How to Use a Geriatric Assessment in an Older Patient With Acute Myeloid Leukemia

A 75-year-old woman presents to the emergency department with pancytopenia and a two-week history of fatigue. Her medical history includes hypertension, hyperlipidemia, diabetes, and osteoarthritis. Her Eastern Cooperative Oncology Group (ECOG) performance status was 1. Laboratory findings were white blood cell count, 2.0 × 10⁹/L with 25 percent blasts; absolute neutrophil count, 1.2 × 10⁹/L; hemoglobin, 10.4 g/dL; and platelets 112 × 10⁹/L. Renal and liver function, prothrombin time/partial thromboplastin time, fibrinogen, uric acid, and phosphorus were normal. Lactate dehydrogenase was elevated (350 U/L). A bone marrow biopsy was performed, which showed acute myeloid leukemia (AML) with no myelodysplastic changes and blast count of 39 percent. Flow cytometry was positive for CD13, CD33, CD34, CD117, and human leukocyte antigen–DR. Cytogenetics were normal, fluorescence in situ hybridization did not show a RUNXITI/RUNX1-, MLL-, PML/RARA-, or CBFB-specific rearrangement. Mutation panel was negative for FLT3, NPM1, CEBPA, or IDH1/2 mutations.

What is the evidence for performing geriatric or fitness assessments?

A geriatric assessment is a multidimensional assessment used to evaluate the health of an older adult. It uses validated measures to assess geriatric domains including physical function, comorbidities, cognitive function, psychological health, nutritional status, medications, and social support. It may sometimes be referred to as a “fitness” or “frailty” assessment since a geriatric assessment can help determine fitness, but these terms are not synonymous. A geriatric assessment uncovers vulnerabilities that are not detected in routine clinical practice¹  and predicts morbidity and mortality,^2,3  which may help with treatment decision-making. A geriatric assessment can also guide supportive care interventions and improve patient-oncologist communication as well as patient and caregiver satisfaction.^4,5 

In a prospective study of older adults with largely good performance status (78% ECOG ≤1) who were scheduled to received intensive therapy for AML, geriatric assessment detected a high prevalence of impairments: 41 percent basic activities of daily living (ADL), 50 percent instrumental ADL (IADL), 52 percent objectively measured physical performance (Short Physical Performance Battery [SPPB] <9), 28 percent cognitive function, 40 to 59 percent psychological distress, and 42 percent Hematopoietic Stem Cell Transplantation Comorbidity Index (HCT-CI) score greater than 1. Both impaired cognition and objective physical performance were associated with worse overall survival.²  Several other studies in older patients with AML and myelodysplastic syndromes (MDS) also found that impaired ADL, impaired IADL, symptoms (fatigue, pain), and higher comorbidity burden were predictive of worse survival.^6-8  In a retrospective study, polypharmacy (prescription medications >4) was associated with higher 30-day mortality and lower odds of achieving complete remission.⁹ 

Concordant evidence supporting the value of geriatric assessment has been demonstrated across varied hematologic malignancies including lymphoma, multiple myeloma, MDS, and in the pretransplantation setting.^10-16  The predictive value extends beyond survival and includes treatment-related outcomes and health care utilization. For example, among older adults with MDS receiving azacitidine, those with impairments in IADL, cognition, and mobility were more likely to discontinue therapy, and those with IADL impairment and higher comorbidity burden had worse survival.¹⁰ 

Among specific domains assessed in a geriatric assessment, measures of physical function and cognition seem to be high yield for predicting increased risk of morbidity or mortality in the context of AML therapy and among those with hematologic malignancies more broadly.^2,6,15,17  Self-reported basic ADLs and instrumental ADLs can identify functional vulnerability associated with worse outcomes among AML patients receiving less intensive therapy.^6,8  IADL impairment has also been associated with higher health-care utilization and worse survival among adults 75 years and older with varied hematologic malignancies.¹⁸  Objective testing of physical function is more sensitive than self-report and adds value to assessment of vulnerability. As mentioned, the SPPB (a composite test including gait speed, chair stands, and balance testing) was predictive of survival among older adults treated intensively for AML. Gait speed alone may be an efficient screening test. Gait speed has been a consistent predictor of health outcomes in older populations.¹⁹  Among adults 75 years or older with varied hematologic malignancies, decrease in gait speed was associated with higher mortality, greater odds of unplanned hospitalizations, and emergency department visits.¹⁵  Similarly, in the same population, a simple five-word recall measure of working memory identified prevalent cognitive impairment which was associated with worse survival.¹⁷ 

Finally, there is evidence to support repeating geriatric assessment measures during treatment to assess the impact of therapy on physical, cognitive, and emotional health. Geriatric detected vulnerabilities are not static but can be influenced positively and negatively by treatment and AML complications. For example, among older adults with AML receiving intensive therapy, clinically meaningful declines in physical function and changes in emotional well-being can be measured at the time of remission, which may influence tolerance to subsequent therapies and quality of life.^20,21  This information can help direct ongoing personalized supportive care and decision-making.

