CDC6 is a prognostic biomarker and correlated with immune infiltrates in glioma.

BACKGROUND: Cell division cycle 6 (CDC6) has been proven to be associated with the initiation and progression of human multiple tumors. However, it's role in glioma, which is ranked as one of the common primary malignant tumor in the central nervous system and is associated with high morbidity and mortality, is unclear. METHODS: In this study, we explored CDC6 gene expression level in pan-cancer. Furthermore, we focused on the relationships between CDC6 expression, its prognostic value, potential biological functions, and immune infiltrates in glioma patients. We also performed vitro experiments to assess the effect of CDC6 expression on proliferative, apoptotic, migrant and invasive abilities of glioma cells. RESULTS: As a result, CDC6 expression was upregulated in multiple types of cancer, including glioma. Moreover, high expression of CDC6 was significantly associated with age, IDH status, 1p/19q codeletion status, WHO grade and histological type in glioma (all p < 0.05). Meanwhile, high CDC6 expression was associated with poor overall survival (OS) in glioma patients, especially in different clinical subgroups. Furthermore, a univariate Cox analysis showed that high CDC6 expression was correlated with poor OS in glioma patients. Functional enrichment analysis indicated that CDC6 was mainly involved in pathways related to DNA transcription and cytokine activity, and Gene Set Enrichment Analysis (GSEA) revealed that MAPK pathway, P53 pathway and NF-κB pathway in cancer were differentially enriched in glioma patients with high CDC6 expression. Single-sample gene set enrichment analysis (ssGSEA) showed CDC6 expression in glioma was positively correlated with Th2 cells, Macrophages and Eosinophils, and negative correlations with plasmacytoid dendritic cells, CD8 T cells and NK CD56bright cells, suggesting its role in regulating tumor immunity. Finally, CCK8 assay, flow cytometry and transwell assays showed that silencing CDC6 could significantly inhibit proliferation, migration, invasion, and promoted apoptosis of U87 cells and U251 cells (p < 0.05). CONCLUSION: In conclusion, high CDC6 expression may serve as a promising biomarker for prognosis and correlated with immune infiltrates, presenting to be a potential immune therapy target in glioma.

[Automatic Delineation of Clinical Target Volume and Organ at Risk by Deep Learning for Prostate Cancer Adaptive Radiotherapy].

Adaptive radiotherapy can modify the treatment plan online based on the clinical target volume (CTV) and organ at risk (OAR) contours on the cone-beam CT (CBCT) before treatment, improving the accuracy of radiotherapy. However, manual delineation of CTV and OAR on CBCT is time-consuming. In this study, a deep neural network-based method based on U-Net was purposed. CBCT images and corresponding mask were used for model training and validation, showing superior performance in terms of the segmentation accuracy. The proposed method could be used in the clinic to support rapid CTV and OAR contouring for prostate adaptive radiotherapy.

[Application of New Posture Fixation Device Compatible with Magnetic Resonance Simulation Positioning in Head and Neck Radiotherapy].

OBJECTIVE: Whether the developed new type of radiotherapy auxiliary fixation device compatible with the head and neck joint coil can improve the quality of magnetic resonance images in radiotherapy and verify whether it can be applied to clinical treatment. METHODS: The clinical trial selected patients with brain metastases and nasopharyngeal cancer patients, using thermoplastic film and head and shoulder molds for posture fixation, and treatment on the ELekta Versa accelerator. SPSS 20.0 statistical software was used to analyze the data. The measurement data were expressed by X±S, and the t test was used. P<0.05 was considered as statistically significant. RESULTS: Considering the influence of the outer contour of the device, the target dose meets the clinical requirements. The setting error is less than 2 mm in the three translation directions, and the rotation error is less than 2o in the three rotation directions. CONCLUSIONS: There is no statistical difference between the treatment results of patients using the new type of fixation device and the conventional method. The target area threatens the organ dose, and the positioning error meets the treatment requirements.

