Automatic gross tumor volume segmentation with failure detection for safe implementation in locally advanced cervical cancer.

Background and Purpose: Automatic segmentation methods have greatly changed the RadioTherapy (RT) workflow, but still need to be extended to target volumes. In this paper, Deep Learning (DL) models were compared for Gross Tumor Volume (GTV) segmentation in locally advanced cervical cancer, and a novel investigation into failure detection was introduced by utilizing radiomic features. Methods and materials: We trained eight DL models (UNet, VNet, SegResNet, SegResNetVAE) for 2D and 3D segmentation. Ensembling individually trained models during cross-validation generated the final segmentation. To detect failures, binary classifiers were trained using radiomic features extracted from segmented GTVs as inputs, aiming to classify contours based on whether their Dice Similarity Coefficient ( DSC ) < T and DSC ⩾ T . Two distinct cohorts of T2-Weighted (T2W) pre-RT MR images captured in 2D sequences were used: one retrospective cohort consisting of 115 LACC patients from 30 scanners, and the other prospective cohort, comprising 51 patients from 7 scanners, used for testing. Results: Segmentation by 2D-SegResNet achieved the best DSC, Surface DSC ( SDSC 3 mm ), and 95th Hausdorff Distance (95HD): DSC = 0.72 ± 0.16, SDSC 3 mm =0.66 ± 0.17, and 95HD = 14.6 ± 9.0 mm without missing segmentation ( M =0) on the test cohort. Failure detection could generate precision ( P = 0.88 ), recall ( R = 0.75 ), F1-score ( F = 0.81 ), and accuracy ( A = 0.86 ) using Logistic Regression (LR) classifier on the test cohort with a threshold T = 0.67 on DSC values. Conclusions: Our study revealed that segmentation accuracy varies slightly among different DL methods, with 2D networks outperforming 3D networks in 2D MRI sequences. Doctors found the time-saving aspect advantageous. The proposed failure detection could guide doctors in sensitive cases.

Image-guided adaptive brachytherapy for advanced cervical cancer spreading to the bladder and/or rectum: Clinical outcome and prognostic factors.

INTRODUCTION: Refinements of brachytherapy techniques have led to better local control of locally advanced cervical cancer (LACC), especially with the development of image-guided adaptive brachytherapy (IGABT). Data on the efficacy of brachytherapy in cervical cancer spreading to adjacent organs are scarce. We report the experience of our institution in the treatment of these advanced tumors with IGABT. MATERIALS AND METHODS: Medical records of patients treated for a LACC spreading to the bladder and/or rectum between 2006 and 2020 at Gustave Roussy Institute were analyzed. Dosimetric parameters were collected and converted into 2 Gy per fraction equivalent doses, including the minimal dose received by 90% of the high-risk target volume (D90 CTVHR) and intermediate-risk target volume (D90 CTVIR), as well as the dose received by the most exposed 2 cm3 of the organs at risk. A Cox regression model was used to study the potential associations between clinical and dosimetric factors with survival endpoints and fistula formation. RESULTS AND STATISTICAL ANALYSIS: A total of 81 patients were identified. All patients received pelvic+/- para-aortic radiotherapy, 45 Gy in 25 fractions +/- boost to gross lymph nodes. Concomitant platinum-based chemotherapy was administered in 93.8% of cases. The median D90 CTVHR dose was 75.5 GyEQD2 (SD: 10.39 GyEQD2) and median CTVHR volume was 47.6 cm3 (SD: 27.9 cm3). Median bladder and rectal D2cm3 dose were 75.04 GyEQD2 (SD: 8.72 GyEQD2) and 64.07 GyEQD2 (SD: 6.68 GyEQD2). After a median follow-up of 27.62 ± 25.10 months, recurrence was found in 34/81 patients (42%). Metastatic failure was the most common pattern of relapse (n = 25). Use of a combined interstitial/intracavitary technique and D90 CTVHR ≥ 75.1 GyEQD2 were prognostic factors for OS in univariate analysis (HR = 0.24, 95%IC: 0.057-1, p = 0.023; HR = 0.2, 95%IC: 0.059-0.68, p = 0.0025, respectively). In multivariate analysis, a D90 CTVHR ≥ 75.1 GyEQD2 was significant for OS (HR = 0.23; 95%IC: 0.07, 0.78, p = 0.018). The occurrence of vesicovaginal fistula (VVF) was the most frequent pattern of local recurrence (HR = 4.6, 95%CI: 1.5-14, p = 0.01). CONCLUSION: Advances in brachytherapy modalities improved local control and survival while reducing toxicities. Enhancing local control through dose escalation and combined intracavitary/interstitial brachytherapy techniques is a major factor in patients cure probability, together with systemic intensification to better control distant events.

[A case of bilateral gestational gigantomastia]. / Un cas de gigantomastie gravidique bilatérale.

We here report the case of a 29-year old gravida 2, para 2 patient with no particular past medical history. Symptoms evolved over 2 months and were marked by bilateral breast growth impairing her daily activities. Clinical examination showed hypertrophied breasts and bilateral breast ulcers. She had a history of 28-weeks amenorrhea. Anatomopathological examination of ulcers showed fleshy bud-like tissue. The patient had high levels of prolactin (1345 µUI/ml). The levels of FSH and LH were normal. The patient underwent bromocriptine therapy without success. Patient's evolution was marked by decrease in size and regression in skin ulcers six months after vaginal birth. Gestational gigantomastia is a breast hypertrophy characterized by a breast volume exceeding 1500 cm3 . Its cause is unknown. Radical treatment is based on bilateral mastectomy.