This includes older adults receiving systemic therapy, including chemotherapy, targeted therapy, or immunotherapy (Type: Evidence based; Benefits outweigh harms; Evidence quality: High; Strength of recommendation: Strong).

The systematic literature review identified nine RCTs^3,4,14-20 that investigated the efficacy of GA with management for a range of primary endpoints in older patients with cancer (the results of RCTs identified by the systematic review are summarized in the Data Supplement). The GAP70+ and GAIN trials evaluated whether integration of GA and GAM would reduce serious (grade 3-5) chemotherapy-related toxic effects. GAP70+, a cluster randomized trial, enrolled 718 patients from 40 community oncology practice clusters; patients were 70 years old or older, had incurable cancer (solid tumors or lymphoma) with at least one identified vulnerability other than polypharmacy, and were receiving a new treatment regimen.³ Patients were randomly assigned to either a usual care group (n = 369) in which the treating oncologists received no GA summary or management recommendations, or to an intervention group (n = 349) in which oncologists received a tailored GA summary and management recommendations. The proportion of patients who had any grade 3-5 toxic effect within 3 months of starting a new high-risk treatment regimen was the primary endpoint of the trial; secondary endpoints included falls and polypharmacy. Analyses revealed that a lower proportion of patients (51%) in the intervention group experienced grade 3-5 toxic effects than patients in the usual care group (71%; relative risk [RR] 0.74 [95% CI, 0.64 to 0.86]; P = .0001). Over 3 months, patients in the intervention group also had fewer falls (12%) than patients in the usual group (21%; adjusted RR, 0.58 [95% CI, 0.40 to 0.84]; P = .0035), and had more medications discontinued (mean adjusted difference, 0.14 [95% CI, 0.03 to 0.25]; P = .015). Two recently published secondary analyses of data from GAP70+ reported, respectively, a significantly lower proportion of stage III and IV patients with lung cancer who experienced grade 3-5 toxicity in the intervention arm versus usual care (53.1% v 71.6%; P = .01)²⁴; and, among 623 patients from GAP70+ with follow-up Patient-Reported Outcomes Common Terminology Criteria for Adverse Events (PRO-CTCAE) data, fewer patients in the GA intervention arm reported grade ≥2 symptomatic toxicity compared to usual care patients (overall: 88.9% v 94.8%; P = .035; core symptoms: 83.4% v 91.7%; P = .001).²⁵

The GAIN randomized trial enrolled 613 patients from a National Cancer Institute (NCI)–designated cancer center. Patients were 65 years old or older with a solid tumor of any stage (71.4% had stage IV disease) who were receiving a new chemotherapy regimen; all patients had a completed GA. Patients were randomly assigned (2:1) to either a standard of care (SOC) arm (n = 203) or to a GAIN arm (n = 402). GA results were provided to treating oncologists for their review within 2 weeks of study enrollment for patients in the SOC arm. In the GAIN arm, a multidisciplinary team reviewed GA results and implemented interventions and referrals based on predefined GA thresholds. Both the patient and the treating oncologist were informed of the plan. The incidence of ≥grade 3 chemotherapy-related toxic effects was the primary endpoint; secondary endpoints included emergency department visits, advance directive completion, average length of stay, unplanned hospitalizations, unplanned hospital readmissions, and chemotherapy dose modifications and early discontinuations. An analysis of overall survival was done up to 12 months after the start of chemotherapy. In the GAIN arm, the ≥grade 3 chemotherapy-related toxic effects was 50.5% (95% CI, 45.6 to 55.4); in the SOC arm, the incidence was 60.6% (95% CI, 53.9 to 67.3), representing a significant 10.1% reduction (95% CI, −1.5 to −18.2; P = .02). With GAIN, there was a significant increase in advance directive completion of 28.4% compared to 13.3% with SOC (P < .001). There were no other differences observed between the two groups in the secondary endpoints evaluated. The reduction in the incidence of grade 3-5 toxicity with comprehensive geriatric assessment (CGA) versus standard care observed in the GAIN and GAP 70+ RCTs is supported by the systematic review and meta-analysis of six RCTs of CGA in older patients with cancer conducted by Chuang and colleagues.²¹

