FLASH Proton Radiation Therapy Mitigates Inflammatory and Fibrotic Pathways and Preserves Cardiac Function in a Preclinical Mouse Model of Radiation-Induced Heart Disease

Abstract

Purpose:

Studies during the past 9 years suggest that delivering radiation at dose rates exceeding 40 Gy/s, known as “FLASH” radiation therapy, enhances the therapeutic index of radiation therapy (RT) by decreasing normal tissue damage while maintaining tumor response compared with conventional (or standard) RT. This study demonstrates the cardioprotective benefits of FLASH proton RT (F-PRT) compared with standard (conventional) proton RT (S-PRT), as evidenced by reduced acute and chronic cardiac toxicities.

Methods and Materials:

Mice were imaged using cone beam computed tomography to precisely determine the heart’s apex as the beam isocenter. Irradiation was conducted using a shoot-through technique with a 5-mm diameter circular collimator. Bulk RNA-sequencing was performed on nonirradiated samples, as well as apexes treated with F-PRT or S-PRT, at 2 weeks after a single 40 Gy dose. Inflammatory responses were assessed through multiplex cytokine/chemokine microbead assay and immunofluorescence analyses. Levels of perivascular fibrosis were quantified using Masson’s Trichrome and Picrosirius red staining. Additionally, cardiac tissue functionality was evaluated by 2-dimensional echocardiograms at 8- and 30-weeks post-PRT.

Results:

Radiation damage was specifically localized to the heart’s apex. RNA profiling of cardiac tissues treated with PRT revealed that S-PRT uniquely upregulated pathways associated with DNA damage response, induction of tumor necrosis factor superfamily, and inflammatory response, and F-PRT primarily affected cytoplasmic translation, mitochondrion organization, and adenosine triphosphate synthesis. Notably, F-PRT led to a milder inflammatory response, accompanied by significantly attenuated changes in transforming growth factor β1 and α smooth muscle actin levels. Critically, F-PRT decreased collagen deposition and better preserved cardiac functionality compared with S-PRT.

Conclusions:

This study demonstrated that F-PRT reduces the induction of an inflammatory environment with lower expression of inflammatory cytokines and profibrotic factors. Importantly, the results indicate that F-PRT better preserves cardiac functionality, as confirmed by echocardiography analysis, while also mitigating the development of long-term fibrosis.

Introduction

Radiation therapy (RT) is a major component of the primary, adjuvant, and palliative treatment of a variety of malignancies, with more than half of all patients with cancer receiving some form of RT during their course of treatment.¹ Although the effectiveness of RT in controlling local tumors is well established, the potential of harm to surrounding normal tissues, which can negatively affect the overall survival and quality of life of cancer survivors, remains a significant concern. Among the malignant tumors located in the thoracic area that are commonly treated with RT, such as breast, lung, gastric, and esophageal cancers, a considerable number of patients are at risk of developing radiation-induced heart disease (RIHD) owing to direct or incidental exposure of the heart to radiation.^2–4 In a retrospective study, patients who received thoracic RT had a 2% higher absolute risk of cardiac morbidity and death at 5 years and a 23% increased absolute risk at 20 years posttreatment compared with nonirradiated (NR) patients.⁵

Although improvements in RT technology, such as intensity modulation RT, daily image guidance, motion mitigation, and proton therapy, have enabled more precise tumor targeting and allowed modest reductions in radiation dose, incidental exposure of the heart to RT and the consequent development of RIHD remain unavoidable.^6–8 Several pathologic processes can contribute to RIHD including cardiomyopathy, conduction disorders, myocardial fibrosis, pericarditis, acute coronary syndrome, congestive heart failure, and valvular disease.^9,10 RIHD often results in radiation-induced myocardial fibrosis (RIMF), which has a high incidence of 20% to 80% in patients who receive RT for thoracic tumors.¹¹ RIMF is a progressive and mostly irreversible process resulting from multicellular interactions mediated by multiple contributory events; however its underlying mechanisms are not yet fully understood. Therefore, new strategies are needed to mitigate or even reverse the course of RIMF.

In recent years, multiple preclinical studies have suggested that delivering RT at high dose rates (termed “FLASH” RT), can improve the therapeutic index by reducing normal tissue damage while maintaining tumor response compared with conventional or standard RT.^12–15 At our institution, we reported the first system to deliver FLASH proton RT (F-PRT) using double scattered protons and showed that it significantly decreased mortality from gastrointestinal fibrosis compared with standard PRT (S-PRT).¹⁶ A recent study demonstrated that F-PRT treatment effectively reduces skin injury, stem cell depletion, and inflammation, thereby protecting normal skin from radiation damage.¹⁷ Furthermore, it was shown that F-PRT induces significantly lower expression of the transforming growth factor β1 (TGF-β1), a key driver of tissue fibrosis, in both murine and canine skin compared with S-PRT.¹⁷ Significantly, in this study, the tumor growth inhibition was indistinguishable between the 2 RT modalities in multiple tumor models (melanoma, pancreatic adenocarcinoma, and sarcomas).

