Medical therapies play an important role in the management of pediatric glaucomas. This is acknowledged from studies such as the British Infantile and Childhood Glaucoma (BIG) Eye study, a national population-based study measuring the incidence, detection patterns, current management, and intraocular pressure (IOP) control at 1 year of children with newly diagnosed glaucoma in the United Kingdom. It reported that medical treatments were used more often in the management of secondary glaucomas compared with primary congenital glaucoma (PCG) cases, with 32% of such patients treated medically only. Eighty-one percent of PCG cases had medication before surgery and an average of 1.7 eye drops were prescribed.1 IOP control of ≤21 mm Hg was achieved in 86% on medications in secondary glaucoma cases.
The European Regulation 1901/2206 (introduced in 2007) changed the clinical and pharmaceutical environment of prescribing medicines to children in Europe.2 It aimed to ensure that pediatric medicines were subject to research of high quality, were appropriately authorized for use in the pediatric population and to improve the information available on the use of medicinal products in the various pediatric population. The Paediatric Glaucoma Service at Moorfields Eye Hospital NHS Foundation Trust is one of the largest subspecialty units serving pediatric glaucoma patients in the world. Hence, this new regulation gave impetus to our study to assess our current clinical practice and to update our knowledge (since the BIG Eye study) about the use of the medical therapies for the pediatric glaucomas. The aims of this study were as follows: to describe the medical treatments currently being used in the treatment of pediatric glaucomas, to assess the IOP-lowering effects of individual medical therapies and to report on the side effects experienced by our patients.
This was a retrospective review of 200 sets of clinical notes of pediatric glaucoma patients (aged 0 to 18 years) under the care of Moorfields Eye Hospital NHS Foundation Trust, approved by and registered with the Trust’s Clinical Audit Assessment Committee. The sample of patients was obtained using the prescribing database of the Moorfields Pharmacy Department (Ascribe Executive Information System) by identifying 200 consecutive patients who had attended the Pediatric Glaucoma Service from April 2006 to March 2007. The clinical details of most pediatric glaucoma patients were hand written onto a pro forma (originally designed by P.T.K.). Information about adherence and any related side effects were systematically recorded at each visit. Any relevant anesthetic details were also verified by checking the medical notes.
The notes were reviewed for the following information:
1. The use of individual medical therapies. Categorization of the methods of prescribing the medicines were as follows:
The medicine was prescribed as a monotherapy.
The medicine was added to other medicines.
The medicine was prescribed simultaneously with other medicines.
The medicine was prescribed at the same time/within a 3-month period of a laser or surgical treatment being performed.
2. The IOP effects of the glaucoma medicines.
3. The reporting of side effects.
IOP measurements were taken from both eyes of 1 patient if medical therapies were used bilaterally using applanation tonometry—either in clinic with a Goldmann Applanation Tonometer (GAT, Haag-Streit) or a Tono Pen (Reichert) or during an examination under ketamine anesthesia with Perkins Tonometry (Haag-Streit). Digital (palpation) tonometry measurements were excluded from the analysis. IOP measurements were excluded if the medication was taken for <7 days or if the medication was prescribed <1 month after a surgical or laser glaucoma procedure.
The Kruskal-Wallis test was used to assess whether there was a difference in IOP-lowering effect between drugs. The Kruskal-Wallis test was used because of evidence of departure from the assumption of equal variances made by analysis of variance; however, a sensitivity analysis was conducted using analysis of variance to assess whether findings were robust to the method adopted. The 2 sample t test with unequal variances was used to assess if the IOP-lowering effect was greater when a combination of glaucoma medicine was prescribed compared with its separate constituents. We did not adjust for multiple testing because these tests were regarded as exploratory hypothesis generating tests and as such no further statistical correction was advisable. In our statistical analysis, we have excluded those prescriptions that were only briefly used (<7 days) before a change in the treatment plan.
The clinical notes of 200 patients who had attended the Paediatric Glaucoma Service at Moorfields Eye Hospital Foundation Trust between April 2006 and March 2007 were reviewed. The demographics of these patients are summarized in Table 1.
Our results have been set out in 3 sections:
1. The prescribing of medicines.
2. The IOP-lowering effect of individual medicines. Any IOP measurements within 1 month of an operative procedure were not included as these measurements might have been affected by the surgery. Most of the IOP measurements were obtained from either Goldmann Applanation or Perkins Tonometry (only 2 Tono Pen IOP measurements were used in this analysis).
