publication date: Oct. 30, 2015

How Much is a Drug Worth? A Provocative Model Puts a Price on Benefit 


A matrix of therapies for non-small cell lung cancer generated by the PACE Continuous Innovation Indicators.

Source: Silvia Paddock/Rose Li and Associates and Jacqueline Ferguson/Lilly Oncology




Eli Lilly & Co. didn’t ask Dan Goldstein, an oncologist at the Winship Cancer Institute at Emory University, to price their drugs, but he volunteered his services anyway.

Indeed, Lilly Oncology is unlikely in the extreme to concur with the price he proposed for necitumumab, a front-line treatment for locally advanced or metastatic squamous non-small cell lung cancer.

Necitumumab, which at this writing is awaiting FDA approval, would be used in combination with a doublet treatment of gemcitabine and cisplatin. This Biologics License Application is all the more important because the treatment of squamous NSCLC hasn’t changed in over 15 years.

In a paper recently published in JAMA Oncology, Goldstein argued that the drug could be worth as little as $563 per month. For perspective, the average launch price for cancer agents is $12,000 a month.

Realistic or not, Goldstein’s paper, “Establishing a Value-Based Cost,” is the first to prospectively model drug prices based on clinical benefit.

At a glance, Goldstein’s model might appear similar to other value assessment tools that have been proposed by researchers and cancer organizations:

• The American Society of Clinical Oncology’s value framework, published in June this year, quantifies clinical benefit, side effects and cost as components of value (The Cancer Letter, June 26).

• Peter Bach’s DrugAbacus, an online tool that allows users to assess the value of 54 cancer drugs that have received U.S. approval since 2001 (The Cancer Letter, June 19).

• The National Comprehensive Cancer Network’s Evidence Blocks, a visual representation of five key value measures that reflects the evolving standard of care for a given disease state (The Cancer Letter, Oct. 16).

These approaches use existing data on the components of value—efficacy, safety, quality of evidence, and affordability—to educate providers and patients. ASCO’s framework, DrugAbacus and NCCN Evidence Blocks incorporate cost analyses, but none suggest prices for drugs before they are approved.

Goldstein does just that. His model differs by solving the cost equation ahead of a drug’s launch using trial results.

However, Lilly Oncology officials aren’t persuaded. Goldstein’s preemptive cost estimates for necitumumab are deflated, they said to The Cancer Letter. “Had Dr. Goldstein chosen more real-world standard of care, that wouldn’t have been the conclusion of the study,” said Jax Ferguson, senior director of global oncology corporate affairs at Lilly Oncology.

Lilly Oncology earlier this year published its own value tool, called PACE Continuous Innovation Indicators. While the model doesn’t deal with cost, it aggregates oncology data in an effort to understand the evolution of value by visualizing progress in 12 disease types over 40 years.

A conversation with Lilly Oncology officials can be found here.

Goldstein’s study is a part of a larger movement to understand the value of drugs, pricing experts say.

“Goldstein and his team’s work is important, because it illustrates the value of prospective economic analyses for prospectively estimating value-based pricing,” said Rena Conti, an assistant professor of health policy and economics in the Department of Pediatrics and Public Health Sciences at the University of Chicago.

Some elements of value aren’t included in Goldstein’s estimates, said Conti, who published an article on new pricing policies on Health Affairs Blog earlier this year.

“Goldstein and his team’s work is one tool for helping payer’ set coverage and patient cost-sharing requirements,” Conti said to The Cancer Letter.

Value-based pricing is becoming a central feature in setting coverage decisions.

“On a large scale, the Centers for Medicare and Medicaid Services has decided to move to value-based payments, which will necessarily incorporate costs,” said Peter Bach, a pulmonologist, health systems researcher, and director of the Center for Health Policy and Outcomes at Memorial Sloan Kettering Cancer Center.

“We’re going to target particular conditions like oncology. We have to understand the landscape of costs in terms of how they’re linked to particular diseases.”

But what is value? And should drugs be priced strictly according to value?

The cost equation is complicated: drug prices aside, researchers making value calculations have to take into account the benefits, harms and toxicities of any particular drug.

“Drug prices are also a constraint that researchers deal with, as they do with the challenges of biology,” Bach said at the 2015 ASCO annual meeting in May. “Why not close this loop? If [the other variables] are equal except for cost, why not just say ‘no’ to a higher priced agent?”


How to Price Necitumumab

Enter Goldstein, armed with the idea to use necitumumab—a highly anticipated lung cancer drug—as a case study.

“What I was interested in was trying to see if we could develop a price for a drug before the drug comes to the market,” Goldstein said to The Cancer Letter. “So we heard about necitumumab and that it was a drug that was likely to get approved by the FDA, so then we used the data from the clinical trial to essentially develop a value-based price for the drug.”

