Comparison of Drugs Used for Adjuvant and Metastatic Therapy of Colon, Breast, and Non–Small Cell Lung Cancers

Key Points Question What is the difference in the number of agents available for metastatic and adjuvant treatment of non–small cell lung cancer, breast cancer, and colon cancer? Findings In this cross-sectional study of available cancer therapies, 69 agents that are recommended for use in metastatic disease were identified, compared with 25 agents for adjuvant use. Drugs used in the adjuvant setting were approved a mean of 10.0 years after drugs approved for the metastatic setting. Meaning There is a substantial difference in the number of agents available for use, as well as the timing of supporting evidence, in the metastatic and adjuvant settings of non–small cell lung cancer, breast cancer, and colon cancer.


Introduction
Adjuvant therapy is the standard of care based on evidence of improved disease-free survival or overall survival in a number of clinical settings. For early-stage non-small cell lung cancer (NSCLC), adjuvant treatment is supported by several large meta-analyses 1,2 demonstrating a 4% to 5% absolute increase in 5-year survival with the addition of systemic therapy following curative surgery.
In stage III colon cancer, adjuvant therapy became widely adopted as evidence of the efficacy of 5-fluorouracil-based adjuvant regimens began to accumulate in the 1990s. [3][4][5] Adjuvant treatment with either endocrine therapy, anti-ERBB2 (formerly HER2/neu)-directed therapy, or chemotherapy is currently recommended for many patients with early-stage breast cancer on the basis of numerous randomized controlled trials and subsequent meta-analyses. 6 Despite the potential for adjuvant therapy to offer benefit, particularly in the aforementioned types of cancer, there have been fewer systemic therapy options added to the armamentarium for this setting when compared with metastatic disease. We sought to describe the degree of difference, as well as any patterns in evidence and experimentation for drug use in both settings, and to identify the number of unique agents that are currently used in metastatic and adjuvant settings.

Methods
This study of publicly available data did not involve personal health information, and institutional review board approval and informed consent were not sought, in accordance with 45 CFR § 46. This study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines for cross-sectional studies.
We limited our study to cancers for which adjuvant therapy is most commonly used: NSCLC, breast cancer, and colon cancer. Using National Comprehensive Cancer Network (NCCN) guidelines that were current as of May 15, 2019, we identified all agents listed as category 1 or 2A recommendations for use in the metastatic or adjuvant settings of their respective cancers. [6][7][8] We classified a unique agent as any single drug listed by NCCN that may or may not be used as part of a combined regimen. For instance, in the case of the commonly used regimen of doxorubicin and cyclophosphamide in breast cancer, cyclophosphamide and doxorubicin were classified as separate agents. Drugs used in multiple cancer types were described as a unique agent for each indication. For example, bevacizumab is recommended for use in NSCLC, breast cancer, and colon cancer, and as such was considered to represent 3 separate agents. For agents that are used as part of more than 1 regimen used in a single cancer type (eg, fluorouracil in both FOLFOX [folinic acid, fluorouracil, and oxaliplatin] and FOLFIRI [folinic acid, fluorouracil, and irinotecan hydrochloride] in colon cancer), the drug in question was counted only once.
We also sought to describe the quality of evidence directly supporting any agent's use in either the metastatic or adjuvant setting. For each agent, we identified the study or studies that led to either their widespread use or US Food and Drug Administration (FDA) approval for the circumstance in question. For FDA-approved agents, the studies cited on the corresponding drug label were used in subsequent qualitative analysis. Labels were obtained by searching the existing database of FDA-approved drugs. In the case of agents that are not FDA approved for use in the specified indication, we evaluated the trials that are cited by NCCN guidelines. In either of these cases, only the cited trials (either via drug label or NCCN guidelines) were used in our study. In a very few instances, however, drug labels from FDA-approved therapies either did not cite any studies or cited studies published after an agent had been adopted into widespread use. In such cases, a literature search was performed using Google Scholar, PubMed, and ClinicalTrials.gov in an effort to identify the trials that ultimately led to routine clinical use. Search queries consisted of "drug name," "cancer type (either breast, NSCLC, or colon cancer)," and "setting (either metastatic or adjuvant)." Of these search results, only the earliest trial leading to routine clinical use was used in our study.
Using these methods, we identified a total of 134 studies published between 1970 and 2019 that were appropriate for further study. We then excluded studies that were phase 1 or phase 2 trials because, by definition, these studies lack comparison of the drug in question against standard treatment. By use of these criteria, we identified 118 trials that were ultimately used in our evaluation.
The outcome of each study was then assessed and qualified as either a positive or nonpositive result.
We considered a positive result as any study in which the agent in question was shown to produce a statistically significant improvement in either progression-free survival (or in the case of studies investigating adjuvant use, disease-free survival) or overall survival. For our study, improved time to progression, objective response rate, or other commonly used end points were not categorized as positive results.
We also aimed to identify the number of agents used in metastatic disease for which there are ongoing trials evaluating efficacy in the adjuvant setting. We performed searches for such studies using Google Scholar, PubMed, and ClinicalTrials.gov. Search queries consisted of "drug name," "cancer type (either breast, NSCLC, or colon cancer)," and "adjuvant."

