Hypoxic tumour cells within the primary tumour have shown prognostic importance for local and metastatic disease control in several cancer sites. Radioresistant hypoxic cells diminish the rate of local control, and the hypoxia driven increase in metastatic potential of the tumour and lowers the rate of distant disease control. DCE MR imaging has been used to quantify the extent of poor perfusion regions within cervical tumours and it has been shown to be a surrogate of hypoxia. Furthermore, a number of studies have demonstrated that DCE MR is predictive of disease failure in cervix cancer.

The EMBRACE II study will implement an imaging sub-study, which will evaluate the value of quantitative MR imaging to identify patients at increased risk of disease recurrence (local, nodal and systemic).

Radiotherapy is an important treatment modality in the management of cancers of the uterine cervix. About half of patients with cervical cancer receive definitive radiotherapy in the course of their disease. Radiotherapy is most often administered as a combination of external beam radiotherapy (EBRT) and brachytherapy (BT). In the late 90's a major breakthrough in cervical cancer radiotherapy took place with the introduction of MR image guidance of BT. By performing MR imaging before each BT implant it is possible adapt the dose given by BT to the anatomy of each individual patient taking into account not only the position of organs at risk (bladder, rectum and sigmoid) but also the tumour regression induced by the preceding EBRT and chemotherapy.

Functional imaging that reflects hypoxia, metabolism, heamodynamics and tissue structure have been applied to locally advanced cervical cancer with the goal to identify imaging markers that may predict outcome early on and improve tissue classification. DCE-MRI may be the most investigated so far for locally advanced cervical cancer. A comprehensive literature review including papers investigating the prognostic value of DCE-MRI in patients with locally advanced cervical cancer identified 20 papers from 10 research groups, with a median number of 30 patients (range 7-102 patients). A total number of 17 papers publish a positive association between pre-treatment DCE-MRI and outcome in terms of local control or disease free survival (1-17). However, not all studies present independent cohorts of patients. Three papers show no effect (18-20) The studies on cervical cancer points in the direction that DCE-MRI has the capability to identify aggressive forms of cervical cancer, and that the pre-treatment measurements may serve as, predictive markers for outcome after chemo-radiotherapy. The largest studies indicate that in particular the tumour fraction with the lowest signal enhancement is an important parameter, though the diversity in methodology is significant.

Diffusion weighted MRI (DWI-MRI) has to a lesser extent than DCE-MRI been investigated in locally advanced cervical cancer. Most studies using DWI-MRI in cervical cancer have investigated its diagnostic capabilities (21-28) all concluding high sensitivity and specificity (review by Kundu et al. (29)). The Toronto group; McVeight et al. (26) and later Gladwish et al. (30) found prior to the onset of treatment that the highest 90th % ADC value correlated with response, similar finding was found by the group in Tianjin; Liu et al. (22). Both groups found that higher ADC value insides the tumour was predictive of poor response to treatment and suggest the higher ADC to be connected to tumour necrosis. When tumour necrosis, occur there is loss of cell membrane integrity and therefore an increase in the extracellular volume and a decrease in intracellular volume effectively increasing the ADC. Conversely, the group from London UK; Harry et al. (31) and Somoye et al. (32) showed no correlation to treatment response at the time prior to treatment. Instead the ADC at 2 weeks (and the change in ADC) into treatment was predictive of treatment response and prognostic of patient outcome. Finally, Marconi et al. (33) found a relation between minimum ADC in the tumor and both DSS and DFS.

This is an observational prospective, non-randomized study in which patients with locally advanced cervical cancer included in the EMBRACE II study can enroll. The study will be carried out in 8-15 EMBRACE centres. MRI will be carried out prior to radiotherapy. The details of the MRI exams will differ from standard clinical practice in the centres, but will be consistent with international guidelines for cervix MRI. The exam will include T1, T2, diffusion, and dynamic contrast-enhanced imaging. At time of brachytherapy, the treatment planning MRI will additionally include DWI and qT2. Patients will be followed up according to the EMBRACE II follow-up schedule.

Inclusion Criteria:

• Patients included in the EMBRACE study (see inclusion criteria in the EMBRACE protocol)
• Patients without previous record of allergic reaction to infusion of protocol related contrast media (Gadolinium-based)
• Patients with sufficient kidney function according to local regulations
• Patient informed consent

Exclusion Criteria:

• According to EMBRACE II protocol
• Patients with active infection or severe medical condition

Yearly, each EMBRACE II centre is expected to enrol 10-20 patients. The inclusion for the imaging protocol in a given centre is expected to be 70% of patients enrolled in EMBRACE II. We expect that at least 8 institutions will enrol 10 patients per year, which will result in minimum 80 patients per year, and thereby at least 320 patients within the 4-year study period. This number is sufficient to meet the required number of patients for the hypotheses in the overall patient population (sample size 196 patients) and in the high-risk patients (sample size 248 patients )