Optical Illusion: Why Genentech Is Winning the Fight Over Lucentis

With Lucentis sales slumping, Genentech took the bold step of cutting off ophthalmologists from supply of Avastin, further angering providers already seething about the company’s maneuvering around an NIH-funded comparative trial. The desperate move of a company painted into a corner? Hardly. Beneath the headlines, it looks like Genentech is going to win the fight to prevent Avastin from becoming standard care for AMD.

By Ramsey Baghdadi

It sure looks like GenentechInc. is on the defensive when it comes to the age-related macular degeneration therapy ranibizumab ( Lucentis).

Genentech’s decision to cut off sales of the vascular endothelial growth factor inhibitor cancer drug bevacizumab ( Avastin) to compounding pharmacists looks like a desperate effort by a company with its back to the wall. Though not approved for ophthalmic use, Avastin has taken almost half of the market for AMD, with the result that Lucentis sales have stalled.

Even worse, a long planned study—organized by Lucentis’ customers to prove the drug isn’t worth a high premium —has finally started enrolling patients.

And then Genentech couldn’t even make the cut off stick. Ophthalmologists, already seething over the company’s efforts to discourage the comparative trial, went ballistic when they learned that Genentech was restricting access to what they consider to be standard-of-care for AMD. So Genentech agreed to delay the cut-off for a month.

By all appearances, Genentech is losing the fight to protect the premium price it thinks Lucentis deserves.

Appearances can be deceiving.

There is no denying the commercial impact of the rebellion by ophthalmologists against Lucentis. But Genentech appears to have laid the groundwork to prevail.

In the short term, the company will regain lost ground when it implements the distribution restrictions on Avastin. Make no mistake: Genentech will implement that change, and should be able to make it stick. That by itself won’t stop off-label use of Avastin, but it will reduce one set of economic incentives that encourage it today.

And, in the long run, Genentech is well-positioned to cut off any serious impact from the comparative trial of Avastin versus Lucentis. Even if the trial results are reported as showing Avastin is as good as Lucentis at one-twentieth of the cost, Genentech will be in a strong position to counterattack, thanks to some big snags in the study design.

Taking on Your Customers

The face-off between ophthalmologists and the biotech company over off-label use of Avastin for wet age-related macular degeneration in place of the specially engineered VEGF inhibitor ranibizumab ( Lucentis) has pitted the company against one of its biggest customer groups. Typically, companies do their best not to anger their biggest customers. However, given the circumstances, Genentech clearly thought they could—and should—take actions that would do just that.

Genentech did everything right when it came to developing Lucentis. They made one critical mistake when the drug was ready to hit the market: vastly overpricing it. ( See "An Eye for an Eye: Lucentis and the New Value System for Pharmaceuticals," The RPM Report, June 2007 (Also see "An Eye for an Eye: Lucentis and the New Pharmaceutical Value System" - Pink Sheet, 1 Jun, 2007.)) The price, about $2,000 per monthly dose, was considered exorbitant by ophthalmologists, who found an alternative they felt was just as effective and much cheaper: microdoses of Avastin that cost roughly $100 per injection.

So eye doctors began flocking to compounding pharmacies in droves a full year before Lucentis was even approved by the Food & Drug Administration in order to get access to lower doses of Avastin.

Not only that, ophthalmologists leveraged the clout of their professional organizations to make the value argument for low-dose Avastin to the reimbursement community, including private insurers and the Centers for Medicare & Medicaid Services. The American Academy of Ophthalmology opened a registry of off-label Avastin use, collecting over 9,000 case reports to support coverage. And it worked.

After all, choosing between spending $50,000 per patient versus $2,400 per patient over the recommended two-year therapy schedule is pretty easy if the drugs are considered equally effective. And that is what ophthalmologists hope to prove more rigorously, after successfully lobbying the National Institutes of Health’s National Eye Institute to conduct a randomized, prospective, head-to-head comparative trial of Lucentis to Avastin.

The result has been a distinctly unusual marketing challenge for Genentech.

" Avastin itself is an unusual and rather difficult competitor to deal with, certainly from our point of view," Genentech EVP for commercial operations Ian Clark dryly understated during a third quarter earnings call on October 15.

The adoption of the off-label use has been staggering. "We’ve quoted the Lucentis market share at approximately 50% and it is fair to say that the vast majority of the rest of the market is Avastin," according to Clark. He acknowledges that OSI Eyetech’s discontinued pegaptinib ( Macugen) and Novartis AG ’s verteporfin ( Visudyne) have "some remaining use."

