Can Geriatric Assessment Measures Be Used to Determine Cancer Treatment Vulnerability in Nongeriatric Patients?

Concurrent chemoradiation, the mainstay of treatment in locally advanced head and neck cancer, is a challenging endeavor even for the fittest of individuals. On one hand, it confers superior therapeutic outcomes compared with either chemotherapy or radiation alone; on the other hand, the high-intensity treatment is associated with substantial treatment toxicity that in turn leads to treatment schedule interruption or incompletion, hospitalization, or even death.^1–4

Several retrospective studies have demonstrated that factors such as medical comorbidities, poorer performance status, sociodemographic characteristics, and older age are associated with impaired ability to tolerate treatment.^5–8 A handful show that tools that screen for geriatric impairments (eg, Vulnerable Elders Survey-13 and G8 screening tool) are able to predict radiation therapy (RT) incompletion and mortality in older patients with head and neck cancer receiving chemoradition.^9–12 Yet, few that prospectively investigated the tolerability of treatment using a comprehensive geriatric assessment (GA) in nongeriatric individuals. In addition, none to our knowledge has compared results on tools and prediction of outcomes between older and younger individuals. Tools that could help clinicians predict the likelihood of such events would be highly valuable for guiding treatment decision-making in cancer patients of any age.

In this issue of the Red Journal, Chou et al sought to determine whether GA measures could be used to predict adverse outcomes in nongeriatric individuals undergoing curative-intent head and neck chemoradiation therapy.¹³ Their report examined a prospective cohort of 461 patients with locally advanced head and neck cancer from 3 hospitals in Taiwan who underwent definitive chemoradiation therapy. The authors used comprehensive GA (CGA) consisting of evidence-based tools to capture geriatric impairments for 7 geriatric domains; the GA was advocated by the Asia-Pacific guidelines for the management of frailty in older adults and validated in a population of community-dwelling older Taiwanese persons, as part of the initial patient evaluation. The questionnaire was given once at the beginning of treatment and consisted of 53 self-reported items across the domains of function, nutrition, comorbidity, mobility/falls, mood, cognition, and medication use, to assess treatment vulnerability. Social support, an important component of any CGA, was not included. “Vulnerability” was defined as impairments present in 2 or more CGA domains. Results demonstrated that patients who were vulnerable were more likely to experience severe treatment toxicities, treatment incompletion, hospitalization (and also a longer stay), tube feeding, and mortality.

The terms vulnerability and frailty are sometimes confusingly interchanged, and it is important to understand the distinction between the 2 concepts, both of which are well-studied in geriatric medicine. Vulnerability is related to frailty but is by no means equivalent. A “vulnerable” individual is a person who is at increased relative risk of morbidity or mortality as a result of decreased reserves or resources in physical, cognitive, psychological, social, or environmental domains.¹⁴ Frailty is defined as a state that affects an individual who experiences an accumulation of deficits in the aforementioned domains, leading to adverse outcomes such as disability or mortality.¹⁴ Frailty refers to a phenotype with 2 commonly used operational descriptions. A physical phenotype was originally described by Fried et al as the presence of unintentional weight loss, exhaustion, weakness (low grip strength), slow walking speed, and low levels of physical activity.¹⁵ A multidomain phenotype has also been described and is based on a CGA that measures cumulative deficits in physical, cognitive, functional, and social domains.¹⁶ In older adults, vulnerability can be conceptualized as a “prefrailty” state.¹⁷ Although using 2 impairments in GA domains to define vulnerability is reasonable, there is no standard definition. Medical oncologists have long recognized that functional, cognitive, and mobility impairments, which are common geriatric conditions, affect the tolerability of chemotherapy. Validated instruments measuring these vulnerabilities are able to predict chemotherapy toxicity better than traditional oncology performance scales and are now part of chemotherapy management guidelines for older adults.¹⁸

In their work, Chou et al collected data from GA measures advocated by guidelines for management of frailty in community-dwelling Taiwanese among nongeriatric head and neck cancer patients to assess treatment vulnerability. Conceptually, this is an interesting and novel idea because it sheds light on the fact that vulnerability to treatment could be more broadly applied beyond older adults and need not be limited to geriatric medicine. It would make sense that certain nongeriatric patients with cancer might be at risk for worse outcomes after (or during) treatment and that the comprehensive health domains commonly evaluated in older adults could potentially apply to younger patients as well (eg, nutritional status [weight loss], polypharmacy, and psychological state). In nongeriatric patients undergoing head and neck surgery, for instance, scores on a modified frailty index have been shown to predict morbidity and mortality.¹⁹ In patients with end-stage liver disease, the Fried frailty phenotype has been shown to manifest in patients as young as 60 years of age and correlate with increased mortality after liver transplantation.²⁰

As proof of concept, the article by Chou et al supports the notion that use of a CGA survey may help guide clinical decision-making regarding intensive cancer therapies where cure is the goal. Interestingly, there were more treatment-related deaths in nonvulnerable individuals compared with vulnerable individuals in the reported study. This phenomenon may have occurred because patients who were deemed not vulnerable at the initial assessment may have become vulnerable during the course of chemoradiation therapy but were not identified as such because follow-up assessments were not conducted during treatment; thus, there was no trigger to alter the treatment plan to accommodate any newly found vulnerability. In addition to understanding which domains are important for predicting vulnerability in younger patients, management is important as the next step. GA should not merely include the evaluation but should also help guide decision-making and geriatric interventions. These interventions could help patients who are younger as well.

Several questions remain as to whether some domains within this assessment tool may be more (or less) relevant than others for younger patients; whether new domains should be added, including a physician-reported or performance-based component; and whether the assessment should be repeated during the course of treatment. For instance, nutritional status was an important driver of vulnerability in younger patients in the studied cohort, whereas comorbidities and cognitive impairment appear to be more important factors affecting vulnerability in older adults. Physical examination is typically an important part of any CGA, but it was not included in this study. A performance-based assessment, even perhaps one that includes grip strength (a key measure of weakness in a frailty assessment) or other physical performance measures such as the Timed Up and Go or the Short Physical Performance Battery, may be useful to aid elucidation of a frail phenotype in nongeriatric patients. In addition, a measure of social support was noticeably lacking in the assessment tool, yet the domain of social support is known to be an important factor in treatment adherence among cancer patients. Finally, the timing of the assessment could be further optimized to allow for a longitudinal examination to capture vulnerabilities that may only manifest after a treatment course is well underway.

In conclusion, this report represents a laudable initial effort to address a gap in our current ability to assess treatment vulnerability in nongeriatric patients with head and neck cancer. There remains ample room to further refine and validate the assessment tool in subsequent studies in order to better fit the younger population being studied. Two key messages, however, are distilled: (1) A comprehensive assessment of vulnerability should be a core element in all new patient evaluations in locally advanced head and neck cancer regardless of age, as it facilitates the delivery of optimal patient-centered care, and (2) further tailoring of the assessment−and interventions−to help younger versus older patients is greatly needed.