An Eye for an Eye: Lucentis and the New Pharmaceutical Value System

A head-to-head trial of Genentech’s Avastin v. Lucentis raises the stakes in the debate over drug pricing. While previous large comparative trials have focused on a number of different drugs—old against new—the Avastin study is evaluating two brand new cutting edge technologies, one for off-label use. Genentech is not alone in finding this unsettling.

By Ramsey Baghdadi

Genentech Inc. did everything right. The company spent hundreds of millions of dollars and a decade developing the vascular endothelial growth factor inhibitor ranibizumab (Lucentis) for age-related macular degeneration.

"The Lucentis clinical trials were the most expensive clinical trials we ever conducted," Genentech spokesperson Dawn Kalmar says. That’s a remarkable statement considering Genentech is viewed primarily as a cancer company.

The big bet certainly seemed to pay off. Although competitor Eyetech Pharmaceuticals, now a division of OSI Pharmaceuticals Inc. [See Deal] was about a year and a half ahead in the race to develop a new treatment for AMD with its own anti-VEGF inhibitor pegaptinib (Macugen), Lucentis was heralded as the next great breakthrough in combating the blinding disease that affects 165,000 new patients each year. The Food & Drug Administration fast-tracked the review of Genentech’s application and Lucentis was approved in June 2006.

Once on the market, Lucentis was a stunning success—and Macugen was an equally stunning failure. After just a few months competing against Genentech, OSI decided to divest its eye business in November 2006. The biotech company now accounts for losses resulting from Macugen under the subhead of "discontinued operations" in its quarterly earnings reports.

Genentech should be beaming after destroying its only competitor in the AMD space and looking forward to years of billion-dollar blockbuster status. (Sales of Lucentis are on track to reach $850 million by the end of this year.)

Instead, Genentech is worried.

Lucentis generated $211 million in sales for the first quarter of the year—not too shabby for a drug that isn’t even in its second year on the market. But that’s a 3% decline compared to the fourth quarter of 2006 and a sharp drop compared to the skyrocketing rate at which the drug was growing right after launch.

What happened? Genentech made one critical mistake: they priced Lucentis too high. Now, they’re paying for it. In an ironic twist, Lucentis obliterated its one FDA-approved competitor, but is struggling to hold its own against a product that isn’t even marketed for use in AMD—in fact, it’s against the law to do so. The stealth competitor? Genentech’s own drug for colorectal cancer, the VEGF inhibitor bevacizumab (Avastin).

Avastin isn’t approved for AMD, but physicians started using miniscule doses of bevacizumab as a less-expensive alternative to Lucentis, and now the older drug is broadly used off-label. And, thanks to unprecedented pressure from physicians, the National Eye Institute is running a head-to-head study that may help make Avastin the standard of care—no matter what Genentech thinks.

Genentech’s Avastin/Lucentis conundrum is a lesson to drug makers that you can execute almost flawlessly throughout the drug development process, but one critical mistake on the commercial side—bold pricing by even today’s standards—can derail a potential blockbuster.

The Avastin vs. Lucentis debate is a nightmare for Genentech, but is also a harbinger of things to come for the rest of the biotech and pharmaceutical industry. Payors are increasingly pushing for more head-to-head studies to determine the value of medicines—and, when they can set the agenda, you can bet they will push for studies to justify choosing the cheaper alternative. The AMD therapies boil the issue down to its simplest form: two novel drugs, one expensive, one not, both of which appear to treat a debilitating disease equally well.

Genentech is not alone in worrying about the implications. The circumstances affecting Lucentis are unusual to say the least, but the intervention of the federal government to run a head-to-head trial is crystallizing a new focus on comparative cost effectiveness across the industry.

Comparative effectiveness was once a niche topic reserved for policy wonks and think tanks, but now emissaries from the largest biopharma companies are being sent out to get up to speed on the different proposals taking shape. "This is our number one issue," according to an official from one of the top three pharmaceutical companies.

