+ Key finding at a glance
Study Overview: Published in The Lancet Digital Health (July 2026), researchers analyzed 5,661 routine brain MRI scans from 881 pediatric brain tumor patients and survivors using an AI tool called iTMT (temporalis muscle thickness).
Prognostic Impact: In high-grade glioma, sarcopenia identified via iTMT pointed to poorer overall survival (16.6 months vs. 23.5 months in non-sarcopenic patients), outperforming traditional anthropometric measures.
High Prevalence: 72.7% of patients developed sarcopenia (low muscle mass) during follow-up, and 29.5% developed sarcopenic overweight.
The BMI Blind Spot: More than 50% of patients with sarcopenia had a normal weight, and nearly 29% were overweight or obese, proving that conventional BMI fails to capture muscle depletion.
Treatment Risks: Craniospinal radiotherapy significantly increased vulnerability, with sarcopenic overweight frequently emerging around five years post-treatment alongside higher rates of endocrine disorders.
A study published in The Lancet Digital Health found that artificial intelligence can identify and monitor sarcopenia in children and young adults with paediatric brain tumours using routine MRI scans. Researchers analysed 5,661 scans from 881 patients and survivors across three clinical cohorts.
The team used an AI tool called iTMT, which measures temporalis muscle thickness from standard brain MRI images. The metric acts as a surrogate marker of lean muscle mass and allows clinicians to track changes over time without additional tests or imaging.
Sarcopenia Often Goes Undetected
Low muscle mass is difficult to assess in paediatric oncology. Clinicians often rely on weight and body mass index (BMI), but these measures cannot distinguish muscle from fat.
The researchers found a weak correlation between BMI and iTMT measurements. As a result, many patients showed signs of sarcopenia despite having a normal or elevated body weight.
Among 730 patients with linked weight data, 531 developed sarcopenia at least once during follow-up. This represented 72.7% of the study population. Another 215 patients, or 29.5%, developed sarcopenic overweight, a condition that combines low muscle mass with excess weight.
More than half of patients identified as sarcopenic had a normal weight when the condition was detected. Nearly 29% were overweight or obese.
Radiotherapy Increased Risk
The study identified radiotherapy as a major risk factor.
Patients who received radiotherapy developed sarcopenia and sarcopenic overweight more often than those who did not. The highest risk appeared in patients treated with craniospinal radiotherapy.
Researchers also observed long-term changes after treatment. Sarcopenic overweight started to emerge around five years after radiotherapy and continued to increase during survivorship.
Children exposed to radiotherapy also showed greater reductions in height percentiles and higher rates of endocrine disorders.
Links with Physical Function and Endocrine Disorders
The investigators examined whether changes in muscle mass correlated with clinical outcomes.
In the UCSF RadArt cohort, lower iTMT values correlated with poorer physical functioning scores on the Paediatric Quality of Life Inventory (PedsQL).
The EMPOWER survivorship cohort produced similar findings. Patients with sarcopenia were more likely to receive a diagnosis of an endocrine disorder than patients without sarcopenia. Thyroid disease and growth hormone deficiencies accounted for much of that difference.
By comparison, excess weight alone showed no significant association with endocrine disorders.
Impact on Survival
The researchers also analysed outcomes in patients with high-grade glioma.
Patients with sarcopenia at diagnosis had poorer overall survival than those with higher muscle mass. Median survival reached 16.6 months in the sarcopenic group compared with 23.5 months in the non-sarcopenic group.
Low BMI and low body weight did not show the same association with survival. According to the authors, iTMT offered stronger prognostic value than conventional anthropometric measurements.
Source
Zapaishchykova A. et al., Artificial intelligence analysis of temporalis muscle thickness for monitoring sarcopenia and clinical outcomes in individuals with paediatric brain tumours: a retrospective cohort study, The Lancet Digital Health, July 2026. Accessed online here
