Safety and Efficacy of 2% Pirenzepine Ophthalmic Gel in Children WithMyopia:  A 1-Year, Multicenter, Double-Masked, Placebo-Controlled ParallelStudy FREE

Myopia, one of the most common ocular disorders in the world, is a significantglobal public health problem. It affects at least 25% of the adult populationin the United States.^1- 3 Myopiais among the 5 conditions that have been identified as immediate prioritiesby the World Health Organization in its Global Initiative for the Eliminationof Avoidable Blindness.⁴

Probably the most widely studied pharmacological treatment for myopiahas been the use of atropine, a nonselective muscarinic antagonist. Severalrecent controlled clinical trials have provided evidence that atropine canretard myopia progression in children.^5- 7 Animalstudies suggest that this retardation occurs independent of the effect ofatropine on accommodation.^8- 12 Pirenzepinehydrochloride, another muscarinic receptor antagonist, has long been usedorally in Europe to treat dyspepsia and pediatric endocrine disorders andhas an extensive clinical history and excellent safety profile.¹³ Unlikeatropine, which is equipotent in binding to M₃ (accommodation andmydriasis) and M₁ muscarinic receptors, pirenzepine is relativelyselective for the M₁ muscarinic receptor^14- 15 andthus less likely than atropine to cause mydriasis and cycloplegia. Since Stoneet al¹² first suggested study of pirenzepineas an agent to decrease myopic progression, animal research has demonstratedthat it reduced the development of deprivation-induced myopia and axial elongationin animals.^16- 18 Basedon previous phase 1 trials of the safety and tolerability of pirenzepine solutionin adults¹⁹ and pirenzepine ophthalmic gelin children,²⁰ we undertook a double-masked,multicenter, placebo-controlled phase 2 trial to evaluate the safety and efficacyof pirenzepine ophthalmic gel in slowing the progression of myopia in school-agedchildren. The maximally tolerated concentration in the previous study, 2%pirenzepine,²⁰ was selected for evaluationin the present study.

This was a parallel-group, placebo-controlled, double-masked study conductedfrom March 1, 2000, to February 28, 2002, at 13 US sites (academic clinicsand private practices). Children were randomized in a 2:1 ratio to receive2% pirenzepine twice daily or placebo control for 1 year. The 2:1 randomizationschema was used to increase the chance of a patient being randomized to theactive treatment arm, while maintaining sufficient power. A central coordinatingcenter was located at Valley Forge Pharmaceuticals Inc (Irvine, Calif). Clinicalresearch monitors for each site were provided via Clinical Studies ManagementGroup, Tulsa, Okla.

Eligible patients were healthy children, aged 8 to 12 years, with myopia,defined as a spherical equivalent (SEQ) of −0.75 to −4.00 diopters(D) and astigmatism of 1.00 D or less in each eye as measured by cycloplegicrefraction. Also required were normal pupils and best-corrected visual acuityof 20/25 or better in each eye by Early Treatment Diabetic Retinopathy Study(ETDRS)charts.²¹ A modified ETDRS procedure was usedto measure visual acuity monocularly at distance and binocularly at near.Starting with the 20/50 line (or 20/100 line when any letter was missed onthe 20/50 line), the child read all 5 letters on each subsequent line untilhe/she missed 3 or more letters on a line. Visual acuity was scored as thelogMAR score for the line plus 0.02 times the total number of letters missedon that line. Ocular exclusion criteria were anisometropia greater than 1.00D in SEQ, any manifest tropia, current use of either contact lenses or bifocals,and a history of ocular surgery, trauma, or chronic ocular disease, includingallergic conjunctivitis. Systemic criteria for exclusion from the study werediseases that required long-term or regular intermittent medication use (eg,asthma, epilepsy); behavioral or neurological disorders that would interferewith the study; participation in any study that involved an investigationaldrug within the month before enrollment; intolerance or hypersensitivity totopical anesthetics, mydriatics, or components of the formulation (eg, benzalkoniumchloride); contraindications to antimuscarinic agents; and pregnancy or plannedpregnancy. Transient pharmacologic therapy for acute diseases was allowed(eg, otitis media, pharyngitis).

The protocol and informed consent and child assent forms were approvedby institutional review boards. The parent or guardian of each study patientgave written informed consent, and the patient provided written assent.

