Cancer in Older Adults

Epidemiology & approach

• Demographics: ~60% of new cancers and ~70% of cancer deaths occur at age ≥65. By 2030 ~20% of the US population (~70 million) will be >65. Median age at diagnosis for most tumors is ~68 to 74 yr.
• Definition: age ≥65 is generally "older" in developed countries; SIOG defines geriatric oncology as age >70 with cancer.
• Heterogeneity: chronologic age is not biologic age. "Fit" vs "vulnerable" vs "frail" guide therapy intensity.
• Higher death rate in older adults: more aggressive biology in some cancers, competing comorbidity, reduced physiologic reserve, undertreatment, and access barriers.
• Core principle: integrate geriatric assessment (GA) before systemic therapy in older adults, especially before intensive treatment.

Aging and cancer: shared biology

• Age is the single most important cancer risk factor. Most hallmarks of cancer overlap with hallmarks of aging.
• Mechanisms linking aging to cancer: longer duration of carcinogen exposure and accumulated somatic mutations; increased susceptibility of aging cells to transformation; impaired DNA repair; oncogene activation / tumor-suppressor loss; telomere shortening and genetic instability (telomeres shorten ~1%/yr from age 30; telomere length predicts death after 60; telomerase reactivates in malignant cells).
• Microenvironment: accumulated senescent cells and higher IL-6 (the "geriatric cytokine") plus a senescence-associated secretory phenotype (IL-1, MMP-3) that can promote premalignant proliferation.
• Decreased immune surveillance (immunosenescence) with age; checkpoint expression on T cells rises with age.

Age-related physiologic changes

• Physiologic decline begins at age 30 (~1%/yr), varying by organ and individual.
• Thinning skin, easy bruising, reduced cardiac reserve and myocytes, increased vascular stiffness, reduced GI motility/absorption, and reduced marrow reserve (greater and longer cytopenias).
• Declining renal and hepatic clearance and increased neurotoxicity sensitivity drive dose adjustment; dose by creatinine clearance (Cockcroft-Gault), not serum creatinine, because lean body mass falls.

Life expectancy estimation

• Wide variability within an age group (e.g., a US woman at 85 has ~10, ~6, or ~3 yr in the top, middle, and bottom quartiles). Providers routinely overestimate prognosis and may underestimate life expectancy in those >75.
• ASCO recommends validated tools (Lee index or Schonberg index, available at ePrognosis) to estimate non-cancer life expectancy, so competing mortality can be weighed before offering screening or intensive treatment.
• Lee index: 12 predictors (age, sex, comorbidities, functional measures, smoking); higher scores predict higher 4-year mortality (0 to 5 points ~≤4%; ≥14 points ~64%).

Treatment approaches

• Surgery: age alone is not a contraindication; assess physiologic reserve, ECOG, comorbidity. Preoperative functional dependence predicts postoperative mortality; postoperative delirium occurs in up to 50%. ERAS and preoperative frailty assessment help.
• Radiotherapy: hypofractionation, brachytherapy, and stereotactic radiosurgery reduce treatment burden and transport needs. Plan dose/duration by organ function and geriatric evaluation.
• Chemotherapy: older patients are underrepresented in trials. Watch renally cleared/toxic agents (methotrexate, cisplatin, ifosfamide, bleomycin); low albumin plus renal impairment predisposes to ifosfamide neurotoxicity; avoid concurrent NSAIDs. Neurotoxicity (taxanes, platinums, vinca, high-dose cytarabine, bortezomib), anthracycline cardiotoxicity, and fluorouracil mucositis are more common/severe; neutropenia is longer.
• Targeted agents: efficacy generally similar to younger patients (e.g., CDK4/6 inhibitors in advanced breast cancer) but with more toxicity and more dose reductions (without loss of benefit); watch CYP450 drug interactions and oral-adherence needs (cognition, support).
• Immune checkpoint inhibitors: similar efficacy (OS and PFS benefit in both age groups) and generally lower toxicity than chemotherapy; irAEs are not clearly more frequent by age but may be harder to manage with comorbidity, and the survival benefit associated with irAEs in younger patients may not extend to older adults.
• Polypharmacy: review with STOPP/START; polypharmacy raises interaction and toxicity risk.

Geriatric assessment (GA / CGA)

• Rationale: ECOG/Karnofsky performance status alone does not capture aging-related domains; GA detects impairments even in patients with good performance status.
• Domains: functional status (ADLs/IADLs, physical performance, falls), comorbidity, cognition (Mini-Cog, MoCA), psychological status (depression), nutrition, polypharmacy, social activity/support, and geriatric syndromes. A full GA takes <30 min (mostly patient-completed).
• ASCO 2023 guideline: perform multidomain GA in all patients aged ≥65 receiving systemic therapy, with GA-guided management for identified impairments (an abnormal G8/VES-13 is not a prerequisite; a practical abbreviated GA is acceptable).

Predicting chemotherapy toxicity: CARG and CRASH

• ASCO recommends using CARG or CRASH to estimate chemotherapy toxicity risk in all adults ≥65 receiving chemotherapy. CARG was developed/validated in solid tumors (no heme); CRASH included lymphoma; neither is validated across all hematologic malignancies and do not predict grade 2 toxicity.