[Improved Design and Acceptance Test of Localization Couch for New Type of CT Simulator].

Accurate CT simulation is the key link of precision radiotherapy, and the performance of the localization couch of CT simulator directly affects the accuracy of radiotherapy. With the rapid development of precision radiotherapy, conventional large aperture radiotherapy special CT simulator is difficult to meet the needs of precision radiotherapy localization, so most radiotherapy centers choose high-end diagnostic CT machines equipped with a flat tabletop for radiotherapy localization. In clinical work, the performance testing of the CT simulator localization couch is easy to be ignored. In addition, there are some problems such as insufficient precision in transforming the cradle-shaped couch top of diagnostic CT into a special flat couch top for radiotherapy. This paper provided an in-depth description of the improved design and performance test of the localization couch of the first special GE Revolution CT simulator for radiotherapy introduced by West China Hospital of Sichuan University. After the improvement, all the acceptance tests of the localization couch are in line with the standard, and the performance meets the high-precision radiotherapy localization needs of patients with different body weight in the center.

[Application of Statistical Process Control in Evaluation of Performance for Beam-matched Medical Electron Linacs].

Clinically, beam matching can greatly improve the flexibility and efficiency of treating patients between different medical electron linacs. However, in addition to the regular quality assurance (QA) test of the machine performance of linacs, there is still a lack of comprehensive evaluation of the clinical radiotherapy performance of beam-matched linacs. In this paper, the performance of volumetric modulated arc therapy (VMAT) between three closely matched linacs was evaluated by statistical process control (SPC) technology. It was found that the average and median γ passing rates of the VMAT QA processes of the three linacs had little difference, but the process capability levels were at three different levels. The results show that SPC technology can effectively evaluate the performance of beam matching for medical electron linacs, improve the patient-specific VMAT QA processes, and guide clinical decision-making.

[Application of Virtual Monochromatic Images Reconstructed by Dual-energy Computed Tomography in Radiotherapy Treatment Planning System].

Virtual monochromatic images (VMI) that reconstructed on dual-energy computed tomography (DECT) have further application prospects in radiotherapy, and there is still a lack of clinical dose verification. In this study, GE Revolution CT scanner was used to perform conventional imaging and gemstone spectral imaging on the simulated head and body phantom. The CT images were imported to radiotherapy treatment planning system (TPS), and the same treatment plans were transplanted to compare the CT value and the dose distribution. The results show that the VMI can be imported into TPS for CT value-relative electron density conversion and dose calculation. Compared to conventional images, the VMI varies from 70 to 140 keV, has little difference in dose distribution of 6 MV photon treatment plan.

[Progress on meiotic gene expression and epigenetic regulation of male sterility in Dzo cattle].

Interspecific hybrid male sterility is a common occurrence in nature and plays an important role in species reproductive isolation. Dzo (cattle-yak), the offspring of interspecific cross between domestic yak (Bos grunniens) and cattle (Bos taurus), is a unique animal model for investigating interspecific hybrid male sterility. Dzo females are completely fertile while the males are sterile. In recent years, molecular studies have demonstrated that the expressions of genes were dysregulated during meiosis in Dzo testis, as compared to those in cattle or yak. Other studies have revealed that epigenetic factors/events, such as DNA methylation, histone modification and non-coding RNA, are also involved in spermatogenesis. This review summarizes the dysregulation of gene expression, DNA methylation, microRNA (miRNA), PIWI-interacting RNA (piRNA), long non-coding RNA (lncRNA), and histone methylation modification during meiosis in Dzo testis. These results highlighted the potential roles of genetic and epigenetic regulations of meiosis in Dzo testis, thereby providing a more detailed understanding on the molecular mechanisms of interspecific hybrid male sterility.

[Accuracy and Influence Analysis of 4D CBCT Automatic Registration Algorithm under the Guidance of Chest Tumor Image].

