Exclusion of Older Adults from Cancer Clinical Trials: Review of the Literature and Future Recommendations

Abstract

In this review, we present the context of older adult (OA) cancer patients within the broader cancer population, including cancer burdens and trial representation. We first describe the proportion of older adults in clinical trials, with studies showing strong evidence that the proportion of OA in cancer trials is much less than the proportion of OA in the overall cancer population. We highlight the lack of generalizability that can lead to challenges in treatment decisions for OA as well as concerns regarding health inequity. We then discuss barriers to OA enrollment related to trial structure/design, physician perspective, and patient/caregiver perspective. We expand on this further by outlining these barriers throughout the process of trial design, patient enrollment/trial implementation, and data analysis in post-market settings. We summarize guidelines from national societies, regulatory agencies, and other institutional bodies, and present a compilation of on-the-ground actionable recommendations to address the challenges of clinical trial design, focusing on geriatric assessments and OA-specific trials. We conclude by providing an outline for future directions, noting specifically the potential impactful role radiotherapy and radiation oncology may plan for clinical trials related to OA patients.

Introduction

Cancer is the second leading cause of death globally and the leading cause of mortality in the United States (U.S.) for patients aged between 60–79 years¹. Older adults (OA), defined by the Centers for Disease Control and Prevention as patients 65 years and older, make up the largest group of patients diagnosed with cancer. The underrepresentation of older patients in cancer therapeutic trials has been a growing concern for several decades. Increasing efforts have sought to identify and address specific barriers that hinder OA enrollment in cancer clinical trials, including exclusionary eligibility criteria, patient and provider perceptions, geographic and financial barriers, and more. The U.S. Food and Drug Administration (FDA) has published guidelines that aim to increase OA representation on cancer trials, with particular emphasis on improving OA enrollment on industry-sponsored studies². Despite these efforts, age disparities continue to be a major issue in cancer clinical trials, with recent data suggesting that age disparities between the general population and clinical trial participants are widening³.

Across myriad cancer types, advanced age remains one of the most important risk factors for both cancer diagnosis and prognosis. The median age of cancer diagnosis in the U.S. is 66 years⁴. Future trend estimates point to an increasingly older cancer population with higher cancer risk: the probability of developing invasive cancer from 2015 to 2017 (within selected age intervals) for adults aged 60 to 69 (13.6%) is more than twice the probability of those who are middle-aged (50 to 59 years, 6.2%) or younger (3.5%)¹. This only adds to the complexity of cancer care management for OA. Furthermore, there are unique challenges that older patients with cancer can face, such as the increased prevalence of comorbid medical conditions, social support issues, limitations in activities of daily living, treatment toxicity, among others. Additionally, these issues may be exacerbated by socioeconomic barriers to care. Here, we examine the current landscape of OA in cancer trials and review the literature on the major barriers to OA enrollment in trials. We highlight prior and ongoing efforts to address such disparities and discuss future directions in trial design, enrollment, and data reporting that aim to improve OA representation in clinical trials.

Current and Projected Cancer Burdens of Older Adults

When examining the underrepresentation of OA in cancer clinical trials, an assessment of current and future cancer trends for OA is crucial. In the past few decades, older adults have comprised an increasing majority of cancer incidence and mortality, and this burden will only increase as the U.S. population grows older. A 2007 study found that 60% of cancer incidence and 70% of cancer mortality occur in adults aged 65 and older⁵. Based on projections from a 2009 study, OA patients are predicted to make up 70% of newly diagnosed cancers by 2030⁶.

In addition, the 2021 cancer estimates provided and analyzed by the American Cancer Society (ACS) provide a current picture of the burden that older patients have concerning cancer incidence and mortality¹. The leading causes of cancer death in patients age 60 years and older are primary lung cancers, and the probability of developing invasive lung cancer is relatively higher than younger patients¹. The ACS estimates also show that cancer is the leading cause of death for older adults age 60 to 79 years¹. These trends underscore the need for clinical trials that adequately answer questions of cancer treatment specific for OA.

