ESTRO 2024 - Abstract Book

S2503

Clinical - Urology

ESTRO 2024

1 National Cancer Center Hospital East, Section of radiation safety and quality assurance, Kashiwa, Japan. 2 University of Tsukuba, Graduate School of Comprehensive Human Sciences, Tsukuba, Japan. 3 Keio University School of Medicine, Department of Radiology, Shinjuku-ku, Japan

Purpose/Objective:

The aim of this study was to develop a normal tissue complication probability (NTCP) model parameters based on variable relative biological effectiveness (RBE)-weighted doses in proton therapy for prostate cancer.

Material/Methods:

We conducted a retrospective review of 150 prostate cancer patients who received passive-scattering proton therapy with a standard RBE of 1.1. For each patient, we recalculated the variable RBE-weighted dose distributions and investigated late rectal complications following proton therapy. Our analysis focused on two grades of rectal bleeding occurring within a minimum of 36 months, grade 1 and grade 2, according to the common toxicity criteria for adverse events version 4.0. We determined Lyman – Kutcher – Burman (LKB) model parameters (n, m, and TD50) using maximum likelihood estimation to assess the relationship between complications and doses with variable and standard RBE. We evaluated parameter errors by calculating the 95% confidence interval (CI) using the profile likelihood method and assessed goodness of fit with Pearson’s chi -square test. Among the 150 patients, 21 occurred grade ≥1, and 4 had grade ≥2 late -stage rectal bleeding. The resulting NTCP model parameters based on variable RBE-weighted dose were as follows: for grade 1, n = 0.31 (95% CI: 0.258 – 0.376), m = 0.24 (0.195 – 0.309), TD50 = 58.9 GyRBE (54.0 – 65.0 GyRBE); for grade 2, n = 0.026 (0.016 – 0.041), m = 0.084 (0.066 – 0.11), TD50 = 93.9 GyRBE (90.2 – 98.7 GyRBE). For a standard RBE, the resulting NTCP model parameters were as follows: for grade 1, n = 0.0062 (95% CI: 0.0062 – 0.010), m = 0.07 (0.056 – 0.089), TD50 = 77.9 GyRBE (76.4 – 79.6 GyRBE); for grade 2, n = 0.0062 (0.0062 – 0.013), m = 0.06 (0.0048 – 0.076), TD50 = 81.0 GyRBE (78.9 – 83.7 GyRBE). According to Pearson’s chi -square test, the p-value is >0.05, indicating good fitting accuracy. The volume effect parameter (n value) was larger for variable RBE compared to standard RBE, and the tolerance dose associated with a 50% probability of complication (TD50 value) was smaller for variable RBE compared to standard RBE. Results:

Conclusion:

We investigated the rectal NTCP model parameters based on variable RBE-weighted doses in prostate proton therapy. The rectal NTCP model parameters for variable RBE-weighted dose distributions were significantly different from the standard RBE dose distribution. The tolerance dose for rectal complication derived from the standard RBE may lead to erroneous decisions when using variable RBE dose distribution. Our NTCP model parameters reveal the dose-response relationship for late rectal complications with variable-RBE proton therapy and may enable more accurate predictions.

Keywords: proton therapy, NTCP modeling, prostate cancer