In Part I of this article, I will focus on our current understanding of drivers of cost for oncology care and the effect of the high cost on patients, as well as on how patients value treatment.
Progress in oncology has resulted in the rapid expansion of more effective, less toxic therapies, due to accumulating insights into cancer biology at the cellular level. However, the rising cost of cancer treatment now competes with the availability of effective therapy as a constraining element in our war on cancer. Patients are often simply unable to afford their personal financial responsibility for the cost of care. This rising cost appears to be due to replacement of less-expensive treatment with higher-cost care, longer patient survival times extending the period of active treatment, demographic changes leading to increased disease prevalence, and changes in the site of care. Oncology drug cost is escalating rapidly. Novel therapies reach patients in an increasingly haphazard payment environment, often placing cost-sharing burdens on patients that are beyond their capacity to pay. About 25% of cancer patients consume most or all of their savings in dealing with their cancer and its treatment. Patients may value treatment in a way that is different from what has been generally perceived, assigning greater importance to the chance of a large survival gain over a more predictable modest benefit.
Three years ago, I counseled a patient after a gastrointestinal stromal tumor had been resected from his stomach. I was pleased to be able to tell him that imatinib (Gleevec), a drug very well tolerated by most patients, would meaningfully reduce his risk of recurrence. Later, we learned that his out-of-pocket expense under his Medicare Part D plan would be several thousands of dollars for a year of treatment. The patient decided the expense was too onerous and that he would forgo treatment. Patient assistance programs are often limited for Medicare beneficiaries, and none could be secured. I pressed to ascertain whether cost was his sole concern, or if there was another unspoken reason for his resistance to proceed with treatment. There was not; his decision was purely due to cost.
How should we view this scenario? Should we think of it simply as the result of cost-sharing arrangements designed to make patients partially responsible for utilization of expensive therapy? Instead, should we decry the millions of dollars already invested in biomedical research and development failing to provide benefits to a patient for whom it was intended?
More than four decades ago, President Nixon signed the National Cancer Act of 1971. Initially, progress was slow. Over the past several years, however, a qualitative change has emerged in both our understanding of cancer biology and the tools we can use to fight it. We can now better discern the biologic mechanisms of the cancer cell itself-the ways in which genetic aberrations drive fundamental properties of malignant cells, namely, altered cell proliferation, capacity for tissue invasion and metastasis, and angiogenesis. Elegant treatments are now being developed that “target” key cellular signals of carcinogenesis or activate endogenous immune mechanisms against malignant cells.
These gains have not come without a price. Research and development is expensive. Patients who survive longer under active therapy generally receive more intense overall treatment; this includes not just the therapy itself but also the radiographic and laboratory surveillance necessary to monitor ongoing treatment response and toxicity. While we strive for the development of more effective, less toxic therapies, this progress may be transforming into a painful paradox: the more we advance scientifically, the more constrained we become economically.
Due to an increasingly robust scientific process, we can now expect therapeutic applications to be identified for dozens of novel cancer therapies over the next few years. However, these new therapies are currently made available to patients in the context of an ever more haphazard payment and coverage environment. Many of these treatments are remarkably expensive despite often providing relatively modest survival gains. These dynamics are forcing us to confront several essential questions regarding the cost of cancer therapy. What is an appropriate cost for cancer treatment? What are the key drivers of cost increases? Who or what mechanism should determine the cost? To what extent should patients be asked to share in the cost of their own care and what do we do when they cannot? What can we do to lower the cost of care?
In Part I of this article, I will focus on our current understanding of drivers of cost for oncology care and the effect of the high cost on patients, as well as on how patients value treatment.
Although cost estimates vary, the National Institutes of Health estimated direct spending on oncology care to be $89 billion in 2007 in the US. This spending represented 5% of total healthcare spending and 0.8% of US gross domestic product. Direct medical costs for oncology care include surgery, hospitalization, physician visits, imaging, radiation therapy, chemotherapy, and biologic therapy. Spending for oncology care tends to follow a U-shaped curve: treatment costs are highest in the year following diagnosis and the last year of life, and lower in the intervening years.
How has this spending changed over time? An analysis by Tangka et al compared spending in 1987 vs spending between 2001 and 2005. Between the late 1980s and the 2000s, spending on cancer care approximately doubled. As a proportion of overall medical expenditures, however, cancer costs remained stable, accounting for 4.8% of total medical expenditures in 1987 and 4.9% between 2001 and 2005. Thus, while oncology draws scrutiny largely because of the high cost of newer therapeutics, the cost of cancer care may not have been rising any faster than overall health spending. It is notable that the proportion of spending for inpatient admissions fell from 64% of total cancer spending in 1987 to 27% in 2001 to 2005. The transfer of oncology care to the outpatient setting has produced substantial savings compared with what the cost would have been had this shift from the hospital inpatient setting not occurred.
