Encrypted login | home

Program Information

Tumor Control Probability in Single-Fraction Stereotactic Brain Radiation Therapy

no image available
M Sharma

M Sharma*, M Cummings , M Schell , M Pacella , D Rosenzweig , M Milano , The University of Rochester Cancer Cen, Farmington, NY


SU-I-GPD-T-520 (Sunday, July 30, 2017) 3:00 PM - 6:00 PM Room: Exhibit Hall

Purpose: The objective of this study was to compute the tumor control probability (TCP) and determine its relation to various biological factors like histology, tumor volume and death or last follow-up.

Methods: A total of 111 patients with 261 metastatic lesions from breast (23), non-small cell lung (NSCLC) (72), melanoma (98) and colorectal (5) primaries were treated with single-fraction LINAC-based stereotactic radiosurgery (SRS) from 2008-2015. The mean treated lesion volume was 2.17 (std., ± 4.4) cc and mean physical dose delivered was 17.5 (std, ± 2) Gy. The median follow-up was 13 months and median follow-up time with MRI imaging was 9.6 months-per-course. The TCP values were computed using the Niemerko's model for inhomogenously irradiated volume, TCP=[1+(TCD₅₀/gEUD)⁽⁴γ₅₀⁾]⁽⁻¹⁾, where TCD₅₀ is the dose that gives response probability of 50% when tumor is heterogeneously irradiated, γ₅₀ is a unit less model parameter specific target and gEUD is the generalized equivalent uniform dose. For the purpose of the study different parameters used were taken from available literature (α/β=8, γ₅₀=2, a=-5 and TCD₅₀=21)

Results: The tumor control probability showed negligible dependence on the histology. The values were 0.96±0.07 for breast, 0.97±0.10 for NSCLC, 0.98±0.06 for melanoma and 0.99±0.01 for colorectal primaries. No significant correlation was found between lesion volume and TCP (Pearson coefficient= 0.012). The mean TCP values showed difference based on death (0.96±0.1) and last follow-up/survival (0.99±0.01).

Conclusion: With the one set of parameters used in the study, similar TCP values were found across all the histologies. When compared to the date of death or last follow-up, lower TCP values were found in the former case. Further work needs to be done with different sets of model parameters to investigate TCP values.

Contact Email: