Pharmaceutique, Département de Pharmacie, Centre Hospitalier Universitaire Sainte-Justine, Montréal, Qué., and †Département d’Ophtalmologie, Centre Hospitalier Universitaire Sainte-Justine, Montréal, Qué.

Correspondence to Jean-François Bussières, MSc, FCSHP, BPharm, Unité de Recherche en Pratique Pharmaceutique (URPP), Département de Pharmacie, Centre Hospitalier Universitaire Sainte-Justine, 3175, chemin de la côte Sainte-Catherine, Montréal QC H3T 1C5; [email protected]

Originally received July 10, 2008. Revised Dec. 30, 2008 Accepted for publication Jan. 22, 2009 Published online May 7, 2009

This article has been peer-reviewed. Cet article a été évalué par les pairs. Can J Ophthalmol 2009;44:323–7 doi:10.3129/i09-065 CAN J OPHTHALMOL—VOL. 44, NO. 3, 2009 

323

Pediatric glaucoma cohort study—Bussières et al. 2000 and monitored at the Ophthalmology Clinic of the CHUSJ and had ocular hypertension or glaucoma in at least 1 eye. Patients were taken from the list of the pedia­ tric glaucoma clinic, which is an exhaustive list because all children with glaucoma are referred to that specific clinic in our department. Glaucoma was defined as an elevated intraocular pressure (IOP) with consequent ocular damages and (or) with a clear etiology. Cases of treated and untreated ocular hypertension were included. This study is a retrospective medical record review. Some of these cases may be in common with those in the Barsoum-Homsy and Chevrette study,4 given the recruitment periods (63 cases between 1975 and 1983). The data gathered concerned patients (age at diagnosis, sex, date of diagnosis, exact diagnosis made, monitoring duration, bilaterality of the ocular involvement, IOP, cup-to-disc horizontal ratio, and cupto-disc vertical ratio), the surgical procedures performed post diagnosis associated with the glaucoma diagnosis (type of surgery and delay after diagnosis), and the drugs prescribed (international nonproprietary names, delay between diagnosis and the start of drug therapy, prescription duration, and route of administration). One eye was randomly chosen to present the IOP, cup-to-disc horizontal ratio, and cup-to-disc vertical ratio results in bilateral glaucoma patients. Cup-to-disc ratio was measured from pictures if available, otherwise from direct clinical examination. Because this is a retrospective study, diagnosis was taken from the chart. Other pathologies of the anterior segment not included under the Axenfeld-Rieger, Peters’, or aniridia categories, but with obvious malformations of the anterior segment, were categorized as anterior segment dysgenesis. Following steroid therapy, glaucoma was considered as uveitic if removal of the corticosteroids did not resolve the elevated IOP, and as steroid responder if it did. No statistical tests were conducted. The results are presented in the form of mean, standard deviation, median, minimum, and maximum values for quantitative variables and percentage for qualitative variables. These records were analyzed with the approval of the institution’s director of professional care services. Results

The study included 163 patients (254 eyes). The CHUSJ’s Department of Ophthalmology monitors most of the glaucoma cases in the province of Quebec. Tables 1, 2, 3, and 4

