PLoS ONE (Jan 2016)

Diagnostic Value and Interreader Agreement of the Pancreaticolienal Gap in Pancreatic Cancer on MDCT.

  • Khoschy Schawkat,
  • Wolfgang Kühn,
  • Daniel Inderbitzin,
  • Beat Gloor,
  • Johannes T Heverhagen,
  • Val Murray Runge,
  • Andreas Christe

DOI
https://doi.org/10.1371/journal.pone.0166003
Journal volume & issue
Vol. 11, no. 11
p. e0166003

Abstract

Read online

The aim of this retrospective study was to evaluate the diagnostic value and measure interreader agreement of the pancreaticolienal gap (PLG) in the assessment of imaging features of pancreatic carcinoma (PC) on contrast-enhanced multi-detector computed tomography (CE-MDCT).CE-MDCT studies in the portal venous phase were retrospectively reviewed for 66 patients with PC. The age- and gender-matched control group comprised 103 healthy individuals. Three radiologists with different levels of experience independently measured the PLG (the minimum distance of the pancreatic tail to the nearest border of the spleen) in the axial plane. The interreader agreement of the PLG and the receiver operating characteristic (ROC) curve was used to calculate the accuracy of the technique.While the control group (n = 103) showed a median PLG of 3 mm (Range: 0 - 39mm) the PC patients had a significantly larger PLG of 15mm (Range: 0 - 53mm)(p 12 mm for PC, with a sensitivity of 58.2% (95% CI = 45.5-70.1), specificity of 84.0% (95% CI = 75.6-90.4) and an area under the ROC curve of 0.714 (95% CI = 0.641 to 0.780). The mean interreader agreement showed correlation coefficient r of 0.9159. The extent of the PLG did not correlate with tumor stage but did correlate with pancreatic density (fatty involution) and age, the density decreased by 4.1 HU and the PLG increased by 0.8 mm within every 10 y.The significant interreader agreement supports the use of the PLG as a characterizing feature of pancreatic cancer independent of the tumor stage on an axial plane. The increase in the PLG with age may represent physiological atrophy of the pancreatic tail.