Although numerous reports have demonstrated the sparing effects of FLASH RT on multiple normal tissue models, its effects on cardiac tissue toxicity have yet to be determined. Using a mouse model of image guided, focal (apex) RT-induced cardiac damage that we previously developed,¹⁸ we demonstrate that compared with S-PRT, treatment of mouse hearts with F-PRT attenuates the induction of an inflammatory environment with lower expression of inflammatory cytokines and profibrotic factors. Studies using RNA-seq analyses identified differential activation of specific gene pathways by F-PRT compared with S-PRT, which provide mechanistic underpinning of the physiological changes. Critically, our studies show that F-PRT better maintains cardiac functionality, as analyzed by echocardiography analysis, while mitigating long-term fibrosis development.

Methods and Materials

Murine studies

Female 9- to 11-week-old C57BL/6 mice purchased from The Jackson Laboratory were maintained within the University of Pennsylvania animal facilities, and all experimental procedures involving live mice were conducted under the guidelines provided by protocols approved by the Institutional Animal Care and Use Committee. Using the Small Animal Radiation Research Platform (SARRP; Xstrahl Life Sciences), mice randomly assigned to treatment groups underwent image guided RT and were irradiated with a single dose of 40 Gy of S-PRT or F-PRT at the cardiac apex with a 5-mm diameter circular collimator. Mice were euthanized via CO₂ asphyxiation at various timepoints to collect heart tissues for further analysis. All experiments in this study are summarized in Figure 1c.

Fig. 1.

Irradiation setup. (a) Irradiation setting; the proton beam angled at 15^o to the vertical (created with BioRender.com). (b) Dose plan in MuriPlan. Frontal view of RT planning and delivery to the cardiac apex at a dose of 40 Gy with a 5-mm diameter circular collimator angled at 15^o to the horizontal. Purple and orange areas indicate the unirradiated and irradiated cardiac areas, respectively. (c) Timeline of experiments conducted in this study. (d) phosphorylated form of histone H2AX immunofluorescence staining in optimal cutting temperature-frozen FLASH proton RT or standard (conventional) proton RT-treated cardiac sections. The red dotted circle outlines the irradiated area. Scale bar, 1 mm. Abbreviation: RT = radiation therapy.

Irradiation setting

Mice were placed vertically on a custom 3-dimensional printed immobilization device with nose cone anesthesia delivery and imaged using the cone beam computed tomography of the SARRP.¹⁹ Imaging was performed to set the beam isocenter at the apex of the heart. Irradiation was performed using the fixed horizontal angle research proton beam line. An IBA Proteus Plus cyclotron was used to deliver a 230-MeV (range ~ 32 g/cm²) beam with a single scattered beam that was collimated with a 5-mm diameter circular collimator (Fig. 1a,b) using a shoot-through technique. The collimator was made of brass and placed with a 3-cm air gap to the mouse skin surface. Dose was measured using EBT-XD Gafchromic film (Ashland) calibrated against ion chamber measurements with an Advanced Markus parallel plate ionization chamber (PTW) with National Institute of Standards and Technology traceable calibration. We employed a National Institute of Standards and Technology-traceable calibrated Advanced Markus for dose measurements, and these measurements were corrected for recombination, polarity, and temperature-pressure variations in accordance with the International Atomic Energy Agency Technical Reports Series 398 protocol. The recombination correction factor applied was 1.004. Cross calibration of the film was performed 3 separate times, with 3-month intervals to verify consistent film response. A batch of EBT XD radiochromic film was calibrated in the proton beam using a large field (26-mm circular diameter) with 1 cm of solid water. Subsequently, the film was scanned using an Epson 1000XL flatbed scanner and a calibration curve between optical density and dose was established, following the procedure outlined in American Association of Physicists in Medicine Task Group 235.²⁰ The same calibrated batch of film was used for each experiment involving mice. Dose measurements were conducted for both standard and FLASH dose rates, revealing a minimal difference in total dose between the 2 treatment modalities (S-PRT, 39.93 ± 0.40; F-PRT, 39.95 ± 0.09). It’s worth noting that the uncertainty in the film dosimetry is estimated at around 5% to 10%, accounting for uncertainties associated with the calibration procedure, the reproducibility of the film scanner, and the setup variations. An online transmission ion chamber was cross-calibrated before each experiment to monitor the dose delivered to each mouse. FLASH ultrahigh dose rates (122.65 ± 2.35 Gy/s) were achieved using a requested cyclotron current of 360 nA, while standard dose rates (0.84 ± 0.07 Gy/s) were achieved using 2 nA. Stability in mouse-to-mouse dose and dose rate was observed within each irradiation session. The beam delivery was monitored and stopped after each irradiation using a monitor chamber with a verified, reproducible response at FLASH irradiation dose rates. Proton flux modulation was achieved through a current generator controlling the beam current regulation unit. Absolute dose was set by interrupting the current pulse to the beam current regulation unit, regulated by a preset counter connected to the monitor chamber. The time structure of the beam was recorded to verify the dose rate. The chamber’s response linearity was verified before use at ultrahigh dose rates.