Results
We identified 69 agents that were either category 1 or 2A per NCCN guidelines for use in the metastatic setting for their respective cancers (24 in NSCLC, 30 in breast cancer, and 15 in colon cancer). This number includes both agents that are currently FDA approved for use and those not approved but widely used; 60 of the 69 agents (87.0%) are recommended by current NCCN guidelines and are FDA approved for use in metastatic disease. Since 2010 there have been 31 new agents approved or recommended for use in metastatic NSCLC, breast cancer, or colon cancer, whereas there have only been 7 agents adopted in the adjuvant setting for these cancers. Only 2 agents (pertuzumab and ado-trastuzumab emtansine in breast cancer) have since been approved for adjuvant use, albeit at a delay of 5.50 and 6.25 years, respectively. Of the 69 identified agents, 24 (34.8%) were later approved or recommended for use in the adjuvant setting. An additional agent (neratinib) was approved for use in the adjuvant setting without preceding approval or recommendation for use in metastatic disease, bringing the total to 25 agents. Fifteen of these 25 agents (60.0%) are currently FDA approved for adjuvant use. The Table represents a summary of the findings. The mean (SD) time difference between use in metastatic and adjuvant settings was 10.0 (7.5) years. The longest difference between adoption in metastatic disease and as adjuvant therapy was 32 years in the case of 5-fluorouracil for colon cancer, whereas the shortest interval between use in these 2 settings was 2 years in the case of cyclophosphamide for breast cancer. For each cancer type, the number of metastatic agents also recommended for adjuvant use was as follows: 7 of 24 agents   have been a number of trials aimed at evaluating additional agents for adjuvant use, most of which have been negative. 13 Irinotecan has been an effective agent in metastatic colon cancer and yet has failed in the adjuvant setting. 14 Bevacizumab is an especially salient example of the lack of correlation between efficacy in adjuvant and metastatic settings of colon cancer, as evidenced in multiple studies. 15,16 In addition, the anti-epidermal growth factor receptor monoclonal antibody cetuximab was found to be no more effective than control when used as adjuvant therapy in colon cancer. 17 Interestingly, there are a number of ongoing trials evaluating other targeted therapies (panitumumab, ramucirumab, and regorafenib) and the checkpoint inhibitor atezolizumab. [18][19][20][21]

JAMA Network Open | Oncology
Comparison of Drugs Used for Therapy of Colon, Breast, and Non-Small Cell Lung Cancers Similar to colon cancer, the promising success of targeted therapies and checkpoint inhibitors in metastatic NSCLC has not been parlayed into positive results when tested in the adjuvant setting of this disease. Although targeted therapies may ultimately serve a role in adjuvant therapy, they have yet to demonstrate a statistically significant benefit, with negative trials evaluating erlotinib and bevacizumab being notable disappointments. 22,23 Studies evaluating afatinib, alectinib, crizotinib, and osimertinib are ongoing; the checkpoint inhibitors atezolizumab and pembrolizumab are also currently being evaluated in the adjuvant setting. [24][25][26][27][28][29] As in both colon cancer and NSCLC, bevacizumab was shown not to be of benefit in the adjuvant setting of breast cancer. 30,31 Other notable negative studies [32][33][34][35]

Limitations
This study has several limitations. First, we included only the 3 cancer types for which adjuvant systemic therapy is most commonly used. Given that our study was retrospective and performed using publicly available data, we thought that restricting our evaluation to these cancers would nonetheless provide an accurate portrayal of the existing differences between systemic therapy options in metastatic and adjuvant settings. Second, we limited our classification of positive trial results to those that demonstrated a statistically significant improvement in either progression-free survival (or in the case of studies investigating adjuvant use, disease-free survival) or overall survival.
Although this classification is necessarily restrictive, we thought that these statistical end points are traditionally the most likely to equate to clinical relevance.

Conclusions
Our study demonstrates that only 34.8% of drugs endorsed in the metastatic setting are also endorsed in the adjuvant setting of NSCLC, breast cancer, and colon cancer. This finding is likely associated with a multitude of factors, especially the higher burden of proof required for adjuvant therapies and a general lack of correlation between efficacy of a drug in advanced cancers and benefit in early disease. Although there are a number of ongoing trials aimed at elucidating additional agents for adjuvant use, there is still a high percentage of drugs that have yet to be studied, suggesting possible future directions.