This study looks at better ways to measure how tumors respond to treatment before surgery. Today, doctors often use RECIST 1.1 (Response Evaluation Criteria in Solid Tumors), which measures the longest diameter across a tumor on an axial scan. This can be unreliable for tumors with irregular shapes. The investigator will compare a new, simple 3-dimensional method called TDTP-RECIST (Three-Dimensional Twelve-Point RECIST) with RECIST 1.1 and other volume-based methods to see which one matches best with what is found in the tumor after it is removed.

This is a retrospective study at Zhongshan Hospital, Fudan University (Shanghai, China), Shenzhen University General Hospital (Shenzhen, China), and Tianjin Cancer Hospital Airport Hospital (Tianjin, China). The investigator will review past medical records and CT or MRI scans from adults (18-80 years) who had biliary tract cancer, breast ductal carcinoma, or pancreatic ductal adenocarcinoma. All included patients received neoadjuvant or "conversion" therapy (treatment given before surgery to shrink the tumor) and then had surgery between 2019 and 2024. The investigator will not contact participants or change their care.

On the scans done before treatment and before surgery, the investigator will calculate tumor response using several methods (RECIST 1.1 and volume-based methods, including TDTP-RECIST). The investigator will compare these imaging results with the "pathology response" seen in the removed tumor tissue (how much living cancer is left). The investigator will also see how well the imaging methods relate to clear surgical margins (R0 resection, meaning no cancer at the cut edge), time without cancer coming back (recurrence-free survival), and overall survival.

About 120 participants are expected. The analysis is planned from August to December 2025. There are no study visits, procedures, or treatments for participants, so risks are minimal. There is no direct benefit to participants. Findings may help doctors choose better tools to judge treatment response for irregular tumors in the future. All data will be de-identified. The hospital's ethics committee has approved this retrospective review and waived individual consent. Sponsor: Zhongshan Hospital, Fudan University. Location: Shanghai, China.

Inclusion Criteria:

1. Pathologically confirmed malignancy of one of the following types: biliary tract cancer (intrahepatic cholangiocarcinoma, hilar cholangiocarcinoma, or gallbladder cancer), breast cancer, or pancreatic ductal adenocarcinoma
2. Age 18 to 80 years (inclusive), any sex
3. Received neoadjuvant or conversion therapy before surgery; no prior anti-cancer therapy before that course
4. Underwent tumor surgical resection or biopsy after completion of neoadjuvant/conversion therapy
5. Available baseline and preoperative imaging with the same modality (contrast-enhanced CT or MRI) at both time points for response evaluation
6. Complete postoperative pathology report and follow-up data available

Exclusion Criteria:

1. History of another malignancy or concurrent multiple primary cancers
2. Missing imaging or inconsistent imaging evaluation across time points (e.g., absent scans or differing modalities that preclude paired assessment)

This study includes adults who were treated and had surgery at Zhongshan Hospital, Fudan University (Shanghai, China) Shenzhen University General Hospital (Shenzhen, China), Tianjin Cancer Hospital Airport Hospital (Tianjin, China), between May 2019 to Dec. 2024. All participants had one of the following cancers: biliary tract cancer (intrahepatic or hilar cholangiocarcinoma, or gallbladder cancer), breast cancer (invasive ductal carcinoma), or pancreatic ductal adenocarcinoma. They received treatment given before surgery to shrink the tumor (neoadjuvant or "conversion" therapy) and then underwent surgical removal of the tumor.