Effectiveness of FLASH vs. Conventional Dose Rate Radiotherapy in a Model of Orthotopic, Murine Breast Cancer

Abstract

Introduction: Radiotherapy is effective for breast cancer treatment but often causes undesirable side effects that impair quality of life. Ultra-high dose rate radiotherapy (FLASH) has shown reduced normal tissue toxicity while achieving comparable tumor growth delay compared to conventional dose rate radiotherapy (CONV). This study evaluated whether FLASH could achieve similar tumor control as CONV with tumor eradication as the primary endpoint, in an orthotopic breast cancer model. Methods: Non-metastatic, orthotopic tumors were generated in the left fourth mammary fat pad using the Py117 mammary tumor cell line in syngeneic C57BL/6J mice. Two sequential irradiation studies were performed using FLASH (93-200 Gy/s) and CONV (0.08 Gy/s) electron beams. Single fractions of 20, 25, or 30 Gy were applied to tumors with varying abdominal wall treatment fields (~3.75 or 2.5 mm treatment margin to tumor). Results: Both FLASH and CONV demonstrated comparable efficacy. Small tumors treated with 30 Gy and larger abdominal wall treatment fields appeared to have complete eradication at 30 days but also exhibited the highest skin toxicity, limiting follow-up and preventing confirmation of eradication. Smaller abdominal wall treatment fields reduced skin toxicity and allowed for extended follow-up, which resulted in 75% tumor-free survival at 48 days. Larger tumors showed growth delay but no eradication. Conclusions: In this preclinical, non-metastatic orthotopic breast cancer model, FLASH and CONV demonstrated equivalent tumor control with single-fraction doses of 20, 25, or 30 Gy. Overall, 30 Gy achieved the highest eradication rate but also resulted in the most pronounced skin toxicity.

Keywords: FLASH radiotherapy; breast cancer; breast conservation; lumpectomy; orthotopic murine breast cancer; radiotherapy; radiotherapy toxicity; syngeneic; ultra-high dose rate radiotherapy.

Figure 1

In vivo breast cancer orthotopic tumor irradiation with FLASH or CONV dose rates. (A) Breast cancer orthotopic tumor-bearing mouse with approximately 30 mm³ tumor at the fourth mammary fat pad (top; black dashed circle and arrow). An anesthetized mouse placed inside the mouse positioner frame with the tumor centered at the bottom of the lateral opening and immobilized with paper tissue (bottom). (B) 3D computer-aided design (CAD) files illustrating the positioning of the mouse positioner on the collimator, the positioning of the radiochromic film during irradiation. (C) Top view of the collimator featuring the mouse positioner in two treatment margins to tumor: (i) Round 1, positioned to expose a 7.5 × 20.0 mm² area of the abdominal wall tissue, 3.75 mm margin or (ii) Round 2, positioned to expose a 5.0 × 20.0 mm² area of the abdominal wall tissue, 2.5 mm margin. (D) 3D CAD file of the 2.0 × 2.0 cm² collimator presented with a lateral cross-section, illustrating mouse radiation shielding. The collimator is filled with a 3 cm layer of aluminum oxide to stop the electrons and a 1 cm layer of tungsten spheres (2.0 mm diameter) to absorb Bremsstrahlung radiation and efficiently shield the rest of the animal’s body (Bremsstrahlung radiation leakage ~0.2%). (E) FLASH and CONV beam geometries. For FLASH irradiations, the collimator and mouse positioner frame are placed inside the treatment head and the beam entrance surface of the mouse is 18.7 cm from the scattering foil. For CONV irradiations, at Round 1, the beam entrance surface of the mouse was 76.1 cm (unmatched geometries) from the scattering foil, and at Round 2 at 18.7 cm (CONV* notation matching geometry with FLASH). The pulses delivered are monitored using an ion chamber, measuring the Bremsstrahlung tail at 11.0 cm solid water depth, and are located 147.5 cm from the scattering foil.

