Incidence of Ocular Side Evects of Topical Β Blockers in the Netherlands

856 Br J Ophthalmol 2000;84:856–859

Incidence of ocular side eVects of topical â Br J Ophthalmol: first published as 10.1136/bjo.84.8.856 on 1 August 2000. Downloaded from blockers in the Netherlands

Luc M van Beek, RobJWdeKeizer, BettineCPPolak, Paul R Elzenaar, Nicolaas J van Haeringen, Aize Kijlstra

Abstract literature. These side eVects include pseudo- Background—Several ocular side eVects ocular cicatricial pemphigoid,5 conjunctival including uveitis, have been reported keratinisation,6 and anterior uveitis.7 Although following topical â blocker treatment for anterior uveitis has been described after glaucoma and ocular hypertension. The timolol8 and betaxolol use,9 its occurrence is incidence of these side eVects was investi- best documented after metipranolol use.7 10–13 gated in the Netherlands. In three diVerent retrospective studies the Methods—A prospective observational reported incidence of metipranolol associated design was used whereby monthly ques- uveitis varied from 6/1000 patients7 to less than tionnaires were sent to all practising oph- 1.87/1000 patient years.14 15 thalmologists in the Netherlands during 3 To our knowledge, except for uveitis, no consecutive months. Questionnaires were incidence ﬁgures of ocular side eVects of topi- returned at the end of each month. Any cal â blockers are available for the general patient whose topical â blocker therapy patient population. Therefore, we undertook a was altered because of an ocular reaction observational, nationwide study to determine was noted on this questionnaire. Ophthal- the incidence of ocular side eVects, including mologists who did not return their ques- uveitis, of topical â blockers in the Nether- tionnaires were interviewed by telephone lands. This study shows that 1.5 cases of ocular at the end of the study period. The number side eVects of topical â blockers per 1000 of patients using topical â blockers was patient years can be expected. derived from drug sales ﬁgures. Results—70% (328/467) of the ophthalmologists in the Netherlands participated Methods Department of in the study. During the 3 month study During three consecutive months (April, May, Ophthalmology, period 34 cases were reported: 15 patients June 1997) all practising ophthalmologists in Leiden University had periorbital dermatitis, in eight pa- the Netherlands received a monthly question- Medical Center, naire. Addresses of ophthalmologists were Netherlands tients eyelids and conjunctiva were af- L M van Beek fected, in seven patients the conjunctiva obtained from the Dutch Ophthalmological http://bjo.bmj.com/ RJWdeKeizer was aVected, and four patients had punc- Society (NOG) of which virtually all ophthal- tate keratitis. The calculated incidence of mologists are member in the Netherlands. At Department of ocular side eVects during topical â blocker the end of each month the questionnaires were Ophthalmology, Free therapy was 1.51 cases/1000 patient years. returned. All patients who were diagnosed as University Hospital, having ocular side eVects associated with topi- Amsterdam, Conclusion—Topical â blocker therapy is Netherlands associated with few clinically important cal â blocker use in that month, were noted on BCPPolak this questionnaire. Information about systemic

ocular side eVects. No cases of uveitis were on September 26, 2021 by guest. Protected copyright. reported. diseases, patient characteristics (age, sex), â Tramedico BV, (Br J Ophthalmol 2000;84:856–859) blocker used, period of time that the â blocker Medical Department, had been used, co-medication, type of ocular Weesp, Netherlands reaction, and improvement after discontinua- P R Elzenaar Topically applied â adrenergic antagonists (â tion of the drug was reported on this question- Netherlands blockers) are the first line of treatment in naire. Ophthalmic Research primary open angle glaucoma and ocular To achieve a high response rate to our ques- Institute hypertension. Safety of topical â blockers is tionnaire, we needed a simple and unambigu- L M van Beek mainly determined by their systemic â block- ous definition of ocular side eVects of topical â N J van Haeringen ing eVects. The bronchoconstriction and car- blockers. Ocular side eVects were defined as A Kijlstra diac arrhythmias caused by topical â blockers any alteration of topical â blocker medication 1 Institute for Animal may be life threatening. However, ocular side because of an ocular reaction. Alterations that Science and Health, eVects may also necessitate discontinuation or were made because of ineYcacy of â blocker Lelystad, Netherlands changing of topical â blocker therapy. therapy were not included in this study. In the A Kijlstra Various ocular side eVects have been re- Netherlands â blockers are available only on ported after the use of topical blockers. medical prescription. Alterations that were Correspondence to: â L M van Beek, MD, Leiden Blepharoconjunctivitis, conjunctivitis, and made because of systemic side eVects of topical University Medical Center, punctate keratitis have been described in clini- â blockers fell beyond the scope of this study. Department of cal trials.2 In these trials, discontinuation of Ocular side eVects can be caused both by the Ophthalmology, PO Box 9600, 2300 RN Leiden, topical â blockers because of ocular side eVects active compound or by one of the auxiliary Netherlands ranged from 0% to 4.3%, independent of dose components in the eye drop. Because the [email protected] or which â blocker was used.34 objective of our study was to determine the Accepted for publication Other ocular side eVects have only become incidence of ocular side eVects of topical â 24 February 2000 apparent through case reports in the medical blockers as they are used in the Dutch patient

