FDA, AAO collaborate to develop endpoints for evaluating premium IOLs

Recent innovations in optics, materials and mechanics have spawned innovative IOL technologies that are designed to correct more than the spherical error at distance, traditionally corrected by monofocal IOLs.

However, development of appropriate endpoints for premium IOLs have not kept up with those innovations, causing government review processes of new IOL products often to be conducted ponderously case by case. Even then, because of a lack of recognized standards and clearly delineated endpoints, only 39% of investigational device exemptions for premium IOLs, reviewed between fiscal years 2005 and 2012, were approved or approved with conditions within the first round, according to Malvina B. Eydelman, MD, director of the Division of Ophthalmic and Ear, Nose and Throat Devices for the U.S. Food and Drug Administration.

To meet the challenge of facilitating the review process, the FDA first identified the drawbacks in the premium IOL process and then convened a joint workshop in March between the Center for Devices and Radiological Health and the American Academy of Ophthalmology to develop novel premium IOL endpoints.

“The current roadblocks can be summarized as a lack of consensus in some preclinical issues, on best clinical trial design, and appropriate safety and effectiveness endpoints,” Eydelman said. “Additionally … we feel there’s a need for new categories for IOLs, based on new optical properties and/or benefits to the patients.”

In particular, the FDA introduced extended depth of focus (EDOF) IOLs as a new category of IOLs intended to improve near and intermediate performance. The FDA’s other goals for the workshop were to discuss requirements for preclinical and clinical testing and to create a foundation for development of a “leap frog” EDOF guidance.

“That is a mechanism via which we can share our initial thoughts regarding the content of premarket submissions for emerging technology, and this will speed development and approval of future EDOF submissions,” Eydelman said.

Workshop groups focused on four subtopics: premium IOL safety assessments, patient-reported outcome measures, objective assessment of accommodation and subjective assessment of accommodation. Group representatives outlined recommendations in breakout sessions.

“I want to highlight the fact that there are no current standards or even draft standards for [EDOF IOLs],” Eydelman said.

IOL safety assessment

In an overview of current safety events with premium IOLs, AAO panelist Douglas D. Koch, MD, presented known events associated with toric, multifocal, accommodating and phakic IOLs.

For toric IOLs, there is the issue of accuracy and resultant misalignment from miscalculation.

“We’ve recently uncovered the importance of the posterior cornea, which can have an effect as high as a diopter,” Koch said, adding that a misaligned lens results in increased postoperative astigmatism at a new meridian that is oblique to the preoperative meridian.

Slight loss of contrast sensitivity and glare are optical concerns associated with multifocal lenses, although patient satisfaction overall is high, according to Koch.

For accommodating lenses, Koch is more concerned with reports of the mechanical issue of anterior dislocations and, in rare instances, posterior dislocations.

“We’re going to have to understand what our acceptable limit will be for these kinds of events, which are inevitably going to occur at a higher rate with most of those lenses,” Koch said, citing the uncertainties that accompany the motion within the eye with accommodating lenses.

Cataract formation and clinically significant opacities are recognized events associated with phakic IOLs, more so in posterior chamber IOLs, Koch said.

“Progressive endothelial cell loss remains an ongoing concern with anterior chamber iris-fixated [lenses] and with any form of an angle-fixated anterior chamber lens,” he said. “And other issues, such as uveitis, glaucoma, peripheral anterior synechiae formation and especially pupil ovalization, can also occur with many of the anterior chamber designs.”

AAO panelist Anne L. Coleman, MD, PhD, described how the FDA-AAO Quintiles Outcome Registry could be used to update information on adverse event rates with premium IOLs and improve surveillance.

“There’s much interest in getting more information than what’s available in the charts or the electronic health records or claims data,” Coleman said. “For the FDA, they’re interested in promoting the development of national and international device registries for selected products.”

The FDA and AAO are using the registry to gather information on patient and provider characteristics such as ocular and systemic comorbidities, medication use, demographics, surgical technique and medical products used in surgery, Coleman said.

“This is information that you’re not going to get from an electronic medical record unless it’s specifically included,” she said.

Recommendations

Using adverse event rates for anterior and posterior chamber IOLs and incorporating IOL safety and performance endpoints in the international IOL clinical standard used by the FDA since 2001, Walter J. Stark, MD, summarized recommendations from the premium IOL safety assessments breakout group.

The group recommended keeping pupillary block and hypopyon as identified safety events.

“We decided it was better to keep those complications, especially pupillary block with some of the new-design lenses that might be a little bit thicker or have two optical plates,” Stark said.

Other recommendations included the need to better define cystoid macular edema and separate secondary surgical interventions by type: optical or mechanical. Patient reporting of glare and halos should be included, and negative dysphotopsia would be another possible subjectively reported adverse event.

For toric IOLs, the group recommended adding rotation of 15° or more; such rotation would result in a 45% reduction in efficacy, Stark said.