In other words, it’s about a 50-50 split between the two drugs, with Genentech holding a virtual monopoly over the total wet AMD market. Lucentis generated $198 million in sales for the third quarter, a 29% increase over the same year-ago period. That sounds great, but in reality, sales are going in the wrong direction : third-quarter revenue represents a 5% decline compared to the second quarter of this year and a 6% drop compared to the first quarter. Not only that, Genentech has been leaving roughly $200 million on the table per quarter, or in the realm of $800 million in unrealized revenue a year, because doctors who would otherwise choose Lucentis are instead using Avastin.

Genentech’s Lucentis/ Avastin nightmare is viewed as a model for how wrong things can go after doing everything right.

But now it looks like Genentech will have the last word. The biotech company is in the process of employing two separate strategies—one targeting physicians and the other targeting Medicare—that should eventually make Genentech the victor in the Lucentis debate. The common thread of both tactics is taking a hard line and increasing the pressure on rebelling physicians and payors. Genentech may compromise its image from being viewed as the gold standard for research in the biotech industry to a company more in the mode of Big Pharma, but that appears to be the pound of flesh necessary to save one of its star revenue producers.

Genentech Cuts Off Avastin Supply

From Genentech’s perspective, there is nothing heavy-handed about the move to cut off compounding of Avastin. The company normally distributes its products by selling to authorized wholesalers, who in turn, sell on to the end user—a physician or hospital. In 2006, Genentech halted all sales to secondary distributors over concerns of counterfeiting, stockpiling and pricing arbitrage. "We don’t think we need it," Clark explained, referring to the secondary market.

However, Genentech left the door open for one product, Avastin. "We felt prior to the approval of Lucentis, there was an unmet need," Clark said. "We don’t think that’s there now."

To ophthalmologists, the decision looked different, like a calculated attempt to make it much tougher to get Avastin for a use Genentech obviously wants to see end. On October 11, Genentech sent a letter to retinal specialists announcing they were completely shutting off supply of Avastin to compounding pharmacies.

"As of November 30, 2007, Genentech will no longer allow compounding pharmacies to purchase this product directly from wholesale distributors," the letter, signed by product development president Susan Desmond-Hellmann, states. Physicians, however, can still acquire inventory from Genentech’s authorized distributors or though hospitals. "We expect them to do so," Clark maintained.

Genentech said a number of issues led to the decision to restrict distribution. In addition to the availability of Lucentis, the company cited an FDA warning letter to a compounding pharmacy related to the sterility and repackaging of Avastin for use in the eye. Also, Genentech says, during a separate, routine FDA inspection of a manufacturing facility, field inspectors raised concerns about the off-label use of the cancer drug because it is not "designed, manufactured or approved" for the eye.

Ophthalmologists, though, certainly didn’t accept that explanation. The warning letter in question was issued to a Massachusetts compounding pharmacy. The inspection of the pharmacy began in September 2004 and was completed in January 2005. The letter was sent to the New England Compounding Center for a number of different violations, including repackaging Avastin, on December 2006. In other words, the letter was close to a year old before Genentech cited it in its October letter to physicians.

As for FDA’s concerns, Genentech did not provide much detail, beyond saying that agency inspectors questioned whether the company was taking sufficient steps to prevent inappropriate reformulation of Avastin in the marketplace. So ophthalmologists suspected that the agency’s concerns probably didn’t translate into a regulatory imperative for Genentech to make the change.

The company’s assiduous concern over off-label use of Avastin for macular degeneration may put Genentech in an inconsistent and awkward position in regards to one of the other commercial areas for the product.

Avastin is used off-label in metastatic breast cancer, a use that contributed to "strong" growth by the product through nine months of 2007. Genentech execs preface remarks about the product’s off-label oncology use with phrases such as "our policies do not allow us to promote Avastin for breast cancer," but they report that the drug is maintaining steady use in that segment. Genentech is not making a well-publicized attempt to make it more difficult to get supplies for that use of Avastin.

Impact of Restricted Distribution

Still, as Genentech points out, physicians will still have plenty of avenues to use to obtain compounded Avastin for their patients. If ophthalmologists want to get Avastin from hospital pharmacies or order direct from authorized wholesalers, those routes will remain open.

But that isn’t making ophthalmologists any happier.

One leading ophthalmologist describes the restricted distribution program as the final move by Genentech to remove all potential liability related to off-label use of Avastin and shift it completely to retinal specialists using the cancer drug to treat AMD.

