Mean intraocular pressure (IOP) in29 patients with ocular hypertension or glaucoma treated with 0.15% unoprostoneisopropyl in one eye and placebo in the contralateral eye, twice daily for28 days. Bars represent standard error of the mean.
Fluorophotometric outflow facilityin 29 patients with ocular hypertension or glaucoma treated unilaterally with0.15% unoprostone isopropyl in one eye and placebo in the contralateral eye,twice daily for 28 days. Bars represent Standard error of the mean.
Toris CB, Zhan G, Camras CB. Increase in Outflow Facility With Unoprostone Treatment in Ocular HypertensivePatients. Arch Ophthalmol. 2004;122(12):1782–1787. doi:10.1001/archopht.122.12.1782
To determine the mechanism by which 0.15% unoprostone isopropyl reducesintraocular pressure (IOP) by studying 33 patients with ocular hypertensionor primary open-angle glaucoma.
At baseline, IOP was determined by pneumatonometry, aqueous flow andoutflow facility by fluorophotometry, episcleral venous pressure by venomanometry,and uveoscleral outflow by mathematical calculation. Unoprostone was administeredto one eye and placebo to the fellow eye of each patient twice daily in arandomized masked fashion. In patients who demonstrated an IOP reduction of3 mm Hg or more in either eye on day 5 ± 1 (n = 29),determinations were repeated on that day and on day 28 ± 2.Treated eyes were compared with control eyes, and treatment days were comparedwith baseline by paired t tests.
Compared with baseline, unoprostone significantly (P<.001) reduced IOP by a mean ± SEM of 5.6 ± 0.4mm Hg and 4.8 ± 0.6 mm Hg on days 5 and 28, respectively.The change from baseline with unoprostone was significantly (P<.001) greater than with placebo by 2.8 ± 0.4mm Hg on day 5 and by 3.2 ± 0.5 mm Hg on day 28. Comparedwith baseline, unoprostone significantly (P≤.001)increased outflow facility by 0.05 ± 0.01 and 0.08 ± 0.02μL·min−1·mm Hg−1 on days5 and 28, respectively. The baseline-adjusted between-treatment differenceswere significant (P≤.04) on day 28 (0.06 ± 0.02μL·min−1·mm Hg−1). Othermeasures were not different from placebo.
In responsive patients, unoprostone decreased IOP by increasing outflowfacility.
Intraocular pressure (IOP) is maintained by the production of aqueoushumor and its drainage through the anterior chamber angle. Current glaucomatherapies lower IOP by reducing aqueous humor production, increasing outflowthrough the uveoscleral pathway, or increasing the facility of trabecularoutflow. Some medications, such as brimonidine tartrate,1 havebeen shown to have multiple mechanisms of action. If target IOP is not reachedafter an appropriate period of monotherapy, combination treatments are usedto achieve the desired IOP-lowering effect, especially combinations of drugswith differing modes of action.2 An understandingof the IOP-lowering mechanism of action of each glaucoma medication wouldhelp predict additivity between drugs.
Unoprostone isopropyl is a structural analogue of prostaglandin (PG)F2α and has been reported to be a docosanoid. It has beenshown to be a safe and efficacious IOP-lowering drug.3- 5 Unoprostoneappears to lower IOP by increasing or facilitating outflow of aqueous humor.An increase in outflow facility6 and uveoscleraloutflow7 has been reported after topical administrationof unoprostone in rabbits. No effect on tonographic outflow facility was foundin healthy humans8 or in patients with glaucoma,9 suggesting that a uveoscleral outflow effect accountedfor the IOP decrease. Recently, Thieme and coworkers10 suggestedthat unoprostone lowers IOP by affecting aqueous outflow through the trabecularmeshwork via inhibition of endothelin-dependent mechanisms.
This study was conducted to determine the effects of unoprostone onaqueous humor dynamics in patients with ocular hypertension (OHT) or primaryopen-angle glaucoma (POAG).
This was a single-center, randomized, double-masked, placebo-controlledstudy in patients with OHT or POAG. The number of patients to enroll was determinedbefore the start of the study by power estimates generated with nQuery AdvisorVersion 2.0 (Statistical Solutions, Boston, Mass). A sample size of 30 subjectswas needed to provide at least 75% power to detect a difference in aqueousflow of 15% between drug-treated and vehicle-treated eyes, assuming a standarddeviation of 0.75 μL/min and a 2-sided significance level of. 05. Thestudy was approved by the University of Nebraska Medical Center InstitutionalReview Board, Omaha, and all patients provided written informed consent beforeinitiation of any study-related assessments.
