Additivity of Bimatoprost or Travoprost to Latanoprost in GlaucomatousMonkey Eyes

Objective  To compare the ocular hypotensive effect of the commercially availablepreparations of bimatoprost or travoprost added to latanoprost in monkey eyeswith laser-induced unilateral glaucoma.

Methods  Four monkeys with unilateral laser-induced glaucoma were used in eachtreatment group and received drops in the glaucomatous eye only. Intraocularpressure (IOP) was measured hourly for 6 hours, beginning at 9:30 AM on day 1 (untreated baseline), days 6 and 7 (single-agent therapy),and days 13 and 14 (2-drug combination therapy). On days 2 through 7, 1 dropof the scheduled single agent was given immediately after the 9:30 AM IOP measurement, and on days 8 through 14, the second scheduleddrug was given 5 minutes after the first. The following 5 different dosingprotocols were studied: latanoprost with bimatoprost added, bimatoprost withlatanoprost added, latanoprost with travoprost added, travoprost with latanoprostadded, and latanoprost with a second dose of latanoprost added.

Results  There were no statistically significant (P =.95) differences among the mean baseline IOPs in any of the 5 treatment groups.When applied as single agents, latanoprost, bimatoprost, and travoprost allproduced significant (P<.05) and equivalent (P = .98) reductions in IOP. The mean ±SEM maximumreduction (P<.05) from baseline IOP was 7.0 ±0.4 mm Hg (20% reduction) with travoprost alone, 6.5 ± 1.6 mm Hg (18%)with bimatoprost alone, and 7.5 ± 1.0 mm Hg (22%) with latanoprostalone. The mean ±SEM maximum additive reductions in IOP were 3.0 ±0.6 mm Hg (P<.05) for travoprost added to latanoprost;2.0 ± 0.4 mm Hg (P<.05) for latanoprostadded to travoprost; 4.8 ± 1.3 mm Hg (P<.05)for bimatoprost added to latanoprost; 4.3 ± 0.6 mm Hg (P<.05) for latanoprost added to bimatoprost; and 0.3 ± 0.5mm Hg (P>.60) for latanoprost added to itself. Thecombination of bimatoprost and latanoprost produced a greater (P<.05) lowering of IOP at trough and peak than the combination oftravoprost and latanoprost.

Conclusions  Latanoprost, bimatoprost, and travoprost used as monotherapy producedsignificant and equivalent reductions in IOP in glaucomatous monkey eyes.The IOP effects of the commercial concentrations of bimatoprost or travoprostwere additive to that of latanoprost, with bimatoprost showing a greater additiveresponse than travoprost.

Clinical Relevance  Because treatment with multiple medications is common among patientswith glaucoma, determining which glaucoma medications produce an additiveocular hypotensive response when used in combination has practical implicationsfor clinicians.

The prostaglandin analogues 0.005% latanoprost, 0.03% bimatoprost, and0.004% travoprost are all potent ocular hypotensive agents. Recent studieshave compared the reduction in intraocular pressure (IOP) achieved by thecommercial preparations of these 3 compounds when used as monotherapy in patientswith ocular hypertension or glaucoma,^1-5 butno published studies to date have examined their potential additive effectson IOP when used in combination. Because treatment with multiple medicationsis common among patients with glaucoma, knowing whether these drugs have anadditive effect on IOP has practical implications for clinicians. Studyingthese drugs in combination may also help elucidate their respective mechanismsof action and receptor profiles, a subject of interest in the literature atpresent.

This study was designed to determine whether an additional loweringof IOP could be achieved by using bimatoprost or travoprost together withlatanoprost in monkeys with unilateral laser-induced glaucoma.

A total of 9 adult female cynomolgus monkeys, each weighing 3 to 5 kgand in which glaucoma had been unilaterally induced by repeated argon or diodelaser photocoagulation of the midtrabecular meshwork, were used in these studies.Eight monkeys were used in the additivity studies of bimatoprost and latanoprost.After a washout of at least 2 weeks, 8 monkeys were used for the additivitystudies of travoprost and latanoprost. A subset group of 4 monkeys was usedfor the control study using latanoprost alone.

In all of the treatment groups, the IOP was measured hourly for 6 hours,beginning at 9:30 AM on day 1 (untreated baseline), on days6 and 7 (single-agent therapy), and days 13 and 14 (2-drug combination therapy).A calibrated pneumatonometer (Model 30 Classic; Mentor, Norwell, Mass) wasused for all measurements. Five minutes before each tonometry measurement,ketamine hydrochloride (1-5 mg/kg) was administered intramuscularly for sedation,and 1 drop of 0.5% proparacaine hydrochloride was topically applied to thestudy eye. On treatment days, the first IOP measurement was taken immediatelybefore the 9:30 AM dosing.

