Radiation and Cancer Biology

SS 39 - Biology 7 - Special Session: Innovative Biologic Approaches to Improve Risk Stratification and Treatment Outcomes

281 - Utilizing the Genomically Adjusted Radiation Dose (GARD) to Model Radiation Dose Personalization

Wednesday, October 24
1:30 PM - 1:40 PM
Location: Room 206

Utilizing the Genomically Adjusted Radiation Dose (GARD) to Model Radiation Dose Personalization
K. A. Ahmed1, Y. Kim2, S. M. H. Naqvi3, A. E. Berglund4, E. Welsh4, A. O. Naghavi1, J. J. Caudell5, L. B. Harrison1, S. A. Eschrich4, J. G. Scott6, and J. F. Torres-Roca5; 1H. Lee Moffitt Cancer Center and Research Institute, Tampa, FL, 2Department of Biostatistics and Bioinformatics, Moffitt Cancer Center, Tampa, FL, 3Moffitt Cancer Center, Tampa, FL, 4H. Lee Moffitt Cancer Center and Research Institute, Department of Biostatistics and Bioinformatics, Tampa, FL, 5H. Lee Moffitt Cancer Center and Research Institute, Department of Radiation Oncology, Tampa, FL, 6Department of Radiation Oncology, Taussig Cancer Institute, Cleveland Clinic, Cleveland, OH

Purpose/Objective(s): We have previously developed the genomically adjusted radiation dose (GARD) as a clinically actionable metric based on the radiosensitivity index (RSI), an extensively validated signature of tumor radiosensitivity. The purpose of this study was to model the effect of radiation dose on GARD to predict tumor control probability across tumor types.

Materials/Methods: A total of 8,271 tumor samples were identified from our institutional IRB-approved prospective tissue collection protocol. Gene expression profiling was performed using the Affymetrix Hu-RSTA-2a520709 platform. The previously tested RSI 10-gene assay was run on tissue samples ranked according to gene expression. GARD was calculated for doses of 45 Gy (subclinical), 60 Gy (microscopic), or ≥ 70 Gy (macroscopic). The 75th percentile of GARD (GARD high) was used to model GARD optimization based on previous validation studies. A generalized linear mixed effect model (GLMM) assessing the effect of radiation dose on the percentage of tumors above GARD high across 20 primary tumor histologies was created.

Results: For all disease sites, the percent of tumor samples achieving GARD high as a function of radiation dose approximated a sigmoidal distribution, a relationship consistent with historical tumor control probability models. Significant differences were noted in the slope of the curves, log odds ratios of achieving GARD high, across tumor sites in the GLMM (F-test p<0.0001). The largest effect of radiation dose escalation/de-escalation was noted in oropharyngeal head and neck (slope parameter = 0.26), thyroid papillary (0.21) and pancreatic adenocarcinoma (0.18). Radiation dose escalation/de-escalation was found to have the least impact on endometrial adenocarcinoma (0.07), sarcoma (0.07), and transitional bladder cell carcinoma (0.08).

Conclusion: In this study, we assess GARD optimization as a function of radiation dose for various tumor types. Significant differences were noted in the model across tumor types revealing the genomic heterogeneity that exists across disease sites and individual tumors. GARD can be used to model the effect of radiation dose escalation/de-escalation on clinical outcome.

Author Disclosure: K.A. Ahmed: None. Y. Kim: None. S. Naqvi: None. A.E. Berglund: None. A.O. Naghavi: None. S.A. Eschrich: Stock; Cvergenx. Executive; Cvergenx. J.F. Torres-Roca: Stock; Cvergenx. Executive; Cvergenx.

Kamran Ahmed, MD

Disclosure:
Employment
Moffitt Cancer Center: Resident: Employee

Presentation(s):

Send Email for Kamran Ahmed


Assets

281 - Utilizing the Genomically Adjusted Radiation Dose (GARD) to Model Radiation Dose Personalization



Attendees who have favorited this

Please enter your access key

The asset you are trying to access is locked. Please enter your access key to unlock.

Send Email for Utilizing the Genomically Adjusted Radiation Dose (GARD) to Model Radiation Dose Personalization