Fixed-interval Dosing of Anti-VEGF for Wet AMD

Is FIDO the AMD patient’s best friend?

MARC C. PEDEN, MD • MARK E. HAMMER, MD • IVAN J. SUÑER, MD

Anti-VEGF treatments have been a game changer for the treatment of exudative age-related macular degeneration, resulting in greatly improved visual outcomes and maintenance of vision in many patients. These treatments come at an opportune time for our patients, with the prevalence of AMD expected to increase by ~250% over the next 15 years.

Such dramatic increases in treatable pathology will pose a strategic challenge for retinal physicians given the need for frequent visits and procedures in this subset of patients. This phenomenon highlights the need for sustained-release treatment modalities and streamlined practice protocols that will focus on clinical efficiency while maintaining the best possible vision in our patients.

Until these options are optimized, determining an appropriate treatment interval strategy that strikes a balance between patient outcomes and treatment burden to patient and physician practice will be increasingly important.

Retinal physicians currently subscribe to one of three treatment regimens: fixed-interval dosing (FIDO), as needed dosing (PRN), or a hybrid, treat-and-extend dosing (TREX).¹ In this article, we will present justification for a fixed-interval approach to treatment for wet AMD.

WHAT THE STUDIES SAY

The two seminal studies, ANCHOR and MARINA, clearly demonstrated the profound benefits of ranibizumab (Lucentis, Genentech, South San Francisco, CA) when administered monthly over the course of two years, with patients gaining 10.7 and 6.6 letters, respectively, at two years in the 0.5-mg cohort.^2-4

Multiple subsequent trials, including PIER and EXCITE, investigated alternative, less frequent dosing regimens, with quarterly dosing schedules following three monthly loading doses, while SAILOR, PrONTO, and HORIZON examined PRN dosing.

With the exception of the 40-patient pilot study PrONTO, which incorporated optical coherence tomography, vision, and examination retreatment criteria, none of the decreased treatment regimens approached the results seen with monthly FIDO.⁵

Most notably, in the extension trial HORIZON, patients previously receiving monthly therapy in ANCHOR, MARINA, and FOCUS lost half of their visual gains in the first year after being converted to PRN dosing.⁶

Furthermore, the chronicity of the disease was highlighted by up to 70% of patients requiring additional therapy despite their initial 24 months of continuous therapy. However, none of these studies had a fixed-interval control arm for comparison.

HEAD-TO-HEAD COMPARISON OF FIXED-INTERVAL TO PRN DOSING

In 2012, the two-year data from CATT were released, which directly compared PRN dosing to monthly dosing. Noninferiority was demonstrated between both agents and treatment regimens, with the exception of PRN bevacizumab (Avastin, Genentech), when compared to monthly ranibizumab.⁷

The two-year HARBOR data also identified, as a secondary outcome, a lack of statistical difference in visual outcomes between the PRN and monthly arms, with an average injection interval of 9.9 weeks, prompting a change to ranibizumab’s package insert labeling.⁸

HARBOR validated PrONTO’s original findings that, when incorporating OCT findings into additional retreatment criteria, similar visual outcomes could be attained with nearly half the number of injections over the course of two years.

Unfortunately, due to a dearth of long-term outcome data, physicians have been forced to extrapolate this short-term data beyond two years, thereby possibly discounting true variances between regimens that manifest with time.

LONG-TERM OUTCOME STUDIES FOR WET AMD

Nine years have passed since the FDA approved Lucentis and the consequent surge in intravitreal procedures. While multiple trials have reported on one- to two-year outcomes with various drugs and treatment regimens, long-term data are only beginning to be reported.

In 2012, Singer et al reported outcomes with five years of follow-up from the HORIZON study.⁶ As previously mentioned, after two years of monthly therapy, patients crossed over to PRN therapy, with which they were followed on average every two months. By five years, mean visual acuity was -0.1 letters less than baseline from a peak 9.0-letter gain.

An additional cohort from HORIZON was further evaluated in the SEVEN-UP study, in which patients exiting HORIZON after four years of combined therapy were followed for an additional three years.⁹ Further vision loss was noted, with patients ending up 8.6 letters worse than baseline; however, the eyes only received an average of 6.8 injections over the 3.4 years after exit from HORIZON.

