The Economics of Anti-VEGF Treatment

With ophthalmic drugs bearing high costs, economic burden can play a key role in treatment.

ADAM WEBER, MD • ANDREW P. SCHACHAT, MD

The escalating cost of health care in the United States is a recurring theme in the nation’s political debate. Physicians cannot ignore that healthcare expenditures accounted for nearly 18% of gross domestic product in 2011, and the growth of healthcare spending outpaces GDP by more than 2% per year.^1,2

One disease affecting a large swath of the population with known ophthalmologic manifestations is diabetes mellitus, and one-third of patients with diabetes age 40 and older have diabetic retinopathy.³ The total cost of diabetes in the United States in 2007 was $174 billion. Diabetes is the leading cause of new onset blindness for people ages 20-74 years.⁴

Further CDC data have indicated that 11% of people with diabetes have some form of visual impairment, with 3.8% having uncorrectable visual pathology.⁵ In a European cohort of type 1 diabetes subjects, the risk of clinically significant macular edema was 18% in 10 years,⁶ and blindness occurred in 9.3% after 25 years.⁷ By 2050, the number of US patients with diabetic retinopathy is expected to climb to 16 million.⁵

These trends of rising health-care costs and disease prevalence have forced physicians into an arena for which the medical education system does not prepare them but that is proving as important as physiology and pathology in delivering quality care to our patients — cost-effectiveness.

DEFINING THE TERMS

To understand and apply cost-effectiveness, we must first have a working definition of the two terms. Cost is commonly conceived in terms of dollars and cents. A physician performs an exam or procedure, and the patient pays a given fee. This simplistic scenario omits the intangible, but equally important, aspects of opportunity costs, costs of managing treatment complications, and lost productivity due to disability, not to mention countless other categories of “cost” that escape an “à la carte” medical pricing paradigm.

Additionally, the definition of effectiveness is not as straightforward as it might first seem. In ophthalmology, the most commonly used metric is lines or letters of vision. However, the literature on cost-effectiveness uses an economic jargon of quality adjusted life-years (QALYs), health state utility values, and the incremental cost-effectiveness ratio (ICER).

VISUAL ACUITY AS A METRIC

Poku et al performed a meta-analysis to investigate the reliability of visual acuity measurements in assessing the true effectiveness of treatments. They found that the most important difference for patients with diabetic retinopathy was between moderate visual impairment (20/50 to 20/100) and vision worse than 20/200.

This finding shows that comparing outcomes to 20/20 vision may skew results to make treatments seem less effective.⁸ It also indicates that a tool other than VA is needed to evaluate treatments.

To be certain, the utility and satisfaction a patient gains from treatment are far more important than the number of letters he or she is able to read on a chart. Patients have different visual demands, and their treatments should be evaluated based on their ability to return the patient to productivity.

These effects may or may not show up as statistically significant differences in tables or charts in a study, and perhaps the most important measuring stick is the patient satisfaction score.

ANTI-VEGF FOR DIABETIC RETINOPATHY

The ophthalmology community has accepted for years that laser is cost-effective for the treatment of proliferative diabetic retinopathy.⁹ Most of the more recent cost-effectiveness literature on DR treatment has focused on macular edema, drawing populations from prior large clinical trials.

In an analysis of the RESTORE trial, Mitchell et al explained that anti-VEGF therapy with ranibizumab (Lucentis, Genentech, South San Francisco, CA), with or without laser, was more economical than laser treatment alone.¹⁰

At one year, patients receiving anti-VEGF therapy had approximately 5 letters more improvement than those in the laser-only treatment group, and they also scored higher on a visual function questionnaire.¹¹

This analysis showed anti-VEGF therapy provided an additional 0.17 QALYs at a cost of approximately $7,050 (£4,191), and ranibizumab was cost-effective for up to 13 injections. Combination therapy with laser and anti-VEGF was less cost-effective, generating 0.13 QALYs more than laser monotherapy for approximately $7,900 (£4,695).