www.bjophthalmol.com Incidence of ocular side eVects of topical â blockers in the Netherlands 857

population, no distinction was made between Results Br J Ophthalmol: first published as 10.1136/bjo.84.8.856 on 1 August 2000. Downloaded from reactions to the diVerent components. Information was obtained from 365 of the 467 Ophthalmologists who did not return their practising ophthalmologists in the Netherlands questionnaires were interviewed by telephone (78.2%). Of the responding ophthalmologists at the end of the three months’ study period. In 37 (10.1%) did not participate in the study for these interviews we asked whether the ophthal- various reasons. Information obtained from mologist had encountered any ocular side 328 (70.1%) ophthalmologists in the Nether- eVect of topical â blocker therapy during the lands could be used in the study. study period. If the answer was negative then During the study period we received infor- the interview was ended, if the answer was mation on 34 patients who met our definition positive the same information as with the ques- of an ocular side eVect. Mean age of the tionnaire was collected. patients was 71.3 years, there were 12 men and In the Netherlands, all sales of pharmaceuti- 22 women. Patients with ocular side eVects had cal wholesalers and pharmaceutical industries used their topical â blocker for a median time to local pharmacies are recorded. Sales figures of 14 months (range 1–144 months, Fig 1). In during the study months of all â blocking drugs 28 patients clinical improvement occurred for ophthalmic use (including all brands and after alteration or cessation of the â blocker. generics) were used to estimate the number of No follow up was available for the remaining patients using topical â blocker medication six cases. (Pharm inform BV, FI-rom, June 1997). It was No cases of uveitis were reported. Most assumed that one patient uses one vial of oph- often the eyelids were aVected (15 patients, thalmic â blocker in 1 month. 44%), these patients usually had periorbital dermatitis or blepharitis. In eight patients 160 (23%) both conjunctiva and eyelids were aVected; these patients had blepharoconjuncti- 140 vitis or periorbital dermatitis combined with conjunctivitis or conjunctival hyperaemia. Of 120 seven patients (21%) only the conjunctiva was aVected; these patients had conjunctivitis or 100 conjunctival hyperaemia. The cornea was aVected in four patients (12%); all these 80 patients had punctate keratitis. Analysis of the alterations of topical â blocker medication showed the following data 60 Time (months) Time (Fig 2). In 12 patients timolol medication was altered. In the majority of these patients (eight) 40 timolol was replaced by a preservative-free timolol formulation. Levobunolol medication 20 Median was changed in five patients. Both betaxolol http://bjo.bmj.com/ (14 months) and metipranolol medication were changed in 0 Timolol Betaxolol Levobunolol Carteolol Metipranolol four patients. Carteolol medication was β Blocker changed in three patients. In one patient the bufenolol medication was replaced by another Figure 1 Time that patients used the topical â blocker before it was altered because of ocular side eVects. For seven patients this information could not be obtained (n=27). â blocker. From the telephone inquiries it Median time the patients used their topical â blocker was 14 months (range 1–144 appeared that ophthalmologists often hold the months). preservative responsible for the occurrence of ocular side eVects. However, in five patients, on September 26, 2021 by guest. Protected copyright. who already used either preservative-free 12 timolol (four patients) or preservative-free Stopped eye medication metipranolol (one patient), ocular side eVects 10 Switched to different topical β blocker necessitated a change in medication. Switched to preservative free topical β blocker Preservative-free â blockers were most often Switched to different topical antiglaucoma drug replaced by a diVerent topical antiglaucoma 8 drug (dorzolamide or latanoprost). Five patients, four on timolol and one on 6 carteolol, used additional topical antiglaucoma drugs besides their topical â blocker. Two No of cases patients used one additional topical antiglau- 4 coma drug and three patients used two additional topical antiglaucoma drugs. These 2 drugs were dipifevrin (two patients) dorzolamide (two patients) carbacholine (one patient), pilocarpine (one patient), apraclonidine (one 0 Timolol Levobunolol Betaxolol Metipranolol Carteolol Bufenolol Preservative patient), and guanethidine/adrenaline (epine- free phrine) combination (one patient). In one Figure 2 Switches made in topical â blocker therapy because of an ocular side eVect patient, only the â blocker medication was (n=34). Preservative free denotes all patients altering their preservative-free â blocker altered, while the additional topical medication medication because of an ocular side eVect. In this latter group, four patients used was continued. In three patients both the topi- preservative-free timolol and one patient used preservative-free metipranolol. Whether the patient was prescribed a diVerent â blocker, a preservative-free â blocker, a diVerent topical cal â blocker and one additional topical medi- antiglaucoma drug, or stopped topical antiglaucoma therapy is also shown. cation were altered. In one patient the â