The commercially available preparation of each drug was used in thisstudy: 0.005% latanoprost (Xalatan; Pfizer, Inc, New York, NY), 0.03% bimatoprost(Lumigan; Allergan, Inc, Irvine, Calif), or 0.004% travoprost (Travatan; AlconLaboratories, Inc, Ft Worth, Tex). One drop of the scheduled medication fromthe commercial bottles was applied topically to the glaucomatous eye only.

Five different treatment groups, each consisting of 4 monkeys, underwenttesting using the following dosing schedule. On days 2 through 7, 1 drop ofthe scheduled single agent was applied to the glaucomatous eye only at 9:30 AM, immediately after the first IOP measurement. On days 8 through14, the second scheduled drug was added to the same eye 5 minutes after theapplication of the first drug. The following 5 different dosing protocolswere studied: latanoprost with bimatoprost added, bimatoprost with latanoprostadded, latanoprost with travoprost added, travoprost with latanoprost added,and latanoprost with a second dose of latanoprost added.

Two-tailed paired and unpaired t tests andanalysis of variance were used to analyze equivalency of baseline IOP in eachgroup; the change in IOP from baseline with latanoprost, bimatoprost, or travoprosttherapy alone; the additivity of bimatoprost with latanoprost, travoprostwith latanoprost, and latanoprost with latanoprost; and the comparative additivityof bimatoprost and latanoprost, travoprost and latanoprost, and latanoprostand latanoprost. A value of P<.05 was consideredstatistically significant. Unless otherwise indicated, data are expressedas mean ± SEM. All experimental studies complied with the Associationfor Research in Vision and Ophthalmology Resolution on the Use of Animalsin Research and were approved by the Mount Sinai School of Medicine, New York,NY, Institutional Animal Care and Utilization Committee.

The mean baseline IOPs of the 5 treatment groups were statistically(P = .95) similar (Table 1). When applied as single agents, latanoprost, bimatoprost,and travoprost all produced significant (P<.05)reductions in IOP (Figure 1). Themean maximum reduction (P<.05) from baseline IOPwas 7.0 ± 0.4 mm Hg (20% reduction) with travoprost alone, 6.5 ±1.6 mm Hg (18%) with bimatoprost alone, and 7.5 ± 1.0 mm Hg (22%) withlatanoprost alone (Table 2). Thedifferences in reduction of IOP from baseline comparing the 3 drugs were notstatistically significant (P = .98) (Figure 1). There was no statistically significant difference inthe mean values of the IOP among the 5 treatment groups at days 6 plus 7 attrough (P>.90) or at peak (P>.60).

The significant (P<.05) mean maximum additionalreductions in IOP were 3.0 ± 0.6 mm Hg (10% reduction) when travoprostwas added to latanoprost and 2.0 ± 0.4 mm Hg (7%) when latanoprostwas added to travoprost (Figure 2).There was no statistically significant difference (P>.90)in reduction of IOP when comparing travoprost added to latanoprost or latanoprostadded to travoprost (Table 2 and Figure 2). The significant (P<.05) maximum additional reductions in IOP were 4.8 ± 1.3mm Hg (15% reduction) when bimatoprost was added to latanoprost and 4.3 ±0.6 mm Hg (14%) when latanoprost was added to bimatoprost (Figure 3). There was no statistically significant difference (P>.90) in the reduction of IOP when comparing bimatoprostadded to latanoprost or latanoprost added to bimatoprost (Table 2 and Figure 3).The maximum additive reduction in IOP, 0.3 ± 0.5 mm Hg, was not significant(P>.60) when a second drop of latanoprost was addedto latanoprost (Table 2 and Figure 4). The additional reduction in IOPwhen the second drug was added in the bimatoprost and latanoprost combinationwas greater (P<.05) than that in the latanoprostand travoprost combination, no matter which drug was given first for eachdrug combination (Table 2).

