Magnetic Resonance Imaging Diagnosis of Uterine Leiomyosarcomas (IDOL)

  • Research type

    Research Study

  • Full title

    Magnetic Resonance Imaging Diagnosis of Uterine Leiomyosarcomas (IDOL)

  • IRAS ID

    251778

  • Contact name

    Christina Messiou

  • Contact email

    christina.messiou@rmh.nhs.uk

  • Sponsor organisation

    The Royal Marsden NHS Foundation Trust

  • Duration of Study in the UK

    1 years, 6 months, 1 days

  • Research summary

    Uterine fibroids are a common benign lesion located in the uterine muscle wall (the myometrium). A very rare and aggressive cancer called a leiomyosarcoma (LMS) also occurs in the myometrium. If fibroids grow rapidly, they present in a similar manner to LMS and can be difficult to distinguish from fibroids both clinically and radiologically. Unfortunately, if common treatments for fibroids are performed on a LMS, this can cause the cancer to spread or be missed thus worsening patient outcome. Achieving a biopsy of a myometrial lesion is challenging, therefore distinguishing the two conditions with imaging (MRI as the gold standard modality) is essential in directing patient management appropriately. Due to the low prevalence of LMS, previous studies have only analysed a small number of cases and have thus failed to create a validated scoring system. We propose undertaking a retrospective study to construct and undertake validation of a magnetic resonance imaging scoring system designed for use in pelvic MR imaging to distinguish uterine LMS from complex fibroids. We will source and anonymise MRI scans already performed on patients with a LMS at the Royal Marsden Hospital and a group of patients with degenerated fibroids from Imperial Healthcare – both with histological confirmation. These cases will be split into a training and validating set. Two experienced radiologists will blind read the training set using a lexicon of definitions for imaging features. Statistical analysis of these imaging features will be used to create a MR scoring system to determine the likelihood of LMS. This scoring system will then be created and tested by two readers with a validation set. If validated, these can be used by gynaecological radiologists to predict the likelihood of malignancy in the future.

  • REC name

    Yorkshire & The Humber - South Yorkshire Research Ethics Committee

  • REC reference

    19/YH/0134

  • Date of REC Opinion

    16 Apr 2019

  • REC opinion

    Favourable Opinion