www.bjophthalmol.com 858 Van Beek, de Keizer, Polak, et al

Table 1 Calculated incidence of ocular side eVects of topical â blockers ever, skin tests were not routinely performed to Br J Ophthalmol: first published as 10.1136/bjo.84.8.856 on 1 August 2000. Downloaded from confirm this diagnosis. Contact hypersensitiv- Vials sold during Incidence (cases/1000 patient years) â Blocker Cases* study period† (95% confidence interval)‡ ity to benzalkonium chloride can cause (kera- to)conjunctivitis and contact dermatitis of the bufenolol 1 2289 5.24 (0, 15.51) eyelids.16 In patients with various forms of con- metipranolol 5 14288 4.20 (0.52, 7.88) carteolol 3 13620 2.64 (0, 5.63) junctivitis or contact dermatitis, a sensitisation levobunolol 5 29611 2.03 (0.25, 3.81) rate of approximately 6% to benzalkonium timolol 16 151434 1.27 (0.65, 1.89) 17 18 betaxolol 4 58478 0.82 (0.02, 1.60) chloride has been found. Patients who total 34 269720 1.51 (1.00, 2.02) discontinued their preservative-free topical â blocker in the present study because of ocular *Cases were defined as any patient whose topical â blocker therapy was altered because of an ocular reaction open to objectification by the ophthalmologist. side eVects may have been sensitised to the â †Sales figures of topical â blockers were used to estimate the number of patients using topical â blocker itself. Contact hypersensitivity has blockers. It was assumed that one patient uses one vial of â blocker in 1 month. been described for all â blocking agents ‡No significant diVerences were found between â blockers. registered for topical use in the Netherlands blocker medication and two diVerent addi- except carteolol.18 tional topical eye medications were altered. Several factors may have biased our results. During the study period, a total of 269 720 Firstly, 30% of the ophthalmologists in the vials of â blocking agents for ophthalmic use Netherlands did not participate. However, it were sold in the Netherlands. â Blockers are may be argued that ophthalmologists who did available only on medical prescription at phar- not encounter any cases of ocular side eVects to macies. Vials are packaged to last 1 month. topical â blocker therapy during the study Extrapolation of these data indicates that period are less likely to return the question- 89 907 individuals in a population of 15 608 naire than ophthalmologists who did encoun- 591 (CBS Voorburg), used a topical â blocker ter such cases. This was supported by the fact during the study period. The incidence of ocu- that telephone interviews with ophthalmolo- lar side eVects to topically applied â blocking gists who had not returned their questionnaire, drugs in this study can thus be calculated as yielded no additional cases of ocular side 1.51/1000 patient years for the total patient eVects. Other than this, we do not know population using a topical â blocker. Incidence whether responding ophthalmologists diVered calculated for the diVerent â blockers, varied from non-responding ophthalmologists, al- from 0.82 (betaxolol) to 5.24 (bufenolol) though we have no reason to think they did. cases/1000 patient years but these diVerences Therefore, bias from non-participation prob- were not statistically significant (Table 1). ably has exerted only a minimal eVect on the results. Another factor of bias may have come from Discussion the fact that some patients used additional This study shows that clinically relevant ocular topical eye medication or systemic medication side eVects associated with topical â blockers besides their topical â blocker medication. In

