CONTROVERSIES IN CARE

The CATT's Whiskers

Edited By Michael Colucciello, MD

Using Avastin for wet AMD is unethical…”

One of my respected colleagues was discussing the use of bevacizumab for wet AMD during a refreshment break four years ago at a retina conference. “I use Avastin frequently,” I told him.

“Michael, until a study is done with Avastin that is powered like MARINA and ANCHOR, which proved efficacy for ranibizumab for wet AMD, it is unethical to use Avastin in that setting based on uncontrolled studies,” he said.

I continued to use bevacizumab, as did most retina specialists, and Avastin became the most commonly used drug in the United States for the treatment of wet AMD.¹

Certainly, the MARINA and ANCHOR studies published in the fall of 2006 provided level 1 evidence for the usage of intravitreal Lucentis for wet AMD.^2,3 However, retina specialists have long suspected that far less costly intravitreal Avastin should work in a similar fashion as Lucentis: It has been shown that Avastin, the larger, full-length humanized monoclonal antibody binding VEGF, still crosses the entire retina, like the smaller, anti-VEGF antibody fragment Lucentis does.⁴ Lucentis was developed from the same anti-VEGF mouse monoclonal antibody as Avastin. Avastin binds VEGF at the same position as Lucentis. Since 2005, many case studies have demonstrated Avastin's efficacy, and retina specialists grew comfortable with its use.

The one-year CATT study results now affirm the usage of Avastin in the setting of wet AMD. These results showed that the monthly use of either Avastin or Lucentis results in about the same outcomes in terms of visual acuity. Monthly and as-needed (with monthly exams) regimens of Lucentis performed equally well, as did as-needed Avastin compared to as-needed Lucentis. Both Avastin and Lucentis provided substantial improvement in 25% to 34% of affected eyes; more than 90% of affected eyes avoided any further loss of vision. A secondary outcome, OCT retinal thickness measurement, favored Lucentis.⁵ Second-year CATT data should help us understand whether the OCT differences will eventually translate into visual acuity differences.

CATT patients given Avastin had more systemic adverse events and hospitalizations than those receiving Lucentis; these events were not associated with organ systems usually identified with anti-VEGF therapy, and the CATT study was insufficiently powered to disclose differences in drug-related adverse systemic events.

A Medicare claims database (n= 77,886) analysis study presented as a poster at the 2011 ARVO meeting (supported by Genentech) showed an 11% higher mortality risk and a 57% higher hemorrhagic stroke risk in the Avastin group; however, the study was limited by incomplete information regarding smoking and lipid and blood pressure levels in the patients analyzed.⁶

The bottom line is that we now have good one-year scientific data on Lucentis/Avastin comparisons. In a responsible fashion, we as retina specialists will now have to substantiate the use of Lucentis, given the vast cost difference between it and Avastin.

Anthony Adamis, MD, global head of ophthalmology at Genentech, Inc., and Hal Barron, MD, executive vice president and chief medical officer at Genentech, have been invited to give their perspective on the results of the CATT trial. A CATT study principal investigator, Pravin Dugel, MD, has also been invited to give his perspective. Their responses follow.

REFERENCES

1. Brechner RJ, Rosenfeld PJ, Babish JD, et al. Pharmacotherapy for neovascular age-related macular degeneration: an analysis of the 100% 2008 Medicare fee for-service Part B claims file. Am J Ophthalmol. 2011;151:887-895.
2. Rosenfeld PJ, Brown DM, Heier JS, et al. MARINA Study Group. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med. 2006;355:1419-1431.
3. Brown DM, Kaiser PK, Michels M, et al. ANCHOR Study Group. Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med. 2006;355:1432-1444.
4. Shahar J, Avery RL, Heilweil G, et al. Electrophysiologic and retinal penetration studies following intravitreal injection of bevacizumab (Avastin) Retina. 2006;26:262-269.
5. Martin D, Maguire M, Ying G, et al. The CATT Research Group. Ranibizumab and bevacizumab for neovascular age-related macular degeneration. N Engl J Med. 2011 Apr 28 [Epub ahead of print].
6. Gower EW, Cassard S, Chu L, et al. Adverse event sates following intravitreal injection of Avastin or Lucentis for treating age-related macular degeneration. Paper presented at: Annual Meeting of the Association for Research in Vision and Ophthalmology; Fort Lauderdale, FL; May 3, 2011.

For AMD Patients, Lucentis Is Safer

Anthony P. Adamis, MD, Hal V. Barron, MD, Genentech, INC.

Genentech congratulates the CATT investigators for providing important information that will enable informed treatment decisions by physicians and patients. As we examine the emerging safety data, we continue to believe that Lucentis is the most appropriate treatment for wet AMD.

