Photon and Proton irradiation in Patient-derived, Three-Dimensional Soft Tissue Sarcoma Models

Soft tissue sarcomas (STS) are a heterogeneous group of rare malignant tumors with more than 70 subtypes listed in the current World Health Organization (WHO) classification [1]. Despite their heterogeneity, guidelines still recommend the same radiotherapy (RT) regimen for all localized high-grade STS subtypes [2‐7]. Standard therapy comprises preoperative RT in daily fractions of 1.8-2.0 Gy to a total dose of 50-50.4 Gy followed by wide resection [2, 4, 5, 8, 9]. The low incidence and diversity of STS make large clinical trials and subtype-specific clinical research particularly challenging [10]. Patient-derived 3D cell cultures (human organoids) have become a valuable tool in the study of human diseases, complementing and in some cases replacing animal studies [11‐13]. In oncology, 3D cell cultures have been successfully used to study tumor development and progression and to test drug sensitivity [14‐20]. Sarcoma patient-derived 3D cell cultures (PD3D) represent an innovative tool to overcome challenges in clinical research and to conduct reproducible subtype-specific analyses on STS [21‐24]. Recently, Haas et al. used 2D cell lines to test subtype-specific radiosensitivity of STS [25]. The 14 well-characterized sarcoma cell lines showed striking differences in radiosensitivity after a single dose of photon irradiation (2–8 Gy). These findings, along with other preclinical data from 2D STS cell lines, suggest significant subtype-specific differences in radiosensitivity [25, 26]. However, 2D cell lines have important limitations. Forcing cells to grow on a 2D culture dish changes cellular morphology which leads to altered gene and protein expression and changes in cellular behavior compared to the tissue of origin [21, 27‐30]. These limitations are partially overcome by PD3D that have a microenvironment similar to that of the donor tissue including the spatial organization, extracellular matrix, nutrient and oxygen gradients that allow tumor cells to have natural cell-cell and cell-matrix interactions [13, 21, 31]. Three-dimensional cell cultures allow the growth of cells in their natural shape and show greater genetical and phenotypical similarity to the tumors in vivo which also translates into improved predictability of the effects of ionizing radiation [32, 33]. Moreover, differences in hypoxia and cell radiosensitivity between tumor core and border can be visually assessed in 3D cell cultures as well [34]. Three-dimensional cell culture models can also be used for multi-omics analyses and single cell sequencing to further individualize tumor treatment (e.g., by drug sensitivity screenings, finding target mutations etc.), which may support clinical decision-making [22, 35‐39]. To the best of our knowledge, there are no data on subtype-specific responses to radiation in STS 3D cell cultures yet. In this pilot study, we present our methodology and preliminary results using STS PD3D that were exposed to increasing doses of photon and proton radiation. We measured tumor cell viability at two different time points after irradiation. Our aim was: (i) to establish a reproducible method for irradiation of STS PD3D and (ii) to explore the differences in tumor cell viability of two different STS subtypes exposed to increasing doses of photon or proton radiation at two different measurement time points.

The current standard RT regimen for localized high-grade STS is far away from the idea of personalized tumor medicine. Instead, all STS subtypes receive the same preoperative RT regimen (1.8-2.0 Gy daily to a total dose of 50-50.4 Gy) despite clinical and preclinical evidence for subtype-specific differences in radiosensitivity [2‐4, 6, 7, 9, 25, 44‐46]. Three dimensional cell cultures have emerged as valuable preclinical models bridging the gap between animal models and humans in cancer medicine [11, 47‐49]. The challenges in clinical research for STS (low incidence, high heterogeneity) are therefore met by using 3D patient-derived cell cultures to understand tumor biology and to test drug or treatment efficacy [21, 22].

