Radioresistance of KRAS/TP53-mutated lung cancer can be overcome by radiation dose escalation or EGFR tyrosine kinase inhibition in vivo

Abstract

Recent clinical data have linked KRAS/TP53 comutation (mut) to resistance to radiotherapy (RT), but supporting laboratory in vivo evidence is lacking. In addition, the ability of different radiation doses, with/without epidermal growth factor receptor (EGFR)-directed treatment, to achieve local tumor control as a function of KRAS status is unknown. Here, we assessed clonogenic radiation survival of a panel of annotated lung cancer cell lines. KRASmut/TP53mut was associated with the highest radioresistance in nonisogenic and isogenic comparisons. To validate these findings, isogenic TP53mut NCI-H1703 models, KRASmut or wild-type (wt), were grown as heterotopic xenografts in nude mice. A clinical RT schedule of 30 fractions over 6 weeks was employed. The dose that controlled 50% of tumors (TCD₅₀ ) was calculated. The TCD₅₀ for KRASwt/TP53mut xenografts was 43.1 Gy whereas KRASmut/TP53mut tumors required a 1.9-fold higher TCD₅₀ of 81.4 Gy. The EGFR inhibitor erlotinib radiosensitized KRASmut but not KRASwt cells and xenografts. The TCD₅₀ associated with adding erlotinib to RT was 58.8 Gy for KRASmut, that is, a ~1.4-fold dose enhancement. However, the EGFR antibody cetuximab did not have a radiosensitizing effect. In conclusion, we demonstrate for the first time that KRASmut in a TP53mut background confers radioresistance when studying a clinical RT schedule and local control rather than tumor growth delay. Despite the known unresponsiveness of KRASmut tumors to EGFR inhibitors, erlotinib radiosensitized KRASmut tumors. Our data highlight KRAS/TP53 comutation as a candidate biomarker of radioresistance that can be at least partially reversed by dose escalation or the addition of a targeted agent.

Keywords: KRAS mutation; cetuximab; erlotinib; radiotherapy.

Figure 1.

Radiation survival of NSCLC cell lines as a function of genotype. (A) Clonogenic survival of 17 NSCLC cell lines treated with single dose irradiation from 2 to 8 Gy. Each data point represents the average survival from typically 3 biological repeats for one cell line. Survival curves were fitted using the LQ formula (see Supplementary Methods) for KRAS-mutant (mut) vs. wild-type (wt) cell lines. Statistical comparison of the curves was performed using the F-test (see Supplementary Methods). (B) A subset of NSCLC cell lines was maintained and irradiated under anchorage-independent growth conditions, and the fraction of cells remaining after irradiation with 6 Gy were estimated using a metabolic assay as described. (C) Heatmap to illustrate the impact of KRAS/TP53 status on radiation survival of NSCLC cell lines. Relative radioresistance is indicated by red colors, representing the ratio of cell survival fraction at a given dose level for a given cell line and the geometric mean of survival fractions for all cell lines at that dose level. Ratios were arbitrarily binned as shown.

Figure 2.

Impact of KRAS status on radioresistance in an isogenic model system. (A) Clonogenic radiation survival of NCI-H1703 cells with endogenous wt KRAS, stably transfected with a KRAS mut (G12V) or a KRAS wt transgene or with an empty vector control. Survival curves were compared with the F-test. (B) Illustration of experimental set-up for determining local tumour control of NCI-H1703 xenografts transplanted on the hind legs of nude mice. (C) Measured local tumour control rates and estimated local tumour control probabilities (TCP) for NCI-H1703 xenografts with or without expression of mutant KRAS protein, where H1703 transfected with vector are designated “wt”. Error bars represent 95% confidence intervals of tumour control dose 50% (TCD₅₀).

Figure 3.

Radiosensitizing effects of erlotinib. (A) Dose modifying factor (DMF) for single dose RT with or without erlotinib in the different NCI-H1703 clones. DMF is the ratio of dose for RT alone and dose of RT plus erlotinib needed to sterilize 50% of cells in a clonogenic survival assay in vitro based on 2–3 biological repeats (see also Fig. S4). Values above 1.0 (dotted line) signify radiosensitization. (B) Measured local tumour control rates and estimated local tumour control probabilities (TCP) for NCI-H1703 xenografts with or without expression of mutant KRAS protein, with and without EGFR-inhibition. Error bars represent 95% confidence intervals of tumour control dose 50% (TCD₅₀). (C) DMF as the ratio of TCD₅₀ values for RT alone and RT plus erlotinib, taken from panel B (see also Fig. S3).