Monitoring response to therapy
Nandita M deSouza1

1Institute of Cancer Research, Surrey, United Kingdom


Meaningful biomarkers/biomarker combinations for response assessment in gynecological malignancy must be i) derived to a specified standard, ii) of known accuracy, iii) sensitive to a specific treatment class and iv) measured at an appropriate time within the treatment pathway.

Monitoring response to therapy in Gynecological Malignancy

Imaging assessment of response to treatment relies on detection of a change in tumor size, morphology or character, either during or at specified time points after treatment. Assessments may be qualitative, semi-quantitative or quantitative. Unless automated, data are highly observer dependent not only when assigning scores for response classification but also when delineating regions-of-interest (ROIs), which impact the quantitative data derived from these ROIs. Moreover, the class of treatment administered (radiotherapy, chemotherapy, immunotherapy, targeted therapy, thermal ablation) determines which imaging biomarker delivers maximal sensitivity for response detection and the time-point after treatment at which it is most useful to record this information. This presentation discusses the most appropriate methods for assessing response to a variety of treatment classes in primary and recurrent gynecological malignancy, their optimal timing in relation to treatment schedule and pitfalls in interpretation.

Choice of measurement

Size Measurements: Classically Response Evaluation Criteria In Solid Tumors (RECIST) criteria are used for tumor response assessment (1) particularly in the context of multicenter clinical trials. Choice of imaging modality on which to make these measurements must be consistent as image contrast and sequence-dependent geometric distortion profoundly affect measurement reproducibility. In ovarian cancer traditionally, CT is used. However, although it lacks the geometric distortion inherent in MR images, the soft-tissue contrast between a tumor lesion and the adjacent bowel wall or other normal anatomy is poor. Therefore, application of standard RECIST criteria may be inadequate for assessment of partial response. In cervical cancer, contrast on CT images is inadequate to separate tumor from normal cervix; T2W MRI is essential to provide sufficient tumor to background contrast for tumor delineation and size measurements.

The advent of immunotherapies and the tissue response they elicit has led to modification of classic RECIST criteria to enable response evaluation to this new class of drugs. iRECIST criteria require longitudinal measurements (at least 2 post treatment scans) before a definitive assessment of disease regression or progression is made (2).

Qualitative and Semi-Quantitative scoring systems: Although GI-RADS has been advocated for adnexal mass classification in sonography (3), its use to score response in cross-sectional imaging has not been advocated.

Quantitative assessments: Standardization of the selected biomarker is an essential requirement. This applies both at the image acquisition stage and during image processing. Analysis methodology for deriving quantified biomarkers require that image acquisition parameters are standardised at the outset. For example, the apparent diffusion coefficient (ADC) extracted from diffusion-weighted MR sequences are influenced by the b- values selected and the mathematical model fitting used (mono, bi-exponential, stretched exponential, ivim) (4). Similarly, for quantitative DCE assessments (Ktrans, Kep, Ve) the method of modelling, dose of contrast agent and temporal resolution of the scans impacts assessment of the arterial input function and the absolute values of the data subsequently derived (5, 6). Normalization of the biomarker to a reference tissue e.g. using myometrial enhancement as a comparator in endometrial cancer, necessitates standardization across examinations so that data is comparative. Finally, biomarker combinations may provide additional information on mechanism of response (metabolism, necrosis) (7, 8, 9), which will increasingly be implemented with MR-PET systems.

Classes of Therapies

The treatment of gynecological malignancy is primarily surgical where hysterectomy, bilateral salpingo-oophorectomy, nodal dissection and debulking procedures appropriately selected clear the majority of disease. In more advanced uterine or cervical cancer, radiotherapy is the primary treatment modality often in combination with chemotherapy. Ovarian cancer usually presents late and warrants neoadjuvant or adjuvant platinum-based chemotherapy. Increasingly, antiangiogenic agents and newer targeted agents (e.g. PARP inhibitors) are being added to these regimens. Focal ablative therapies with cyberknife radiotherapy and high-intensity focused ultrasound are being deployed in recurrent disease. Each of these management strategies requires smart adaptation of the imaging techniques used for response assessment. Functional measurements, for example, are an increasingly important consideration for targeted agents where size criteria are insensitive to early response (10). Dynamic contrast enhanced parameters may be most appropriate where antiangiogenic agents are used (11).

