Unencrypted login | home

Program Information

Prediction in Clinical Response to Radiation Therapy for Head and Neck Cancer


Z Huang

Z Huang1*, N Mayr2, S Lo3, S Winkler1, K Yuh2, C Lok1, T Liu4, S Stephenson1,R McLawhorn1, K Rasmussen1, J Rice1, W Yuh2, (1) East Carolina University, Greenville, NC, (2) Ohio State University, Columbus, OH, (3) Case Comprehensive Cancer Center, Cleveland, OH, (4) Baylor College of Medicine, Houston, TX

TU-G-108-1 Tuesday 4:30PM - 6:00PM Room: 108

Purpose:
A tumor control probability (TCP) model incorporating a generalized linear-quadratic (gLQ) model was investigated to predict clinical outcome after radiation therapy for patients. The clinical utility of the model incorporating the following individually measured radiobiology parameters: intrinsic radiosensitivity, proliferation and number of clonogenic cells, was evaluated. The hypothesis in the study was that the incorporation of individually measured tumor parameters into the TCP model would increase its reliability in predicting treatment outcome compared with the use of average population derived data.

Methods:
46 patients with head and neck tumors were included and most of them received both external beam radiotherapy and brachytherapy (as published in Acta Oncologica, 2009; 48:584-590). Primary tumor size was drawn from case records and pre-treatment CT scans or MRI using two or three dimensional measurements to calculate tumor volume. Surviving fraction after 2 Gy (SF2), reflecting intrinsic radiosensitivity, was estimated by a soft-agar clonogenic assay (as published in Int J Radiat Oncol Bio Phys, 2000; 46:13-19). Eighteen patients receiving external beam treatment alone were used to perform statistical analyses. Local control was determined by follow-up records.

Results:
SF2 correlated significantly with tumor response (p=0.04, Mann-Whitney test), but initial tumor volume did not (p=0.6). Eight of the 18 patients had a >95% calculated tumor control probability and none developed a local recurrence, yielding a negative predictive value of 100%, compared with 67% for population-derived data. There was a statistically significant difference in local control levels bewteen the patient group with >95% vs. <5% predicted probability of local control (p<0.001).

Conclusion:
The results suggest that incorporation of measured biological data within a TCP radiobiological model would improve its ability to predict radiation therapy outcome.

Contact Email: