Roche (Genentech)’s Lucentis receives FDA approval for DME – August 17, 2012

Roche announced last week that FDA approved the 0.3 mg dose of Roche (Genentech)’s Lucentis (intravitreally injected ranibizumab) for the treatment of diabetic macular edema (DME). This is the first FDA approved pharmaceutical treatment for DME; the current standard of care is laser therapy (and has been for the last 25 years), though some clinicians have been treating DME off-label with Avastin (bevacizumab, an anti-VEGF cancer therapy structurally similar to, but far cheaper than, Lucentis and also owned by Roche [Genentech]) or with off-label Lucentis (which had already been approved for wet age-related macular edema [wAMD] and macular edema following retinal vein occlusion [RVO] in the US). Notably, on July 26, 2012, an FDA advisory committee voted unanimously to recommend the 0.3 mg dose of Lucentis for DME and 8-2 for the 0.5 mg dose based on data from two phase 3 trials, RISE (n=377) and RIDE (n=382). Surprisingly, the 0.3 mg dose performed even better than the 0.5 mg dose in RISE. In general, the ocular safety profile of Lucentis in RISE/RIDE was found to be similar to its safety profile in wAMD and RVO; no evidence of increased arteriothromboembolic events (ATEs) was observed (in contrast to previous studies that had suggested Lucentis might be associated with increased risk of ATEs), but a slight increase in fatality was observed with the higher dose. Seeking approval for the 0.3 mg dose of Lucentis for DME may have been a strategic pricing decision on Genentech’s part given that it will be priced at $1170/monthly dose (compared the$1950/0.5 mg monthly dose used in wAMD and RVO), which places its price below that ofBayer/Regeneron’s Eylea ($1850/dose) and at a price that will be more palatable, presumably, to payors. That said, $14,000/year for treatment is quite expensive and well above the cost of a year’s worth of most branded oral or injected (but not in the eye) drugs, for example, which would be closer to$2,500 - $4,000/year.

We are curious to see how patients and physicians currently using Avastin off-label to treat DME will respond to Lucentis’s FDA approval – the cost of Lucentis has been highly prohibitive in the past for DME patients seeking an alternative to laser therapy, and ideally Lucentis’s FDA approval will improve patient access to this valuable DME treatment through insurance coverage. With regard to Genentech’s Lucentis/Avastin debacle, we are continually fascinated and perplexed by the dilemma that both Genentech and physicians face. In our view, doctors will be less likely to use Avastin going forward since there is now an approved drug, Lucentis, and the legal liability in the event of a safety problem with Avastin may not be worth it to many doctors (remote though the risk may be). It has always been the case that a physician’s job is to present patients with the best available treatment options, not to make prescribing decisions based on cost to society - physicians now have one approved option and one unapproved, less expensive option, and it will be interesting to see directionally how physicians move forward. Hopefully Lucentis’s DME approval provides a bit more clarity for physicians for now, while the upcoming three-way head-to-head trial comparing Lucentis, Avastin, and Eylea in DME will finally put cost-effectiveness questions (although not necessarily all safety questions) to rest so that the DME field can move forward to innovate even better options for patients.

