Oncoger at the XXX National Congress of Geriatrics and Gerontology: towards precision oncogeriatrics

At the XXX National Congress of Geriatrics and Gerontology, Dr. Gonzalo Navarrete —oncologist and geriatrician at FALP and the University of Chile Clinical Hospital, assistant professor at the University of Chile School of Medicine and co-founder of Oncoger— delivered the lecture “International advances towards precision oncogeriatrics”. The talk traced the path from the biology of ageing to artificial intelligence applied to the treatment decision, and was the occasion to present PROTEGER to the national geriatric community.

We gather here the central ideas of the presentation in a summarised, interactive version, followed by an account of its main elements.

The scale of the problem

Cancer is, increasingly, a disease of older people. In the United States, the estimated prevalence of cancer will rise from 15.5 million people in 2016 to 26.1 million in 2040, with growth concentrated in the groups aged 65 and over. By 2050, 80% of older people with cancer will live in low- and middle-income countries. The consequences of treatment in this group are significant: among those over 65, 53% experience moderate to severe toxicity from chemotherapy, and up to 48% lose their entire wealth because of the illness. In Chile, the problem also has an access dimension: delays in GES cancer guarantees increase with age and reach 194.8 per 100,000 inhabitants in people aged 60 and over, the highest rate of all age groups.

Cancer in the context of the ageing host

The biology of cancer has undergone a paradigm shift. The updated hallmarks framework no longer understands the tumour as a cell with mutations, but as a dynamic ecosystem that interacts with its microenvironment and with the whole organism; a single diagnosis may have different “ecotypes” and therefore a different response to treatment. In older people, that ecosystem is modulated by ageing: chronic inflammation, immunosenescence, cellular senescence, an altered microenvironment, metabolic and microbiome changes, and reduced physiological reserve. The clinical consequence is greater biological heterogeneity and one central fact: toxicity and benefit do not depend on the tumour alone. Ageing, moreover, is not only chronological: accelerated biological age raises cancer risk by around 9% per standard deviation, and treatments themselves accelerate the biological ageing of survivors.

The gap between clinical trials and the real geriatric patient

“We prescribe 21st-century treatments based on data that ignore the real geriatric patient,” Dr. Navarrete argued. Of 129 cancer drugs approved by the FDA between 2013 and 2023, 41% do not improve overall survival or quality of life at five years. In immunotherapy trials, only 42.3% of participants were 65 or older and just 11.1% were 75 or older; 98% of trials exclude patients with ECOG greater than 1 and none includes a baseline geriatric assessment. Real-world data show the distance: while the pivotal pembrolizumab trial in lung cancer reports a median survival of 26.3 months, older people with ECOG 2–3 treated in practice reach 1.6 months. In metastatic pancreatic cancer, the phase III standard offers 8.7 months; in vulnerable older patients, the observed survival was about 4.5 months, with no difference between regimens.

What the oncogeriatric assessment adds

The intervention evidence is solid. Three randomised trials show that comprehensive geriatric assessment with recommendations to the oncologist changes outcomes: GAIN reduced grade 3 or higher toxicity from 60.6% to 50.5%; GAP70+ reduced it from 71% to 51%; and INTEGERATE achieved 41% fewer unplanned hospitalisations, 39% fewer emergency visits and better quality of life. The assessment also generates economic value, with an estimated net incremental monetary benefit of CAD 2,231 per patient for GAP70+ and CAD 2,104 for INTEGERATE. In Chile, the expert consensus published in the Journal of Geriatric Oncology —co-authored by Dr. Navarrete— recommends applying CGA in people aged 65 and over with cancer who will receive systemic therapy, using G8 as screening when access to CGA is limited, and a set of scales by domain.

The guidelines, however, orient the intervention but do not complete the clinical decision: they do not classify the patient as robust, vulnerable or frail, do not state who should not receive treatment, and do not indicate in whom to start with a dose reduction. Two key messages from the talk: frailty is contextual —the same patient may be frail for FOLFIRINOX, robust for abiraterone and vulnerable for definitive chemoradiotherapy— and frailty does not mean “not treatable”. In vulnerable patients, a reduced starting dose with escalation according to tolerance is a reasonable standard, and the starting dose can be understood as a therapeutic trial: start, observe and adjust.

PROTEGER: artificial intelligence as decision support

That is the space PROTEGER seeks to fill: artificial-intelligence algorithms to assist oncology and geriatric teams in treatment decisions for older people with cancer. Published in the Journal of Geriatric Oncology and registered under TRIPOD-AI, the model was developed from the Chilean Oncogeriatric Telecommittee, integrating oncological and CGA variables, and was externally validated in two tertiary centres with an eligibility AUC of 0.84 against the expert recommendation. Its classification separates groups with clearly different survival and reproduces that of the oncogeriatric team. The national cohort now brings together 629 patients from six public centres from Antofagasta to Coyhaique, with a median age of 79 years, 55% at stage IV and 54% with cognitive impairment: a sample that reflects the complexity of real public oncology. Clinical validation continues in centres in Chile, Peru and Brazil.

Better assessment means better treatment

Chronological age is not enough. The oncogeriatric decision integrates tumour biology, functional reserve, expected benefit, risk and the patient's goals, and leads to three possible paths: treat, optimise or not treat. As Dr. Navarrete summed up in his guiding principle: “Technology supports; the final decision remains with the clinical team and, together, with the patient”.

References: Navarrete et al., J Geriatr Oncol 2026;17:103031 · Quilodrán Loyola et al., J Geriatr Oncol 2026;17:102847 · Li et al., JAMA Oncol 2021 (GAIN) · Mohile et al., Lancet 2021 (GAP70+) · Soo et al., Lancet Healthy Longev 2022 (INTEGERATE) · Hall et al., JAMA Oncol 2021 (GO2) · Ferrat et al., JCO 2017 · Hanahan, Cell 2026.