The ophthalmologist notes that there is an ongoing study of Medicare macular degeneration claims at Duke, the AWARE study under Scott Cousins, to try to pick up the frequency of untoward events from excessive anti-VEGF from injection in the eye. The study uses date from the Chronic Condition Warehouse database, managed by a Medicare contractor, the Iowa Foundation for Medical Care (IFMC), a Medicare contractor. "Genentech knows more [about adverse events related to Avastin in the eye] than they’re saying publicly," he says. Genentech claims that Lucentis has been designed as an antibody fragment to bind more specifically in the eye and avoid appearing systematically.

In addition, the restricted distribution plan shifts a large inventory risk to Genentech’s customers: the wholesalers and ophthalmologists. The firm says it is not changing payment terms from its current 85-day billing cycle for the product. It could have extended the payment terms to soften the blow of forcing more doctors to the higher-priced version of anti-VEGF treatment. The higher priced product also puts the eye doctors in the uncomfortable position of trying to collect average co-pays in the $400 per month range.

So ophthalmologists were furious with Genentech after the letter hit. Top company officials, including Desmond-Hellmann, flew to Washington on October 26 to meet with representatives from the American Academy of Ophthalmology (AAO) and the American Society of Retina Specialists (ASRS) to discuss the distribution decision.

Three days later, the company issued a second letter to retinal specialists, which sounds conciliatory. However, the end result of discussions with AAO and ASRS was to give physicians an extra four weeks to get used to the new distribution program. "We made this change so that affected physicians will have sufficient time to develop and implement transition plans to assure that patient care is not adversely affected," the second letter says.

Genentech says it agreed to delay the plan after the October meeting to allow physicians to transition their patients. "We felt these were reasonable requests so we granted them," Genentech spokesperson Dawn Kalmar explains.

Genentech also promised ophthalmologists it would lift the restricted distribution plan and begin shipping Avastin to compounding pharmacies again if FDA gave the company "legal and regulatory authorization to do so."

Simply put, that is not going to happen.

The company has no plans to file a supplemental BLA for Avastin for wet AMD, nor does it plan on securing the multiple regulatory, compliance and manufacturing authorizations necessary to make Avastin available to compounding pharmacies in the future. Why would it? Every sale of compounded Avastin diminishes Genentech’s return on the investment in developing Lucentis.

How Genentech Held Up NIH, CMS

The hard line taken by Genentech on the supply side follows the extended and contentious argument that Genentech has been engaged in with the ophthalmologic community over a federally funded comparative trial of Avastin and Lucentis.

Physicians persuaded the federal government to pursue the study despite the loud opposition of Genentech. In that fight, NEI has been the physicians’ most powerful ally. CMS has also supported ophthalmologists, with officials estimating a switch to Avastin could save Medicare up to $3 billion a year. CMS has recently received political pressure to cut back on Lucentis spending. ( See Exhibit 1).

But there have been some serious roadblocks within the Medicare agency, according to principal investigator Daniel Martin, Emory University School of Medicine. And those roadblocks may have doomed any potential positive outcome for ophthalmologists before the study even starts.

The design of the Comparison of Age-Related Macular Degeneration Treatments Trials (CATT) began in 2005. The study would evaluate 1,200 patients in four dosing arms: fixed doses of Lucentis, fixed doses of Avastin, Lucentis as needed, and Avastin as needed.

The $50 million study was supposed to start enrolling patients in 2006 at 47 US sites but was delayed for a number of reasons. First and foremost, Genentech refused to participate in the study by supplying Lucentis, a decision that started the NEI-sponsored investigators on a torturous route to find funding and sources of supply of drugs for the trial.

NEI initially agreed to fund the study with $16.2 million over four years for the infrastructure of the trial, including a coordinating center, clinical coordinators, and optical coherence tomography (OCT) and photo reading analysis.

NEI also agreed to foot the bill for the $1 million cost of Avastin for the trial. But that leaves the cost of Lucentis unaccounted for. In most studies, of course, the manufacturer provides the drug for free. Genentech’s refusal left NEI a long way from being able to begin the trial: the cost of Lucentis for the trial will be $22 to $25 million depending on how much is used in the as-needed dosing arms.

"Given that the percentage of trials funded by NIH is already at an all time low I don’t think it is reasonable to expect NIH to cover the cost of Lucentis, especially since CMS is already responsible for care in the majority of these patients," Emory’s Martin said during a CMS town hall meeting on clinical trial policies September 20. The purpose of the meeting was to discuss the effects of coverage and payment on clinical trial retention, participation and conduct.

CMS director of coverage and analysis Steve Phurrough explained that CMS was grappling with a number of issues. "Do we pay too soon for something that prevents someone from then enrolling in the trial that’s going to determine the benefit of that particular technology? Do our payment policies affect various trial design issues, such as randomization and blinding?"