describe, respectively, the characteristics and diagnosis data of 163 monitored patients, the 374 surgical proce­dures performed, and the 2885 antiglaucoma drug therapies. For the 4 most frequent pathologies, the monitoring durations were 8.4 (SD 4.2) years for aphakic or pseudophakic glaucoma (range 0.3–17.2 years), 10.0 (SD  5.5) years for congenital glaucoma (range 0.8–18.5 years), 9.0 (SD 5.2) years for Axenfeld-Rieger syndrome (range 0.3–18.3 years) and 7.5  (SD 3.4) years for uveitic glaucoma (range 1.3– 13.0  years). Certain patients may have presented with an abnormally low IOP reading when they first visited the CHUSJ because they were diagnosed and treated before their monitoring and drug therapy at CHUSJ. In total, 113 patients had at least 1 surgery (69.3%). Type of surgery among diagnoses is described in Table 5. The 4 most frequent diagnoses were also the ones that recorded the greatest number of surgical procedures as a group (81 for 38 patients with aphakic or pseudophakic glaucoma, 125 cumulative surgeries for 36 patients with congenital glaucoma, 48 for 20 patients with Axenfeld-Rieger syndrome, and 53 for 19 patients with uveitic glaucoma). For these pathologies, between 63.2% and 94.4% of the patients had at least 1 surgical procedure. For the pathologies with more than 5 patients, the diagnoses that led to the greatest number of surgical procedures per patient were congenital glaucoma (3.5 operations per patient and 2.2 operations per eye), uveitic glaucoma (2.8 and 2.1, respectively), and Axenfeld-Rieger syndrome (2.4 and 1.4, respectively). Among the 50 patients who did not have any surgery, 14 (28.0%) had an aphakic or pseudophakic glaucoma, 5 (10.0%) had uveitic glaucoma, and 5 (10.0%) had Axenfeld-Rieger syndrome. The 2885 antiglaucoma drug therapies were given to 161 patients: 2 patients (1.2%) with ocular hypertension did not receive any drug therapy nor had any surgical procedure performed. Before 1985, only timolol, pilocarpine, epinephrine, acetazolamide, and dipivefrin were used. Other beta blockers then appeared (betaxolol, levobunolol between 1985 and 1990, and the timolol-pilocarpine association between 1990 and 1995). After 1995, we saw the arrival of a new class of prostaglandin F2 alpha analogues with latanoprost and other carbonic anhydrase inhibitors such as dorzolamide and brinzolamide (Fig. 1). Brimonidine was prescribed for 32 patients between 6.6 (SD 4.2) years (from 0 to 17.6 years), 3 of whom were under 2 years. Furthermore, 1042 corticoid prescriptions were in the con-

Table 1—Characteristics of the 163 patients Characteristic

Mean

Standard deviation

Median

Min-Max

Age at diagnosis (y) Length of monitoring* (y) IOP at the time of diagnosis (mm Hg)† CDH at the time of diagnosis‡ CDV at the time of diagnosis‡

5.7 7.9 29.1 0.5 0.5

5.1 4.7 8.7 0.2 0.2

5.2 8.0 28.0 0.5 0.5

1 d–19.4 118 d–20.2 5.0–52.0 0.1–1.0 0.1–1.0

*Data available for 161 patients (98.8%). † Data available for 124 patients (76.1%). ‡ Data available for 73 patients (44.8%). Note: Min, minimum; Max, maximum; y, years; d, days; IOP, intraocular pressure; CDH, cup-to-disc horizontal ratio; CDV, cup-to-disc vertical ratio.

324 

CAN J OPHTHALMOL—VOL. 44, NO. 3, 2009

Pediatric glaucoma cohort study—Bussières et al. text of uveitic glaucoma or short-term and perioperatory treatment (they are not compiled in Table 4). This study presents a pediatric glaucoma cohort from Quebec. An earlier cohort from Quebec was published in 1986.4 Barsoum-Homsy and Chevrette4 reported 63 pediatric glaucoma cases monitored at the CHUSJ between 1975 and 1983 (95 eyes). The other published cohorts reported 51 patients (83 eyes) monitored between 1980 and 2000 for primary infantile glaucoma by MacKinnon et al.5 in Australia and 306 children (520 eyes) monitored between 1974 and 1995 by Taylor et al.6 in Toronto. Age at diagnosis varied according to the studies and inclusion criteria. For MacKinnon et al.,5 who studied primary infantile glaucoma, the age at diagnosis was always

under 1 year. On the other hand, for Taylor et al.6 and Barsoum-Homsy and Chevrette,4 the diagnosis could be made up to the end of adolescence, as was the case in our study. Glaucoma always affects boys more often than girls (61% to 66%  of the cases are boys), but the proportion varies depending on the type of glaucoma.1,4,5 Furthermore, the type of glaucoma found in children varies according to the study. Some reported that the most frequent glaucoma was congenital glaucoma, up to twice as frequent as aphakic/ pseudophakic glaucoma for Taylor et al.,6 whereas in our cohort, we had as many aphakic/pseudophakic glaucoma cases as congenital glaucoma cases (comparison per patient).4 Similarly, the patient age breakdown according to Taylor et al.6 sometimes differed from our data, depending

Table 2—Diagnosis data* of the 163 patients Age at diagnosis, y (mean [SD])