But should we be solving for price anyway?

“I think we do need to do exactly what Dan Goldstein just presented,” Bach said, discussing Goldstein’s presentation at the ASCO 2015 annual meeting. “He actually goes ahead and solves for price.”

Necitumumab, a first-line therapy for metastatic squamous non-small cell lung cancer, is worth no more than $1,300 per cycle, according to Goldstein’s study. Based on the value it provides, necitumumab could fetch as little as $563 per cycle.

“What we’re trying to do is trying to link the benefit to the price,” Goldstein said. “For example, a drug like necitumumab, because there’s a low level of efficacy, it should have a relatively low price.”

This means that, at a conservative estimate of $50,000 per quality-adjusted life year, necitumumab should cost $500 a month, Bach said. At $200,000 per QALY, it’s $2,700 a month.

“This sounds crazy, a drug approved maybe this year, for only $2,700 a month?” Bach said. “But let me remind you that that is close to Gleevec’s launch price in 2001—in today’s dollars—and Gleevec is many times more effective. But the truth is it sounds crazy today.”

Based on the contemporary average launch price for cancer drugs, necitumumab would cost $900,000 per QALY, Bach said.

Necitumumab, when used in combination with gemcitabine and cisplatin, is the first regimen in the first-line setting to show significant improvement in overall survival over chemotherapy alone for patients with advanced squamous NSCLC.

No first-line therapies exist for this indication—necitumumab promises to fill an unmet need in a disease with a five-year survival rate of less than 5 percent.

The drug received overwhelmingly positive comments from the FDA Oncologic Drugs Advisory Committee July 9 (The Cancer Letter, July 10).

No formal vote was taken by the committee—in what appears to be the new FDA approach to getting advice on oncology treatments—but The Cancer Letter’s analysis of the public comments made by ODAC members suggests that, had a vote been tallied, necitumumab would have received an overwhelming 11:1 vote in favor of approval.

FDA is expected to announce its decision later this year.


Goldstein: Lilly Drug Does Not Deserve High Prices

How does all this translate into value?

Necitumumab provides a “minimal” incremental benefit for patients who have metastatic squamous NSCLC, said Goldstein.

“They’re going to die anyway, but it basically just extends life by about six weeks,” Goldstein said to The Cancer Letter.

Goldstein has authored multiple studies on the cost effectiveness of different cancer treatments—many new agents are not very cost effective, he concluded.

“Value is benefit divided by cost, and cost is an essential component when discussing value,” Goldstein said. “For instance, a drug that is developed for an unmet need is more beneficial than a drug that is developed for a need that has already been met. If the two drugs are priced the same, the drug for the unmet need is therefore more valuable.”

Necitumumab’s efficacy was tested in a phase III trial, called SQUIRE, that randomized 1,093 patients to receive necitumumab plus gemcitabine and cisplatin (n=545) or gemcitabine and cisplatin (n=548).

Necitumumab is a second-generation, recombinant human monoclonal immunoglobin G1 EGFR antibody that binds to the extracellular domain of the human epidermal growth factor receptor and blocks interaction between EGFR and its ligands.

The trial demonstrated that adding necitumumab to chemotherapy in the first-line setting increased median overall survival by 1.6 months and progression-free survival by 0.2 months.

Two drugs were recently approved for the squamous NSCLC indication, but not in the front-line and not in combination with chemotherapy.

The new second-line drugs are:

• Ramucirumab, approved in December 2014 in combination with docetaxel, for treatment of metastatic NSCLC (both squamous and non-squamous) with disease progression on or after platinum-based chemotherapy, and

• Nivolumab, approved in March 2015 for patients with squamous NSCLC with progression on or after platinum-based chemotherapy.

For necitumumab, Goldstein and his team of investigators developed a Markov model using data from multiple sources, including the SQUIRE trial, to evaluate the costs and patient life expectancies associated with each regimen.

In the analysis, patients were modeled to receive gemcitabine and cisplatin for six cycles or gemcitabine, cisplatin and necitumumab for six cycles, followed by maintenance necitumumab.

The cost inputs included drug costs, based on the Medicare average sale prices, and costs for drug administration and management of adverse events, based on Medicare reimbursement rates.

“With value-based pricing, if it were a drug that’s essentially a game-changer drug—something that really cures the disease or really changes the course of the disease—drugs like that would still carry a high price,” Goldstein said.

“Drugs that come to mind that would warrant a high price are drugs such as imatinib or the new drugs to treat hepatitis C. These drugs truly change the course of a disease. They’re truly game changing, and they warrant high prices.