This past year, our FIDO revealed that initial visual gains could be maintained over the long term, with patients maintaining, on average, 12.1 letters above baseline at seven years.¹⁰ Furthermore, half of the eyes maintained 3-line visual gains at seven years in FIDO, compared to 12.3% in SEVEN-UP.^9-10

It is important to note that the PRN protocols used in HORIZON/SEVEN-UP are quite different than the standard PrONTO-based monitoring and retreatment criteria, which rely heavily on closer observation and OCT guidance.

More stringent PRN protocols with OCT-guided retreatment protocols may likely result in superior outcomes to these reported PRN outcomes; however, long-term data for this protocol are not currently available.

Similarly, TREX protocols may also provide improved outcomes compared to those seen in SEVEN-UP, and the three-year outcomes appear promising. However, only 34.4% of eyes gained 3 lines or better at three years, compared to 52.3% of eyes at three years in FIDO, again demonstrating that extending too early may result in delayed gainers not receiving their full visual potential, as discussed below.^10-11

ADVANTAGES OF FIXED-INTERVAL DOSING

The basic tenet of FIDO is to maintain therapeutic levels of suppression of VEGF in an effort to help minimize new vessel formation, leakage, bleeding, and fibrosis through timely, continuous redosing. In our FIDO study, patients continued to experience gains out to three years (Figure 1), again demonstrating that switching to less frequent dosing intervals simply based on the appearance of quiescent, stable disease may result in some delayed gainers not realizing their full visual improvement.

Figure 1. Distribution of eyes in study that improved, stabilized, or lost vision compared to baseline. Improvement was defined as eyes that gained ≥15 letters, stabilization as eyes that gained <15 letters but lost <15 letters, and loss as eyes that lost ≥15 letters.

While TREX regimens appear promising in that they meld aspects of PRN and fixed-interval treatment, thereby capitalizing on some of the pros of both modalities, extension may occur before full visual gains are realized. With FIDO, a more conservative and gradual extension is usually employed, thereby limiting the risk of recurrent leakage and devastating submacular hemorrhage.

While no strict guidelines are employed in our practice with regard to the rate of extension, the majority of patients in FIDO received monthly dosing for the first two years. If the vision and OCT were stable, the injection interval was gradually extended by one week every six months out to eight weeks, as long as the patients showed no signs of recurrence.

Patients who demonstrated extensive vision loss in their fellow eye, especially if from wet AMD, were usually kept on monthly therapy. Naturally, these treatment guidelines are guided by input from patients, as well as physician recommendation.

While the previously mentioned PrONTO and HARBOR studies showed essentially equivalent efficacy between PRN and FIDO over the course of two years, CATT showed inferior visual gains when PRN dosing with bevacizumab was employed compared to monthly ranibizumab after the first year.

The development of such discrepancies over time raises the question of whether additional temporal divergences might be seen in this chronic disease. Given the expense of conducting large randomized trials, treatment data have essentially been limited to two-year outcomes, with the few exceptions mentioned above. Only with these more recent publications from HORIZON, SEVEN-UP, and FIDO have we begun to see that a divergence does indeed play out with time.

During the first two years of SEVEN-UP, patients received monthly dosing and were then crossed over to PRN upon enrollment in HORIZON. Visual gains were notable in both the HORIZON cohort and FIDO cohort over the first two years, with 11.2 letters gained in the former and 16.1 letters in the latter.

Over the ensuing five years, patients in the SEVEN-UP cohort lost 19.8 letters, compared to a 4-letter loss seen in those receiving fixed-interval treatment in FIDO (Figure 2). With FIDO, visual gains were attained and maintained, with 50% of eyes showing a >15-letter improvement at the end of seven years, compared to 12.3% in SEVEN-UP.^9-10

Figure 2. Mean letter change from baseline at each time point for HORIZON, SEVEN-UP, and FIDO studies. All eyes received fixed-interval dosing for year 1. The HORIZON and SEVEN-up cohorts were crossed over to PRN dosing after year 2, and initial gains are lost in contrast to FIDO where fixed-interval doing maintained gains with a nonsignificant decrease out to 7 years.