Patients were scheduled for 4 visits, consisting of screening, baseline,day 5 ± 1 of treatment, and day 28 ± 2 oftreatment. The protocol is summarized in Table1.
At visit 1 (screening), a medical history was collected from each patientand a complete ophthalmic examination was performed. Main inclusion criteriaat visit 1 included diagnosis of bilateral POAG or OHT for at least 1 yearand corrected distance visual acuity of 20/200 or better (Early TreatmentDiabetic Retinopathy Study visual acuity chart). Main exclusion criteria atvisit 1 consisted of any visual field defect, known hypersensitivity to study-relatedmedication, previous glaucoma filtering procedure, cataract or laser surgerywithin the past year, ocular infection or inflammation within the past 3 months,or history of elevated IOP caused by processes other than POAG or OHT. Beforethe baseline visit (visit 2), subjects who had been taking medication to treatthe elevated IOP discontinued these drugs before the baseline visit. The medicationsincluded latanoprost (n = 7), timolol maleate (n = 4),betaxolol hydrochloride (n = 2), latanoprost plus timolol (n = 2),dorzolamide hydrochloride (n = 1), bimatoprost (n = 1),latanoprost plus dorzolamide (n = 1), latanoprost plus brimonidine(n = 1), and timolol plus travoprost (n = 1). The washoutperiod was 3 days for dorzolamide, 15 days for brimonidine, and 28 days forlatanoprost, bimatoprost, travoprost, and timolol.
Between 9 PM and 4 AM the night beforevisit 2, patients instilled 1 drop of 2% fluorescein at 5-minute intervalsuntil 6 to 10 drops were instilled in each eye.
At visit 2, central corneal thickness and anterior chamber depth weremeasured by slitlamp pachymetry. From these measurements, the anterior chambervolume was calculated for each eye.11 Fourpairs of duplicate fluorophotometric scans of the cornea and anterior chamberwere collected at 45-minute intervals between 8 AM and 11 AM, with the use of a scanning fluorophotometer (OcuMetrics, PaloAlto, Calif). These values were used to calculate baseline aqueous flow.12 An episcleral venomanometer (Eyetech, Morton Grove,Ill) was used to measure episcleral venous pressure.13
All IOP measurements were done with a pneumatonometer (Medtronic Xomed,Jacksonville, Fla). To be eligible for the study, IOP had to be between 23and 30 mm Hg in both eyes at 11 AM with no greater than a 5-mmHg difference between eyes on visit 2. Eligible patients then received 1 dropof 0.5% timolol maleate in each eye or acetazolamide, 250 mg orally. Thesedrugs reduce aqueous flow and IOP to enable calculation of fluorophotometricoutflow facility (Cfl), which provides an estimate of trabecular outflow facility.14
The following formula was used to calculate Cfl:
Cflx = [(Fa − Fax)/(IOP − IOPx)],
where Fa indicatesaqueous flow rate before treatment with acetazolamide or timolol; Fax, aqueous flow rate at intervals x = 1, 2,and 3 after acetazolamide-timolol; IOP, theIOP just before acetazolamide-timolol administration; IOPx, average of IOP values taken at the beginning and endof intervals x = 1, 2, and 3; and Cflx, Cfl at intervals x = 1, 2, and 3. The 3 calculated Cflx measurements were averaged to obtain thereported Cfl values.
Uveoscleral outflow (Fu) was calculated by means of the followingformula:
Fu = Fa − Cfl(IOP − Pev),
where Pev is episcleral venous pressure.
Safety assessments were performed and eligible patients received 2 bottlesidentical in appearance and labeled only by the patient’s identificationnumber and the words right eye or left eye. One bottle contained 0.15% unoprostone isopropyl (Rescula;Novartis Ophthalmics, East Hanover, NJ) and the other contained vehicle (placebo).Patients were instructed to instill 1 drop in the appropriate eye twice daily(8 AM and 8 PM) for 4 to 6 days in a double-masked,randomized fashion. Patients were asked to record on a log sheet the timeof each drug instillation and any omissions or errors in treatment. The nightbefore visit 3, fluorescein was administered as before.