Older Adult Disparities in Cancer Clinical Trial Participation

Proportion of OA enrolled in clinical trials in the United States

Over the last few decades, multiple investigations have demonstrated the pervasive underrepresentation of OA in clinical trials^7–9. The 2004 landmark paper by Murthy and colleagues showed that older patients were underrepresented in cooperative group trials compared to younger participants. The study examined the four most common cancers (i.e. breast, lung, colorectal, and prostate) by incidence and mortality and found that OA participants made up one-third of trial participants in these cancer types, despite accounting for two-thirds of the cancer population for these four disease sites¹⁰. Expanding on this effort, a 2019 study³ examined a comprehensive collection of phase III cancer clinical trials, including both NCI-sponsored studies as well as industry-supported trials. Results from this analysis showed that trial participants were much younger than the disease-specific population groups; the average difference in median age (DMA) for trial participants was 6.5 years younger than the population median age. Furthermore, the study also showed that age disparities widened over time, with the DMA in trial enrollment decreasing by 0.2 years every year. These disparities were particularly evident in industry-funded trials, which make up the vast and increasing majority of late-phase cancer trials. As the proportion of industry-funded trials increases over time, it will be important to quantify why age disparities are worsening in industry trials versus cooperative group trials.

While this work has focused on late-phase trials, there also remains much to be understood about the incidence and impact of age disparities in early-phase studies, and how this may translate into questions regarding generalizability, validity, and success of later-phase studies. One recent study demonstrated age disparities in accrual to cooperative group-sponosored phase II trials¹¹, suggesting that earlier phase studies may have wider disparities than later phases, especially as these trials are early markers for safety and efficacy. Underrepresentation of OA in early-phase trials may amplify age-related differences in subsequent late-phase trials, the results of which are often used to guide clinical decisions and establish standard of care practices. As a result, the generalizability of findings from randomized trials may be limited for OA with cancer, thus complicating optimal treatment decisions for these patients.

Proportion of OA enrolled in FDA Registration Trials

Many retrospective studies have evaluated the age-related enrollment of patients with cancer in registration trials (trials that are moving forward for review by the FDA as a new drug or as an expansion to new indications). Compared to the mid-1990s, there was a slight increase (25% to 32%) in the proportion of enrolled OA in clinical trials from 1997 to 2000⁵. Overall, one study found that between 1995 to 2002, patients 65 years or older enrolled in trials of new cancer drugs with FDA approval made up 36% of the study population, as compared to the corresponding 60% of the U.S. cancer population⁷. Furthermore, the discrepancy between OA enrollment and representation in the U.S. cancer population increased with age, with those 70 and 75 years of age representing 20% and 9% of the trial population, while comprising 46% and 31% of the U.S. cancer population, respectively.

A more recent analysis of FDA-approved agents from 2005 through 2015 demonstrated a similar pattern of OA enrollment, with particularly marked disparities among the oldest subgroups of patients¹². OA patients continued to be underrepresented in this later set of registration trials (2005 – 2015)^12,13. Singh and colleagues noted that patients aged 70 years or older make less than a quarter of participants in FDA registration trials¹². Most strikingly, patients over 80 years accounted for approximately 16% of the population incidence of the relevant malignances (based on NCI Surveillance, Epidemiology, End Results [SEER] data), but only 4% of the FDA registration trial population¹², highlighting unique challenges in generalizability of regulatory decisions to older real-world patients.

Landscape of Clinical Trials with Older Adults

Estimation of risks and benefits of cancer therapies is often derived from clinical studies of younger and/or healthier patients, which can provide misleading treatment information for clinicians. A 2017 study¹⁴ points out that although OA-specific therapeutic trials exist for advanced non-small cell lung cancer (NSCLC), much of the data on NSCLC comes from larger studies that focused on younger populations (i.e. younger than 70 years). A recent analysis¹⁵ has shown that several newly-registered cancer drugs are not necessarily well-tolerated in OA patients. For example, an external group examined the BOLERO-2 trial’s findings for the general OA tolerability for the combination treatment everolimus and exemestane for endocrine-sensitive advanced breast cancer and found that OA patients experienced more frequent adverse events than younger patients. Such results have clear and important ramifications for OA with cancer and their physicians when attempting to make informed shared decisions regarding therapeutic options. These results further highlight the importance of OA representation in cancer clinical trials as a means to improve the generalizability of data and the quality of decision-making.

