Radiomics applies machine learning algorithms to quantitative imaging data to characterise the tumour phenotype and predict clinical outcome. For the development of radiomics risk models, a variety of different algorithms is available and it is not clear which one gives optimal results. Therefore, we assessed the performance of 11 machine learning algorithms combined with 12 feature selection methods by the concordance index (C-Index), to predict loco-regional tumour control (LRC) and overall survival for patients with head and neck squamous cell carcinoma. The considered algorithms are able to deal with continuous time-to-event survival data. Feature selection and model building were performed on a multicentre cohort (213 patients) and validated using an independent cohort (80 patients). We found several combinations of machine learning algorithms and feature selection methods which achieve similar results,e.g., MSR-RF: C-Index = 0.71 and BT-COX: C-Index = 0.70 in combination with Spearman feature selection. Using the best performing models, patients were stratified into groups of low and high risk of recurrence. Significant differences in LRC were obtained between both groups on the validation cohort. Based on the presented analysis, we identified a subset of algorithms which should be considered in future radiomics studies to develop stable and clinically relevant predictive models for time-to-event endpoints.

Cisplatin is an extensively used chemotherapeutic drug for lung cancer, but the development of resistance decreases its effectiveness in the treatments of non-small cell lung cancer (NSCLC). In this study, we examined the effects of metformin, a widely used antidiabetic drug, on cisplatin radiosensitization in NSCLC cell lines. Human NSCLC cell lines, A549 (cisplatin-resistant) and H460 (cisplatin-sensitive), were treated with metformin, cisplatin or a combination of both drugs before ionizing radiation. Cell proliferation, clonogenic assays, western blotting, cisplatin-DNA adduct formation and immunocytochemistry were used to characterize the treatments effects. Metformin increased the radiosensitivity of NSCLC cells. Metformin showed additive and over-additive effects in combination with cisplatin and the radiation response in the clonogenic assay in H460 and A549 cell lines (p = 0.018 for the interaction effect between cisplatin and metformin), respectively. At the molecular level, metformin led to a significant increase in cisplatin-DNA adduct formation compared with cisplatin alone (p < 0.01, ANOVA-F test). This was accompanied by a decreased expression of the excision repair cross-complementation 1 expression (ERCC1), a key enzyme in nucleotide excision repair pathway. Furthermore, compared with each treatment alone metformin in combination with cisplatin yielded the lowest level of radiation-induced Rad51 foci, an essential protein of homologous recombination repair. Ionizing radiation-induced γ-H2AX and 53BP1 foci persisted longer in both cell lines in the presence of metformin. Pharmacological inhibition of AMP-activated protein kinase (AMPK) demonstrated that metformin enhances the radiosensitizing effect of cisplatin through an AMPK-dependent pathway only in H460 but not in A549 cells. Our results suggest that metformin can enhance the effect of combined cisplatin and radiotherapy in NSCLC and can sensitize these cells to radiation that are not sensitized by cisplatin alone.

Uveal melanoma (UM) is the most common intraocular tumor in adults. Despite local tumor control, no effective therapy has been found to prevent metastasis, resulting in a high mortality rate. In the present study, we evaluated the anti-tumor potential of non-selective ß-blockers in 3D tumor spheroids grown from UM cell lines. Of the various ß-blockers tested, carvedilol and its enantiomers were most potent in decreasing the viability of Mel270 spheroids. Carvedilol at a concentration of 10–50 µM significantly elicited cytotoxicity and induced apoptosis in spheroid cells. In result, carvedilol inhibited tumor spheroid growth and compactness, and furthermore prevented the long-term survival and repopulation of spreading spheroid cells. The drug sensitivity of the different spheroids grown from Mel270, 92-1, UPMD2, or UPMM3 cell lines was dependent on 3D morphology rather than on high-risk cytogenetic profile or adrenergic receptor expression levels. In fact, the monosomy-3-containing UPMM3 cell line was most responsive to carvedilol treatment compared to the other cell lines. The concurrent treatment of UPMM3 spheroids with carvedilol and 5 or 10 Gy irradiation revealed additive cytotoxic effects that provided tumor control. Collectively, our data demonstrate the anti-tumor properties of carvedilol and its enantiomers, which may serve as candidates for the co-adjuvant therapy of UM.