Patient data

n

%

Male gender Bilaterality of the ocular involvement Diagnosis years†   1980–1984   1985–1989   1990–1994   1995–2000 Diagnosis per patient   Aphakic/pseudophakic glaucoma   Congenital glaucoma   Glaucoma associated with Axenfeld-Rieger syndrome   Uveitic glaucoma   Nevus flammeus/Sturge-Weber syndrome with glaucoma   Open-angle juvenile glaucoma   Posttraumatic glaucoma   Glaucoma associated with aniridia   Glaucoma associated with the use of corticosteroids   Glaucoma associated with anterior segment dysgenesis not   otherwise specified   Glaucoma associated with PHPV   Glaucoma associated with neurofibromatosis   Other types of glaucoma‡   Ocular hypertension

93 91

57.1 55.8

38 49 53 23

23.3 30.1 32.5 14.1

— — — — — — —

38 36 20 19 11 8 7 5 4 2

23.3 22.0 12.3 11.7 6.8 4.9 4.3 3.1 2.4 1.2

6.6 (4.4) 1.5 (2.8) 3.6 (4.1) 9.2 (3.6) 4.3 (5.4) 9.9 (3.5) 11.6 (2.6) 3.9 (4.1) 10.0 (1.4) 16.1 (4.6)

2 2 6 3

1.2 1.2 3.7 1.8

15.0 (3.0) 0.3 (0.3) 8.2 (3.8)

*The data are presented per patient. † Years cited included in the interval. ‡ Marfan syndrome, glaucoma associated with Lowe’s syndrome, glaucoma associated with oculodentodigital syndrome, glaucoma associated with corneal transplantation, glaucoma associated with retinopathy of prematurity, glaucoma associated with Peters’ anomaly. Note: y, years; PHPV, persistent hyperplastic primary vitreous.

Table 3—Characteristics of the 374 surgeries for the 113 patients Value

Median

Min-Max

Delay between diagnosis and surgery (mo) (mean, SD) Number of surgical procedures per patient (mean, SD)

31.9 (38.2) 3.3 (2.3)

13.8 3.0

0–164.4 1–12

Type of surgical procedure (n, %)*   Trabeculectomy   Goniotomy   Trabeculectomy with mitomycin   Trabeculotomy   Seton drainage   Holmium laser sclerostomy   Trabeculotomy with trabeculectomy   Cyclocryotherapy   Review of trabeculectomy   Iridectomy   Other surgeries†

105 (28.1) 103 (27.5) 47 (12.6) 35 (9.4) 21 (5.4) 20 (5.2) 13 (3.5) 11 (2.9) 8 (2.1) 7 (1.9) 4 (1.1)

— — — — — — — — — — —

— — — — — — — — — — —

*Calculated in terms of the total number of surgical procedures. † Sclerostomy, synechialysis, goniosynechialysis, review of Molteno implant. Note: Min, minimum; Max, maximum; mo, months.

CAN J OPHTHALMOL—VOL. 44, NO. 3, 2009 

325

Pediatric glaucoma cohort study—Bussières et al. on the diagnoses. This may be explained by certain categories of glaucoma with small numbers of cases or a slightly different classification of diagnoses and groups. Among the surgical procedures observed in the literature, goniotomy was always the most frequent surgery and it is often considered the gold standard treatment, particularly for congenital glaucoma. MacKinnon et al.5 reported that goniotomy was used as a primary therapy for 50/83 eyes (69.4%) and was effective for 28/50 eyes. The use of goni-

otomy is considered a good prognosis of glaucoma. Among the congenital glaucomas, Taylor et al.6 reported that standard (or simple) goniotomy was also sufficient for 42.3% of the patients (51.4% of the eyes). In the preceding CHUSJ cohort, 25/63 patients (39.7%) had not had any surgery and received only drug therapies.4 This proportion is a little less significant in our cohort (50/163 patients, i.e., 30.7%). The number of surgical procedures per patient was similar in our study to that found in the Toronto cohort, with 2.9

