Calculating the Value of Cancer Drugs

A Conversation With Peter B. Bach, MD, MAPP

Get Permission

Peter B. Bach, MD, MAPP

How can we quantify the key issues in value, and what about the value of hope and other constructs that seem hopelessly abstract?

—Peter B. Bach, MD, MAPP

For nearly a decade, Peter B. Bach, MD, MAPP, Director of the Center for Health Policy and Outcomes at Memorial Sloan Kettering Cancer Center in New York, has been a leading voice in sounding the alarm over the escalating cost of cancer drugs and in seeking a solution to the problem. In 2012, Dr. Bach and his colleagues at Memorial, Leonard B. Saltz, MD, Chief, Gastrointestinal Oncology Service and Head, Colorectal Oncology Section, and Robert E. Wittes, MD, then Physician-in-Chief, made headlines when they announced in an op-ed1 in The New York Times that Memorial Sloan Kettering would no longer prescribe ziv-aflibercept (Zaltrap) to its patients with advanced colorectal cancer because the drug cost more than twice as much—$11,063 for a month of treatment—as bevacizumab (Avastin) but provided the same survival benefit, prolonging patients’ lives for a median of 1.4 months. (A few weeks later, Sanofi, the manufacturer of ziv-aflibercept, reduced the cost of the drug by 50%.)

The following year, the chorus of protests over the high cost of cancer drugs grew louder after an international group of 120 experts in chronic myeloid leukemia authored an article in Blood,2 arguing that it was important to lower the prices of cancer drugs so more patients could afford them.

In an article published earlier this year,3 Dr. Bach and his colleagues examined the history of cancer drug pricing in the United States and found that the “average price of anticancer drugs, adjusted for inflation and health benefits, increased by 10% annually—or an average of $8,500 per year—from 1995 to 2013.”

“The notion that drug prices are linked to anything coherent like value or benefits has been debunked,” said Dr. Bach. “Drug pricing is illogical because as drugs gained competitors, their prices went up; as drugs gained market share, their prices went up; and as they gained new indications, their prices went up. There is no relationship between the cost of drugs and benefit or how novel the compound is. Pricing is just based on what the market will bear.”

Calculating Drug Value

ASCO has been investigating how to define and ensure the delivery of high-value oncology care for more than 10 years, and in June, the Society released details of its findings in a conceptual framework for assessing value in cancer care in the Journal of Clinical Oncology,4 a move praised by Dr. Bach. He is incorporating some of the core principles used in ASCO’s value framework in an interactive calculator he has developed called ­DrugAbacus.

DrugAbacus ( is a research tool that Dr. Bach has created to compare the cost of 54 oncology drugs approved by the U.S. Food and Drug Administration (FDA) since 2001 with what their prices would be if factors such as drug side effects, survival benefit, and whether the treatment was expensive to develop were added to the equation to determine actual value-based pricing.

The ASCO Post talked with Dr. Bach about the development of the ­DrugAbacus and his pursuit of value-based oncology drug pricing.

Calculating Value

Why did you create the DrugAbacus, and what is its purpose?

The Drug Abacus is essentially a proof-of-principle research tool. I developed it because it seemed like the appropriate next step in trying to determine value in drug pricing. The issue that prices are rising very rapidly has been pretty clearly documented by our group at Memorial Sloan Kettering and by folks at other institutions. And I’ve published a few articles, as have a number of other people, raising questions about how drug prices might be determined more coherently.

Drug pricing should have some relation to benefit, but it doesn’t. It should have some relation to how novel the compound is, but it doesn’t. In some articles I’ve written, I’ve done head-to-head comparisons to determine a rationale for price disparity. How can it be that ceritinib (Zykadia) for late-stage lung cancer costs about $1,700 more than crizotinib (Xalkori)—$13,200 per month vs $11,500 per month, respectively?

The notion that we should be thinking about value when it comes to drug pricing is becoming entrenched, but there is the concern of how to define value or apply it in the real world of caring for our patients. Everybody’s stake in this effort is different. How can we quantify the key issues in value, and what about the value of hope and other constructs that seem hopelessly abstract?

In the course of listening to concerns about what to take into consideration when determining value, I thought that if we could wrangle these notions into empiric structures, maybe we could come up with a value equation. Since there was no disagreement on which components go into the equation but, rather, how much they matter, I converted the information into a mathematical way of thinking about value in drug pricing.

Interactive Tool

How does the DrugAbacus work?

The home page of the DrugAbacus shows an image and price of pembrolizumab (Keytruda, $2,500) and one of imatinib (Gleevec, $9,000) and asks: Are these the right prices; should they cost more or less? Inside the DrugAbacus is all the information that might be relevant to the value of the drug, including efficacy, toxicity, novelty, cost of development, disease rarity, and population burden of disease.

