Avedro’s Eye Drug-Device Combo Haunted By Past Sponsor’s Decisions

FDA advisory committees will weigh clinical trial design, low study enrollment and device changes in review of application for two orphan eye conditions that Avedro didn’t have control over.

The original sponsor of Photrexa didn’t heed FDA’s advice on its Phase III trials’ primary endpoint and that decision is just one of a handful that could come back to hamper Avedro Inc.’s chances of success at its Feb. 24 advisory committee panel.

A joint panel of the Dermatologic and Ophthalmic Drugs and the Ophthalmic Devices advisory committees will also weigh in on the lower-than-planned enrollment in the late-stage studies, including the adequacy of the safety database, the minimum age supported by the application’s pediatric data and the differences between the device the company will market and the one studied in trials, among other issues on an especially lengthy 10-question agenda.

Avedro is proposing the combination product Photrexa Viscous (riboflavin ophthalmic solution) with dextran and Photrexa without dextran combined with KXL-System, a UVA light source for corneal collagen cross-linking, be indicated for the treatment of progressive keratoconus and corneal ectasia following refractory surgery. Both indications have orphan drug designation.

The joint panel will vote on each indication.

The goal of the drug-device combo is corneal collagen crosslinking, which occurs when the Photrexa solutions are exposed to the KXL UVA light. Collagen crosslinking strengthens and stabilizes the cornea to delay progression of the deformation associated with both conditions.

Keratoconus is a naturally-occurring ocular condition characterized by progressive thinning and protrusion of the cornea, resulting in corneal optical irregularities with increasing myopia, irregular astigmatism and consequential loss of best corrected visual acuity. Corneal ectasia can occur after refractive surgical procedures and is characterized by progressive thinning and protrusion of the cornea, resulting in corneal optical irregularities and loss of both uncorrected visual acuity and best corrected visual acuity.

Late Primary Endpoint Change Under Scrutiny

A single investigator first opened the IND and made critical decisions that affected trial conduct. Avedro acquired rights to the product in 2010. According to Avedro’s briefing documents, study UVX-001 was conducted under the Investigational New Drug application 78,933, an investigator initiated study submitted by R. Doyle Stulting at Emory University. All rights to the data generated in UVX-001 were transferred to Avedro in a licensing agreement dated 10 September 2010. Studies UVX-002 and UVX-003 were multi-center studies conducted in the United States under IND 77,882 originally held by Peschke Meditrade GmbH.

The primary endpoint analysis will likely play a key role in the advisory committees’ discussion. FDA’s briefing documents said that the original applicant declined to follow the agency’s recommendations to evaluate the primary endpoint in the Phase III studies at one year.

Instead, the applicant proposed evaluating the endpoint at three months but did agree to extend follow-up evaluations through one year.

The final statistical analysis plan submitted by Avedro changed the primary time point of analysis for the endpoint to month 12. This change was made after study completion, but prior to conducing formal analyses.

Avedro ’s briefing documents state that at the time of UVX-001 study planning, a review of literature suggested the primary efficacy endpoint could be analyzed at three months post-procedure, but additional literature suggested that later time points were better suited for evaluating long-term clinical significance of the procedure because the healing responses following the procedure require six to 12 months to stabilize.

Because patients in the control arm were allowed to crossover to the treatment arm after month three, Avedro did not have 12 month data for most control patients. It used the last observation carried forward approach for subjects in the control group who switched arms and had missing data after Month 3 to conduct the six and 12 month analyses.

It told FDA that this is a biologically plausible approach because of the progressive nature of keratoconus. Since there are no data to indicate these patients experience spontaneous remission, become free of disease or improvement, this LOCF approach would provide a conservative measure of success, the company said.

The primary efficacy evaluation was based on corneal curvature as measured by maximum keratometry (Kmax) evaluated over time. Success was defined as a difference of at least 1 diopter in the mean change in Kmax from baseline to months 3 and 12 between the treatment group and the control group.

All Phase III studies nearly identical in design except that study UVX-001 included both progressive keratoconus and corneal ectasia patients, UVX-002 only evaluated keratoconus patients and UVX-003 only studied corneal ectasia patients. The studies were analyzed separately but data was also pooled based on indication.

While the sponsor met its primary endpoints at month 12 in all studies, for the progressive keratoconus population, statistical significance was not achieved at month 3.

FDA wants the advisory panel to discuss the strengths and weaknesses of the trial design, and consider the potential introduction of bias, the use of LOCF and the stability of corneal response to treatment.

Investigator Pulls Out, Leaving Study Under Populated

The agency also wants input on the number of subjects studied versus what was originally planned.

In particular, enrollment in UVX-001 was far under the planned 160 patients as it was terminated early when the investigator left the study site. This left only 58 keratoconus patients enrolled in that trial and 49 corneal ectasia subjects.

Furthermore, a high percentage of these patients dropped out because the investigator left the site. Nearly 60% of control keratoconus patients and 31% of keratoconus patients on treatment discontinued. For corneal ectasia, 56% of control patients and 17% of on-treatment patients discontinued, most due to the investigator leaving the study.

FDA wants comments on both the efficacy implications of the enrolled population size as well as comments on the size of the safety database.

It also wants input on the minimum age supported by the data based on the number of pediatric patients enrolled and the applicability of extrapolating from adult data.

There were no pediatric corneal ectasia patients enrolled. There were 33 keratoconus patients who were 21 or younger and 10 keratoconus patients who were 16 or younger.

Device Not Studied In Phase III

Another key hurdle for the sponsor may be the device component as the studies were conducted on a different device than the one proposed to be marketed.

Differences FDA highlighted in its questions to the committee between the two devices include the illumination diameter (aperture) and UV focal alignment.

FDA’s briefing documents include a three and a half page table outlining further similarities and differences between two devices, with two and half of those pages focusing on differences.

Given the differences and lack of any data collected using the KXL system, FDA wants to know whether the data it has is adequate to assess the safety and efficacy of the KXL system.

This is Avedro’s second shot at its product’s approval. It first submitted the application in September 2013 and received a “complete response” letter in March 2014. FDA’s briefing documents said the complete response asked for additional information on the drug constituent part, the drug facility inspections, the device constituent part, clinical/statistical information, and clinical site inspections among other comments. A complete response was submitted Sept. 29, 2014.

There are no FDA- approved drugs for either condition, while one device INTACS is approved for keratoconus [See Deal]. FDA said it is aware that there are “practice of medicine use” or study of other products that are not approved for these uses.