We formulated 2% pirenzepine with hydroxypropyl methylcellulose andpreserved it with 0.005% benzalkonium chloride. Both pirenzepine and placebo(identical formulation except for the pirenzepine) were packaged in identicaltubes, the identity of which was masked from the children, parents, and investigators.The product that contained active drug and the vehicle were identical in appearance.Study medications were administered twice daily as an approximately ¼-in(6-mm) strip in the cul-de-sac of the lower eyelid. Approximately 9 monthsafter study began, an electronic monitoring device (MEMS SmartCap system;Aardex, Union City, Calif) was introduced into the trial. The study medicationwas placed into an outer standard medication bottle. To access the medication,the SmartCap had to be removed. The device recorded and stored a bottle openingevent (limit 1 per 15-minute period to reduce double counts). SmartCap datawere retrieved during the patient’s regularly scheduled visit.

Height and weight were recorded, and a baseline, predrug symptom queryfor symptoms that existed before study drug instillation was administered.Monocular and binocular best-corrected visual acuity was measured at distanceand near using the ETDRS charts. A comprehensive eye examination, includingmeasurement of intraocular pressure, was performed to rule out any exclusioncriteria. Autorefraction (Speedy 1; Nikon, Melville, NY) was performed 30to 60 minutes after instillation of 0.5% proparacaine hydrochloride, 1.0%cyclopentolate hydrochloride, and 1.0% tropicamide in each eye with 1 minuteof eyelid closure. A-scan ultrasonography was used to measure axial lengthvia the standard fashion of each site. All qualified patients were providedwith a demonstration by the clinical staff of how to properly instill thestudy medication using a tube of placebo gel. They were then provided witha tube of the placebo gel to take home to practice instillation to ensurethat the child and parents were able to use the gel. The child then returnedwithin 1 week for randomization.

Before randomization, individuals were queried regarding ability toadminister the agent and any potential adverseeffects. No child or parentreported any problems with administration or use of the vehicle gel. Eligiblepatients were randomized either to active or placebo treatment using a sponsor-prepared,computer-generated randomization list stratified by site (PROC PLAN; SAS version8, SAS Institute Inc, Cary, NC). Study medication was administered by studypersonnel, and 10 and 60 minutes later, a symptom query was given and vitalsigns were measured. At 60 minutes, pupil size was recorded and biomicroscopyperformed. Subsequent visits were scheduled at 15 days and 1, 3, 6, 9, and12 months. At each visit, a symptom query was administered and visual acuity,pupil size, anterior segment evaluation, intraocular pressure, heart rate,and blood pressure were measured. At months 3, 6, and 12, cycloplegic autorefractionand A-scan ultrasonography were performed. Autorefraction was also performedat month 9.

The primary outcome measure was SEQ. We assumed that the progressionin the placebo group would be at least −0.3 ± 0.3 D/y(Karla Zadnik, OD, PhD, Orinda Longitudinal Study of Myopia, National Institutesof Health, National Eye Institute, grant 10-EY08893, unpublished data, 1989-2000),with an arbitrary 50% difference between treatments in change from baselinein SEQ cycloplegic refraction. The study had 90% power to detect a mean differencebetween treatment groups of approximately 0.17 D in myopia (α = .05,1-tailed, SD = 0.3 D,t test), assuming144 evaluable children at 1 year (96 pirenzepine patients and 48 placebo patients).The secondary outcome measure was axial length.

In this bilateral treatment study, continuous measures (eg, sphere,pupil size) were averaged between eyes for analysis. Change from baselinerefractive error and axial length was analyzed by a mixed-model analysis ofcovariance with baseline as the covariate; treatment group, site, and theirinteraction as between-patient factors in the model; and the repeated measures(visits and right and left eyes) and their interaction with the between-patientfactors as within-patient factors. Nonsignificant interactions were droppedfrom the model used to estimate treatment effect. Safety measures made oncontinuous scales were analyzed in a manner similar to the efficacy measures.The analysis of safety measures made on categorical or frequency scales wasbased on χ² statistics. All analyses were performed usingPC-SAS (version 8.1; SAS Institute Inc). A planned interim analysis was conductedwhen 50% of patients reached the primary end point (12 months). The null hypothesisof no treatment effect was tested with a P valueof.005 and not rejected. The criterion P value forstatistical significance in the final analysis was adjusted to.048 to maintainan experiment-wise type I error of less than.05.²²

Demographics and baseline characteristics of the 174 patients enrolledare given in Table 1.

