Oncology Drugs Add Promise, Stresses to Healthcare Delivery

Over the past decade, the slow trickle of oncology drug candidates has become a torrent: Close to 750 new cancer drugs (and new indications for existing agents) are in clinical development, according to PhRMA. 2009 Drug Report, an annual study by Medco, cited targeted oral drugs, combination treatments, personalized medicine, monoclonal antibodies, and expanding indications for existing drugs as leading trends in the oncology drug marketplace.

Medco predicts that new oncology drug approvals will decline in 2009 but rapidly recover thereafter. Oral agents slated for approval in 2010-2011 include Zactima (AstraZeneca’s vandetanib, for lung and breast cancers), axitinib (Pfizer; thyroid), alvocidib (Sanof-Aventis; leukemia, arthritis), motesanib (Amgen; breast, thyroid), and lonafarnib (Schering/Merck; breast). All are small-molecule, oral drugs with novel mechanisms of action. This trend, together with changes in oncology therapy practices, heralds a changed market for delivering cancer care.

One cancer, many pathways

Unlike conventional chemotherapy, which is administered in a doctor’s office or hospital, targeted oral drugs are self-administered in pill or capsule form. Patients find taking pills more convenient than waiting for hours in a clinic for chemotherapy, but what is truly exciting about current-generation drugs is they target biological pathways that cancer cells depend on to survive. Chemo, by contrast, kills all dividing cells, which is the source of side effects like nausea and hair loss. Targeted therapies at their best home in on mechanisms that affect only (or primarily) cancer.

Cancer is driven by a relatively small number of such pathways that stop working properly. Each general pathway type comes in several varieties, which nature mixes and matches from one cancer patient to the next. Some experts believe that curing cancer, or keeping it at bay for long stretches of time, will entail knocking out multiple pathways with several drugs administered in combination. Determining which pathways are operative, and designing drug cocktails to knock them out, is the future of personalized medicine, says Ken Bloom, MD, a pathologist and chief medical officer at cancer diagnostics firm Clarient (Aliso Viejo, CA).

Give me a sign

One favorite target pathway is biological signaling, which is carried out by proteins known as kinases. Drugs that disrupt the activity of kinases, known as kinase inhibitors, are among the most successful new oral cancer agents.

They include the oral drugs Gleevec (Novartis: imatinib), Sutent (Pfizer: sunitinib maleate), Nexavar (Onyx Pharma: sorafenib), Tykerb (GSK: lapatinib), Tarceva (Genentech: erlotinib), and Tasigna (Novartis: nilotinib), all of which are significantly more effective and less toxic than traditional chemotherapy. On the down side these agents work in relatively narrow patient populations.

Another “pathway” class to watch is the vascular endothelial growth factor (VEGF) inhibitors, which work by slowing the formation of blood vessels that feed tumors. Genentech’s Avastin (bevacizumab), arguably the most successful VEGF agent, is an injected biologic for treating advanced cancers of the colon, lung, breast, and most recently kidney; a brain cancer indication is expected in late 2009. GlaxoSmithKline’s pazopanib, an oral VEGF agent close to being approved, is effective against advanced kidney cancer.

Kinases and VEGF are popular targets for cancer drug development because every tumor depends on pathways associated with them. Less popular approaches include agents that induce or prevent apoptosis (programmed cell death), cell cyle, enzyme inhibition, immune stimulation/suppression, and receptor binding.

Some worry that kinase and VEGF agent classes are already suffering from over-crowding. When Phase III data on pazopanib were presented at an American Society of Clinical Oncology meeting in June, Primo Lara, MD, from the University of California, Davis, called pazopanib a “me-too” drug for its similarity to agents already on the market. Newer drugs, he argued, should work better, cause fewer side effects, or cost less. “Pazopanib has not met any of these criteria,” Lara said.

The danger of not approving what appear to be copycat cancer treatments is the possibility of losing a drug that might work wonders in the right patient group or in a different cancer. Phase III data for Genentech’s breast cancer drug, Herceptin (traztuzumab) were uninspiring because the study failed to focus on patients who over-expressed the Her2neu biomarker. Once that criterion was applied, Herceptin became a blockbuster.

Similarly the monoclonal antibody ipilimumab (BMS/Medarex), which works by stimulating the immune system, was originally tested in melanoma patients but early clinical results were mixed. In one study released in 2007 the drug was no more effective than placebo in treating advanced skin cancer. But in June 2009, the Mayo Clinic announced that two men with advanced, aggressive prostate cancer were cured by the drug.

Antibodies remain an active area of R&D for oncology despite falling short of “magic bullet” status early developers had hoped for. Today, most cancer antibodies (and other biologics) are administered by injection or infusion, alongside conventional chemotherapy. Antibodies target tumors by killing cancer cells that carry specific protein markers. Thanks to the high specificity of today’s antibodies, there has been renewed interest in antibody drugs that carry a toxic “payload” directly to cancer cells. Antibodies are considered targeted therapies but, like chemotherapy drugs, they operate on cancer cells rather than underlying biological pathways.