Timing of response assessment

The timing of response assessment is critical in the context of the treatment being administered because of potential adaptive treatment modifications. Imaging during radiotherapy for example may be advantageous if there is an opportunity to adapt plans to changes in tumour size and shape, particularly with the advent of linear accelerators integrated with MR scanners (MR Linac). Stopping potentially toxic agents or introduction of targeted agents may also form part of new adaptive chemotherapeutic strategies. The timing of response assessment scans may also be used to predict long-term outcome (12, 13).

In summary, meaningful biomarkers/ biomarker combinations for response assessment in gynecological malignancy must be i) derived to a specified standard, ii) of known accuracy, iii) sensitive to a specific treatment class and iv) measured at an appropriate time within the treatment pathway.


CRUK and EPSRC support to the Cancer Imaging Centre at ICR and RMH in association with MRC and Department of Health C1060/A10334, C1060/A16464 and NHS funding to the NIHR Biomedical Research Centre and the Clinical Research Facility in Imaging.




3. Zhang T, Li F, Liu J, Zhang S. Diagnostic performance of the Gynecology Imaging Reporting and Data System for malignant adnexal masses. Int J Gynaecol Obstet. 2017;137: 325-331.

4. Winfield JM, deSouza NM, Priest AN, Wakefield JC, Hodgkin C, Freeman S, Orton MR, Collins DJ. Modelling DW-MRI data from primary and metastatic ovarian tumours. Eur Radiol. 2015; 25: 2033-2040.

5. Winfield JM, Payne GS, Weller A, deSouza NM. DCE-MRI, DW-MRI, and MRS in Cancer: Challenges and Advantages of Implementing Qualitative and Quantitative Multi-parametric Imaging in the Clinic. Top Magn Reson Imaging. 2016;25: 245-254.

6. Sala E, Rockall A, Rangarajan D, Kubik-Huch RA. The role of dynamic contrast-enhanced and diffusion weighted magnetic resonance imaging in the female pelvis. Eur J Radiol. 2010; 76: 367-385.

7. Brandmaier P, Purz S, Bremicker K, Höckel M, Barthel H, Kluge R, Kahn T, Sabri O, Stumpp P. Simultaneous [18F]FDG-PET/MRI: Correlation of Apparent Diffusion Coefficient (ADC) and Standardized Uptake Value (SUV) in Primary and Recurrent Cervical Cancer. PLoS One. 2015; 10: e0141684.

8. Sarabhai T, Tschischka A, Stebner V, Nensa F, Wetter A, Kimmig R, Forsting M, Herrmann K, Umutlu L, Grueneisen J. Simultaneous multiparametric PET/MRI for the assessment of therapeutic response to chemotherapy or concurrent chemoradiotherapy of cervical cancer patients: Preliminary results. Clin Imaging. 2018; 49: 163-168.

9. Oldan JD, Shah SN, Rose TL. Applications of PET/MR Imaging in Urogynecologic and Genitourinary Cancers. Magn Reson Imaging Clin N Am. 2017; 25: 335-350.

10. Tirkes T, Hollar MA, Tann M, Kohli MD, Akisik F, Sandrasegaran K. Response criteria in oncologic imaging: review of traditional and new criteria. Radiographics. 2013; 33: 1323-1341.

11. Chase DM, Sill MW, Monk BJ, Chambers MD, Darcy KM, Han ES, Buening BJ, Sorosky JI, Fruehauf JP, Burger RA. Changes in tumor blood flow as measured by Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) may predict activity of single agent bevacizumab in recurrent epithelial ovarian(EOC) and primary peritoneal cancer (PPC) patients: an exploratory analysis of a Gynecologic Oncology Group Phase II study. Gynecol Oncol. 2012; 126: 375-380.

12. Makino H, Kato H, Furui T, Morishige K, Kanematsu M. Predictive value of diffusion-weighted magnetic resonance imaging during chemoradiotherapy for uterine cervical cancer. J Obstet Gynaecol Res. 2014; 40: 1098-1104.

13. Minkoff D, Gill BS, Kang J, Beriwal S. Cervical cancer outcome prediction to high-dose rate brachytherapy using quantitative magnetic resonance imaging analysis of tumor response to external beam radiotherapy. Radiother Oncol. 2015; 115: 78-83.

Proc. Intl. Soc. Mag. Reson. Med. 26 (2018)