Incidence of occult uterine sarcoma and other unexpected pathologies in patients having surgery for presumed myomas: A retrospective observational study

The automatic diagnosis artificial intelligence system for preoperative magnetic resonance imaging of uterine sarcoma

OBJECTIVE

Magnetic resonance imaging (MRI) is efficient for the diagnosis of preoperative uterine sarcoma; however, misdiagnoses may occur. In this study, we developed a new artificial intelligence (AI) system to overcome the limitations of requiring specialists to manually process datasets and a large amount of computer resources.

METHODS

The AI system comprises a tumor image filter, which extracts MRI slices containing tumors, and sarcoma evaluator, which diagnoses uterine sarcomas. We used 15 types of MRI patient sequences to train deep neural network (DNN) models used by tumor filter and sarcoma evaluator with 8 cross-validation sets. We implemented tumor filter and sarcoma evaluator using ensemble prediction technique with 9 DNN models. Ten tumor filters and sarcoma evaluator sets were developed to evaluate fluctuation accuracy. Finally, AutoDiag-AI was used to evaluate the new validation dataset, including 8 cases of sarcomas and 24 leiomyomas.

RESULTS

Tumor image filter and sarcoma evaluator accuracies were 92.68% and 90.50%, respectively. AutoDiag-AI with the original dataset accuracy was 89.32%, with 90.47% sensitivity and 88.95% specificity, whereas AutoDiag-AI with the new validation dataset accuracy was 92.44%, with 92.25% sensitivity and 92.50% specificity.

CONCLUSION

Our newly established AI system automatically extracts tumor sites from MRI images and diagnoses them as uterine sarcomas without human intervention. Its accuracy is comparable to that of a radiologist. With further validation, the system could be applied for diagnosis of other diseases. Further improvement of the system's accuracy may enable its clinical application in the future.

Gynecological sarcomas, surgical management: primary, metastatic, and recurrent disease

Adult-type gynecological soft tissue and visceral sarcomas are rare tumors, with an estimated incidence of 13% of all sarcomas and 4% of all gynecological malignancies. They most often develop in the uterus (83%), followed by the ovaries (8%), vulva and vagina (5%), and other gynecological organs (2%). The objective of this review is to provide an overview of the current management of gynecological sarcomas, according to international guidelines. The management of gynecological sarcomas should follow the recommendations for the management of soft tissue and visceral sarcomas. Centralizing cases in expert centers improves patient survival, both for the diagnostic phase and for multidisciplinary therapeutic management. In the case of pelvic soft tissue sarcomas, a radiological biopsy is essential before any surgical decision is taken. In the case of a myometrial tumour which may correspond to a sarcoma, if conservative surgery such as myomectomy or morcellation is planned, an ultrasound-guided biopsy with pathological analysis including comparative genomic hybridization analysis must be carried out. In all cases, en bloc surgery, without rupture, is mandatory. Many rare histological subtypes require specific surgical management.

Rate of Leiomyosarcomas during Surgery for Uterine Fibroids: 8-Year Experience of a Single Center

The aim of this study is to investigate the prevalence of occult malignant mesenchymal tumors in patients operated on for uterine fibroids in relation to the surgical approach and type of operation. A retrospective review of all patients that underwent surgery for uterine fibroids (January 2011-December 2018) at the 1st Department of Obstetrics & Gynecology at "Papageorgiou" Hospital. The surgical approach and clinicopathological characteristics were analyzed. A total of 803 patients were operated on: 603 (75.1%) with laparotomy, 187 (23.3%) laparoscopically, and 13 (1.6%) vaginally. Furthermore, 423 (52.7%) patients underwent hysterectomy and 380 (47.3%) myomectomies. Laparoscopy and myomectomy were offered to younger patients with fewer smaller uterine fibroids and were associated with statistically significant shorter hospitalization. The pathological reports revealed: 690 (86%) benign leiomyomas, 32 (4%) cellular leiomyomas, 29 (3.6%) degenerated leiomyomas, 22 (2.7%) adenomyomas, 18 (2.2%) atypical-bizarre leiomyomas, 1 (0.1%) STUMP, 5 (0.65%) endometrial stromal sarcomas, and 6 (0.75%) cases of leiomyosarcomas (LMS). All LMS were preoperatively characterized as suspicious and underwent abdominal hysterectomy. Morcellation was offered in two cases of atypical leiomyomas, with no morcellation-associated complication. Laparoscopy as a valuable surgical approach for young patients with fewer in number and smaller in size fibroids is associated with shorter hospitalization. The risk of unintended morcellation of LMS seems to be very low and can be reduced with careful preoperative work-up but not eliminated.

Percutaneous Uterine Needle Biopsy with Microscopic and Array-CGH Analyses for Preoperative Sarcoma Diagnosis in Patients with Suspicious Myometrial Tumors on MRI: A Prospective Pilot Study (SARCGYN)

BACKGROUND

Unlike for soft tissue sarcomas, percutaneous biopsy is not validated for uterine myometrial tumors, leading to leiomyosarcoma inadvertent morcellation and overtreatment in childbearing patients. This study aimed to evaluate preoperative percutaneous uterine needle biopsy (PUB) with microscopic examination (M-PUB) and array-comparative genomic hybridization (MCGH-PUB).

METHODS

This was a prospective single-center cohort study including all consecutive patients who were candidates for hysterectomy because of suspected uterine leiomyosarcoma on magnetic resonance imaging (MRI) who received PUB. Microscopic and array-CGH analyses with genomic index (GI) counts were performed to guide the therapeutic strategy. Smooth-muscle tumors with suspect features with a GI above 15 were deemed malignant, as were tumors without microscopic malignant features with a complex genomic profile (GI above 30 or malignant profile). Preoperative diagnoses based on M-PUB and MCGH-PUB were compared with the postsurgical pathological specimen or follow-up.

RESULTS

From November 2016 to February 2022, 34 patients were included. Based on the surgical specimen (N = 23) or follow-up (N = 11), final diagnoses were 11 sarcomas and 23 non-sarcomas. The median follow-up was 12 months (IQR 6-37). The diagnostic accuracies of M-PUB and MCGH-PUB were 94% and 100%, respectively. The sensitivity, specificity, and negative predictive value of MCGH-PUB were 100%, 100%, and 100%, respectively. A high GI was significantly associated with malignancy (P < 0.001). Genomic analyses allowed malignancy upgrades for four tumors. There were no complications and no dissemination along the biopsy track.

CONCLUSION

MCGH-PUB is safe and accurate for preoperatively diagnosing uterine sarcomas and should be used routinely after suspicious MRI to tailor surgery.