Quantitative echocardiography

Ultrasound examination of the left ventricular (LV) was performed at 8 and 30 weeks post-PRT using the Visual Sonic Vevo 2100 system. Mice were anesthetized by intraperitoneal injection of 0.005 mL/g of 2% Avertin (2,2,2-tribromoethanol; Sigma-Aldrich). The mouse is placed on a warm platform in the supine position to maintain the body temperature at 37°C. The chest hair is shaved using hair removal gel cream (Nair). Warm ultrasound gel was applied to the scan field as couplant. Care was taken to maintain adequate contact while avoiding excessive pressure on the chest. To evaluate the LV systolic function, 2-dimensional (2D) LV long-axis, LV short-axis, and 2D guided LV M-mode images were obtained. Transmitral inflow Doppler spectra and Doppler waveforms were also recorded in an apical 4-chamber view to evaluate the LV diastolic function. After completion of the imaging studies, mice were monitored for full recovery and returned to their cages. Echo images were downloaded and analyzed offline using Visual Sonic Vevo 2100. Analyses were performed blinded to treatment group.

Immunofluorescence staining

Heart tissues were harvested and stored in optimal cutting temperature medium before being cut in 10-µm-thick sections, labeled, and stored at —80°C. Slides were allowed to equilibrate to room temperature before being subsequently fixed with 2% paraformaldehyde for 20 minutes. After 3 washes with phosphate buffered saline (PBS), tissues were enclosed with a hydrophobic barrier and blocked with 8% bovine serum albumin in PBS-TT (0.1% Tween-20, 0.025% Triton X-100) at room temperature for 1 to 2 hours. The primary antibodies against CD31 (1:50; BD Biosciences, 550274), a smooth muscle actin (SMA) (1:50; Sigma, C6198), gamma-histone H2AX (1:50; Millipore Sigma, 16–202A), and TGF-β1 (1:50; Proteintech, 21898–1-AP) were used to incubate the slides at 4°C. A negative control without primary antibody was performed in all cases. After overnight incubation, the slides were washed 3 times with PBS-TT before being incubated with the secondary antibodies (1:200; Invitrogen, A11006, A11007) for 1 to 2 hours at room temperature. Slides were washed again for 3 × 5 min in PBS-TT, and tissues were stained with Hoechst (1:1000; Invitrogen, H3570) for 30 min at room temperature and washed for 2 × 5 min in PBS-TT. Lastly, coverslips were mounted with an antifade mounting medium (Vector, H-1700) before further processing.

Preparation of heart tissue homogenates

To prepare heart homogenates for protein analysis, the irradiated region of the heart (apex) was collected and homogenized in 200 µL of radio-immunoprecipitation assay buffer, consisting of 10% protease inhibitor cocktail (MilliPore, 11836153001) and 1% Tween-20 using a homogenizer (Janke & Kunkel Ultra Turrax). Homogenates were then centrifuged at 14,000 rpm for 15 minutes at 4°C, and the supernatant was then collected and stored at —80°C until needed for further cytokine analysis.

Cytokine analysis: Multibead assay

Cardiac tissue lysates from the irradiated portions of mice hearts (specifically, the apex area covering approximately the lower third of the cardiac tissue) at 3-weeks post-PRT were used to analyze the expression of various cytokines through a multiplex assay (Millipore). Following the manufacturer’s protocol for the Luminex platform, expression for tumor necrosis factor α, interferon-g, interleukin 2, interleukin 6, vascular endothelial growth factor, and Placenta Growth Factor-2 were analyzed. Results from the multiplex assay were normalized to total protein concentration calculated by detergent compatible protein assay (Biorad).

Histologic staining

After 30 weeks, mice were euthanized, and the entire hearts were harvested for histologic evaluation. The isolated tissues were fixed in a 10% formalin solution and subsequently embedded in paraffin for further processing. Upon cutting, tissues were subjected to Masson’s Trichrome and Picrosirius Red staining.