The user can then decide how much each of these factors should matter in a drug’s value; the abacus calculates the user’s “abacus” price and compares it to the drug’s actual price. The users of the DrugAbacus are people who are concerned that we need a better model for drug pricing.

This version of the DrugAbacus offers a very rough draft of how we could put a value on drug pricing. I think of it as a sort of version 0.1, but going from 0 to 0.1 matters because it takes us from asking how could we do this to this is how we do it—now let’s work out the details. There is hard work ahead, but there is coherence with this model, and it gives us a start.

I am not making a judgment about what is the right way to go about this. The DrugAbacus is a proof of concept that helps users create a value equation and actually map drugs to it to come up with a value-based price, but I don’t expect it to be used at the point of care. It is a policy tool aimed at addressing the concern that the drug-pricing system is completely broken right now and endeavors to help find what a better pricing system might look like.

Patients Left Behind

I read that you were inspired to develop the DrugAbacus after hearing from patients who didn’t fill their prescriptions because they couldn’t afford the insurance copayments. Please talk about that experience.

I still harbor the belief that the reason we are all in the business of biomedical innovation is to benefit patients. I think the reality of that marketplace, however, is that it largely benefits a number of parties involved in the process between the scientists and the point of care but that the end users—the patients—are regularly left behind.

The goal of my research and the reason others are attempting to clarify drug value and pricing is to reduce human suffering. Undoubtedly, you can blame insurance cost-sharing structures for making it difficult for patients to afford these drugs, and I do. But insurance cost-sharing structures are a consequence of high drug prices, and if you tack on cost-sharing charges for a drug like blinatumomab (Blincyto), which costs $180,000 and has to be administered in the hospital, you ensure that some patients with acute lymphoblastic leukemia will not get the drug.

That scenario is also true of imatinib, the most important drug to be developed in my professional lifetime. Novartis has raised the price of the drug nearly threefold since it was approved by the FDA in 2001 for chronic myeloid leukemia—and more recently in the treatment of childhood acute lymphoblastic leukemia—and some patients aren’t getting their prescription filled because of insurance copayments. I can’t look at that situation and say our system is working.

In 2014, the FDA approved 10 drugs for cancer treatment and that many or more will be approved this year, which is great news. But if these new treatments don’t get to the patients we are developing them for because they can’t afford them, something is broken, and what’s broken is drug pricing. So the focus has to be on solving this problem.

Total Price Flexibility

What has been the reaction from drug manufacturers to your concerns about drug pricing?

The reality is that many of these companies are full of people who believe in the same mission that I just articulated and believe that they are “cannibalizing their young” in the pursuit of near-term profits. Some companies are structured around capitalization of the current broken health-care system, and other companies are in business for the long haul, and they have very different reactions to the goal of trying to rationalize the pricing system into a sustainable model that benefits patients.

The companies that are in business for the long haul and have been around for a long time are believers in their model of innovation and are fully on board with the notion that we need indication- and performance-specific drug pricing. That is a model that I developed and one that Express Scripts Holding Company is now using to set pricing for some cancer drugs based on how well they work.

But there are other companies that are in the arbitrage business. They buy up an asset, try to get it cheap, use market leverage to jack up the price as much as possible without adding any value, and use tools to wall off competition and the entry of generics into the marketplace. Their goal is to maximize their near-term profits, and they have very little interest in any other approach to pricing.

The question I keep asking, and I’ve been asking it ever since Sanofi lowered its price for [ziv-aflibercept] by 50% after we wrote one op-ed in The New York Times, is why are we acting like these drug prices are real? The prices are products of markets, so we have indulged pricing flexibility, and our marketplace tolerates unlimited pricing indiscretion by the pharmaceutical industry. Point to another industry where an op-ed can lead to a 50% price reduction. That response proved total price flexibility.

The questions we should be asking are how do we plan to pay for all the drugs that will come out over the next 20 years, and how can drugs like sofosbuvir (Sovaldi) for the treatment of hepatitis C cost $84,000 per 12-week course in the first place? And the answers, respectively, are, we can’t, and because they can. ■

Disclosure: Dr. Bach reported no potential conflicts of interest.


1. Bach PB, Saltz LB, Wittes RE: In cancer care, cost matters. The New York Times. October 14, 2012. Available at Accessed July 27, 2015.

2. Experts in Chronic Myeloid Leukemia: The price of drugs for chronic myeloid leukemia (CML) is a reflection of the unsustainable prices of cancer drugs: From the perspective of a large group of CML experts. Blood 121:4439-4442, 2013.

3. Howard DH, Bach PB, Berndt ER, et al: Pricing in the market for anticancer drugs. J Econ Perspect 29:139-162, 2015.

4. Schnipper LE, Davidson NE, Wollins DS, et al: American Society of Clinical Oncology statement: A conceptual framework to assess the value of cancer treatment options. J Clin Oncol 33:2563-2577, 2015.