This is a retrospective study using images acquired routinely for diagnosis of rectal cancer to see if these could be used to predict responses to radiotherapy treatment and if it can, whether the treatment can be optimised to produce better outcome for patients. Using a clinical database, patients who have had neo-adjuvant chemo-radiotherapy will be recruited, their diagnostic images and radiotherapy planning scan will be obtained. By use of imaging registration and clinical information, the question of why some patients respond well to radiotherapy and some don't could be answered.

The treatment of locally advanced rectal cancer is primarily radiotherapy (+/- chemotherapy) followed by surgery. The reason for radiotherapy is to reduce the risk of recurrence after surgery or to shrink the tumour first so that when surgery is done all the cancer will be successfully removed otherwise some will be left behind to grow. More than 80% of patients require a stoma after their surgery, some get reversed by two years after surgery but a third has it lifelong. About 15% of patients after radiotherapy have no disease left (clinical complete response) and can be monitored closely after radiotherapy and will not need to have surgery or stoma. Surgery carries a risk of death and complications, having a stoma have a lot of complications and have an effect on patient's quality of life, most end up not going out much and withdraws from friends and family due to the risk of accidents in public places. This study is aimed at looking at ways to increase the number of patients that do not require surgery after radiotherapy by looking at the differences between those that responded well to radiotherapy and the ones that did not by comparing their diagnostic and treatment scans. The main question to answer is why some patients have complete response to radiotherapy and others don't. Is there a way to increase the number of these patients through changes in radiotherapy? The study will be looking at the diagnostic images and radiotherapy planning scans to compare these two groups. Is there a way of predicting who will respond to radiotherapy treatment? If there is, modifications could be made to the type of treatment given. This study will be looking to radiomics techniques to develop this. This retrospective study will only make use of scans that patients have already had for their diagnosis and treatment so no patient intervention is required. Patients will be recruited using the clinical and research database of the Christie hospital which is the largest cancer centre in the UK. The study is funded by the charitable fund of the Christie Hospital NHS Foundation Trust.

Inclusion Criteria:

• Histologically confirmed rectal adenocarcinoma.
• Received pelvic radiotherapy (+/- chemotherapy) as neo-adjuvant treatment
• Age 18 and above

Exclusion Criteria:

• Other rectal pathologies.
• Patients less than 18yrs at diagnosis.

Patients who are treated with neo-adjuvant chemo-radiotherapy recruited from the clinical data of the Christie Hospital NHS databases