Barriers to Enrollment of OA in Clinical Trials

There is extensive literature that explores specific barriers to the adequate enrollment of older patients in cancer trials. A recent systematic review examined studies that assessed barriers to OA participation and found multiple barriers under four main categories: system-based (e.g. eligibility criteria, consent form language), provider (toxicity concerns, lack of personnel), patient (knowledge, transportation), and caregiver (preferences, burden)¹⁶. Figure 1 summarizes the various barriers to the enrollment of OA throughout the trial process, from design to post-implementation.

Figure 1.

Barriers Throughout Clinical Trial Development, Implementation, & Post-Market Settings

Notably, although there have been many proposed barriers to OA accrual, quantitative validation and independent association of OA representation with individual factors such as specific eligibility criteria have not been demonstrated. In a 2020 study, Ludmir and colleagues found that only 10% of late-phase cancer trials use eligibility criteria that explicitly excluded patients of a certain age, and found no association between age disparities and use of exclusionary criteria¹⁷. Furthermore, trials with the widest age disparities did not utilize explicit age-restrictive eligibility criteria. This suggests that there are other factors involved and the need for more research on potential actionable trial-level factors, including exclusionary eligibility criteria based on past medical history, laboratory parameters, and more, that predict OA exclusion and enrollment.

Turning toward barriers related to communication, shared decision-making and patient education are fundamental requirements for clinical trial consent and participation; therefore barriers to effective patient-provider communication may compromise the ability of patients to equitably enroll in clinical studies. The OA population may face unique challenges in this regard, particularly following the rapid rise in use of telemedicine technologies in response to the COVID-19 pandemic¹⁸. As patients increasingly rely on electronic and internet resources not only to learn about their disease but also available treatment options, including potential clinical trials, eHealth literacy will play a growing role in patient decision-making. However, a significant body of literature has demonstrated that OA are more likely to have limited eHealth literacy compared to their younger counterparts, thus creating an early structural barrier to clinical trial enrollment¹⁹. Efforts to improve the effectiveness of patient education, including interventions aimed at improving eHealth literacy among OA, could represent an important opportunity to increase the age diversity of cancer clinical trial participants. Furthermore, early adoption of formal telemedicine training for providers may improve the ability of OA patients to navigate an increasingly virtual healthcare environment. Lastly, trial investigators may consider deliberate use or allowance of traditional forms of communication such as phone calls or postal service with the aim of specifically catering to individuals with limited eHealth literacy, so as not to limit enrollment of OA in cancer clinical trials.

Beyond trial system/structural factors, are other barriers to trial enrollment that begin at the provider level with recruitment and accrual of OA patients for cancer trials. One study administered an anonymous survey to members of the Alliance for Clinical Trials in Oncology cooperative group, including physicians, nurses, and other research staff to determine awareness of OA enrollment issues in clinical trials²⁰. The most commonly reported perceived reason for exclusion of OA patients in clinical practice was that older patients usually do not meet eligibility requirements, specifically due to comorbidities or tumor characteristics. Other major reasons included regimen toxicity, transportation and time issues, patient/family enrollment preferences, concern for limited life expectancy, lack of relevant trials for OA, and lack of patient/family knowledge on RCTs. One study in Canada surveyed medical oncologists to analyze the proportion of OA and younger patients offered a clinical trial and explore why certain patients were not offered a trial. Investigators found that a common reason that OA patients were not offered trial participation is due to a lack of applicable RCTs for a patient’s cancer and stage²¹. This may point to OA patients having fewer trial options than younger patients, which reflects the barriers in trial design and trial portfolio for specific disease sites and stages. It could also reflect OA patients’ lack of accessibility to large urban academic centers, as a majority of patients are treated in the community setting, often in suburban or rural settings²². Furthermore, while OA were less likely to be offered trial enrollment than younger adults, they were just as likely to accept when offered the opportunity to participate in clinical studies²¹.

Furthermore, these structural and professional barriers⁷ are compounded with existing systemic and individual factors that historically underrepresented groups have to navigate, including socioeconomic status, access to care, caregiver characteristics, geographic barriers, and age itself. Nevertheless, it is important to note that patient perception seems favorable towards enrollment; one study pointed out that if offered enrollment in a breast cancer trial, OA were just as likely to accept enrollment compared to their younger counterparts²³.

