We report a new technique of brachytherapy consisting of intracavitary combined with computed tomography (CT)-guided interstitial brachytherapy for locally advanced cervical cancer. A Fletcher-Suit applicator and trocar point needles were used for performing high-dose rate brachytherapy under in-room CT guidance. First, a tandem and ovoids were implanted into the patient's vagina and uterus by conventional brachytherapy method. Based on clinical examination and MRI/CT imaging, operating radiation oncologists decided the positions of insertion in the tumor and the depth of the needles from the upper surface of the ovoid. Insertion of the needle applicator was performed from the vaginal vault inside the ovoid within the tumor under CT guidance. In treatment planning, dwell positions and time adaptations within the tandem and ovoids were performed first for optimization based on the Manchester system, and then stepwise addition of dwell positions within the needle was continued. Finally, dwell positions and dwell weights were manually modified until dose-volume constraints were optimally matched. In our pilot case, the dose of D90 to high-risk clinical target volume was improved from 3.5 Gy to 6.1 Gy by using our hybrid method on the dose-volume histogram. D1cc of the rectum, bladder and sigmoid colon by our hybrid method was 4.8 Gy, 6.4 Gy and 3.5 Gy, respectively. This method consists of advanced image-guided brachytherapy that can be performed safely and accurately. This approach has the potential of increasing target coverage, treated volume, and total dose without increasing the dose to organs at risk.

Background: Carbon-ion radiotherapy (CIRT) for prostate cancer was initiated at Kanagawa Cancer Center in 2015. The present study analyzed the preliminary clinical outcomes of CIRT for prostate cancer. Methods: The clinical outcomes of 253 patients with prostate cancer who were treated with CIRT delivered using the spot scanning method between December 2015 and December 2017 were retrospectively analyzed. The irradiation dose was set at 51.6 Gy (relative biological effectiveness) delivered in 12 fractions over 3 weeks. Biochemical relapse was defined using the Phoenix definition. Toxicities were assessed according to CTCAE version 4.0. Results: The median patient age was 70 (47-86) years. The median follow-up duration was 35.3 (4.1-52.9) months. According to the D'Amico classification system, 8, 88, and 157 patients were classified as having low, intermediate, and high risks, respectively. Androgen deprivation therapy was administered in 244 patients. The biochemical relapse-free rate in the low-, intermediate-, and high-risk groups at 3 years was 87.5, 88.0, and 97.5%, respectively (P = 0.036). Grade 2 acute urinary toxicity was observed in 12 (4.7%) patients. Grade 2 acute rectal toxicity was not observed. Grade 2 late urinary toxicity and grade 2 late rectal toxicity were observed in 17 (6.7%) and 3 patients (1.2%), respectively. Previous transurethral resection of the prostate was significantly associated with late grade 2 toxicity in univariate analysis. The predictive factor for late rectal toxicity was not detected. Conclusion: The present study demonstrated that CIRT using the spot scanning method for prostate cancer produces favorable outcomes.

BackgroundThe reliability of DeepL Translator (DeepL GmbH, Cologne, Germany) for the translation for medical articles has not been verified yet. In this study, we investigated the accuracy of machine translation from Japanese to English for a medical article using the DeepL Translator. MethodologyThe subject of this study was an English-language medical article translated from Japanese, which had already been published. The original Japanese manuscript was translated into English using DeepL Translator. The translated English article was then back-translated into Japanese by three researchers. In turn, three other researchers compared the back-translated Japanese sentences with the original Japanese manuscript and calculated the percentage of sentences that retained the intended meaning. ResultsThe mean ± standard deviation of the match rate for the entire article was 94.0 ± 2.9%. The match rate in the Results section was significantly higher than that in the other sections; while the match rate in the Materials and Methods section was significantly lower than the rate in the other sections. Compound sentences and sentences with an unclear subject and predicate appeared to be significant predictors for mismatched translation. ConclusionsThe translation for a medical article from Japanese to English was performed accurately by DeepL Translator.

