Variable Dosing With Ranibizumab
Lessons learned from the PrONTO study.

ANNE E. FUNG MD, GEETA A. LALWANI, MD, & PHILIP J. ROSENFELD MD, PhD

In August 2004, the Prospective OCT Imaging of Patients with Neovascular AMD Treated with Intra-Ocular Lucentis (PrONTO) study was initiated to explore an optical coherence tomography (OCT)-guided variable-dosing regimen with ranibizumab (Lucentis, Genentech) for the treatment of neovascular age-related macular degeneration (AMD). In this protocol, patients received 3 consecutive monthly injections and then received additional injections only if certain retreatment criteria were fulfilled. In contrast to this variable-dosing regimen, the ranibizumab phase 3 trials used fixed-dosing regimens; the MARINA and ANCHOR trials used a fixed monthly dosing regimen and the PIER study used 3 consecutive monthly injections followed by fixed dosing every 3 months.

The goal of the PrONTO study was to determine whether OCT-guided retreatment was feasible and then evaluate the results from the PrONTO Study alongside those of the ranibizumab phase 3 fixed-dosing regimens. This evaluation would help to establish clinically useful OCT guidelines for a larger, phase 4 head-to-head study comparing an OCT-guided variable dosing regimen with a fixed-dosing regimen. The underlying premise for this investigation was that patients are different with respect to their need for retreatment. Some patients may require fewer retreatments compared with consecutive monthly dosing. In the long run, the hope is to decrease the treatment and follow-up burden on the patient and the physician without sacrificing visual acuity (VA) outcomes.

In the following article, we will share the vision and OCT trends of PrONTO patients at 1 year and discuss current clinical applications. Useful lessons learned about Stratus OCT (Carl Zeiss Meditec, Dublin, Calif) imaging while following patients for ranibizumab therapy will also be discussed.

PRONTO DESIGN AND METHODS

Forty patients with neovascular AMD involving the fovea were enrolled between August 2004 and April 2005 in this 2-year prospective investigator-sponsored trial. To be eligible for the study, patients had to have macular neovascularization with an OCT central retinal thickness (CRT) of at least 300 μm. All patients were imaged with both fast and slow macular Stratus OCT scans at each visit. The
1-mm central macular thickness measurements were obtained from the 6 diagonal fast scans (low resolution scans at 128 a-scans per diagonal), and all qualitative assessments were performed on the 6 diagonal slow scans (high resolution scans at 512 a-scans per diagonal).

The treatment protocol was designed to give all patients 3 monthly injections (baseline, month 1, and month 2) followed by retreatment as guided by the following 5 predetermined criteria (Table).
 

PRONTO RESULTS AT 1 YEAR¹

Rapid improvements in VA and OCT were seen during the first 3 months when 3 consecutive monthly injections of ranibizumab were performed. VA improved about
2 Snellen lines through the first 3 months. The improvement in VA was mirrored by rapid quantitative and qualitative improvements on OCT. The excess macular fluid, as reflected in OCT CRT measurements, continued to decrease over 3 months. The mean CRT measurements at month 3 decreased by nearly 200 μm.

Improvements in VA and OCT measurements were sustained at 1 year using this OCT-guided variable-dosing approach. Over the course of the first year, when the initial 3 injections are included, patients received 5 to 6 injections on average. Of the 40 patients, 7 patients did not require any additional injections beyond the first 3. No ocular or systemic adverse events occurred during the first year of study. No thromboembolic events occurred and no new cases of hypertension were observed during year one.

PRACTICAL LESSONS FROM THE PRONTO STUDY

Optimizing OCT Information

Technical expertise is required to obtain high-quality OCT images that are free of artifact and can be used for reliable CRT measurements, as well as for assessing qualitative changes between visits. During the PrONTO study, constant vigilance was necessary to ensure that the thickness measurements were accurate. The same experienced technician was responsible for most of the images obtained. This technician would repeat the scans until artifacts were absent and image quality was optimal. Several factors proved challenging: 1) patients with poor central fixation due to AMD had difficulty keeping stationary for radial line slow scans,
2) less experienced technicians had difficulty locating the foveal center in patients with large AMD lesions, and 3) the internal algorithm of the Stratus OCT had difficulty demarcating the inner retinal boundary and the retinal pigment epithelial (RPE) layer in maculae with abnormal anatomy secondary to neovascularization and scarring. As a result, artifacts were common, resulting in variable CRT measurements. This could only be controlled by reviewing all scans at the time of the visit to ensure the boundaries were appropriately demarcated. Reproducibility was particularly crucial for the PrONTO study because change in the CRT measurement was a criterion for retreatment.

Understanding the OCT: CRT Measurements

The CRT measurements are taken with a fast macular map. Six fast diagonal line scans at 30Þ intervals (like
6 cuts into a pie) are taken in rapid succession to minimize patient movement between scans. The lines should overlap in the fovea to provide the central 1-mm retinal thickness measurement, which is an average of those
6 over-lapping points (Figure 1). The Stratus OCT has an internal algorithm that determines the CRT by tracing the boundaries of the inner retina (vitreous/retinal interface) and the outer retina (the RPE/photoreceptor junction) and calculating the distance in microns between these 2 boundaries. If the boundary tracings are inaccurate, the algorithm calculation will also be inaccurate. For this reason, OCT technicians must learn to identify the appropriate demarcation boundaries within the image to ensure that the thickness measurements are accurate. Technicians should repeat the fast macular map until all
6 scans are correctly traced.

