Many drugs show little benefit
The JAMA Oncology study builds on other data that raise doubts about how the FDA determines value when making approval decisions for cancer drugs.
Overall survival is the most direct measure of clinical benefit used to determine value. But OS as an endpoint in clinical trials generally requires larger patient numbers and increased time for follow-up, thereby delaying approvals. This is likely why the use of surrogate endpoints to approve drugs has increased. Dr. Nabhan’s study found the use of surrogate endpoints during trials grew from 67% during the period of 2008 through 2012 to 76% during the period of 2011 through 2017. Surrogate endpoints can include tumor shrinkage, time to progression, or time to reappearance of disease.
The use of surrogate endpoints to determine value however, has long come under scrutiny. A 2019 analysis published in JAMA Internal Medicine found that most cancer drugs approved by the FDA based on response rate (RR) – the percentage of patients who experience tumor shrinkage – have less than transformational response rates, and that such indications do not have confirmed clinical benefit.
Of 59 oncology drugs with 85 unique indications, most had a response rate ranging from 20% to 59%. Of 81 available indications, the median complete response rate – defined as the percentage of patients with no visible disease and normalization of lymph nodes – was 6%. (Complete response data were not reported for four drug indications.) Investigators also reported that many of the drugs have remained on the market for years without subsequent confirmatory data.
In addition, a 2018 review of randomized clinical trials published in JAMA Internal Medicine that analyzed progression-free survival in cancer patients found the endpoint did not improve patients’ lives. A quantitative analysis of 52 articles that covered 38 randomized clinical trials involving 13,979 patients across 12 cancer types found no significant association between progression-free survival and health-related quality of life.
“These findings raise questions regarding the assumption that interventions prolonging [progression-free survival] also improve [health-related quality of life] in patients with cancer,” the study authors wrote. “Therefore, to ensure that patients are truly obtaining important benefit from cancer therapies, clinical trial investigators should measure [health-related quality of life] directly and accurately, ensuring adequate duration and follow-up.
PROs, which can encompass health-related quality of life, is a measure the FDA has encouraged investigators to use during trials – when the benefit exists. In the analysis by Dr. Nabhan, trials evaluated PROs during pivotal studies supporting initial approval in 50 of the 71 indications. Before approval, 14 drugs demonstrated a statistically significant improvement in at least one PRO, but only 1 of the 14 – ruxolitinib – was granted a PRO labeling claim at the time of approval. Post approval, a statistically significant improvement in PROs was shown for only 18 of the 71 (25%) initial indications.
Meanwhile, although overall survival is considered the optimal yardstick with which to measure drug benefit, the value of longevity – and how it should be weighted – poses further questions. A number of new oncology drugs approved based on survival data lengthen lives by a very short time, Dr. Saltz noted. In the study of 71 indications for example, the median OS gain for drugs approved based on survival data was 1.7 months.
“We crossed the absurdity boundary a long time ago with drugs in the range of $10,000 to $20,000 a month with median survival benefits of less than 2 months,” Dr. Saltz said. “That’s not very much. Then we get into a rather circuitous argument that doesn’t settle anything as to whether it’s ‘worth it or not.’ The question becomes to whom is it worth what, and who’s paying for it?”