Skip to main content
. Author manuscript; available in PMC: 2019 Sep 1.
Published in final edited form as: Expert Rev Anticancer Ther. 2018 Jul 17;18(9):837–860. doi: 10.1080/14737140.2018.1496822

Table 3:

CT scan proposed criteria for identifying aggressive panNETs (Part A), correlating with tumor grade/differentiating High grade G3 from G1/G2 panNETs (Part B) or differentiation of panNETs from adenocarcinomas (Part C)

A. CT scan findings favoring aggressive over non-aggressive panNETs (1)
  1. Presence of pancreatic ductal dilation(p=0.014) [54]; (p<0.05)[49]

  2. Increased tumor size(p=0.003)[54]; p<0,0001 [56]

  3. Presence of vascular involvement (p=0.003)[54]

  4. Presence of lymphadenopathy p=0.002) [54]

  5. The texture parameter entropy (p=0.003)[54]

  6. Tumor shape with less round, more lobulated in advanced grades[56]

  7. On multivariate analysis, size>3cm, (p=0.006); portal enhancement ratio (≤1.1) (p=0.001); hepatic metastases(p=0.003) predicted worse recurrence-free survival [278]

  8. The contrast enhancement pattern of panNETs correlates with the histological classification [279]. None of benign panNETs had early contrast enhancement with rapid wash-out, while panNETs of uncertain behavior or that were NET carcinomas frequent have either even or only late contrast enhancement[279]

B. CT scan findings correlating with grade or distinguishing PanNETs with G3 over G1/G2
  1. G2 over G1 was favored by larger tumor size(p=0.029); tumor conspicuity [non-hyperattenuation compared to pancreatic parenchyma during the portal venous phases] (p=0.016), presence of distant metastases. In a panNET≥2cm, M grade(M1), tumor conspicuity accuracy of a G2 diagnosis was 71%, 61%,71% and all together=825[50]

  2. Presence of iso/hypo-attenuation (43% of panNETs) correlated with higher grading [51].

  3. The CT ratio (proportion of the quantification value in tumor versus parenchyma in arterial phase) predicted G3 grades in panNETs with a sensitivity-100%. specificity-94% and correlated with microvessel density(p<0.001) [52]

  4. Increased tumor grade correlated with increasing tumor size [52,54]; with ill-defined tumor margins[49,53]; lower sphericity, higher skewness of arterial 2D analysis [53]; heterogeneous enhancing[49,56]; hypervascularity[49]

  5. G2 favored over G1 by a lower attenuation value, and ROC analysis showed this had sensitivity of 83%, specificity=92%with AUC=853[55]. G2 was also favored by irregular tumor margins, vessel involvement, cystic degeneration/necrosis, but less that tumor size or CT attenuation[55]

  6. Grade 3 favored over G1/G2 by: portal enhancement ratio (<1) [sensit=92%, specif=81%]; poorly defined margin, tumor size>3cm. bile duct dilation and vascular invasion. When at least 2 of 5 criteria present sensit=92% and specif=88% for G3[57]

C. CT scan findings favoring panNETs over pancreatic ductal adenocarcinomas
  1. Well circumscribed, homogeneously enhancing and hypervascular appearance favor pNETs[49]

  2. Pancreatic duct dilation more frequent in pancreatic cancer[49]

  3. The uncommon features on CT in a panNET of ill-defined, heterogeneously enhancing and hypovascular appearance with duct dilatation could be differentiated from PDAC with 0.76–81 diagnostic performance [49]