• As a reminder, Lucentis is a VEGF-A antibody that had previously been approved for wet age-related macular degeneration (wAMD) and macular edema following retinal vein occlusion (RVO). Roche (Genentech) holds rights to the drugs in the US and Novartis ex-US. The 0.5 mg dose of Lucentis was approved for the treatment of wAMD in 2006 and for RVO in 2010, but has been losing market share in the wAMD field to Regeneron/Bayer’s Eylea (intravitreal aflibercept, another VEGF-A inhibitor). Presumably, Roche hopes that the DME indication for Lucentis will help the drug regain traction after several quarters of slowed growth. The prevalence of DME in the US is > 560,000, and the 2012 Vision Problems in the US report states that there are nearly eight million cases of diabetic retinopathy in the US >40 population – this represents a striking 89% increase since 2000. We are not certain what has driven the increase and believe that it must in part be due to better diagnosis rates.
• In 2Q12, Roche-recognized Lucentis sales totaled 354 million CHF (~$378 million), representing Lucentis’ first quarter of negative operational growth in at least five quarters (growth rates have steadily deteriorated each quarter from 35% in 1Q11, to 29% in 2Q11, 17% in 3Q11, 13% in 4Q11, and flat in 1Q12), but of course we cannot determine the subsets of revenue for RVO vs. wAMD. Outside of the US, Lucentis is widely approved for wAMD and has been approved for diabetic macular edema (DME) and RVO in Europe and Canada. Ex-US Novartis-recognized Lucentis sales grew 20% year-on-year in 2Q12 with Novartis estimating that new launches in DME and RVO accounted for 15-20% of sales in reimbursed markets.
• On July 26, 2012, an FDA advisory committee voted unanimously to recommend the mg dose of Lucentis for DME and 8-2 for the 0.5 mg dose based on data from two phase 3 trials, RISE (n=377) and RIDE (n=382). Genentech presented data demonstrating that monthly Lucentis treatment conferred a greater likelihood of achieving a 15- letter gain (three lines on an eye chart) in best-corrected visual acuity (BCVA) than sham injection. In both of these studies, patients stratified by baseline state were randomized in a 1:1:1 ratio to receive monthly sham injection, 0.3 mg Lucentis, or 0.5 mg Lucentis for 24 months. Both doses conferred significant vision gains as early as seven days after the first injection. Correspondingly, after 24 months, more patients receiving Lucentis gained >15 letters in BCVA than patients receiving sham injections (primary endpoint; p<0.001 for either dose vs. sham), and mean vision gain for patients receiving Lucentis was greater than that for control patients (see table below; p<0.0001 for either dose vs. sham). Surprisingly, the 0.3 mg dose performed even better than the 0.5 mg dose in RISE. The study authors acknowledge that an even stronger study might have attempted to demonstrate superiority of Lucentis over laser treatment (the previous standard of care for DME prior to Lucentis approval), but noted that Novartis’s 2011 RESTORE study and a 2010 study by the DRCR.net demonstrated superiority of Lucentis to laser over 12 months with regard to visual acuity outcomes for DME patients.