At visit 3 (day 5 ± 1), only patients whose morningIOP was reduced by at least 3 mm Hg from baseline in at least one eye, anddid not exceed 30 mm Hg, continued in the study. If the IOP criteria werenot met, this visit served as an exit visit and further investigations werenot performed. The morning dose of unoprostone or placebo was administeredto the appropriate eye by the investigator in the clinic, and all measurementswere repeated as at visit 2. That evening, patients continued treatment with0.15% unoprostone twice daily in one eye and placebo twice daily in the felloweye until visit 4. The night before visit 4, fluorescein was administeredto each eye by the patient as before. At visit 4, the final dose of unoprostoneor placebo was administered by the investigator immediately after the firstIOP measurement. All measurements were repeated as at visit 2.
The efficacy variables were the change from baseline in aqueous humorflow, fluorophotometric outflow facility, uveoscleral outflow, episcleralvenous pressure, and IOP. These variables were assessed at visits 2, 3, and4. Baseline values for statistical purposes were the assessments taken onday 0 (visit 2). In addition, treated eyes were compared with contralateralcontrol eyes at each visit. The primary time point for evaluating the effectsof treatment on aqueous humor dynamics was day 28 (visit 4).
Both an intent-to-treat and per-protocol set of patients were analyzedfor efficacy. For the between-treatment comparison of each efficacy variable,the null hypothesis to be tested was that there was no difference betweenthe 2 treatment groups. The alternative hypothesis was that there was a differencebetween the groups.
Within each treatment group, tests were made to determine whether themean change from baseline differed significantly from zero. The change-from-baselinevalues were used to determine whether the 2 treatment groups were different.These analyses were performed by means of a paired t test.As there was one primary time point (day 28) compared with baseline, therewere no adjustments made for multiple comparisons. All statistical tests inthis trial were 2 sided, and all tests with a corresponding P≤.05 were considered statistically significant. All values arereported as mean ± SEM.
Changes from baseline in brachial artery blood pressure, radial pulse,and visual acuity were tested within each treatment group by a paired t test. Changes from baseline in the slitlamp examination,ocular symptoms, and ophthalmoscopy were evaluated.
Thirty-three patients were enrolled in the double-masked treatment periodof this study. Twenty-nine patients completed the study. Four patients werediscontinued on day 5 (visit 3) because of insufficient IOP reduction in eithereye.
The average age of the patients was 57.7 ± 2.0 years(range, 32-84 years). Thirteen (39%) of the patients were male and 25 (76%)were white. Fifteen patients (45%) had dark-colored (black or brown) iridesand 18 patients (55%) had light-colored (hazel, green, blue, or gray) irides.Mean IOP was 15.6 ± 0.4 mm Hg at screening, ranging from11 to 22 mm Hg before washout (intent-to-treat data set). All patients enrolledwere diagnosed as having OHT except for one patient who was diagnosed as havingPOAG in the right eye and OHT in the left eye.
In patients who completed the study, unoprostone significantly reducedIOP at days 5 and 28 compared with baseline and with placebo (Figure 1). Mean baseline IOP values were 25.5 ± 0.6mm Hg and 25.7 ± 0.7 mm Hg in the unoprostone-treated eyesand placebo-treated eyes, respectively. Average reduction from baseline ineyes treated with unoprostone was 5.6 ± 0.4 mm Hg (P<.001) and 4.8 ± 0.6 mm Hg (P<.001) on days 5 and 28, respectively, whereas theaverage reduction in the placebo-treated eyes was 2.5 ± 0.04mm Hg (P<.001) and 1.7 ± 0.1mm Hg (P = .008; Figure 1), respectively. The baseline-adjusted between-treatmentdifferences were statistically significant on day 5 (2.8 ± 0.4mm Hg; P<.001) and on day 28 (3.2 ± 0.5mm Hg; P<.001).
Compared with baseline values, both unoprostone and placebo significantlydecreased aqueous humor flow on day 5 of treatment but not on day 28 (Table 2). The baseline-adjusted between-treatmentdifferences were not statistically significant at either day 5 (0.14 ± 0.11μL/min; P = .21) or day 28 (0.18 ± 0.14μL/min; P = .22).
