Effects on Intraocular Pressure and Side Effects of 0.005% Latanoprost Applied Once Daily, Evening or Morning A Comparison with Timolol Albert Alm, MD,l ]ohan Stjernschantz, MD, 2 the Scandinavian Latanoprost Study Group Purpose: To compare the effect on intraocular pressure (lOP) and side effects of 0.005% latanoprost applied once daily, morning or evening, with 0.5% timolol applied twice daily. Methods: A 6-month randomized, double-masked, multicenter study with three parallel groups was undertaken. Two hundred sixty-seven patients were randomized, 84 to timolol, 89 to latanoprost in the morning for 3 months and then in the evening for another 3 months, and 94 to latanoprost with the treatment schedule reversed. Results: After 6 months, timolol reduced diurnal lOP from 24.6 to 17.9 mmHg (27%); latanoprost applied in the morning, from 25.5 to 17.7 mmHg (31%); and latanoprost applied in the evening, from 24.8 to 16.2 mmHg (35%). The efficacy of latanoprost applied in the evening was statistically superior to latanoprost applied in the morning and to timolol (P < 0.001). Latanoprost induced a slight increase in conjunctival hyperemia in 31.4% of treated patients, compared with 15.9% for timolol. Sporadic episodes of mild punctate corneal epithelial erosions were three times as frequent in latanoprosttreated eyes as in timolol-treated eyes. The most significant ocular side effect was increased pigmentation of the iris observed in five and suspected in seven more latanoprost-treated eyes. All these eyes had a mixed green-brown or blue/gray-brown iris color. Timolol reduced heart rate by 3 beats/minute (P < 0.005). Conclusions: The effect on diurnal lOP of latanoprost applied once daily in the evening is superior to that of timolol. The main difference in side effects is increased pigmentation of the iris induced by latanoprost, most likely due to stimulation of melanogenesis in iris stromal melanocytes. Ophthalmology 1995;102:1743-1752
Originally received: October 30, 1994. Revision accepted: August 9, 1995. I Department of Ophthalmology, University Hospital, Uppsala, Sweden. 2 Pharmacia Ophthalmics AB, Uppsala, Sweden. Presented in part at the International Society of Glaucoma Meeting, Quebec, June 1994, and the American Academy of Ophthalmology Annual Meeting, San Francisco, November 1994. Supported by Pharmacia AB, Uppsala, Sweden. Dr. Aim is a consultant to Pharmacia AB. Dr. Stjernschantz is employed by Pharmacia AB. Each author states that he has no proprietary interest in the development or marketing oflatanoprost.
Latanoprost (13,14-dihydro- 17-phenyl-l 8,19,20-trinorprostaglandin F;",-isopropyl-ester-previously PhXA41) is a new phenyl-substituted prostaglandin analogue.F Latanoprost or its epimeric mixture, PhXA34, has proven to be an effective ocular hypotensive drug in healthy eyes3- 5 and in eyes with elevated intraocular pressure (IOp).6-12 From these studies, it has been concluded that 0.005% latanoprost given once daily is the optimal dose for achievement of Reprint requests to Albert Aim, MD, Department of Ophthalmology, University Hospital, S-751 85 Uppsala, Sweden.
1743
Ophthalmology
Volume 102, Number 12, December 1995
maximal lOP reduction. It has also been established that the mechanism of action is on outflow, with little or no effect on aqueous flow. 3,4.11 The major effect seems to be increased uveoscleral outflow.11.13-15 Unlike its mother compound, PGF 2,,-isopropylester, latanoprost has been well tolerated , and ocular irritation and conjunctival hyperemia have not been clinical problems in previous studies. In addition, latanoprost and PhXA34 do not cause an initial increase in IOP,5.16 as has been observed with PGF2a-isoprop ylester.17 In several studies that have used laser flare meter or fluorophotometry, it has been shown that neither latanoprost nor PhXA34 has any significant effect on the integrity of the blood-aqueous barrier.3- 5.8 In toxicologic studies (unpublished observations), increased iris pigmentation has been observed in cynomolgus monkeys during long-term treatment with latanoprost, but no change 'in iris color was observed in Dutch belted rabbits, even with 12 months' treatment, or in humans with 3 months' treatment. 12 To obtain more information on the clinical usefulness of latanoprost as a glaucoma drug, we designed the current study as a multicenter, masked, long-term comparison oflatanoprost and timolol in patients with newly diagnosed open-angle glaucoma or ocular hypertension who had received little or no previous glaucoma treatment. Part of the study included a comparison of morning and evening application of latanoprost. Photographs of the iris were included to demonstrate any change in iris color.
