Clinicopathological correlation of radiologic measurement of post-therapy tumor size and tumor volume for pancreatic ductal adenocarcinoma

Dongguang Wei, Mohamed M. Zaid, Matthew H. Katz, Laura R. Prakash, Michael Kim, Ching Wei D. Tzeng, Jeffrey E. Lee, Anshuman Agrawal, Asif Rashid, Hua Wang, Gauri Varadhachary, Robert A. Wolff, Eric P. Tamm, Priya R. Bhosale, Anirban Maitra, Eugene J. Koay, Huamin Wang

Research output: Contribution to journalArticlepeer-review

5 Scopus citations


OBJECTIVES: Tumor size measurement is critical for accurate tumor staging in patients with pancreatic ductal adenocarcinoma (PDAC). However, accurate tumor size measurement is challenging in patients who received neoadjuvant therapy before resection, due to treatment-induced fibrosis and tumor invasion beyond the grossly identified tumor area. In this study, we evaluated the correlation between the tumor size and tumor volume measured on post-therapy computed tomography (CT) scans and the pathological measurement. Also, we investigated the correlation between these measurements and clinicopathological parameters and survival.

MATERIALS AND METHODS: Retrospectively, we evaluated 343 patients with PDAC who received neoadjuvant therapy, followed by pancreaticoduodenectomy and had pre-operative pancreatic protocol CT imaging. We measured the longest tumor diameter (RadL) and the radiological tumor volume (RadV) on the post-therapy CT scan, then we categorized RadL into four radiologic tumor stages (RTS) based on the current AJCC staging (8th edition) protocol and RadV based on the median. Pearson correlation or Spearman's coefficient (δ), T-test and ANOVA was used to test the correlation between the radiological and pathological measurement. Chi-square analysis was used to test the correlation with the tumor pathological response, lymph-node metastasis and margin status and Kaplan-Meier and Cox-proportional hazard for survival analysis. P-value < 0.05 was considered significant.

RESULTS: As a continuous variable, RadL showed a positive linear correlation with the post-therapy pathologic tumor size in the overall patient population (Pearson correlation coefficient: 0.72, P < 0.001) and RadV (δ: 0.63, p < 0.0001). However, there was no correlation between RadL and pathologic tumor size in patients with ypT0 and those with pathologic tumor size of ≤1.0 cm. Post-therapy RTS and RadV group correlated with ypT stage, tumor response grades using either CAP or MDA grading system, distance of superior mesenteric artery margin and tumor recurrence/metastasis.

CONCLUSION: Although RadL tends to understage ypT in PDAC patients who had no radiologically detectable tumor or small tumors (RTS0 or RTS1), radiologic measurement of post-therapy tumor size may be used as a marker for the pathologic tumor staging and tumor response to neoadjuvant therapy.

Original languageEnglish (US)
Pages (from-to)200-207
Number of pages8
Issue number1
StatePublished - Jan 2021


  • Pancreatic cancer
  • Radiologic tumor size
  • Radiologic tumor volume
  • Tumor response grade
  • Tumor stage
  • Carcinoma, Pancreatic Ductal/diagnostic imaging
  • Humans
  • Middle Aged
  • Kaplan-Meier Estimate
  • Male
  • Tomography, X-Ray Computed
  • Treatment Outcome
  • Lymphatic Metastasis
  • Pancreaticoduodenectomy
  • Pancreatic Neoplasms/diagnostic imaging
  • Disease-Free Survival
  • Survival Analysis
  • Aged, 80 and over
  • Adult
  • Female
  • Neoadjuvant Therapy
  • Aged
  • Retrospective Studies
  • Neoplasm Staging

ASJC Scopus subject areas

  • Endocrinology
  • Endocrinology, Diabetes and Metabolism
  • Hepatology


Dive into the research topics of 'Clinicopathological correlation of radiologic measurement of post-therapy tumor size and tumor volume for pancreatic ductal adenocarcinoma'. Together they form a unique fingerprint.

Cite this