Figure 2

Radiochromic film dosimetry of delivered doses and film-derived collimator characterization. (A) Grouped scatter plot of film-derived mean dose of all animals combined (Sm for small 20–40 mm³ and Lg for large 250–800 mm³ tumor volumes). Red arrows indicate missed pulses, which resulted in elimination of these animals from the analysis. (B) Top view of experimental set up illustrating the film positioning and exposure. (C) Representative exposed films with either FLASH or CONV 25 Gy single fraction target dose group, illustrating the direction of the beam profiles. (D) Film-derived X (transverse) and Y (craniocaudal) profiles of the unmatched geometries between FLASH and CONV with 7.5 × 20 mm² abdominal wall treatment field used in Round 1 of irradiation, and (E) profiles from the matched geometries with the 5 × 20 mm² abdominal wall treatment field used in Round 2. Salmon highlights represent the exposure of the abdominal wall in the X direction. (F) Film-derived percentage depth-dose curve (PDD) from FLASH and both geometries of CONV configurations, using films parallel to the direction of the irradiation beam. Overall, the delivered doses between FLASH and CONV were comparable and despite the difference in source-to-surface (SSD), profiles and PDDs of the two modalities remain comparable within the tumor volumes.

Figure 3

(A–D) Tumor measurements with calipers plotted as response curves of breast cancer orthotopic tumors of small volume (20–40 mm³) irradiated with 20, 25, and 30 Gy single fraction with either FLASH or CONV dose rates. (A) Tumor volumes from animals irradiated with 3.75 and 2.5 mm treatment margin combined (n = 7 FLASH, n = 8 CONV). FLASH group had one animal excluded due to a missed pulse. Tumors were controlled for the first 4 weeks and regressed thereafter, with no significant differences between groups at any time point post-irradiation (Mann–Whitney U test, all p > 0.05). (B) Tumors targeted with 25 Gy and 2.5 mm treatment margin (n = 7 FLASH, n = 8 CONV); one FLASH exclusion. Tumors remain controlled for 4 weeks and regressed thereafter, with no differences between groups (Mann–Whitney U test, all p > 0.05). (C) Tumor treated with 30 Gy and 3.75 mm treatment margin (n = 4 per group) were controlled by day 30, but severe tissue toxicity led to study termination. No significant differences observed (all p > 0.05). (D) Tumors treated with 30 Gy and 2.5 mm treatment margin remained controlled for the first 4 weeks (all p > 0.05). By day 48, only one tumor per group showed regrowth (mean rank difference = 0.583 mm³; Mann–Whitney U test, p = 0.829). Overall, there was no significant difference between FLASH and CONV in tumor growth delay or eradication of small tumor volumes (20–40 mm³) with single fractions of 20, 25, and 30 Gy.

Figure 4

Tumor measurements of breast cancer orthotopic tumors with calipers plotted from non-irradiated controls and tumors of large volume (250–800 mm³) irradiated with 30 Gy single fraction with either FLASH or CONV dose rates. (A) Tumor growth curve of unirradiated breast cancer orthotopic tumors at the fourth mammary fat pad (n = 7). Small size tumor volumes were selected for irradiation on day 5 (20–40 mm³; light blue arrow), and larger range of tumor volumes at day 24 (250–800 mm³; red arrow). Tumors grow moderately for the first 2 weeks and exponentially thereafter. (B) Tumor response curve of breast cancer orthotopic tumors of large volume range irradiated with 30 Gy single fraction of either FLASH or CONV dose rates (n = 6 per group). Tumor volumes were suppressed for the first 2 weeks and regressed thereafter with no statistically significant differences observed between FLASH and CONV at any time point post-irradiation (Mann–Whitney U test, all p > 0.05). There was no significant difference between FLASH and CONV in tumor growth delay of large tumor volumes (250–800 mm³) with a single fraction of 30 Gy.