Materials and Methods The study was designed as a randomized, three-group , double-masked comparison of latanoprost and timolol over 6 months in a multicenter study involving thirteen Scandinavian eye clinics. Approval from the appropriate regulatory authorities and ethics committees was obtained. The patients were informed about the increased iris pigmentation observed in cynomolgus monkeys, and a signed informed consent was obtained from all patients before study entrance. Two groups received 0.005% latanoprost once daily and one group received 0.5% timolol twice daily. The groups that received latanoprost daily initially received either morning or evening application and then switched after 3 months' treatment. All patients received two identical dropper bottles, labeled m orning or evening. For patients treated with timolol, both bottles contained timolol, whereas for those treated with latanoprost, one contained the vehicle. The patients were allocated to treatment groups according to a computer-generated scheme prepared by Pharmacia (Uppsala, Sweden). Patients were included after assessment for eligibility within 1 month before study start. Inclusion criteria were that the patients be of either sex older than 40 years of age with unilateral or bilateral primary open-angle glaucoma, capsular glaucoma, pigmentary glaucoma, or ocular hypertension (defined as an lOP of at least 22 mm Hg measured during the prestudy period). Patients requiring
1744
bilateral treatment had to fulfill eligibility criteria for both eyes. There was a washout period of previous treatment of 2 weeks for adrenergic agonists and 5 days for pilocarpine or carbonic anhydrase inhibitors. Patients who had been treated with topical beta-adrenergic blockers within the last 6 months or longer than 3 months at any time were excluded. Exclusion criteria included patients with a history of acute angle-closure glaucoma, severe trauma at any time , intraocular surgery or argon laser trabeculoplasty within the last 6 months, current use of contact lenses, history of severe dry eye syndrome, ocular inflammation within the last 3 months, any systemic disease that prevented the use of beta-adrenergic blockers and patients for whom the investigator judged that monotherapy would be insufficient regarding the optic nerve head and/or visual field. During the study, six visits were scheduled at baseline and after 2, 6, 12, 18, and 26 weeks' treatment. Examinations were made at 8 AM, noon , and 4 PM at baseline and after 12 and 26 weeks' treatment and only in the morning for the other visits. The prestudy examination included an ocular and medical history; gonioscopy unless documented in the case notes; at least two visual fields, one of them within the last 6 months and one during the I-month prestudy period; symptomatology; visual acuit y and refraction; visual fields determined with an automatic perimeter; slit-lamp examination; evaluation of conjunctival hyperemia; ophthalmoscopy; lOP determination; en face and iris photographs; blood pressure; heart rate and blood and urine samples. On baseline and on all following examinations, symptomatology, visual acuity, refraction (if visual acuit y had changed), slit-lamp examination, evaluation of conjunctival hyperemia, and lOP determinations were performed. Blood pressure and heart rate were determined at baseline and after 2, 12, and 26 weeks. Photographs ofthe iris were taken after 12, 18,and 26 weeks, and the final examination included ophthalmoscopy, a new visual field and new blood and urine samples. A Goldmann tonometer (Haag-Streit, Bern, Switzerland) was used to determine lOP . Three measurements were taken in each eye, and the mean of the three measurements was used in the statistical analysis. Conjunctival hyperemia was graded by the investigator by comparing observations with a set of standard photographs correspond ing to no, mild, moderate, or marked hyperemia. Visual fieldswere determined with the same automated perimetry threshold program . Humphrey 24:2 or 30:2 (Humphrey Instrument, Inc, San Leandro , CA), Octopus G I (lnterzeag, Schlieren, Switzerland), or Competer (Bara Elektronik AB, Lund, Sweden). Blood pressure and heart rate were measured while the patient was seated and after he or she had rested for at least 10 minutes. Photographs of the iris were classified according to color, and an y change in color was judged by an independent evaluator. Laboratory safety assessments included routine samples for hematology, blood chemistry (serum lipids, albumin, and glucose), kidney and liver function, electrolyte and fluid balance and urine analysis of glucose and protein.
Alm et al . lOP and Latanoprost versus Timolol Table 1. Characteristics of Patients Included in Latanoprost Group 1 (starting with latanoprost in the morning), Latanoprost Group 2 (starting in the evening), and Timolol Characteristics
Latanoprost Latanoprost Group 1 Group 2 Timolol Total
Men Women Age (yrs) Mean Range POAG Capsular Pigmentary OH Mixed· POAG
=
39 50
43 51
67 40-84 27 15
67 44-85 31 14 1 44 4
a
43 4
primary open-angle glaucoma; OH
34 50
=
116 151
66 67 42-84 40-85 33 91 43 14 a 1 36 123 1 9 ocular hypertension.
• Different diagnosis in the two eyes.
Results Two hundred sixty-seven patients were included in the study. Table 1 presents patient demographics. Two hundred sixty-six patients were white and 1 was ofIndian ancestry. One hundred seventy-one patients required bilateral treatment: 116 received latanoprost and 55 received timolol. Thirteen patients had previously received pilocarpine (7 were taking latanoprost, 6 were taking timolol); 5, epinephrine (3 latanoprost, 2 timolol); and 3, oral carbonic anhydrase inhibitors (21atanoprost, 1timolol). Sixteen patients in the latanoprost groups and 3 in the timolol group were given oral beta-adrenergic blockers throughout the study, and only 1 of the patients had received topical beta-adrenergic blockers before entering the study. Two hundred forty-eight patients remained at the conclusion of the study. The reasons for withdrawals are presented in Table 2. The most common reason for withdrawal was the information given to patients and investigators that one case of increased iris pigmentation had been observed, which prompted one of the centers to drop out of the study at that time. Intraocular Pressure
Statistical Methods The comparison of the two treatments was performed with use of a parallel design. Analysis of covariance (with treatment group and study center as factors and baseline lOP as covariate and their interactions) was performed. If both eyes were included, the average lOP of the two eyes was used. The diurnal lOP (defined as the average of lOPs determined at 8 AM, noon, and 4 PM) after 6 months' treatment was used to compare the efficacy of the two drugs. The test of equivalence between the two applications of latanoprost was performed with use of a crossover design. Analysis of variance (with treatment sequence, patient within sequence, period and treatment group as factors) was performed. The diurnal lOP at 3 and 6 months was used.
