The growing worldwide population of older adults has been underrepresented in clinical trials that set the standards for oncology care. In the Southwest Oncology Group (SWOG) analysis of data on 16,396 patients enrolled in 164 trials during the 1990s, patients with OC older than 65 accounted for only 30 % of all included patients [15]. Similarly, in a recent Surveillance, Epidemiology, and End Results (SEER) survey, only 9 % of patients with cancer older than 75 were included in clinical trials of new therapies [16]. Although in the last decade improvement in the enrollment of older adults in clinical trials has occurred, the representation of older adults on trials is still far from ideal, and the enrollment fraction of patients 75 and older in clinical trials (i.e., the number enrolled divided by estimated cancer cases) worsens with increasing age. Perhaps the most concerning statistics is that older adults are underrepresented in registration trials of new drugs approved by regulatory authorities, despite the FDA’s recommendation to include older adults in clinical trials [16]. As a result, the standard drug dosing and expected side effect profile is primarily derived from a younger cohort of patients.

Two possible strategies may be adopted: to increase the number of elderly patients in conventional trials for adult or to design trials specific for elderly people who are not candidates for conventional therapies. However, how can we define the patients that can enter and those not suitable for conventional clinical trials?

Chronological age is no longer a valid exclusion criterion, and the majority of adult cooperative group clinical trials no longer specify an upper age limit. In fact, it is believed that the clinical applicability of the results of a cancer treatment trial depends largely on whether the study participants are representative of the population of interest [17]. Hence, to ensure that clinical trial results are generalizable to older patients, trials should include them in numbers proportional to their distribution among the cancer population.

However, the majority of cancer trials prohibit participation of people with hematologic, hepatic, renal, or cardiac abnormalities. Exclusions on the basis of hypertension, cardiac disease, hematologic, or pulmonary function abnormalities resulted in 8.6 %, 5.3 %, 14.1 %, and 9.3 % lower enrollment of older patients, respectively, in National Cancer Institute (NCI) trials from 1997 to 2000 [18]. Aging is associated with several physiologic changes in organ function that could alter drug pharmacokinetics and have an impact on cytotoxic chemotherapy tolerability and toxicity [19]. Renal function, as indicated by the glomerular filtration rate, is reduced with age [20]. Bone marrow reserves diminish with increasing age, and myelotoxicity can be substantially increased [21].

Multiple comorbid diseases are common in elderly cancer patients. The overall burden of comorbidities has a negative impact on patient’s survival [22, 23] and on the patient’s ability to tolerate treatment or may be a contraindication for cancer treatment (e.g., trastuzumab in patients with congestive heart failure).

In an observational study of 17,712 patients with a new primary diagnosis of cancer, survival was found to be inversely related to age, and comorbid medical conditions had an impact on survival independent of cancer stage [23].

Thus, the only way to encourage the participation of elderly people in conventional clinical trials would be to clearly define in the protocol, when there is no clear contraindication, dose adjustment, and titration strategies for patients with mild functional abnormalities related to comorbidities. In fact comorbidity conditions and toxicity are considered by physicians as the most frequently potential barriers to trial enrollment [18, 24]. Other factors that physicians consider as a barrier to enrollment are the lack of support for the older patient to manage side effects at home, transportation needs in case of toxicity, or simply that life expectancy of some patients is too short to justify participation in clinical trials [18, 25].

However, even in the case where the protocol clearly defines rules to encourage the participation of elderly people, including wider inclusion criteria and defining clear dose adjustment rules according to comorbidities, there is a proportion of aged patients that we do not feel to be safely included in conventional trials and that in our view should be included in dedicated elderly trials.

In other words these patients are vulnerable or fragile and at serious risk of excess toxicity from surgery or standard medical treatments.

Frailty is defined as a state of decreased physiological reserve and increased susceptibility to suffer disability in response to stressors [26–28]. How frailty is measured varies between a physical assessment and a more physiological assessment. Makary and colleagues used a validated scoring system of physical frailty in a prospective study of patients older than 65 presenting for elective surgery. The frailty score was based on five domains, which included weight loss, weakness, exhaustion, low physical activity, and slowed walking speed [26]. This frailty score was then compared with more conventional preoperative assessments such as the ASA score, Lee’s revised cardiac index, and Eagle score. Frailty was found to correlate with postoperative complications with an adjusted odds ratio of 1.78–2.13 for intermediately frail patients and 2.48–3.15 in frail patients. Frailty was also correlated with increased length of stay (65–89 % longer stays for frail patients) and nontraditional discharge [28]. Other investigators have looked at physiological frailty as a predictor of outcome in preoperative patients. Domains studied included comorbidity, function, nutrition, cognition, geriatric syndrome assessment, and extrinsic frailty (social support) [28]. Using this assessment of frailty, following factors found were associated most closely with 6-month mortality: cognitive dysfunction, lower albumin, having fallen in the previous 6 months, lower hematocrit, functional dependence, and increased comorbidities. Having four or more of these markers predicted 6-month mortality with a sensitivity of 86 % [28]. Although older adults have been identified as being vulnerable to side effects from cancer therapy, few oncology studies to date have specifically incorporated baseline metrics for measuring health conditions other than functional status (Eastern Cooperative Oncology Group [ECOG] or Karnofsky performance status [PS]) to identify individuals most at risk. Until recently, a consistent definition of frailty remained elusive.

Thus, the clear definition in the protocol of the type of elderly people that are enrolled is the first mandatory step when designing a clinical trial in this setting.

The cutoff point at which an adult is considered “old” has not been well defined. Traditionally, a chronological landmark has been considered the age of 70. After this age, it is reported an increased incidence of age-related physiological changes, risk factor for altered pharmacokinetics and pharmacodynamics, potentially leading to increased treatment-related toxicity [29].

However, this cutoff is arbitrary. Aging is a highly individualized process, and all changes involved in this process cannot be predicted solely on the basis of chronological age. Aging has been defined as a loss of “entropy and fractality” [30]. Loss of “entropy” implies a progressive reduction in an individual’s functional reserve, whereas loss of “fractality” implies a progressive reduction in the ability to coordinate different activity and negotiate the environment.

Trials should be designed to capture functional reserve and ability of elderly patients. Furthermore, coexisting medical problems also led to significant use of medication; polypharmacy has been reported as a significant factor which contributes to increased chemotherapy toxicity.

Older adults may be particularly susceptible to polypharmacy, given the increased number of comorbidities in this population. Polypharmacy raises the likelihood of adverse drug reactions in older cancer patients and has been associated with increased mortality [31].

A strict assessment of comorbidities and polypharmacy in clinical trials is required because it will determine the patients’ life expectancy and clarify the generalizability of the results of the trial. CRF must be clearly designed to capture this scenario.