Purpose or Learning Objective Evaluating whether radiomic analysis applied to 68Ga-DOTANOC PET/CT examinations could allow detecting misassigned bioptic grading, when discriminating grade 1 (G1) and grade 2 (G2) pancreatic neuroendocrine tumours (panNET). Methods or Background Thirty-three patients with low grade primary panNET assessed through pathological bioptic samples underwent 68Ga-DOTANOC PET/CT. Seven lesions with volume lower than 1cm3 were excluded from the study and standardized uptake value (SUV) maps were extracted for the remaining twenty-six patients (M:F=11:15; mean age:64 y.o. [48-85]; G1:G2=18:8). 60- first and second-order radiomic features (RFs) were computed on SUV maps of the whole tumour volume, discarding those showing high linear correlation. A discriminative radiomic model (RM) was generated for each of the surviving RFs. The area under the curve (AUC) were computed on the receiver operating characteristic curve. To evaluate models’ discrimination capability, the two-tail Wilcoxon rank-sum test was applied (p-value≤0.05) on the features space and the RF with the lowest p-values and the highest AUC were selected. Results or Findings The best performing RM is based on second-order correlation which provides a significant (p-value=0.028) separation between G2 and G1 panNET, with AUC=0.78, sensitivity=88%, specificity=78%, accuracy=81%. More importantly, the only FN is a lesion with borderline ki67=3 index, detected as FN in all the other RMs, this leading to questioning if it was correctly discriminated by the bioptical sample examination. Conclusion This preliminary study showed that this single RF well discriminates low grade panNET. Considering that tumour grade based on bioptical samples is not representative of the entire lesion, RFs may be used as tool to overcome this limitation, assigning a correct grade in all patients, especially in cases with borderline ki67. Limitations Reduced patients’ sample size and lack of grading assessment on tumour surgically excided.

Radiomic analysis could allow detecting misassigned bioptic grading when discriminating G1 and G2 primary pancreatic neuroendocrine tumours imaged with 68Ga-DOTANOC PET/CT

Silvia Malavasi;Diletta Calabrò;Alessandro Bevilacqua;Claudio Ricci;Riccardo Casadei;Davide Campana;Stefano Fanti;Valentina Ambrosini
2021

Abstract

Purpose or Learning Objective Evaluating whether radiomic analysis applied to 68Ga-DOTANOC PET/CT examinations could allow detecting misassigned bioptic grading, when discriminating grade 1 (G1) and grade 2 (G2) pancreatic neuroendocrine tumours (panNET). Methods or Background Thirty-three patients with low grade primary panNET assessed through pathological bioptic samples underwent 68Ga-DOTANOC PET/CT. Seven lesions with volume lower than 1cm3 were excluded from the study and standardized uptake value (SUV) maps were extracted for the remaining twenty-six patients (M:F=11:15; mean age:64 y.o. [48-85]; G1:G2=18:8). 60- first and second-order radiomic features (RFs) were computed on SUV maps of the whole tumour volume, discarding those showing high linear correlation. A discriminative radiomic model (RM) was generated for each of the surviving RFs. The area under the curve (AUC) were computed on the receiver operating characteristic curve. To evaluate models’ discrimination capability, the two-tail Wilcoxon rank-sum test was applied (p-value≤0.05) on the features space and the RF with the lowest p-values and the highest AUC were selected. Results or Findings The best performing RM is based on second-order correlation which provides a significant (p-value=0.028) separation between G2 and G1 panNET, with AUC=0.78, sensitivity=88%, specificity=78%, accuracy=81%. More importantly, the only FN is a lesion with borderline ki67=3 index, detected as FN in all the other RMs, this leading to questioning if it was correctly discriminated by the bioptical sample examination. Conclusion This preliminary study showed that this single RF well discriminates low grade panNET. Considering that tumour grade based on bioptical samples is not representative of the entire lesion, RFs may be used as tool to overcome this limitation, assigning a correct grade in all patients, especially in cases with borderline ki67. Limitations Reduced patients’ sample size and lack of grading assessment on tumour surgically excided.
ECR 2021 – BOOK OF ABSTRACTS
1
1
Silvia Malavasi, Diletta Calabrò, Alessandro Bevilacqua, Claudio Ricci, Riccardo Casadei, Davide Campana, Stefano Fanti, Valentina Ambrosini
File in questo prodotto:
Eventuali allegati, non sono esposti

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11585/785453
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact