logo SBA


Digital archive of theses discussed at the University of Pisa


Thesis etd-07302009-161736

Thesis type
Tesi di dottorato di ricerca
Thesis title
Academic discipline
Course of study
tutor Prof. Caramella, Davide
  • caratterizzazione
  • neoplasie neuroendocrine
  • pancreas
  • TC multidetettore
Graduation session start date
Release date
The role of imaging in the evaluation of neuroendocrine pancreatic neoplasms includes the localization of functioning neoplasms, the differential diagnosis with other pancreatic lesions, the identification of signs of malignancy and the loco-regional staging, that may help in suggesting the best therapeutic approach.
At the state of the art, pre-operative suspicion of lesion nature arises both from clinical and imaging finding. If clinical criterium generally adopted in order to formulate a diagnosis of benign lesion is related to the presence and the type of secreted hormone, imaging suggestion of lesion nature is based quite only on nodular dimension and on the presence of local and distant spreading.
Aim of our study was to evaluate the possibility to determine the nature of neuroendocrine pancreatic neoplasms by analysing lesions enhancement pattern at MDCT.
Between January 2005 and September 2008, 36 patients referred to our Institution because of the suspicion of neuroendocrine pancreatic neoplasm. Clinically, 9 patients complained symptoms due to hormonal secretion.
All preoperative CT examinations were performed by a multidetector CT. CT studies were performed by acquiring basal as well as post-contrastographic scans, after intravenous administration of high-concentration iodinated contrast medium (Iomeron® 400) at 4 up to 5ml/min flow rate. Post-contrastographic study included 4 phases: early arterial phase (delay 15-20”); pancreatic phase (delay 35”); venous phase (delay 70”) and late phase (delay 180”). At MDCT, we performed a qualitatively analysis of the post-contrastographic behaviour in different subsequent phases. Particularly, two different patterns of enhancement were defined: pattern A, including lesions showing an early enhancement (during early arterial or pancreatic phase) and a rapid wash-out; pattern B, including lesions showing wash-in in the early arterial or pancreatic phase with no wash-out nor in the late phase (pattern B1), and lesions showing enhancement only in the venous and/or late phases (pattern B2).
At MDCT, the greatest lesion dimension was also measured.
Data obtained from MDCT examination were then related to those obtained at histology after surgical resection.
At MDCT, 55 lesions were detected in 36 patients. At MDCT, 29 out of 55 (53%) lesions showed a pattern of type A; diameters of such lesions ranged between 5 and 20mm (average maximum diameter-a.m.d: 13mm). Twenty-six out of 55 (47%) lesions showed instead pattern B; diameters of such lesions ranged between 12 and 100mm (a.m.d: 41mm); 18 lesions presented a pattern of type B1 (a.m.d: 38mm) and 8 a pattern of type B2 (a.m.d: 49mm). Pathology assessed the neuroendocrine nature of all 55 lesions identified at MDCT. Particularly, 25 out of 55 (45%) lesions were classified as benign neuroendocrine pancreatic neoplasms, 9 out of 55 (16%) as borderline neoplasms and 21 out of 55 (39%) as low grade carcinomas.
Pathological analysis stated that 23 out of 29 lesions showing pattern A (79%; a.m.d 12mm) resulted to be benign neuroendocrine pancreatic neoplasms (a.m.d: 8mm), 5 resulted to be borderline neuroendocrine pancreatic neoplasms (a.m.d: 16mm) and 1 resulted to be a low grade carcinoma (12mm).
Twenty out of 26 lesions (77%) showing pattern B resulted to be low grade carcinomas (a.m.d: 39mm), 4 borderline neuroendocrine neoplasms (a.m.d: 18mm) and 2 lesions, measuring respectively 15mm and 20mm, benign neuroendocrine neoplasms; overall average greatest diameter of lesions showing pattern B was 32mm. In particular, among the lesions showing pattern B1, 11 out of 18 (61%), resulted to be low grade carcinomas, 5 out of 18 (28%) borderline neuroendocrine neoplasms and 2 out of 18 (11%) resulted to be benign. All the 8 lesions (100%) showing pattern B2 resulted to be low grade carcinomas at histology.
Overall, pattern A showed a positive predictive value (PPV) in assessing benignancy of 79%, while pattern B showed a PPV of 77% in assessing malignancy.
Regarding the correlation between MDCT and histology about lesions dimensions, we found an excellent correlation (R=0.96) in case of lesions with pattern A and a very good (R= 0.85) correlation in lesions with pattern B.
In conclusion, MDCT may suggest the nature of a neuroendocrine pancreatic neoplasm and particularly, the lesion post-contrastographic pattern may represent a further criterium for suspecting lesion malignancy. Anyway, lesion dimension greater than 2cm can be confirmed as a reliable cut off for suspecting the lack of benignancy.