RNA-sequencing: Sample collection, library preparation, and sequencing

At 2 weeks after PRT, we harvested the irradiated cardiac tissue (apex only) and processed it for RNA isolation. In brief, the tissue was minced, homogenized, embedded in a lysing solution, and total RNA was isolated using a Macherey-Nagel kit. The RNA samples were then submitted to the CHOP High Throughput Sequencing Core for additional processing and sequencing. RNA concentrations were measured with Qubit Flourometric Quantification (Theromo Fisher Scientific), and RNA integrity number was determined with Bioanalyzer on RNA 6000 Nano Kit (Agilent Technologies); all samples were high quality with RNA integrity number > 9. RNA-seq libraries were generated using NEBNext Poly (A) + ultra−Directional RNA library Prep Kit for Illumina (New England BioLabs) according to manufacturer’s instructions. Unique Dual Indexes Primer Pair Set 1 (New England BioLabs) was incorporated for multiplexed high-throughput sequencing. The final product was assessed for its size distribution and concentration using BioAnalyzer High Sensitivity DNA Kit (Agilent Technologies). Shortly Poly(A) was purified from 300 ng of total RNA using Oligo (deoxythymine) beads. The purified mRNA was fragmented, then subjected to reverse transcription, end repair and 3^՚-end adenylation, and adaptor ligation, followed by indexing through 9 cycles of polymerase chain reaction amplification followed by bead purification (Beckman Coulter). The resulting libraries were pooled, diluted to 2 nM using 10 mM Tris-HCl, pH 8.5, denatured, and loaded onto a P1–100 (PE50) flow cell on an Illumina NextSeq 1000 (Illumina, Inc) according to the manufacturer’s instructions. Demultiplexed and adapter-trimmed sequencing reads were generated using bcl2fastq.

RNA-sequencing data analysis

Raw RNA-Seq data sets were quality checked and preprocessed using FastQC (www.bioinformatics.babraham.ac.uk/projects/fastqc/) and Cutadapt.²¹ Alignment of the trimmed sequencing reads was conducted against the Mus musculus (version GRCm39) genome using the STAR aligner.²² Gene-level quantification was performed using RNA-Seq by Expectation-Maximization,²³ and expected counts for each gene were extracted for downstream analysis. Trimmed mean of M-values normalization²⁴ was applied to RNA-Seq by Expectation-Maximization read counts, and differential expression analysis was conducted using edgeR.²⁵ Differentially expressed genes were filtered based on the false discovery rate (<= 0.05). Pathway enrichment analysis (Gene ontology [GO] biologic processes) was performed using the clusterProfiler package²⁶ and the hypergeometric test module for upregulated and downregulated genes separately. Pairwise functional similarity and association of GO terms were calculated on the top 10 most significant pathways (in terms of false discovery rate) using naviGO.²⁷ Every GO term paired with a Lin’s semantic similarity >= 0.75 was connected. Visualizations were created using R version 4.0.2. ggplot2^28,29 and the draw.io software.

Digital imaging and quantification

Digital images of immunofluorescence staining were captured on a Zeiss Observer.Z1 inverted microscope (Zeiss) using the same exposure settings for all conditions. Immunofluorescent images were taken from the irradiated area (apex), with each sample being represented ideally by 10 ×10 magnification fields. Images from histochemical (Masson Trichrome and Picrosirius Red) staining were also captured from the irradiated area only (apex), using brightfield optics and a Nikon DSRi2 color camera at ×20 magnification. For each ×20 magnification field of Masson Trichrome stain, we located all blood vessels and conducted 10 to 15 measurements (thickness of positive histochemical staining in mm) to calculate the average vessel wall thickness per mouse, using the Aperio ImageScope software.¹⁸ For each ×20 magnification field of Picrosirius Red stain, we calculated the percent of positive stain per field using the area function within the ImageJ Fiji software.³⁰ The signal of the immunofluorescent images was subjected to image analysis through ImageJ Fiji.³⁰ Briefly, the background for the channel of interest was subtracted using negative controls for each antibody. Subsequently, automatic thresholding was applied to each channel. Fluorescence intensity and area of all thresholded regions were measured. Quantification is presented as integrated density (for TGF-β1) and area (for aSMA) normalized to total nuclei area.

Statistical analysis

Statistical analysis was conducted using GraphPad Prism 9.1.2 software. Data are presented as box and whisker plots showing all points (min to max) with each dot representing quantitative value from a 10× field. To compare the groups (pairwise comparisons), we used a 2-tailed t test.

Data availability

RNA-Seq raw data sets were deposited in the Gene Expression Omnibus database at National Center for Biotechnology Information under accession number GSE261367.