The average changes in outflow facility from baseline values in eyestreated with unoprostone were statistically significant (P≤.001) on days 5 and 28, whereas the average changes in the placebo-treatedeyes were not significant (Table 2, Figure 2). The baseline-adjusted between-treatmentdifferences were statistically significant on day 28 (0.06 ± 0.03μL·min−1·mm Hg−1; P = .04) but not day 5.
Unoprostone and placebo did not significantly alter episcleral venouspressure (Table 2). Both unoprostoneand placebo reduced uveoscleral outflow on day 28 compared with baseline (P≤.04; Table 2).However, the baseline-adjusted between-treatment differences were not statisticallysignificant at day 5 or 28.
There were no serious adverse events and no clinical concerns detectedduring the comprehensive ophthalmic examinations. The incidence of burning,stinging, or conjunctival hyperemia on drug instillation was more frequentwith unoprostone than with placebo treatment. Most reported adverse eventswere mild and ocular. No patients were discontinued from the trial becauseof adverse events.
In the patients who completed the current study, unoprostone reducedIOP in a clinically significant manner similar to previous studies.3,4 This reduction in IOP appeared to beprimarily the result of increased outflow facility. Compared with baseline,unoprostone increased outflow facility by 67% at 5 days and 100% at 28 days.The increase in outflow facility was confirmed by the between-treatment comparisons.
Data presented herein support the view of Yamamoto et al,15 whohad hypothesized that the IOP-lowering effect of unoprostone may be due tofactors other than increasing the rate of uveoscleral outflow. An effect ofunoprostone on the trabecular meshwork is one possibility. An increase inoutflow facility was found in rabbits treated with 1 drop of unoprostone.6,7 In vitro findings10 suggestedthat the IOP-lowering effect of unoprostone was likely due to its effect oncalcium-gated potassium channels, intracellular calcium, and some degree ofsmooth-muscle relaxation in the trabecular meshwork. Taken together with datafrom the current study, these findings suggest that one possible mechanismof the IOP-lowering effect of unoprostone is modulation of the trabecularmeshwork function.
Contrary to experiments in rabbits6,7 andthe current clinical study, previous investigations of 0.12% unoprostone inocular normotensive volunteers8 and patientswith POAG9 did not find an effect on outflowfacility when measured by tonography. These apparent discrepancies may bedue to differences in study design, method of measurement, and concentrationof drug. The current study enrolled only patients who responded to treatmentwith at least a 3-mm Hg decrease in IOP. Unoprostone appears to work wellin some individuals and not well in others16 (unpublishedpilot data, C.B.T., G.L.Z., and C.B.C., December 1998). Enrolling only respondersincreased the power of the study to find differences in the aqueous humordynamic parameters under investigation and to identify the mechanism by whichthe IOP was decreased. The fluorophotometric method to assess outflow facilitywas used, which detects differences not always found by tonography.17 The 0.15% unoprostone in the current study was expectedto provide a greater effect on IOP than the 0.12% used in some earlier studies.All of these differences in study design made it possible to detect an increasein outflow facility with unoprostone treatment.
The PGF2α analogues increase outflow predominantly,or at least partially, through the uveoscleral pathway.18,19 Latanoprostin normotensive and hypertensive human eyes primarily affects uveoscleraloutflow,18- 22 thoughit and other PGF2α analogues, including bimatoprost, havebeen found to increase outflow facility as well.18,23- 25 Theoutflow facility increase with bimatoprost was insufficient to account forthe entire IOP decrease, suggesting that uveoscleral outflow also was increased,24 similar to the effects of latanoprost.21,23 Travoprostincreases uveoscleral outflow in monkeys without affecting other parametersof aqueous humor dynamics,26 but its effectsin humans have not been reported.
A reduction of aqueous flow in unoprostone-treated eyes occurred after5 days of treatment, but a similar decrease was also noted in placebo-treatedeyes compared with baseline. The effect on aqueous flow disappeared after28 days of dosing. This suggests a possible short-term, contralateral effectof unoprostone on aqueous flow, a finding that requires confirmation. At nopoint during the study was the aqueous flow difference between treated andcontralateral control eyes statistically significant. Therefore, a unoprostone-inducedeffect on aqueous flow cannot account for the IOP reduction in the treatedcompared with contralateral control eyes.
