Contemporary pharmaceutical industry pricing practices are threatening to undermine the health care industry’s and policymakers’ efforts at cost control. Egregious as the price hikes are, the hedge fund managers who gained notoriety for exorbitant price increases on some generics are the least of the problem. The 2 biggest drivers of high drug costs in recent years have been the steady increase in prices for existing on-patent drugs, which account for more than 70% of all drug spending, and the 6-figure retail prices set for the latest generation of specialty and cancer therapeutics. To counter increasing public alarm over these high prices, industry leaders continue to assert that the substantial investment in researching and developing new products and the riskiness of that enterprise justify charging Americans the highest prices in the world for medicines. To support the assertion, the industry’s trade group relies on an industry-funded study first produced in 1979 by the Tufts University Center for the Study of Drug Development.1 The most recent iteration of the study, which was last updated in 2014, claims it takes more than 10 years and nearly $2.7 billion in capital to develop a single drug.2 In inflation-adjusted dollars, the study’s estimate for developing a new drug has more than doubled in the past decade and is more than 10 times the original 1979 figure. Industry critics and journalists have repeatedly questioned the assumptions in the Tufts study. They argue its authors arbitrarily inflate the value placed on the cost of capital spent on research and development, which is a current expense and not an investment for accounting and tax purposes; fail to distinguish between new products that are truly innovative and those that merely replicate drugs already on the market; and ignore the fact that the industry consistently generates the highest profit margins among all US industries, which suggests the pricing power afforded by patent exclusivity far outweighs the inherent riskiness of pharmaceutical research and development. Critics also lament the Tufts study’s lack of transparency because it relies on industry-supplied data that the authors refuse to make public. Over the years, none of the critiques have had much political impact. Most politicians from both political parties accept the Tufts study’s basic premise because it provides them with a rationale for failing to enact countermeasures, which could include giving Medicare the right to negotiate prices, allowing drug importation or establishing reference pricing or value-based pricing schemes for new products.3 It took just one meeting with industry leaders for President Donald Trump, who had campaigned against high drug prices, to put the issue on the backburner. “Fifteen years, $2.5 billion to come up with a product where there’s not even a safety problem. So it’s crazy,” he said.4 However, the issue of increasing drug prices continues, as do the repercussions. Total retail prescription drug spending increased by 4.8% in 2016, about the same pace as overall health spending, after increasing 12.4% in 2014 and 9.0% in 2015. The outlook for the next 5 years is for 4% to 7% annual growth as new specialty and rare disease drugs, which now account for 42.9% of all drug spending, come on line.5 Therefore, it is timely to revisit the issue of the cost of pharmaceutical research and development and its association with drug pricing. In this issue of JAMA Internal Medicine, Prasad and Mailankody6 report the results of their study of compiled research and development data from the Securities and Exchange Commission filings of start-up companies with a single approved product. By tracking the pharmaceutical companies’ publicly reported investments, the authors found that the actual cost of developing a new drug is approximately one-fourth the Tufts study estimate. Their fresh perspective is powerful because the authors chose not to challenge the core assumptions of the Tufts study, specifically, including the cost of failed experiments, giving equal value to new products that mimic the therapeutic approach of previously approved drugs (so-called me-too drugs), and treating research and development expenses as an investment. Prasad and Mailankody account for the cost of failure because all the companies in their sample were developing multiple drugs at the time of their first approval. They do not discriminate based on the value of the newly approved medications (just half their sample of 10 drugs received a priority review from the US Food and Drug Administration, which is reserved for substantial breakthroughs). In addition, like the Tufts study authors, they increase the total estimated cost of developing a new drug by adjusting the value of expenditures made during the early stages of development for inflation and what those investments would have earned if placed in an alternative investment. To estimate this latter factor, Prasad and Mailankody used an expected rate of return (the cost of capital) that was 3.5 percentage points below the rate reported in the Tufts study. However, the latter’s 10.5% rate may be unrealistic given the industry’s very low tax rate, very low debt to equity ratio, and mean stock price volatility.7 In any case, using the higher Tufts study rate would raise their overall estimate of the cost to develop a new drug by only approximately 10%, which would not alter the markedly different outcomes of the 2 studies. The authors note as a weakness that their study looked only at cancer drugs, whose high cost may not translate to other disease states. However, the method works well with another new drug whose high price generated headlines in recent years: sofosbuvir (Sovaldi) for hepatitis C. Developed by an antiviral drug development start-up company called Pharmasset, the company’s US Securities and Exchange Commission filings Author Audio Interview