Results

Mouse model of localized heart damage post-FLASH and S-PRT

We previously developed and characterized a pathophysiologically relevant mouse model of radiation-induced cardiotoxicity involving targeted in situ irradiation of the cardiac apex.¹⁸ The first objective of this study was to test the extent of localized cardiac apex toxicity (partial heart irradiation) after F-PRT or S-PRT. Anesthetized mice were placed on a stage at a vertical position in front of the proton beam line (Fig. 1a). Using the SARRP’s Control Interface, contrast-enhanced cone beam computed tomography images were initiated and an RT plan was designed accordingly (Fig. 1b). Single doses of 40 Gy were precisely delivered to the cardiac apex minimizing exposure to adjacent lung tissue. No deaths or skin toxicities were recorded during or at any timepoint post-PRT in any of the treatment schemes and modalities, and all mice remained healthy after RT. To evaluate the localized cardiac damage post-PRT, we stained frozen cardiac tissues harvested at 1 hour post-PRT for ghistone H2AX, a sensitive and quantitative method for DNA double-strand breaks. Results from the immunofluorescence staining confirmed that DNA damage was restricted to the ~20 to 30 mm² area in the lower third of the heart (cardiac apex; Fig. 1d), leaving the surrounding tissues intact. We conclude that both treatment modalities cause comparable localized damage to the apex of the mouse heart.

RNA profiling of PRT-treated cardiac tissue reveals uniquely regulated pathways in each treatment modality

To gain a better understanding of the effects of each treatment modality on cardiac tissue, we performed bulk RNA-sequencing on NR samples, as well as apexes treated with F-PRT or S-PRT at 2 weeks after a single dose of 40 Gy (Fig. 2a,b). Our focus on this early timepoint was aimed at capturing the differential gene expression changes that may potentially manifest as long-term effects driven by each treatment modality. We identified several pathways uniquely upregulated by S-PRT but not in the F-PRT group (Fig. 2c), including pathways associated with DNA damage response, p53 signaling, induction of tumor necrosis factor superfamily, and inflammatory response. Moreover, certain pathways exhibited downregulation in response to S-PRT while remaining unaltered by F-PRT (Fig. E1a), particularly those involving vascular wound healing, regulation of endothelial and epithelial cell proliferation, and heart regeneration. These pathways and corresponding genes are of significant interest as they potentially signify cardiac tissue damage introduced by S-PRT. Conversely, the top 10 pathways upregulated in F-PRT, yet unaffected by S-PRT (Fig. 2d), primarily pertain to cytoplasmic translation, mitochondrion organization, and adenosine triphosphate synthesis. Furthermore, pathways uniquely downregulated by F-PRT included several associated with tissue morphogenesis and regulation of developmental growth (Fig. E1b). Together, these findings suggest that F-PRT elicits a gene expression response that leans toward a state of tissue repair, in contrast to the gene expression profile induced by S-PRT, which is associated with a more “damagelike” and inflammatory signature.

Fig. 2.

RNA-seq analysis on S-PRT and F-PRT treated hearts (apex) at 2 weeks post-proton RT. (a) Groups of samples (n = 4 for nonirradiated; n = 4 for S-PRT; and n = 3 for F-PRT) analyzed using multidimensional scaling plots. (b) Venn diagram indicating the number of significant (false discovery rate < 0.01) differentially expressed genes across 2 key comparisons (F-PRT, S-PRT) and the overlap between each set of genes. (c, d) Gene ontology enrichment analysis of the differentially expressed genes (upregulated) in the heart of S-PRT treated mice compared with F-PRT treated mice; n = 3–4 per group. Abbreviations: F-PRT = FLASH proton RT; RT = radiation therapy; S-PRT = standard (conventional) proton RT.

F-PRT ameliorates cardiac inflammation and profibrotic responses induced by proton radiation

