Oxygen-enhanced MRI in patients with uterine cervical carcinoma: feasibility and treatment-associated changes in...

Tumour hypoxia causes resistance against radiotherapy and chemotherapy in locally advanced cervical cancer (LACC). Non-invasive magnetic resonance imaging (MRI) biomarkers have the potential to guide hypoxia-targeted therapy but require technical development and validation. This study evaluated the feasibility and early treatment response monitoring capabilities of oxygen-enhanced MRI (OE-MRI) in healthy female volunteers and patients with LACC.

Healthy volunteers underwent OE-MRI to demonstrate the feasibility of the technique and to compare the onset, magnitude, and duration of change in longitudinal relaxation rate (ΔR ₁) for breathing either 100% oxygen or carbogen. Uterine cervix and body tissues were examined. Next, patients with biopsy-proven LACC underwent OE-MRI pre-treatment, and then, at 3 and 5 weeks into chemotherapy, it was combined with external beam radiotherapy.

Feasibility was demonstrated in 97% (31/32) of uterine tissue regions of interest, measured in 18 volunteers. ΔR ₁ values in the uterine cervix and body were similar across gas types with a cohort ΔR ₁ of between 0.026 and 0.033 s^-1. In patients with LACC, significant increases in tumour ΔR ₁ were observed during treatment by week 3 (mean increase, 0.0143 s^-1; 95% CI 0.0014 to 0.0273; p = 0.035). At week 5, six of eight evaluable patients showed a >50% treatment-associated increase in tumour ΔR ₁. Changes persisted in most patients to the end of treatment.

This first dedicated OE-MRI study in cervical cancer shows technique feasibility in healthy volunteers and patients with LACC. Treatment-associated increases in tumour ΔR ₁ were demonstrated during chemoradiotherapy and likely reflect hypoxia modification. These findings support further evaluation of OE-MRI alongside independent hypoxia biomarkers in larger cervical cancer cohorts.