3. The side effects associated with glaucoma medicines.
Section 1: The Prescribing of Medicines
Medicines were prescribed a total of 1592 times in 200 patients as summarized in Table 2. Overall, the prostaglandin analogue and prostamide medicines were the most commonly prescribed topical medicines accounting for 27% of all prescriptions, of which latanoprost 50 mcg/mL (Pfizer) was the most commonly prescribed prostaglandin analogue. The α-agonists brimonidine tartrate 0.2% (Allergan) and apraclonidine 0.5% (Alcon) were the next most commonly prescribed glaucoma medicines, accounting for 22% of all prescriptions.
The prostaglandin analogues and prostamides were the most commonly prescribed monotherapy (accounting for 39% of monotherapy prescriptions), followed by a topical β-blocker (21% of all monotherapy prescriptions). Brimonidine tartrate 0.2% or apraclondine 0.5% were the most commonly prescribed medicines when required for additional treatment, accounting for 32% of this type of prescribing. Several β-blockers were prescribed: the most commonly prescribed were timolol maleate 0.1% (Novartis) and timolol maleate 0.25% accounting for 26.0% and 33.7% of the total β-blocker prescriptions, respectively. Betaxolol hydrochloride 0.5% (Alcon) accounted for 21.3% of the β-blocker prescriptions. Betaxolol hydrochloride 0.25% (Alcon) was used in 10.7% of prescriptions and levobunolol hydrochloride 0.5% (Allergan) was used in 8.3% prescriptions.
Sixty-four percent of the prostaglandin analogue and prostamide medicines were prescribed (as monotherapy or when added to other medicines) after glaucoma surgery. Fifty-eight percent of topical β-blocker prescriptions (as monotherapy or when added to other medicines) were issued after glaucoma surgery. Thirty-five percent of all prescriptions (556/1592 prescriptions) were of the type where multiple medicines were started simultaneously, hence no IOP analysis was attempted with any of these prescriptions. Pilocarpine hydrochloride 2% or pilocarpine hydrochloride 4% (Alcon) made up 14% of all types of prescriptions and were mostly used in combination with surgical treatments accounting for 23% of this type of prescribing (mostly after goniotomy surgery for PCG patients). Brimonidine tartrate 0.2% or apraclonidine 0.5% accounted for 17% of prescriptions associated with surgery.
The most frequently prescribed combination was dorzolamide hydrochloride 2%, timolol maleate 0.5% (Merck, Sharp & Dohme), which accounted for 15% of all types of prescriptions. The second most frequently used combination, latanoprost 50 mcg/mL, timolol maleate 5 mg/mL (Pfizer) was prescribed 44 times in 15.5% (31/200) of patients. The combinations of brimonidine tartrate 0.2%, timolol maleate 0.5%, and bimatoprost 300 mcg/mL, timolol maleate 5 mg/mL were prescribed only occasionally. Acetazolamide (Goldshield) was prescribed for a total of 8% (16/200) of patients at some point in the course of their management.
Section 2: The IOP-lowering Effect of Individual Medicines
Table 3 represents a summary of the IOP-lowering effects of the glaucoma medicines. We excluded from our analysis those prescriptions that were only used for <7 days before a change in the treatment plan. Eleven percent or less of the patients studied (when prescribed medication as a monotherapy or when medication was added) had undergone glaucoma surgery elsewhere before being referred to Moorfields Eye Hospital. The mean IOP-lowering effect of topical β-blockers (when used as a monotherapy) was −5.1 (SD 7.4) mm Hg after 36 prescriptions. The mean IOP-lowering effect of the prostaglandin analogue and prostamide medicines (monotherapy prescriptions only) was −1.8 (SD 5.7) mm Hg after 66 prescriptions. The median percentage IOP-lowering effect of topical β-blockers and the prostaglandin analogue, latanoprost 50 mcg/mL (when prescribed as a monotherapy) was −17.7% and −17.2%, respectively. The mean IOP-lowering effect of topical β-blockers when added to other medical therapies was −2.6 (SD 6.5) mm Hg after 31 prescriptions. The mean IOP-lowering effect of the prostaglandin analogue and prostamide medicines when added to other medical therapies was −3.1 (SD 8.3) mm Hg after 109 prescriptions.