We then sought to investigate whether the differences in gene expression between the treatment modalities correspond to distinct inflammatory responses. Consequently, we analyzed heart lysates from F-PRT or S-PRT treated mice at 3-weeks post 40 Gy of PRT by employing a mouse-specific multiplex cytokine/chemokine microbead assay. Intriguingly, the levels of the inflammatory cytokine tumor necrosis factor a (levels were undetectable in a sample from each treatment modality, resulting in n = 4) were significantly decreased in the F-PRT-irradiated hearts compared with hearts from the S-PRT group (Fig. 3a). Moreover, the levels of interferon g decreased (although not significantly) in the F-PRT group compared with the S-PRT group (Fig. E2a), indicating possibly a difference in kinetics between the inflammatory cytokines. These cytokines have also been linked to triggering or amplifying the fibrotic response.^31,32 Analysis of the levels of additional cytokines did not show any difference between the 2 treatment modalities (Fig. E2b-e), possibly because of the early timepoint analyzed. Additionally, immunofluorescence staining for the profibrotic marker TGF-β1 at 2 weeks post-PRT revealed substantially higher levels in irradiated hearts from S-PRT-treated mice compared with the NR and F-PRT groups, and F-PRT mice were almost indistinguishable from the NR group (Fig. 3b). The levels of TGF-β1 were further increased in S-PRT group at 3-weeks post-PRT, and in the F-PRT group, they were maintained at low levels (Fig. 3c-e), further supporting the notion of a less fibrotic environment in F-PRT-treated mice. The induction of TGF-β1 promotes the activation of fibroblasts to produce excessive amounts of collagen and other proteins in the extracellular matrix.³³ Therefore, we sought to investigate if F-PRT or S-PRT affect the fibroblast activation post-PRT by costaining sections of irradiated hearts from F-PRT or S-PRT treated mice for CD31 (a marker of endothelial cells) and aSMA (Acta2; a common marker of fibroblast activation). Notably, levels of aSMA in the F-PRT group were nearly equal to those in the untreated (NR group) hearts compared with the substantially high levels expressed in the S-PRT group (Fig. 3f,g). Interestingly, the high levels of aSMA expression in S-PRT treated hearts were primarily restricted in the perivascular area (denoted by the CD31 staining), indicating that this fibroblast subpopulation may interact with the damaged endothelium to drive the fibrogenic events postfocal PRT, which are evident in the Trichrome and Picrosirius Red staining of the tissues.

Fig. 3.

Significant reduction of inflammatory and profibrotic responses in F-PRT treated hearts compared with S-PRT group. (a) Fold change (normalized to nonirradiated, n = 4) of tumor necrosis factor a from heart lysates at 3 weeks post 40 Gy of focal PRT (n = 4 for S-PRT and n = 4 for F-PRT). (b) Quantification of integrated density of TGF-β1 production at 2 weeks and (c) at 3 weeks post-PRT (n = 5 for S-PRT and n = 5 for F-PRT). Each dot represents quantitative value from a ×10 field. (d) Representative images of TGF-β1 staining. (e) Mean TGF-β1 integrated density at various timepoints post-PRT from (b) and (c). (f) Quantification of %α-smooth muscle actin + area at 3 weeks post-PRT (n = 5 for S-PRT and n = 5 for F-PRT). (g) Representative images of aSMA/CD31 costaining. Scale bar, 100 mm. Box and whisker plots. P values calculated with 2-tailed t test. *P < .05, **P < .01, ***P < .001. Abbreviations: F-PRT = FLASH proton RT; n.s.= not significant; PRT = proton RT; RT = radiation therapy; S-PRT = standard (conventional) proton RT; TGF = transforming growth factor.

F-PRT reduces collagen deposition and muscle layer thickness compared with S-PRT

To test the effects of PRT on the long-term consequences of an inflammatory and fibrotic environment post-RT, we irradiated the cardiac apexes with a single dose of 40 Gy of F-PRT or S-PRT. Thirty weeks later, we harvested the treated hearts and evaluated the levels of fibrosis collagen and deposition. Quantitative analysis of Masson’s trichrome-stained tissues (vessel wall thickness) showed a significant reduction in the perivascular fibrosis in F-PRT treated hearts compared with S-PRT group (Fig. 4a,b). Surprisingly, the levels of perivascular fibrosis in the F-PRT group were comparable to the levels expressed in the NR group (Fig. 4b). Additionally, the increased levels of collagen, mostly deposited in the perivascular area, defined by the Picrosirius Red staining, again were substantially lower in the F-PRT group compared with the S-PRT treated hearts (Fig. 4c,d). These findings indicate that the F-PRT significantly reduces the RIMF by lowering the levels of a chronic inflammatory environment post-PRT.

Fig. 4.

Significant reduction of myocardial fibrosis in F-PRT treated mice at 30 weeks post-PRT. (a) Representative Masson’s trichrome stain images (n = 2) of formalin fixed heart sections irradiated with 40 Gy at 30 weeks post-PRT. (b) Quantification of fibrosis formation by calculating the average muscle layer thickness (n = 3 for nonirradiated; n = 5 for standard [conventional] proton RT; and n = 5 for F-PRT). (c) Representative Picrosirius red stain images (n = 2) of formalin fixed heart sections irradiated with 40 Gy at 30 weeks post-PRT. (d) Quantification of the percentage of collagen area in irradiated heart areas (n = 3 for nonirradiated; n = 5 for standard [conventional] proton RT; and n = 5 for F-PRT). Box and whisker plots. P values calculated with 2-tailed t test. *P < .05, **P < .01, ***P < .001; Scale bar, 200 mm. Abbreviations: F-PRT = FLASH proton RT; n. s. = not significant; PRT = proton RT; RT = radiation therapy.