The diurnal lOPs in the three treatment groups at baseline and after 3 and 6 months' treatment are presented in Table 3, and the lOPs at 8 AM on each visit are shown in Figure 1. Both drugs reduced lOP significantly, with the least reduction for timolol at 6 months (27%) and the greatest reduction for latanoprost group 2 (evening application) at 3 months (36%). There was no difference in diurnal lOP between latanoprost applied in the morning and timolol, but latanoprost applied in the evening reduced diurnal lOP significantly better than either latanoprost applied in the morning or timolol. Thus, lOP reduction after 6 months of latanoprost application in group 1 was statistically significantly better than that of timolol at 6 months (P < 0.001), and the lOP reduction at 3 months for latanoprost group 2 was better than that of timolol at
Table 2. Reasons for Withdrawal in the Groups Treated with Latanoprost or Timolol Reasons
Latanoprost
Timolol
Repeated corneal erosions lOP not controlled Retinal arterial embolus Central retinal vein thrombosis Increased irispigmentation Information about irispigmentation Decreased visual acuity due to diabetes Burning sensation in the tongue Headache Spreading metastasis of cancer mammae
1 1 1 1 1 2 1 1
a
1
1
2 1
a a
1
a
lOP
=
a
a a
a 3
1
Total
1 1 5 1 1 1 1
intraocular pressure.
1745
Ophthalmology
Volume 102, Number 12, December 1995
Table 3. Diurnal Intraocular Pressure at Baseline and after 3 and 6 Months of Treatment Baseline
3 Mos
6Mos
Treatment
Mean ± SE (no. of patients)
Mean ± SE (no. of patients)
Mean ± SE (no. of patients)
Latanoprost I" Latanoprost 2t Timolol'[
24.8 ± 0.4 (89) 25.5 ± 0.3 (94) 24.6 ± 0.3 (84)
17.1 ± 0.3 (80) 16.4± 0.3 (85) 17.3 ± 0.3 (78)
16.2 ± 0.3 (84) 17.7 ± 0.3 (85) 17.9± 0.3 (79)
SE
=
standard error of the mean .
• Patients applied 0.005% latanoprost in the morning and placebo in the evening for 3 months and then latanoprost in the evening and placebo in the morning.
t t
Patients started with evening applications and then switched after 3 months. Patients applied 0.5% timolol twice daily throughout the study.
3 months (P < 0.01). At the end of the study, 58 of 84 patients (69%)treated with latanoprost in the evening had a diurnal lOP of 17 mm Hg or lower, compared with 27 of 79 patients (34%) receiving timolol (P < 0.001, chisquare analysis). Reduction ofIOP observed after 2 weeks' treatment was usually preserved throughout the study with little loss of effect. However, the increase in mean diurnal lOP observed in the timolol group between the 3- and 6month examinations, 0.6 mm Hg, was statistically significant (P < 0.001). This may be due to a tendency toward increased outflow resistance with time as well as longterm drift of the effect oftimolol. No similar comparison can be made for the latanoprost groups, because their application schedules were switched after 3 months' treatment, but there was no significant long-term drift of the
latanoprost effect during the two 3-month periods of treatment in the two groups. There was no significant effect on lOP in the fellow eye in patients with unilateral treatment with latanoprost. However, timolol caused a slight (1.1 ± 0.3 mm Hg) but statistically significant (P < 0.01) drop in lOP of the untreated fellow eyes. The effect of both drugs was similar regarding sex, different age groups, different diagnosis, or different iris color.
Ocular Side Effects Figure 2 shows the average maximal hyperemia observed in the three treatment groups, and Table 4 presents the observed increments in conjunctival hyperemia at the last
(mmHg)
30 Figure 1. The intraocular pressure (mean ± standard error of the mean) at 8 AM on baseline and after 2,6, 12, 18 and 26 weeks' treatment with latanoprost or timolol. Latanoprost group 1 (n = 89) started with 0.005% latanoprost in the morning for the first 12 weeks, latanoprost group 2 (n = 94) in the evening, and then the two groups switched application schedule. Throughout the study, 0.5% timolol (n = 84) was applied twice daily.