The results of this study demonstrate that latanoprost, bimatoprost,and travoprost produce significant and equivalent reductions in IOP when usedas single agents at the indicated clinical dosage in glaucomatous monkey eyes.This finding appears consistent with those of some clinical studies in humansthat have compared the effectiveness of these 3 drugs.^1-4 A3-way comparative study by Parrish and colleagues¹ showsthat latanoprost, bimatoprost, and travoprost are equivalent in lowering IOPin patients with glaucoma or ocular hypertension. Three other recent studies^2-4 comparing bimatoprostor travoprost to latanoprost also demonstrate comparable efficacy when differencesin baseline IOP are taken into account.⁶ Incontrast, one study has found bimatoprost to be more effective than latanoprostin lowering IOP.⁵ However, as has been discussedelsewhere,¹ the reduction in IOP from baselinewith latanoprost is lower than expected, and the number of patients who werenonresponders to latanoprost is higher than expected in this one study comparedwith data from most other published studies. These differences may, in part,explain the significantly lower IOP achieved by bimatoprost relative to latanoprostin this single study.

Our results also show that, regardless of which drug is used duringthe first week of the protocol, significant additional lowering of IOP occurswhen bimatoprost or travoprost is used with latanoprost in the second weekof therapy, strongly suggesting true additivity. The fact that no additionalreduction of IOP takes place when we double the latanoprost dose in the secondweek of treatment implies that the single application of the drug is at orclose to a maximum effective dose. To date, no published studies have examinedwhether combinations of these 3 drugs are additive in humans, although additivitystudies with latanoprost and other prostaglandin derivatives have been performed.In 2 studies of unoprostone isopropyl ester and latanoprost in patients withglaucoma or ocular hypertension, no additional reduction of IOP was observedwhen unoprostone was added to latanoprost, although lowering was observedwhen latanoprost was added to unoprostone.^7,8 Thus,latanoprost has greater efficacy to lower IOP than unoprostone, but the effectsof the 2 drugs are not additive. In contrast, the IOP effects of latanoprostand 8-iso prostaglandin-E₂, an isoprostane, appear to be additivein glaucomatous monkey eyes, a finding possibly due in part to pharmacologicallydifferent mechanisms of action in lowering IOP related to the unique structureof 8-iso prostaglandin-E₂.⁹

Our finding that the ocular hypotensive effect of bimatoprost or travoprost,administered as a commercial preparation, is additive to that of latanoprostis somewhat unanticipated given that in most studies these 3 compounds havebeen shown to be agonists at the same prostaglandin-FP receptor.^10-15 Onestudy claims a different receptor profile and metabolism for bimatoprost.¹⁶ In addition, most studies report that all 3 drugslower IOP primarily by the same mechanism of increased uveoscleral outflow. ^17-21 Despitethese similarities, differences must exist in their mechanisms of action toaccount for the additivity that we have shown. Because we have demonstratedequivalent ocular hypotensive efficacy with monotherapy with all 3 commercialcompounds, differences in drug distribution within the glaucomatous monkeyeye, differences in prodrug metabolism, or a greater receptor affinity orintrinsic activity of bimatoprost or travoprost relative to latanoprost areunlikely to explain our additivity results. However, the drugs added to latanoprostin combination therapy may have an additional and distinct mechanism of action.

Thus, a possible explanation for the additivity of bimatoprost or travoprostto latanoprost is relative differences in their effects on trabecular anduveoscleral outflow. In studies performed in normotensive and ocular hypertensivehuman eyes, latanoprost was shown to lower IOP by increasing uveoscleral outflowwithout substantially altering aqueous flow or trabecular outflow facility.¹⁷ However, in another study, latanoprost substantiallyincreased tonographically measured outflow facility in humans, in additionto its effect on uveoscleral outflow.¹⁸ Theeffects of travoprost on outflow facility have not been reported in humans,but in a study in normotensive and glaucomatous monkey eyes, travoprost significantlyincreased uveoscleral outflow in the normotensive eyes.¹⁹ Anincrease in uveoscleral outflow in the ocular hypertensive eyes was also measuredbut did not reach statistical significance. No significant alterations werefound in aqueous flow or trabecular outflow facility in the normal or theglaucomatous eyes for travoprost. Similarly, in a study of aqueous humor dynamicsin normal human eyes treated with bimatoprost, Brubaker and colleagues²⁰ found a substantial increase in uveoscleral outflow,consistent with the effects on aqueous dynamics of latanoprost and travoprost.However, a statistically significant increase in tonographic facility of outflowcompared with baseline was also measured. Although that study was performedin normotensive eyes only, an increase in both trabecular and uveoscleraloutflow by bimatoprost may be consistent with our findings of additivity ofbimatoprost and latanoprost, if latanoprost has no effect on traditional trabecularfunction. However, the increase in both tonographic outflow facility and uveoscleraloutflow in another study of normotensive and ocular hypertensive human eyestreated with latanoprost belies this hypothesis.¹⁸ Inaddition, tonography appears to measure more than just trabecular resistance.These conflicting findings on latanoprost and aqueous dynamics by variousinvestigators need to be explained. Nevertheless, it is conceivable that relativedifferences among latanoprost, bimatoprost, and travoprost in their effectson trabecular and uveoscleral outflow may contribute to their combined ocularhypotensive effect when used together.