OBJECTIVE: In order to provide guidance for clinical use of four-dimensional cone-beam CT (4D CBCT), the accuracy of image registration and its influencing factors were analyzed using the automatic registration method when 4D CBCT was used as an image guidance strategy for patients with chest tumors. METHODS: The respiratory motion model and two kinds of lung plug-ins were used to simulate two types of tumors and their movements in the chest. 4D CT was scanned for each kind of simulated tumor, and 4D CBCT was scanned under various artificial positioning errors. For the registration of 4D CBCT, the manual and automatic registration methods were used for each group. RESULTS: There were more obvious mismatches in the intrapulmonary adhesion tumor group. When the masks were created based on the size of the target area or expanding the target area by 0.5 cm, the results between the automatic registration and manual registration were statistically different. There were no significant mismatches in the isolated lung tumor group, and there was no statistical difference between the results of automatic registration and manual registration. CONCLUSIONS: When 4D CBCT is used as an image guidance strategy for patients with chest tumors, the automatic registration procedure should not be used for tumors adhering to chest wall and mediastinum. For solitary lung tumors, the automatic registration method and the manual registration method have similar registration accuracy, but significant mismatches need to be excluded.

[Preliminary study on monitoring patient-specific volumetric modulated arc therapy quality assurance process with statistical process control methodology on the basis of TG-218 report].

Patient-specific volumetric modulated arc therapy (VMAT) quality assurance (QA) process is an important component of the implementation process of clinical radiotherapy. The tolerance limit and action limit of discrepancies between the calculated dose and the delivered radiation dose are the key parts of the VMAT QA processes as recognized by the AAPM TG-218 report, however, there is no unified standard for these two values among radiotherapy centers. In this study, based on the operational recommendations given in the AAPM TG-218 report, treatment site-specific tolerance limits and action limits of gamma pass rate in VMAT QA processes when using ArcCHECK for dose verification were established by statistical process control (SPC) methodology. The tolerance limit and action limit were calculated based on the first 25 in-control VMAT QA for each site. The individual control charts were drawn to continuously monitor the VMAT QA process with 287 VMAT plans and analyze the causes of VMAT QA out of control. The tolerance limits for brain, head and neck, abdomen and pelvic VMAT QA processes were 94.56%, 94.68%, 94.34%, and 92.97%, respectively, and the action limits were 93.82%, 92.54%, 93.23%, and 90.29%, respectively. Except for pelvic, the tolerance limits for the brain, head and neck, and abdomen were close to the universal tolerance limit of TG-218 (95%), and the action limits for all sites were higher than the universal action limit of TG-218 (90%). The out-of-control VMAT QAs were detected by the individual control chart, including one case of head and neck, two of the abdomen and two of the pelvic site. Four of them were affected by the setup error, and one was affected by the calibration of ArcCHECK. The results show that the SPC methodology can effectively monitor the IMRT/VMAT QA processes. Setting treatment site-specific tolerance limits is helpful to investigate the cause of out-of-control VMAT QA.

[Feasibility Research of the New Fixation Device Compatible with Head and Neck Coil of MRI for Radiotherapy].

MRI simulation images quality of head and neck coil scanning is better than that of radiotherapy surface coil, but currently the head and neck coil is not compatible with radiotherapy positioning devices. In this paper, a new fixation device is developed based on computer reverse engineering technology, which can be used in combination with head and neck coil. This article focuses on discussing the feasibility of the new device in radiotherapy. The obtained ACR phantom and Cat phantom 504 images were used to analyze MR and CT images quality assurance indicators. The dose attenuation of 6 MV photons was measured using the ionization chamber. The results showed each index met the clinical application requirements of intracranial tumor radiotherapy, thereby it can be used in intracranial tumor radiotherapy.

[A review of progress of real-time tumor tracking radiotherapy technology based on dynamic multi-leaf collimator].

While radiation treatment to patients with tumors in thorax and abdomen is being performed, further improvement of radiation accuracy is restricted by the tumor intra-fractional motion due to respiration. Real-time tumor tracking radiation is an optimal solution to tumor intra-fractional motion. A review of the progress of real-time dynamic multi-leaf collimator(DMLC) tracking is provided in the present review, including DMLC tracking method, time lag of DMLC tracking system, and dosimetric verification.