0--0 Latanoprost 1 Latanoprost 2 0---0 Timolol
• ---e 25
20
---=-
... -
15
10 -1--.....- - . . . . . , - - - - - . , . . . - - - - . . . , . - - - - - . _ _ -
o
2
6
12
18
Duration of treatment
1746
26
(Weeks)
Alm et al . lOP and Latanoprost versus Timolol visit compared with baseline. There was a slight increase in average maximal hyperemia in all three groups with treatment. The increase was slightly more pronounced in the two groups receiving latanoprost than in the group receiving timolol, but the average hyperemia was considered to be less than mild in all three groups at all times. Increased hyperemia, compared to baseline, was observed in 31.4% of patients treated with latanoprost and in 15.9% of patients treated with timolol. Conjunctival hyperemia was reported as an adverse event in two patients and reported in the case-report form in another four patients treated with latanoprost (Table 5), but there was no patient with conjunctival hyperemia above level 2 (moderate) at either the 3- or the 6-month visit. Table 5 summarizes the ocular side effects observed in the study. Surprisingly, ocular discomfort was as common among patients treated with timolol as it was among patients treated with latanoprost. Apart from hyperemia, there are two striking differences between the two drugs. Punctate epithelial keratopathy was observed in 12 patients treated with latanoprost compared with 2 patients treated with timolol. Three patients (2 receiving latanoprost and 1 receiving timolol) were reported as adverse reactions, and I of these patients taking latanoprost was withdrawn from the study when a second episode of punctate erosions was detected. In the remaining 11 patients, the punctate epithelial keratopathy was graded as mild by the investigator and was observed only on one or two examinations. Mild episodes of punctate epithelial erosions were also observed during the study in 3 of 96 untreated fellow eyes. The other difference in ocular side effects observed was an increased pigmentation of the iris, observed in 5 patients and suspected in 7 more patients, all of whom were
Table
4. Percent of Patients with Different Degree of Increase in Conjunctival Hyperemia at the Last Examination*
Level of Increase
Latanoprost
Timolol
No increase Barely detectable (0.5)
68.6 26.3 5.1
84.1 9.8
Mild (1.0)
6.1
• A total of 175 patients treated with 0.005% latanoprost and 82 patients treated with 0.5% timolol were evaluated. Hyperemia was graded from standard photographs corresponding to no (0), mild (1), moderate (2), or severe (3) hyperemia. Increments in hyperemia level from baseline are in steps of 0.5.
treated with latanoprost for at least 3 months (Fig 3). Inadequate quality of some iris photographs made the diagnosis difficult, but it seems reasonable to assume that the 7 suspected cases represent mild cases of increased pigmentation. All cases of increased iris pigmentation occurred in irides with a mixed green-brown or blue/graybrown iris (not including nevi or freckles). Thus, of the 12 cases, 9 were found in green-brown irides and 3 in blue/gray-brown irides. A breakdown of iris color of the entire patient population is presented in Table 6. No aqueous flare was observed in any patient, and there was no drug-related loss of visual acuity or change in refraction. Visual fields and optic discs were evaluated at baseline and at the end of the study. A significant deterioration of the visual field was observed in one eye only (receiving timolol), and no significant change in appearance of the optic discs, apart from disc hemorrhages (Table 5), was observed.
Arbitrary units
2.0 0--0 Latanoprost 1 .---e Latanoprost 2 D---D Timolol
1.5
Figure 2. Average maximal hyperemia (mean ± standard error of the mean) observed at baseline and after 3 and 6 months' treatment. Hyperemia was graded from standard photographs corresponding to no hyperemia (0), slight (1), moderate (2), or marked (3).
1.0
...........
-'-.-~~~
__._._._._--e----_.. _---------£)
--.....,-----.,....----...,.-----r--
O+--__
o
2
6
12
18
26
(Weeks)
Duration of treatment
1747
Ophthalmology
Volume 102, Number 12, December 1995
Table 5. Ocular Side Effects: Number of Patients with Reported Adverse Events (excluding serious adverse events) or Reported Symptoms or Signs among 183 Patients Treated with Latanoprost and 84 Patients Treated with Timolol Type Ocular discomfort Corneal punctate erosions Blurred vision/vision disturbance Hyperemia Conjunctivitis Eyelid pain and discomfort Hordeolum Disc hemorrhage Increased pigmentation Other']
Latanoprost (n = 183)
Timolol (n = 84)
40 (8)
21 (4)
12 (2) 6 (2) 6 (2) 2 (1) 2 (0) 2 (0) 3 (3)
2 (1) 6 (0)
7 (5)* 13 (2)
o
2 (2) 1 (0)
o
3 (3)
o
6 (3)
• Increased pigmentation was suspected in another five cases when the iris photographs were evaluated.
t For latanoprost, it includes one case each of pinguecula, hyposfagma, eyelid erythema, eyelid tic, repeat surgery for ectropion, dots on corneal endothelium, corneal epithelial haze, vitreous detachment, senile retinoschisis, retinal pigmentary degeneration, dilation of iridial vessels, retrobulbar pain, and darker eyelashes. For timolol, it includes one case each of hyposfagma, eyelid edema, eyelid contusion, corneal epithelial infiltrates, vitreous detachment, and vitreous hemorrhage. Values in parentheses indicate those reported as adverse events.