Looking more broadly at the entire landscape of prospective clinical research beyond oncology, OA enrollment disparities have been shown in clinical trials for other chronic conditions. Older patients carry high burdens of cardiovascular disease, yet are underrepresented in clinical trials for cardiovascular therapies and interventions²⁴. This again raises the issue of generalizability and translability of trial findings to the real world, making treatment decisions more complex and challenging for older patients often irrespective of underlying diagnosis²⁵. Translating the therapeutic results from major phase III cancer trials to patient treatment will continue to be a challenge if these results are not adequately representative. As the burden of cancer continues to increase for OA, concerns regarding health equity should come to the forefront in addressing unequal access to clinical trial participation and subsequent evidence-based practice. These disparities will continue to pose serious implications for the future of OA health outcomes.

Addressing Clinical Trial Disparities

There are ongoing efforts from major institutional bodies, including national/international professional societies, government regulatory agencies, cooperative groups, and others, to increase the inclusion of OA in oncology trials (Table 1). A 2013 report from the Institute of Medicine emphasized the importance of improving the evidence base for treating OA patients with cancer by increasing enrollment of OA in trials and increasing the depth of data evaluations. Throughout the last few years, there have been workshops, meetings, and collaborations²⁶ between the American Society of Clinical Oncology, the NCI-NIA, and the FDA to produce the most robust guidelines to date. Draft guidance in 2020 was developed by the FDA²⁷ specifically for industry and industry-sponsored trials, addressing trial practices and methodological concerns to promote the inclusion of OA in biopharmaceutical-industry-sponsored cancer clinical trials.

Table 1.

Institutional Guidelines for Increasing Enrollment of Older Adults in Clinical Trials

Institutional Guidelines for Increasing Enrollment of OA in Clinical Trials
Organization	Year	Summary of Initiative	Specific Aim(s)	Type of Intervention	Barriers Addressed
FDA	1989	published guidelines51 for industry, “study of drugs likely to be used in the elderly”	information related to pharmacokinetics and conditions related to age should be collected	clinical trials	system/structural
	1994	published guidelines52 for industry, “studies in support of special populations: geriatrics”	formally defined older patients as ages 65 or older	clinical trials	system/structural
	2017	identified strategies to generate evidence at the workshop in the FDA/ASCO Geriatric Oncology Symposium	increase incentives and removing barriers to trial enrollment, including modifying eligibility criteria, expanding trial locations, and addressing patient concerns; expand type of evidence in trials, including outcomes; expand use of RWD by designing RWD studies (submitting to CancerLinQ) and improving quality of EHRs; strengthen collaboration between stakeholders to develop policy and advocacy solutions	clinical trials	system/structural
	2020	drafted guidance27 for industry called “Inclusion of OA in Cancer Clinical Trials” for better evaluation of the benefit-risk profile of cancer drugs in the OA population	recommended that sponsors should enroll OA in early phase studies and evaluate drug-drug interactions in these studies, incorporate age stratification in trial design, collect additional information for OA such as adverse event monitoring, and develop a plan to collect data on OA in the postmarket setting through trials or collection of RWD	clinical trials	system/structural
IOM	2013	published53 “Delivering High-Quality Cancer Care” found majority of cancers occur in OA and projected shortage of healthcare providers with sufficient expertise in geriatric care	increase breadth and depth of data by enrolling more OA into clinical trials and characterizing the study population through evaluation tools such as GA	clinical trials	system/structural
ASCO	2014	Reviewed IOM’s 2013 report with added recommendations	build on existing investment in quality monitoring and IT to enhance delivery of high value care	health systems management	system/structural, physician perspective
	2014	established ASCO’s Addressing Cancer Health Disparities Among Older Adults Task Force	create expert panels to develop guidelines and incorporate best practices	clinical trials	system/structural
	2015	published54 statement in JCO called “Improving the Evidence Base for Treating Older Adults With Cancer”	improve research environment by increasing FDA authority to incentivize and require research involving OA and use journal policy to improve reporting of age distribution and health risk profiles of OA participants	clinical trials	system/structural
	2017, 2021	generated joint research statement55 with Friends of Cancer Research and produced four work group manuscripts, then an impact analysis paper after the second round	recommended addressing eligibility criteria in age requirements for trial enrollment, organ dysfunction, and prior and concurrent malignancies	clinical trials	system/structural
	2018	JNCI publication22 that highlighted four new action items from an FDA-ASCO workshop based on previous recommendations	recommended that FDA work with sponsors to develop plans for enrolling OA, for sponsors to use ASCO-FDA-Friends criteria for organ dysfunction and malignancies, and for sponsors to work with social and behavioral scientists to consider needs of OA participants; NCI should implement the NIH Inclusion Across the Lifespan Policy	clinical trials	system/structural
	2019	joint meeting with FDA that had a fourth public workshop on “clinical outcome assessments in cancer clinical trials”	explored use of physical function as an outcome measure and to explore standardization of data collection, COA tools, and visualization of physical function as decision-making tools	clinical trials	system/structural
SIOG	2015	published56 guidelines from multiple task forces of SIOG	guidelines to fill the need for a more objective baseline assessment of OA cancer patients (and better estimate physiologic reserve)	clinical trials	system/structural
NCI, NIA, CARG	2010–2015	U13 conferences that discussed current level of research evidence, outline knowledge gaps, and propose updated and evidence-based research design	recommended inclusion of GA and physiologic/biologic markers in trials, increase enrollment for frail patients/patients who do not meet eligibility criteria, developing intervention studies to improve quality of survival in OA and frail adults	clinical trials	system/structural
NASEM Workshop	2021	virtual workshop that discussed the challenges and opportunities to improve the evidence base for treating OA cancer patients	discussed study designs based on prior recommendations, such as early phase drug development trials, registration trials, and post-marketing strategies; explored policy opportunities and RWD opportunities (such as administrative claims data)	clinical trials	system/structural