Long-term follow-up results of simultaneous integrated or late course accelerated boost with external beam radiotherapy to vaginal cuff for high risk cervical cancer patients after radical hysterectomy.

BACKGROUND: To assess the safety and efficacy of simultaneous integrated boost (SIB) or late course accelerated boost (LCAB) with external beam radiotherapy (EBRT) to the vaginal cuff for high risk cervical cancer patients after radical hysterectomy. METHODS: Between October 2009 and January 2012, patients with high risk cervical cancer who had undergone radical surgery followed by EBRT to the vaginal cuff were enrolled. Patients were treated with either intensity modulated radiotherapy (IMRT)/volumetric modulated arc therapy (VMAT) with SIB (arm A) or IMRT/VMAT to the pelvis followed by LCAB (arm B) to vaginal cuff. In arm A, the pelvic and boost doses were 50.4 Gy and 60.2 Gy in 28 fractions, respectively. In arm B, pelvic irradiation to 50 Gy in 25 fractions followed by a boost of 9 Gy in 3 fractions were delivered. Chemotherapy was given concurrently. RESULTS: Overall, 80 patients were analyzed in this study (42 in arm A, 38 in arm B). In arm A and B, median follow-up was 37 and 32 months, respectively. The 3-year disease-free survival and overall survival in arms A vs B were 88.7% vs. 93.4% (p = 0.89), and 91.8% vs.100% (p = 0.21), respectively. The 3-year local-regional control and distant failure were 97.6% vs. 100% (p = 0.34), and 4.8% vs. 5.3% (p = 0.92), respectively. Grade 3-4 acute leukopenia and dermatitis were seen in 11 (26.2%) and 8 (19.0%) patients in Arm A, vs. 7 (17.8%) and 6 (15.8%) patients in Arm B, respectively (p > 0.05). Only Grade 1-2 chronic gastrointestinal (GI) and genitourinary (GU) toxicities were observed. CONCLUSIONS: Our results indicate that both SIB and LCAB to vaginal cuff for high risk cervical cancer patients after radical hysterectomy are associated with excellent survival, local control and low toxicity.

Stereotactic body radiation therapy for post-pulmonary lobectomy isolated lung metastasis of thoracic tumor: survival and side effects.

BACKGROUND: Stereotactic body radiation therapy (SBRT) has emerged as an alternative treatment for patients with early stage non-small cell lung cancer (NSCLC) or metastatic pulmonary tumors. However, for isolated lung metastasis (ILM) of thoracic malignances after pulmonary lobectomy, reported outcomes of SBRT have been limited. This study evaluates the role of SBRT in the treatment of such patients. METHODS: A retrospective search of the SBRT database was conducted in three hospitals. The parameters analyzed in the treated patients were local control, progression-free survival (PFS), overall survival (OS), and the treatment-related side-effects. RESULTS: In total, 23 patients with single ILM after pulmonary lobectomy treated with SBRT were identified and the median follow-up time was 14 months (range: 6.0-47.0 months). Local recurrences were observed in two patients during follow-up and the 1-year local control rate was 91.3%. Median PFS and OS for the studied cohort were 10.0 months [95% confidence interval (CI) 5.1-14.9 months] and 21.0 months (95% CI 11.4-30.6 months), respectively. Acute radiation pneumonitis (RP) of grade 2 or worse was observed in five (21.7%) and three (13.0%) patients, respectively. Other treatment-related toxicities included chest wall pain in one patient (4.3%) and acute esophagitis in two patients (8.7%). By Pearson correlation analysis, the planning target volume (PTV) volume and the volume of the ipsilateral lung exposed to a minimum dose of 5 Gy (IpV5) were significantly related to the acute RP of grade 2 or worse in present study (p < 0.05). The optimal thresholds of the PTV and IpV5 to predict RP of acute grade 2 or worse RP were 59 cm3 and 51% respectively, according to the receiver-operating characteristics curve analysis, with sensitivity/specificity of 75.0%/80.0% and 62.5%/80.0%. CONCLUSIONS: SBRT for post-lobectomy ILM was effective and well tolerated. The major reason for disease progression was distant failure but not local recurrence. The PTV and IpV5 are potential predictors of acute RP of grade 2 or higher and should be considered in treatment planning for such patients.