Cardiac functionality is better preserved after F-PRT

To further assess the functionality of cardiac tissue and correlate with the levels of fibrotic tissue, we used 2D echocardiograms (2D Echo) at 8 and 30 weeks after 40 Gy of F-PRT or S-PRT. Interestingly, at 8 weeks, heart functionality of F-PRT mice closely resembled that of NR mice whereas S-PRT mice showed significant reduction in multiple cardiac parameters, including LV internal dimension (diastole [LVIDd] and systole), end-systolic LV volume, and so forth (Fig. 5a-c; Fig. E3a). Intriguingly, 2D Echo at 30 weeks after PRT showed that, despite 1 mouse in the F-PRT group succumbing to death during anesthesia (reducing the group to n = 4), the remaining mice maintained levels of most critical cardiac parameters that were similar to those of the NR group (Fig. 5d-h). In contrast, S-PRT treated hearts displayed a substantial deterioration described by the significant decrease in LVIDd, LV mass, and ejection time and increase in relative wall thickness (Fig. 5d-h). Collectively, these results demonstrate that F-PRT better preserved the cardiac functionality than S-PRT even at this longer timepoint.

Fig. 5.

Preservation of heart functionality in F-PRT treated mice. (a-c) Diastolic 2-dimensional echocardiography on S-PRT and F-PRT treated hearts at 8 weeks post-PRT (n = 5 for nonirradiated; n = 5 for S-PRT; and n = 5 for F-PRT). (d-g) Diastolic 2-dimensional echocardiography on S-PRT and F-PRT treated hearts at 30 weeks post-PRT (n = 5 for nonirradiated; n = 5 for S-PRT; and n = 4 for F-PRT). Box and whisker plots. P values calculated with 2-tailed t test. *P < .05, **P < .01, ***P < .001. (h) Representative images of relative wall thickness at 30-weeks post-PRT. Abbreviations: ESV = end-systolic left ventricular volume (mL); ET = ejection time; F-PRT = FLASH proton RT; LVIDd = LV internal dimension, diastole (mm); LVIDs = LV internal dimension, systole (mm); LVM = LV mass (mg); n.s.= not significant; PRT = proton RT; RT = radiation therapy; S-PRT = standard (conventional) proton RT.

Discussion

In this study, we have compared, for the first time, the effects of F-PRT and S-PRT on radiation-induced cardiac toxicities with an emphasis on myocardial fibrosis. Using our mouse model of image guided, focal (apex) RT-induced cardiac damage,¹⁸ we delivered 40 Gy of focal F-PRT and S-PRT and demonstrated that F-PRT induces a unique gene signature that correlates with tissue repair and reduced inflammation compared with S-PRT. Critically, we found that F-PRT mitigates late-stage RIHD (primarily fibrosis) and preserves cardiac functionality compared with S-PRT.

Despite the progress made in delivery of RT for thoracic malignancies, RIHD is becoming an increasingly significant clinical concern. Although recent research has offered new insights into the salient features of this condition, there is still lack of knowledge regarding the molecular and pathophysiologic mechanisms of cardiovascular toxicity as well as optimal approaches for the treatment of patients with cardiovascular dysfunction. Until recently, animal models for studying irradiation-induced heart injury were limited to whole thorax or whole heart with partial lung radiation.³⁴ Here, we used a pathophysiologically relevant mouse model of radiation-induced cardiotoxicity that we previously developed¹⁸ and describe our successful implementation of targeted and image guided proton ultrahigh dose rate FLASH delivery to the heart. This study builds upon previous work involving small animal image guided proton irradiator and proton FLASH techniques to treat skin and intestinal tissues.^16,17,19 To achieve our goal, we developed a comprehensive protocol encompassing dosimetry, positioning, and targeting procedures. This protocol ensured precise and reproducible small-field proton irradiation specifically focused on the apex of a mouse heart.

The mechanisms of RIHD are multifactorial, involving direct DNA damage, increased oxidative stress, injury to vascular endothelial cells, persistent inflammation, and fibrosis.^31,35–37 Even though cardiomyocytes have shown relative resistance to radiation,³⁸ they can still experience degeneration owing to oxidative stress and metabolic abnormalities post-RT.³⁹ The initial response to RIHD involves the activation of the coronary microvascular endothelium, primarily through the nuclear factor-kB signaling pathway.

Consequently, this leads to increased endothelial dysfunction, vascular permeability, and deposition of fibrin and other blood components in the interstitium, contributing to microvascular obstruction.⁴⁰ This is subsequently followed by an acute inflammatory response and by the release of multiple cytokines and chemokines, an effect observed in other tissues exposed to RT.^41,42 In this study, we employed RNA-seq and immunofluorescence analyses of tissue sections to explore the underlying mechanisms for the differential effects of FLASH relative to standard dose rate.

