Related Experiment Video
Updated: May 3, 2026
An Orthotopic Resectional Mouse Model of Pancreatic Cancer
Published on: September 24, 2020
Normal tissue complication probability model for severe radiation-induced lymphopenia in patients with pancreatic
Fuki Koizumi1, Norio Katoh1, Takahiro Kanehira2
1Department of Radiation Oncology, Hokkaido University Faculty of Medicine and Graduate School of Medicine, North 15 West 7, Kita-ku, Sapporo, Hokkaido 060-8638, Japan.
Background And Purpose:
Radiation-induced lymphopenia (RIL) may be associated with a worse prognosis in pancreatic cancer. This study aimed to develop a normal tissue complication probability (NTCP) model to predict severe RIL in patients with pancreatic cancer undergoing concurrent chemoradiotherapy (CCRT).
Materials And Methods:
We reviewed pancreatic cancer patients treated at our facility for model training and internal validation. Subsequently, we reviewed data from three other facilities to validate model fit externally. An absolute lymphocyte count (ALC) of ＜0.5 × 103/μL during CCRT was defined as severe RIL. An NTCP model was trained using a least absolute shrinkage and selection operator (LASSO)-based logistic model. The model's predictive performance was evaluated using the receiver operating characteristic area under the curve (AUC), scaled Brier score, and calibration plots.
Results:
Among the 114 patients in the training set, 78 had severe RIL. LASSO showed that low baseline ALC, large planning target volume, and high percentage of bilateral kidneys receiving ≥ 5Gy were selected as parameters of the NTCP model for severe RIL. The AUC and scaled Brier score were 0.91 and 0.49, respectively. Internal validation yielded an average AUC of 0.92. In the external validation with 68 patients, the AUC and scaled Brier score was 0.83 and 0.30, respectively. Calibration plots showed good conformity.
Conclusions:
The NTCP model for severe RIL during CCRT for pancreatic cancer, developed and validated in this study, demonstrated good predictive performance. This model can be used to evaluate and compare the risk of RIL.