[Effect of respiratory movement on cone beam computed tomography images].

To investigate the impact of respiratory movement to determine the target volume on cone beam CT (CBCT) for lung tumor, we used CIRS dynamic thorax phantom (Model-CIRS008) to simulate the sinusoidal motion of lung tumor. With a constant amplitude, the ratio of the time of near-end-expiratory and near-end-inspiratory (E/I) changed when it was scanned with CBCT. We analyzed the contrast changes of target by extracting the CT value of each pixel on the center line of the target movement direction. The targets were contoured with region growing method and compared with the motion volume generated by the tumor trajectory method. The result showed that the contrast of near-end-expiratory increased and the contrast of near-end-inspiratory decreased with increasing E/I. The contoured volume generated by region growing method decreased with increasing E/I. When E/I = 4, the amplitude A = 1 cm, diameter of 1 cm and 3 cm target volumes were reduced by 48.2% and 22.7%. The study showed that CBCT was not suitable to be used to accurately determine the range of lung tumor movement. The internal target volume (ITV) may be underestimated in CBCT images.

Causal analysis of radiotherapy safety incidents based on a hybrid model of HFACS and Bayesian network.

Objective: This research investigates the role of human factors of all hierarchical levels in radiotherapy safety incidents and examines their interconnections. Methods: Utilizing the human factor analysis and classification system (HFACS) and Bayesian network (BN) methodologies, we created a BN-HFACS model to comprehensively analyze human factors, integrating the hierarchical structure. We examined 81 radiotherapy incidents from the radiation oncology incident learning system (RO-ILS), conducting a qualitative analysis using HFACS. Subsequently, parametric learning was applied to the derived data, and the prior probabilities of human factors were calculated at each BN-HFACS model level. Finally, a sensitivity analysis was conducted to identify the human factors with the greatest influence on unsafe acts. Results: The majority of safety incidents reported on RO-ILS were traced back to the treatment planning phase, with skill errors and habitual violations being the primary unsafe acts causing these incidents. The sensitivity analysis highlighted that the condition of the operators, personnel factors, and environmental factors significantly influenced the occurrence of incidents. Additionally, it underscored the importance of organizational climate and organizational process in triggering unsafe acts. Conclusion: Our findings suggest a strong association between upper-level human factors and unsafe acts among radiotherapy incidents in RO-ILS. To enhance radiation therapy safety and reduce incidents, interventions targeting these key factors are recommended.

Evaluation of the sensitivity of two 3D diode array dosimetry systems to setup error for quality assurance (QA) of volumetric-modulated arc therapy (VMAT).