We screened 277 patients and enrolled 174 patients, of which 145 completedthe trial. Twenty-six (22%) of 117 patients randomized to pirenzepine didnot complete the study compared with 3 (5%) of the 57 patients in the placebogroup (Figure 1, P = .005). Thirteen of the 26 pirenzepine-treated patientsstopped participation in the study due to an adverse event: allergic reactionor conjunctivitis, 7; accommodation or blurred vision at near, 5; and stinging,1. Thirteen patients (11%) randomized to pirenzepine did not complete thestudy for reasons other than adverse events: noncompliance with medication, 3;lost to follow-up, 2; and other reasons, 8 (missed visits, withdrew consent,“didn’t like study medication,” moved, use of prohibitedconcomitant systemic medication, or personal reasons). Three patients (5%)randomized to placebo did not complete the study due to the following reasons:noncompliance with medication, 1; and other reasons, 2 (patientunable to accept ultrasound, patient “got tired of using the gel”).For 5 of the 29 patients, discontinuation occurred within the first few daysdue to either intolerance to the study medication or the study procedures.However, for the most part, discontinuation occurred after the patients werein the study for several months or more. When a compliance measurement wasintroduced partway through the study, the mean compliance ratio for each groupwas 79%. No patient’s treatment was unmasked early.

At study entry, mean ± SD SEQ refraction was −2.098 ± 0.903D for the pirenzepine group and −1.933 ± 0.825 D forthe placebo group (P = .22). As shown in Figure 2, during the 1-year study, the mean SEQrefraction became more myopic in both treatment groups. At 3, 6, 9, and 12months, the mean increase in myopia was significantly higher (P ≤ .006) in the placebo group compared with the pirenzepinegroup (Table 2). At 12 months, therewas a mean increase in myopia of 0.26 D in the pirenzepine group vs 0.53 Din the placebo group for treatment effect in favor of pirenzepine of 0.260D (P < .001). There was no evidencefor a difference among study sites (analysis of variance test for homogeneityacross treatment by investigative site, P = .52at baseline and P = .95 for change frombaseline). Age, sex, iris color, and refraction at entry were not predictiveof treatment difference. Additional methods were used to impute missing valuesdue to patients who discontinued the study. In all 3 methods used (last observationcarried forward, visit-to-visit extrapolation using median of respective treatment,and visit-to-visit extrapolation using median of placebo group), the treatmenteffect was similar to or greater than in the primary analysis method (0.262-0.343D). This is not surprising for the first 2 methods, because they correct forthe dropout of primarily pirenzepine-treated patients, who had, on average,lower rates of myopic progression.

The proportion of patients whose myopia progressed by at least 0.75D was 4% (4/99) in the pirenzepine group and 20% (11/56) in the placebo groupat 6 months (P = .004). At 1 year, theproportions were 11% (10/92) vs 31% (17/54), respectively (P = .006, Figure 3).

At study entry, there was no significant difference in mean axial lengthbetween groups (Table 3, 23.88 mm vs23.77 mm, P = .41). At 12 months, therewas a mean increase in axial length of 0.19 mm in the pirenzepine group and0.23 mm in the placebo group. This difference of approximately 0.04 mm, infavor of less myopic progression in the active treatment group, was not statisticallysignificant (P = .60).

Adverse events for both groups are provided in Table 4. In general, they were mild or moderate in severity. The3 most frequent systemic events were headache, common cold, and flu syndrome.The 3 nonophthalmic adverse events that differed statistically at the conservative P = .15 cutoff level were common cold, rhinitis,and sinusitis, which were actually more frequent in the placebo treatmentgroup. The 3 most frequent ocular events were mydriasis, erythema of the eyelids,and ocular itching. Six ophthalmic adverse events differed at the criterion P value: (1) symptoms of decreased accommodation, (2) increasein papillae and follicles, (3) decreased visual acuity (primarily near), (4)mydriasis and (5) eye discomfort, which were more common in the pirenzepinegroup, and (6) medication residue on eyelids or eyelashes, which was morecommon in the placebo group. Most reports of blurred or decreased vision (primarilynear) and other symptoms of accommodative abnormalities began during the firstmonth of the study. There was 1 serious adverse event reported during thestudy; a patient in the pirenzepine group was thrown from a horse and hospitalizedfor surgical repair of a fractured right arm. During the study, no differencesof note occurred between groups with respect to height, weight, heart rate,or blood pressure.