Two of the nine GAM RCTs identified by the guideline update systematic review investigated the effect of GAM on the completion of scheduled chemotherapy. Lund et al¹⁴ reported the results of GERICO, a phase III RCT of comprehensive GA-based interventions (n = 71) versus standard care (n = 71) in vulnerable (Geriatric-8 [G-8] questionnaire ≤14 points) patients ≥70 years old who were receiving 3-6 months of adjuvant or first-line palliative chemotherapy for stage II-IV colorectal cancer. Patients in the intervention group received the GA at or around the start of chemotherapy; the GA consisted of an assessment of comorbidity, a medication review, determination of psychocognitive function and nutritional, functional, and physical status with corresponding interventions such as referral to a dietitian, and a program of physical exercise. The primary endpoint of the trial was completion of planned chemotherapy without later dose modifications or delays; secondary endpoints included toxicity, quality of life, and survival. Compared with standard care patients, more patients in the GA intervention group completed scheduled chemotherapy (45% v 28%; P = .0366). Aspects of quality of life improved versus controls, with decreased burden of illness (P = .048) and improved self-reported mobility (P = .008) observed among intervention group patients.

Ørum et al¹⁹ similarly investigated the effect of GA on the completion of planned treatment among frail and vulnerable older (≥70 years) patients with cancer. In a single-center, randomized phase III trial, Ørum et al compared GA with a TFU intervention by a geriatric multidisciplinary team (gMDT; n = 152) to GA without (n = 149) a TFU intervention. The control group received a baseline GA with recommendations for interventions but with no TFU in the subsequent 90 days on any interventions that were initiated. The intervention group received a baseline GA with recommendations for interventions, and TFU on GA-guided interventions initiated. The form of the TFU varied among intervention group patients, but could involve multiple telephone or in-person contacts (home visits or at-hospital visits) between the patient and the gMDT. The primary outcome was the ability to complete initially proposed cancer therapy within 90 days; daily life activities, functional status, and need for hospitalization were secondary outcomes. There were no statistically significant differences between the study group in the proportion of patients who completed planned treatment: 61% of patients in the intervention group and 52% of patients in control group completed treatment (risk rate, 1.16 [95% CI, 0.95 to 1.4]; P = .14). There were also no differences between the groups observed in daily life activities, 90 days physical performance, or hospital admissions (55% of controls v 47% of intervention group patients; risk rate, 0.86 [95% CI, 0.69 to 1.07]; P = .19).

Two identified studies evaluated whether GA can improve quality-of-life outcomes in older patients receiving cancer treatment. In the multicenter, open-label, INTEGERATE (Integrated Geriatric Assessment and Treatment Effectiveness) phase III RCT, Soo et al¹⁶ randomly assigned (1:1) 154 patients to integrated oncogeriatric care (n = 76) or usual care (n = 78). Patients were ≥70 years old and had solid cancer or diffuse large B-cell lymphoma for which they planned to receive systemic anticancer treatment (chemotherapy, targeted therapy, or immunotherapy). The integrated oncogeriatric care intervention consisted of GA (review of medications, comorbidities; physical, cognitive, psychological, and social functioning, falls, frailty, nutrition, sensory impairment, advanced care planning, chemotherapy toxicity risk, and immunization status) followed by geriatrician consultation and implementation of GA-guided interventions. In 96% (68/71) of cases, patients in the intervention group received the GA after treatment initiation (median time to CGA was 14.5 days [IQR, 6-21] after the start of treatment). Participants in the usual care group did not receive the study-specific GA but could be referred to a geriatrician by their clinician. The primary endpoint of the trial was change in health-related quality of life (HRQOL) over 24 weeks; HRQOL, measured with the Elderly Functional Index (ELFI), was assessed at baseline, at week 12, at week 18, and at week 24. A range of secondary endpoints was included that assessed additional measures of functioning, mood, nutrition, treatment modification, health care utilization outcomes, and overall survival. Soo et al found that, compared to usual care, integrated oncogeriatric care resulted in improved HRQOL (overall main effect of group: t = 2.1, df = 213; P = .039; effect size = 0.38); the maximal between-group differences in HRQOL were observed at week 18 (mean difference in change, 9.8 [95% CI, 2.4 to 17.2]; P = .010, corrected P = .030, effect size = 0.48). There were also fewer unplanned hospital admissions (multivariable-adjusted incidence rate ratio, 0.60 [95% CI, 0.42 to 0.87]; P = .0066) with integrated oncogeriatric care versus usual care. No difference between the two groups in overall survival was observed.