The purpose of this study is to evaluate the sensitivities of 3D diode arrays to setup error for patient-specific quality assurance (QA) of volumetric-modulated arc therapy (VMAT). Translational setup errors of ± 1, ± 2, and ± 3 mm in the RL, SI, and AP directions and rotational setup errors of ± 1° and ± 2° in the pitch, roll, and yaw directions were set up in two phantom systems, ArcCHECK and Delta4, with VMAT plans for 11 patients. Cone-beam computed tomography (CBCT) followed by automatic correction using a HexaPOD 6D treatment couch ensured the position accuracy. Dose distributions of the two phantoms were compared in order to evaluate the agreement between calculated and measured values by using γ analysis with 3%/3 mm, 3%/2 mm, and 2%/2 mm criteria. To determine the impact on setup error for VMAT QA, we evaluated the sensitivity of results acquired by both 3D diode array systems to setup errors in translation and rotation. For the VMAT QA of all patients, the pass rate with the 3%/3 mm criteria exceeded 95% using either phantom. For setup errors of 3 mm and 2°, respectively, the pass rates with the 3%/3mm criteria decreased by a maximum of 14.0% and 23.5% using ArcCHECK, and 14.4% and 5.0% using Delta4. Both systems are sensitive to setup error, and do not have mechanisms to account for setup errors in the software. The sensitivity of both VMAT QA systems was strongly dependent on the patient-specific plan. The sensitivity of ArcCHECK to the rotational error was higher than that of Delta4. In order to achieve less than 3% mean pass rate reduction of VMAT plan QA with the 3%/3 mm criteria, a setup accuracy of 2 mm/1° and 2 mm/2° is required for ArcCheck and Delta4 devices, respectively. The cumulative effect of the combined 2 mm translational and 1° rotational errors caused 3.8% and 2.4% mean pass rates reduction with 3%/3 mm criteria, respectively, for ArcCHECK and Delta4 systems. For QA of VMAT plans for nasopharyngeal cancer (NPC) using the ArcCHECK system, the setup should be more accurate.

[Primary study of the threatening of unfixed planning of image guided radiotherapy to the volume margin of neck tumor].

Some patients who have neck tumor but cannot tolerate the thermoplastic immobilization may be supported by simple cushions, and are marked on the neck skin during CT simulation. We therefore set 5 mm as the spinal cord-planning risk volume margin in the intensity-modulated radiotherapy plans in our Centre. Cone beam CT (CBCT) scans were acquired for three times, and matched with the simulation CT images in each radiotherapy. The mean and the standard deviation of the individual, the root mean-square and the standard deviation of the individual were calculated. The matched results of the third CBCT were used to calculate the spinal cord- planning risk volume margin. The results showed that the interfraction error was significantly reduced and the intrafraction error was stable by CBCT guiding. CBCT and 5 mm spinal cord-planning organ is feasible and safe without threatening volume margin to high dose radiotherapy for the patients with neck tumor and not able to tolerate thermoplastic immobilization.

A Script-Based Automatic Intensity Modulated Radiation Therapy Planning Method With Robust Optimization for Craniospinal Irradiation.

This report describes a script-based automatic planning method with robust optimization for craniospinal irradiation (CSI) to reduce sensitivity to field matching errors and increase planning efficiency. The data of 10 CSI patients with planning target volume (PTV) lengths between 49.8 and 85.0 cm were retrospectively studied. Robust intensity modulated radiation therapy plans with ±5-mm longitudinal position uncertainty were generated by the automatic planning script. A simple dose prediction model and a self-adjusting method were implied in the automatic plans. The plans' robustness against setup errors was evaluated by deliberately shifting the middle beamset ±5 mm in the superior-inferior direction. Manual and nonrobust plans were also created to evaluate the automatic robust plans' quality, efficiency, and robustness. There were no significant differences between the manual and automatic plans in terms of homogeneity index; conformity index; D1%, D2%, and D98% of PTV; and average doses of organs at risk. However, the D99% of the PTV in the automatic plans was slightly inferior to that in the manual plans. Compared with the manual plans, the automatic plans greatly increased efficiency, with a reduction in planning time of approximately 48%. When ±5-mm superior-inferior errors were introduced, the average deviations of the maximum dose D1% and minimum dose D99% to the spinal cord were 4.9% (±1.1%) and -3.4% (±1.3%), respectively. However, the corresponding values of the nonrobust plans were 20.0% (±5.4%) and -21.2 (±6.3%), respectively. The script-based automatic CSI planning method, combining robust optimization and a dose prediction model, efficiently created a good-quality plan that was robust to setup errors.

Clinical evaluation on automatic segmentation results of convolutional neural networks in rectal cancer radiotherapy.

