Extensive clinical testing of Deep Learning Segmentation models for thorax and breast cancer radiotherapy planning.

BACKGROUND: The performance of deep learning segmentation (DLS) models for automatic organ extraction from CT images in the thorax and breast regions was investigated. Furthermore, the readiness and feasibility of integrating DLS into clinical practice were addressed by measuring the potential time savings and dosimetric impact. MATERIAL AND METHODS: Thirty patients referred to radiotherapy for breast cancer were prospectively included. A total of 23 clinically relevant left- and right-sided organs were contoured manually on CT images according to ESTRO guidelines. Next, auto-segmentation was executed, and the geometric agreement between the auto-segmented and manually contoured organs was qualitatively assessed applying a scale in the range [0-not acceptable, 3-no corrections]. A quantitative validation was carried out by calculating Dice coefficients (DSC) and the 95% percentile of Hausdorff distances (HD95). The dosimetric impact of optimizing the treatment plans on the uncorrected DLS contours, was investigated from a dose coverage analysis using DVH values of the manually delineated contours as references. RESULTS: The qualitative analysis showed that 93% of the DLS generated OAR contours did not need corrections, except for the heart where 67% of the contours needed corrections. The majority of DLS generated CTVs needed corrections, whereas a minority were deemed not acceptable. Still, using the DLS-model for CTV and heart delineation is on average 14 minutes faster. An average DSC=0.91 and H95=9.8 mm were found for the left and right breasts, respectively. Likewise, and average DSC in the range [0.66, 0.76]mm and HD95 in the range [7.04, 12.05]mm were found for the lymph nodes. CONCLUSION: The validation showed that the DLS generated OAR contours can be used clinically. Corrections were required to most of the DLS generated CTVs, and therefore warrants more attention before possibly implementing the DLS models clinically.

Surface-guided positioning eliminates the need for skin markers in radiotherapy of right sided breast cancer: A single center randomized crossover trial.

BACKGROUND AND PURPOSE: To prospectively investigate whether surface guided setup of right sided breast cancer patients can increase efficiency and accuracy compared to traditional skin marker/tattoo based setup. MATERIAL AND METHODS: Twenty-five patients were included in this study. Each patient was positioned using skin marks and tattoos (procedure A) for half of the fractions and surface guidance using AlignRT (procedure B) for the other half of the fractions. The order of the two procedures was randomized. Pretreatment CBCT was acquired at every fraction for both setup procedures. A total of ten time points were recorded during every treatment session. Applied couch shifts after CBCT match were recorded and used for potential error calculations if no CBCT had been used. RESULTS: In the vertical direction procedure B showed significant smaller population based systematic (Æ©) and random (σ) errors. However, a significant larger systematic error on the individual patient level (M) was also shown. This was found to be due to patient relaxation between setup and CBCT matching. Procedure B also showed a significant smaller random error in the lateral direction, while no significant differences were seen in the longitudinal direction. No significant difference in setup time was found between the two procedures. CONCLUSION: Setup of right sided breast cancer patients using surface guidance yields higher accuracy than setup using skin marks/tattoos and lasers with the same setup time. Patient alignment for this patient group can safely be done without the use of permanent tattoos and skin marks when utilizing surface-guided patient positioning. However, CBCT should still be used as final setup verification.

Low dose-rate irradiation with [3H]-labelled valine to selectively target hypoxic cells in a human colorectal cancer xenograft model.

BACKGROUND: Earlier in vitro studies show that irradiation with an ultra-low dose-rate of 15 mGy/h delivered with [3H]-valine leads to loss of clonogenicity in hypoxic T-47D cells. Here, the aim was to determine if [3H]-valine could be used to deliver low dose-rate irradiation in a colorectal cancer model. METHODS: Clonogenicity was measured in cultured cancer cell line HT29 irradiated with 15 mGy/h combined with intermittent hypoxia. Mice with HT29 xenografts were irradiated by repeated injections of [3H]-valine intravenously. The activity in the tumor tissue was measured by scintillation counting and tumor growth, hypoxic fraction and tritium distribution within tumors were assessed by pimonidazole staining and autoradiography. RESULTS: Ultra-low dose-rate irradiation decreased clonogenicity in hypoxic colorectal cancer cells. In vivo, the tumor growth, hypoxic fraction and weight of the mice were similar between the treated and untreated group. Autoradiography showed no [3H]-valine uptake in hypoxic tumor regions in contrast to aerobic tissue. CONCLUSION: Continuous low-dose-rate irradiation was well tolerated by aerobic tissue. This indicates a potential use of low dose-rate irradiation to target hypoxic tumor cells in combination with high dose-rate irradiation to eradicate the well oxygenated tumor regions. However, [3H]-valine is not the appropriate method to deliver ultra-low dose-rate irradiation in vivo.

Separation of two sub-groups with different DNA content after treatment of T-47D breast cancer cells with low dose-rate irradiation and intermittent hypoxia.

Background Previous studies have shown that combined treatment with internal ultra-low dose-rate irradiation selectively inactivated hypoxic T-47D breast cancer cells after three to five weeks of treatment. However, 2-3% of the hypoxic cells were found to survive and restart proliferation upon re-oxygenation. Purpose To investigate the metastatic potential and characteristics of radiosensitivity of these surviving cells, named T - 47 DS. Material and Methods The T - 47 DS cells were grown in ambient air without irradiation. A cloning experiment identified two sub-groups with different DNA content ([Formula: see text] and [Formula: see text]). Furthermore, radiosensitivity and presence of hyper-radiosensitivity (HRS) was measured by Co-60 challenge irradiation and relative migration was determined by scratch assays. Results The two subpopulations of T - 47 DS had different DNA content; one had abnormally high DNA content ([Formula: see text]) and one had DNA content similar to wild-type T-47D cells ([Formula: see text]). HRS was surprisingly present in cells of the cloned population [Formula: see text], but was absent in cells of both [Formula: see text] and T - 47 DS. The radio response of T - 47 DS, [Formula: see text] at higher radiation doses were similar to that of T-47D cells, and neither subpopulation showed increased migration compared with wild-type T-47D. Conclusion No increase in the risk of metastasis was found and only slight changes in radiosensitivity in response to conventional clinical doses was observed. Thus, the data suggest that if ultra-low dose-rate irradiation is used for targeting the hypoxic tumor fraction, conventional high dose-rate irradiation can be used to eradicate eventual surviving cells as well as cells in the well oxygenated areas of the tumor.