This study proposeto integrate a variety of imaging quantitative indicators to establish a new MRI-based tumor response regression(mrTRG) classification method. The accuracy of the established mrTRG classification method according pathology TRG(pTRG) will be tested. The ability of the established mrTRG classification method to predict prognosis will also be tested.

The study included two part: one is a retrospective single-center cohort for analyzing quantitative MRI factors associated with pTRG, from which a new MRI-based tumor response regression(mrTRG) classification method will be established. The other is a prospective multi-center cohort for testing the accuracy of the established mrTRG classification according to pTRG; and also for validating the established mrTRG classification associated with survival.

Inclusion Criteria:

• biopsy-proved primary rectal cancer
• baseline MRI confirmed LARC (cT3 and/or cN+)
• scheduled to take neoadjuvant therapy
• receive baseline and preoperative pelvic MRI
• receive TME surgery after NAT
• have routine follow-up after the operation

Exclusion Criteria:

• history of malignancy or multiple primary cancers
• receive surgery directly without NAT
• don't receive TME surgery
• don't complete NAT
• simultaneous distant metastasis at the baseline diagnosis
• lack of baseline or preoperative pelvic MRI
• incomplete or unavailable pathological information
• lost to follow-up within three months after surgery (lost since the first follow-up visit)

biopsy-proved primary locally advanced rectal adenocarcinoma(cT3 and/or cN+)