We can see that anti-VEGF monotherapy had a 64% likelihood of being more cost-effective than laser alone, and combination therapy had a 42% probability of being more cost-effective.

The QALY gain from anti-VEGF treatment increased if a younger patient population was substituted into the model, providing 0.26 QALYs if the baseline age was adjusted from ages 63 to 40 years. Adverse events did not occur with sufficient frequency to affect cost-effectiveness.¹⁰

JUSTIFYING HIGH COST

Based on this analysis, the increased cost of ranibizumab is justified by the QALYs delivered, and this benefit is amplified when applied to younger patients who have greater longevity to enjoy better vision.

The test population and metrics used may limit wider application of these findings. The average hemoglobin A1c (HbA1c) in the group analyzed was 7.3%. As evidenced by the Diabetes Control and Complications Trial¹² and the United Kingdom Prospective Diabetes Study,¹³ patients with better glycemic control do not encounter the same rates or severity of diabetic eye disease. We should also question whether the QALY benefits described would persist in populations with more poorly controlled disease.

Additionally, the worse-seeing eye was used for classification in 67.2% of the patients. As demonstrated by Poku et al, effectiveness measurements are more accurate and representative when tied to the better-seeing eye.⁸

The CATT group’s finding of therapeutic equivalence for ranibizumab and bevacizumab (Avastin, Genentech) in treating AMD¹⁴ has been mirrored in the cost-effectiveness literature for diabetic eye disease. Ranibizumab costs approximately seven times as much as bevacizumab per injection, including pharmaceutical and office charges.

Stein et al’s model predicted that this price difference allowed for 12 injections of bevacizumab, compared to less than seven of ranibizumab, to meet a cost-effectiveness goal of $50,000/QALY in treating DME.¹⁵

ADDING ADVERSE EVENTS TO THE EQUATION

Genentech, which manufactures both medications, has stated that ranibizumab should be the preferred agent for intravitreal injection despite its higher cost, due to a decreased risk of systemic side effects.^16,17

However, the literature demonstrating increased cardiovascular side effects from bevacizumab has been based on patients receiving systemic administration for gastrointestinal cancer.¹⁸

Carniero et al provided some reason for pause with intravitreal therapy, showing lower levels of systemic VEGF following three monthly intravitreal injections of bevacizumab, compared to ranibizumab.¹⁹ However, in AMD patients, the CATT group was unable to find any statistically significant difference in systemic adverse outcomes between the two medications.¹⁴

Stein et al concluded that bevacizumab was cost-effective as long as the risk of stroke or heart attack was less than 4%, and ranibizumab was cost-effective if that rate was less than 2%.¹⁵

We must concede that AMD patients and diabetic patients may have different susceptibilities for cardiovascular complications, but Stein et al determined that the risk for major cardiovascular events would have to be greater than 1.5% more for bevacizumab, compared to ranibizumab, for ranibizumab to prove more cost-effective.¹⁵

COSTS ACROSS PATIENT GROUPS

William Smiddy, MD, also applied economic modeling to clinical trial populations with the added caveat of subdividing that population based on presenting VA and lens status. In patients with VA of 20/32 or better, the only treatment modality with published results on which to base cost-effectiveness analyses for DME was focal grid laser.

Other large clinical trials have excluded patients who see this well. Laser treatment costs $1,758 over one year, offering a 92% savings over anti-VEGF therapy in this subset of patients with good VA, assuming that this subpopulation incurs costs equivalent to the average of the entire patient population. Pseudophakic patients avoided a very common complication of intravitreal steroid therapy: cataracts.²⁰

Dewan et al confirmed this unique setting for the advantageous use of intravitreal triamcinolone acetonide.²¹ With this consideration in mind, intravitreal triamcinolone (Kenalog, Bristol–Myers Squibb, New York, NY), with or without laser treatment, offered 85% to 88% greater savings over ranibizumab therapy. For phakic patients with poor presenting VA, using bevacizumab over ranibizumab could provide 85% savings.²⁰

While both medications were treated as therapeutic equivalents in this analysis, patients treated with bevacizumab required slightly more frequent injections, compared to ranibizumab.