There appeared to be no IOP-lowering effect when brimonidine tartrate 0.2% or apraclonidine 0.5% were used as monotherapies—with a mean IOP effect of +0.75 (SD 4.9) mm Hg when prescribed 11 times. Brimonidine tartrate 0.2% and apraclonidine 0.5% appeared to have an IOP-lowering effect when added to other therapies with a mean reduction of −2.2 (SD 7.4) mm Hg after 134 prescriptions. Overall our statistical analysis indicated that there was little evidence against the null hypothesis; that is, the average IOP-lowering effect was the same whatever drug was used (Kruskal-Wallis P-value 0.19). Of the prescriptions using the combination, dorzolamide hydrochloride 2%, timolol maleate 0.5%, there was a mean IOP reduction of −4.3 (SD7.6) mm Hg (after 29 prescriptions when used alone) and a mean of −4.3 (SD5.7) mm Hg (after 69 prescriptions when added to other medicines). The IOP-lowering effect of this combination was not statistically greater to the effect observed with the other medicines (two sample T tests >0.05).
Medication persistence that is the act of continuing a treatment for the prescribed duration (and measured by the duration of time from initiation to discontinuation of the treatment) is reported in Table 4. The topical β-blockers (when prescribed as a monotherapy) appeared to have the least persistence (92 d). The combination of dorzolamide hydrochloride 2%, timolol maleate 0.5% (when prescribed as a monotherapy) appeared to have the greatest persistence of 294 days. All the medicines when prescribed as an addition to other medicines appeared to have similar persistence except the combination of dorzolamide hydrochloride 2%, timolol maleate 0.5%, which again appeared to have the greatest persistence.
Section 3: The Side Effects Associated With Glaucoma Medicines
Side effects were recorded in the notes of 19.5% (39/200) of patients as summarized in Table 5. In our study, brimonidine tartrate 0.2% accounted for the highest rate of systemic side effects, as 17.4% (6/46) patients reported fatigue; of which 1 was found to be an infant (age 0.7 y) and the other 5 children with an age range from 6.6 to 14.4 years with a mean age of 10.1 years. Only 1.7% (2/117) of patients prescribed apraclonidine 0.5% experienced a systemic side effect. We found that 6.1% (8/132) of patients prescribed topical β-blockers experienced systemic side effects, of which 4 of the 8 appeared to be related to respiratory problems. Similarly, 4.6% (6/131) of patients experienced systemic side effects with the combination of dorzolamide hydrochloride 2%, timolol maleate 0.5%, of which 4 of the 6 appeared to be associated with respiratory problems. The fewest number of side effects occurred with the prostaglandin analogue and prostamide medicines [3.8% (6/158) patients] and the pilocarpines [3.7% (4/109) patients].
This study has attempted to update and further substantiate the findings of the BIG Eye study, which described the medical management (at 1 year of follow-up) of newly diagnosed pediatric glaucomas. Only 11% or less of the patients studied (when prescribed medication as a monotherapy or when added to other medicines) had undergone glaucoma surgery elsewhere (Table 3). This was not surprising as pediatric glaucoma patients are largely treated at tertiary centers in the United Kingdom (89% in the BIG Eye study).1 Therefore, this group of patients represented a reasonably wide range of pediatric glaucoma patients; from those with relatively straightforward glaucoma to those with more advanced or aggressive glaucoma. This should be borne in mind when applying the findings of this study, as one potential limitation was that the IOP-lowering effect of the glaucoma drugs might have been sometimes masked by patients with more aggressive disease.