Electrochemotherapy (ECT) is emerging as a complementary treatment modality for local tumor control in various cancer entities. Irradiation is an established therapeutic option for oncologic patients, which is commonly combined with chemotherapy due to its insufficient targeting ability. The efficiency of radiotherapy for tumors can be enhanced with different radiosensitizers. ECT can potentiate the radiosensitizing effect of chemotherapeutic agents such as bleomycin. The present study aims to evaluate the radiosensitizing effect of concomitant ECT with bleomycin on 3D tumor spheroids with primary and radioresistant uveal melanoma cell lines (UPMD2, UPMM3, UM92.1, Mel270) and irradiation. The changes in the spheroid growth and the cell viability as well the cytotoxic long-term effect of the combination treatment were evaluated with various combinations of electroporation settings and bleomycin concentrations as well as radiotherapy doses. A broad range of radiosensitivity was documented among the spheroids from different uveal melanoma cell lines. The primary cell lines showed a higher radiosensitivity and required lower irradiation and bleomycin doses. The maximal tumor control with a reduction of cell survival <10% was achieved with a 5 Gy irradiation only in the primary uveal melanoma cell lines and in combination with all tested ECT settings, whereas the same result could be obtained in UM92.1 spheroids only after ECT with 20 Gy irradiation. Based on the spheroid growth and the measurement of the cross-sectional area, the Mel270 spheroids, originating from a previously irradiated recurrent uveal melanoma, required higher doses of bleomycin and ECT settings after irradiation with 5 Gy in order to achieve a significant growth reduction. No significant difference could be demonstrated for the reduction of cell viability in the combination therapy with 20 Gy and 1000 V/cm between 1 and 2.5 µg/mL bleomycin even in Mel270 spheroids, underlying the importance of a drug delivery system to potentiate the radiosensitizing effect of agents in lower doses. ECT should be further assessed for its applicability in clinical settings as a therapeutic radiosensitizing option for radioresistant tumors and a sufficient local tumor control with lower chemotherapy and irradiation doses.

Glioblastomas are resistant to conventional therapies, including radiotherapy. Our previous study proved that epigenetic regulation influences the radiation response of glioma cells. This study evaluated the role of the acetyltransferase Tip60 on the radiation response. Tip60 expression was down-regulated by transfecting specific siRNA's in A7 and MO59K cells with high and low expression of Tip60, respectively, and its effect on survival was assessed. DNA repair was analyzed by foci scoring (γH2AX, Rad51, 53BP1, pATM). The interaction of Tip60 with ATM and DNA-PK was investigated using the specific inhibitors KU55933 and NU7441, respectively. Knockdown of Tip60 significantly (p < .001) reduced survival in both cell lines, but the effect was more pronounced in A7 cells. ATMi and DNA-PKi significantly reduced the surviving fraction following irradiation. However, no further effect of siTip60 on the radiosensitivity of ATMi treated A7 cells was observed. In contrast, DNA-PKi effectively enhanced the sensitizing effect of siTip60. Mechanistically, siTip60 reduced the number of initial Rad51 and ATM foci formation after irradiation and prevented their dissolution at 24 h. siTip60 had no impact on the formation of 53BP1 and γH2AX foci and did not further affect these end-points if combined with ATMi or DNA-PKi. Downregulation of Tip60 enhances the radiation sensitivity of both glioma cells and markedly elevates the radiation sensitivity when combined with DNA-PKi. Therefore, treatment with DNA-PK inhibitors represents a promising approach to augment the radiation sensitivity of glioma cell lines with deficient Tip60 activity in a synergistic manner.

Background: Concurrent cisplatin and radiotherapy frequently cause severe acute hematologic toxicity. This study investigated whether lymphocyte apoptosis, quantified via the γ-H2AX pan-nuclear immunostaining assay, can serve as a reliable biomarker for the acute hematologic side effects of combined chemoradiation. Methods: Blood samples were collected from cancer patients and healthy controls. Lymphocyte apoptosis was evaluated using the γ-H2AX pan-nuclear assay following in vitro and in vivo exposure to cisplatin and ionizing radiation (IR). In PBMCs, γ-H2AX results strongly correlated with conventional caspase-3 ( r = 0.9736) and annexin V ( r = 0.9780) apoptosis assays. Additionally, in vitro experiments assessed how phytohemagglutinin-L (PHA-L)-induced lymphocyte proliferation influences the combined effects of cisplatin and IR on apoptosis, alongside the regulatory role of the ATM protein. Results: In vitro, cisplatin increased lymphocyte apoptosis asymptotically, plateauing at 10 µg/ml (11.6 ± 5.0%). Combined cisplatin and IR treatment increased apoptosis in a strictly additive manner (radiation effect: p < 0.0001), with no significant synergistic interaction effect between the two modalities ( p = 0.22). The apoptotic effect level of in vivo cisplatin administration did not differ significantly from the maximum effect observed following a 1-hour in vitro exposure to high cisplatin concentrations. In vivo, the additive effect of cisplatin on IR-induced apoptosis peaked up to 3 days post-administration before returning to baseline by day 6. Consistent with these findings, cisplatin and radiation induced a significant, additive increase in apoptosis within PHA-L-stimulated lymphocytes. Inhibition of ATM has no effect on apoptosis following treatment with cisplatin and IR. Conclusions: The γ-H2AX pan-nuclear assay provides a valid, reliable method for quantifying radiation- and chemotherapy-induced lymphocyte damage. The time-dependent, additive toxicity observed indicates a critical clinical window (up to 3 days post-cisplatin) for maximum hematologic damage, allowing for optimization of treatment schedules. Clinical trial number: Not applicable.