Michaelides et al administered an average of nine bevacizumab injections in one year to treat macular edema,²² and patients treated with ranibizumab received an average of 9.3 injections over two years.²³ The opportunity cost associated with increased clinic visits was not included in the cost calculation.²⁴

EVOLVING CALCULATIONS

Aflibercept (Eylea, Regeneron, Tarrytown, NY), a VEGF-trap medication, is one of the newest treatment modalities on the market. Insufficient data yet exist in the literature to allow for similar cost-effectiveness modeling as discussed above.

However, the published data show that the medication is effective at treating DME and is well tolerated, with no increased major systemic side effects, compared to laser photocoagulation.²⁵ As more clinical data become available, a QALY-based analysis of this medication will allow for comparison to bevacizumab and ranibizumab.

Physicians and patients have multiple appropriate treatment modalities to offer patients with diabetic eye disease. Perhaps most important, but most difficult to obtain, is better glycemic control. The CDC estimates that proper blood sugar control could prevent approximately half of vision loss due to diabetes.⁵ According to a 2002 estimate, intensive management of blood sugar cost $4,500 per patient per year,²⁶ and we can assume the cost will be similar or greater in 2014.

In some cases, one therapy is most likely to deliver the greatest clinical benefit to the patient, but in others, the only difference borne out in clinical studies may be the cost of therapy.

Such a situation involves the physician in the discussion of value judgments. A commonly accepted target for cost-effectiveness is $50,000/QALY,²⁷ mirrored by £30,000 in UK-based studies. However, several practices have far exceeded this cutoff: mammography in women younger than 50 and PET scans for lung cancer diagnosis can cost up to $200,000/QALY.^28,29

THE RETINAL PHYSICIAN’S ROLE

While recent political rhetoric has pushed for more responsible health-care spending, the Patient Protection and Affordable Care Act prohibits the use of QALYs in making coverage decisions.³⁰ This seeming contradiction forces health-care providers to aim at an amorphous and moving target for cost control. To engage in cost cutting with no defined budgetary goal could have disastrous consequences for patient health and for the patient-physician relationship.

The cost-effectiveness literature does answer some questions, but some of the questions it raises are intricate, thorny, and compelling. Paramount is who should make the decision of which treatment to select. Treatment selection can be a difficult enough decision when the only cost-benefit analysis pertains to patient health and well-being. Adding economics to the picture adds a whole new layer of complexity.

As physicians, we are charged to explain the options available to the patient and to serve as consultants, helping the patient to weigh the pros and cons of those choices. Of course, when it comes to more recent treatment developments, we lack the long-term evidence to provide reliable information.

Very few patients make therapy decisions based on cost alone, largely because most patients do not pay directly for the cost of the care that they receive. If presented with two options with outcomes that differ by even just one letter, few if any patients would choose the option with marginally worse results, regardless of the cost differential, although such a small difference is not clinically relevant and the individual would not perceived it.

Patients, and for that matter most healthcare providers, do not acknowledge or address that health care is both a scarce and inelastic good. Patients are not choosing whether they want a steak or a hamburger but whether they want to see the next line down on the eye chart. One would be hard-pressed to find a patient or physician willing to accept poorer vision to save a third-party payer a few dollars.

CONCLUSION

A need exists for continued cost-effectiveness research, especially in ophthalmology, with so many burgeoning new therapeutic and diagnostic modalities. It is important that physicians be active participants in discussions of how to maintain an economically viable healthcare system, not only out of self-preservation, but even more so as advocates for patients.

Decisions regarding patient health are better made by patients with the assistance of physicians and not in a boardroom numb to the individual nuance of each patient. RP

REFERENCES