ASCO = American Society of Clinical Oncology; CARG = Cancer & Aging Research Group; FDA = U.S. Food & Drug Administration; IOM = Institute of Medicine; JNCI = Journal of the NCI; NCI = National Cancer Institute; NASEM = National Academies of Sciences, Engineering, & Medicine; NIA = National Institute on Aging RWD = real world data; GA = geriatric assessment; EHR = electronic health records

Directed efforts to test strategies that improve older patient enrollment have been limited. One recent review found only one relevant RCT, with results reported in 2005, that studied a physician-focused educational course and its impact on OA trial enrollment; unfortunately this RCT was negative with regard to improving OA enrollment²⁸. This signals a need to improve the quality of evidence on certain recommendations and guidelines. Below, we expand on certain areas for improvement in enrollment that touch on broad domains. Examining the impact of modernized eligibility criteria that incorporates geriatric assessment tools would be a relevant area for research. Another area of interest is exploring the potential of OA-specific trials. Furthermore, there are additional factors that affect enrollment outside of trial design, such as patient interest and knowledge, which certainly play a gatekeeping role in trial participation in this population. Efforts to improve the availability and quality of educational interventions designed specifically for OA may decrease initial barriers to participation.

Potential Utilization of Geriatric Assessments in Cancer Trials

Identifying ways that a robust geriatric assessment could aid in patient recruitment would be valuable. A comprehensive geriatric assessment is a multidisciplinary process that identifies patient-reported measures of medical, psychosocial, and functional limitations of OA to develop a plan to maximize treatment and overall health. Generally, this tool is referred to as a geriatric assessment (GA) when used to determine physiologic age, frailty, and fitness²⁹. Its use has risen considerably in geriatric oncology care, as it aims to improve management of age-sensitive impairments and concerns that are often overlooked. Information from these assessments could potentially predict treatment toxicity, providing both physicians and trial sponsors a better understanding of OA-specific conditions and comorbidities. Common domains³⁰ of GA most utilized in cancer care include functional status, comorbidity, depression, and cognition, and these have been recommended by the International Society of Geriatric Oncology (SIOG) to inform treatment course. Certain GAs, such as the Cancer and Aging Research Group (CARG) score³¹ and the Chemotherapy Risk Assessment Scale for High-Age Patients (CRASH) score³² can help predict toxicity risks before and during chemotherapy. There have been ongoing efforts to standardize which models of GA should be used in oncology, but to date no consensus single GA method has been recommended. The most common screening tools for patients who are referred for a GA are the G8 and the abbreviated comprehensive GA, both of which are designed specifically for the assessment of frailty in OA cancer patients³⁰.

