July 2017 Edition Vol.11, Issue 7

Latest in Cost and Value Discussion at ASCO ’17

By Christina Bennett, MS

Alongside the breadth of clinical research reported at the 2017 American Society of Clinical Oncology Annual Meeting are cost and value, aspects to care that are gaining attention and posing to be formidable challenges. Highlighted here are four important studies from the annual meeting that report the latest cost and value findings.


Opportunity for cost savings

In the United States, the cost of Keytruda (pembrolizumab) for a certain group of lung cancer patients could be cut by an estimated $0.8 billion—if dosing behavior is changed from fixed to personalized, according to study results now published in the Journal of the National Cancer Institute.1 Although the change sounds simple, it is anything but: The study calls for legislative action and additional effort in the clinic to make these cost savings possible.

Pembrolizumab is approved for several indications, one being first-line treatment of PD-L1–expressing metastatic non–small cell lung cancer (NSCLC). Currently the indication is for a fixed dose of 200 mg every three weeks because that was the dose used in the pivotal KEYNOTE 024 trial that garnered the FDA approval.2

KEYNOTE 024 used a fixed dose despite pembrolizumab first being evaluated in clinical trials at a weight-based—or personalized—dose and equivalency studies showing that 2 mg/kg and 10 mg/kg have the same efficacy and safety, supporting the use of 2 mg/kg.The switch from weight-based to fixed dosing was the result of policy changes within the manufacturer, Merck.

The reason for the changes was “unclear,” according to lead study author Daniel Goldstein, MD, senior physician in medical oncology at Rabin Medical Center, Israel. If administered at 2 mg/kg, less drug is usually needed. According to the study, the average metastatic NSCLC patient in the United States weighs 76 kg and therefore needs only 152 mg of pembrolizumab, not 200 mg.

To calculate total cost of pembrolizumab for both dosing methods, researchers incorporated several variables into their base case analysis, including target population number, drug cost, patient weight ranges, and total amount of drug used. Total annual cost in the United States for fixed dosing was $3,440,127,429 and $2,614,496,846 for personalized, translating to an estimated annual saving of $825,630,583—or 25%. These savings could impact payers and out-of-pocket costs for patients, depending on the patient’s plan. Personalized, not fixed, dosing should be used for these lung cancer patients, the authors concluded.

But one challenge to implementing personalized dosing and making these cost savings realized in the United States is the lack of a cost-effective vial size.

Originally, Merck provided pembrolizumab in 50-mg vials in the US market, but then in 2015, they replaced 50-mg vials with 100-mg vials. “[50-mg vials] are still available in the European market place, but there is a plan to remove them from there as well,” Dr. Goldstein said. Once opened, a vial lasts for only 24 hours refrigerated. “By removing the 50-mg vials, it makes it more difficult to give these vials without having wastage,” Dr. Goldstein said. The decision to remove 50-mg vials from the US market and the potential removal from the Europe market, according to Dr. Goldstein, is “clearly a business decision.” Adding 50-mg vials back into the US market would help but not necessarily fix this scenario. “What we would argue for pembrolizumab is, yes, bring back the 50-mg vials—and also give us 10-mg vials,” Dr. Goldstein said.

A few actions can be taken to bring personalized dosing into the clinic and thereby cut costs. One is to propose legislation that requires manufacturers to provide appropriate vial sizes. Another is to convince the FDA to change the indication from 200-mg fixed dosing to 2 mg/kg weight-based. In addition, clinicians can practice effective vial sharing to prevent wastage. At smaller practices, where low numbers of patients receiving pembrolizumab can make this difficult, study authors recommended scheduling infusions only on certain days of the week. “Physicians and prescribers should be responsible stewards of societal resources to prescribe only the dose that is required to provide the efficacy that is required,” Dr. Goldstein stressed.


Call for rationale gene panel reimbursement policies

Next-generation sequencing (NGS)—an integral part of precision medicine—is used to inform treatment decisions; however, whether NGS is reimbursed varies among payers, making access to this personalized diagnostic test difficult. The results of a study further demonstrate how rationale reimbursement policies are needed.“While drug costs are the greatest challenge in personalized—or precision—medicine, cost and reimbursement are substantial barriers to genomic profiling with NGS,” said lead study author Thomas Brown, MD, MBA, executive director of Swedish Cancer Institute in Seattle, WA. Dr. Brown said the list price for NGS is currently $4,000 and, for patients paying out-of-pocket, the discounted price is $1,800.

