Discrepancy Between Radiographic and Pathological Response Assessment in Neoadjuvant Treatment for Pancreatic Cancer: A Comparison Between Neoadjuvant Chemotherapy and Neoadjuvant Chemoradiotherapy
Executive Summary
This briefing document analyzes the findings of a retrospective study conducted at Osaka University Hospital regarding the accuracy of radiological assessments following neoadjuvant treatment (NAT) for pancreatic cancer (PC). The study highlights a critical discrepancy between radiographic responses, evaluated via Response Evaluation Criteria in Solid Tumors (RECIST), and pathological responses, evaluated via the Evans classification.
The most significant finding is that radiological evaluations frequently underestimate the effectiveness of neoadjuvant chemoradiotherapy (NACRT). While radiographic findings may indicate "Stable Disease" (SD), pathological examination often reveals a "favorable response" (destruction of >51% of tumor cells) or even a "pathological complete response" (pCR). This discrepancy is primarily attributed to radiotherapy-induced fibrosis, which mimics residual tumor on imaging. Consequently, the study suggests that clinicians should assess surgical eligibility more proactively in NACRT patients, even when radiographic changes are minimal, particularly when supported by biological markers like CA19-9 and metabolic markers like SUVmax.
Overview of Neoadjuvant Treatment (NAT) in Pancreatic Cancer
Curative resection with pathologically negative margins (R0) remains the primary hope for a cure in pancreatic cancer. Traditionally, surgical resectability was determined by preoperative imaging. However, the emergence of NAT—including neoadjuvant chemotherapy (NAC) and neoadjuvant chemoradiotherapy (NACRT)—has shifted the paradigm.
Key Assessment Tools
RECIST 1.1: A size-based imaging criterion used to categorize tumor response as Progressive Disease (PD), Stable Disease (SD), Partial Response (PR), or clinical Complete Response (cCR).
Evans Classification: A histological grading system used to evaluate the pathological response in resected specimens.
Unfavorable Response (Grade I–IIa): Destruction of <50% of tumor cells.
Favorable Response (Grade IIb–IV): Destruction of >51% of tumor cells or complete absence of viable cells.
Comparative Analysis: NAC vs. NACRT
The study analyzed 243 patients (125 receiving NAC and 118 receiving NACRT). While both groups showed similar results under RECIST criteria, their pathological outcomes differed significantly.
Pathological and Clinical Findings
*Note: The study notes a disparity in histological types and nodal involvement between the groups.
The "Stable Disease" (SD) Discrepancy
The most pronounced discrepancy occurred in patients radiologically classified as having Stable Disease.
NAC Group: Only 12.7% of SD patients showed a favorable pathologic response.
NACRT Group: 36.4% of SD patients showed a favorable pathologic response (p < 0.001).
Underestimation: Patients in the NACRT group were approximately three times more likely to have a favorable pathologic response than NAC patients, despite both being classified as SD by imaging.
Causes of Radiographic Underestimation
The discrepancy between imaging and pathology is largely attributed to the unique anatomopathological nature of pancreatic cancer.
Dense Fibrotic Stroma: PC consists of tumor cells embedded in dense fibrosis, making it difficult to distinguish viable tissue from non-viable tissue via CT or MRI.
Radiotherapy-Induced Changes: NACRT enhances local treatment effectiveness but also induces extensive fibrosis and inflammatory responses. This fibrosis retains the mass's physical size on a scan even after the cancer cells have been destroyed.
Underestimated Viability: In NACRT cases classified as SD, pathological specimens often show "sparse residual tumor cells with extensive fibrosis," whereas NAC cases classified as SD tend to show "minimal pathological changes."
Pathological Complete Response (pCR) in SD Cases
Notably, all instances of pCR (Evans Grade IV) in the NACRT group were associated with residual lesions identified during preoperative imaging. This means patients who appeared to still have a tumor on a CT scan actually had no viable cancer cells remaining.
Biological and Metabolic Indicators of Response
Because anatomical imaging (RECIST) is limited, the study emphasizes the importance of incorporating non-anatomical "biological" factors to predict pathological response.
CA19-9 Dynamics: A reduction of ≥50% in CA19-9 levels was significantly associated with a favorable pathologic response in the NACRT SD group (Odds Ratio: 9.913).
FDG-PET (SUVmax): A reduction in SUVmax of ≥50% was a strong predictor of a major pathologic response in NACRT patients who otherwise showed modest radiographic changes (Odds Ratio: 4.886).
Clinical Implications and Conclusions
The study concludes that the type of NAT administered—NAC or NACRT—must dictate how clinicians interpret preoperative imaging.
NACRT Strategy: Clinical teams should be more proactive in recommending surgery for NACRT patients, even if imaging shows "Stable Disease," provided that biological markers (CA19-9) or metabolic markers (PET-CT) indicate a response. Relying solely on RECIST may lead to the unnecessary exclusion of surgical candidates who have actually achieved a favorable response.
NAC Strategy: In contrast, radiographic findings in the NAC group more closely align with pathological reality; therefore, a more cautious evaluation of surgical indications is warranted when radiographic changes are minimal.
Prognostic Value: Pathological response (Evans classification) remains a superior predictor of Disease-Free Survival (DFS) compared to RECIST. In this cohort, patients with a favorable response (Evans IIb–IV) had a median DFS of 48.5 months, compared to 15.2 months for those with an unfavorable response.
Study Limitations
Retrospective Design: Potential for inherent biases and non-randomized treatment assignment.
Regimen Variation: The NAC group primarily received multi-drug regimens (FOLFIRINOX/GnP), while the NACRT group mostly received single-agent or dual-agent therapies (S-1/Gemcitabine).
Sample Size: The number of patients achieving pCR was limited, requiring larger-scale surveillance to fully understand long-term prognostic impacts.