rarely occur. Reactions that did occur in our four patients not only was the topical â blocker http://bjo.bmj.com/ study most often involved the eyelids and con- medication altered but also one or two junctiva, the cornea was aVected in only a few additional eye medications. The ocular side cases. No cases of uveitis were reported eVects of these four patients, therefore, may whereas Akingbehin and Villada7 have re- not have been caused by the topical â blocker. ported an incidence of metipranolol associated Another two patients used systemic as well as uveitis of 6/1000 patients. However, high topical â blockers. Systemic â blockers have incidence rates for metipranolol associated been reported to reduce tear production19 and uveitis have never been found outside of the glaucoma itself has also been associated with a on September 26, 2021 by guest. Protected copyright. United Kingdom. Kessler and Christ14 found decreased tear production.20 This may aggra- an incidence rate of 0.5/1000 patient years in vate ocular side eVects caused by the eye drops. Germany and in the United States, Beck et al15 Another factor of bias may have been misdi- did not ﬁnd any case of uveitis associated with agnosis. Because the study was designed to metipranolol use in a sample of 1972 patient interfere as little as possible with the day to day years. Similarly, in our sample of 1191 patient practice of the responding ophthalmologists, years of metipranolol use no cases of uveitis the diagnosis of ocular side eVect was not veri- were reported. ﬁed by a rechallenge or skin testing. However, Conjunctivitis and periorbital dermatitis symptoms improved in all patients, in whom after topical â blocker use can be caused by follow up was available (n=28), after the hypersensitivity to the preservative or to the â suspected â blocker was changed for another blocker itself.16 Because the aim of our study antiglaucoma therapy. was to determine the incidence of ocular side In estimating the number of patients using eVects of topical â blockers as they are used in topical â blockers in the Netherlands the the general patient population, no distinction following bias may have been introduced. We was made between reactions to these two assumed that patients use one vial of â blockers diVerent components. Benzalkonium chloride in 1 month. However, patients may use a vial is used as preservative in all commercially for a longer or shorter period. In the Dutch available preserved â blockers for topical use. health care system costs for â blocker therapy In the present study patients switching to a are completely reimbursed, except for preservative-free formulation are presumably preservative-free formulations, which are diagnosed by the ophthalmologist as being partly reimbursed. â Blockers are packaged to hypersensitive to benzalkonium chloride. How- last a month and it is recommended to patients

www.bjophthalmol.com Incidence of ocular side eVects of topical â blockers in the Netherlands 859