Systemic Side Effects Table 7 lists systemic adverse events reported during the study. Compared with patients treated with timolol, more cases of muscle and joint pain and of nonocular allergy and eczema were reported among patients treated with latanoprost. The cases with muscle or joint pain were mainly transient episodes of lumbago (6 cases) or pain
due to rheumatoid arthritis or arthrosis. Additionally, the eczematous or allergic skin reactions were, with one exception (latanoprost), transient episodes seen on one examination only. The serious adverse effects that occurred during the study are shown in Table 8. In total, 14 serious adverse effects were recorded among 13 patients, II in the pooled latanoprost groups and 2 in the timolol group. Apart from a statistically significant reduction of mean heart rate from 72.0 to 69.3 beats/minute at 6 months in timolol-treated patients (P < 0.005), there was no significant effect on either blood pressure or heart rate in any group. Blood and urine analysis before and after the study did not reveal any significant effect on the hematologic, urinary or clinical chemistry for either drug.
Discussion Efficiency In previous studies on PGF2,,-isopropylester, there was a transient increase in lOP 30 minutes after application. I? In the current study, the first lOP measurement was made 4 hours after application, and an early transient increase in lOP with latanoprost could have escaped detection. However, it is unlikely that such a biphasic lOP response took place. The lOP increase observed with PGF2,,-isopropylester coincided with a marked conjunctival hyperemia, and we assumed that the lOP increase was due to an intra-ocular vasodilation. The early conjunctival hyperemia observed with PGF 2,,-isopropylester has not been observed with latanoprost or its epimeric mixture, PhXA34, with which a slight hyperemia was seen 4 hours after application of a 10-JIg dose.' Also, in two previous studies, one on PhXA34 and one on latanoprost, lOP determinations 30 minutes after application did not reveal any early transient lOP elevation.P'!" Both drugs effectively reduced lOP. When applied in the evening, latanoprost reduced mean diurnal lOP
Figure 3. The left eye of one patient treated with latanoprost (left) before and (right) after 6 months' treatment. The iris was classified as blue-brown, with a brown zone surrounding the .pupil before treatment.
1748
Alm et al . lOP and Latanoprost versus Timolol Table 6. Number of Patients with Different Iris Colors According to the Classification Used in the Current Study" Latanoprost (n = 183)
Iris Color
65 58 22 I 2 26 1
Blue/gray Blue/gray with slightly brown Blue/gray-brown Green Green with slightly brown Green -brown Yellow-brown, whites Brown, whites Brown, non-whites
7 1
Timolol
Total
(n = 84)
(n = 267)
(%)
27
92
37
95
(34.5) (35.6)
12
34 1
o o
(12.7) (0.4) (0.7)
2
8
o o o
34 1
(12.7)
7
(2.6)
1
(0.4)
(0.4)
, The classification was prompted by the discovery of one case of increased iris pigmentation, and the groups blue/ gray with slightly brown and green with sligh tly brown were used for borderline cases with very little brown. Yellow-brown is used to classify eyes with a brown iris with a pale yellowish periph ery. In the current study increased pigmentation was seen or suspected only in blue /gray-brown or green-brown eyes.
significantly better than did timolol (P < 0 .0 I) , whereas the efficacy of morning application of latanoprost was very similar to that oftimolol. This difference in efficacy Table 7. Systemic Side Effects: Number of Patients with Reported Adverse Events (excluding serious adverse events) or Reported Symptoms or Signs among 183 Patients Treated with Latanoprost and 84 Patients Treated with Timolol Type Upper respiratory tract infection Muscle and joint pain Eczema/ allergy Urinary tract infection Headache Vertigo Paresthesia Arterial hypertension Diarrhea Fever Tr auma Postsurgical symptoms Other'
Latanoprost (n = 183) 19 (7) 12 (4) 7 (6) 4 (2) 4 (1) 4(0) 2 (0) . 3 (2) 1 (0) 2 (0) 2 (1) 2 (0)
17(7)
Timolol
(n = 84)
17 (6) 2 (0) 1 (1) 1 (1) 6 (4)
o
1 (0)
o
2 (1)
o
2 (2) 1 (0) 5 (2)
, For latanoprost, it includes one case each of angina pectoris, ulcus ventr iculi, stom ach pain , sneezing after eye drops, n octuria, fatigue , malaise, infected woun d, dry mouth, symptoms due to tum or investigation, uncomfortable feeling after eye drops, itching of lower extremities, sleeping probl ems, anx iety, nightmares, a transient increase in serum aspartate amino tra nsferase and alanine amino transferase, and a transient increase in serum bilirubin. For timolol , it includes one case each of thrombophleb itis, diverticulitis, subcutaneous abscess, tendinitis, and reduced libido. V alues in parentheses indic ate th ose reported as adverse events.
between the two dose regimens is probably due to the time course of the latanoprost response. The maximum effect is not reached until up to 12 hours after application, 16 and the lOP measurements were made 12, 16, and 20 hours after evening application but 4 , 8, and 24 hours after morning application. Comparing the effect on mean diurnal lOP is a reasonable evaluation of clinical efficacy. Both drugs have long duration, and there are no pronounced peak and trough effects. However, the 8 AM values can be regarded as trough effects for
Table
8.
