The Cost and Value of Cancer Medicine

Thomas Roberts, MD, Farallon Capital Management, San Francisco, California, USA


Fellow Summary


Authored by Rebecca Epperly

Dr. Roberts’ talk challenged participants to consider how healthcare is valued. He started by encouraging us to expect adversity, focus objectively on our ideas, and take risks when the ideas are worthy. He described the current economic landscape of drug development. Regulatory advances including breakthrough designation, accelerated approval, fast track, priority review, orphan drug designation, and special protocol assessment are mechanisms that have been in place prior to 2017 and which have helped streamline the drug approval process. Since 2017, regulatory innovation has focused on adaptive clinical trial design, digital innovation, generic competition, opioid abuse, implementation of the 21st century cures act, and supporting FDA culture change. These innovations have contributed to a median approval time of less than 10 months from submission. There are now more than 100 drugs whose mechanism involves targeting specific molecular subgroups. Because targeted therapeutics can provide an exceptional response in a small subset of patients but have no activity in another, they are challenging to value.

Dr. Roberts highlighted the following structural challenges impacting the field which contribute to increased drug prices:

  • With the expectation for a constant rate of innovation, what role should “fast followers” or “me too” drugs play?
  • Drug development remains risky in an atmosphere where only about 10% of drug studies in the phase I setting ultimately get approved. Those that do reach approval require an average of 9 years and $2.6 billion dollars in development.
  • Large pharmaceutical companies struggle for continuous innovation.
  • Research and development is now focused in small companies, with only 10% of new compounds coming from large pharmaceutical companies. In contrast, large companies used to generate 2/3 of new compounds.
  • The pipeline is currently crowded, particularly in the fields of oncology and inflammatory disorders.
  • The field is trending towards smaller and more targeted drug launches, with about 1/3 of new drugs qualifying as orphan drugs. While these drugs address a key need, valuation is challenged by lower anticipated peak sales.

Dr. Roberts cited Michael Porter, who defines value in healthcare as healthcare outcomes that matter to patients divided by the cost of delivering those outcomes. The price of a new drug does not necessarily correlate with improved survival. For example, in 1995 the cost of a new drug averaged $54,000/year of life gained while in 2017 the cost was $475,000/year of life gained. In the current economic and political landscape, several methods have been proposed for re-shaping the way in which value is applied in healthcare pricing. This includes value-based pricing models tying reimbursement to outcome and the international price index, which sets prices based on an international benchmark. Dr. Roberts encouraged that in order to address value in drug development moving forward, we need to use real world data incorporating data science and informatics, consider new payment models, and strive for leadership in value.


Fellow Summary


Authored by Brian Grieb

As a researcher, it is easy to only concern yourself with Western blots and Kaplan-Meier curves.  However, Dr. Roberts challenged us to consider the financial implications of drug development costs and pricing.  The financial burden of developing Novel Molecular Entities (NMEs) is undeniable.  Pharmaceutical spending on research and development (R&D) increased from $12 billion in 1995 to $65 billion in 2017.  Drug development is also risky.  Approximately 21 in 23 NMEs tested in phase I clinical trials are never approved by the FDA, not to mention the many pre-clinical NMEs that never reach human testing.

Sunk costs as well as the sheer expense of successful drug development and production are used to justify increasing cost of medications.  While some expensive medications have improved patient outcomes, they do so with decreasing efficiency.  The average cost per year of life saved in 1995 was $54,100, doubling to $139,100 in 2005, and ballooning to $475,000 in 2017.  Moreover, there is no correlation between improvement in overall survival and cost of a therapy.  These trends combined with pharmaceutical industry scandals like profit-focused Valeant and exorbitant pyrimethamine (Daraprim) pricing by Turing Pharmaceuticals have created public outcry for stricter regulations by government.

Politicians also feel the financial pinch.  With the aging US population, Medicare covers an increasing proportion of pharmaceutical sales at increasing cost.  Medicare covered 18% of US prescription drug costs in 2006 at $224 billion, but this increased to 30% in 2017 costing $343 billion.  By 2024, it is projected Medicare will cover 34% of prescription drugs costs at $564 billion.  As such, pharmaceutical pricing is a growing political football. 

Moving forward, Dr. Roberts proposed several initiatives to promote value-based pharmaceutical pricing, considering the impact on health outcomes that matter to patients and payers as well as the cost of development and production.  For example, government agencies could use comparative effectiveness studies to determine the relative value of a medication, intelligently informing coverage decisions.  Drug prices could be set by an international benchmark, so the entirety of the developed world can share in drug development cost, overcoming current international pricing discrepancies.  Lastly, payers and pharmaceutical companies could explore shared-risk models.  Interestingly, in this spirit, Novartis signed a pay-for-performance contract in 2016 with insurers for sacubitril/valsartan (Entresto), in which a base rebate fluctuates depending on patient outcomes.  As Entresto promises to reduce heart failure hospitalizations and deaths by 20%, high performance would benefit both patient well-being and insurers’ bottom line. 