Biotech companies are also feeling the pressure. "If you were to ask government affairs people in Washington: in 2007 how much time would you be spending on the issue of comparative cost effectiveness, most all of them would have said, ‘probably not much,’" Amgen Inc. executive director for global government affairs Andrew Swire said at the Biotechnology Industry Organization annual meeting in Boston. "The Democrats have shown there’s a real strong inducement to get something done, possibly this year."

For now, comparative effectiveness efforts are largely fragmented among a number of small organizations and academic institutions. But legislation floating through Congress would establish a centralized comparative effectiveness center. Similar ideas have been proposed by Blue Cross Blue Shield and Project Hope.

There are a number of ways to evaluate comparative effectiveness, including literature reviews and claims data analysis. However, nothing sends shivers through pharmaceutical companies like the idea of third-party-sponsored head-to-head clinical trial pitting their biggest brand against a cheaper competitor. And that’s where the trend seems headed.

Is Lucentis 20 Times Better than Avastin?

For a payor, the case against Lucentis is pretty cut and dry. A monthly dose of the drug costs roughly $2,000 per injection. Genentech recommends four initial monthly doses, and the possibility of less frequent quarterly dosing afterward; however, the company cautions the benefits of taking the drug could be reduced with fewer injections.

So over the 24-month recommended therapy schedule, Lucentis can cost between $30,000 and $50,000 per patient. But ophthalmologists concerned by the perceived exorbitant cost of the drug have been using pharmacy-compounded, off-label micro doses of Avastin to treat the same form of AMD for which Lucentis is labeled. Avastin is itself notoriously expensive as a cancer therapy, but because such tiny doses are used for ophthalmic use, it only costs about $100 an injection, or between $1,000 and $2,400 during the two-year period.

Nobody questions the value of Lucentis as a medical breakthrough. (See Exhibit 1.) The agent was successful in halting the progression of blindness in 95% of patients taking it compared to placebo. Approximately 40% of patients improved their vision using the biologic.

So the decision to widely use Avastin off-label really comes down to the difference between price tags. Decision makers—whether they are physicians, patients, insurers, or the US government—question whether Lucentis is roughly 20 times more valuable than Avastin for treating AMD.

And they have every reason to think the answer is no, given how closely related the two molecules are. The American Academy of Ophthalmology, for example, describes Lucentis as a fragment of the full monoclonal antibody that is marketed as Avastin.

Adding insult to injury for Genentech is the fact that if it weren’t for Lucentis, no one would have tried Avastin for AMD in the first place. "It is quite possible that the off-label use of Avastin would not have happened had it not been for...Lucentis because it was only after the positive results from the Lucentis Phase III clinical trial that physicians started using Avastin off-label," Genentech’s Kalmar says.

Off-label use was primarily driven by positive data announced in May 2005 from the first Phase III Lucentis study. Ophthalmologists, blown away by the data, immediately concluded that Lucentis was a better therapy than Macugen, the only FDA-approved anti-VEGF therapy at the time. However, there was a 12-month lag time between the unveiling of the Lucentis results and the June 2006 approval. So before Lucentis was even approved for use in the US, doctors had already decided to take matters into their own hands by using compounded Avastin.

"While we did go ahead and open up a 5,000-patient study [allowing access to Lucentis] to help bridge the gap, there were still a lot of patients that needed help, and that’s what drove the off-label use of Avastin," Kalmar says. Genentech claims it doesn’t know how many physicians are currently using Avastin off-label.

You can’t blame Genentech if they failed to anticipate that physicians would look at Avastin as a low-cost alternative. At the time the off-label use began, Genentech was being raked over the coals for what some saw as an outlandish price tag for Avastin in colorectal cancer. Voices opposing Genentech’s pricing practices became so loud, the firm chose to put a $55,000 cap on the annual cost of the drug when Avastin was approved as an adjunct therapy in lung cancer in October 2006. (See "Spending Caps: A Desperate Measure," The RPM Report, July 2006 (Also see "Spending Caps: A Desperate Measure?" - Pink Sheet, 1 Jul, 2006.).)

Indeed, it is the ophthalmologists who are responsible for bringing the off-label use of Avastin forward. Payors are eager to benefit now, but the first ophthalmologists to try Avastin for AMD encountered barriers to coverage, since off-label indications are not routinely covered by Medicare or private payors. So the American Academy of Ophthalmology opened a registry of off-label Avastin use, collecting more than 9,000 case reports to support coverage.