Lucentis was specifically designed for intraocular use. It lacks the Fc region of monoclonal antibodies and fusion proteins, and therefore it is cleared from the systemic circulation more quickly.¹ The short systemic half-life of Lucentis is particularly important when considering the older patient populations being treated. Systemic VEGF inhibition can result in serious adverse events (SAEs), which include bleeding, thrombosis, myocardial infarction, stroke and death. When given intravitreally (IVT), systemic Lucentis levels can reach more than 10 times normal circulating VEGF levels.^2,3 Thus, Genentech continuously monitors both the ocular and systemic safety of Lucentis. For example, when an interim safety analysis revealed a higher incidence of stroke with Lucentis, Genentech promptly informed healthcare providers,⁴ updated the label⁵ and presented the data to the scientific community.⁶

A growing body of data suggests that IVT Avastin may pose greater risks than Lucentis. Clinical trials are not typically powered to detect differences in rare SAEs. To do so would require a study with tens of thousands of patients. However, two large retrospective Medicare claims analyses recently studied the requisite number of patients. The first study, from Duke University, demonstrated a statistically significantly higher risk of death and stroke with the use of Avastin vs Lucentis in AMD.⁷ When an analysis was performed to control for potential socioeconomic confounders, statistical significance was lost, perhaps due to the smaller sample size. A second study from Johns Hopkins (Genentech-supported) also showed a statistically significant 11% increased risk of death and a 57% increased risk of hemorrhagic stroke.⁸ In this study, the numbers were greater and significance was maintained after adjusting for confounders.

Although not powered to detect rare SAEs, CATT nonetheless found a 29% increased risk for systemic SAEs with Avastin (P=.04), with over 80% of patients having SAEs requiring hospitalization. There were also more deaths with Avastin (15 vs nine). Our analysis of the safety data⁹ focused on SAEs potentially attributable to VEGF inhibition. Utilizing the same adverse-event classification system used in oncology, our exploratory analysis suggests that roughly one-third of the systemic SAE difference between the drugs was due to events associated with VEGF inhibition. The imbalance in SAEs associated with VEGF inhibition was greatest in the monthly arms (21 vs 11), in which the exposure was greatest.

Biological plausibility for these safety findings exists, as studies have suggested that IVT Avastin may suppress systemic VEGF levels for up to 28 days,^10,11 whereas Lucentis does not.¹¹ Recently, the Rotterdam study found a six-fold increase in the risk of a hemorrhagic cardiovascular accidents in advanced AMD patients vs controls.¹² This finding suggests that wet AMD may pose a stroke risk, independent of systemic VEGF inhibition. We speculate that systemic VEGF inhibition may be additive to the AMD-associated stroke risk.

While the above data are not yet conclusive, the combined evidence indicates that the safety profiles of Avastin and Lucentis differ and that the use of IVT Avastin is likely associated with an increased risk of systemic side effects.

REFERENCES

1. Ferrara N, Damico L, Shams N, Lowman H, Kim R. Development of ranibizumab, an anti-vascular endothelial growth factor antigen-binding fragment, as therapy for neovascular age-related macular degeneration. Retina. 2006;26:859-870.
2. Larsson A, Skoldenberg E, Ericson H. Serum and plasma levels of FGF-2 and VEGF in healthy blood donors. Angiogenesis. 2002;5:107-110.
3. Xu L, Lu T, Tuomi L, Jumbe N, Lu J, Eppler S, Kuebler P, Damico-Beyer L, Joshi A. Pharmakokinetics of ranibizumab in patients with neovascular age-related macular degeneration—a population approach. Manuscript submitted.
4. Barron H. Dear health care provider letter. Genetech Web site. at http://www.gene.com/gene/products/information/pdf/lucentis_sailor_letter.pdf. Accessed June 21, 2011.
5. Lucentis US product label. Genentech Web site. http://www.gene.com/gene/products/information/pdf/lucentis-prescribing.pdf. Accessed June 21, 2011.
6. Boyer D, Chung CY, Rundle AC, Tuomi L. A safety overview of ranibizumab in patients with wet AMD: ANCHOR, MARINA, PIER and SAILOR studies. Paper presented at: Annual Meeting of the American Academy of Ophthalmology; Atlanta, GA; November 8-11, 2008.
7. Curtis L, Hammill B, Schulman K, Cousins SW. Risks of mortality, myocardial infarction, bleeding and stroke associated with therapies for age-related macular degeneration. Arch Ophthalmol. 2010;128:1273-1279.
8. Gower EW, Cassard S, Chu L, et al. Adverse event sates following intravitreal injection of Avastin or Lucentis for treating age-related macular degeneration. Paper presented at: Annual Meeting of the Association for Research in Vision and Ophthalmology; Fort Lauderdale, FL; May 3, 2011.
9. Martin D, Maguire M, Ying G, et al. The CATT Research Group. Ranibizumab and bevacizumab for neovascular age-related macular degeneration. N Engl J Med. 2011 Apr 28 [Epub ahead of print].
10. Matsuyama K, Ogata N, Matsuoka M, Waka M, Takahashi K, Nishimura T. Plasma levels of vascular endothelial growth factor and pigment epithelium-derived factor before and after intravitreal injection of bevacizumab. Br J Ophthalmol. 2010; 94:1215-1218.
11. Barros-Pereira R, Costa R, Falcao M, et al. Plasma levels of vascular endothelial growth factor before and after intravitreal injection of bevacizumab or ranibizumab in the treatment of neovascular age-related macular degeneration. Paper presented at: Annual Meeting of the Association for Research in Vision and Ophthalmology; Fort Lauderdale, FL; April 29-May 5, 2011.
12. Wieberdink RG. Age-related macular degeneration and the risk of stroke: The Rotterdam StudyPaper presented at: American Stroke Association International Stroke Conference; Los Angeles, CA; February 9, 2011.