Dr. Roberts showed how early decisions in the drug development pipeline have exponentially expanding ramifications on the value of a drug.  This reflected two commonly repeated themes throughout the conference: to only “work on good drug targets”, and to conduct research that makes a “meaningful clinical difference.”  In other words, instead of increasing R&D expenses by pushing a drug with a poor target forward and spending millions in clinical trials for a 30- to 32-month increase in PFS, we too must be value-driven and strive for substantive and valuable clinical impact.


Fellow Summary


Authored by Kamaneh Montazeri

Dr. Thomas Roberts highlighted the recent advances in drug development as well as the costs and challenges involved. In recent years, the pace of development and FDA approval for new molecular entities (NMEs) has become significantly faster, with the average rate of 30-50 NMEs per year. This is partly due to the regulatory advances at the FDA since 2017 (including digital innovations, enhancing adaptive clinical trials and genetic competition, implementing 21st Century Cures Act) as well as pre-2017 advances in the setting of accelerated approvals. In addition, companies are becoming more and more selective with regards to choosing their study population, which has resulted in decrease in the number of addressable population, but increase in the probability of approval as well as the value of the drugs. That said, drug development remains risky. It takes 2.6 billion dollars on average to develop a drug, while only 10% of phase I drugs proceed to FDA approval. Large pharmaceutical companies struggle to innovate; for example, Merck, as one of the most productive companies, brings one drug per year to registration. Moreover, pipelines are crowded with over 1200 solid tumor NMEs. Another issue is the significantly rising cost of drugs. Value-based evaluation addresses this issue, which is defined by healthcare outcomes over the cost of delivering those over time. Monthly costs of cancer drugs have been rising exponentially, without any clear correlation between the price and efficacy. Some of the methods for applying value-based pricing models in oncology include direct negotiation of the price by Medicare, using international price index (IPI), allowing the government to use comparative effectiveness research, encouraging companies to be transparent about net price, and allowing pharmacies to import drugs from Canada. The pay-for-performance contract between Novartis and Cigna/Aetna for Entresto is one example using the performance-based pricing method. In summary, we need better real world data to validate the trial results and to help improve payment models and control drug costs.


Fellow Summary


Authored by Yonina Murciano-Goroff

Dr. Roberts’s fascinating talk entitled “New Drug Development in 2020: Progress, Challenges, and Prices” began with career advice.  Dr. Roberts emphasized the importance of following one’s ideas, but seeing them as “hypotheses to be tested” rather than as truths to be admired.

He went on to give a landscape vision of the state of drug development in 2020.  He noted that the FDA has approved about 20-30 new molecular entities (NMEs) per year recently, of which the majority have been in oncology. The turn-around-time for approvals has also gone down with support from the FDA and with many drugs receiving accelerated approval based on surrogate endpoints.

He noted that we are spending more on R&D, raising the question of how these funds can be used most efficiently.  Drug development remains costly, with only about 10% of drugs progressing from phase I to approval.  As companies are liable to invest roughly $2.6 billion in each new drug developed and to continue to invest in R&D for the drug post-approval, it becomes necessary to thoughtfully analyze which drugs are likely to offer the highest value to our patients. 

In reflecting on how we can bolster oncology’s successes, he emphasized that biomarkers that predict response to therapy are first and foremost important for matching patients to optimal treatments and sparing them from unnecessary toxicities.  While one might have thought that a strong biomarker would cut into the market for a drug by decreasing the target population, the higher likelihood of responses on trial and of drug approval actually leads to decreased costs in biomarker-selected studies. Interestingly, biologics have had high success rates as well, possibly due to their highly validated targets and fewer off-target side effects. 

Thinking about how we define value, such as using Michael Porter’s definition in terms of outcomes that matter to patients over costs, also becomes important. Thankfully, oncology drugs lately have had a relatively high success rate in terms of R&D productivity, but returns on R&D appear to be falling.  In turn, the costs of drug prices for patients have gone up exponentially, as Peter Bach has documented, and this is especially the case in the US.  Overall, a small number of drugs account for a large percentage of spending.  Changes in pricing models, Dr. Roberts explained, may help quell unsustainable pricing while helping accelerate the development of those drugs that are most meaningful for patient outcomes.