In the two-group parallel (1:1) single-blind, multicenter (eight involved hospitals), 5C phase III RCT, Puts et al¹⁷ investigated the effectiveness of GA and management (GAM; n = 173) at 6 and 12 months versus usual oncologic care (n = 177) on quality of life (QOL) in older (≥70 years) patients diagnosed with solid tumor, lymphoma, or myeloma, and referred for first- or second-line palliative or adjuvant, curative (>54% of patients) chemotherapy, targeted therapy, or immunotherapy. The GAM intervention consisted of completion of a GA (functional status, cognition, mood, medications, mobility and falls, nutritional status, social support, and comorbidity) at baseline on or after treatment initiation for most patients; followed by GA-guided, evidence-based management interventions; and at least monthly follow-up calls from the intervention team registered nurse. The global QOL subscale of the European Organisation for the Research and Treatment of Cancer QOL Questionnaire core version 30 items (QLQ C30)³⁵ was the primary endpoint. Overall survival, functional status, grade 3-5 treatment toxicity, satisfaction, cancer treatment plan modification, and health care use were the secondary endpoints. Analyses indicated that GAM intervention did not improve QOL (difference in global QOL of 4.4 points [95% CI, 0.9 to 8.0] favoring the control arm); and there were no differences between the two groups in change in treatment plan, overall survival, treatment toxicity, or unplanned hospitalization and/or emergency department visits.

Two RCTs evaluated the effect of GA with management on short-term (6-month, 1-year) survival outcomes. In a phase III RCT, DuMontier et al²⁰ compared the impact of consultation with a geriatrician combined with standard oncologic care (n = 60) to standard oncologic care alone (n = 100) for older (≥75 years), prefrail or frail, transplant-ineligible adults with hematologic malignancies (lymphoma, leukemia, or multiple myeloma). Intervention group patients received embedded geriatric consultation, including a GA with individualized management and interventions, with a licensed geriatrician; the patient's hematologic oncologist provided standard oncologic care. One-year overall survival was the primary outcome; documented end-of-life (EOL) goals-of-care discussions and unplanned care utilization within 6 months of follow-up were the secondary outcomes. Forty-eight of the 60 patients (80%) in the intervention group completed ≥1 visit with a geriatrician. Consultation with a geriatrician combined with standard oncologic care did not improve 1-year overall survival compared to standard oncologic care (difference: 2.9% [95% CI, –9.5 to 15.2]; P = .65), and did not significantly reduce the incidence of hospital admissions, days in the hospital, or emergency department visits. The consultation intervention did, however, improve the odds of having EOL goals-of-care discussions (odds ratio, 3.12 [95% CI, 1.03 to 9.41]).

The EGeSOR (Effectiveness of Geriatric Assessment-Driven Interventions on Survival and Functional and Nutritional Status in Older Patients with Head and Neck Cancer) open-label, multicenter, randomized, parallel-group, controlled trial evaluated the efficacy of GA-driven intervention and follow-up in older (≥65 years old) patients with head and neck squamous cell carcinoma.¹⁸ Patients were randomly assigned (1:1) to receive GA-driven interventions and follow-up (n = 238) or SOC (n = 237). The intervention consisted of a pretreatment GA conducted by a geriatrician with oncology expertise, involvement of the geriatrician in shaping the cancer treatment plan, GA-driven intervention recommended by the geriatrician, and standardized geriatric follow-up for 2 years. The primary endpoint of the trial was a composite outcome that included 6-month overall survival, functional status, and nutritional status. There was no statistically significant differences between the study groups in the primary, composite outcome (41.0% v 38.0%; P = .53), or in any of the individual, component outcomes of death (31 [13%] v 27 [11.4%]; P = .48), weight loss (69 [29%] v 65 [27.4%]; P = .73), or functional status (fall in ADL score ≥2; 9 [3.8%] v 13 [5.5%]; P = .35).