Purpose: Image segmentation can be time-consuming and lacks consistency between different oncologists, which is essential in conformal radiotherapy techniques. We aimed to evaluate automatic delineation results generated by convolutional neural networks (CNNs) from geometry and dosimetry perspectives and explore the reliability of these segmentation tools in rectal cancer. Methods: Forty-seven rectal cancer cases treated from February 2018 to April 2019 were randomly collected retrospectively in our cancer center. The oncologists delineated regions of interest (ROIs) on planning CT images as the ground truth, including clinical target volume (CTV), bladder, small intestine, and femoral heads. The corresponding automatic segmentation results were generated by DeepLabv3+ and ResUNet, and we also used Atlas-Based Autosegmentation (ABAS) software for comparison. The geometry evaluation was carried out using the volumetric Dice similarity coefficient (DSC) and surface DSC, and critical dose parameters were assessed based on replanning optimized by clinically approved or automatically generated CTVs and organs at risk (OARs), i.e., the Planref and Plantest. Pearson test was used to explore the correlation between geometric metrics and dose parameters. Results: In geometric evaluation, DeepLabv3+ performed better in DCS metrics for the CTV (volumetric DSC, mean = 0.96, P< 0.01; surface DSC, mean = 0.78, P< 0.01) and small intestine (volumetric DSC, mean = 0.91, P< 0.01; surface DSC, mean = 0.62, P< 0.01), ResUNet had advantages in volumetric DSC of the bladder (mean = 0.97, P< 0.05). For critical dose parameters analysis between Planref and Plantest, there was a significant difference for target volumes (P< 0.01), and no significant difference was found for the ResUNet-generated small intestine (P > 0.05). For the correlation test, a negative correlation was found between DSC metrics (volumetric, surface DSC) and dosimetric parameters (δD95, δD95, HI, CI) for target volumes (P< 0.05), and no significant correlation was found for most tests of OARs (P > 0.05). Conclusions: CNNs show remarkable repeatability and time-saving in automatic segmentation, and their accuracy also has a certain potential in clinical practice. Meanwhile, clinical aspects, such as dose distribution, may need to be considered when comparing the performance of auto-segmentation methods.

Predictive value of magnetic resonance imaging radiomics-based machine learning for disease progression in patients with high-grade glioma.

Background: Accurately predicting the prognosis of patients with high-grade glioma (HGG) is potentially important for treatment. However, the predictive value of images of various magnetic resonance imaging (MRI) sequences for prognosis at different time points is unknown. We established predictive machine learning models of HGG disease progression and recurrence using MRI radiomics and explored the factors influencing prediction accuracy. Methods: Radiomics features were extracted from T1-weighted (T1WI), contrast-enhanced T1-weighted (CE-T1WI), T2-weighted (T2WI), and fluid-attenuated inversion recovery (FLAIR) images (postoperative radiotherapy planning MRI images) obtained from 162 patients with HGG. The Mann-Whitney U test and least absolute shrinkage and selection operator (LASSO) algorithm were used for feature selection. Machine learning models were used to build prediction models to estimate disease progression or recurrence. The influence of different MRI sequences, regions of interest (ROIs), and prediction time points was also explored. The receiver operating characteristic (ROC) curve was used to evaluate the discriminative performance of each model, and the DeLong test was employed to compare the ROC curves. Results: Radiomics features from T2WI and FLAIR demonstrated greater predictive value for disease progression compared with T1WI or CE-TIWI. The best predictive models, with areas under the ROC curves (AUCs) of 0.70, 0.68, 0.78, 0.78, and 0.78 for predicting disease progression at the 6th, 9th, 12th, 15th, and 18th month after radiotherapy, respectively, were obtained by combining clinical features with gross tumor volume (GTV) and clinical target volume (CTV) features extracted from T2WI and FLAIR. Conclusions: Structural MRI obtained before radiotherapy can be used to predict the disease progression or posttreatment recurrence of HGG. When using MRI radiomics to predict long-term outcomes as opposed to short-term outcomes, better predictive results may be obtained.