Artificial intelligence analysis of temporalis muscle thickness for monitoring sarcopenia and clinical outcomes in individuals with paediatric brain tumours: a retrospective cohort study.
Authors
Affiliations (13)
Affiliations (13)
- Mass General Brigham Artificial Intelligence in Medicine Program, Harvard Medical School, Boston, MA, USA; Department of Radiation Oncology, Dana-Farber Cancer Institute and Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA; Radiology and Nuclear Medicine, CARIM & GROW, Maastricht University, Maastricht, Netherlands.
- Mass General Brigham Artificial Intelligence in Medicine Program, Harvard Medical School, Boston, MA, USA; Department of Radiation Oncology, Dana-Farber Cancer Institute and Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA.
- Boston Children's Hospital, Boston, MA, USA.
- Department of Radiation Oncology, Dana-Farber Cancer Institute and Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA; Boston Children's Hospital, Boston, MA, USA.
- Division of Sleep and Circadian Disorders, Departments of Medicine and Neurology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA.
- Department of Data Science, Dana-Farber Cancer Institute, Boston, MA, USA.
- Mass General Brigham Artificial Intelligence in Medicine Program, Harvard Medical School, Boston, MA, USA; Boston Children's Hospital, Boston, MA, USA.
- Department of Neurology, Neurosurgery and Pediatrics, University of California, San Francisco, CA, USA.
- Department of Radiology, University of Pennsylvania, Philadelphia, PA, USA.
- Medical Oncology, Dana-Farber Cancer Institute and Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA.
- Pathology Department, Dana-Farber Cancer Institute and Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA.
- Mass General Brigham Artificial Intelligence in Medicine Program, Harvard Medical School, Boston, MA, USA; Department of Radiation Oncology, Dana-Farber Cancer Institute and Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA; Boston Children's Hospital, Boston, MA, USA.
- Mass General Brigham Artificial Intelligence in Medicine Program, Harvard Medical School, Boston, MA, USA; Department of Radiation Oncology, Dana-Farber Cancer Institute and Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA. Electronic address: [email protected].
Abstract
People with and who have survived paediatric brain tumour (PBT) have a poor quality of life due to physiological frailty, a primary component of which is sarcopenia (ie, low lean muscle mass) and the associated condition, sarcopenic overweight. MRI-based temporalis muscle thickness (TMT) is an image-derived biomarker for lean muscle, frailty, and survival in adult cancers. Here, we evaluated artificial intelligence-based TMT measurements (iTMT) to track sarcopenia in people with PBT at scale and identify trends, risk factors, and associations with morbidity and mortality. We conducted the secondary analyses of three cohorts (one prospective trial and two retrospective databases). iTMT was applied to all MRIs from diagnosis until the last follow-up or tumour recurrence to generate longitudinal, patient-level iTMT percentile curves. We investigated the rates of iTMT-defined sarcopenia (iTMT <15th percentile) and sarcopenic overweight (iTMT <15th percentile and BMI or weight >85th percentile) and associations with physical functioning (Pediatric Quality of Life Inventory [PedsQoL] version 4.0), endocrine disorders, and survival via multivariable analyses, using generalised additive models for longitudinal data. From all three databases, there were 5661 MRIs from 881 individuals. Of 730 patients with linked iTMT and weight (mean age 13·4 [SD 5·8]), 531 (73%) developed iTMT sarcopenia and 215 (29%) developed sarcopenic overweight at least once (median time: 5·1 years [IQR 1·4-10·6]). Of those with sarcopenia, 294 (55·4%) had normal weight and 153 (28·8%) were overweight or obese at time of sarcopenia. Radiotherapy was associated with iTMT sarcopenic overweight (p<0·0001), with the highest risk in craniospinal radiotherapy (p=0·013). iTMT sarcopenia was associated with poor physical functioning (p=0·058) and endocrine disorder diagnosis (p=0·035) in survivorship. In high-grade glioma, iTMT sarcopenia at diagnosis was associated with worse survival (log-rank p=0·046). iTMT sarcopenia is an image-derived biomarker for morbidity and mortality in people with and who have survived PBT that is common and cannot be reliably predicted by BMI. Incorporating iTMT into practice would enable the routine monitoring of sarcopenia and help triage individuals for supportive interventions to mitigate sarcopenia and associated morbidity. National Institutes of Health/National Cancer Institute, and Botha-Chan Low-Grade Glioma Consortium, St Baldrick's Research Foundation.