PATIENTS AND METHODS

NTG patients with progressive optic disc cupping/visual field defects not associated with other disease, new disc haemorrhage, or glaucomatous field defects that threatened fixation were recruited between November 2000 and September 2001. Diagnosis of NTG is based on reproducible field defects of a retinal nerve fibre type (Humphrey visual field analyser, program 24-2), corresponding disc findings, open angles, and normal IOP. IOP was <22 mm Hg on multiple measurements including one diurnal phasing (every 2 hours from 8 am to 8 pm). Exclusion criteria were corneal abnormality, a history of corneal surgery, other diseases that might cause field defects, ocular trauma, ocular laser/surgery within 12 months, advanced glaucoma with impaired visual acuity, or severe systemic disease.

Patients taking glaucoma medication before the study underwent a washout period of 4 weeks for topical β blockers and latanoprost, 2 weeks for topical adrenergic agonists, and 1 week for cholinergic agents and carbonic anhydrase inhibitors.^18,19 Topical lubricant (Artelac; Dr Mann Pharma, Berlin, Germany) was given during the washout period. Patients were then randomly allocated to two groups. Group 1 received latanoprost once daily at 9 pm for 4 weeks, lubricant for 4 weeks, and then brimonidine twice daily at 9 am and 9 pm for 4 weeks. Group 2 followed the same schedule but with the two active regimens in reverse order.

Patients were examined on day 1, week 4, week 8, and week 12. One masked ophthalmologist measured POBF at 8 am, 12 noon, and 4 pm at each visit. On each occasion, POBF was measured three times with one OBF tonometer (Paradigm Medical Industries, Inc, UT, USA). Institutional review board approval was obtained, and verbal and written consent was obtained from all subjects.

The eye with more advanced glaucoma was selected for analysis if both eyes of one patient were eligible. Data were analysed using the statistical software stata (Stata Corporation, College station, TX, USA). The normality of data distribution was checked with the Shapiro-Wilk test. Comparisons between baseline and after treatment were performed with paired Student’s t test. Multivariate linear regression with generalised estimation equation was used to adjust the effects of IOP on POBF.^20,21 Treatment effect comparisons between two regimens were assessed with a crossover designed analysis of variance (ANOVA). A p value of less than 0.05 was considered as statistically significant.

RESULTS

Thirty two Asian patients were enrolled; one withdrew after developing a facial rash at the first treatment period. Another six patients were excluded because a digital readout of the POBF results could not be obtained consistently on each occasion at each visit. Data of 25 patients (18 men and seven women) with a mean age of 64.4 (SD 12.5) years (range 31–79 years) were analysed. Eighteen patients had no previous glaucoma treatment, and seven patients had been treated with one modality (betaxolol-S in two patients, dorzolamide in one patient, latanoprost in three patients, and laser trabeculoplasty in one patient). Except one eye that received phacoemulsification 10 years before, none had a history of incisional surgery. The mean spherical equivalent was −1.0 (2.5) dioptre (range −6.5 to +3 dioptres). The average visual field mean deviation was −9.66 (7.0) dB (range −3.08 to −24.35 dB), and the mean pattern standard deviation was 8.73 (4.17) dB (range 2.92 to 16.07 dB). No significant side effects were encountered. Ten patients were taking systemic cardiovascular medication (α blockers in two patients, β blockers in three patients, calcium channel blockers in three patients, and angiotensin converting enzyme inhibitors in three patients), and all kept using the same medication during the study.

The data of IOP, mean blood pressure, and pulse rate are shown in Table 1. Brimonidine was associated with a reduction in average mean blood pressure and pulse rate (p = 0.026, p = 0.021, respectively). There was no significant difference (p = 0.768) in baseline POBF measured at 8 am, 12 noon, and 4 pm. POBF increased at all time points with latanoprost, while it increased only at 8 am with brimonidine treatment (Table 2 and Fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The pulsatile ocular blood flow (POBF) (mean (SD)) at different times of day at baseline and after treatment.

Average POBF increased by 213 (257) μl/min (22.8%) from a baseline of 936 (265) μl/min (p <0.001) after latanoprost and increased by 97 (183) μl/min (10.4%) (p = 0.014) after brimonidine treatment. The treatment effect is different in favour of latanoprost (p = 0.04). After adjusting for the factor of IOP, the increase of mean POBF associated with both regimens no longer reached statistical significance (p = 0.424 and p = 0.345, respectively).