Serious Adverse Events
Nature of Adverse Event
T reatment
Recovered
Double vision Back pain due to spinal meningioma Pancreatic adenocarcinoma Cancer mammae, metastasis spreading Esophageal tumor Deep venous thrombo sis, left leg Venous thrombosis, right leg Angina pectoris Vitreous hemorrhage due to diabetes Cardiac arrhythmia Central retinal arterial embolus Central retinal venous thrombos is Branch retinal vein occlusion Visual field deterioration
Latanoprost Latanoprost
Yes' No'
Latanoprost Latanoprost
No No
Latanoprost Latanoprost Latanoprost Latanoprost Latanoprost
No Yes Yes Yes Yes
Latanoprost Latanoprost Latanoprost
Yes Yes No
T imolol Timolol
No No
, Adverse events occurred in the same patient.
1749
Ophthalmology
Volume 102, Number 12, December 1995
latanoprost applied in the morning as well as for timolol. As shown in Figure 1, the trough effect is almost identical for the two drugs. The noon measurements for timolol and for latanoprost applied in the evening can be used to compare peak effects. Timolol reduced noon lOP from 24.7 ± 0.4 to 17.5 ± 0.4 mm Hg, compared with from 25.5 ± 0.3 to 16.2 ± 0.3 for latanoprost at 3 months. The corresponding figures at 6 months were from 24.7 ± 0.4 to 17.9 ± 0.4 for timolol and 24.7 ± 0.4 to 16.2 ± 0.3 mm Hg for latanoprost. Thus, the main difference in efficacy between the two drugs is a better peak effect for latanoprost. Latanoprost contains twice the amount of benzalkonium chloride as timolol. The use of a preserved vehicle to mask the study may have influenced the permeability of the cornea. However, an improved therapeutic effect due to increased penetration oflatanoprost seems unlikely. Latanoprost is an isopropylester and thus highly soluble in lipids, and in a previous study twice-daily applications of 0.006% latanoprost caused a loss rather than enhancement of the effect on IOPY There was no apparent difference in lOP response due to sex, age, type of glaucoma or color of the iris for either drug. Thus, it appears that patients respond to latanoprost and timolol uniformly, and the results indicate that latanoprost given once daily should be an effective lOP reducing agent in glaucoma therapy.
Ocular Side Effects Considering the nature of prostaglandins, previous studies have focused on three possible side effects oflatanoprost; ocular discomfort, conjunctival hyperemia and permeability of the blood-aqueous barrier. All of these possible side effects can be regarded as part of an inflammatory response. An inflammatory response is a complicated chain of events requiring many factors besides prostaglandins, such as other eicosanoids, autacoids, sensory nerves, and neuropeptides.Pr" There are considerable differences among species differences regarding the ocular effects of these agents, and the rabbit eye is particularly sensitive." However, even in rabbits, breakdown of the blood-aqueous barrier is seen mainly with prostaglandins with affinity to the EP2 receptor, whereas FP agonists do not disrupt the barrier.F Latanoprost is a selective FPagonist with little affinity for the EP2 receptor." Most natural prostaglandins, including PGF2a-isopropylester, have affinity for several prostanoid receptors. The higher degree of selectivity of latanoprost probably explains the lack or irritative response and the absence of the early, marked conjunctival hyperemia seen with PGF2a-isopropylester. Thus, there was no difference in ocular discomfort induced by the two drugs, no flare was seen in any patient, and the general examination of the eyes did not reveal any pathologic findings in the latanoprost-treated eyes compared with findings at baseline. In all treatment groups, some conjunctival hyperemia was registered. We must emphasize that the maximum hyperemia observed during the day is reported here. A comparison of the number of patients with an increase in conjunctival hy-
1750
peremia during treatment (Table 4) shows that latanoprost produced more conjunctival hyperemia than timolol, but the difference was not large. Still, in a few patients treated with latanoprost, a more pronounced hyperemia may occur; approximately 1% of the patients treated with latanoprost exhibited enough hyperemia to be reported as an adverse effect (Table 5). There was no difference in maximum conjunctival hyperemia during the day between evening and morning application of latanoprost. Thus, the possibility that a peak in hyperemia would occur during the night in patients applying latanoprost in the evening is remote. The frequency of punctate epithelial keratopathy was approximately three times higher in the latanoprost groups than in the timolol group. A higher concentration of benzalkonium chloride in the latanoprost eyedrops, 200 mg/ ml compared with 100 mg/ml in timolol eyedrops, may contribute to this difference. The patients in the latanoprost group were given twice as much benzalkonium chloride as were the patients in the timolol group, because the vehicle used as a placebo contained benzalkonium chloride as well. We concluded that factors other than the drug are involved in punctate epithelial erosions because most episodes were mild and isolated, and such episodes were also observed in about 3% of untreated fellow eyes. The frequency of punctate epithelial erosions did not differ between patients applying latanoprost in the morning and those applying it in the evening. Therefore, it is unlikely that the frequency of punctate epithelial erosions was underestimated in the group applying latanoprost in the evening. Similarly, there were no other obvious differences between morning and evening application oflatanoprost regarding ocular side effects, thereby suggesting that the preserved vehicle had no therapeutic effectthat might have concealed some ocular side effect. An increase in iris pigmentation has been observed in chronic toxicity studies of latanoprost eyedrops in cynomolgus monkeys. In the current masked 6-month study, in 6.6% ofthe patients treated with latanoprost, including suspected cases, iris pigmentation increased. No such case was noted in the timolol group. By classifying the iris colors, we believe it possible to predict if a change in iris color is likely to occur. The increased pigmentation was seen only in irides of a mixed green-brown or blue/graybrown color. No increase of pigmentation or change of color was seen in blue, gray or green irides or in irides that were homogenously brown from the pupil to the iris root. It is also noteworthy that nevi or freckles of the iris did not change shape or color during treatment. The first change of iris pigmentation was observed in some patients after 3 months' treatment, but in most cases it occurred later. Treatment was terminated in all patients exhibiting increased pigmentation of the iris. The mechanism behind the increased iris pigmentation is unknown, but preclinical studies have shown that a similar increased iris pigmentation is induced by naturally occurring prostaglandins and that the most likely mechanism is a stimulation of the melanocytes in the iris stroma to produce melanin (unpublished observation). The path leading to increased pigmentation in monkeys and in humans seems to be
AIm et al . lOP and Latanoprost versus Timolol similar; it takes approximately 3 to 6 months of continuous treatment with latanoprost at the clinical dose before a clear-cut change in iridial pigmentation is detected in monkeys. Slit-lamp examinations of the patients with increased iris pigmentation have not revealed any structural changes in the iris. In a follow-up recovery study, there have been no changes or signs ofreversal up to 18 months after termination of the treatment, which indicates that the change in iris color is permanent or very slowly reversible.