The systematic literature review identified one trial that assessed whether providing a GA summary and corresponding GA-guided recommendations to oncologists could enhance communication about aging-related concerns. COACH (Improving Communication in Older Cancer Patients and Their Caregivers),¹⁵ a cluster-randomized trial, enrolled 131 oncologists, 541 patients, and 414 caregivers from 31 community oncology practices; patients were age 70 years or older with an advanced solid malignant tumor or lymphoma who had at least one impaired GA domain and were receiving cancer treatment with palliative intent. In addition, patients chose one caregiver to participate in the study. Community oncology practices were randomized to receive either the intervention (n = 17 practice sites) or usual care (n = 14 practice sites). The intervention involved providing a tailored GA summary with GA-guided recommendations to treating oncologists for each enrolled patient; usual care provided alerts to oncologists only if patients met criteria for cognitive impairment or depression. Patient satisfaction with communication about aging-related concerns, measured after the initial oncology visit, was the primary outcome. The number of aging-related concerns discussed during the visit, QOL, and caregiver satisfaction with communication about aging-related patient concerns were the secondary outcomes. Compared to the usual care group patients, intervention group patients were more satisfied with communication about aging-related concerns after the visit (difference in mean score, 1.09 points [95% CI, 0.05 to 2.13 points]; P = .04); and satisfaction with communication about aging-related concerns continued to be higher among intervention group patients over 6 months (difference in mean score, 1.10 [95% CI, 0.04 to 2.16]; P = .04). The intervention group's visits included more aging-related conversations than the usual care group's visit (difference, 3.59 [95% CI, 2.22 to 4.95]; P < .001). Finally, in the intervention group, caregivers were more satisfied with communication after the visit (difference, 1.05 [95% CI, 0.12 to 1.98]; P = .03). Two recently published secondary analyses of data from COACH reported that, compared to usual care, providing oncologists with a GA summary with tailored recommendations was associated, respectively, with an increase in oncologist-initiated conversations concerning physical performance and functional status with corresponding recommendations to address these concerns³⁶; and with an increased number of conversations regarding comorbidities per patient, with having a greater number of concerns acknowledged, and with a greater chance of having those comorbidity concerns addressed.²⁶

Clinically, the most important conclusion is that it is essential to do a GA for older adults with cancer to provide appropriate care when considering systemic therapy; when GAM is compared with SOC, it clearly leads to significantly less chemotherapy toxicity and improves adherence to chemotherapy. It also improves important patient-centered outcomes and communication, particularly patient and caregiver satisfaction with care, communications about aging concerns, and completion of advanced directives. These benefits are especially strong for patients who are older and are most vulnerable. These recommendations are strongest for older adults receiving chemotherapy, but the panel still recommends them for any systemic therapy, based on early evidence of similar benefits. We note that the GAP70+ study included patients who received immunotherapy in combination with chemotherapy as they were being treated in the NCORP sites. The Soo et al trial summarized in the guideline manuscript also included patients who received immunotherapy, which was 6 of the 154 total patients (4%; three patients per arm). A nonrandomized study of older adults receiving immunotherapy, while not including a GA, notes that such therapies were discontinued due to adverse events more than twice as often among patients age 90 years or older compared to others, suggesting some possible safety concerns that would benefit from GA.³⁷ An observational study comparing older (70+ years) with younger (<70 years) patients receiving immunotherapy showed that older adults who screened positive for frailty using the G-8 (60%) experienced higher hospitalization rates and shorter survival.³⁸ Other rigorous studies specifically including immunotherapies are ongoing. While the data for GAM are growing, it is reasonable to consider GAM recommendations as relevant for all older adults receiving systemic therapy since they are developed from geriatrics guidelines (eg, fall prevention for older adults who are falling). Questions regarding other systemic therapies, such as the rapidly growing field of immunotherapy,³⁹ are still being developed. Many patients who are older will receive immunotherapies, and they are also likely to benefit from a GA to identify management needs.⁴⁰ There is mixed evidence that GAM can improve QOL as well, although this is inconsistent across studies, depending on the interventions. Overall survival is not adversely impacted by GAM care, although the GA-guided interventions reduced toxicity rates and improved patient-reported outcomes. In these higher-resourced settings over the time intervals studied, GAM does not seem to consistently alter other outcomes, such as hospitalizations or emergency department visits. The evidence is sufficient that GAM, using formal GA tools, provides significant benefits to older patients with cancer and caregivers, especially the most vulnerable patients, and that GA should be conducted in such patients.