Applying GAs in the context of cancer trials has emerged as an integral component of geriatric oncology. In the early 2000s, there were few studies^33,34 that incorporated GA.. One group examined³⁵ the implementation of GA in OA-specific clinical trials, and found that while the increase in use was significant (1% to 11%), the overall level of GA utilization remained low even at the end of the study period. Some barriers to implementing a GA include time and resources required, of which there are potential solutions30,34 such as utilizing a screening GA. More recently, several groups³⁶ have studied incorporating electronic versions of GAs into national cooperative group studies, and demonstrated feasibility. Muiltiple new and ongoing cooperative groups studies have begun using GAs as part of screening criteria. For example, the Cancer and Leukemia Group B (CALGB) has implemented GA successfully in multiple cancer cliniclal trials including CALGB 361006 and CALGB 360401, which investigated interventions for OA with hematologic malignancies^34,37. Similar methods are being developed and implemented in other diseases sites as well, including pancreatic, colorectal and lung cancers^38–40.

GAs have the potential to elucidate certain characteristics of an OA study population that could guide trial methodology and outcome. What’s more, utilization of GA could also widen the selection for patients who are frail and/or fare poorly on a GA, instead of only including the fittest OA patients based on GA results. With multiple ongoing prospective trials utilizing GAs, forthcoming results over the next several years will likely clarify and potentially validate the role of GAs in OA recruitment, enrollment, and stratification in cancer clinical trials.

Clinical Trials Exclusive to OA

Cancer trials that are specifically designed for OA represent another potential avenue to fill the unmet need for trial results that better inform the unique treatment challenges that older cancer patients face. Despite underrepresentation across cancer clinical trials in general, several late-phase trials have been designed specifically for OA, as shown in Table 2^41–45. OA-specificity allows for clinical questions to be asked and answered that more directly address tolerability of therapies, oncologic non-inferiority of alternative interventions with improved quality of life, and similar. This conceptual framework of directly examining tolerability and adverse events among a specific patient subgroup is to some extent a function of underreporting of subgroup-specific toxicity profiles. One review⁴⁶ examined the reporting of treatment efficacy and adverse events for OA cancer patients in phase III clinical trials and their results revealed inadequate reporting for both areas. Looking at over 150 clinical trials, this study found that only 39.9% and 8.9% reported treatment efficacy and adverse events by age. OA-exclusive trials allow room for standard-of-care practices and age-related changes in cancer pharmacology, while advancing clinical knowledge on adverse events and tolerability¹⁶. The potential for utilizing OA trials extends beyond age, but also age-related factors, such as poorer performance status and optimal treatment courses (e.g. de-escalating therapy).

Table 2.

Recommendations to Improve Older Adult Enrollment Mechanisms

Recommendations to Improve OA Enrollment Mechanisms
Domain	Specific Strategy	Examples
Trial Design	decrease reliance on selecting patients with relatively good performance status	- incorporate GAs
		- create trials for patients with poor GA scores and/or include patients with low and high scores on GA
		- include specific subsets of OA that with different levels of treatment tolerance
	incorporate real world data into trial design
	design OA-specific trials or trials that focus on OA	- trial only includes OA
		- include specific subgroups that incorporate OA without age restrictions and set a percentage goal for OA participants
Trial Enrollment	improve communication training for oncologists	- increase the quality of physician-patient discussion on trial participation
	improve patient education	- increase availability of public facing education material
		- assess eHealth literacy of OA patients and trial participants
Data Monitoring & Reporting	improve the accessibility of protocol methodology	- simplify PRO forms and trial protocols
	encourage development of early phase trials that include OA	- include reporting of stratified analyses of age cohorts
	incorporate patient-driven outcomes and needs	- establish improved PROs and QoL metrics specifically targeted for OA
	intensify side-effect monitoring	- design specific monitoring criteria for adverse events in OA to ensure safety

GA = geriatric assessment; PRO = patient reported outcomes; QoL = quality of life

