TY - JOUR
T1 - Effects From Nonuniform Dose Distribution in the Spinal Nerves of Pigs
T2 - Analysis of Normal Tissue Complication Probability Models
AU - Hrycushko, Brian
AU - Medin, Paul M.
N1 - Funding Information:
This project was funded by the Cancer Prevention & Research Institute of Texas (RP150356). Disclosures: P.M. and B.H. were investigators and received salary support for this project funded by a research grant paid to their institution (University of Texas Southwestern Medical Center) from the Cancer Prevention & Research Institute of Texas. P.M. has taught radiosurgery courses sponsored by BrainLAB Inc. Data-sharing statement: Research data are not available at this time.
Funding Information:
This project was funded by the Cancer Prevention & Research Institute of Texas ( RP150356 ).
Publisher Copyright:
© 2020 Elsevier Inc.
PY - 2021/4/1
Y1 - 2021/4/1
N2 - Purpose: Our purpose was to evaluate normal tissue complication probability (NTCP) models for their ability to describe the increase in tolerance as the length of irradiated spinal nerve is reduced in a pig. Methods and Materials: Common phenomenological and semimechanistic NTCP models were fit using the maximum likelihood estimate method to dose-response data from spinal nerve irradiation studies in pigs. Statistical analysis was used to compare how well each model fit the data. Model parameters were then applied to a previously published dose distribution used for spinal cord irradiation in rats under the assumption of a similar dose-response. Results: The Lyman-Kutcher-Burman model, relative seriality, and critical volume model fit the spinal nerve data equally well, but the mean dose logistic and relative seriality models gave the best fit after penalizing for the number of model parameters. The minimum dose logistic regression model was the only model showing a lack of fit. When extrapolated to a 0.5-cm simulated square-wave–like dose distribution, the serial behaving models showed negligible increase in dose-response curve. The Lyman-Kutcher-Burman model and relative seriality models showed significant shifting of NTCP curves due to parallel behaving parameters. The critical volume model gave the closest match to the rat data. Conclusions: Several phenomenological and semimechanistic models were observed to adequately describe the increase in the radiation tolerance of the spinal nerves when changing the irradiated length from 1.5 to 0.5 cm. Contrary to common perception, model parameters suggest parallel behaving tissue architecture. Under the assumption that the spinal nerve response to radiation is similar to that of the spinal cord, only the critical volume model was robust when extrapolating to outcome data from a 0.5-cm square-wave–like dose distribution, as was delivered in rodent spinal cord irradiation research.
AB - Purpose: Our purpose was to evaluate normal tissue complication probability (NTCP) models for their ability to describe the increase in tolerance as the length of irradiated spinal nerve is reduced in a pig. Methods and Materials: Common phenomenological and semimechanistic NTCP models were fit using the maximum likelihood estimate method to dose-response data from spinal nerve irradiation studies in pigs. Statistical analysis was used to compare how well each model fit the data. Model parameters were then applied to a previously published dose distribution used for spinal cord irradiation in rats under the assumption of a similar dose-response. Results: The Lyman-Kutcher-Burman model, relative seriality, and critical volume model fit the spinal nerve data equally well, but the mean dose logistic and relative seriality models gave the best fit after penalizing for the number of model parameters. The minimum dose logistic regression model was the only model showing a lack of fit. When extrapolated to a 0.5-cm simulated square-wave–like dose distribution, the serial behaving models showed negligible increase in dose-response curve. The Lyman-Kutcher-Burman model and relative seriality models showed significant shifting of NTCP curves due to parallel behaving parameters. The critical volume model gave the closest match to the rat data. Conclusions: Several phenomenological and semimechanistic models were observed to adequately describe the increase in the radiation tolerance of the spinal nerves when changing the irradiated length from 1.5 to 0.5 cm. Contrary to common perception, model parameters suggest parallel behaving tissue architecture. Under the assumption that the spinal nerve response to radiation is similar to that of the spinal cord, only the critical volume model was robust when extrapolating to outcome data from a 0.5-cm square-wave–like dose distribution, as was delivered in rodent spinal cord irradiation research.
UR - http://www.scopus.com/inward/record.url?scp=85097221275&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85097221275&partnerID=8YFLogxK
U2 - 10.1016/j.ijrobp.2020.11.002
DO - 10.1016/j.ijrobp.2020.11.002
M3 - Article
C2 - 33171201
AN - SCOPUS:85097221275
SN - 0360-3016
VL - 109
SP - 1570
EP - 1579
JO - International Journal of Radiation Oncology Biology Physics
JF - International Journal of Radiation Oncology Biology Physics
IS - 5
ER -