If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author: Department of Epidemiology and Biostatistics, UCSF Mission Bay Campus, Mission Hall: Global Health & Clinical Sciences Building, 550 16th St, 2nd Fl, San Francisco, CA 94158, USA.
Division of Hematology and Medical Oncology, Knight Cancer Institute, Oregon Health and Science University, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA
Fifty drugs (55%) approved for a genomic indication had a study evaluating OS benefit.
•
Twenty-two randomised studies (24%) demonstrated an improvement in OS.
•
The median improvement in OS for drugs with a genomic indication was 4.7 months.
Abstract
Aim
Drug approvals for genome-informed indications have been increasing in recent years, but it is unknown how many of them have demonstrated an improvement in overall survival (OS). We assessed the frequency of approved genome-informed drugs demonstrating improvements in OS and progression-free survival (PFS) and whether the frequencies differed by cancer type.
Materials and methods
We searched all Food and Drug Administration approvals from 2006 to 2020, and for each drug that was approved for a genomic indication, we then searched on PubMed for randomised studies examining OS or PFS.
Results
We found 53 drugs approved for 92 unique indications from 2006 to 2020. We found that 50 drugs (55%) approved for a genomic indication had a randomised study evaluating OS benefit, and of those, only 22 demonstrated an improvement in OS. Similarly, 52 drugs (57%) evaluated PFS benefit, and 51 of these studies demonstrated an improvement in PFS. Drugs approved for BRAF V600 melanoma demonstrated an improvement in OS more often than drugs approved for ALK non–small cell lung cancer. The median improvement in OS was 4.7 months (range 1.5 months–49.1 months).
Conclusion
Although there is widespread enthusiasm for this class of agents, and many demonstrate impressive response rates, further trials or post-marketing studies are needed to ascertain the impact on survival and quality of life, the magnitude of these gains, and the cost-effectiveness of these agents.
]. However, only one in seven oncology drugs that are approved on a surrogate outcome is later shown to improve overall survival (OS) in extended follow-up, post-marketing, or subsequent studies [
Cancer drugs approved on the basis of a surrogate end point and subsequent overall survival: an analysis of 5 years of US food and drug administration approvals.
Approvals for genome-informed indications, which are often based on ORR, have been increasing in recent years. Currently, 13.6% of US cancer patients are eligible for genome targeted drugs and 7% of cancer patients may respond [
]. However, it is unknown what percentage of these drugs have proven OS gains for these indications and what the magnitude of those differences are. In this study, we sought to determine the percentage of drugs approved for a targeted indication that have studies reporting on OS and PFS and whether this varies by cancer type.
2. Methods
We searched the Food and Drug Administration (FDA) website to find all oncology drugs approved for a genetically targeted indication for advanced, metastatic, or unresectable cancers (January 2006 through December 2020). We then extracted data regarding the indication, ORR, and date of approval. For each approved drug, we searched PubMed for articles reporting on randomised clinical studies that tested whether the approved drug improved OS, PFS, or ORR, compared with standard of care for each approved indication. The search terms included the study drug, the tumour type, and the genetic indication, filtering by ‘clinical trials’. In some cases where there were a lot of search results produced, we used the Boolean operator of ‘not’ to remove studies that were not relevant to our study (e.g. adjuvant if the drug was approved first line). We searched for studies published through May 25, 2021.
For each study, we extracted data relating to the efficacy for both the intervention and control groups in the study (e.g. median times of PFS and OS, hazard ratios, ORR, P values, and/or confidence intervals). We then classified each drug as having a randomised controlled trial that reported on these outcomes, and if so, whether the results were positive or negative/null for each outcome type (OS, PFS, or overall ORR).
We then calculated frequencies for each outcome by trial result positivity. We did this for all cancers combined and for four cancers with the most frequent drug approvals. For drugs that demonstrated OS improvement and reported median OS, we calculated the median OS improvement. All analyses were done using Excel and R software, version 3.6.1. In accordance with 45 CFR §46.102(f), this study was not submitted for institutional review board approval because it involved publicly available data and did not involve individual patient data.
3. Results
During the time interval 2006–2020, we found 53 drugs approved for 92 unique indications. Genetic indications with the most approvals include drugs targeting the Philadelphia chromosome (PH)+ (14 approvals), EGFR (12 approvals), HER2 (10 approvals), ALK (8 approvals), and BRAF V600 (9 approvals). Cancer types with the most drug/indication approvals were non–small cell lung cancer (NSCLC; 23 approvals), breast (12 approvals), chronic myeloid leukaemia (10 approvals), colorectal (8 approvals), and melanoma (7 approvals). Fifty-eight (63%) approvals were regular, and 34 (37%) were accelerated. There were 18 approvals in 2020, 8 in 2019, 14 in 2018, and 11 in 2017. The remaining 41 approvals were made in 2016 or before, with 16 of them being accelerated.
