External Validation of PathFx 3.0 Model on a Single-Centre Portuguese Population.

Area of Science:

• Oncology
• Medical Statistics
• Surgical Oncology

Background:

• The PathFx 3.0 prognostic model predicts survival in patients with bone metastatic disease.
• External validation is crucial for assessing model applicability across diverse healthcare systems.
• Previous validation of PathFx 3.0 has been extensive, but its performance in specific national cohorts requires further investigation.

Purpose of the Study:

• To externally validate the PathFx 3.0 prognostic model in a Portuguese patient cohort undergoing surgery for skeletal metastases.
• To assess the model's predictive accuracy and calibration in a real-world clinical setting outside its original development population.

Main Methods:

• Retrospective analysis of 49 patients who had surgery for skeletal metastases between January 2018 and April 2023.
• Survival probabilities were calculated at 1, 3, 6, 12, and 24 months post-surgery.
• Performance was evaluated using Area Under the Receiver Operating Characteristic Curve (AUC-ROC), Brier scores, and calibration analysis.

Main Results:

• The PathFx 3.0 model demonstrated improved discrimination over time, with AUC values ranging from 0.495 at 1 month to 0.791 at 12 months.
• Brier scores indicated moderate predictive accuracy, with values ranging from 0.172 to 0.246.
• Calibration analysis revealed good agreement between predicted and observed survival, except in the immediate post-operative phase.

Conclusions:

• PathFx 3.0 shows good prognostic performance in the studied Portuguese population, supporting its use in clinical decision-making for patients with metastatic bone disease.
• The model's accuracy for early survival prediction (1 month) is limited, but it becomes valuable for longer-term prognostication.
• Further validation in larger, multicenter Portuguese cohorts is recommended to confirm these findings and enhance generalizability.