In trying to launch CATT, the first obstacle was trying to decide which government agency to work with, in which order. Martin said there was a vigorous debate about whether to work out the potential issues with CMS first or whether the trial should be funded first. In the end, NEI agreed to fund the trial in October 2006.

Study investigators had their first face-to-face meeting with the coverage and analysis group in July 2006.

A big financial obstacle arose when CMS told the study investigators that they did not think Medicare could pay for Lucentis without a change in the clinical trial policy. "This was a surprise to us, particularly when Lucentis is already FDA-approved, and routine care is covered in a clinical trial, with routine care defined as items or services typically covered absent a clinical trial." Martin maintained.

Still, Lucentis was deemed an investigational therapy under terms of the trial coverage policy. The problem was eventually resolved when CMS revised its clinical trial coverage policy, but the result was a delay in the start of the head-to-head trial ( See "CMS: The Clinical Trials Agency," The RPM Report, June 2007 (Also see "An Eye for an Eye: Lucentis and the New Pharmaceutical Value System" - Pink Sheet, 1 Jun, 2007.)).

Un-Masking the CATT Study: Genentech’s Big Win

Once CMS determined that Lucentis would be covered by Medicare in the study, that still left the matter of patient copays. Medicare only pays for 80% of the drug costs, leaving study participants with a potentially large bill even though they are enrolled in a clinical trial.

That isn’t the only problem: the size of the copay would serve to unblind the study, since patients would know immediately that they were on Lucentis when they got their bills. That is not how the investigators intended the study to work; they went to considerable trouble to design the trial as a rigorous double-blind study, even to the point of obtaining an IND to use compounded Avastin in special vials with an overwrap to look identical to Lucentis, complete with an 11-month shelf life.

All that work would go for naught if study participants received a monthly bill from their Medicare carrier telling them what drug they got and how much they owed in copays.

Even worse, Martin noted, the difference in the copays would surely serve to jeopardize the integrity of the study by causing differential dropouts based on ability to pay.

At first, the investigators hoped that CMS would simply waive the copay, arguing that the Medicare program would still be saving money since many patients who otherwise would be generating Lucentis claims would be treated with Avastin at NIH’s cost. But the strict language of the Medicare statute made that a no-go.

So instead the study investigators asked CMS to make an upfront outlay of $25 million to centrally purchase the drug and then distribute it locally to the different sites. The study investigators proposed to bill patients for a blended average of the copay between the two drugs and then provide a rigorous accounting back to CMS. That would mean CMS would collect its 20% copay and the masking would be preserved.

The lead investigators met with CMS staff and turned to creating a demonstration project that would use a payment structure similar to a research grant-sponsored study, with an upfront payment to purchase and distribute the drug, eliminate bills and provide centralized recordkeeping.

"Asking for $25 million we didn’t think was so unreasonable because this project did not expand existing coverage; this was not going to be a net increase in dollars paid out by CMS," Martin said. "In fact, there was going to be enormous cost savings: for the 1,200 patients participating in the study…the savings to CMS is $25 million, it’s real simple math. These are not cooked numbers."

Over the next four months, a number of different plans were developed and subsequently shot down, further frustrating the NEI group, according to Martin. "One project would be developed and it would go up for legal review and come back and say ‘no, that won’t work,’ and another would go up and they’d say ‘no, do it the other way, the way you did it the first time.’"

Finally, the demonstration project eventually received approval from all the key stakeholders within CMS. Lucentis and Avastin would be billed by clinics using a special G code and identified on notices to beneficiaries as " Lucentis/ Avastin study drug." The total price was the average cost of the new drugs "plus a small margin" that had been built in on the basis of the assumed imbalance from the "as needed" dosing arms. NEI agreed to pay all co-pays and it was determined that it was legally permissible for the institute to do so after Medicare and supplemental policies had paid.

Then-CMS Administrator Leslie Norwalk signed off on the project in May 2007. It was sent to HHS and the Office of Management and Budget for approval. The CATT study principals had previously discussed the projects with both government agencies and received strong support in return, according to Emory’s Martin. Three months went by without a word. Then, seemingly out of nowhere, the Office of the General Counsel at HHS led by former Sidley Austin attorney Daniel Meron deemed the project unapprovable.

The trial investigators were stunned. "The justification provided was that ‘it was obvious that the demo project will improve the quality of the clinical trial with Medicare beneficiaries’ participation in it." Therefore, the HHS general counsel maintained "that we did not need to do the project to prove it,’" Martin said. "I don’t think any scientist would ever accept that argument as a valid argument."

NEI is going to go forward with the study anyway, with the first patient expected to be enrolled by the end of the year. The institute will cover the co-pays, but patients will be informed of the drug they actually receive when the get their Medicare coverage notices or supplemental insurance notices. In other words, examiners and treating physicians will be masked, but patients will not be.