As presented in Fig. 1, 50 drugs (55%) approved for a genomic indication had a randomised study evaluating OS benefit; 52 drugs (57%) evaluated PFS benefit. These results, stratified by haematologic versus solid tumour indications, are presented in Fig. 1.
Fig. 1Percentage of studies with positive, negative/null, or no outcomes for oncology outcomes in randomised trials of genome-informed drugs approved by the US Food and Drug Association, overall and stratified by solid versus haematologic cancers.
The percentage of drugs/indications with positive study results was 24% for OS (n = 22) and 55% for PFS (n = 51). The percentage of drugs with negative or null study results was 30% for OS (n = 28) and 1% for PFS (n = 1).
These results vary by cancer type, with NSCLC drugs having a low percentage of studies reporting positive results for OS (n = 3; 12% of studies for NSCLC drugs) and melanoma drugs having a higher percentage of studies confirming OS benefit (n = 5; 71% of studies for melanoma drugs; data not shown). Table 1 shows the drugs/indications for which we found a study showing improved OS.
Table 1Drugs that are FDA approved for a genomic indication that have shown to improve overall survival.
Drug
Date of approval
Cancer type
Genomic target
Median overall survival times (intervention versus control)
Gilteritinib
11/28/18
AML
FLT3
9.3 versus 5.6 months
Midostaurin
4/28/17
AML
FLT3
74.7 versus 25.6 months
Trastuzumab
9/25/1998
Breast
HER2
25.4 versus 20.3 months
Tucatinib in combination with trastuzumab and capecitabine
4/17/20
Breast
HER2
21.9 versus 17.4 months
Pertuzumab
6/8/12
Breast
HER2
56.5 versus 40.8 months
Ado-trastuzumab emtansine
2/22/13
Breast
HER2
30.9 versus 25.1 months
Ibrutinib
7/28/14
CLL
17p
90% versus 81% at 12 months
Nilotinib
6/17/10
CML
Ph+
98.5% versus 95.2% at 3 years
Encorafenib in combination with cetuximab
4/8/20
CRC
BRAF V600E
8.4 versus 5.4 months
Cetuximab in combination with FOLFIRI
7/9/12
CRC (1st line)
EGFR (KRAS)
23.5 versus 20.0 months
Cetuximab
2/1/04
CRC (later line)
EGFR (KRAS)
6.1 versus 4.6 months
Panitumumab
9/27/06
CRC
KRAS
10.0 versus 7.4 months
Trastuzumab
10/20/10
Gastric
HER2
13.8 versus 11.1 months
Imatinib
12/19/08
GIST
GIST
92% versus 82% at 5 year
Encorafenib and binimetinib
6/27/18
Melanoma
BRAF V600 E or K
33.6 versus 16.9 months
Cobimetinib in combination with vemurafenib
11/10/15
Melanoma
BRAF V600 E or K
22.3 versus 17.4 months
Trametinib and dabrafenib
1/10/14
Melanoma
BRAF V600 E or K
25.1 versus 18.7 months
Trametinib
5/29/13
Melanoma
BRAF V600 E or K
81% versus 67% at 6 months
Vemurafenib
8/17/11
Melanoma
BRAF V600E
13.6 versus 9.7 months
Dacomitinib
9/27/2018
NSCLC
EGFR 19/21
34.1 versus 26.8 months
Osimertinib
4/19/18
NSCLC
EGFR 19/21
38.6 versus 31.8 months
Nivolumab
10/9/15
NSCLC
EGFR or ALK
12.0 versus 9.6 months
FDA = Food and Drug Administration; AML = acute myeloid leukemia; CLL = chronic lymphocytic leukemia; CML = chronic myeloid leukemia; CRC = colorectal cancer; GIST = gastrointestinal stromal tumour; NSCLC = non-small cell lung cancer.
For drugs that were shown to improve OS and reported median OS times (n = 17), the median improvement in OS was 4.7 months. The improvement ranged from 1.5 months for EGFR relapsed/refractory colorectal cancer to 49.1 months for FLT3 acute myeloid leukaemia. More than half (59%) of the approvals that demonstrated an improvement in OS were for BRAF V600 melanoma and EGFR NSCLC or colorectal cancer.
We found 34 drugs/indications that failed to improve OS, RR, or PFS, 15 drugs/indications improved all three of these outcomes, 18 drugs/indications had an improvement in PFS only, three had an improvement in ORR only, two had an improvement in OS only, 15 improved both PFS and ORR, three improved OS and PFS, and two improved OS and ORR (Supplemental Table). Of the drugs that had no data on OS, 16 (38%) drug approvals were for haematologic indications.
