Related Experiment Video
Updated: Aug 14, 2026
Pretargeted Radioimmunotherapy Based on the Inverse Electron Demand Diels-Alder Reaction
Published on: January 29, 2019
Analysis of the Radiosensitivity Index in Paired Preneoadjuvant and Postneoadjuvant Therapy Triple-Negative Breast
Shane R Stecklein1, Roberto Salgado2, Julia R White3
1Department of Radiation Oncology, University of Kansas Medical Center, Kansas City, Kansas; Department of Pathology and Laboratory Medicine, University of Kansas Medical Center, Kansas City, Kansas; Department of Cancer Biology, University of Kansas Medical Center, Kansas City, Kansas; The University of Kansas Cancer Center, Kansas City, Kansas; Department of Breast Radiation Oncology, University of Texas MD Anderson Cancer Center, Houston, Texas.
Purpose:
The radiosensitivity index (RSI) is a validated gene expression-based biomarker that can predict intrinsic radiosensitivity and has been shown to be associated with local control in patients with triple-negative breast cancer (TNBC) treated with upfront surgery. Currently, most patients with TNBC receive neoadjuvant systemic therapy (NAST). Whether the RSI predicts response to NAST and how the RSI and predicted radiosensitivity are altered by exposure to NAST is unknown.
Methods And Materials:
Total RNA was extracted from pretreatment core needle biopsy specimens from 197 patients with TNBC treated on the NeoSTOP (NCT02413320) and NeoPACT (NCT03639948) trials. Total RNA was also extracted from paired post-NAST (residual disease) tumor tissue in 58 patients. A published algorithm using 10 genes was used to compute the RSI for each tumor. Stromal tumor-infiltrating lymphocytes (sTILs) were scored on pretreatment and post-NAST samples by an expert breast pathologist according to international consensus guidelines. CIBERSORTx was used to impute leukocyte fractions in samples using RNA-sequencing data.
Results:
Cluster analysis of RSI genes in pretreatment samples revealed immune-depleted and immune-enriched groups, and this classification was strongly associated with sTIL infiltration (P < .0001) and likelihood of achieving pathologic complete response (pCR) (P = .001). RSI showed associations (false discovery rate q < 0.01) with M0 and M1 macrophages, CD4+ memory resting, CD4+ memory activated, CD8+, and follicular helper T-cells, activated natural killer cells, naïve and memory B cells, and resting dendritic cells on CIBERSORTx leukocyte deconvolution. In the entire cohort, NAST-induced change in RSI was variable, but among initially RSI-immune-enriched tumors that did not achieve pCR, there was a significant decrease in predicted radiosensitivity between paired pretreatment and post-NAST samples. NAST-induced reduction in sTILs and naïve B cells may be associated with this decrease in radiosensitivity.
Conclusions:
Pretreatment RSI cluster identity is associated with the degree of immune enrichment and response to NAST in TNBC. Initially immune-enriched TNBCs that do not achieve a pCR to NAST exhibit a decrease in predicted radiosensitivity compared with paired pretreatment tumors.
