Optimizing Delayed-Phase CT Strategies for Early Pancreatic Ductal Adenocarcinoma Detection.

RATIONALE AND

OBJECTIVES

To optimize pancreatic CT protocols, particularly for early-stage pancreatic ductal adenocarcinoma (PDAC), focusing on the timing and reconstruction parameters of delayed-phase (DP) imaging.

MATERIALS AND

METHODS

This retrospective, single-center, case-control study included consecutive patients who underwent four-phase pancreatic CT-pancreatic phase (PP), portal venous phase (PVP), 3-min DP, and 5-min DP-between January 2014 and June 2024. Nine image sets were generated-Set 1 (standard protocol: PP and PVP at 120 kVp or 70 keV) and Sets 2-9 incorporating DP (3- or 5-min) and/or dual-energy reconstructions (40 keV images with or without iodine concentration maps). Diagnostic performance for PDAC was compared using McNemar's test, including subgroup analysis for small tumors (≤ 20 mm).

RESULTS

Overall, 886 patients were included (677 PDAC, 209 controls; median age, 70 years [interquartile range, 63-75]). In Set 1, average sensitivity was 88.0% overall and 59.4% for small tumors. Adding DP (Sets 2-3) increased sensitivity, with Set 3 (5-min DP) outperforming Set 2 (3-min DP). Dual-energy reconstructions during PP/PVP (Sets 4-5) also improved sensitivity but remained inferior to DP-inclusive protocols (Sets 2-3). Triple-phase protocols incorporating 3- or 5-min DP with dual-energy reconstruction (Sets 6-9) achieved the highest sensitivity (overall, up to 96.6-97.4%; small tumors, up to 85.9-88.8%), with negligible differences among these sets. Specificity remained high across all sets (99.4-100%).

CONCLUSION

Integrating optimized DP imaging into pancreatic CT protocols improves PDAC detection. A 3-min DP acquisition reconstructed at 40 keV appears most practical, balancing diagnostic performance with workflow efficiency.