At study entry, the mean pupil diameter in room light was 5.0 mm inboth treatment groups (P = .62). Sixtyminutes after the first instillation, the mean mydriasis in pupil diameterwas 1.5 mm in the pirenzepine group and 0.2 mm in the placebo group (P < .001). A similar mydriatic effect wasseen 60 minutes after instillation at month 1. Twelve hours after the lastdose, the mean mydriasis in pupil diameter was approximately 0.5 mm in thepirenzepine group and ±0.2 mm in the placebo group (P ≤ .005, Figure 4).For the most part, there were few reports of abnormal biomicroscopic resultsthat were not present at baseline. Mild conjunctival erythema was noted in4 pirenzepine patients and 1 patient randomized to placebo at 1 or more visits.Conjunctival edema was noted in 2 pirenzepine patients at 6 months, and theseverity was mild and moderate; mild edema was noted in 1 pirenzepine patientat 9 months. No edema was reported in placebo patients.

Medication residue was similar in both treatment groups and was describedas mild to moderate for both groups (45 [38%] of 117 pirenzepine patientsand 30 [53%] of 57 placebo patients). Only one instance of severe conjunctivalerythema was reported during the study; the observation occurred in a pirenzepinepatient at 3 months. No abnormalities of the lens or posterior segment werenoted at any time in any patient. Of the patients with objective conjunctivalpapillae and follicles, few reports of ocular symptoms of discomfort occurred(11 [9%] of 117 pirenzepine patients and 2 [4%] of 57 placebo patients). Ofthe patients found to have increased conjunctival follicles and papillae,9 (19%) of 47 in the pirenzepine group and 2 (20%) of 10 in the placebo groupexperienced an increase of 2 grades or more. From a mean baseline intraocularpressure of 15 mm Hg, there was a mean decrease of up to 1 to 2 mm Hg in eacheye for both treatment groups, with little apparent effect of treatment duration(1 day vs 1 year) or interval from instillation (0 or 60 minutes).

At study entry, mean best-corrected distance visual acuity was approximately0.01 logMAR (equivalent to approximately 20/20 Snellen) in both treatmentgroups. During the study, mean changes in distance visual acuity were muchless than 1 line ETDRS, with mean acuity at 12 months similar to entry −0.03 ± 0.089(pirenzepine) and −0.01 ± 0.118 (placebo). Near visualacuity was approximately 0.06 at entry in both groups, and mean changes wereless than 1 line ETDRS. At each visit, the proportion of patients who reportedloss of 3 or more lines in distance or near vision was 0% to 3% (all in thepirenzepine group).

In this large, placebo-controlled study in children 8 to 12 years ofage, pirenzepine was more effective than the placebo gel in retarding boththe mean increase in SEQ and proportion of patients who showed myopic progression.These clinical data are consistent with the efficacy observed in animal models.^16- 18 No serious adverseevents related to drug use occurred, and mydriasis and cycloplegia, as mightbe expected from a nonselective muscarinic antagonist, prompted withdrawalin only 5 (4%) of the 117 pirenzepine-treated patients. Conjunctival allergicreactions were more common in the pirenzepine group but again prompted withdrawalin only 7 (6%) of these 117 patients. Medication residue was frequently observedby the investigator and also reported by patients in the preceding study.²⁰ Conjunctival follicles, observed in both study arms,occur with high frequency in this age group²³;it is possible that in many cases both the appearance of follicles and changesin grade are related to environmental and individual immune factors ratherthan to either the study drug or the vehicle. Mydriasis, although expectedand consistent with the pharmacology of pirenzepine, was relatively smallin magnitude (average dilation of 1.3 mm at 15 minutes after dosing and 0.5mm at 12 hours after dosing). As demonstrated in a previous study,²⁰ pupils in children treated with pirenzepine remainreactive to light, unlike what typically occurs with atropine. Therefore,patients in this study were not required to use photochromic spectacles andconcern for lens and/or retinal phototoxicity was minimized.