Notable cancer antibodies include Genetech’s Rituxan (rituximab; for treating non-Hodgkin’s lymphoma) and Herceptin, and Erbitux (cetuximab; BMS/Merck/Imclone) for pancreatic and lung cancers. Two cancer antibodies, ipilimumab (BMS/Medarex; melanoma, prostate cancer), and Arzerra (GSK’s and Genmab’s ofatumumab, for lymphoma and leukemia), are in late-stage testing.

Implications of oral cancer treatments

The advent of cancer pills that patients take at home is revolutionizing the oncology marketplace on several levels. Just 10% of marketed oncology medicines are taken orally, but this segment is expected to comprise 25% of treatments within five years. Oral, targeted cancer drugs tend to be more effective and far less toxic than old-line chemotherapies, but greater effectiveness comes at a price.

For example, BristolMyers Squibb’s Sprycel (dasatinib) leukemia drug costs close to $4,000 for a month’s supply, and Pfizer’s Sutent kidney cancer medicine costs $5000 for 28 pills. It is not unusual for drug plans to impose significant copays—20% to 30%—for these treatments, or refuse to cover them when they are used off-label.

Oral treatments cut into oncology practice profitability as well. Until fairly recently, oncologists acted as de facto pharmacists, stockpiling chemo drugs and re-selling them for whatever price insurers would pay. This practice, known as “buy and bill,” accounted for 30-50% of oncologists’ income. The only good thing one could say about the practice is that chemo was generally covered 100% as a medical treatment, whereas patients pay significant sums, sometimes 100%, for cancer pills. In 2005 Medicare limited the mark-up to 6% and private plans quickly followed suit. Oncologists complain that 6% barely covers costs associated with purchasing, inventorying, and tracking intravenous drugs, although they may still charge the going rate for dispensing them.

Cancer treatment decisions are optimally based on clinical factors. But to remain profitable in today’s economic environment oncologists must familiarize themselves with such terms as working capital, human resource costs, practice efficiency, and opportunity costs, observes Susan Dennis, an internal oncology expert with healthcare analytics firm SDI (Plymouth Meeting, PA). In some instances medical decisions may be based on what insurers pay doctors for conventional chemo administered in a clinic, and what they will reimburse patients for pills taken at home. “The typical community oncologist must re-evaluate every facet of their practice to recoup the revenue loss incurred with reimbursement changes,” Dennis says.

Since doctors make no money on medicines patients obtain from a pharmacy, cancer pills will cause oncologists’ incomes to deteriorate even further. One need not be too cynical to detect the potential for conflict of interest, as doctors have a financial disincentive to prescribe what may be a more effective medicine. In fact oncologists are keenly aware of the economics. A January 2007 study in the Journal of Clinical Oncology reported that 42% of oncologists regularly raised drug-cost issues with patients, 23% said cost affects treatment decisions, and 16% don’t mention very expensive drugs to patients they think cannot afford them.

Advocacy groups, including the American Cancer Society, are lobbying states to mandate equivalent reimbursement for oral and infused cancer drugs. Twenty-five states are considering such laws but only Oregon has one on the books. Equivalent reimbursement, while appealing to our sense of fairness and compassion, will raise healthcare premiums and likely exacerbate the perverse incentives for oncologists to prescribe less-effective drugs.

“The economics of oncology therapy are just goofy,” observes Clarient’s Kenneth Bloom. Oncology drug prescribing needs gatekeepers, Bloom argues, and pathologists—physicians who examine cancer tissue from biopsies and surgeries—are in the best position to serve in that capacity. By examining cells microscopically and testing them for various attributes, they have a better grasp on the biology of cancer than do medical oncologists, who prescribe based on a tumor’s organ of origin. Biology- or pathway-based prescribing will be essential for optimal utilization of the emerging classes of oral cancer drugs. “The key is breaking out beyond organ classification for cancer, to a molecular definition,” says Richard Wagner, PhD, senior director at MattsonJack (Foster City, CA), an industry consulting firm.

Bloom is also critical of how cancer drugs are tested against organ-based diagnoses instead of cancer pathways. That, he says, is one reason why half of all cancer drugs end up being used off-label, a practice he believes serves neither patients nor the healthcare system at large. No scientific evidence exists, for example, for 70% of off-label chemotherapy use in breast cancer, according to a study presented at the recently held ASCO meeting.

An article in the May issue of Managed Care quoted John Frick, director of pharmacy at Blue Cross & Blue Shield Assn. (Chicago), on off-label use of cancer drugs: “Instead of saying ‘It couldn’t hurt,’ we should be asking ‘How will this help?’ That is the real question.”