And now payors are very much on board. Barry Straube, the chief medical officer for the Centers for Medicare & Medicaid Services, told the Food & Drug Law Institute annual meeting in April that using Avastin off-label would save the health care system $1 billion to $3 billion per year. Of course, that will only happen, Straube noted, if Avastin can be shown to be as effective as Lucentis.

Genentech Refuses to Participate

That is what the National Eye Institute hopes to prove. The federal research agency is conducting a randomized, prospective, head-to-head clinical study of Lucentis vs. Avastin to compare safety and efficacy of the two agents. The study is called the Comparison of Age-Related Macular Degeneration Treatments Trials, or CATT. But a better acronym might be COST.

The study follows high-profile "old versus new" comparative effectiveness trials run by the National Institutes of Health, including the ALLHAT trial, which evaluated anti-hypertensive drugs, and the CATIE study on antipsychotics. (See "CATIE: Learning to Live With Large Comparative Trials," The RPM Report, December 2005 (Also see "CATIE: Learning to Live With large Comparative Trials" - Pink Sheet, 1 Dec, 2005.).)

Unlike ALLHAT and CATIE, which tested many different agents in broadly diagnosed indications, the CATT study is truly head-to-head. And it is hard to see any good outcome for Genentech. If the conclusion is that Avastin is just as good (or better) than Lucentis, payors should have all they need to refuse to pay more for Lucentis. If Lucentis is better, Genentech may still have a challenge in demonstrating that it is better enough to justify the 20-fold price premium.

Genentech says it is not participating in the planning, implementation or donation of drugs for the CATT study. That decision is a gamble from a public relations perspective since it allows Genentech to be painted as obstructionist while simultaneously limiting any opportunity for the company to manage the release of the findings from the trial.

Genentech says it isn’t supporting the trial because it isn’t designed to meet an unmet medical need. The company also says the study is not sufficiently powered to detect safety differences between the two drugs.

Genentech officials aren’t the only ones to raise the question of safety. In November 2005, before Lucentis was approved, the Medicare Coverage Advisory Committee convened to hear evidence on emerging AMD therapies. During the meeting, concerns over the safety of Avastin for use in the eye were a key point of discussion.

"The use of intravenous Avastin on label for colon cancer has a safety profile that’s not so good," CMS medical officer Ross Brechner, MD, said, noting that the cancer drug has a "black box" warning on the label for potentially fatal gastrointestinal perforations.

"Avastin might be okay for [cancer patients], but for us to make the assumption that because we are putting such a small amount of drug inside the eye when we’re using intravitreal Avastin, that there is no way that this could cause any serious side effects, I think this might be very misleading." Brechner recommended against evaluating Avastin in Medicare beneficiaries in any kind of safety trial.

Genentech director for vascular and ophthalmic medicine Charles Semba pointed out to the committee that the company had already done all of the work to develop a proven safe and effective agent for AMD.

"We are aware that physicians are interested in exploring the off-label use of Avastin," he said. "Avastin and Lucentis are different molecules designed for vastly different indications." Semba highlighted Lucentis’ more targeted engineering and that the drug had been followed in 1,400 patients for three years. "[Lucentis] was safe and well tolerated; the overall benefits outweigh any potential risks."

The Pros and Cons of Head-to-Head Trials

Genentech is not alone in questioning the role of the federal government—and especially the National Institutes of Health—in running large head-to-head trials. (See "CMS: The Clinical Trials Agency?" in this issue. (Also see "CMS: The Clinical Trials Agency? " - Pink Sheet, 1 Jun, 2007.))

Critics of CATIE and ALLHAT say the government didn’t get an adequate return on its investment. Long-term comparative trials are undeniably expensive to run: CATIE and ALLHAT cost the government $64 million and $128 million, for what resulted in little change in physician prescribing patterns after the data were released.

Such studies also cannot answer questions about the value of therapies launched after the trial is designed. CATIE, for example, didn’t include Bristol-Myers Squibb & Co.’s aripiprazole (Abilify).

