Mixed results for neovascular AMD gene therapy

February 29, 2016

In a Phase IIa study of patients with neovascular age-related macular degeneration, a single subretinal injection of rAAV.sFlt-1 gene therapy (AVA-101) demonstrated acceptable safety, but not a complete or durable response. Additional preclinical research is underway.

Reviewed by Jeffrey S. Heier, MD

In a phase IIa study, rAAV.sFlt-1 gene therapy (AVA-101, Avalanche Biotechnologies) for neovascular age-related macular degeneration (AMD) demonstrated acceptable safety, but failed to provide a complete or durable anti- vascular endothelial growth factor (VEGF) response.

The results are disappointing, but do not necessarily spell the end of the investigational product’s development, according to Jeffrey S. Heier, MD, director of the Vitreoretinal Service, Ophthalmic Consultants of Boston, MA.

‘I still believe gene therapy is well suited for the management of retinal disease. It has a proven track record in humans and, for a variety of reasons, the retina is ideally suited for developing gene-based therapies. Furthermore, there is a strong rationale for developing a sustained delivery product that could help with the ongoing treatment burden,’ commented Dr. Heier.

‘Numerous factors are being studied in an effort to fully understand the responses observed in the Phase IIa study of AVA-101, and more preclinical research is underway that will hopefully provide insight into the variables that can improve outcomes.’

AVA-101 uses an adeno-associated viral vector to deliver a gene that produces naturally occurring soluble VEGF receptor-1 (sFlt-1). The study enrolled patients with treatment-naive or previously treated neovascular AMD. In total, 32 patients were randomised 2:1 to receive a single subretinal injection of AVA-101 or control.

All patients received an intravitreal injection of ranibizumab 0.5 mg (Lucentis; Genentech, San Francisco, CA) at day 0 and month 1. The gene-therapy group underwent core vitrectomy and received the AVA-101 subretinal injection adjacent to the macula on day 7, and all patients returned for monthly follow-up visits until week 52. During the course of the study, all patients were eligible for rescue treatment with ranibizumab, which was given as needed based on pre-specified retreatment criteria.

At week 52, mean visual acuity increased from baseline by 2.2 letters in the AVA-101 group, while the control arm showed a mean loss of 9.3 letters. ‘There was a statistically significant difference between the treatment and control arms, but this was largely driven by poor performance in the control arm due to three subjects losing at least four lines of visual acuity from baseline,’ Dr. Heier said.

Anatomic outcomes were better in the control arm. Centre point thickness (CPT) at one year showed a 27 micron increase from baseline in the AVA-101 arm but an 85 micron decrease from baseline in the control arm.

There were no serious ocular adverse events associated with the investigational therapy and no systemic safety signals attributed to AVA-101. All patients in the AVA-101 arm who were phakic at baseline developed a cataract, and three AVA-101 patients (14%) developed a moderate vitreous haemorrhage that was related to the sclerotomy or trocar insertion used for delivery.

Lessons and questions

 

Lessons and questions

Dr. Heier said that when results do not turn out as expected, it is necessary to go back to basics and try to understand the underlying reasons. He pointed out that there was significant variation in the number of rescue ranibizumab injections received by patients in the treatment arm.

‘For example, there may be something we can learn from determining why some patients received very few ranibizumab injections and others received many more,’ he said.

Other issues being evaluated include whether the subretinal delivery route and the AAV vector itself are optimal – not all vectors are created equal, Dr. Heier said.

He also noted that the experience of this study underscores the challenges in conducting surgical treatment trials.

‘Surgical trials for treating medical diseases are extremely difficult to design,’ Dr. Heier said.

Issues to consider pertain to patient selection criteria, what represents an appropriate sham control and masking of patients and examiners.

Dr. Heier also made the point that it is difficult to implement the control arm of a surgical trial providing extended delivery treatment for neovascular AMD. Although standard of care today mandates anti-VEGF therapy for any level of fluid because it represents early disease recurrence, subtle fluid may not be an indicator of disease recurrence in an eye treated with an extended delivery therapy.

‘Perhaps the latter patients are at less risk of vision loss when subtle fluid is seen,’ Dr. Heier said.

In addition, a surgical approach introduces confounders that can be difficult to control for within a clinical trial.

‘It is much more difficult to control for differences in eyes in a surgical study than in a study of an intravitreal injection. There are also differences in surgeon skills and other issues to consider, such as where to place the bleb and how to manage cataracts,’ Dr. Heier explained.