• In general, the ocular safety profile of Lucentis in RISE/RIDE was found to be similar to its safety profile in wAMD and RVO; no evidence of increased arteriothromboembolic events (ATEs) was observed (which previous studies had suggested might be a safety concern for Lucentis), but a slight increase in fatality was observed. The safety profile of Lucentis can be divided into three categories: 1) ocular safety events (relating the eye that receives Lucentis injections); 2) events potentially related to systemic exposure of a VEGF inhibitor; and 3) manufacturing. The most common ocular severe adverse event (SAE) affecting study eyes was vitreous hemorrhage, which did not occur more in Lucentis-treated eyes than in sham-treated eyes. Increased intraocular pressure after injection was more common in the Lucentis arms than control arms, but that was not surprising given that sham-injections did not contain anything. Overall rates of systemic SAEs were similar between treatment groups. However, according to Genentech’s slides presented at the advisory committee meeting, pooled strokes and deaths from any cause were slightly more common in the 0.5 mg Lucentis groups than the control groups (deaths, 1.2% of sham and 4.4% of 0.5 mg Lucentis- treated patients; strokes, 1.6% of sham and 3.2% of the 0.5 mg of Lucentis-treated patients), but still low. Similar increases in deaths and strokes were not observed in the RESTORE or DRCR.net studies comparing 0.5 mg Lucentis to laser or laser + sham in DME patients. Presumably, the lack of clinical benefit of the 0.5 mg dose over the 0.3 mg dose resulted in the approval of only the 0.3 mg dose. As we understand it, RISE and RIDE were not large enough to statistically establish the safety of one Lucentis dose over another, but no glaring safety signals were detected for the 0.5 mg dose. Given that diabetes patients are already at elevated CV risk, the concern for any systemic exposure to anti-VEGF treatment was, presumably, enough to warrant exclusion of the 0.5 mg dose. The FDA has reviewed three-year data from RISE/RIDE, and those results will be publically released Monday, August 27, 2012 when they are presented at a major eye meeting in Las Vegas. On the manufacturing front, some have said that there have been problems with manufacturing of Avastin for the eye and that compounding pharmacies have created problems – see below for more details on this front.
• Seeking approval for the 0.3 mg dose of Lucentis for DME may have been a strategic pricing decision on Genentech’s part given that it will be priced at $1170/monthly dose (compared the $1950/0.5 mg monthly dose used in wAMD and RVO). The$1170/dose price places 0.3 mg Lucentis below the $1850/dose price of Bayer/Regeneron’s Eylea (intravitreal aflibercept, another VEGF-A inhibitor, currently approved for wAMD). Eylea is currently in phase 3 trials for DME (see below), though for wAMD it is prescribed as a monthly injection for the first three months and subsequently every two months, whereas Lucentis is prescribed as a monthly injection indefinitely. Therefore, one year of Eylea, assuming the DME dosing were to remain the same as wAMD dosing, would still cost slightly less than one year of Lucentis treatment for DME ($12,950 vs. $14,040). Eylea is the closest competitor for Lucentis to market in the US, and either choice is a relatively pricey for a diabetes treatment, though blindness is certainly difficult to price.
• We are curious to see how patients and physicians currently using Avastin off-label to treat DME will respond to Lucentis’s FDA approval. The cost of Lucentis has been highly prohibitive in the past for DME patients seeking an alternative to laser therapy, so hopefully Lucentis’s FDA approval will improve patient access to this valuable DME treatment through insurance coverage. However, given that Avastin’s lower price ($50/0.5 mg dose) has been a driving factor in off-label Avastin use for ocular conditions, we believe that many physicians may still prescribe it for patients whose insurance does not cover Lucentis or forpatients who do not have insurance (in this case, Avastin is seen as far better than nothing). We wonder, however, if concerns over legal liability will be enough to drive more physicians to switch to Lucentis. Dr. David Brown, a lead investigator of the RISE/RIDE trials and the CATT study, asserted in a recent conversation with us, “ … If it weren’t for the price, I don’t think anyone would recommend use of Avastin…” now that Lucentis has an FDA approval for DME and Avastin does not – “Avastin just was not meant to be put in the eye!” Dr. Brown expressed concern that some insurance plans would still tacitly pressure physicians to prescribe Avastin over Lucentis for DME due to incentives to minimize cost for care (under such directives, a physician might be indirectly penalized for switching from Avastin to Lucentis if he/she achieves the same outcome for a greater price). As a reminder, Avastin is an anti-VEGF antibody approved for the treatment of various cancers and is also manufactured by Genentech.
  • The two-year CATT study demonstrated that Lucentis and Avastin had similar efficacy in the treatment of wAMD, but Avastin use was associated with slightly higher safety risks. Due to differences in disease state and patient population, these results could not be directly applied to DME, but more clarity will be provided from the much-anticipated upcoming study led by the DRCR.net and sponsored by Genentech, Regeneron, and the National Eye Institute (slated to begin August 2012) examining the comparative efficacy of Eylea, Avastin, and Lucentis in DME (clinicaltrials.gov identifier: NCT01627249).
  • One safety concern over Avastin is its longer systemic half-life. As a colorectal cancer drug, Avastin was designed to act more systemically than Lucentis. When either drug is injected intraocularly, some small amount will escape into systemic circulation, but while Lucentis’s systemic half-life is about two hours, Avastin’s is ~20 days.
  • Another safety concern with Avastin is that the FDA does not have full oversight over compounding pharmacies, where Avastin is re-packaged into small doses for off-label intraocular use. When Avastin is administered as a cancer treatment, doses range from 5 mg/kg to 15 mg/kg, and so the drug is packaged as either a 100 mg/4 ml or 400 mg/16 ml single use vial. Thus, each vial is divided into at least 200 parts (the 100mg/4 ml vial were split into 0.5 mg doses) for re-distribution. In 2011, at least 17 cases of endophthalmitis resulting in vision loss were traced back to contamination of Avastin re-packaged at compounding pharmacies. Full regulation of compounding pharmacies is a state function (not an FDA function), and we were extremely surprised to learn that some states do not even require compounding pharmacists to wear a mask.
• As a reminder, various other potential DME treatments are in late-stage clinical development. Bayer/Regeneron’s Eylea (intravitreal aflibercept, another VEGF-A inhibitor) is in three phase 3 studies with primary completion dates of May, September, and December 2013, suggesting that the earliest Eylea could be approved for DME would be in 2014. As we understand it, three-year data will ultimately be required for approval, but Regeneron may attempt submission to the FDA with the two-year data. Alimera Sciences/pSividia’s Iluvien, an implantable device that releases a corticosteroid received an FDA complete response letter in November citing safety and side-effect concerns; Alimera noted that addressing the FDA’s concerns would be financially prohibitive, and we have not heard any updates on the status of resubmission in the US. This is disheartening given that Iluvien would be a one-time implantation and much easier for patients. Genentech is also currently developing a long-acting device that would release Lucentis over time. The device remains in phase 1 trials, but we are eager to followits progress since a decreased injection frequency would certainly make the drug easier to take. For more details on recent updates in the DME field (including updates on Macugen, ActiveSite Pharmaceuticals’ plasma kallikrein inhibitor, and iCo Therapeutics/JDRF’s stage 2 candidate iCo-007) please see our Roche 2Q12 report here: http://www.closeconcerns.com/knowledgebase/r/ce2825be additionally, we wrote anindustry update in 2011 that has more historical information and related interviews – see that at http://www.closeconcerns.com/knowledgebase/r/ad1f4c9d.

-- by Jessica Dong, Kira Maker, and Kelly Close