This prospective imaging data provides heterogeneity imaging assessment of the tumor volume (cervical cancer) with both functional MRI and 18FDG PET in parallel. The resulting imaging data sets allow detailed study of the evolution of functional/molecular tumor heterogeneity before, across and after completion of the radiation therapy course. This is a unique data set (not available elsewhere to our knowledge). The prospectively obtained and standardized imaging is not only timed according to the radiation dose level and time course, but also correlated with ultimate the treatment outcome (tumor response, local tumor control, survival).

Cervical cancer is an ideal model to study vascular, cellular and molecular tumor properties across the radiation treatment course as the tumor undergoes radiation-induced functional/biological changes within its heterogeneous volume. The understanding of such biological processes occurring during the radiation course remains a major knowledge gap in radiation oncology. Such phenomena are challenging to study with histologic biopsy due to the inability to sample the entire tumor, and due to the challenge and risk in performing multiple tumor biopsies during radiation therapy in clinical patients.

In contrast functional/molecular imaging can assess this biological heterogeneity encompassing the entire tumor volume. Our study does so longitudinally across the time course of radiation therapy. These intra-treatment functional phenomena may translate into important actionable imaging biomarkers of long-term treatment outcome. For example, vascular tumor properties (as identified by DCE MRI) are highly significant for tumor oxygenation, which in turn profoundly influences radiation response. For example, our early investigations suggest that typically dynamic contrast enhancement (indicative of tumor microvasculature) improves early in the treatment course and heterogeneity decreases, particularly in responders; while FDG heterogeneity tends to show reduced metabolic activity that becomes evident later in the radiation treatment course.

Our imaging data in cervical cancer may complement the TCIA's existing disease portfolio, which currently to our knowledge contains only one other collection in cervical cancer (imaging–genomic atlas).

We believe our functional/molecular imaging data set can serve a unique resource, that should be made available to the wider scientific community, so that it will continue to be available to longitudinally (across a cytotoxic therapy course) interrogate sequential functional and molecular tumor properties – well into the future as image analysis and radiomics techniques will continue to advance, well beyond the time course of our own studies.

Translation to clinical practice: The imaging sequences, while complex and functional, were designed upfront such that they are translatable and widely applicable to community settings. This is important because cervical cancer is a disease largely treated in the community setting, often in underserved communities.