Study researchers prospectively evaluated reimbursement amount and frequency for a 68-gene panel test among a cohort of 602 participants with a range of solid tumor types and from all stages of cancer. The 68 gene mutations had actionability, meaning a targeted therapy could be paired with the mutation. Reimbursement was evaluated by payer type, patient age, patient gender, disease site, disease stage, and actionability of gene panel results. “Gene alterations were categorized as actionable, that is associated with on-label indication, or applicable, that is associated with an off-label and/or clinical trial indication,” Dr. Brown said.

A total of 568 patients (95%) had a gene alteration, and of those about three-quarters had an actionable or applicable mutation and a quarter had a mutation of unknown significance. For both private insurance and Medicare, health maintenance organizations (HMOs) provided higher and more frequent reimbursement than non-HMO programs. For non-HMO Medicare, no reimbursement was given; most Medicare patients in this study (75%) had non-HMO coverage. This makes sense, Dr. Brown explained, because current policy for non-HMO Medicare is to not reimburse for gene panels; reimbursement is only done for Medicare HMO.

“Most intriguing,” according to Gary Lyman, MD, MPH, co-director of the Hutchinson Institute for Cancer Outcomes Research, was the finding that reimbursement frequency and payment were inversely associated with actionability, meaning reimbursement was more likely for test results that had no clinical utility than for results that did. He stressed the need for validation of this finding in other studies.

Dr. Brown too found this finding “surprising” and was not convinced of this relationship either because, although the study was prospective, this analysis was retrospective and the reason some patients were denied reimbursement could have been for having had prior single gene or panel testing. Prior testing status was unknown at the time of analysis.

Olivier Elemento, PhD, associate director of the Institute for Computational Biomedicine at Weill Cornell Medicine, agreed that prior gene testing could explain this inverse relationship. Dr. Brown and colleagues plan to determine prior testing status before publication of the study results. “It’s definitely a very interesting study,” Dr. Elemento said. The “most interesting aspect” is that a proportion of patients did receive reimbursement “because a couple of years ago it was 0%.” About one-third of patients received some reimbursement.

“These data demonstrate the need for a rationale, transparent, consistent reimbursement policy or set of policies,” Dr. Brown said. “That doesn’t exist now.”


Drug cost and value disconnect

As cancer drug costs trend upward and value-based care enters the discussion, a clear connection between drug costs and their value is needed. A study found no such correlation to exist between out-of-pocket (OOP) costs for oral oncolytics and their value.“This study was very important to do,” said Eric Nadler, MD, MPP, medical oncologist at Texas Oncology–Baylor Charles A. Sammons Cancer Center. “It’s important to understand how pharmacoeconomics and value are playing out within the oncology space.” Study researchers retrospectively evaluated insurance claims from a commercial insurance database for cancer patients who had filled an oral oncolytic prescription between 2007 and 2014. The researchers then calculated monthly OOP costs.

For this study, value was defined by a single metric, drug efficacy—that is, progression-free survival and overall survival. Other contributors to value, such as drug safety and ease of use, were not included. Oral oncolytics were categorized into groups according to the efficacy stated in their FDA-approved indication. There was also group for off-label use. In total, 44,113 patients and 731,354 filled prescriptions for oral oncolytics were extracted from the database. The mean OOP payment each month was $83 (standard deviation, $353), ranging from $0 to $14,167. The median was $26.60. More than half of patients (55%) received either Gleevec (imatinib) or Revlimid (lenalidomide).

“Those two drugs are some of the most beneficial drugs in regard to survival and progression-free survival and toxicity that we have,” Dr. Nadler said, adding that this characteristic makes the study particularly interesting.

“The punchline of this study is really that marginal out-of-pocket payment didn’t differ in an obvious way for drug indication combinations in almost any survival group versus the marginal out-of-pocket payment for drugs prescribed off-label,” said lead study author Lisa Rotenstein, MD, MBA, resident physician in Internal Medicine at Brigham and Women’s Hospital.

The lack of difference means the amount patients pay for a drug indicated to have high clinical value could be the same as that for a drug prescribed off-label.