Pathways associated with DNA damage and inflammatory responses exhibited upregulation with S-PRT whereas they were unaltered in F-PRT at 2 weeks after 40 Gy of PRT. Through the evaluation of cytokine/chemokine levels and immunofluorescence staining, we corroborated the significantly higher expression of inflammatory and fibrotic markers resulting from S-PRT treatment. Other studies using electron, proton, and x-ray FLASH RT have reported significant reduction in inflammatory responses compared with conventional RT.^17,43,44 Although there were no significant differences in the levels of other cytokines/chemokines between the 2 treatment modalities, we believe that additional timepoints post-PRT are required to establish a more detailed inflammatory profile.

The damaged microvasculature and persistent inflammatory response in the damaged area further trigger the release of TGF-β, platelet-derived growth factor, and connective tissue growth factor, which stimulate the proliferation of endothelial cells and fibroblasts.³³ Recently, it was demonstrated that F-PRT spares normal skin by reducing skin injury, stem cell depletion, and inflammation.¹⁷ Moreover, the same study revealed that F-PRT leads to notably lower TGF-β1 expression compared with S-PRT in both murine and canine skin, indicating a potentially reduced risk of fibrosis development after F-PRT treatment. Similarly, in vitro and in vivo studies have reported substantial reduction in TGF-β1 expression followed by FLASH RT compared with conventional treatment.^12,45 As part of this study, we also assessed the levels of TGF-β1, as a common mediator of inflammation and fibrosis post-RT.^35,46 Interestingly, we found that S-PRT caused significant upregulation of TGF-β1 at 2-weeks post-PRT, which further increased at 3-weeks post-PRT. In contrast, TGF-β1 expression was maintained at very low levels after F-PRT treatment. The differential expression of TGF-β1 between the treatment modalities is reflected in the activation of fibroblasts, as shown by the substantially higher levels of aSMA expression in the S-PRT group compared with the F-PRT group, which were nearly equal to the unirradiated hearts (NR group). It is well-established that the activated myofibroblasts result in excessive collagen production and the development of myocardial fibrosis.^32,33 A significant finding is that the perivascular fibrosis development was markedly elevated in hearts treated with S-PRT, whereas in the F-PRT group, the levels were comparable to those observed in the NR group. To further examine the effect of fibrosis development on heart functionality, we used quantitative echocardiography and identified systolic and diastolic dysfunction. Surprisingly, we found that compared with the S-PRT treated hearts, the F-PRT focally irradiated hearts preserved many crucial parameters, including LVIDd, LV mass, ejection time, and relative wall thickness, with levels closely resembling those of nonirradiated mice. This finding is of high importance because the long-term FLASH sparing effects are not well documented.

Increased heart radiation dose is strongly associated with noncancer-related morbidity and mortality in patients receiving thoracic RT, with a range of heart doses proposed to relate to toxicity, including volume of heart receiving >50 Gy and mean heart dose >10 Gy.^47,48 Importantly, the time scale for these effects is in months to 2 to 3 years after completion of RT, and the mechanism(s) by which heart RT leads to increased morbidity and mortality remains an area of intense active investigation. In our previous work, we found that partial heart volume irradiation with a single fraction of 40 Gy produces pathologic and physiological heart damage within 8 to 16 weeks that mimics clinical multifraction RT-induced heart disease.¹⁸ Furthermore, we emphasize the increasing relevance of this model to the clinical use of ultrahypofractionated (1–5 fraction) regimens in thoracic RT, as well as in clinical and preclinical investigations of F-PRT.^49–52 Nevertheless, we acknowledge that this single fraction, 8- to 16-week model may limit our ability to observe other potentially clinically relevant aspects of RIHD. Having established the FLASH sparing effects through this model, the future steps include the implementation of additional single doses to generate dose-response curves for various endpoints and to investigate the effect of 3 to 5 fraction F-PRT regimens with longer follow-up periods. In addition, our future studies will explore the effect of focal PRT on the unirradiated base structure of the mouse heart, given its sensitivity to radiation. Despite these limitations, we believe that our study holds important ramifications for patient care and that, ultimately, F-PRT should be tested in larger animal models (eg, mini pigs) characterized by longer lifespans and potentially improved radiobiological modeling of human tissues.

Conclusion

In this study, we demonstrate that F-PRT spares cardiac apex from both acute and chronic toxicities compared with S-PRT. F-PRT initiates a less intense inflammatory response, accompanied by a minimal change in TGF-β1 lev els. RNA-seq data and immunofluorescence analysis, in turn, suggest that F-PRT mitigates RIMF and maintains cardiac functionality. Overall, this study set the base to further test the ability of F-PRT to enhance the management of thoracic malignancies compared with traditional PRT and to initiate human clinical trials for patients with cancer undergoing RT.

Data Sharing Statement:

Research data are stored in an institutional repository and will be shared upon request to the corresponding author.