PIPO

PIPO predicts overall survival (OS) outcomes for patients considered for inclusion in phase 1 oncology clinical trials treated with immune checkpoint inhibitors (ICIs) or targeted agents (TAs), to provide individualized prognostic estimates for patient selection.

Key Features:

• Prognostic model development: The model was trained on a cohort of 558 patients and validated on an independent cohort of 241 patients using baseline clinical parameters to predict median overall survival (mOS) and the 3-month OS rate.
• Predictive variables: Multivariable analysis identified six independent prognostic variables: Eastern Cooperative Oncology Group performance status (ECOG), number of metastatic sites, presence of liver metastases, derived neutrophils/(leukocytes minus neutrophils) ratio [dNLR], albumin levels, and lactate dehydrogenase (LDH) levels.
• Performance metrics: Model discrimination in the validation set yielded a C-statistic of 0.71 for overall survival and an accuracy of 87.2% for predicting 3-month OS outcomes.
• Treatment context: Predictions are specified for patients receiving immune checkpoint inhibitors (ICIs) or targeted agents (TAs) in phase 1 trials.

Scientific Applications:

• Patient selection for phase 1 trials: Provides objective survival predictions to inform enrollment decisions for early-phase oncology trials using ICIs or TAs.
• Trial stratification and prognostic assessment: Enables stratification and individualized prognostic assessment to support interpretation of outcomes in early-phase clinical studies and personalized medicine approaches.

Methodology:

Model development used baseline clinical parameters with multivariable analysis to identify six prognostic variables; the model predicts mOS and 3-month OS and was evaluated by C-statistic and accuracy on a validation cohort of 241 patients following training on 558 patients.