“We also found that more recent year of fill was associated with lower out-of-pocket spending, meaning that patients are paying less out-of-pocket over time for targeted oral oncolytics, and that patients’ out-of-pocket payments are in line with what is being paid overall for the drug—if there’s higher total spending for a drug, it’s associated with higher out-of-pocket spending,” Dr. Rotenstein said. The data suggest that despite increasing discussions of value-based payment for drugs in the popular press by payers, a trend towards value-based payment for patients hasn’t yet happened, Dr. Rotenstein said.

However, Dr. Nadler asked, was the study actually measuring the true OOP cost to the patient, or were some of these patients receiving rebates from drug assistance programs, thereby mitigating some of that cost?

Study authors said the information about the use of drug assistance programs was not available in their data. Nevertheless, their data highlight what was paid—either by the patient or by a drug assistance program—and their conclusion remains the same: Cost-sharing programs are not currently designed to encourage more use of higher-value drugs and less use of lower-value drugs.


Talking cost

Patients and providers are becoming increasingly sensitive to cost of care. Despite this, bringing cost into the appointment conversation can be met with a range of barriers and how to overcome these barriers is not straightforward, a study showed.The study was done qualitatively to identify barriers and facilitators to discussing cost. The study involved conducting a series of semi-structured interviews with 22 patients and 19 providers. Patients were predominantly female, white, elderly, low-income, and from community practices. Twelve providers were from community practices and the remaining 7 were from academic medical centers.

“We found that there were several pragmatic barriers to conversations about cost, and probably the number one was the difficulty in providing accurate out-of-pocket cost estimates to patients in a timely manner,” said lead study author Erin Aakhus, MD, instructor at University of Pennsylvania, Perelman School of Medicine. Another barrier was time constraints during an appointment.

Dr. Aakhus said that one of the more “surprising” barriers was the “very strong aversion” on both the patient and the provider side to making health cost trade-offs. “There was a strong feeling that engaging in cost discussions might lead to more trade-offs and inferior treatments for some patients, and kind of in the same vein, the patients and providers were very concerned that increasing the frequency of discussions about cost might lead to socioeconomic or sociodemographic discrimination and then worsening of disparities.”

In addition, providers and patients were concerned about financial conflicts of interest and how that influences provider recommendations.

“We did ask patients and doctors to describe how they think this conversation about cost would happen in an ideal world,” Dr. Aakhus said. The result was a “significant heterogeneity of opinions.” For instance, many patients wanted to discuss cost early on in care while some wanted to wait until later. Also, some patients were comfortable talking about cost with their physician while most preferred a third-party financial expert. The study researchers are doing a follow-up survey on a larger population to generate more representative data.



  1. Goldstein DS, Gordon N, Davidescu M, et al. A phamacoeconomic analysis of personalized dosing versus fixed dosing of pembrolizumab in first-line PD-L1 positive non-small cell lung cancer. J Natl Cancer Inst. 2017;109:djx063.
  2. Pembrolizumab (KEYTRUDA) Checkpoint Inhibitor. https://www.fda.gov/drugs/informationondrugs/approveddrugs/ucm526430.htm. Accessed June 12, 2017.
  3. Garon EB, Rizvi NA, Hui R, et al. Pembrolizumab for the treatment of nonsmall-cell lung cancer. N Engl J Med. 2015;372:2018-2028.
  4. Brown TD, Tameishi M, Liu X, et al. Analysis of reimbursement (R) for next generation sequencing (NGS) on patients’ tumors in the context of a personalized medicine program. J Clin Oncol. 2017;35:(suppl; abstr 6506).
  5. Rotenstein L, Dusetzina S, Keating NL. Relationship between out-of-pocket spending and drug value for oral oncolytics reimbursed by commercial insurers: 2007-2014. J Clin Oncol. 2017;35:(suppl; abstr 6524).
  6. Aakhus E, Rosenstein A, Joffe S, Bradbury AR. Implementing cost transparency in oncology: A qualitative study of barriers, facilitators, and patient preferences. J Clin Oncol. 2017;35:(suppl; abstr 6597).

Post a Comment

OBR Archives

To view previous issues of OBR green you can visit our archives. The entire library of OBR green articles is searchable.