Understanding the OCT: Placement of
Diagonal Scans

Optical coherence tomography also provides important qualitative information regarding the fluid in the macula and its response to ranibizumab therapy. The precise placement of diagonal scans is crucial to qualitative OCT assessments. Variability in the placement of each diagonal scan may result in apparent qualitative changes between visits where no change actually exists. For example, a radial line scan that was incorrectly centered could miss a pocket of subretinal fluid present at a previous visit resulting in the misdiagnosis of a fluid-free macula and therefore could lead to the inappropriate cessation of therapy. To prevent this possibility, OCT technicians must learn to identify the appropriate center of the macula. Unfortunately, using current OCT technology, the technician's job is made even more difficult by having only a blurry, low-contrast, black-and-white video image of the macula to guide positioning. This is particularly problematic when the patient's vision is poor and the macula is distorted with fibrosis or fluid. For this reason, landmarks such as the optic nerve, the retinal vessels, or notable scars or pigment play a crucial role in identifying the fovea. Future OCT technology will help resolve these issues, but until then, it is essential for the retina specialists to play an active role in training OCT technicians and reviewing the scans prior to accepting any changes in CRT measurements or macular fluid between visits.

Understanding the OCT: Follow the Fluid

Differentiating between the different fluid compartments in the macula is important. For example, cystic changes within the retina and subretinal fluid resolve much faster than the fluid in a retinal pigment epithelial detachment (PED). In order to obtain the best qualitative information, 6 radial slow diagonal scans were performed in PrONTO. Technical expertise was required to reliably and reproducibly center all 6 scans on the foveal center so that sequential comparisons between visits were possible. In evaluating the data, the PrONTO study showed that fluid in different macular compartments resolve in a predictable pattern. First, the cystic changes in the retina resolve and make the largest contribution to the dramatic and rapid reduction in CRT. This change is seen usually following just 1 ranibizumab treatment. Pockets of subretinal fluid are the next component to reabsorb, and finally there is resolution of the PED. While most PEDs improve over time, they are variable in their response to ranibizumab treatment and some never completely resolve.

With this attention to OCT, a frequently asked question is whether or not fluorescein angiography (FA) should be abandoned. In our current clinical practice, FAs are performed at baseline to document and confirm the diagnosis of neovascular AMD and again as needed in cases of unexplained vision loss during ranibizumab therapy. FA is particularly useful for establishing the diagnosis of a RPE tear, which is an occasional sequela of neovascular AMD. While impending or early tears of the RPE can be recognized clinically or on OCT (eg, attenuation or corrugation of the the RPE/Bruch's complex), FA is still the gold standard for establishing this diagnosis once a tear of the RPE has evolved. However, OCT is generally the most useful tool for following the fluid and for revealing other causes of vision loss such as atrophy of the photoreceptor outer layer, vitreomacular traction, or epiretinal membrane.

A review of those patients retreated based on criteria of new intraretinal hemorrhage or new classic component on FA revealed that all patients exhibited qualitative changes on OCT consistent with recurrent fluid. For this reason, a sixth criterion for reinjection has been added to the PrONTO study for the remainder of the second year: patients will be retreated with ranibizumab for any qualitative change on OCT suggesting new fluid in the macula including the reappearance of cysts or subretinal fluid or an increase in size of a PED.

Evaluating Patient Progress During
Ranibizumab Therapy

When monitoring patients on ranibizumab therapy, the observations of changes in VA and OCT measurements from the PrONTO study should prove to be clinically useful. Changes in OCT occurred within the first day following initiation of ranibizumab therapy and steadily improved (decrease in average CRT and qualitative improvements) throughout the first 3 to 4 months. Average VA also improved within 2 weeks and steadily increased during the first few months.

Generally, changes in OCT preceded changes in VA including the resolution of macular edema prior to measurable improvements in VA. Similarly, the recurrence of edema on OCT in a previously fluid-free macula was seen before VA deterioration. The PrONTO Study confirmed the suspicion that once fluid appears, more fluid follows and VA eventually worsens (although vision loss may lag a few months). With this information, it now makes sense to retreat as soon as fluid first reappears in the macula. In summary, the reappearance of fluid in the macula is the harbinger of future vision loss thus treatment should not be delayed.

Frequency of Treatment

The first-year PrONTO data suggest that not all patients require consecutive monthly ranibizumab treatments and that patients develop individual patterns of activity. Seven of 40 patients (17.5%) in PrONTO did not require additional injections during the first year of therapy yet maintained and continued in their vision gains. However, 1 patient required all 13 injections. The remainder of patients developed a pattern of retreatments generally between 2 and 4 months apart.

SUMMARY

Visual acuity measurements, clinical examination, and OCT evaluations are all that appear to be necessary for the successful management of neovascular AMD patients treated with intravitreal ranibizumab. A head-to-head trial comparing this OCT-guided variable dosing regimen with fixed monthly dosing is required to demonstrate that the 2 strategies result in comparable VA outcomes. However, until this study is done, it seems reasonable to use VA with fast and slow OCT scans to monitor the response to therapy and determine whether retreatment with ranibizumab is necessary.

REFERENCE

1. Fung AE, Lalwani GA, Rosenfeld PJ, et al. A variable dosing regimen with intra-vitreal lucentis (ranibizumab) for neovascular age-related macular degeneration: one year results of the PrONTO study. Am J Ophthlamol. In Press.

Philip J. Rosenfeld, MD, PhD, is professor of ophthalmology at Bascom Palmer Eye Institute in Miami. Geeta A. Lalwani, MD, is a vitreoretinal fellow at Bascom Palmer. Anne E. Fung, MD, is a medical retina consultant at Pacific Eye Associates in San Francisco.
Drs. Rosenfeld and Fung both serve on Genentech's advisory board and speakers' bureau and have received clinical research funding from Genentech.