The prescribing patterns for the pediatric glaucomas in this study were complex, partly because this was a heterogenous group of patients. The prostaglandin analogue and the prostamide medicines were prescribed more frequently than the topical β-blockers (27% vs. 12% of the time, for all types of prescriptions, respectively) and were the most commonly prescribed monotherapy (accounting for 39% of all monotherapy prescriptions). This prescribing behavior was unexpected as the clinicians would have followed the Moorfields’ Guidelines for Prescribing for Children, which recommend topical β-blockers as first-line therapy and prostaglandin analogues as second-line therapy. One explanation for this is that the clinicians’ prescribing choices would have been limited or influenced by the drug history of the patients who had been referred to Moorfields. The clinicians would also have referred to international guidelines such as the World Glaucoma Association’s “Medical Management of Glaucoma in Infants and Children,” which recommend that both topical β-blockers and prostaglandin analogues are effective treatments, with the caveat that the latter are less effective in children compared with adults.3
There did not appear to be a significant difference when comparing the IOP-lowering effect of the topical β-blockers versus the prostaglandin analogue and the prostamide medicines (with a P value of 0.19). It might be expected that if a medicine were more commonly used after glaucoma surgery then its IOP-lowering effect could appear more favorable than a medicine more commonly used before glaucoma surgery. This potential bias does not appear to be applicable in this study as the topical β-blockers and the prostaglandin analogue and prostamide medicines were prescribed to a similar extent (58% vs. 64% after surgery respectively). This type of prescribing (after surgery) also illustrated how glaucoma medicines played an essential role in keeping a patient’s IOP under control over the years, after partially successful surgical treatments.
An early study by Enyedi et al and a more recent randomized controlled trial by Plager et al used a 15% reduction in IOP to define a responder rate.4,5 Enyedi and colleagues reported only 19% patients (6/31 eyes) were responders to latanoprost 50 mg/mL with an IOP reduction of 8.5 mm Hg, that is the majority of patients were nonresponders. Plager and colleagues more recently reported a responder rate between 38% and 47% for betaxolol hydrochloride ophthalmic suspension 0.25% and timolol gel-forming solution 0.25% and 0.5%, respectively. Our study agreed with Plager and colleagues’ findings, in that 52.8% (28/53) of patients responded to latanoprost 50 mg/mL as monotherapy. A similar responder rate of 58.3% (21/36) of patients was measured when topical β-blockers were prescribed as monotherapy, indicating that both classes of medicines seemed to have a similar IOP-lowering effect. Moreover, our statistical analysis did not support any differences between the medicines, that is the average IOP-lowering effect was the same whatever drug was used (P value 0.19). There were far fewer prescriptions for some of the drugs compared with the prostaglandin analogue and prostamide medicines and the topical β-blockers, therefore, it should be noted that the lack of any statistical difference in IOP effect between drugs might be explained by the possibility of the study being underpowered to detect this difference.
In this study, acetazolamide was only used in 8% (16/200) of patients and no patient reported any side effects. Topical carbonic anhydrase inhibitors accounted for 9% of all prescriptions. The use of these medicines was relatively modest considering the World Glaucoma Association’s “Medical Management of Glaucoma in Infants and Children” recommendation that systemic and topical carbonic anhydrase inhibitors “can be safe and effective.” The combination, dorzolamide hydrochloride 2%, timolol maleate 0.5% accounted for 15% of all types of prescriptions and therefore probably affected the prescribing of the separate medicines. The efficacy and safety of carbonic anhydrase inhibitors have been reported in recent randomized controlled trials, namely by Whitson et al, Ott et al, and an earlier crossover study by Portellos et al.6–8 Although the combination of dorzolamide hydrochloride 2%, timolol maleate 0.5% did not appear to have a significantly greater effect on IOP compared with its separate constituents (P value of >0.05), it did have the greatest persistence compared with the other medicines. This may have been because patients adhered to combination medication better than the separate medications. Pilocarpine hydrochloride 2% and 4% when prescribed as a monotherapy appeared to have limited persistence mostly because they were prescribed shortly before glaucoma surgery.
This study had limitations as it was a retrospective study and designing it in terms of estimating its power was not undertaken. Although our IOP measurements were not standardized, the measurements included in the IOP analysis were only those taken using applanation tonometry methods. In addition, the IOP measurements may have been subject to regression toward the mean and the effect of nonadherence on IOP effect could not be assessed in this study. Moreover, a subanalysis of drug effects depending on disease severity was not undertaken as the sample size would have been insufficient. Nevertheless, the authors suggest that useful information in terms of the overall prescribing behavior of clinicians, insight into the overall IOP effect of medicines when used in this challenging group of patients and the side effects experienced have been gained from this study.
