Two articles in the most recent New England Journal of Medicine should get you thinking about where we are as a profession. Both have to do with prostate cancer, and while unrelated in their subject matter they both require us to address the same issue: whom do we treat?
The first, by Bill-Axelson and colleagues (NEJM 364 (18): 1708-17), is a randomized trial of radical prostatectomy versus watchful waiting in early stage prostate cancer. It may seem strange to think that it took so long to prove that local therapy improves prostate cancer survival: didn’t we know that all along? Apparently not, at least not in the prospective randomized trial sense of scientific proof. I do not think that this is a trial we could have performed in the United States, but our Swedish colleagues were able to carry it off. The first report of this trial was published in 2002; this paper gives 15-year follow-up.
At 15 years the difference in survival between watchful waiting and immediate prostatectomy is 6.1%. The survival benefit was confined to men younger than age 65, and the number needed to treat to avoid one death was 15 overall and 7 for men younger than 65. The age results are not too surprising, of course: competing causes of mortality come into play for every medical intervention in older patients. Survival benefits were seen even (especially?) in low-risk prostate cancer. As this trial had very few PSA-detected prostate cancers, its applicability to such men is uncertain.
I am sensitive to these questions as the son of a man who died of prostate cancer at the age of 71. It still impresses me how many unanswered questions swirl around screening and treatment of early stage prostate cancer. But the questions almost pale in significance compared to advanced prostate cancer, as another paper by Chambers and Neumann (NEJM 364 (18): 1687-9, 2011) suggests.
These authors discuss the Provenge story. Provenge (sipuleucel-T) is a relatively new cellular immunotherapy approach demonstrated to improve median survival by 4.1 months in patients with hormone-refractory metastatic prostate cancer. The FDA approved Provenge last year, and the company that produces it charges $31,000 per treatment for a total of three treatments. The Centers for Medicare and Medicaid Services (CMS) opened a national coverage analysis before eventually agreeing to pay for Provenge on March 30th.
Provenge has become something of a poster child for the health economics of cancer: high-tech medicine that is as expensive as all get-out which only modestly improves the survival of patients with metastatic disease. I know that had the drug been available, my father would have taken it in a heartbeat: he wanted every last second of life that modern medicine had to offer. I know that his oncologist would have offered it to him: as doctors our first responsibility is to our patients, and improving overall survival is our gold standard for offering a treatment.
But I also know that the health care system can only survive so many “break the bank” treatments before it goes belly up, and that $93,000 for four months of life seems over the top. When CMS launched its coverage review of Provenge last year, ASCO objected. We objected not because we love expensively-priced treatments—we are quite concerned about the unsupportable trends in the cost of cancer drugs—but because we did not feel that the government should institute an American version of Britain’s NICE through the back door, and because the FDA had already performed due diligence. We do need some rational means of allocating scarce resources if our health care system is to survive. We do need a rational political discourse on this subject.
But to get back to the question we started with: whom do we treat? In both early stage and late stage prostate cancer, as with many cancers, knowing who benefits should be the first step to knowing whom to treat. Our “rationing” of health care, in an ideal world, would be primarily science-based. For every treatment there are winners and losers, and knowing who does not benefit (ER-negative = no treatment benefit with tamoxifen) is as valuable as knowing that some do benefit. Knowing who does not benefit, almost always the majority of treated patients across the spectrum of human cancers, would save the healthcare system untold billions, and would save patients significant toxicity. Value (defined as effective care at a reasonable cost) is a societal as well as a personal good, and has the potential to transform sterile debates over the cost of health care.
Determining who will benefit is the job of our community. It will undoubtedly require us to understand tumor biology at a deeper level and to translate the lessons of the lab to the clinic. Along these lines I would also suggest, for your reading entertainment, a recent article on the genomic complexity of prostate cancer by Michael Berger and colleagues in Nature (Nature 470: 214-20, 2011). This paper compares the complete sequence of seven prostate cancers with their normal host DNA and points to several molecular events that may prove important for prostate carcinogenesis. Cool stuff, but hopefully useful stuff as well in the near future.