Practical Geriatric Assessment in Older Adults With Non-Small Cell Lung Cancer Undergoing Stereotactic Body Radiation Therapy

Overview

National guidelines recommend that older adults with cancer undergo a special health assessment before starting cancer treatment. This type of assessment evaluates physical function, nutrition, social support, psychological well-being, medical conditions (both cancer-related and non-cancer-related), and cognitive function. The results can help doctors make better treatment decisions and determine whether additional support services-such as nutrition counseling, physical therapy, or social work-would be beneficial. Even though these assessments are recommended, they are not typically used because they need to be performed by a specialist and can take over an hour to complete. Given these challenges, a 10-15-minute assessment called the Practical Geriatric Assessment (PGA) was recently developed. The PGA can be completed by any healthcare provider and helps identify older adults who may need extra support alongside their cancer treatment. While the PGA has the potential to make geriatric assessments more accessible, the investigators do not yet know whether patients will find it useful or easy to complete. Additionally, it is unclear whether using the PGA will lead to more referrals for recommended supportive care services. This study aims to address these questions. The investigators will evaluate whether using the PGA impacts the number of patients referred to recommended supportive care services. Investigators will also evaluate how participants feel about completing the PGA, including how easy or difficult it is, and to assess the feasibility of implementing this survey on a larger scale. Finally, the investigators will use facial photographs and audio-visual data from the PGA to develop and evaluate artificial intelligence algorithm(s) to identify vulnerable patients who might benefit from additional supportive care services; namely, FaceAge, a validated deep learning model capable of estimating biological age from still facial images.

Comprehensive geriatric assessment (CGA) is one proven mechanism for delineating baseline care needs and improving outcomes in older adults with lung cancer. This type of geriatrician-led assessment, which captures functional ability, health, and socio-environmental situation, can be used to identify vulnerable older adults for whom tailored interventions might optimize care. However, CGA can be resource-intensive to perform, and may not practically possible in all settings, particularly given national shortages in geriatricians. At BWH/DFCI, retrospective work among patients with stage I-II NSCLC suggests fewer than 5% of patients receive CGA, despite 76% meeting national guidelines for this type of evaluation. These findings underscore the practical challenges of assessing geriatric needs, even in high-resource settings. To address these barriers, abridged instruments capturing the key domains of the CGA which can be completed by any provider within 10-25 minutes have recently been developed, including the practical geriatric assessment (PGA) tool. This screening tool, which is now recommended by American Society of Clinical Oncology (ASCO), the International Society for Geriatric Oncology (SIOG), and the Cancer \& Aging Research Group (CARG), has the potential capacity to delineate relevant baseline features in geriatric populations without creating undue provider burden. However, the feasibility of implementing the PGA and its acceptability to patients remain unclear. Further, relationships between the PGA and salient outcomes, including subsequent patterns of recommended care delivery, remain underexplored. To improve outcomes among high-risk subgroups interfacing with radiation oncology, including older adults with NSCLC undergoing SBRT, further interrogation of these practical factors is needed. Finally, the investigators will use facial photographs and audio-visual data from the PGA to develop and evaluate artificial intelligence algorithm(s) to identify vulnerable patients who might benefit from additional supportive care services; namely, FaceAge, a validated deep learning model capable of estimating biological age from still facial images.