To avoid the possible effects of systemic cardiovascular medication on POBF, data were further analysed after excluding patients with such medications. The results remained similar. Changes in POBF with latanoprost and brimonidine were not correlated with the visual field mean deviation (p = 0.987, p = 0.087, respectively) or pattern standard deviation (p = 0.500, p = 0.734, respectively). No significant period effects or carryover effects were observed.

DISCUSSION

The measurement of POBF is based on the pressure waveform from continuous IOP recording and a pressure/volume relation as described by Langham and colleagues.^12,13 The validity of the theoretical basis for POBF determination has been shown to be acceptable and its application may result in reasonable approximations.²² Measurements with the OBF tonometer are reproducible with a coefficient of reliability around 0.92.^23,24

The prostaglandin F_2α can cause vasoconstriction in isolated bovine retinal arterioles.²⁵ Latanoprost, being a prostaglandin F_2α analogue, might exert vasomotor effects on vessels that supply the optic disc. Similarly, as an α₂ adrenergic agonist, brimonidine may affect ocular circulation, particularly at the level of terminal arteriole.²⁶ We found that the increase in POBF after latanoprost and brimonidine is related to their IOP lowering effect. After adjusting for the factor of IOP, the changes in POBF no longer reached statistical significance. This suggests that either they do not exert a direct vasomotor effect to influence POBF, or the effects are too trivial to be revealed by OBF tonometer. The ocular haemodynamic effects of these agents have been investigated with a variety of methods, such as radioactive/colour microspheres, colour Doppler ultrasonography, and scanning laser Doppler flowmetry.^27–31 No significant alteration of blood flow or blood velocity was identified in the retina, optic disc, and retrobulbar vessels. None of the studies, including ours, demonstrated that latanoprost or brimonidine impaired ocular circulation in primates.

Our findings with latanoprost agree with previous studies that show an increased POBF after treatment.^15–17 McKibbin and Menage reported an average POBF increase of 21% in NTG eyes receiving latanoprost.¹⁷ Our results are very similar to theirs despite the fact that our patients had a lower baseline IOP. It is noteworthy that non-pulsatile blood flow is not estimated by this technique, and the ratio of pulsatile to non-pulsatile flow may change with alteration in the heart rate, IOP, and systemic blood pressure. Whether an increase in POBF is beneficial to the optic disc remains debatable.

Fontana and colleagues found a lower POBF in eyes of NTG patients with visual field loss than their fellow eyes with normal field, indicating that POBF may provide an indirect estimation of optic disc blood flow.³² Furthermore, in a study to evaluate the effect on POBF of trabeculectomy in patients with high tension open angle glaucoma, James found a mean increase of 29% in POBF after surgery. This suggests that an increased POBF may be associated with favourable prognosis of glaucoma.³³

To the best of our knowledge, the effect of brimonidine on POBF has not yet been reported. Although a significant increase was identified in our study, the 10.4% increase after brimonidine treatment may not be attributable to drug effects. One study investigated the variation in repeat measurement of POBF in normal subjects 3 months apart and found a variation up to 12.4% (Crowhurst C and Massey AD, data of OBF Labs Ltd), even in experienced hands. The change in POBF should exceed the variability resulting from measurement and physiological variation to be attributable to drug effects.

We noted a statistically significant but clinically insignificant reduction in mean blood pressure (3.4 (7.2) mm Hg) and pulse rate (2.1 (4.2) beats/minute) after brimonidine treatment, findings in accordance with several reports.^34–36 Treatment with 0.2% brimonidine may reduce the systolic blood pressure up to 7.1 mm Hg and decrease the pulse rate up to 3 beats/minute.^34–36 In a study on ocular normotensive subjects with implanted cardiac pacemakers, Trew et al found that the mean POBF increased by about 14% as the heart rate increased from 60–90 beats/minute.³⁷ Although the decreased pulse rate after brimonidine treatment may in itself reduce POBF, the effect is likely to be insignificant.

In conclusion, 4 week treatment with latanoprost increases POBF, which is related to its IOP lowering effect. The increase in POBF noted after brimonidine is within the range of long term variation and may not be attributable to the drug effect. The clinical significance of enhanced POBF remains to be clarified.