Unlike most other PGF2α analogues, unoprostone didnot increase uveoscleral outflow in the present study. Instead, a reductionin uveoscleral outflow was observed compared with baseline measurements. Becausea reduction in uveoscleral outflow also was observed in the contralateralplacebo-treated eyes, with no difference between treated and the contralateralcontrol eyes, the effect was not considered to be clinically important. Thereduction of uveoscleral outflow is more likely the indirect effect of increasedoutflow facility than a direct effect on uveoscleral outflow. The balancein resistance factors between the trabecular meshwork and uveoscleral pathwaymay have shifted in favor of the trabecular meshwork, causing a redirectionof some fluid from the uveoscleral pathway into the trabecular meshwork. Inother words, the fluid took the path of least resistance.
The finding that unoprostone did not affect uveoscleral outflow is contraryto published studies in humans8 and rabbits.7 It should be noted, however, that the earlier clinicalstudy concluded an effect on uveoscleral outflow only when an effect on aqueousflow and outflow facility was not detected. Our pilot study (unpublished data),which included some nonresponders and patients with relatively low baselineIOPs, also failed to find a significant effect on aqueous flow and outflowfacility, and when calculated mathematically, uveoscleral outflow remainedunchanged as well. It is possible that the power of the earlier studies wasinsufficient to detect changes in outflow facility. The increase in uveoscleraloutflow with unoprostone treatment in rabbits not found in humans might beexplained by species differences in the structures of the anterior chamberangle and the unique sensitivity of the rabbit blood-aqueous barrier,27,28 especially to topical PGs.29 Breakdown of the blood-aqueous barrier alone canincrease uveoscleral outflow.30 Rabbit eyesdo not respond well to topical latanoprost,31 yetthis drug has become the gold standard for IOP reduction in humans. The rabbitis a poor model for the study of PGs and aqueous humor dynamics in humans.
It is always a concern in studies of this nature that patients may haveinstilled some of their drops in the incorrect eye at some time during thetreatment period. These errors might account for apparent contralateral effects.Each patient was informed repeatedly of the need for accurate adherence totheir regimen and the need to report any errors in drug administration. Allpatients filled out a daily log reporting the exact times of each drop applicationand any problems. Patients rarely reported omissions, delays in administrationof drops, or administration of the wrong drop to an eye. The drops were administeredby the investigator on each day of measurements to ensure that the treatmentwas correct while data were being collected. It is unlikely that sufficientnumbers of patients administered the drops erroneously to account for thecontralateral effects.
In clinical practice, IOP-lowering drugs often are combined to achievetarget IOPs. Drugs that increase facility of outflow might be used in combinationtherapy with IOP-lowering drugs that inhibit inflow. Unoprostone and timolol(an aqueous flow suppressant) have been found to be additive in several clinicaltrials.4,16,32 Onthe other hand, combination therapy of unoprostone with drugs that increaseaqueous humor outflow may or may not be effective. There is no apparent additivityof unoprostone with latanoprost.9,33- 35
In conclusion, unoprostone significantly reduced IOP in patients withocular hypertension by increasing the facility of outflow through the trabecularmeshwork. Unoprostone was safe and well tolerated, and may be a suitable adjunctdrug to aqueous flow suppressants.
Correspondence: Carol B. Toris, PhD, Departmentof Ophthalmology, 985840 Nebraska Medical Center, Omaha, NE 68198-5840 (email@example.com).
Submitted for Publication: July 14, 2003; finalrevision received April 1, 2004; accepted June 1, 2004.
Funding/Support: This study was supported inpart by Novartis Ophthalmics and by an unrestricted grant from Research toPrevent Blindness Inc, New York, NY. Dr Camras was a Research to Prevent BlindnessSenior Scientific Investigator.
Acknowledgment: Generous assistance in analyzingthe data and reviewing the manuscript was provided by Reza M. Haque, MD, PhD,Naveed Shams, MD, PhD, Christine Kubilus, RN, Kim Truett, MS, Jennifer Johnson,Barry Kapik, MS, Natalia Yannoulis, PhD, Lindsay Cook, MA, and Ken Green,PhD, of Novartis Ophthalmics.
Financial Disclosure: Dr Camras was a consultantfor the Pharmacia Corp.