How does GAM improve these important outcomes without adversely impacting overall survival? The primary GAM interventions are care optimization in response to GA.^41,42 One is changes in decision making, primarily changes in treatments. For example, clinicians may choose to change their approach to management choices, such as decreasing chemotherapy doses,³ helping improve adherence to treatment, or more clearly defining care goals. There is evidence supporting each of these strategies being used successfully. Another consequence is a significant increase in the number of multidisciplinary interventions enacted based on the GA. These interventions include referrals to physical therapy, supportive care, social work, or nutrition when appropriate thresholds from validated tests are met. Clearly, there are interactions between these two optimization strategies; for example, adherence might be improved through dosing adjustments combined with a referral to nutrition to improve weight. In short, GAM results in improved decision-making, better targeting of interventions, and improved communications with patients and families, who are more satisfied with their care. In this regard, the updated systematic review of 65 publications (61 studies) conducted by Hamaker et al²² demonstrated that GA can lead to changes in cancer treatment plans and nononcologic interventions and improve communication about care planning and aging-related concerns. The specific mechanisms and details of how GAM works in different contexts and circumstances continue to be investigated.

As the evidence strengthens supporting the value of GAM for patients, families, and providers, the lack of uptake^5,6 remains a primary barrier to realizing the benefits. Implementation barriers to GAM include lack of understanding about available GA tools, perceived lack of resources, need for training, poor documentation, and system barriers.⁴³ Further improvements in the use of these tools require addressing these remaining barriers.

We note that, while the evidence is there for conducting GA in all older adults over 65 years old receiving systemic therapy, this may not be feasible in certain settings or certain populations. We, therefore, leave to the judgment of providers and practices how to adapt the GA to their own specific practice, based on volume, resources, and personnel. As described below, we take seriously these implementation barriers, and the Panel has developed tools, training materials (https://old-prod.asco.org/sites/new-www.asco.org/files/content-files/practice-patients/documents/2023-PGA-Final.pdf; https://youtu.be/jnaQIjOz2Dw; https://youtu.be/nZXtwaGh0Z0), and an updated ASCO website to aid in making the implementation as easy as possible.

Studies of clinical uptake of GA and GA implementation barriers. Because guideline-recommended GA use has been inconsistent,⁴³ the Expert Panel revisited the question from the 2018 guideline¹ of which GA tools clinicians should use to predict adverse outcomes in older patients who are receiving systemic treatment for cancer. The corresponding literature search and associated inclusion criteria were broad, designed to capture articles addressing the uptake by clinicians of GA implementation and perceived barriers to GA use in everyday clinical practice. The review identified 11 relevant publications.^{5,6,12,27,29-34} This research has consistently shown that the uptake of GA is generally modest. Thus, in a study of the use and knowledge of GA instruments among US community-based oncologists, Gajra et al⁶ found that just 13% of the 349 oncologists surveyed used GA for all of their older patients; 60% of oncologists did not use a formal GA for any of their geriatric patients; and 19% of oncologists reported that they were not aware of any validated GA instruments. Dale et al,⁵ based on the results of a survey of 1,277 providers (70% US-based, 63% in academic medicine, and 35% in private practice) who treated adults with cancer, reported that just 21% of respondents indicated that they performed a multidimensional GA using validated tools always or most of the time; 22% performed multidimensional GA some of the time; and 57% performed multidimensional GA rarely or never. A greater frequency of using a multidimensional GA with validated tools was associated with awareness of the 2018 ASCO GA guideline (aware of the ASCO guideline v unaware: 55% v 31%; P < .01). Fifty-three percent of respondents had indicated that they were aware of the ASCO guideline. In a web-based survey, Mishra et al³³ assessed the use of GA among transplantation physicians and the barriers to routine GA implementation in clinical practice to help determine candidacy for allogeneic hematopoietic cell transplantation among older (≥60 years) patients. The most common barriers to GA use in this study were uncertainty about which GA instruments to use and the lack of training in or knowledge of GA assessment tools. Other barriers included lack of time and a lack of adequate clinical support to implement routine GA.

