April 22, 2011

New Cancer Treatments: Hype or Hope?

If you are a man, you have a 1 in 6 chance of developing prostate cancer during your lifetime. The good news is that most men will die with prostate cancer, not of prostate cancer. The bad news is that some men will succumb to a virulent form of prostate cancer: even with surgery and/or radiation therapy, tens of thousands will develop metastatic disease. Initially, prostate cancer typically responds to hormonal therapy (the cancerous cells need male sex hormones to thrive so treatment that lowers the level of these hormones can hold the cancer at bay). But if it stops responding to hormonal treatment and new metastases develop, patients are in trouble. The cancer often travels to bones, causing pain and fractures. Sometimes it causes spinal cord compression, which can cause paralysis. According to the National Cancer Institute, about 32,000 men die of prostate cancer in the US every year.

When the Seattle-based biotech company Dendreon announced a radical new approach to the treatment of prostate cancer, the enthusiasm was palpable. As the company reports on its website, it is in the business of "transforming lives through the discovery, development, and commercialization of novel therapeutics." Its new strategy for prostate cancer is more like a vaccine than a drug and it involves a highly individualized approach rather than one-size-fits-all chemotherapy. Treatment entails drawing blood from a patient, removing the white blood cells, treating them in the laboratory to endow them with special anti-prostate cancer fighting properties, and then returning them to the patient. Three randomized controlled trials suggested that the new approach, called autologous cellular immunotherapy or sipuleucel-T (brand name Provenge) is effective with relatively modest side effects. The Food and Drug Administration approved sipuleucel-T in April, 2010. Now Medicare is trying to decide whether to pay for the therapy. Medicare has commissioned a "technology assessment" to evaluate the evidence that the treatment is effective; it has carefully reviewed all the available data, and it has assembled its advisory committee to make a recommendation. Much of the reason for all this scrutiny is cost: a full course of treatment (three infusions) is just under $100,000.

Since the FDA has found sipuleucel-T "safe and effective," isn't any restriction of usage by Medicare simply government interference in the practice of medicine? That's one way to spin Medicare's concern. It's certainly the view of the company that makes the sipuleucel-T system. But a sober assessment of the facts suggests the alleged benefits of immunotherapy may well be exaggerated: the system has only been shown to work in a small fraction of men with so-called castration-resistant metastatic prostate cancer; it does not prolong the length of time men go without progression of the disease; and its efficacy may depend on the use of conventional chemotherapy.

Which men stand to benefit from sipuleucel-T?
To be eligible to participate in one of the 3 clinical trials of sipuleucel-T, men had to meet a whole slew of criteria. They had to have low testosterone levels (indicating their disease was progressive despite lowering of the male sex hormone), they had to be in generally good health (as indicated by independence in their daily activities), they couldn't be experiencing pain or taking narcotics for pain, and they couldn't have lung, liver, or brain metastases. In this carefully selected group, sipuleucel-T increased median survival by 4 months. We simply don't know whether the observed effectiveness of sipuleucel-T applies to men who do not meet these criteria.

What exactly does sipuleucel-T do?
The first two clinical trials of sipuleucel-T concentrated on measuring the length of time before a patient's cancer progressed. The assumption was that even if the new treatment didn't cure prostate cancer-it didn't-it might extend quality of life, if not length of life, by offering patients a period in which they had no measurable tumor growth. Neither study found any such effect: the time until progression of the disease was in the range of 10-11 weeks with either sipuleucel-T or placebo. But surprisingly, median overall survival was 4 months longer with sipuleucel-T. A third study, undertaken after the first two and published in the New England Journal of Medicine, deliberately chose overall survival as the main outcome measure (presumably since using the time to progression hadn't shown that the $100,000 therapy offered an advantage). Consistent with its predecessors, this study also found survival was better with sipuleucel-T (25.8 months versus 21.7 months), although again, median time to progression was identical in both groups (14.6 weeks versus 14.4 weeks). Biologically, this result was puzzling: as an accompanying editorial commented, it is hard to understand how sipuleucel-T could prolong life without having any measurable effect on tumor growth.

Why do men with castration- refractory metastatic prostate cancer live longer with sipuleucel-T than with placebo? It's not clear. But one curious observation is that once the tumor progressed-which it did in 90% of people during the course of the study-physicians were free to treat patients with other anti-cancer regimens such as conventional chemotherapy, and more of the patients who had gotten sipuleucel-T opted for such treatment. It's possible that it was in fact the chemotherapy that contributed to the survival advantage in this group of otherwise healthy older men, although the mathematical models generated by the authors to adjust for differences in treatment after relapse dispute this interpretation.

It seems that the latest and greatest treatment for prostate cancer may not be so wonderful after all. And if it is approved by Medicare, it will almost certainly be used "off label," that is for other patients with castration-resistant prostate cancer, including those on narcotics for bone pain and with lung or liver metastases, patients who were excluded from the original studies. Physicians are legally allowed to use an approved treatment modality in any circumstance they think it might be effective. If Medicare approves immunotherapy without restrictions, the treatment is likely to be applied to thousands of men who are far older, sicker, and more impaired than the men in whom the treatment was tested. Whether these men will benefit at all is unknown-there are simply no studies that address that question. But if sipuleucel-T is of uncertain benefit in the healthiest men, it is even less likely to help those who have multiple other medical problems. Treating 1000 men with immunotherapy will cost Medicare $100 million. But what if it's 10,000 men? Do we really want to spend a billion dollars on a treatment that, at best, might extend life a few months and, at worst, will have no benefit at all but may cause fevers, chills, and tremors?

Politicians have finally recognized that we need to curb the rate of rise of expenditures in the Medicare program. The only sensible way to achieve this end is to require Medicare to function within a budget. And if we don't want to do this by simply limiting the amount of money that Medicare pays out to private insurance companies in the form of vouchers-an approach that will have no effect on the rate of rise of medical expenditures and will merely shift the cost from the federal government to the consumer-we will need to finally insist that Medicare make coverage decisions responsibly, taking cost into account. Medicare is constrained by law to cover treatments that are "reasonable and necessary." Surely it is not reasonable to approve a treatment that costs $100,000 unless it has clearly been shown to work. If the therapy is effective in some patients, its use should be restricted to the types of patients in whom it has proved useful. Finally, Medicare should be able to set the price it is willing to pay for such treatment based on its cost-effectiveness compared to alternative treatments. That's not government intrusion into the practice of medicine; it's common sense.