A prospective study examining cachexia predictors in patients with incurable cancer

Area of Science:

• Oncology
• Palliative Care
• Clinical Prediction Models

Background:

• Cachexia significantly impacts patient outcomes in incurable cancer.
• Early intervention requires accurate prediction of cachexia development.
• This study aimed to identify predictors and develop a predictive model for cancer cachexia.

Purpose of the Study:

• To identify significant predictors of cachexia development in cancer patients.
• To construct and validate a predictive model for cancer cachexia.
• To improve early intervention strategies through accurate risk stratification.

Main Methods:

• Secondary analysis of a prospective, observational, multicentre study.
• Inclusion criteria: palliative care patients with incurable cancer, no baseline cachexia.
• Cachexia definition: weight loss >5% (6 months) or weight loss >2% with BMI <20 kg/m².
• Predictors evaluated using Cox analysis; model developed with classification and regression tree analysis.

Main Results:

• 159 (25%) of 628 patients developed cachexia during follow-up.
• Significant predictors included initial weight loss, cancer type, appetite, and COPD.
• A five-level predictive model demonstrated 76% accuracy in predicting cachexia at 3 months.
• Risk-level 5 patients (e.g., 3-5% weight loss) developed cachexia in a median of 51 days.

Conclusions:

• Key predictors for cancer cachexia have been identified.
• A novel, validated predictive model for cancer cachexia has been developed.
• This model can aid in timely intervention for patients at high risk of cachexia.