24-Month Phase I/II Clinical Trial of Bimatoprost Sustained-Release Implant (Bimatoprost SR) in Glaucoma Patients

In this 24-month, phase I/II, open-label, multicenter study, a paired-eye comparison of Bimatoprost SR (four dose strengths) was performed in successive cohorts of patients with OAG (ClinicalTrials.gov identifier NCT01157364). The study followed an adaptive design in which expansion of the ongoing trial was guided by careful consideration of the safety and efficacy data emerging from the trial by a data review committee, which determined the dose strengths to be evaluated as the study progressed [28]. Details of the study design and methodology have been reported previously [28].

Patients were enrolled from 23 September 2010 to 5 August 2014 at 24 clinical sites in six countries: Australia, Canada, Israel, Philippines, Singapore, and the USA. The study was performed in compliance with good clinical practice, the principles of the Declaration of Helsinki, or the laws and regulations of the country in which the study was conducted. Institutional review board or independent ethics committee approval was obtained at each site before the study began, and all patients provided written informed consent before undergoing any study-related procedure.

Detailed patient eligibility criteria were previously described in Lewis et al. [28]. Briefly, patients aged ≥ 18 years were eligible to participate if they had a diagnosis of OAG with > 1 dB and < 17 dB of mean deviation visual field loss in the study eye, a history of a minimum of 20% IOP-lowering response to a topical PGA ocular hypotensive medication, diagnosis of OAG or ocular hypertension in the fellow eye that could be treated adequately with topical bimatoprost 0.03% monotherapy, and an iridocorneal angle inferiorly in the study eye of Shaffer grade ≥ 3 (by gonioscopy) to accommodate the implant without corneal endothelial contact. IOP in both eyes after washout was required to be between 22 and 36 mmHg, inclusively, at hour 0 (8:00 a.m. ± 1 h) at the baseline visit. Key inclusion and exclusion criteria are summarized in Online Resource 3. Key exclusion criteria included a history of posterior capsule tear during cataract surgery, any ocular surface finding of greater than trace severity on biomicroscopic examination at baseline, history of conjunctival hyperemia of greater than mild severity with bimatoprost or other PGA use, history of narrow-angle or closed-angle glaucoma, central corneal thickness < 470 µm or > 630 µm (or a difference between eyes > 70 µm), and central endothelial cell count < 2000 cells/mm² by specular microscopy.

Patients who were using topical IOP-lowering medication in either eye at screening (days − 42 to − 4) discontinued the medication during the screening period washout window. The required period of washout before the baseline visit was dependent on and prespecified by drug class/product: parasympathomimetic and carbonic anhydrase inhibitor agents, sympathomimetic and α-adrenergic agonists, β-adrenergic antagonists, PGA, and fixed-combination products. If both eyes met the study eye entry criteria at baseline, the eye with the higher IOP (or if the IOP was the same in both eyes, the eye with the larger iridocorneal angle width on optical coherence tomography (OCT) as determined by an independent reading center) was selected as the study eye. Patients were assigned to receive Bimatoprost SR 6, 10, 15, or 20 µg and were masked to the dose strength that they received. The site staff who measured IOP were also masked to the study treatment dose strength assignment. On day 1, the Bimatoprost SR implant was administered intracamerally to the study eye using the preloaded, single-use applicator described above. The 6-, 10-, and 15-µg dose strengths were available as a single implant that varied in size; two 10-µg implants were administered concurrently via the same applicator to achieve the 20-µg dose strength. The study eye was prepared for administration according to standard clinical practice: a broad-spectrum topical antibiotic and a topical anesthetic were administered, and the eye was prepped with 5% povidone–iodine ophthalmic solution and draped in the usual manner for sterile intraocular procedures. The entrance site for the applicator needle was anterior to the insertion of the conjunctiva through the clear cornea in the superior or temporal quadrant of the cornea, with the trajectory of the needle parallel to the iris plane. After administration, the needle track was checked for aqueous leakage and to ensure it was self-sealing, and a topical broad-spectrum antibiotic was applied.