Factors Driving Escalating Costs of Cancer Care
Other sources estimate a higher current increase in oncology spending. A recent Institute of Medicine Report estimates that oncology spending is growing at more than 15% annually. Many factors have been implicated as the cause of the upward trend (Table 1). These include replacement of less expensive treatments with more expensive ones; provision of more aggressive care by physicians; and longer patient survival times, leading to prolongation of the period of treatment. Spending on cancer drugs has been estimated to comprise 10% to 15% of total spending on oncology care, with this cost rising between 15% and 20% per year. Tangka et al concluded that the key driver of escalating oncology spending from 1987 through 2005 has been the increased number people living with cancer, not the cost per treated cancer case.
Changes in the site of care may be acting as a more recent cost driver. A recent Milliman report examined the total cost of care for Medicare patients, depending on the site of chemotherapy infusion. The study examined the fee-for-service population from the Medicare Limited Data Set from 2006 to 2009. The per-patient-per-month cost was $4,361 in the physician office setting compared with $4,981 in the hospital outpatient setting, a 14% difference. A similar recent study by Avalere Health examined 3 years of commercial health plan data to determine the total episode cost for cancer patients treated either in the physician office setting or hospital outpatient department. When adjusted for age, sex, and a prior history of cancer, the average cost of an office-based episode was approximately $28,000 compared with $35,000 for a hospital outpatient department–managed episode, a difference of 24%. The unadjusted difference was 34%.
The relative contribution of the various components of spending to the overall oncology cost curve is challenging to disentangle. Furthermore, proportional and absolute spending is a dynamic process. Analysis of historical data may not sufficiently inform us on how to best address the current drivers of the cost curve in an era of rapid scientific progress and evolving regional models of health delivery.
Mariotto et al have modeled how much cancer spending may increase in the future. This modeling requires estimates regarding incidence, mortality, prevalence, and costs of treating individual patients, particularly in the initial year after diagnosis and last year of life. The authors estimated a range of cost increases between 2010 and 2020. Assuming no increased expense of treating an individual patient, total spending can be estimated to rise by 27% over this 10-year period, largely driven by the aging and growth of the US population. If individualized treatment costs were assumed to increase at 5% annually, total spending would then increase by 66%.
The rising price of new oncologic pharmaceuticals attracts an enormous amount of scrutiny. As stated previously, approximately 10% to 15% of oncology spending is devoted to drugs. The rising rate of spending on oncology drugs appears to be increasing faster than other areas of healthcare.
The evolution in treatment patterns for management of colon cancer over the last 15 years has been a commonly explored example of changes in drug cost.[10,11] Over the past 15 years, the median survival for patients with metastatic colorectal cancer has improved from approximately 8 months to nearly 2 years. This progress has not resulted from use of one particular drug or drug combination but rather from the sequential use of more than one active regimen. The newer drugs used in these regimens include irinotecan, oxaliplatin (Eloxatin), capecitabine (Xeloda), bevacizumab (Avastin), cetuximab (Erbitux), and panitumumab (Vectibix). The cost of 8 weeks of the Mayo regimen using the older drug combination fluorouracil (5-FU)/leucovorin is approximately $63. This cost increased to almost $12,000 with FOLFOX (leucovorin, 5-FU, oxaliplatin), to more than $21,000 with FOLFOX/bevacizumab, and to over $30,000 with FOLFIRI (leucovorin, 5-FU, irinotecan)/cetuximab. Whether the extended survival achieved with modern regimens is commensurate with their added cost is a central concern among many payers, clinicians, and policy makers.
Currently, it is estimated that there are more than 100 new oncology drugs in phase III trials. It is logical to conclude that the proportion of spending on oncology drugs is likely to be a greater driver of cost in the future. Many new agents are expected to have narrower applications as compared with novel agents developed over the last one to two decades, which may mitigate some of the effect on total drug spending.
It is notable that the pharmaceutical industry retains the capability to unilaterally determine the price of drugs coming to market in the US. This contrasts sharply with hospitals and physicians, who historically have had much less control over the price of their services. Additionally, no cost-effectiveness standards are included in the FDA approval process for new drugs. Prices of new therapeutics can easily exceed $5,000 to $8,000 per month.
Nevertheless, new drug development is the sine qua non for progress in medical oncology. While we aggressively seek the benefits of biomedical research, we are ambivalent about paying the price. Biomedical research, and particularly oncology research, is enormously expensive. Potentially as few as 5% of cancer drugs tested in phase I trials ultimately go to market, making oncology the laggard of any therapeutic category in this regard.