"This is not what we initially thought to do," Martin maintained. "We remain confident that we can do a clinical trial that can produce robust clinical data, but it is not optimal."

Martin is optimistic that the results will not be compromised, at least when it comes to the two standardized dosing arms. In those cases, the physician and patients are essentially receiving "robotic" therapy, Martin says, with no provider discretion that would jeopardize a rigorous comparison between Lucentis and Avastin.

Giving CATT the CATIE Treatment

Genentech won’t see it that way.

The company now has a significant flaw in the study to exploit in the future, assuming the data show Avastin and Lucentis to be equally effective. Genentech can simply point to the failure to mask AMD patients taking part in the study as a serious deficiency of the overall clinical trial that makes any interpretation suspect.

If the study shows Lucentis to be more effective than Avastin, great. If it doesn’t, the study wasn’t properly controlled. Either way, Genentech is in the driver’s seat.

The fact is that even without the built-in design flaw, Genentech would probably be able to weather the CATT study. Biopharmaceutical companies are becoming increasingly savvy at defusing the results of large-scale, head-to-head government-funded studies aimed at debunking the value of their newer therapies.

For example, the National Institute of Mental Health-funded CATIE study concluded with great fanfare that older, generic antipsychotic drugs were just as effective as newer, more expensive therapies ( 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.)).

But the results had little negative impact on the scripts of those more expensive atypical antipsychotic drugs. Why? Eli Lilly & Co. ( Zyprexa), Johnson & Johnson ( Risperdal), AstraZeneca PLC ( Seroquel), and Bristol-Myers Squibb Inc. ( Abilify) each found a loophole to build a strategy around: a subset of data to tout as a positive result, a small study to counter the CATIE results, highlight a dose-effect or take the message directly to prescribing physicians. And it worked.

The National Heart, Lung and Blood Institute-funded ALLHAT study had a similar outcome, except in the antihypertensive category: older and cheaper was deemed just as good as newer and more expensive. Despite the clinical data, huge patient population and high expense, the impact of the study on physician prescribing patterns was negligible.

The CATT study likely would present Genentech with similar avenues to discuss with physicians no matter the outcome. For example, there is certain to be a most effective dosing schedule of Lucentis in the "as needed" arm which will generate calls for further research and continued updates to prescribers.

Now Genentech has an even simpler case. The study simply may not be viewed as robust enough to drive any change in treatment decisions by physicians, and certainly would be a dubious base for coverage policy changes.

Getting Off "The Government’s Timeline"

During the clinical trial meeting, Martin said CMS still wanted to find a way to work out a better process that could potentially allow patient masking. However, Martin and his colleagues decided not to wait any longer. "We’ve been on the government’s timeline for the last year and…it’s not a good place to be." However, despite their decision to drive forward, the trial has yet to start. An NEI spokesperson says April 2008 is now the target date to begin enrollment but that could change at any time.

For all the time and trouble, the CATT study has a clear and potentially fatal weakness.

That means Genentech is on the verge of winning the Lucentis/ Avastin value battle. By refusing to supply Lucentis upfront, Genentech delayed the trial by more than a year and counting and indirectly tainted any results that could hurt them down the road. On another front, the company’s decision to cut off distribution of Avastin to compounding pharmacies stops the bleeding for one of its most commercially promising therapies.

Genentech is not declaring victory publicly just yet—but that may change in early 2008. "I think [the difficult environment] is going to persist certainly through the end of this year and quite possibly into the first quarter of the following year," Genentech’s Clark said during the quarterly call. "I would then hope that maybe when that starts to fall behind us, we will see some return to growth."

Genentech’s victory won’t come without a price.

The collateral damage caused by Genentech’s hard-line approach—angering its ophthalmology customer base—is the one X-factor that will be hard to predict in the future.

The company says it intends to rebuild relationships in the retinal community, but that’s a difficult proposition considering the lengths the company went to in order to deny access to Avastin. Other companies may be able to capitalize on that pent up resentment. Regeneron Pharmaceuticals Inc. , for example, is developing a VEGF inhibitor for AMD that just successfully finished Phase II trials. Regeneron is partnering with Bayer AG outside the US, but holds exclusive rights to market the drug in the US.

But for now, Genentech appears to have emerged from between a rock and a hard place with the Lucentis franchise intact. The company is using unforgiving tactics, sacrificing some of its biotech good-guy image to protect a premium price. It has been a tough fight; but, in an era of disappearing blockbuster drugs, Genentech may have saved one from extinction.