to start on a new vial every month. If patients ophthalmic solution, 1978–1985. Am J Ophthalmol Br J Ophthalmol: first published as 10.1136/bjo.84.8.856 on 1 August 2000. Downloaded from use a vial longer than one month the incidence 1986;102:606–11. 2 Weinreb RN, Caldwell DR, Goode SM, et al. A double- per patient years will decrease; obviously, the masked three-month comparison between 0.25% betaxolol incidence per vial remains the same. suspension and 0.5% betaxolol ophthalmic solution. Am J Ophthalmol 1990;110:189–92. Furthermore, a factor that may have influ- 3 Krieglstein GK, Novack GD, Voepel E, et al. Levobunolol enced our result was that we based our calcula- and metipranolol: comparative ocular hypotensive eYcacy, tion of the number of patients on topical â safety, and comfort. Br J Ophthalmol 1987;71:250–3. 4DenVer H. EYcacy and tolerance of metipranolol: results of blocker medication on drug sales from indus- a multi-center long term study. In: Merte HJ, ed. tries and wholesalers to local pharmacies. Metipranolol. New York: Springer-Verlag,1983:121–5. However, these sales figures reliably reflect the 5 Fiore PM, Jacobs IH, Goldberg DB. Drug-induced pemphigoid. A spectrum of diseases. Arch Ophthalmol sales to patients because in the Netherlands 1987;105:1660–3. local pharmacies do not keep large stocks, 6 Derous D, de Keizer RJ, de WolV-Rouendaal D, et al. Con- junctival keratinisation, an abnormal reaction to an ocular thanks to a dense distribution network. This beta-blocker. Acta Ophthalmol 1989;67:333–8. was confirmed by the fact that no substantial 7 Akingbehin T, Villada JR. Metipranolol-associated granulo- diVerences were noticed in quarterly sales matous anterior uveitis. Br J Ophthalmol 1991;75:519–23. 8 Zimmerman TJ, Baumann JD, Hetherington J Jr. Side figures of topical â blockers throughout the eVects of timolol. Surv Ophthalmol 1983;28(Suppl):243– year. As mentioned previously, in the Nether- 51. lands patients can only obtain topical block- 9 Jain S. Betaxolol-associated anterior uveitis. Eye 1994;8: â 708–9. ers on medical prescription via local pharma- 10 Burvenich H. Metipranolol associated granulomatous ante- cies. rior uveitis: not so uncommon as thought. Bull Soc Belg Ophtalmol 1995;257:63–6. Ophthalmologists who failed to return their 11 Melles RB, Wong IG. Metipranolol-associated granuloma- questionnaires were interviewed at the end of tous iritis. Am J Ophthalmol 1994;118:712–15. the 3 month study period. In recalling ocular 12 Watanabe TM, Hodes BL. Bilateral anterior uveitis associated with a brand of metipranolol. Arch Ophthalmol side eVects from the past 3 months, bias may 1997;115:421–2. have been introduced to our results. 13 Patel NP, Patel KH, Moster MR, et al. Metipranolol- The overall eVect of bias in our study may associated nongranulomatous anterior uveitis. Am J Ophthalmol 1997;123:843–3. have resulted in a small underestimation of the 14 Kessler C, Christ T. Incidence of uveitis in glaucoma incidence. The incidence we have found of patients using metipranolol. J Glaucoma 1993;2:166–70. 1.51/1000 patient years should therefore be 15 Beck RW, Moke P, Blair RC, et al. Uveitis associated with topical beta-blockers. Arch Ophthalmol 1996;114:1181–2. considered as a minimum. 16 Herbst RA, Maibach HI. Contact dermatitis caused by We conclude that topical â blockers cause allergy to ophthalmics: an update. Contact Dermatitis 1992; 27:335–6. 1.51 cases/1000 patient years of ocular side 17 Rudzki E, Kecik T, Rebandel P, et al. Frequency of contact eVects. In our study none of the topical â sensitivity to drugs and preservatives in patients with con- blockers gave rise to (anterior) uveitis. When junctivitis. Contact Dermatitis 1995;33:270. 18 Perrenoud D, Bircher A, Hunziker T, et al. Frequency of new drugs for the treatment of glaucoma are sensitization to 13 common preservatives in Switzerland. compared with â blockers, safety, tolerance, Swiss Contact Dermatitis Research Group. Contact Derma- comfort, and costs should be considered as titis 1994;30:276–9. 19 O’Donnell BF, Foulds IS. Contact allergy to beta-blocking well as clinical eYcacy. agents in ophthalmic preparations. Contact Dermatitis 1993; :121–2. 28 http://bjo.bmj.com/ 20 Petounis AD, Akritopoulos P. Influence of topical and This research was supported by a grant from Dr Gerhard Mann systemic beta-blockers on tear production. Int Ophthalmol GmbH, Germany, and Tramedico BV, Netherlands. 1989;13:75–80. 21 Kuppens EV, van Best JA, Sterk CC, et al. Decreased basal 1 Nelson WL, Fraunfelder FT, Sills JM, et al. Adverse respira- tear turnover in patients with untreated primary open- tory and cardiovascular events attributed to timolol angle glaucoma. Am J Ophthalmol 1995;120:41–6. on September 26, 2021 by guest. Protected copyright.