We developed predictive models for lung adenocarcinoma and colorectal carcinoma that integrate tumor genetics, disease stage, functional status, and baseline weight to estimate individualized cachexia risk over one to three years, potentially enabling earlier interventions.

A consistent blood signature accompanies cachexia across cancer types, featuring lower albumin, anemia indicators (low hemoglobin and red blood cell count), elevated liver enzymes, and higher white blood cell and neutrophil counts, while albumin declines may stay within normal ranges.

The research was enabled by MSK’s Cancer Data Science Initiative and funded by philanthropic sources, foundations, and Cancer Grand Challenges, highlighting the value of large-scale, anonymized clinical data for advancing cachexia understanding.

Cachexia occurs across nearly all cancer types in episodes rather than a single gradual decline, with episodes often aligning with disease progression but sometimes arising independently, pointing to multiple contributing factors.

The study indicates cachexia subtypes and advocates a precision medicine approach to prevention and treatment, aiming to tailor interventions to individual risk profiles.

Specific tumor genetic mutations relate to cachexia risk, including TP53 mutations linked to higher risk across cancers, with STK11 in lung cancer, SMAD4 and KRAS in colorectal cancer; some mutations like ARID1A and KMT2B may be protective.

MSK clinicians and collaborators analyzed routine data from nearly 60,000 cancer patients, including over 2 million weight measurements and millions of blood tests, producing the largest human study of cancer cachexia to date.