中文翻译：

---

急需的药物开发成本纠正措施

当代制药行业的定价做法​​正威胁着医疗保健行业和政策制定者在成本控制方面的努力。尽管价格上涨令人震惊，但因某些仿制药价格过高而臭名昭著的对冲基金经理是最不重要的问题。近年来药物成本居高不下的两个最大驱动因素是现有专利药物价格的稳步上涨，占所有药物支出的 70% 以上，以及为最新一代药物设定的 6 位数零售价。专业和癌症治疗。为了应对公众对这些高价的日益担忧，行业领导者继续断言，在研究和开发新产品方面的大量投资以及该企业的风险证明向美国人收取世界上最高的药品价格是合理的。为了支持这一说法，该行业的贸易组织依赖于 1979 年由塔夫茨大学药物开发研究中心首次进行的一项行业资助研究。 1 该研究的最新迭代（最后一次更新于 2014 年）声称它开发一种药物需要 10 多年的时间和近 27 亿美元的资金。2 以通货膨胀调整后的美元计算，该研究对开发新药的估计在过去十年中增加了一倍多，是 1979 年原始数字的 10 倍以上. 行业评论家和记者一再质疑塔夫茨研究中的假设。他们认为，其作者任意夸大了用于研究和开发的资本成本的价值，这是一项当期费用，而不是出于会计和税收目的的投资；无法区分真正具有创新性的新产品和仅复制市场上已有药物的新产品；并且忽略了该行业一直在美国所有行业中创造最高利润率的事实，这表明专利独占性提供的定价能力远远超过了药物研发的内在风险。批评者还对塔夫茨大学的研究缺乏透明度表示遗憾，因为它依赖于作者拒绝公开的行业提供的数据。多年来，这些批评都没有产生太大的政治影响。来自两党的大多数政治家都接受塔夫茨研究的基本前提，因为它为他们未能制定对策提供了理由，其中可能包括赋予医疗保险谈判价格的权利，允许进口药物或建立参考定价或基于价值的定价计划新产品。3 与反对高药价的总统唐纳德·特朗普仅会面一次，就将这个问题搁置一旁。“15 年，花费 25 亿美元来开发一种甚至没有安全问题的产品。所以这太疯狂了，”他说。4 然而，药价上涨的问题仍然存在，其后果也是如此。2016 年零售处方药总支出增长 4.8%，与整体医疗支出增长速度大致相同，2014 年和 9 年分别增长 12.4%。2015 年为 0%。未来 5 年的前景是每年 4% 至 7% 的增长，因为现在占所有药物支出 42.9% 的新专科和罕见病药物上线。5 因此，这是及时重新审视药物研发成本及其与药物定价的关联问题。在本期 JAMA Internal Medicine 上，Prasad 和 Mailankody6 报告了他们对美国证券交易委员会提交的具有单一批准产品的初创公司的研究和开发数据汇总的研究结果。通过跟踪制药公司公开报告的投资，作者发现开发一种新药的实际成本大约是 Tufts 研究估计的四分之一。他们的新观点是强大的，因为作者选择不挑战塔夫茨研究的核心假设，特别是包括失败实验的成本，为模仿先前批准药物治疗方法的新产品（所谓的 me-太药物），并将研发费用视为投资。Prasad 和 Mailankody 解释了失败的成本，因为他们样本中的所有公司在第一次获得批准时都在开发多种药物。他们不会根据新批准药物的价值进行歧视（他们的 10 种药物样本中只有一半获得了美国食品和药物管理局的优先审查，这是为重大突破保留的）。此外，像塔夫茨研究的作者一样，它们通过调整在开发早期阶段为通货膨胀而支出的价值以及如果将这些投资置于另类投资中将获得的收益，从而增加了开发新药的总估计成本。为了估计后一个因素，Prasad 和 Mailankody 使用的预期回报率（资本成本）比 Tufts 研究报告的回报率低 3.5 个百分点。然而，鉴于该行业的税率极低、债务权益比率极低以及平均股价波动，后者 10.5% 的税率可能不切实际。 7 无论如何，使用更高的塔夫茨研究率会提高他们对成本的总体估计开发一种新药的成本只有大约 10%，这不会改变 2 项研究明显不同的结果。作者指出，他们的研究仅着眼于抗癌药物是一个弱点，其高成本可能不会转化为其他疾病状态。然而，该方法与另一种近年来高价成为头条新闻的新药配合得很好：用于丙型肝炎的索非布韦（Sovaldi）。 由一家名为 Pharmasset 的抗病毒药物开发初创公司开发，该公司向美国证券交易委员会提交的文件作者音频面试