Too Soon to Tell

Pravin U. Dugel, MD

I will confine my comments only to the CATT study, as I do not believe any retrospective, data-based study (biased by uncontrolled, confoundable variables) can be compared to a well-controlled, prospective, randomized, multicenter study. The level of evidence in the latter (considered definitive studies) far exceeds the former (considered hypothesis-building studies).

CATT, to me, is the definitive one-year study showing that Avastin is noninferior to Lucentis, both in efficacy and in safety. Note, however, that wet AMD is not a one-year disease. It is a chronic disease, with most patients living more than a dozen years after diagnosis. This is an important distinction: although, in my opinion, the CATT study has recently provided us with definitive one-year results, definitive long-term AMD study results are forthcoming and not yet available.

The CATT study is a noninferiority study with a five-letter noninferiority limit and 99.2% confidence interval. The five-letter limit was chosen as it was less than the difference observed for previously successful treatments, such as thermal laser, PDT and Macugen. Typically in drug studies, a 95% confidence interval is chosen. However, in the CATT study, a self-imposed higher bar of 99.2% was chosen.

The results show noninferiority in the vision outcome. Surprisingly, the as-needed arms were only approximately two letters less than the monthly arms. These are the best results for as-needed arms reported in any trial. However, the retreatment criteria were very strict, and patients received more injections than in any other as-needed protocol. It is important not to misinterpret these results. It might be tempting to conclude that as-needed therapy is just about as good as monthly injections. This was not a treat-and-extend comparison (this analysis will be forthcoming in the Norwegian LUCAS trial, led by Karina Berg). In the CATT study, patients were seen strictly on a monthly basis, even in the as-needed arm. The results cannot be extrapolated to treat-and-extend protocols.

Much has been made of the OCT and safety results. Only time-domain OCT results were analyzed in the first year. The OCT results showed that patients in the monthly Lucentis group had on average a 30-µm thinner retina. Clearly, this result did not translate into a functional advantage. Many have opined that this small difference will prove important in the two-year results. That may or may not be true. Note that the other arms had foveal thickness closer to normal foveal thickness. The opposite argument could also be made. At this point, any extrapolation of this finding in either direction is purely speculative.

This study was not powered primarily to assess safety. However, because safety is obviously of paramount importance, it is critical to recognize that the CATT study extended its evaluation to two-year data. Historically, all anti-VEGF studies have focused on deaths and cardiovascular events for sound physiologic reasons. In these two categories, there was no difference between the two drugs. There was a difference in serious adverse events (SAEs) in the Avastin as-needed arm. This finding has created controversy—partly by misunderstanding and, unfortunately, partly by misinformation.

In most studies, the total SAE bucket is not reported. I commend the CATT leaders for being totally transparent (the entire patient information in several hundred pages is available online). Nonetheless, the SAEs must be properly understood. Any and all adverse events are included in this bucket. For instance, if a patient develops sepsis, breaks an ankle or has congestive heart failure, it is included here. In other words, there is much room for “statistical noise” here.

Is it really just statistical noise? Consider that, in general oncology, doses that are magnitudes greater are given, without any concerns regarding the recorded SAEs. Consider that the increased in SAEs was driven by the Avastin as-needed group and not the monthly group, as would be expected from more frequent administration of the drug. Consider that, by chance (despite randomization), the patients in the Avastin as-needed group were older and sicker. The bottom line is, as with the OCT results, we do not know what this SAE difference means. We will have clarity in the upcoming years. An alarm need not be sounded. These data must be understood in their proper context.

The CATT study thankfully will continue to provide us with valuable information in its second year by answering the following questions:

1. Do visual acuity findings at one year remain through year two?

2. Do the anatomical differences noted in year one alter long-term visual outcomes?

3. What is the effect of PRN dosing after one year of monthly dosing?

4. Will OCT result in more fluid detection and subsequent injections?

5. Are there genetic findings that predict outcomes, durability of response or number of injections required?

The CATT study leaders ought to be congratulated for conducting the best designed, best executed and most transparent study in which I have had the privilege to participate. However, it must be stated that CATT does not address our most formidable challenge: treatment burden. All patients in all randomized arms were seen monthly. With Lucentis or Avastin, our overburdened treatment model makes anti-VEGF monotherapy simply unsustainable. Urgent work is needed to find a sustainable treatment model.

In this chronic disease, more follow-up is needed to have definitive long-term results. Until then, retinal physicians should have access to both drugs, without regulatory restrictions. Additionally, we all must clearly understand this study and not be misled by misinformation. We owe this to our patients. RP