Patients were instructed to instill 1 drop of bimatoprost 0.03% ophthalmic solution in the fellow (control) eye once daily in the evening throughout the 2-year duration of the study. Rescue treatment using topical IOP-lowering drops could be initiated in any eye (study and/or fellow eyes) at the investigator’s discretion if the eye failed to attain the target IOP (as determined by the investigator) on consecutive visits ≥ 1 week apart or if it was considered to be in the best interest of the patient.

A data review committee assessed the ongoing safety in patients who received one or two (20-µg dose strength) implant(s) and concluded it was safe to administer an additional implant if needed for IOP control. As a result, a protocol amendment was implemented in 2013 allowing patients who had received a ≤ 15-µg dose strength of Bimatoprost SR to receive a single repeat treatment with the same dose strength in the study eye between day 90 and month 12 (Online Resource 4) if they met all of the following criteria: neither eye had received rescue therapy, the study eye failed to maintain a ≥ 20% reduction in IOP from baseline at 8:00 a.m. on consecutive visits ≥ 1 week apart, and the initial implant demonstrated adequate safety and did not contact the corneal endothelium. Patients who received a repeat treatment with implant had follow-up visits for a minimum of 12 months after the second administration. Patients could complete the study and be exited at or after the month 18 visit if they had received rescue treatment in the study eye and the implant(s) were not visible to the investigator on gonioscopic examination.

IOP was measured at 8:00 a.m. by Goldmann applanation tonometry at all visits using a two-person masked reading method. At baseline, weeks 4, 8, 12, 16, and 20, and month 6, additional measurements were taken at 10:00 a.m., 12:00 p.m., 2:00 p.m., and 4:00 p.m. Patients who had not been rescued in the study eye had additional diurnal measurements at months 9, 12, 15, 18, 21, and 24.

Safety assessments included AEs, biomicroscopy, ophthalmoscopy, macroscopic conjunctival hyperemia, iris color, gonioscopy, visual acuity, visual fields, central corneal endothelial cell density (CECD) by specular microscopy, corneal thickness by pachymetry, and OCT of the macula.

Patients completed a questionnaire on day 8 and week 4 after the first and second (when applicable) administrations. This questionnaire assessed the burdensomeness of the study eye procedure compared with patient expectations. Possible answers were that the procedure was much less burdensome, somewhat less burdensome, as burdensome, somewhat more burdensome, and much more burdensome than expected. A second questionnaire administered at week 12 (after the first administration) and at month 24 or early exit assessed the likelihood that the patient would agree to receive another implant if given the choice and the likelihood that the patient would recommend the implant to individuals with the same disorder. Possible answers were extremely likely, very likely, somewhat unlikely, very unlikely, and extremely unlikely.

Secondary endpoints included the mean IOP, use of IOP-lowering rescue treatment or second Bimatoprost SR administration, and patient-reported outcomes. The number of rescue medications used in the study and fellow eye at each time point was also evaluated.

All analyses were performed using SAS software version 9.3 or newer (SAS Institute Inc., Cary, NC, USA). All analyses of IOP used observed data with no imputation for missing data in the modified intent-to-treat population of all treated patients with at least one IOP measurement at baseline and at least one post-baseline IOP measurement through week 16. Data were censored at rescue/retreatment in some analyses. Safety parameters were evaluated using observed values in the safety population of all patients who received the study treatment.

Sample size determination was conducted by a data review committee, which reviewed the available efficacy and safety data throughout the study and determined the dose strengths of Bimatoprost SR to be evaluated and the number of patients to be enrolled for each dose strength. Because of the adaptive nature of this study, the sample size of 75 patients was determined empirically rather than selected to provide power for statistical comparisons between implant dose strengths or between study and fellow eyes.

In total, 75 patients were enrolled and received the study treatment, and 63 (84.0%) completed the study. Reasons for study discontinuation included AEs in two (2.7%) patients (cataract, considered treatment-related; adenocarcinoma, considered unrelated to treatment); personal reasons in two (2.7%) patients; lack of efficacy in one (1.3%) patient; loss to follow-up in two (2.7%) patients; and “other” reasons in five patients (6.7%). Rates of discontinuation were similar among the 10- and 15-µg dose strengths (Online Resource 5).