Challenges of Clinical Trial Design

The challenges of designing and implementing cancer clinical trials with adequate OA representation are multifactorial (Figure 2). They are preceded by an initial step²¹ of enrollment: consideration by a physician, which is an important factor that influences a patient’s decision to participate in a trial. Trial design and subsequent implementation also affect OA inclusion and represent an important opportunity to improve the age diversity of patient enrollment. One common aspect of trial enrollment is the use of performance status, which arose in part to help assess how fit a patient is to determine a drug’s therapeutic index. The main metric used in many countries, including the U.S., has been the Eastern Cooperative Oncology Group Performance Status⁴⁷ (ECOG PS), which measures the patient’s functioning and capacity to perform activities of daily living. One review⁴⁸ in the United Kingdom highlighted that although the ECOG PS has been widely utilized and is known for its ease of usage, it can also be quite subjective and reductive. A recent paper⁴⁹ found that among a large cohort of phase III cancer trials (600 trials), over 96% of enrolled patients had ECOG PS 0–1; similarly, a vast majority of trials (88%) included exclusion eligibility criteria based on PS. Among trials that subseuqently led to FDA approval, fewer than 5% of enrolled patients had ECOG PS <1⁴⁹, again highlighting concerns regarding generalizability of findings and applicability of regulatory approvals for older patients or patients with poor PS. Perhaps the most transparent influence on OA representation is the presence, or lack thereof, of age-specific eligibility criteria and stratified age-specific data analysis. Multiple points in this selection process can affect OA cancer outcomes based on trial results and subsequent inapplicable clinical guidelines. To tackle these challenges, trial design should move away from solely relying on good PS for patient enrollment and move towards adapting trial designs to be more OA-inclusive; enrollment could include Gas and feature OA-specific design. Elucidating OA-specific results from clinical trials also requires relevant metrics, such as patient-reported outcomes and considerations regarding both oncologic / disease-control outcomes as well as quality of life metrics.

Figure 2.

General Timeline of Trial Design, Implementation, & Outcomes

Future Directions

Although there is increasing recognition of the importance of OA participation in cancer clinical trials, there remains a significant need for improved representation of this population in current and future trials. There is a need to bridge the gap between implementing trial design improvements and updated recommendations. Although there are guidance documents from regulatory agencies, there still is no clear mechanism for enforcement or accountability. Particular emphasis should be placed on industry-sponsored trials, which account for the vast and growing majority of cancer clinical trials. A recent paper⁵⁰ highlighted Pfizer’s efforts to be more inclusive of OA participants which include promoting diversity in early-phase trials, utilizing real-world evidence to examine quality-of-life outcomes, and incorporating GA screening tools. As shown in Table 1, many groups and organizations have published guidance to provide a blueprint for trial sponsors to follow in promoting OA enrollment.

A final area to highlight is the unique opportunity within radiation oncology to address issues of geriatric oncology and representation in clinical trials. Radiotherapy represents a treatment option that is often effective and may have a favorable toxicity profile specifically among older patients as compared with cytotoxic and/or multiagent systemic therapy. Given the comorbidities that often accompany older patient subgroups, noninvasive or minimally-invasive local therapies (such as radiotherapy) may afford patients the opportunity to optimize disease control without excess toxicity in light of concurrent comorbidities. Similarly, the potential for radiotherapy de-escalation and/or hypofractionation may allow for treatment completion in shorter times without additive toxicity or compromised disease control. In disease sites such as breast cancer, consideration should be given to the relative toxicity and quality-of-life profiles for multi-year administration of adjuvant endocrine therapy, compared with emerging utilization of very-short-course (1 week) adjuvant radiotherapy. To that end, we strongly encourage multidisciplinary design of OA-tailored clinical trials; we contend that discussion and design of such trials should heavily incorporate input from the radiation oncology community.

Conclusions

The last three decades of examining underrepresentation of OA cancer patients in clinical trials have not produced extensive progress to mitigate these disparities. Despite continued efforts to improve enrollment, age disparities within cancer trials may be widening over time. This problem is complex and requires a multi-stakeholder approach from the biopharmaceutical industry, oncologists, and other governing bodies. Results from FDA-approved clinical trials ultimately inform the course of cancer care; therefore, improving trial design and implementation is the key to better reflect the true proportion of the population affected by cancer. Future endeavors to expand eligibility criteria and include more OA in trials, while monitoring toxicity and treatment tolerance, remains crucial to improving cancer care.