TRIAL ENDPOINTS

The US FDA classifies clinical trial endpoints as “results, conditions or events associated with individual study patients that are used to assess study treatments”.
However, clinical trial endpoints must primarily support the intended hypothesis of the trial and therefore the features of a disease and how it might be altered with particular treatments need to be addressed for a drug to get approved, said Prof Aniz Girach, Chief Medical Officer at NightstaRx, and Honorary Professor of Ophthalmology at Wills Eye Hospital, Philadelphia, USA, in his lecture to delegates at the 5th EURETINA Winter Meeting at Merton College Oxford, UK.
In short, a drug seeking regulatory approval should ideally improve either survival, or improve patients’ functional outcomes or prevent a disease from progressing, and therefore a clinical trial endpoint must serve one or more of these purposes for the endpoint to be ultimately usable and validated from a regulatory point of view. 
In terms of retinal disease endpoints, Prof Girach explained that such measurements need to be clinically relevant, “true or hard endpoints”, and in the ophthalmic world visual acuity (VA) is probably the most common measurement in regular use today. 
Prof Girach explained that endpoints differ depending on the phase of the trial. For example, in phase 1, safety and tolerability evaluations are the main objective and much effort is focused on these measures, consequently there are few if any efficacy endpoints. 
While an efficacy signal is always welcomed in phase 1 studies, such early stage assessments are rarely powered for such detection. As treatments move into phase 2, a better picture of efficacy begins to build and more elaborate endpoints are included, both primary and secondary, running parallel to the background of safety and tolerability. 
Finally, as a drug progresses into phase 3 the focus shifts to efficacy, as ultimately this is what drives the indication for which approval is received. The secondary endpoints will provide additional data that may also end up on the labelling, however it is the primary endpoint that is key and fundamental in the phase 3 stage. 
As such, significant time and effort is generally invested in designing, validating and negotiating such endpoints with the regulatory agencies. Established retinal endpoints used in clinical trials can be split into the functional and anatomic endpoints, with VA as the functional endpoint that is used most frequently. 
“Any element of visual function testing is accepted by the regulators as an established endpoint, as long as you can show a statistical and clinical relevance in that parameter of visual function at more than one time point,” according to Prof Girach. 
Although VA is one of the key parameters, there are different ways to measure VA. Some trials use mean VA – detecting a mean VA difference from baseline on ETDRS scores between treatment and control groups, and the mean difference between the groups is the critical measure that the regulators look for. 
Anything beyond a mean difference of five or more letters is generally accepted as being clinically significant. Categorical analysis for VA is also a common measurement, so for example, “two or three line gain/loss in VA on the ETDRS Chart” has been accepted in a number of publications to provide a clinically meaningful change and the regulators are becoming increasingly attuned to such a VA measurement. 
In terms of anatomic retinal endpoints, the progression of non-proliferative diabetic retinopathy via a three-step or more progression on the ETDRS retinopathy grading scale has now being established for some time. 
Usually the duration of such diabetes-related trials must be three years, primarily due to the DCCT data (Diabetes Control and Complications Trial) which showed that glycaemic control influences the switch over occurring at the 18-month stage of outcomes in terms of retinopathy progression. 
However, more recently the FDA has become more flexible in allowing shorter duration trials that look at the same endpoints, provided that one can show that the glycaemic control has not been altered. 
The most recent example of anatomic endpoint approvals in the retinal field was with Jetrea (ocriplasmin), which was achieved by a team led by Prof Girach himself.
Aniz Girach: a.girach@nightstarx.com