Demographics and study eye characteristics were similar among the groups treated with different dose strengths (Online Resource 6). All patients were diagnosed with primary OAG in both eyes. Overall, mean ± standard deviation (SD) age was 63.4 ± 11.7 years, 45% were aged > 65 years, 51% were female, and 71% were White. Iris color was dark (brown or dark brown) in 53% of study eyes and light in 47% (Online Resource 6). All patients in the Bimatoprost SR treatment groups required washout of prior medications and had a mean IOP increase of 7.8 ± 4.0 mmHg from screening to after washout (baseline).

All dose strengths of Bimatoprost SR reduced the IOP in study eyes. In all dose groups combined, 44 patients required rescue treatment with eye drops; mean IOP (hour 0) prior to initiation of rescue was 23.7 ± 4.8 mmHg [median 23.3 mmHg (range 14–34)]. Figure 2 shows the primary endpoint of time-matched IOP change from baseline at hour 0 (8:00 a.m.). In the preplanned analysis, data were censored from the start of rescue/retreatment, such that the results reflect IOP lowering of a single Bimatoprost SR administration. A sustained IOP-lowering effect that persisted up to month 24 was seen, with the mean IOP reduction from baseline at 24 months in study eyes that had not been rescued or retreated ranging from 5.7 to 7.4 mmHg across all Bimatoprost SR dose strengths (Fig. 2a).

Subsequent analysis of the IOP change from baseline used all observed IOP data in the modified intent-to-treat population, including patients who required rescue treatment with eye drops (Fig. 2b). Although no statistical comparison of the preplanned (Fig. 2a) and post hoc (Fig. 2b) analyses were performed, the overall results indicated similar IOP reductions from baseline between eyes remaining on the initial treatment and rescued eyes for both the Bimatoprost SR-treated eyes and the pooled fellow eyes treated with topical bimatoprost. Analysis of all available data showed a trend toward numerically larger mean IOP reductions in rescued eyes when compared with results of the censored data analysis (Fig. 2a, b). During the first 16 weeks of the study, a dose response of Bimatoprost SR was generally evident in both analyses (Fig. 2); after week 16, the magnitude of IOP reduction with each dose strength was similar to that of a topical PGA. The 8:00 a.m. mean overall study eye IOP reduction from baseline through week 12, the period used in the ARTEMIS registration trials (all observations weighted equally), ranged from 7.4 to 9.8 mmHg across the Bimatoprost SR dose strengths, compared with 8.4 mmHg in the pooled fellow eyes treated with topical bimatoprost 0.03% (Table 1). Mean IOP reductions from baseline at month 24 for all available (observed) data were 7.5, 7.3, 7.3, and 8.9 mmHg for the 6-, 10-, 15-, and 20-µg dose strengths in study eyes, respectively, versus 8.2 mmHg in all pooled fellow eyes, with a mean number of additional medications used (including rescue topical IOP-lowering medications or implant retreatment) of 0.64, 0.75, 0.67, and 0.90 in study eyes versus 0.05 in pooled fellow eyes (Fig. 2b).

IOP was controlled without rescue or re-administration in 51 (68%), 30 (40%), and 21 (28%) study eyes up to 6, 12, and 24 months, respectively (Fig. 3). Subgroup analysis of the patients who completed the study without receiving rescue or re-administration in the study eye showed persistent IOP-lowering effects of a single implant administration in the absence of additional therapy. Mean IOP in study eyes that were not rescued or retreated was reduced from baseline values of 23.6, 23.4, 23.2, and 24.4 mmHg to 18.0, 16.0, 15.9, and 17.8 mmHg at 24 months after a single administration of Bimatoprost SR 6 µg, 10 µg, 15 µg, and 20 µg, respectively (Table 2).

Among study eyes that received a second administration of Bimatoprost SR, mean baseline IOP values (hour 0) were 25.2 (n = 6), 25.0 (n = 8), and 26.0 (n = 10) mmHg for eyes in the 6-µg, 10-µg, and 15-µg dose strength groups, respectively. IOP lowering post-treatment in these eyes was similar after the first and second implants. Mean ± SD reduction in IOP from baseline (data censored at rescue/retreatment) averaged over 12 weeks after administration was 7.9 ± 3.6 mmHg after the first implant and 8.4 ± 4.2 mmHg after the second implant for the combined 6-, 10-, and 15-µg dose strengths.