How expensive is drug development? DiMasi et al examined this question nearly 10 years ago. Drug development includes several phases: discovery and preclinical development; and phase I, phase II, and phase III trial testing. To calculate drug development costs, confidential survey information was collected from pharmaceutical firms. Development costs were analyzed in expected value terms since it is necessary to include the costs of drugs that fail to come to market with those of the few that do. A total out-of-pocket expense for the various phases of development and testing was determined. Of the four phases, phase III testing is most expensive. Since drug development takes place over several years, the financial outlay for development must be capitalized at a chosen discount rate. This recognizes that the large capital invested could have been profitably invested in other areas during the period of research and development. The authors used a discount rate of 11%.
The full estimate of out-of-pocket expense per new drug was $403 million. When including the 11% discount rate to capitalize costs, the full cost was estimated to be $802 million. The study focused on development data from the 1980s and 1990s. Given the economic slowdown over the last several years, it may be reasonable to apply a lower discount rate, since alternative investments may not be predicted to return as much as 11%. However, it must also be considered that increases in development costs over time could offset the effect of a lower discount rate. Furthermore, development costs are variable across classes of drugs. The cost for oncology drugs has been estimated to be 20% higher compared with the cost average for all drugs.
Novel therapies reach patients in an increasingly haphazard payment-and-coverage environment. Patient out-of-pocket expense is often prohibitive, and coverage is often restricted. Oral drugs are typically covered under pharmacy benefits, with copayment or coinsurance amounts often far too great for patients to bear. Oncology drugs are often placed on the highest tier by private health insurers. Higher-tier coverage is ostensibly designed to encourage use of lower-cost generics. However, generic substitutes for oral oncolytics are often nonexistent. For drugs infused in the office setting, Medicare patients without supplemental secondary insurance policies face copayment amounts of 20%, which can translate to tens of thousands of dollars per year for more expensive regimens.
As many as 10% of patients abandon their first oral oncology drug prescription. This rate increases to 25% when patient cost-sharing exceeds $500. Medicare patients are substantially more likely to have a cost-sharing amount above $500. Other factors associated with abandonment of oral oncology drugs include lower income and increased prescription activity.
Direct costs extend beyond drugs and include procedures, hospitalizations, radiology studies, and professional services. Indirect costs include lost wages due to either reduced hours worked or loss of employment altogether. More than half of personal bankruptcies are caused at least in part by medical illness and medical bills. Three-quarters of those going bankrupt had health insurance and faced average out-of-pocket expenses exceeding $17,000. Hospital bills were most frequently the largest expense. A national survey found that 25% of patients consumed most or all of their savings in dealing with their cancer and its treatment.
A key benefit of health insurance is to protect individuals from risk of medical costs due to potential future illness. Concerns exist that because beneficiaries are not required to pay the full cost of medical care, they may be encouraged to utilize treatments with limited benefit or cost-effectiveness. Cost-sharing arrangements in which patients bear a proportion of costs are structured to mitigate this problem. The data described above would suggest that this is a highly imperfect process, with often tragic consequences for patients and their families. Patients, particularly those with lower income levels, often simply cannot afford the proportion of cost allocated to them.
The benefit of some oncology therapeutics is often viewed as marginal relative to their cost. For example, sipuleucel-T (Provenge) extends median survival of men with hormone-refractory metastatic prostate cancer by 4 months, but at a cost exceeding $90,000 for a course of treatment. It is useful to ask how patients judge the value of therapies when considering coverage and payment issues.
Behavioral science broadly, and prospect theory specifically, have provided constructive models that help to uncover what may matter most to cancer patients. Prospect theory contends that the absolute value of an outcome is not determinative of an individual’s healthcare decisions. Instead, patients value a treatment intervention according to their individual reference point. For someone who expects to live for 1 to 2 years, a 4-month survival benefit may have more meaning than it would for those of us who may expect to live for the next 20 to 40 years. In short, the world begins to look different for patients with cancer. Their reference point changes as they become accustomed to their change in health.
When deciding the value of therapies, those of us without cancer typically try to put ourselves in our patients’ shoes. This process of predicting how we would feel in a given situation is known as affective forecasting. It turns out that we are not very good at this. We tend to exaggerate in both directions. We anticipate a favorable circumstance to be better than it truly is and an unfavorable circumstance to be worse than it truly is. All of this urges some caution for those of us who are not patients when making judgments regarding the value of therapies.
Median gains in overall survival may be a flawed metric in deciding the value of a particular therapy. For many cancer therapies, the concept of small probabilities of large gains may be more appropriate to a patient’s actual decision-making.[4,20] High-dose intereukin-2 for metastatic renal cell carcinoma has been a long-standing example of this proposition. Only a minority of patients benefit from this potentially toxic therapy, but the few that do often achieve survival extending many years. Ipilimumab (Yervoy), a novel agent for advanced melanoma, may be similar in this regard. Whereas median overall survival benefit may be approximately 4 months, patients demonstrating a response to ipilimumab may maintain this response for years. Patients confronting a fatal disease may prefer a “hopeful gamble” over a sure bet. In fact, patients appear to place independent value on a chance of large benefit over a more predictable but more modest improvement in average survival.