Systemic Side Effects Neither latanoprost nor timolol had any significant effect on systolic or diastolic blood pressure measured 12 to 24 hours after application. However, in the timolol group, there was a reduction in heart rate of about 3 beats/minute at the end of the treatment period. Because this reduction was statistically significant (P < 0.005) and about twice as large as the changes seen with latanoprost, this probably represents a true drug-induced effect, as has been shown previously for timolol eyedrops." Thus it appears that in contrast to timolol, latanoprost has no or only minor effects on the cardiovascular system. There were a large number of systemic adverse effects, symptoms and signs reported for both drugs (Tables 7 and 8). During a 6-month follow-up in this age group, one must expect a certain number of symptoms and signs to appear regardless of treatment. A conclusion concerning a relation between drug and systemic effect is unlikely to emerge from a single study, unless the systemic effect can be connected to a known pharmacologic effect of the drug. There is no such known effect of latanoprost suggesting a relation between the reported systemic effects and latanoprost. When compared with timolol, latanoprost application led to more patient reports of muscle and joint pain and of allergic or eczematous skin reactions. However, apart from two patients, all these reports concerned transient episodes of usually mild or moderate severity with complete recovery despite continued treatment. Of the two patients treated with latanoprost who were the exceptions, one had a swollen, painful arm and one had eczema, both lasting for the last 3 to 4 months of the study. Neither was considered drug related, and both patients completed the study. The more numerous serious adverse effects reported for latanoprost cannot be explained by known properties of the drug, but monitoring of more patients will be required before any definite conclusions about systemic side effects of latanoprost can be drawn. In conclusion, we have found that the efficacy of 0.005%latanoprost administered once daily in the evening is at least as good as and probably superior to 0.5% timolol administered twice daily. Latanoprost tended to cause slight conjunctival hyperemia and more punctate epithelial keratopathy. The latter effect may be due to the preservative benzalkonium chloride, which was overdosed because placebo drops were also used to mask the study. The most important side effect of latanoprost was increased pigmentation of the iris, which occurred in 6.6%
of the patients, all of them with irides of mixed greenbrown or mixed blue/gray-brown color. Although it seems possible to predict if a patient will develop this side effect, it is important that we clarify the underlying mechanism and the possible consequences of this increase in iris pigmentation.
Appendix Members of the Scandinavian Latanoprost Study Group. Sweden: Ingmar Widengard, MD, Department of Ophthalmology, Uppsala; Christina Linden, MD, Department of Ophthalmology, Umea University Hospital; Mats Soderstrom, MD, Department of Ophthalmology, Huddinge University Hospital; Sven Erik Nilsson, MD, and Bjorn Fristrom, MD, Department of Ophthalmology, University Hospital, Linkoping; Bertil Lindblom, MD, Department of Ophthalmology, Sahlgrenska Hospital, Gothenburg; Anders Heijl, MD, and Kjell-Gunnar Gundersen, MD, Department of Ophthalmology, Malmo General Hospital; Berndt Ehinger, MD, Catharina Holmin, MD, and Elisabeth Bengtsson-Stigmar, MD, Department of Ophthalmology, Lund University Hospital. Norway: Henry Aasved, MD, and Pal Jangard, MD, Department of Ophthalmology, Bergen; Amund Ringvold, MD, and Torgeir Vegge, MD, Department ofOphthalmology, Ulleval Hospital, Oslo; Reidar Halseide, MD, Department of Ophthalmology, Regional Hospital, Trondheim. Denmark: Henrik Lund-Andersen, MD, and Per Flesner, MD, Department of Ophthalmology, KAS Gentofte, Hellerup; John Thygesen, Department of Ophthalmology, Vejle Hospital. Finland: Juhani Airaksinen, MD, and Anja Tuulonen, MD, Department of Ophthalmology, Oulu University Hospital.
References I. Stjernschantz J, Resul B. Phenyl substituted prostaglandin
2. 3.