“Additional adverse events should be added for all IOLs,” Stark said. “We need to add negative dysphotopsia as an adverse event and get that under either just as negative dysphotopsia or patient-reported subjective problems.”

For phakic or aphakic anterior chamber IOLs, the group recommended adding rotation, endothelial cell loss, pupil ovalization, peripheral anterior synechiae, and patient-reported symptoms or problems.

For posterior chamber phakic IOLs, the group recommended adding rotation, endothelial cell loss and pupil ovalization.

Another suggestion was adding lens malposition, subluxation and decentration resulting in a loss of two lines of best corrected visual acuity.

Patient-reported outcomes

FDA panelist Michelle Tarver, MD, PhD, discussed the use of questionnaires to record patient-reported outcomes (PROs).

“First, define the diseased population. Define the context of use. Select the concepts of measurement that will define treatment benefit or the safety concern,” Tarver said. “If you cannot observe it, you will need a PRO to capture that information. Once you decide that you want to use a PRO in your clinical study, you have to understand exactly what concept you’re trying to capture. And in order to inform that, you should conduct patient interviews, focus groups, as well as qualitative cognitive interviewing. Acquisition of empirical data will help revise and format the actual conceptual framework.”

Interpretability of questionnaire results, precision, reliability, reproducibility, responsiveness and content validity are also critical, Tarver said.

“Content validity should be determined after you know exactly what concept you’re trying to measure and whether the context of use is appropriate. If an existing instrument is being used for a new context of use, we may need additional content validity evidence,” she said.

Patient questionnaires need to differentiate visual function from visual disturbance, invited panelist Oliver D. Schein, MD, MPH, MBA, said.

“I would define a visual disturbance as some kind of visual side effect or limitation to quality of vision that is not wanted,” Schein said. “Not all disturbances yield a drop in visual function. Unfortunately, when patients do complain about visual disturbances, they don’t correlate so well with objective measures.”

Patient satisfaction is also an important patient-reported outcome, Schein said.

Recommendations

The PRO group agreed that a questionnaire should include separate measurements, Schein said.

“The consensus is that issues such as quality of vision and visual function, although there is an overlap and there is clearly some association, are different,” he said. “You can succeed in one and fail in the other. So, the questionnaires should not merely be subsets of each other and should be different.”

The group also concluded that visual function should be measured across a range of vision, Schein said.

More work needs to be done to establish standard definitions of clinical parameters such as visual disturbance and patient satisfaction, Schein said.

“We agreed that the PROs had an important role in this domain,” he said. “We agreed the concepts of quality of vision and visual function are of independent importance.”

Objective assessment of accommodation

Objective measurements are needed to show optical and biometric changes associated with accommodation, Adrian Glasser, PhD, AAO panelist, said.

“Accommodation is an active process that includes an optical change, and it includes physical biometric changes. Accommodation is not simply the ability of a distance-corrected eye to see clearly at near — therefore, the importance of objective measures of accommodation,” Glasser said. “Objective measurements are absolutely necessary, and practical solutions do exist.”

Refractive changes and biometric changes that occur in accommodation should be measured, Glasser said.

“If no accommodation is stimulated, no accommodative optical or biometric change will occur and none will be measured. If the subject for some reason chooses not to accommodate to [an] accommodative stimulus, you will see no accommodation. You will measure no accommodation,” he said.

The International Organization for Standardization (ISO) recommends only optical methods, while the American National Standards Institute (ANSI) calls for optical or biometric methods, Glasser said.

Depth of field and depth of focus influence accommodation, Glasser said.

“By definition, the depth of field of the eye is the range of movement of an object in object space over which there is no perceptible change in image on the retina. That’s depth of field,” he said. “Depth of focus is related, and that is the range of movement of the image in image space over the retina over which there is no perceptible change in image focus.”

Stimulating accommodation may be difficult in patients with large depth of focus, according to Glasser.

“Many factors affect the ability to stimulate accommodation, and they include the blur perception, contrast sensitivity, depth of focus, ocular aberrations,” he said.

Pilocarpine may be used to stimulate accommodation in an eye with a large depth of focus or one implanted with an accommodating IOL with some multifocality, Glasser said.

However, “pilocarpine is not perfect. There are many problems,” he said.

Glasser showed results of a study in which patients with light irises had faster and more dramatic pupil constriction with pilocarpine stimulation compared with those with dark irises.

Recommendations

Glasser said his group recommended that IOLs cannot be assessed only by optical methods and that biometric methods may be appropriate in some cases.

Concerning whether any biometric methods may be inappropriate for measuring accommodation, the group decided that it would be “incumbent upon the sponsor to choose a method that affords the appropriate resolution to measure the IOL being investigated,” Glasser said. “It was felt that no general statement could be made that would be applicable to all IOLs with regards to biometric methods.”