How are geriatric or fitness assessments performed in the clinic or inpatient setting, and how much time do they take?

A cancer-specific geriatric assessment generally takes approximately 15 to 30 minutes, keeping in mind that most of the assessments are self-reported and can be administered without assistance from a health-care professional²⁰  Objective assessments (i.e., gait speed and cognition testing) are administered by staff and typically take approximately five minutes, including instructions if performed routinely.^20,1,22,23  Geriatric assessment has been successfully implemented in AML trials conducted in the cooperative group setting.²⁰  The length of time depends on the number of assessments included. In general, it is recommended that each of the geriatric domains be assessed, but some of the domains may be obtained from the electronic medical record (e.g., comorbidities, medications, and weight).

Practices that use geriatric assessment often have patients complete the self-reported portion of the assessment prior to clinic visits in the outpatient setting. These are mailed out or sent electronically in advance to patients for completion at home, or they are administered as part of a check-in process prior to the clinic visits. In the inpatient setting, paper surveys or electronic surveys via a tablet can be administered. Any clinic/hospital staff or health-care professional (e.g., patient care technician, nurse, advanced practice professional) can be trained to perform the objective assessments.

If time and resources prohibit the incorporation of a full geriatric assessment, and there are no geriatrics or geriatric oncology services available, certain geriatric domains or assessments that are more prognostic may be considered. Physical function, cognition, and comorbidities have been shown to be of highest yield. These assessments may include the following: 1) physical function (self-reported questions such as ADL and IADL or objective testing such as a SPPB, Timed Up and Go, or gait speed), 2) comorbidity burden (HCT-CI or Cumulative Illness Rating Scale [CIRS] or CIRS-Geriatric), or 3) cognition (Mini-Mental State Examination or Blessed Orientation-Memory-Concentration or 5-word recall).

How do geriatric or fitness assessments guide interventions?

There is increasing interest in using geriatric assessment to allocate treatment for older adults. At this time, the evidence supporting an optimal assessment strategy to allocate treatment in AML is insufficient given the complex interplay of fitness and disease biology. Randomized trials that include geriatric assessment measures are lacking. Instead, a geriatric assessment provides information to consider during an informed discussion about treatment and can guide supportive care interventions. If vulnerabilities are identified earlier in the disease course, interventions can be instituted to reverse impairments, with the potential to improve treatment tolerance and outcomes. For example, physical therapy or an exercise program could be recommended to address functional impairment.^24,25  For cognitive impairment, strategies to prevent delirium should be considered, as well as engaging family, caregivers, and nurse navigators during treatment. De-prescribing can be instituted for polypharmacy to reduce drug interactions, and depressive symptoms can be addressed with counseling and/or pharmacologic therapy.

The patient underwent a geriatric assessment. She was independent in her ADLs and most IADLs, except shopping. She reported one fall in the previous six months. Her HCT-CI was 1 (diabetes mellitus; recent HbA1c, 7.5%). Medications were amlodipine, lisinopril, atorvastatin, acetaminophen, metformin, aspirin, omeprazole, and vitamin D. She had not lost weight. She is married with two daughters. She lives at home with her husband but both daughters live within 10 miles. She scored 26 out of 30 on the Mini-Mental State Examination and 2 out of 15 on the Geriatric Depression Scale (negative screen for both). She scored 7 out of 12 on the SPPB (<9 indicates physical vulnerability).

The patient’s geriatric assessment revealed several vulnerabilities, including impairment in IADL and SPPB, polypharmacy, and a fall. These impairments indicate potential higher risk for complications and possibly shorter survival compared to an older adult without these vulnerabilities. After informed discussion, she was enrolled onto a clinical trial evaluating the combination of azacitidine and venetoclax with a novel therapeutic agent. Several supportive care interventions were recommended based on her geriatric assessment. She was referred to physical therapy and started an exercise program. A gait evaluation was performed, and a cane was recommended. Home safety evaluation was performed, and the patient received a medical alert bracelet due to risk of falls. Her daughter was asked to check in with the patient daily. After medication review with the oncologist and pharmacist, omeprazole was discontinued.

As demonstrated in this case, geriatric assessment can inform management of older adults with AML. Routine collection of geriatric assessment measures in clinical trials will further inform treatment selection and personalized care.