GA also predicts mortality and functional decline

• Mortality: baseline GA measures predict OS and early death across solid and hematologic cancers. Simplified GA (age, comorbidity, ADL/IADL) stratifies fit vs intermediate vs frail (Palumbo, myeloma: 3-yr OS 84% vs 76% vs 57%), and underlies the Elderly Prognostic Index in DLBCL.
• Functional decline: depression and IADL impairment predict early decline during chemotherapy; VES-13 identifies women with breast cancer at risk.

GA-guided management improves outcomes

• GAP70+ (Mohile Lancet 2021; N=718): GA summary plus recommendations reduced grade 3 to 5 toxicity 51% vs 71% (RR 0.74) without compromising OS, largely via cycle-1 dose reductions.
• GAIN (Li JAMA Oncol 2021; N=613): GA-guided management reduced grade 3 to 5 toxicity ~10% (50.5% vs 60.6%) and increased advance-directive completion.
• GERICO (CRC, G8-prescreened): GA-guided interventions raised chemotherapy completion and improved QoL/mobility.
• INTEGERATE: integrated geriatrician-oncologist care improved QoL, cut hospital use ~39%, and reduced early treatment discontinuation (32.9% vs 53.2%).
• ESOGIA (advanced NSCLC): GA-based allocation did not improve treatment-failure-free survival or OS but reduced toxicity and toxicity-related failures.
• Across cohorts, GA changes treatment decisions ~20 to 47% of the time, usually toward less intensive therapy.

Concept of frailty

Specific older-adult evidence

• PRIME II: women ≥65 with HR-positive N0 tumors ≤3 cm, RT can be omitted on adjuvant endocrine therapy.
• Single-fraction RT for bone metastases: 8 Gy x1 well tolerated.
• VIALE-A: venetoclax plus azacitidine in older/unfit AML improved OS (mOS 14.7 vs 9.6 mo).
• DLBCL: R-mini-CHOP for patients ≥80 (prospective phase II, Peyrade 2011; also tested in SENIOR); full R-CHOP if fit.
• Older NSCLC: single-agent pembrolizumab for PD-L1 ≥50%; chemo-IO if able.
• Older mCRPC: choose by prior therapy, fitness, biomarkers (not age alone); ARPI if ARPI-naive, but after ARPI progression favor a class switch (taxane/other); fit older adults tolerate taxanes.

Goals of care & palliative integration

• Discuss goals (longevity vs QoL); older adults highly value functional independence and cognition (in one study, 74% and 89% would decline treatment if the outcome were severe functional or cognitive impairment).
• Palliative care throughout the disease course, not just terminally; use symptom tools (Edmonton Symptom Assessment Scale).
• Caregiver support (the older patient often has an older caregiver).

2024-2026 geriatric oncology updates

• ASCO 2023 GA guideline update (Dale JCO 2023): GA plus GA-guided management is a Grade A recommendation for all patients ≥65 on systemic therapy; validated by GAP70+ (Mohile Lancet 2021) and GAIN (Li JAMA Oncol 2021).
• GAP70+ mature analyses (Mohile 2024): GA-driven interventions reduced grade 3+ toxicity (51% vs 71%; RR 0.74) without compromising OS; trial population was older adults with at least one impaired GA domain.
• BiTE/CAR-T in older adults: real-world MM data (Sidana Blood Adv 2024), selected older adults (including ≥75) can benefit from CAR-T (cilta-cel, ide-cel) and bispecifics (tec/elra), but evidence is agent-specific and limited in ≥75; assess frailty, infection risk, cytopenias, and nonrelapse mortality individually (do not assume uniformly increased CRS/ICANS mortality); recommend careful selection and extended monitoring (product doses and step-up schedules are fixed, not frailty-adjusted).
• Cardio-oncology geriatric assessment: baseline echo/NT-proBNP before HER2-directed, anthracycline, or IO combinations; SGLT2i and ARNi considered for cancer-therapy-related cardiac dysfunction (2022 ESC cardio-onc guideline).
• Deprescribing in older cancer patients: STOPP/START v3 (2023) for older adults; OncPal deprescribing guideline for advanced cancer/limited life expectancy.

High-yield older adults pearls

• GA (not gated by a G8 or VES-13 screen) for all ≥65 starting systemic therapy, not only intensive chemotherapy; it is prognostic and Grade A per ASCO.
• CARG and CRASH predict chemotherapy toxicity; GA-guided management (GAP70+) cuts grade 3 to 5 toxicity without hurting survival.
• Estimate non-cancer life expectancy (Lee/Schonberg, ePrognosis) before screening or intensive treatment.
• Frailty: Fried phenotype (5 criteria) vs Rockwood cumulative-deficit index.
• PRIME II: omit RT in elderly low-risk breast on endocrine therapy. Ven + aza for older AML. R-mini-CHOP for frail DLBCL.
• Dose by creatinine clearance, not creatinine; review polypharmacy (STOPP/START); watch falls with neuropathy.
• Hippocampal-sparing WBRT plus memantine for cognitive preservation.