Herein, we present our pilot data on cell viability after photon and proton irradiation for PD3D of high-grade STS. The UPS PD3D showed less radiosensitivity after four days in all doses and radiation modalities compared to the PLS cell culture. While in the UPS cell culture 16 Gy of photon and proton irradiation barely caused a cell-killing effect of 50% after eight days, in the PLS cell culture half of the cells were eradicated by 8 Gy (photon and proton) only four days after irradiation [6]. These findings do correlate with typical clinical outcomes for both entities. While both tumors - UPS and PLS - have mediocre clinical prognoses compared to other STS entities, UPS tumors show a higher tendency for metastases and worse survival outcomes [50‐52]. Clinically, the high sensitivity of the PLS PD3D was also evident in the patient the cells were derived from. The patient received preoperative radiochemotherapy with 50.4 Gy in 28 daily fractions and surgical resection; the specimen has shown distinctly reduced proportions of viable tumor cells of 10–20%. To what extent the addition of chemotherapy has contributed to that effect cannot be assessed retrospectively. The UPS patient did not receive RT.

In both cell cultures and radiation modalities, the cell-killing effect of irradiation sustained after eight days. Although the survival measurement points in the 16 Gy dosage (photon and proton) of the UPS culture did show some degree of flattening at eight days, there was no sign of net cell number increase (Fig. 3).

The PLS cell culture on the other hand displayed a more dynamic cell turnover. After eight days, the PLS cells in the 0 Gy group overgrew in the organoid culture. Therefore, no valuable conclusions could be drawn from the results of the PLS group at eight days. Nevertheless, the PLS culture did show remarkable decreases after photon and proton irradiation at four days. This strikingly different and rather dynamic cell turnover of PLS compared to the UPS cells underlines the differences in radiosensitivity seen among both STS tumor cell cultures [6].

The physical advantages of proton irradiation are the characteristic energy deposition peak (‘Bragg peak’), behind which the energy drops towards near zero, minimizing dose deposition (“exit radiation”) in the tissue behind the Bragg peak [53]. Thereby, protons can deliver similar radiation doses to the target with 50–60% less integral or total radiation dose compared to photon intensity-modulated RT  [54‐58]. These normal tissue-sparing features of proton radiation make them particularly suitable for delicate tumor locations (skull-base, spine etc.) and for pediatric patients where the risk for late toxicity or radiation-associated malignancy is the highest [53, 59‐61]. Although there are no large well-matched trials comparing photon to proton therapy for sarcomas yet, many retrospective data analyses and phase II trials suggest promising low normal tissue toxicity and comparable local tumor control [59, 62‐66]. In line with these findings, we only observed minimal differences in cell-killing properties between photon and proton radiation in both tumor entities at both time points (except for the PLS cell culture at eight days, which is not analyzable due to cell overgrowth). It may therefore be interesting to assess normal tissue side effects irradiation by co-culturing STS tumor cells with physiological connective tissue cells, compare viability and additionally analyze established cellular radiation-induced DNA damage response markers such as γH2AX after irradiation (photon vs. proton) at different time points [67, 68].

Our data open up a new subfield of translational sarcoma research using sarcoma PD3D to individually assess radiosensitivity in different STS entities and patients individually. The results within our own study were reproducible and stable. A potential limitation of the approach is the absence of stromal and immune cells in current 3D cell culture models, as immune cells contribute to a variety of diseases and stromal cells have become important targets in cancer drug discovery [30, 69, 70]. However, the steady evolution of such multicellular 3D tissue models with immune cells will most likely lead to new models that will be able to mimic such complex interactions [30, 71]. Another limitation is the small number of two STS PD3D used and the lack of regular re-analysis of the phenotypic stability of the PD3D. The PD3D were initially analyzed histopathologically to confirm the diagnosis of the patient they were derived from (supplementary Figs. 1 and 2). However, the analysis was not repeated to assess the maintenance of phenotopic stability as cells grew and were passaged. Moreover, certain refinements are necessary in our model such as finding the appropriate tumor cell number for the rapidly overgrowing PLS culture. Further studies with more STS entities for the 3D patient-derived sarcoma cell culture radiosensitivity assay presented herein are warranted to extent our knowledge and eventually prepare clinical-translational validation.