Table 4—Characteristics of the 2885 drugs prescribed for the 161 patients Number of drugs per patient* (mean, SD) Number of drugs per patient-year* (mean, SD) Prescription duration for each prescription drug (mo)† (mean, SD) Routes of drug administration (n, %)   Ophthalmic   Per os   Intravenous Prescribed drugs   Systemic route    Acetazolamide per os    Methazolamide per os    Glycerol per os    Intravenous mannitol   Ophthalmic route    Beta blockers     Timolol     Betaxolol     Levobunolol     Timolol and pilocarpine     Timolol and dorzolamide    Sympathomimetics     Dipivefrin     Epinephrine    Topical carbonic acid anhydrase inhibitors     Dorzolamide     Brinzolamide    Alpha agonists     Apraclonidine     Brimonidine    PGF2a analogues     Latanoprost    Parasympathomimetics     Pilocarpine     Phospholine

Value

Median

Min-Max

18.0 (16.9) 3.2 (4.0) 8.5 (14.6)

12,0 1.9 2.7

1–98 0.1–27.9 0–136.0

2093 (72.6) 765 (26.5) 27 (0.9)

— — —

— — —

741 19 5 27

(25.7) (0.7) (0.2) (0.9)

— — — —

— — — —

551 49 29 97 22

(19.1) (1.7) (1.0) (3.4) (0.8)

— — — — —

— — — — —

291 (10.1) 150 (5.2)

— —

— —

79 (2.7) 28 (1.0)

— —

— —

66 (2.3) 61 (2.1)

— —

— —

37 (1.3)

—

—

627 (21.7) 6 (0.2)

— —

— —

*The number of drug therapies (1 drug, 1 length of treatment) is the sum of all drugs prescribed to 1 patient during the follow-up (simultaneously or no t). † Data available for 2595 drugs (89.9%). Note: Min, minimum; Max, maximum; mo, months; PGF2a, prostaglandin F2 alpha.

Table 5—Surgical interventions (n = 374) among diagnoses Goniotomy Aphakic/pseudophakic glaucoma Congenital glaucoma Glaucoma associated with Axenfeld-Rieger syndrome Uveitic glaucoma Nevus flammeus/Sturge-Weber syndrome with glaucoma Open-angle juvenile glaucoma Posttraumatic glaucoma Glaucoma associated with aniridia Glaucoma associated with neurofibromatosis Other

Trabeculectomy*

Seton drainage

Holmium laser sclerostomy

Other

n

%

n

%

n

%

n

%

n

%

n

%

11 51 19

13.6 40.8 39.6

35 38 18

43.2 30.4 37.5

4 23 5

4.9 18.4 10.4

5 4 1

6.2 3.2 2.1

6 2 2

7.4 1.6 4.2

20 7 3

24.7 5.6 6.3

10 4

18.9 26.7

28 7

52.8 46.7

— —

— —

5 1

9.4 6.7

7 2

13.2 13.3

3 1

5.7 6.7

1 1 2 3 1

10.0 7.7 15.4 50.0 10.0

8 5 6 2 5

80.0 38.5 46.1 33.3 50.0

— — 2 — 1

— — 15.4 — 10.0

— 2 2 — 1

— 15.4 15.4 — 10.0

— 2 — — —

— 15.4 — — —

1 3 1 1 2

10.0 23.1 7.7 16.7 20.0

*Trabeculectomy with and without mitomycin. Note: Data and percentage are given per diagnosis.

326 

Trabeculotomy

CAN J OPHTHALMOL—VOL. 44, NO. 3, 2009

Pediatric glaucoma cohort study—Bussières et al. (SD 3.6) operations per patient.6 For 3 of the most frequent pathologies in our study (congenital, aphakic, and uveitic), the percentage of patients with at least 1 surgical procedure was between 42% and 98%.6 We were unable to observe the impact of the surgical interventions on IOP or the number of drugs received by the children, unlike in other studies, because our patients had sometimes undergone a number of surgical procedures, the diagnoses were varied, and the length of the monitoring was extensive.7 We are also unable to provide any outcome issue for our patients regarding data collected. No other published cohorts precisely describe the drug therapies that were used. Our study describes the variety of drugs prescribed for children with glaucoma. Quigley8 described the development of glaucoma drug therapies. He reported the replacement of the carbonic anhydrase inhibitors per os by topical agents and the watershed event in the treatment of glaucoma with the introduction of prostaglandin analogues and their daily intake, which made it possible to improve compliance. Fig. 1 shows the impact of marketing novel drugs on the development of drug prescriptions with the arrival after 1995 of prostaglandin F2 alpha analogues (latanoprost), a-agonists (apraclonidine, brimonidine), and drug combinations (timolol-pilocarpine, timolol-dorzolamide). Children are seldom included in clinical studies for glaucoma marketing, whereas in practice they are prescribed glaucoma drugs.2,3 The efficacy of latanoprost seems less obvious based on the age of the children compared with adults, and further studies are required to confirm the ideal conditions of use.9 The pediatric use of brimonidine has been associated with systemic adverse events (fatigue, drowsiness, apnea, bradycardia, hypotension, hypotonia) according to the company’s data and a number of publications.10–12 The most serious side effects were described in children younger than 2 years, but older children can experience drowsiness and extreme fatigue as well. Only 3/32 children who received brimonidine were under 2 years in our study. Although no side effects occurred in our patients, we do not recommend the pediatric use of brimonidine. Patient cohorts like the one presented in this paper allow us to present the practical pediatric use of these drugs. Our