The aim of this study was to assess the feasibility of planning dose–volume histogram (DVH) parameters in computed tomography-based 3D image-guided brachytherapy for locally advanced cervical cancer. In a prospective multi-institutional study, 60 patients with stage IIA2–IVA cervical cancer from eight institutions were treated with external beam radiotherapy using central shielding and intracavitary or hybrid (combined intracavitary/interstitial) brachytherapy (HBT). The dose constraints were set as a cumulative linear quadratic equivalent dose (EQD2) of at least 60 Gy for high-risk clinical target volume (HR-CTV) D90, D2cc ≤ 75 Gy for rectum, D2cc ≤ 90 Gy for bladder and D2cc ≤ 75 Gy for sigmoid. The median HR-CTV D90 was 70.0 Gy (range, 62.8–83.7 Gy) in EQD2. The median D2cc of rectum, bladder and sigmoid was 57.1 Gy (range, 39.8–72.1 Gy), 68.9 Gy (range, 46.5–84.9 Gy) and 57.2 Gy (range, 39.2–71.2 Gy) in EQD2, respectively. In 76 of 233 sessions (33%), 23 patients underwent HBT, and the median number of interstitial needles was 2 (range, 1–5). HBT for a bulky HR-CTV (≥40 cm3) significantly improved the HR-CTV D90 compared with intracavitary brachytherapy alone (P = 0.010). All patients fulfilled the dose constrains for target and at risk organs by undergoing HBT in one-third of sessions. We conclude that the planning DVH parameters used in our protocol are clinically feasible.

The purpose of this study is to evaluate the prediction and classification performances of the gamma passing rate (GPR) for different machine learning models and to select the best model for achieving machine learningbased patient-specific quality assurance (PSQA). Methods: The measurement verification of 356 head-and-neck volumetric modulated arc therapy plans was performed using a diode array phantom (Delta4 Phantom), and GPR values at 2%/2 mm with global normalization and 3%/2 mm with local normalization were calculated. Machine learning models, including ridge regression (RIDGE), random forest (RF), support vector regression (SVR), and stacked generalization (STACKING), were used to predict the GPR. Each machine learning model was trained using 260 plans, and the prediction accuracy was evaluated using the remaining 96 plans. The prediction error between the measured and predicted GPR was evaluated. For the classification evaluation, the lower control limit for the measured GPR and lower control limit for predicted GPR (LCL p ) was defined to identify whether the GPR values represent a "pass" or a "fail." LCL p values with 99% and 99.9% confidence levels were calculated as the upper prediction limits for the GPR estimated from the linear regression between the measured and predicted GPR. Results: There was an overestimation trend of the low measured GPR. The maximum prediction errors for RIDGE, RF, SVR, and STACKING were 3.2%, 2.9%, 2.3%, and 2.2% at the global 2%/2 mm and 6.3%, 6.6%, 6.1%, and 5.5% at the local 3%/2 mm, respectively. In the global 2%/2 mm, the sensitivity was 100% for all the machine learning models except RIDGE when using 99% LCL p . The specificity was 76.1% for RIDGE, RF, and SVR and 66.3% for STACKING; however, the specificity decreased dramatically when 99.9% LCL p was used. In the local 3%/2 mm, however, only STACKING showed 100% sensitivity when using 99% LCL p . The decrease in the specificity using 99.9% LCL p was smaller than that in the global 2%/2 mm, and the specificity for RIDGE, RF, SVR, and STACKING was 61.3%, 61.3%, 72.0%, and 66.8%, respectively. Conclusions: STACKING had better prediction accuracy for low GPR values than other machine learning models. Applying LCL p to a regression model enabled the consistent evaluation of quantitative and qualitative GPR predictions. Adjusting the confidence level of the LCL p helped improve the balance between the sensitivity and specificity. We suggest that STACKING can assist the safe and efficient operation of PSQA.