Patient adherence worries

Off-label use is rampant in cancer therapy, but one can understand that physicians and patients feel obligated to “do something,” even when science suggests otherwise. Less comprehensible is poor compliance to drug regimens, which appears to be as prevalent in cancer as in other diseases. Cancer patients, fearing side effects, often fail to show up for chemotherapy sessions. Cancer pills, notes Paul Boni, chief research officer at Grail Research (Cambridge, MA) are easier to quantify despite the well-known adherence gap for self-administered medications.

Manufacturers claim that adherence to newer cancer treatment is good, but research indicates this may be wishful thinking. Adherence to drug regimens depends to a large extent on the patient’s perception of their disease. A 1987 study found compliance to oral medication regimens in newly-diagnosed leukemia patients to be “remarkably low” at the stage when patients might not feel sick. Little has changed in two decades, as adherence for self-administered drugs still hovers at around 50% six months after the initial prescription is filled. While adherence is high for patients in supervised settings, or during the acute stage of disease, manufacturers worry that pills will go untaken as indications are broadened. Patients are less likely to stick to their treatment in situations where dosing is unsupervised, the objective is to prevent recurrence, patients do not perceive an immediate benefit, or they are taking several other medications. “Patients are trying to process lots of things patients at the same time,” observes Joanna Schwartz, PharmD, an oncology pharmacist and professor of pharmacy practice at the Albany College of Pharmacy (Colchester, VT).

Of all the factors affecting compliance, drug costs are perhaps the most intractable. With oral cancer drugs priced at $2,000 to $8,000 per month, patients skip doses or fail to fill prescriptions altogether. A report in the journal Blood earlier this year estimated that one-third of patients taking Gleevec for leukemia were non-adherent, and just 14% of patients fully complied with their drug regimen. Gleevec costs $3,600 per month. The numbers are astounding when you consider that this drug is a model of effectiveness for a once-fatal illness: Gleevec has kept some patients alive for ten years.

Pricing — elastic, up to a point

Paul Boni refers to oncology as a “privileged market” because of cancer’s seriousness, and the relative dearth of truly effective medicines. “There hasn’t been much push-back or resistance from payers,” a fact that has caused cancer drug sales to rise faster other medicines’.

Nevertheless, companies realize that no matter how well their products work they cannot expect unlimited reimbursement. Many of the most expensive cancer drugs now carry per-patient cost caps, which provide patients with as many doses as they need but limit how much insurers pay.

Pricing pressures are originating from the payer side as well, particularly with respect to “specialty pharma”—drugs that require special handling before administration. “It used to be off the radar screen; payers had other things to worry about. But now it’s hitting the breaking point,” Boni says. Payers are hitting back not with traditional tiering, but with requirements for prior authorizations and larger copays, which grow even more with off-label usage.

Oral cancer drugs are expensive, “but so are many other aspects of cancer care including diagnostic radiology and lab, surgeries, and radiation therapies,” observes Lee Blansett, SVP at MattsonJack. And diabetes, that poster-child of chronic illnesses, reportedly costs the U.S. healthcare system $191 billion per year, a figure that continues to rise. Exorbitant costs have been explained by the relatively small patient populations for which the drugs are approved, high late-stage clinical trial failure rates, and high development costs. The economics are actually quite simple: development costs are roughly the same whether the drug is indicated for 20 million people or 20,000. Perhaps the most charitable thing that can be said of high-priced cancer drugs is that patients do not survive long enough to bankrupt pharmacy benefit plans. “Very few cancer patients get to spend $80,000 on any one drug in one year,” Blansett notes.

One could argue that the price differential for cancer pills shrinks when total utilization costs and value are factored in. The point of targeted therapies, after all, is that only susceptible patients receive them, and insurers need only reimburse when the medicines are used appropriately.

On the value side, some oral cancer drugs have been shown to double or triple survival time compared with less expensive chemotherapy. Novartis’s Gleevec, which works in about 90% of cases of chronic myeloid leukemia, has transformed this once-deadly cancer into a manageable disease. A National Bureau of Economic Research working paper, authored by Cornell researchers and released in July,* calculated that the value-adjusted cost of colorectal cancer drugs was 30% lower than a decade ago if longevity and quality of life are factored in. The study examined data on a combination regimen of Genentech’s Avastin with two chemo drugs, compared with a single chemotherapy agent that was the standard of care in the mid-1990s. It concluded, based on actual prices and effectiveness, that Avastin’s $36,000 cost for 24 weeks of treatment was “lower” than that of the single chemo drug.

This line of reasoning works now perhaps, when the number of targeted therapies is small and most add just a few extra weeks or months to patients’ lives. The Cornell researchers will need to dust off their calculators as more such drugs are approved, and second- and third-generation therapies routinely extend life by years or decades for broad swathes of cancer patients. PC

*A Quality-Adjusted Price Index for Colorectal Cancer Drugs,” by Claudio Lucarelli and Sean Nicholson, Cornell Univ. Dept. of Policy Analysis & Management; NBER Working Paper No. w15174, available at

nber.org.