Potential limitations of the study include lack of formal accommodationtesting and phoria quantification in this protocol. However, accommodationwas measured in the preceding study,²⁰ andno individual with any manifest strabismus was permitted to enroll. Therewas a slight difference in baseline myopia between the pirenzepine and placebogroups. However, this difference was not statistically significant, but myopicprogression between the 2 groups was. Additionally, the entry level of myopiawas not predictive of degree of progression to a statistically significantlevel. Also, there was no difference in family history of myopia between the2 groups. Thus, despite these limitations, we believe that the results ofour study are applicable to the general population.

Although there is little doubt that preventing progression of myopiabeyond 5 or 6 D decreases the risk of retinal detachment, peripheral retinaldegenerations, and glaucoma, preventing even lesser degrees of myopia is alsoof great clinical significance. From a public health standpoint, the incidenceof these myopic complications certainly increases with refractive errors between−1.00 and −5.00 D, although not as significantly as at higherrefractive errors (eg, a patient with −2.00-D myopia has less risk ofdeveloping retinal detachment than a patient with −4.00-D myopia).^24- 25 Additionally, quality-of-life issuesbecome important with increasing dependence on constant refractive correctionwith higher degrees of myopia. A child with a refractive error of −1.00D may do fine scholastically and athletically with only wearing glasses parttime, but a patient with −2.00-D myopia needs correction all the time.

The magnitude of efficacy seen in the present study, approximately 50%reduction in progression or 0.3 D during 12 months in this US population,is of similar magnitude to that reported by atropine in an Asian population⁵ and greater than that reported by the use of progressiveaddition lenses.²⁶ The M₃ muscarinicantagonistic ocular effects of 0.1% to 0.5% atropine (eg, mydriasis and lossof accommodation) would be expected to be greater than that seen with 2% pirenzepinein the present study. Also, patients in the present study were not allowedto wear bifocals. The apparent ability of long-term pirenzepine treatmentto retard the development of myopia without predominant M₃ muscarinicantagonism is contrary to an accommodative basis for the development of pediatricmyopia. Rather, like studies with dopaminergic agonists,^27- 29 theefficacy of muscarinic agents in chicks in which the intraocular muscles arenicotinic in nature, and the Correction of Myopia Evaluation Trial (COMET)using progressive addition lenses,²⁶ this implicatesa muscarinic, cholinergic pathway that does not involve M₃ receptoractivation and is supportive of a neural mechanism that may be based in theretina. A scleral-based mechanism, however, is also conceivable; cholinergicamacrine cells may not be required for the progression of form-deprivationmyopia in chicks,⁹ and muscarinic antagonists(in somewhat high concentrations) inhibit chick scleral chondrocytes.³⁰ Alternatively, both retinal and scleral sites mayplay a role. Regardless of the site(s) of action, the results of this studyand prior work with pirenzepine argue against an accommodative mechanism inthe development of pediatric myopia and also serve to validate the use ofform-deprivation myopia to study this entity.

In conclusion, results of this study serve to establish the safety andefficacy of administration of pirenzepine ophthalmic gel in myopic childrenin slowing the progression of myopia during a 1-year treatment period. Weare continuing to treat and evaluate these patients, because 84 have electedto continue with a second year of controlled evaluation.

Correspondence: R. Michael Siatkowski, MD,Dean A. McGee Eye Institute, 608 Stanton L. Young Blvd, Oklahoma City, OK73104 (Rmichael-Siatkowski@ouhsc.edu).

Submitted for Publication: June 3, 2003; finalrevision received February 12, 2004; accepted May 14, 2004.

Financial Disclosure: This study was supportedby Valley Forge Pharmaceuticals Inc, the developers of pirenzepine ophthalmicgel, and Novartis Ophthalmics AG, its licensee. The study organizers wereemployees of Valley Forge Pharmaceuticals Inc at the time this study was conducted.Drs Crockett and Novack are consultants to Valley Forge Pharmaceuticals Inc.The clinical investigators were contractors to Valley Forge PharmaceuticalsInc for this clinical study. Drs Siatkowski and Miller have received honorariafrom Valley Forge Pharmaceuticals Inc; neither are stockholders nor have anyother investigators received consulting fees from Valley Forge PharmaceuticalsInc.