And they produce vast amounts of data subject to many competing interpretations. Critics argue that both CATIE and ALLHAT were announced with overly simplistic messages about the benefits of off-patent medicines. Whether that analysis is fair, it is clear that neither study had a dramatic effect in terms of changing the practice of medicine.

Other researchers vigorously disagree. Large comparative trials may not generate black-and-white answers, they acknowledge, but at least they provide new data to inform clinical practice.

"Even if you don’t come up with an answer that fits in a headline that can never be debated again, I think there is huge value in having done the ALLHAT study and I would make the same argument about the CATIE study," says Sean Tunis, the former chief medical officer at the Centers for Medicare & Medicaid Services and founder of the Center for Medical Technology Policy.

"If you had done ALLHAT using a massive retrospective study from a large claims database, it would have been dismissed in a heartbeat. There is a very different level of seriousness" when you do a prospective clinical trial, Tunis said during an Avalere Health evidence-based medicine conference in Washington, D.C. May 22.

With the growing trend of head-to-head comparative trials, the question becomes: how will the data be used? Harvard researcher and director of the Institute for Comparative Economic Research Steve Pearson, MD, has developed a scoring system to standardize the flow of information that could be coming down the pike and turn it into a decision-making tool (See Exhibit 2).

For example, the system assigns an "A," "B," or "C" in descending order to determine comparative clinical effectiveness and an "a," "b," or "c" rating for comparative value. What’s more valuable, an "Ac" rated drug or a "Bb" rating? It’s unclear. And that ambiguity is where problems could arise.

Still, says Tunis, "the downward pressure on technology is going to come in one form or another, it’s just whether it’s going to be rational or irrational. We’re going to ration care one way or another; it’s a question of whether we’re going to be clinically sophisticated about it or not."

Industry Gets Ready

The advancement of comparative cost-effectiveness in policy circles has companies already rethinking their internal R&D. "Internally, as we begin to prepare for these challenges we will face in the foreseeable future, we have to realign our strategies to fulfill those needs," Roche Diagnostics director for government affairs John Ridge said at the BIO meeting. "We have to look beyond safe and efficacious."

Amgen’s Swire agrees. "The regulatory pathway approach now requires a much more robust evidence pathway beyond just the approval of a drug or biologic. It’s something you have to be involved in, probably as early as in Phase I studies."

There are real concerns that innovation will be severely stifled if the focus on comparative cost-effectiveness becomes too great. In the past, there was a high premium payment—five-fold or ten-fold increases—given to novel therapies just because they were new. The rationale by drug manufacturers is that the bonus goes back into R&D for the development of new products.

An alternate approach is if a drug is 10% or 20% more effective, the agent should be priced at a 10% or 20% markup. "That’s a lot less of a bonus for innovation to feedback into R&D and it also makes the whole proposition a whole lot less attractive," acknowledges Tunis. "That’s not an irrelevant problem."

Former CMS advisor and Memorial Sloan-Kettering Cancer Center researcher Peter Bach, MD, argues that in order for drug makers to justify the prices of their drugs going forward, comparative effectiveness will have to be patient-centered.

"If you’re going to spend X amount of dollars for a drug, it has to give the patient something," he said during the BIO meeting. In cancer, for instance, average prolongation of survival or average prolongation of quality of life are patient-focused clinical endpoints that warrant bigger price increases—not a surrogate endpoint like progression-free survival, Bach says. (See "Is the Oncology Bubble Ready to Pop?," The RPM Report, March 2006 (Also see "Is the Oncology Bubble Ready to Pop?" - Pink Sheet, 1 Mar, 2006.).)

And that brings us back to Lucentis. The drug delivered on its promises of safety and breakthrough efficacy for patients with few options compared to other therapies already on the market. But Genentech overreached on price.

The company, however, is sticking to its guns. CEO Arthur Levinson was unapologetic in a June interview with the Wall Street Journal and that message has trickled down in the company. "We believe the price reflects the years of clinical study and the value that it brings to patients," Kalmar says. "It’s inappropriate to make comparisons of the price of Lucentis to the price of Avastin."