4.
5.
6.
analogs for glaucoma treatment. Drugs Future 1992;17:691704. Resul B, Stjernschantz J, No K, et al. Phenylsubstituted prostaglandins: potent and selective antiglaucoma agents. J Med Chern 1993;36:243-48. Aim A, Villumsen J. PhXA34, a new potent ocular hypotensive drug. A study on dose-response relationship and on aqueous humor dynamics in healthy volunteers. Arch Ophthalmol 1991;109:1564-68. Ziai N, Dolan JW, Kacere RD, Brubaker RF. The effects on aqueous dynamics of PhXA41, a new prostaglandin F2a analogue, after topical application in normal and ocular hypertensive human eyes. Arch Ophthalmol 1993;Ill: 1351-58. Hotehama Y, Mishima HK. Clinical efficacy of PhXA34 and PhXA41, two novel prostaglandin F2a-isopropyl ester analogues for glaucoma treatment. Jpn J Ophthalmol 1993;37:259-69. Aim A, Villumsen J, Tornquist P, et al. Intraocular pressurereducing effect of PhXA4l in patients with increased eye
1751
Ophthalmology
7.
8.
9.
10.
11.
12.
13.
14. 15.
Volume 102, Number 12, December 1995
pressure: a one-month study. Ophthalmology 1993;100: 1312-17. Fristrorn B, Nilsson SEG. Interaction of PhXA41, a new prostaglandin analogue, with pilocarpine. A study on patients with elevated intraocular pressure. Arch Ophthalmol 1993;111:662-65. Hotehama Y, Mishima HK, Kitazawa Y, Masuda K. Ocular hypotensive effect of PhXA41 in patients with ocular hypertension or primary open-angle glaucoma. Jpn J Ophthalmol 1993;37:270-74. Nagasubramanian S, Sheth GP, Hitchings RA, Stjernschantz J. Intraocular pressure-reducing effect ofPhXA41 in ocular hypertension: comparison ofdose regimens. Ophthalmology 1993;100:1305-11. Racz P, Ruzsonyi MR, Nagy ZT, Bito LZ. Maintained intraocular pressure reduction with once-a-day application of a new prostaglandin Fz" analogue (PhXA41). An in-hospital, placebo-controlled study. Arch OphthalmoI1993;lll:65761. Toris CB, Camras CB, Yablonski ME. Effects of PhXA41 , a new prostaglandin F z" analog, on aqueous humor dynamics in human eyes. Ophthalmology 1993;100:1297304. Aim A, Widengard I, Kjellgren D, et al. Latanoprost administered once daily causes a maintained reduction of intraocular pressure in glaucoma patients treated concomitantly with timolol. Br J Ophthalmol 1995;79:12-16. Crawford K, Kaufman PL. Pilocarpine antagonizes prostaglandin Fz,,-induced ocular hypotension in monkeys: evidence for enhancement of uveoscleral outflow by prostaglandin FZa" Arch Ophthalmol 1987;105:1112-16. Gabelt BT, Kaufman PL. Prostaglandin Fz" increases uveoscleral outflow in the cynomolgus monkey. Exp Eye Res 1989;49:389-402. Nilsson SFE, Samuelsson M, Bill A, Stjernschantz J. In-
1752
16. 17. 18.
19.
20. 21. 22.
23. 24.
creased uveoscleral outflow as a possible mechanism of ocular hypotension caused by prostaglandin F z,,-I-isopropylester in the cynomolgus monkey. Exp Eye Res 1989;48: 707-16. Villumsen J, Aim A. PhXA34-a prostaglandin Fz" analogue. Effect on intraocular pressure in patients with ocular hypertension. Br J Ophthalmol 1992;76:214-17. Villumsen J, Aim A. Prostaglandin Fz,,-isopropylester eye drops. Effects in normal human eyes. Br J Ophthalmol 1989;73:419-26. Camras CB, Schumer RA, Marsk A, et al. Intraocular pressure reduction with PhXA34, a new prostaglandin analogue, in patients with ocular hypertension. Arch Ophthalmol 1992;110: 1733-38. Hotehama Y, Mishima HK. Clinical efficacy of PhXA34 and PhXA41: two novel prostaglandin Fz,,-isopropyl ester analogues for glaucoma treatment. Jpn J Ophthalmol 1993;37:259-69. Stjernschantz J, Bito LZ. The ocular effects of eicosanoids and other autacoids: Historic background and the need for a broader perspective. Prog Clin Bioi Res 1989;312: 1-13. Unger WG. Mediation of the ocular response to injury and irritation: Peptides versus prostaglandins. Prog Clin Bioi Res 1989;293-328. Protzman CE, Woodward DF. Prostanoid-induced bloodaqueous barrier breakdown in rabbits involves the EP2 receptor subtype. Invest Ophthalmol Vis Sci 1990;31:246366. Stjernschantz J, Resul B. Phenyl substituted prostaglandin analogs for glaucoma treatment. Drugs Future 1992;17:691704. • Boger WP III, Puliafito CA, Steinert RF, Langston DP. Longterm experience with timolol ophthalmic solution in patients with open-angle glaucoma. Ophthalmology 1978;85:25967.