KEI Reaction to Astellas Comment Regarding Public and Private Sector Contributions to the Development of Xtandi

(More on the Xtandi case here:

Astellas has issued a statement, dated February 9, 2022, on the Bayh-Dole Act and the petition to exercise march-in and government use rights in the prostate cancer drug Xtandi. It was brought to KEI’s attention today.

Link to Astellas’ statement:,-2022

The statement issued by Astellas includes the following:

“While the U.S. government contributed approximately $500,000 to the initial discovery of the molecule that eventually became XTANDI, Astellas has invested more than $1.4 billion to date in research and development efforts for XTANDI on behalf of the patients we serve in the U.S. and around the world. Astellas and its partners supported extensive clinical trials and research necessary to demonstrate that XTANDI is safe and efficacious for patients, support the application that led to FDA approval, and further understand which patients will benefit from treatment.”

There is a lot to unpack here, both on the estimate of the public sector investment and on the claims about Astellas’s contribution.

Public Sector Investment

It’s not clear from the Astellas statement, but the $500,000 figure cited may be related to the two grants that are listed on the Xtandi patents, one from the US Army and one from the NIH.

We will have to wait to see exactly what Astellas is referring to, and we encourage them to transparently publish any and all data on which their estimates rely, since they believe this information is relevant to the outcome of the petition. The public has a clear stake and interest in the outcome, evidenced by the 1,500 letters received by the NIH in support of the petition and 16,000 individuals who signed a letter supporting the petition.

A common technique by opponents of march-in rights is to minimize the public sector investment by focusing the cost of one or two grants only, and not the risk-adjusted value of the investments. When industry investments in preclinical costs are studied, there is generally a recognition of the high risks of failures. Joseph DiMasi, for example, in his widely cited study of drug development costs (J.A. DiMasi et al. / Journal of Health Economics 47 (2016) 20–33), includes in his total estimate the risk-adjusted costs of bringing one successful product to market. These risks are difficult to model or measure in the preclinical stage. One typical example of modeling risks would be the 2004 paper by Kraljevic et al, which states, “Of 5,000 compounds that enter pre-clinical testing, only five, on average, are tested in human trials, and only one of these five receives approval for therapeutic use.“ This example is a 1,000 to one success rate to even progress to testing in humans, and a 5,000 to one odds of FDA approval. If $500,000 in federal grants have a one in a thousand chance of success, the risk-adjusted value of the investment is $500,000,000. If the odds were really 5,000 to one for FDA approval, the risk-adjusted value would be $2.5 billion.

When DiMasi estimates the economic cost of pre-clinical research, he also adds the cost of capital. Given that preclinical work on Xtandi took place from “early” 2000 (UCLA’s press release on the sale of its royalty interest) to the beginning of trials in 2007, for an FDA approval in 2012, this would mean that depending upon the discount rate, the economic value of preclinical outlays would be more than doubled. For comparison, in his 2016 paper, DiMasi valued preclinical research at $430 million after risk adjustment and $1.098 billion after adding the cost of capital, an increase of 155 percent over the risk adjusted costs.

There are also other ways of considering the public sector investments in Xtandi.

The U.S. Army grant to UCLA was part of the US Department of Defense’s Congressionally Directed Medical Research Program (CDMRP). The cumulative CDMRP outlays on prostate cancer research were more than $2 billion by fiscal year 2021, and averaged $84 million per year from fiscal year 2000 to 2007.

The NIH outlays on prostate cancer-related research were even more substantial. A search of NIH RePORTER using the keyword “prostate” identifies more than $11 billion in grants, including grant totals of $200 million to $500 million per year from fiscal years 2000 to 2007.

If one looks at the funding for the NIH grant listed on the Xtandi patent, from 2000 to 2012 (the year Xtandi was approved by the FDA), the total is $18 million.

Charles Sawyers, the lead inventor on the Xtandi patents, is listed as the principal investigator for $27.7 million in NIH-funded projects and $8.3 million in subprojects, beginning in 1985.

Michael Jung, another of the Xtandi inventors, was the principal investigator for 42 NIH grants involving $5.99 million.

Like the industry, the government places a lot of bets in medical research, in hopes that some will come through. The Astellas statement deliberately and misleadingly understates the government’s role.

Astellas and Other Private Sector Investments

Astellas makes the claim that it has “invested more than $1.4 billion to date in research and development efforts for XTANDI on behalf of the patients we serve in the U.S. and around the world.”

First of all, once Xtandi was approved by the FDA on August 31, 2012, it became a commercial and profitable product that has generated more than $20 billion in global sales. Most drug development studies look at the costs incurred before FDA approval. That, for example, was the approach by DiMasi in his widely quoted 2016 paper, the one PhRMA loves to quote. DiMasi provides estimates of the costs of clinical testing, which are considered on the high side, and are broken down by phase, through FDA approval. His mean cost figure, in 2013 dollars (Xtandi was FDA approved in 2012), was $25.3 million for Phase 1, $58.6 million for Phase 2 and $255.4 million for Phase 3 (the median cost figures were lower). This is a total of $339.3 million, more than a billion less than the $1.4 billion figure in the Astellas statement.

The FDA’s Medical Review for the 2012 approval is here:

In the Medical Review for Xtandi, the FDA relies on Phase 3 testing involving 1,199 patients (800 patients using enzalutamide and 399 using a placebo). This is not a huge trial, even for a cancer drug.

A 2015 study by Battelle, conducted for PhRMA and published on the PhRMA web page, estimated the per patient costs of oncology trials, by Phase in Table A-1. Phase 3 trials were estimated to have a cost of $69,000 per patient, in 2013 dollars. This would put the cost of the pivotal phase 3 trial cost at $82.7 million, which is significant, but a long way from $1.4 billion.

Biopharmaceutical Industry-Sponsored Clinical Trials: Impact on State Economies, Battelle, 2015.

We are sure that Astellas will point to various things that cost Astellas money, to get to the purported $1.4 billion number, and we look forward to a thorough and transparent review of any data they want to present in a proceeding on the Bayh-Dole petition, including, for example, any efforts to pass off asset acquisition costs as R&D costs.