Implants were typically observed to initially swell as they biodegraded, and the majority of those received at the first administration visit (day 1) were estimated by the investigator to be 76–125% of their initial size at assessments through month 6. By 12 months, the majority of those implants had either totally biodegraded (n = 6) or were estimated to be ≤ 25% of their initial size (n = 26). At month 24, a total of 34 patients were evaluated with gonioscopy, and the implant administered on day 1 had biodegraded and was no longer visible in 26.5% (9/34) of these patients. Of all implants administered on day 1 (including two implants per study eye for the 20-µg dose strength), 30 implants (in 25 eyes) were still visible at month 24. Size assessment was missing for 6.7% (2/30) of these implants; of those assessed, the implant size was estimated to be ≤ 25% of initial size for 75.0% (21/28) of the implants and 26–50% of initial size for 25.0% (7/28) of the implants. Analysis of residual implant at month 24 was limited to assessments for patients who did not meet criteria of total implant biodegradation allowing early study completion (at or after month 18).

Among patients who were not rescued/retreated or discontinued from the study, 28.6% (6/21) had no visible implant in the study eye on gonioscopic examination at month 24, yet IOP remained controlled (mean IOP of 16.9 mmHg) at month 24.

One or more ocular AEs were reported in 64.0% of study eyes (48/75) and in 48.0% of fellow eyes (36/75) (Table 3). Consistent with the 6-month results [28], many of the AEs in study eyes occurred within 2 days after implant administration, i.e., within the immediate post-administration period, and were likely associated with the administration procedure itself. These AEs were typically mild and resolved quickly. The most common AE in the immediate post-administration period was conjunctival hyperemia, which could likely be attributed to the application of ophthalmic povidone–iodine during the preparation for administration.

By contrast, more than 2 days after the procedure (outside of the immediate post-administration period), AEs showed an equivalent overall incidence in the study and fellow eyes, and the incidence of some AEs typically associated with topical PGA use was lower in the eyes treated with Bimatoprost SR versus eyes treated with topical bimatoprost. For example, of AEs with onset later than 2 days after Bimatoprost SR administration, conjunctival hyperemia was reported in 17.3 versus 28.0%; eyelash growth in 0 versus 6.7%; and iris hyperpigmentation in 0 versus 4.0% of all study eyes and topically treated fellow eyes, respectively (Table 3).

No study eyes and two fellow eyes treated with topical bimatoprost were found to have periorbital fat atrophy upon biomicroscopic examination, and one of these findings was reported as an AE. There were no serious ocular AEs in the study eyes, and the only serious ocular AE reported during the study was retinal detachment in a topically treated fellow eye. No treatment-related serious AEs or non-ocular AEs related to the study treatment were reported in any patient, and no patient required implant removal for safety reasons. Furthermore, no additional safety concerns were observed in patients who were retreated with a second implant or in patients in the 20-µg dose strength group who received two 10-µg implants on day 1.

The baseline mean ± SD CECD was 2698.6 ± 196.7 and 2678.5 ± 199.0 cells/mm² in study and fellow eyes, respectively. At month 24, the mean ± SD CECD was 2607.0 ± 203.0 and 2651.6 ± 196.4 cells/mm², respectively, reflecting mean ± SD reductions from baseline of 88.9 ± 135.3 and 22.2 ± 127.8 cells/mm² over 2 years in the study and fellow eyes, respectively (Fig. 4). There was no deterioration in the mean visual field mean deviation over time in Bimatoprost SR-treated eyes. At month 24, the mean ± SD change in mean deviation from baseline was + 0.20 ± 3.16 dB in Bimatoprost SR-treated study eyes compared with − 0.35 ± 3.04 dB in topical bimatoprost-treated fellow eyes (n = 59) (Fig. 5). There were no notable differences between study eyes and fellow eyes in the change in central corneal thickness from baseline (Online Resource 6).

Most patients reported that the procedure was somewhat or much less burdensome than expected (67–100% across treatment groups and time points). At month 24 across all dose strengths, 82.9% of patients reported that they were extremely or very likely to have another implant procedure if given the choice, and 88.6% of patients reported they would recommend the procedure to someone else with the same eye condition (Table 4).