Previous Presentation: This study was presentedin part as a poster at the Association of Research in Vision and Ophthalmologymeeting; May 8, 2003; Fort Lauderdale, Fla.

Study Investigators: Geoffrey E. Bradford,MD (principal investigator), Terry Schwartz, MD (investigator), Kenneth Gainer,MD (investigator), and Paula Fenstermacher (coordinator), West Virginia University(Morgantown); Don L. Bremer, MD (principal investigator), Gary L. Rogers,MD (investigator), Mary Lou McGregor, MD (investigator), Jennifer Fogt, OD(investigator), and Rae Fellows (investigator), Columbus Children’sHospital (Columbus, Ohio); Gerhard W. Cibis, MD (principal investigator),Denise Hug, MD (investigator), Timothy Hug, OD (investigator), Debra Kirk,OD (investigator), Marcia Bray, OD (investigator), Renae Borden, OD (investigator),Dirck DeKeyser, OD (investigator), and Dana Kinney (coordinator) (OverlandPark, Kans); Susan Cotter, OD (principal investigator), Soonsi Kwon, OD (investigator),Raymond H. Chu, OD (investigator), Richard Yardley, OD (investigator), SusanShin, OD (investigator), Jennifer Slutsky, OD (investigator), John Maher,MD (investigator), Sherene Fort, OD (investigator), and Yvonne Flores (coordinator),Southern California College of Optometry (Fullerton); Arlene V. Drack, MD(principal investigator), Scott Lambert, MD (investigator), and Cameile MooreMartin (coordinator), Emory University School of Medicine (Atlanta, Ga); RichardM. Evans, MD (principal investigator), Timothy Dunham, MD (investigator),Lynnell coordinator. Lowry, MD (investigator), and Naomi Quesada (coordinator),Med Center Ophthalmology Associates; Marjean A. Taylor Kulp, OD (principalinvestigator), Michael J. Earley, OD, PhD (investigator), Stephan Katz, MD(investigator), and Tracy Kitts (coordinator), Ohio State University (Columbus);Robert J. Noecker, MD (principal investigator), Joseph M. Miller, MD, MPH(investigator), Rand Siekert OD (investigator), and Toby Aparisi (coordinator),University of Arizona (Tucson); Thomas R. Walters, MD (principal investigator),Douglas Lewis, MD (investigator), James E. Montgomery, MD (investigator),and Cindy Jasek (coordinator), Texan Eye Care (Austin, Tex); Colin A. Scher,MD (principal investigator), Yvette Jockin, MD (investigator), Henry O’Halloran,MD (investigator), and Lynn Nelles (coordinator), Children’s Hospitaland Health Center, San Diego (San Diego, Calif); R. Michael Siatkowski, MD(principal investigator), Dana M. Jones, OD (investigator), and Lisa Ogilbee(coordinator), Dean A. McGee Eye Institute/University of Oklahoma Departmentof Ophthalmology (Oklahoma City); P. Sarita Soni, OD (principal investigator),William Rainey, OD (investigator), James Laughlin, MD (investigator), TracyNguyen, OD (investigator), Brent Shelly, OD (investigator), and Donna Carter(coordinator), Indiana University School of Optometry (Bloomington); JohnH. Sullivan, MD (principal investigator), Douglas R. Fredrick, MD (investigator),Shaku Nagpal, MD (investigator), and Pat Pascoe (coordinator), Eye MedicalClinic of Santa Clara Valley (San Jose, Calif).

Study Organization: Marcia Edmondson, BS, AnneButeyn, BS.

Biostatistics: R. Stephens Crockett, PhD (D.A.T.A.,Inc, Mobile, Ala); the writing committee consisted of Susan Cotter, OD (SouthernCalifornia College of Optometry, Fullerton), Joseph M. Miller, MD, MPH (Universityof Arizona, Tucson), Colin A. Scher, MD (Children’s Hospital and HealthCenter, San Diego, Calif), R. Michael Siatkowski, MD (Dean A. McGee Eye Institute/Universityof Oklahoma, Oklahoma City), R. Stephens Crockett, PhD (D.A.T.A., Inc), andGary D. Novack, PhD (PharmaLogic Development Inc, San Rafael, Calif).