ARGUMENTS AGAINST FIXED-INTERVAL DOSING

While FIDO is felt by many to provide the best coverage and visual outcomes, others are adverse to this regimen, feeling that it incurs onerous burdens on patients and practices, excessive costs to the healthcare system, and possible progression of geographic atrophy.

Burden on Patients and Practice

While the commitment to FIDO is significant, it seems that we, as treating physicians, may be overly paternalistic in our assumptions about how burdensome this regimen truly is for patients.

Not surprisingly, most patients are happy to accept the rigors of treatment in exchange for the increased potential of maintained vision, especially those who already have lost vision in a fellow eye due to disciform disease.

Frequent dosing may present more of a burden to physicians, especially in today’s environment of diminishing reimbursements for injections and the ever-growing population of patients that will require chronic treatment. Clinic schedules can quickly become saturated with these procedures, which remunerate less each year and occupy slots that could otherwise be filled by new patients.

With FIDO, the need for frequent, time-consuming examinations can be lessened, thereby streamlining clinics. As additional intravitreally administered treatments become available, it is quite possible that we may see the incorporation of physician extenders, such as physician assistants or nurse practitioners, into the retina practice for administering injections, thereby helping to offset some of the physician burden. (Actually, we know of two practices where this is already being done.)

Expense

Due to the more frequent dosing schedule with fixed-interval treatment, many are inclined to believe that this treatment regimen is much more costly than other methods. However, this expense is entirely dependent on the agent used and, in fact, may provide cost savings for bevacizumab users.

In our practice, patients on a monthly treatment interval receive 12 injections per year with four clinical examinations and OCTs. PrONTO demonstrated approximately half the number of injections required with monthly visits, and similar to HARBOR, the PRN arm was reduced to 7.7 injections per year with monthly visits and OCTs.

Assuming six injections in a PRN scenario with monthly visits, the cost of FIDO with bevacizumab is $2,491.49/year, compared to $2,909.04/year in the PRN arm (Table). These cost advantages are quickly depleted when using other anti-VEGF agents, with FIDO costing $25,726.96/year with monthly treatment and $14,526.78/year with PRN treatment.

This cost discrepancy is an issue that may be addressed by government reform of drug pricing practices, because it puts physicians in the uncomfortable position of saving money by using non–FDA-approved therapies.

Are We Making Geographic Atrophy Worse?

While VEGF can lead to the deleterious effects seen with choroidal neovascular complex formation, it also provides some level of neurotrophism to the retinal photoreceptors and retinal pigment epithelium that are felt to be important to overall retinal health.

Indeed, VEGF-A–knockout mice demonstrated progressive ablation of the choriocapillaris, with ensuing photoreceptor dysfunction and loss.¹² Subanalyses from CATT reported a 1.59 times greater risk of geographic atrophy in eyes receiving monthly therapy, compared to PRN.⁷

Other studies have confirmed GA, and in fact, it appears to be nearly ubiquitous, with 98% of patients in SEVEN-UP displaying atrophy at 7.3 years.⁹ However, despite the increased prevalence of GA in patients receiving more sustained VEGF suppression, direct causation cannot be inferred.

A recent study by Tanaka et al demonstrated that, after 3.5 years, GA did not tend to occur outside the boundaries of the initial CNV, unless eyes had GA outside this area at baseline.¹³

In essence, GA at the site of CNV may be a desired and anticipated physiologic response that confers response to treatment, rather than a deleterious effect of VEGF suppression. In fact, in FIDO, GA resulted in vision loss in only three of 44 eyes, with 93% of the eyes stabilizing or improving.¹⁰

To infer that sustained treatment results in vision loss from GA is a dangerous position to advocate and may actually result in patients losing vision due to undertreatment.

THE FUTURE

Regardless of the dosing regimen we employ in clinic, we are fortunate to have such successful agents for treatment of wet AMD. FIDO provides the best long-term visual outcomes reported, and it most closely models what we may expect to see with sustained-release treatments.

Despite many misconceptions, FIDO is well tolerated by patients, can increase practice efficiency, and will provide cost savings. As we go forward, it is likely that we will see more data on the long-term outcomes of various treatment paradigms, including TREX and OCT-guided PRN regimens.

However, it is more likely that we will see new treatment regimens emerge, with combination therapies such as PDGF, as well as longer-term VEGF suppression by sustained-release devices or viral vectors. RP

REFERENCES