Surveys of oncology providers in Australia, Canada, the Netherlands, and Mexico revealed similarly low or variable levels of GA use. In a survey of members of the Medical Oncology Group of Australia (N = 69), To et al³² found that GA had been requested by just 56% of respondents; 71% of respondents perceived that GA added value to clinical assessment alone. Puts et al³⁴ conducted an online survey of Canadian health care professionals' geriatric oncology learning needs and reported on the biggest challenge these clinicians faced in caring for older adults with cancer. These included a lack of resources (eg, a lack of geriatricians) and challenges related to pretreatment assessment and decision-making (eg, lack of knowledge regarding how to conduct a needs assessment of older adults with cancer). Comprehensive GA was used infrequently (n = 3) across centers in the study. Driessen et al³⁰ conducted a survey of GA practices among pulmonologists and radiation oncologists who treated older patients with non–small-cell lung cancer in the Netherlands, and they reported that use of GA in clinical practice varied widely across centers (n = 15) included in their study. All agreed that GA added value for treatment decision-making. Finally, in the study by Verduzco-Aguirre et al,¹² a sequential mixed-methods study of Mexican oncology professionals (N = 196) consisting of an online survey to cancer specialists followed by semistructured interviews of respondents based on their reported GA use, just 37 physicians or 18.9% of respondents reported routinely performing a GA.

The 11 articles identified by the literature search for this clinical question all addressed perceived or measured barriers to implementation of GA in oncology clinical practice. The most commonly cited barriers were the time required to perform GA,^{5,6,12,27,29-32} the lack of adequate resources (qualified staff and financial support) to integrate GA into routine clinical practice,^5,27-29,32 and the lack of relevant knowledge or training.^5,12,29,30 The growing understanding of barriers to GA implementation in everyday practice gleaned from this research has naturally led to calls for innovative and more practicable approaches to GA and GAM. The PGA that is recommended by the Expert Panel as an option for GAM is one such approach.

A GA is a multidimensional assessment to identify patient risks, prognosticate for outcomes, and identify targets for interventions. In order to perform these functions, a GA must include a sufficient number of domains, which minimally include physical and cognitive function, emotional health, comorbid conditions, polypharmacy, nutrition, and social support. The PGA is an attempt to translate the evidence that has been generated for the GA, and adapt it for use by most oncologists, with an emphasis on the practical considerations for its widespread adoption. The Panel recognizes that there is no perfect one-size-fits-all solution to the adoption of a GA in specific practices, as different practices settings will have different resources, different personnel, and different comfort levels for conducting GA.

Having said this, the Panel wanted to offer a potential solution to the challenge of adoption that could be integrated into variably resourced clinical settings. The tools chosen through the Delphi process⁶⁹ provide usable information on which to act. The ASCO Older Adults Task Force, in conjunction with SIOG and CARG, has also created additional resources, including a companion article⁹¹ detailing the PGA with scoring instructions, suggestions for actions to take based on the scoring, video guidance on how to conduct the PGA, and references for evidence to support its use. Through ASCO's website, there are updated details on how to use the information generated. Finally, if there is widespread adoption of the PGA in multiple locations, the opportunity for real-world data collection and evidence based on a consistent measurement will ensure continued refinement of the PGA into the future.

The Panel recognizes the challenges of feasibility, including of the PGA. The PGA represents the cumulative wisdom from two groups of cancer and aging experts from ASCO and CARG. It is the most concise version of the GA that is evidence-based and aligns with the available data. It has been shown that the GA is less burdensome than other common interventions to oncology practice.⁷⁰ Given that clinical judgment has not proven reliable in assessing frailty, the PGA is recommended to be used as the appropriate evaluation for older adults. Specific questions on the use of the PGA are also contained in a companion article by Williams et al⁹¹ in JCO Oncology Practice. The Panel acknowledges the challenges of adopting a new practice, and recognizes that judgment will be exercised by individual practices on how best to incorporate these assessments as much a possible into their practices.