Prediction Model for Brain Metastasis in Patients With Metastatic Germ-Cell Tumors.

Area of Science:

• Oncology
• Medical Statistics
• Clinical Prediction Modeling

Background:

• Brain metastasis (BM) is a significant negative prognostic factor in metastatic germ cell tumors (mGCT).
• Effective and practical prediction models for BM in mGCT are needed to improve patient outcomes.

Purpose of the Study:

• To establish and validate an effective and practical prediction model for brain metastasis (BM) in patients with metastatic germ cell tumors (mGCT).
• To identify clinical and pathological factors associated with BM occurrence in mGCT.

Main Methods:

• Retrospective analysis of 2291 mGCT patients treated between 1990 and 2017.
• Patients categorized into BM present (154) and BM absent (2137) groups.
• Logistic regression and stepwise selection used to develop a predictive model, validated on a separate dataset.

Main Results:

• A prediction model was developed incorporating age ≥40, choriocarcinoma histology, hCG ≥5000, pulmonary metastasis size (<3 or ≥3 cm), and bone metastasis.
• The model demonstrated increasing BM probability with higher scores, ranging from 0.6% to 90% for scores 0 to 8.
• Patients with BM had significantly worse 2-year progression-free survival (17% vs 65%) and overall survival (62% vs 91%) compared to those without BM.

Conclusions:

• The developed prediction model effectively discriminates the occurrence of brain metastasis in mGCT.
• This model can be utilized to identify mGCT patients at high risk for developing brain metastasis.