Fig. 2 shows the percentage of drugs demonstrating an improvement in OS in oncology drugs approved for a genetic indication, by cancer type and genetic target, for the more common cancer types. The number and percentage of studies finding an improvement in OS varied by tumour type and genetic indication. We found that for the seven drugs approved for BRAF V600 melanoma, five (71%) improved OS, and all eight (100%) of the drugs approved for ALK NSCLC failed to show an improvement in OS. Of the three drugs approved for MSI/MMR colorectal cancer, none (0%) had studies reporting on OS.
Fig. 2Percentage of drugs demonstrating an improvement in overall survival, by tumour type and genetic target, in oncology drugs approved for a genetic indication.
We found that only about half of oncology drugs FDA approved for a genomic indication had studies reporting on OS, and only about one-fifth of them demonstrated an improvement in OS. Our results are slightly lower than another study that report that 32% of drugs approved for an oncology indication had studies showing an improvement in OS in either pre- or post-marketing studies [
]. With a focused effort on increasing earlier treatment options for cancer patients, surrogate markers such as PFS or ORR, which are thought to bring drugs to market faster, are being increasingly used in drug approvals. However, to justify the high cost of these drugs, an improvement in OS should also be demonstrated. Here, we show that most approved targeted drugs have yet to show OS benefit.
We found that drugs approved for a genomic indication, which improved OS, did so by a median of 4.7 months, which is slightly more than other studies that have evaluated all cancer drugs approved between 2003 and 2013 [
Unintended consequences of expensive cancer therapeutics—the pursuit of marginal indications and a me-too mentality that stifles innovation and creativity: the John Conley Lecture.
JAMA Otolaryngol Head Neck Surg.2014; 12: 1225-1236
The effect of targeted therapy on overall survival in advanced renal cancer: a study of the national surveillance epidemiology and end results registry database.
]. This is not unexpected given that previous observational studies have shown that for certain cancers, patients treated with targeted therapies have better OS than patients treated without targeted therapies [
]. A concern for targeted drugs is that trial-level data overestimates OS because participants who are less likely to participate in clinical trials are also less likely to have an improved OS in real-world data [
One of the limitations to our analysis is that we only used PubMed to find articles demonstrating an improvement in outcomes. We may have found more studies finding improvement in these outcomes had we used other search engines. However, increasing the number of studies examining a given outcome also increases the likelihood of finding benefit, real or not. A second limitation is that some data were immature, and the final results for these studies may be different from the preliminary data. In these instances, we used the immature findings because that is all we had available. Consequently, our results apply to current knowledge and may not be generalisable to future findings.
5. Conclusion
About half of FDA-approved drugs for a genome-informed oncology indication have had studies evaluating OS benefit. Only about one-fifth of drugs for these indications have had randomised studies reporting positive OS outcomes, whereas over half of drugs have been shown to improve PFS. Although there is widespread enthusiasm for this class of agents, and many demonstrate impressive response rates [
], further trials or post-marketing studies are needed to ascertain the impact on survival and quality of life, the magnitude of these gains, and the cost-effectiveness of these agents.
Authors' contribution
V.P., A.H., and M.S.K. conceptualised study design. A.H. and M.S.K. reviewed and abstracted data. V.P. reviewed and confirmed abstracted data. A.H. wrote the first draft of the article. All authors reviewed and revised subsequent and finalised draft of the article.
Funding
This study was funded by Arnold Ventures.
Conflict of interest statement
V.P. received research funding from Arnold Ventures, royalties from Johns Hopkins Press and Medscape, and honoraria from Grand Rounds/lectures from universities, medical centres, non-profits, and professional societies; is a consultant at UnitedHealthcare; and received speaking fees from Evicore. (Other) Plenary Session podcast has Patreon backers. All other authors have no financial nor non-financial conflicts of interest to report.
Appendix A. Supplementary data
The following is the Supplementary data to this article:
Cancer drugs approved on the basis of a surrogate end point and subsequent overall survival: an analysis of 5 years of US food and drug administration approvals.
Unintended consequences of expensive cancer therapeutics—the pursuit of marginal indications and a me-too mentality that stifles innovation and creativity: the John Conley Lecture.
JAMA Otolaryngol Head Neck Surg.2014; 12: 1225-1236
The effect of targeted therapy on overall survival in advanced renal cancer: a study of the national surveillance epidemiology and end results registry database.
01186-2&id=gr1.jpg){kind=link}
01186-2&id=gr2.jpg){kind=link}