According to a study by Chung et al,9 all of the glaucoma medicines described in our study were listed as not recommended (animal studies or anecdotal cases of adverse drug reactions reported in the pediatric population) or not established (insufficient evidence to support a specific pediatric indication). Children may be at greater risk of systemic side effects compared with adults mainly due to differences in drug handling or metabolism, due to the ocular dosing not being adjusted for differences in weight and for some infants, their immature blood-brain barrier. The new European Regulation 1901/2206 on medicinal products for pediatric use (enforced since January 2007) and the World Health Organisation “Make Medicines Child Size” highlighted the importance of safe and effective drugs for children, hence one of the motivating factors of this study was to establish to what extent our patients were experiencing side effects.2,10 Although questions relating to problems with medication were asked at each visit, the occurrence of side effects in 19.5% of patients was likely to be an underestimate due to under reporting by carers and children and incomplete record keeping by the clinicians. The prostaglandin analogues were associated with local side effects only and our findings concur with the side effects experienced by adults where systemic problems are relatively uncommon.11 The side effect of cystoid macular edema induced by prostaglandin analogues did not appear to be experienced in this sample of patients; it is an uncommon side effect (suspected if there had been a change in vision) and is mostly a risk in aphakic and uveitic patients. This side effect could have been underestimated in, for example, the very young where visual measurements and macular structural investigations would have been difficult to obtain. The pilocarpines were also found to be associated with a low rate of systemic side effects (3.7% patients).
In our study, brimonidine tartrate 0.2% accounted for the highest rate of systemic side effects. Seventeen percent (6/46) of patients reported fatigue, of which 1 was found to be an infant (age 0.7 y) and the other 5 children were aged 6.6 to 14.4 years (mean age of 10.1 y). Brimonidine tartrate 0.2% should not have been prescribed in the child of 0.7 years; it is contraindicated in children under the age of 2 years in the World Glaucoma Association’s “Medical Management of Glaucoma in Infants and Children” and not recommended in children under the age of 6 years in the Moorfields Guidelines for Prescribing for Children. Brimonidine tartrate 0.2% can cause drowsiness due to CNS depression as its lipophilic structure allows it to cross the blood-brain barrier of infants. In 1998, Allergan issued a notification of labeling change, recommending that the drug not be used in children after 2 infants suffered bradycardia, hypotension, and apnea. Excessive sleepiness and lethargy with the frequency of these side effects increasing with low weight (<20 kg) and in the young (<6 y) have been reported by numerous studies.12–14 Our study agreed with the findings of earlier studies, indicating that Brimonidine tartrate 0.2% still caused side effects in older children and should be used with caution. In contrast, far more children in our study (117) were prescribed Apraclonidine 0.5% and only 2/117 (1.7%) experienced a systemic side effect. It appeared to be a much safer drug to prescribe in children.
It is well known that topical β-blockers may cause adverse respiratory effects. In our study, we found that 6.1% (8/132) of patients prescribed topical β-blockers experienced systemic side effects of which 4 of the 8 were related to respiratory problems. None of the patients in this review were known to have any preexisting history of asthma. Two drops of Timolol maleate 0.25% may equate to a 10 mg oral dose capable of lowering blood pressure significantly in the elderly. Unrecognized respiratory impairment can be a significant problem. One study reported improvements in spirometry and exercise tolerance of 8% and 11%, respectively, in patients over 60 years with no known history of airways disease when switched from timolol to other glaucoma drugs of the time.15 Diggory et al16 concluded that “By performing spirometry before starting topical β-antagonist therapy and repeating it after 1 month most patients at risk of respiratory impairment can be identified.” This would be a prudent way to introduce topical β-blockers in children who can perform this test. Systemic side effects with the topical carbonic anhydrase inhibitors and the combination of dorzolamide hydrochloride 2%, timolol maleate 0.5% were experienced in 5.5% and 4.6% of patients, respectively, slightly higher than reported in other studies.6,7
Our study has tried to provide further insight into the use of glaucoma drugs in children, describing the currently prescribed medications and attempting to provide some quantitative confirmation of the effect of medical treatments in pediatric glaucomas. The prescribing choices in this study were in line with international guidelines such as the World Glaucoma Association's Medical Management of Glaucoma in Infants and Children. Although it is generally acknowledged that pediatric glaucomas are more often than not treated surgically, our study showed that all the glaucoma medicines play an important role in reducing IOP after glaucoma surgery. The findings of this study suggested that there was no statistically significant difference in IOP-lowering effect between the different glaucoma medicines, and that the topical β-blockers and prostaglandin analogue and prostamide medicines appeared to have similar responder rates. Hence given that the latter class of medicines appeared to have the lowest occurrence of systemic side effects, it would be advisable to consider the prostaglandin analogue and prostamide medicines as a good first or second choice. The combination of dorzolamide hydrochloride 2%, timolol maleate 0.5% had the greatest persistence (of just <1 y) suggesting that its ability to lower IOP effectively can be sustained. We acknowledge that this study (due to its retrospective nature) was an exploratory rather than a definitive study to assess the IOP-lowering effects of glaucoma medicines in children, hence it is hoped that this information will support other and future studies.