The group failed to reach consensus on the development of standard operating procedures for use of pilocarpine or cycloplegia to stimulate accommodation.

Subjective assessment of accommodation

One key parameter in subjective assessment of accommodation is the defocus curve. Recommendations from the ANSI on the defocus curve include the study of best corrected visual acuity in 30 patients and 30 controls, Scott MacRae, MD, AAO panelist, said.

“Measuring that visual acuity in half-diopter steps for the near testing between +2 D and –5 D is recommended, with frequent chart changes,” MacRae said. “That, from a practical standpoint, is very labor intensive, and one must also specify the range of power and the number of chart changes. As you can imagine, it requires a lot of patient-examiner time.”

Pupil size is also critical in subjective assessment of accommodation, MacRae said.

“What’s not commonly known [is that], as the pupil dilates, astigmatism becomes much more of an issue, particularly in multifocal patients. So, a 3-mm pupil is a much different problem with astigmatism than a 6-mm pupil. You get much more aberration from that 6-mm pupil,” he said.

Optical barriers to subjective assessment include poor lighting. Pupil constriction, pinhole aperture, undercorrection and overcorrection also need to be addressed, MacRae said.

“Some of these, I think, can be helped with tablet technology, iPads and things like that,” he said.

Other limits of subjective testing include human variables such as patient effort, psychological effort, fatigue and literacy, MacRae said.

Concerning subjective accommodating and multifocal IOL testing, intermediate distance and luminance should be defined. MacRae noted that a webcam or iPad can be used to conduct subjective vision testing.

“You can control the lighting. You can probably control distance with that,” he said. “I think we’ll see apps that will allow us to do this type of testing in a standardized way. We can also test with non-letter items such as animals, non-animals, faces or objects such as cars.”

MacRae summarized the lack of standards for subjective testing of accommodation.

“In terms of subjective testing, there’s really no standardized time for testing, no standardized near vision testing, no consensus on binocular vs. monocular. Contrast sensitivity is not standardized, and we have large standard deviations with this. Higher-order aberrations and scatter are greater with multifocals, and that may enhance or reduce the effectiveness of these systems,” he said. “What else is needed? We need to correlate our subjective tests and our objective tests. We also need to correlate that with subjective questionnaires. Patients may be able to perform short-term reading tests, but are they having marked fatigue or are they having headaches because of the blurred imagery? And how satisfactory is this?”

Recommendations

AAO panelist Elizabeth Yeu, MD, summarized recommendations for subjective assessment of accommodation. The group agreed with ANSI and ISO standards that defocus curve testing should involve the vision chart target in the distance for EDOF lenses and 40 cm for accommodating lenses.

“So, instead of the more commonly used visual threshold for defocus curves of 20/30, or a logMAR 0.3, it would be a little tighter, closer to a visual acuity of a 20/30 target and logMAR of 0.2,” Yeu said.

In addition, the group recommended standard luminance of 80 cd/m² to 100 cd/m² with almost 100% contrast sensitivity and monocular testing. All pupil sizes should be measured and classified as less than 2.5 mm, 2.5 mm to 4 mm, or greater than 4 mm, Yeu said.

“The minimum goal of the testing should demonstrate that there’s a minimum of a half diopter of greater depth of focus over the control, “ she said. “Specific to accommodative IOLs, beyond just monocular testing, there should be binocular defocus curve testing and, again, as I mentioned, a near vision target. And companies with different lenses can also seek modified labeling utilizing different contrast levels or mesopic and photopic lighting.”

The group recommended performing tests with a standardized computer visual acuity chart that randomizes letters, developing a standard operating protocol and, when performing defocus curves, starting subjective testing in the plus range and moving into the minus range, Yeu said.

Regarding the question of which methods, along with potential modifications to minimize bias, should be prioritized, the group decided that the top priority is to “establish the subjective methodology that can be standardized for testing,” Yeu said. “Standardizing and coming up with a protocol for the defocus method with a computerized vision chart and randomized letters is the recommendation here.”

Concerning which preclinical laboratory methods to characterize depth of focus associated with an EDOF IOL are most viable, the group stated that two methods should be used: a multifocal optical bench with white light only and an adaptive optics bench test, Yeu said.

“This will help objectively measure the image quality over the defocus curve for a specific model eye in a real-world situation,” she said. “This does utilize white light. There does not need to be a designated standard model eye, so long as this model eye can be defended for whichever corneal aberrations are being utilized based on established information.”

The group also recommended against a new subgroup of lenses that may be accommodating and have a pseudoaccommodating component because of potential clinical confusion to patients.

“We tried different names,” Yeu said. “Even if you called it an ‘enhanced accommodative IOL’ or some hybrid form of an IOL, those types of adjectives are both misleading as well as ambiguous in nature, so we are recommending against having a separate group.” – by Matt Hasson