Fig. 1—Prescription of the various drug classes over the years. (CCAI, carbonic acid anhydrase inhibitors; PGF2a, prostaglandin F2 alpha.)

cohort, however, does not allow us to draw a conclusion as to the efficacy of these therapies or surgical procedures, but we do describe the development of prescription practices and pediatric glaucoma management. This study also illustrates the great variety of pediatric glaucomas. We decided not to present IOP development for these patients because the results are too complex to interpret, given the numerous parameters that come into play, including the type and number of surgical procedures, drug therapies, compliance, and diagnoses. Conclusions

This study presents a cohort of 163 pediatric glaucoma patients with a total of 374 surgical procedures and the use of 2885 antiglaucoma drug therapies from 1980 to 2000. This study illustrates the great variety of glaucoma diagnostic subgroups, and the use of surgery and drug therapies. The research group would like to thank Dr. Line Chevrette, MD, FRCSC, and Dr. Magda Barsoum-Homsy, who were involved in the treatment of the majority of these patients during their careers at the Centre Hospitalier Universitaire Sainte-Justine. The authors have no proprietary or commercial interest in any materials discussed in this article.

References 1. Papadopoulos M, Cable N, Rahi J, Khaw PT; BIG Eye Study Investigators. The British Infantile and Childhood Glaucoma (BIG) Eye Study. Invest Ophthalmol Vis Sci 2007;48:4100–6. 2. Papadopoulos M, Khaw PT. Advances in the management of paediatric glaucoma. Eye 2007;21:1319–25. 3. Moore W, Nischal KK. Pharmacologic management of glaucoma in childhood. Paediatr Drugs 2007;9:71–9. 4. Barsoum-Homsy M, Chevrette L. Incidence and prognosis of childhood glaucoma. A study of 63 cases. Ophthalmology 1986;93:1323–7. 5. MacKinnon JR, Giubilato A, Elder JE, Craig JE, Mackey DA. Primary infantile glaucoma in an Australian population. Clin Experiment Ophthalmol 2004;32:14–8. 6. Taylor RH, Ainsworth JR, Evans AR, Levin AV. The epidemiology of pediatric glaucoma: the Toronto experience. J AAPOS 1999;3:308–15. 7. Hill R, Ohanesian R, Voskanyan L, Malayan A. The Armenian Eye Care Project: surgical outcomes of complicated paediatric glaucoma. Br J Ophthalmol 2003;87:673–6. 8. Quigley HA. Advances in glaucoma medication during the 1990s and their effects. J Glaucoma 2001;10:S71–2. 9. Enyedi LB, Freedman SF. Latanoprost for the treatment of pediatric glaucoma. Surv Ophthalmol 2002;47:S129–32. 10. Carlsen JO, Zabriskie NA, Kwon YH, Barbe ME, Scott WE. Apparent central nervous system depression in infants after the use of topical brimonidine. Am J Ophthalmol 1999;128:255–6. 11. Enyedi LB, Freedman SF. Safety and efficacy of brimonidine in children with glaucoma. J AAPOS 2001;5:281–4. 12. Al-Shahwan S, Al-Torbak AA, Turkmani S, Al-Omran M, AlJadaan I, Edward DP. Side-effect profile of brimonidine tartrate in children. Ophthalmology 2005;112:2143. Keywords: pediatrics, open-angle glaucoma CAN J OPHTHALMOL—VOL. 44, NO. 3, 2009 

327