The authors thank Special Trustees, National Institute for Health Research Biomedical Research Centre (NIHR BMRC), Unresricted grant Pfizer, Medicines for Children Research Network.
1. Papadopoulos M, Cable N, Rahi J, et al..The British Infantile and Childhood Glaucoma (BIG) Eye Study.Invest Ophthalmol Vis Sci.2007;48:4100–4106.
2. Official Journal of the European Union Regulation (EC) No 1901/2006 of the European Parliament and of the Council of the 12 December 2009 on medicinal products for paediatric use and amending Regulation (EEC) No 1768/92, Directive 2001/20/EC, Directive 2001/83/EC and Regulation EC (No) 726/2004; 27.12.2006: L378/1.
3. World Glaucoma Association’s “Medical Treatment of Glaucoma” *Medical Management of Glaucoma in Infants and Children”Weinreb RN, Araie M, Susanna R, et al..Medical Treatment of Glaucoma.Consensus Series 7.2010:Chapter 10.Amsterdam, The Netherlands:World Glaucoma Association Kugler Publications;267–271.
4. Enyedi LB, Freedman SF, Buckley EG, et al..The effectiveness of latanoprost for the treatment of paediatric glaucoma.J AAPOS.1999;3:33–39.
5. Plager DA, Whitson JT, Netland PA, et al..Betaxolol hydrochloride ophthalmic suspension 0.25% and timolol gel-forming solution 0.25% and 0.5% in paediatric glaucoma: a randomised clinical trial.J AAPOS.2009;13:384–390.
6. Whitson JT, Roarty JD, Vijaya L, et al..Efficacy of brinzolamide and levobetaxolol in pediatric glaucomas: a randomized clinical trial.J AAPOS.2008;12:239–246.e3.
7. Ott EZ, Mills MD, Arango S, et al..A randomized trial assessing dorzolamide in patients with glaucoma who are younger than 6 years.Arch Ophthal.2005;123:1177–1186.
8. Portellos M, Buckley EG, Freedman SF, et al..Topical versus oral carbonic anhydrase inhibitor therapy for paediatric glaucoma.J AAPOS.1998;2:43–47.
9. Chung I, Buhr V, Chung I, et al..Topical ophthalmic drugs and the paediatric patient.Optometry.2000;71:511–518.
10. World Health Organisation. Promoting Safety of Medicines For Children. ISBN 978-92-4-156343-7.
11. Alm A, Grierson I, Shields MB, et al..Side effects associated with prostaglandin analog therapy.Surv Ophthalmol.2008;53suppl 1S93–S105.
12. Al-Shahwan S, Al-Torbak AA, Turkmani S, et al..Side-effect profile of brimonidine tartrate in children.Ophthalmology.2005;112:2143–2148.
13. Enyedi LB, Freedman SF, Enyedi LB, et al..Safety and efficacy of brimonidine in children with glaucoma.J AAPOS.2001;5:281–284.
14. Bowman RJ, Cope J, Nischal KK, et al..Ocular and systemic side effects of brimonidine 0.2% eye drops (Alphagan) in children.Eye.2004;18:24–26.
15. Diggory P, Cassels-Brown A, Vail A, et al..Avoiding unsuspected respiratory side effects of topical timolol with cardioselective or sympathomimetic agents.Lancet.1995;345:1604–1606.
16. Diggory P, Cassels-Brown A, Vail A, et al..Randomised, controlled trial of spirometric changes in elderly people receiving timolol or betaxolol as initial treatment for glaucoma.Br J Ophthalmol.1998;82:146–149.