Although latanoprost, bimatoprost, and travoprost lower IOP primarilyby increasing uveoscleral outflow, the precise mechanisms by which this occursare not yet known. Various theories to explain the observed increase in uveoscleraloutflow have been investigated and include ciliary muscle relaxation, vasodilationof the ciliary body, and alterations in the extracellular matrix of the ciliarymuscle by several different mechanisms.²² Thus,prostaglandins may lower IOP primarily through an increase in uveoscleraloutflow, but it is possible that they accomplish this through different pathways.It is therefore conceivable that the prostaglandins we studied increase uveoscleraloutflow via parallel but distinct mechanisms and thereby demonstrate additivitywhen used together.

For these drugs to act by different or parallel mechanisms, there maybe differences in their receptor profiles. The respective free acids of all3 parent compounds have been shown to have potent agonist activity at theFP receptor, which has been identified in human cell cultures of trabecularmeshwork, ciliary epithelium, and ciliary muscle.^10-15 Bimatoprostitself is an FP receptor agonist.¹⁵ It is possiblethat differences in FP receptor subtypes may exist that would account fordistinct actions among these 3 prostaglandin analogues.²² Itis also possible that bimatoprost binds to an unidentified receptor,¹⁶ in addition to its known activity as an FP receptoragonist. A unique receptor profile for bimatoprost would be consistent withour finding that the combination of bimatoprost and latanoprost results ina greater lowering of IOP than that of travoprost and latanoprost.

It should be emphasized that our findings of additivity were demonstratedin nonhuman primates. There may be species differences in the onset, duration,and extent of the reduction in IOP produced by these 3 prostaglandins. Althoughit is unclear what the maximum effective doses of latanoprost, bimatoprost,or travoprost are in nonhuman primates, the lack of any additional reductionin IOP when we doubled the dose of latanoprost implies that the single applicationof the commercially available preparation of the drug is at or close to themaximum effective dose in glaucomatous monkeys.

The differences in the second baseline IOP, the results in days 6 plus7 after monotherapy, among the 5 groups could account for apparent differencesin additivity. Ocular hypotensive drugs are well known to be more effectivein eyes with higher baseline IOPs. The variability in our untreated baselineIOPs and in our treated baselines at days 6 plus 7 after monotherapy (Table 1) may be due in part to the smallnumber of monkeys in each treatment group, as well as to the wide fluctuationsthat are known to occur in the laser-induced glaucomatous monkey eye. However,the untreated mean baseline IOPs in the 5 treatment groups are statisticallysimilar, as are the mean IOPs after monotherapy.

The mechanisms responsible for the apparent pharmacologic additivityof bimatoprost or travoprost to latanoprost that we have shown in this studyare not known. Further investigations into the methods by which these drugslower IOP, their effects on aqueous humor dynamics, and differences in theirreceptor profiles may be helpful in elucidating our results.

Latanoprost, bimatoprost, and travoprost demonstrated equivalent effectivenessin lowering IOP when used as single agents in glaucomatous monkey eyes, consistentwith the results of the only reported 3-way clinical trial in humans.¹ In monkeys, bimatoprost and travoprost showed an additionalocular hypotensive response when used with latanoprost, regardless of whichdrug was used first, suggesting pharmacologic additivity. These results suggestthat bimatoprost, travoprost, and latanoprost may have relatively differentreceptor profiles and mechanisms of IOP reduction, which require further study.These results also suggest that combination therapy with these drugs may proveto be beneficial in some patients with glaucoma and that controlled additivitystudies in patients with ocular hypertension or glaucoma may be worth pursuingfor these clinically available prostaglandins.

Correspondence: Steven M. Podos, MD, Box 1183, Mount Sinai Schoolof Medicine, One Gustave L. Levy Place, New York, NY 10029.

Submitted for publication August 12, 2003; final revision received January22, 2004; accepted January 29, 2004.

This study was supported in part by grant EY01867 from the NationalInstitutes of Health, Bethesda, Md, and an unrestricted grant from Researchto Prevent Blindness, Inc, New York, NY.