In Part II of this article on the cost of oncology care, I will examine what measures can be taken to reduce cost while avoiding or minimizing the sacrifices of quality, access, and innovation. This will include both broad, conceptual viewpoints being proffered and specific, actionable programs that are currently being piloted.
Financial Disclosure:The author has no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
1. Greenberg D, Earle C, Chi-Hui F, et al. When is cancer care cost-effective? A systematic overview of cost-utility analyses in oncology. J Natl Cancer Inst. 2010;102:82-8.
2. Pauly M. Is high and growing spending on cancer treatment and prevention harmful to the United States economy? J Clin Oncol. 2007:25:171-4.
3. Yabroff K, Lund J, Kepka D, Mariotto A. Economic burden of cancer in the United States: estimates, projections, and future research. Cancer Epidemiol Biomarkers Prev. 2011;20:2006-14.
4. Tangka F, Trogdon J, Richardson R, et al. Cancer treatment costs in the United States. Cancer. 2010;116:3477-84.
5. Institute of Medicine: Assessing and improving value in cancer care: workshop summary. Washington, DC, The National Academies Press, 2009. Available from http://www.nap.edu/catalog.php?record_id=12644.
6. Site of service cost differences for Medicare patients receiving chemotherapy. Milliman Client Report. Oct 19, 2011. Available http://www.communityoncology.org/UserFiles/pdfs/milliman-site-of-service-cost-differences-medicare-report.pdf.
7. Total cost of cancer care by site of service: physician office vs outpatient hospital. March 2012. Available http://admin.communityoncology.org/UserFiles/files/Avalere_Cost%20of_Cancer_Care_StudyF(2).pdf.
8. Mariotto A, Yabroff R, Shao Y, et al. Projections of the cost of cancer care in the United States: 2010-2020. J Natl Cancer Inst. 2011;103:117-28.
9. Bach P. Limits on Medicare’s ability to control rising spending on cancer drugs. N Engl J Med. 2009;360: 626-33.
10. Schrag D. The price tag on progress – chemotherapy for colorectal cancer. N Engl J Med. 2004;351:317-19.
11. Karaca-Mandic P, McCullough J, Siddiqui M, et al. Impact of new drugs and biologicals on colorectal cancer treatment and costs. J Oncol Pract. 2011;7(3 Suppl):e30s-7s.
12. DiMasi J, Hansen R, Grabowski H. The price of innovation: new estimates of drug development costs. J Health Econ. 2003;22:151-85.
13. Adams C, Brantner V. Estimating the cost of new drug development: is it really 802 million dollars? Health Aff (Millwood). 2006;25:420-8.
14. Streeter S, Schwartzberg L, Husain N, Johnsrud M. Patient and plan characteristics affecting abandonment of oral oncology prescriptions. J Oncol Pract. 2011;7(3 Suppl):46s-51s.
15. Kim P. Cost of cancer care: the patient perspective. J Clin Oncol. 2007;25:228-32.
16. Himmelstein D, Thorne D, Warren E, Woolhandler S. Medical bankruptcy in the United States, 2007: results of a national study. Am J Med. 2009;122:741-6.
17. USA Today, Kaiser Family Foundation, Harvard School of Public Health: National survey of households affected by cancer. November 2006. Available from http://www.kff.org/kaiserpolls/7590.cfm.
18. Romley JA, Sanchez Y, Penron JR, Goldman DP. Survey results show that adults are willing to pay higher insurance premiums for generous coverage of specialty drugs. Health Aff (Millwood). 2012;31:683-90.
19. Weinfurt K. Value of high-cost cancer care: a behavorial science perspective. J Clin Oncol. 2007; 25:223-7.
20. Lakdawalla D, Romley J, Sanchez Y, et al. How cancer patients value hope and the implications for cost-effectiveness assessments of high-cost cancer therapies. Health Aff (Millwood). 2012;31:676-82.
21. Klapper JA, Downey SG, Smith FO, et al. High-dose interleukin-2 for the treatment of metastatic renal cell carcinoma: a retrospective analysis of response and survival in patients treated in the surgery branch at the National Cancer Institute between 1986 and 2006. Cancer. 2008;113:293.
22. Prieto PA, Yang JC, Sherry RM, et al. CTLA-4 Blockade with ipilimumab: long-term follow up of 177 patients with metastatic melanoma. Clin Cancer Res. 2012;18:2039.