“And that’s the good thing, in my mind, because value-based pricing will continue to incentivize the development of truly game-changing innovations.”

Goldstein establishes a threshold for what is cost effective with a system called the Incremental Cost Effective Ratio, or ICER—that is, the number of dollars it costs to gain one year of life or one year of quality-adjusted life by a specific treatment.

“So the quality of quality-adjusted life here is taking into account of the fact that patients with advanced cancer, their quality of life is not as good as patients who don’t have cancer,” Goldstein said. “There is no absolutely accepted value in the United States, but it is somewhere between $50,000 to $150,000 to gain one quality-adjusted life year.

“And so we use that as essentially the end result, and work backwards to find out how much necitumumab would have to cost in order for it to be cost effective by those thresholds.”


Eli Lilly: Goldstein Model Not “Real-World”

Goldstein’s study is methodologically sound, Lilly Oncology’s Ferguson said.

“However, the comparator that he chose in his clinical trial—he didn’t actually take into account the real range of treatment options available and reflect real-world choices,” Ferguson said to The Cancer Letter. “So unfortunately he chose gemcitabine and cisplatin and generic comparators, which resulted in the deflated annual costs and per cycle costs of necitumumab.”

Using a different set of comparators would not have made a difference, Goldstein said.

“Our sensitivity analyses make the study extremely robust—such criticism does not hold water, and it’s an excuse to avoid the real issue—that prices should be related to the benefit that a drug provides,” Goldstein said.

“If we had used—as the comparator arm—other regimens aside from gemcitabine and cisplatin, such as carboplatin and paclitaxel, it would have had essentially zero impact on the model since the cost of carboplatin and paclitaxel is very similar to gemcitabine and cisplatin, and this wouldn’t have impacted the model results significantly.

“Our study was very robust in that it had multivariate sensitivity analyses, and it took account of any variation in the cost of the comparator arm, so this potential difference would not have affected the results of this study at all.”

Goldstein said he hopes his findings will help payers, the pharmaceutical industry and the U.S. government address the issue of high cancer drug prices.

“Currently, FDA has no mandate to assess cost of value. However, it would appear that that some governmental intervention may be required,” Goldstein said. “Some people have argued that we should just bring down prices in kind of a simplistic way.

“However, I feel that we shouldn’t just simply bring down prices but more that the price should be linked to the benefit that the drug provides, and that’s a value-based price.

“I hope that industry will take notice of this study and recognize the importance of it—and if and when necitumumab gets approved, it will be priced appropriately by the manufacturer.

“I’m also hoping that governmental institutions will take note in recognizing the importance of developing some type of independent institution to aid in the development of drug prices prior to them reaching the marketplace.”


Bach: Answers that Lead to Questions

Researchers studying value-based pricing have—in this era—somehow decided on a hierarchy of comparative decision-making factors in oncology, MSKCC’s Bach said in his critique of Goldstein’s presentation at the 2015 ASCO annual meeting.

“That begins with efficacy, and then toxicity, and then costs,” Bach said. “I don’t know how we ended up there.”

There is no rationale for abiding by that hierarchy when, like Goldstein, researchers can calculate prices to solve the cost equation, Bach said.

“Does efficacy always drown out cost? Would we really pay an infinite amount for a microscopic benefit?”

What are the next steps?

Before value-based pricing can be used a comprehensive tool for understanding costs, there are many questions that need to be answered, Bach said.

For instance, researchers have to methodically quantify numerous confounding variables, and decide on which measurements are most reliable.

“Do we use the average life expectancy or do we start looking for this tail?” Bach said. “Do we look for the cures? Do we look for the lottery winners? Is a drug worth a lot more because there’s a small group that might benefit a lot?

“Do we discount for toxicities? Or do we just use QALY? What about cost replacement? If a drug prevents or adds a hospital day, does that go into the price?

“Should we pay more for rare conditions? Should we pay more for unmet needs in populations or diseases that have no treatment? What about first-in-class? Does new science get a reward?

“What about the cost of R&D? If it costs a lot to develop a drug, should a company be able to recoup that to some extent, because it’s hard to find populations?

“Are we paying for the past or the future? Are we paying so that people keep rushing into the void and filling needs that we need for our patients? Are we going to pay extra for that? Or are we going to just say, ‘whatever we have today, let’s price it out.’

“For measuring benefits, should we use life years or QALYs?” Bach said. “Dan Goldstein showed both of them. I’m actually not sure which way to go. Do we use the mean benefit? The median underestimates in general.

“I mean these as open questions, not as cynical statements. They are answerable, but we’re not there now.”

Copyright (c) 2021 The Cancer Letter Inc.