The value of dual probes in tracerpositron emission tomography/magnetic resonance imaging in the grading diagnosis of pancreatic neuroendocrine neoplasms
白亚亚,黄新韵,孟宏平,王思文,张敏,金佳斌,李彪,林晓珠
DOI: https://doi.org/10.3760/cma.j.cn311367-20220315-00115
2022-01-01
Abstract:Objective:To explore the diagnostic and grading value of combination of 68Ga -1, 4, 7, 10-tetraazacyclododecane-1, 4, 7, 10-tetraacetic acid- D-Phe1-Tyr3-Thr8-octreotide ( 68Ga-DOTA-TATE) and 18F-flurodeoxyglucose ( 18F-FDG) dual probes in multi-parameter positron emission tomography (PET)/magnetic resonance (MR) imaging in pancreatic neuroendocrine neoplasm (PNEN). Methods:From April 9th, 2020 to February 24th, 2022, in Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine, the clinical data and the imaging of 68Ga-DOTA-TATE PET/MR and 18F-FDG PET/MR of 59 patients with pancreatic tumors (27 male, 32 female, aged 22 to 75 years old(51.8±13.3) years old), confirmed by surgical or biopsy pathology were retrospectively analyzed. All the cases were divided into PNEN group (42 cases) and non-PNEN group (17 cases) according to pathological results. Among which 39 patients with PNET were further divided into grade 1 group (G1 group, 27 cases) and grade 2 group (G2 group, 12 cases). Non-zero parameters were selected via the least absolute shrinkage and selection operator (LASSO) regression approach, and a logistic regression model was established by combination of the selected features and the corresponding non-zero coefficients. The measurement data with non-normal distribution were compared by Mann-Whitney U test. The receiver operating characteristic (ROC) curve were used to detemine the optimal cut off value to assess the dignostic efficiency. Results:Compared with those of non-PNEN group, the parameters of PNEN group increased, which included maximum standard uptake value of 68Ga-DOTA-TATE(SUV Gmax, 46.70 (22.37, 76.35) vs. 7.12 (4.75, 8.64)), mean standard uptake value of 68Ga-DOTA-TATE(SUV Gmean, 25.50 (13.18, 43.90) vs. 3.65 (2.89, 4.69)), peak standard uptake value of 68Ga-DOTA-TATE (SUV Gpeak, 27.17 (12.39, 46.97) vs. 5.46 (4.12, 6.56)), total lesion somatostatin receptor (SSR) expression (TLSRE, 68.21 (32.52, 440.96) vs. 26.02 (14.87, 69.57)), SUV Gmax/maximum standard uptake value of 18F-FDG (SUV Fmax, 12.71 (3.80, 21.70) vs. 1.10 (0.52, 2.35)), tumor to background ratio of 68Ga-DOTA-TATE (TBR G, 13.31 (5.54, 22.38) vs. 1.57 (1.31, 2.66)), tumor to liver ratio of 68Ga-DOTA-TATE(T/L G, 6.54 (2.90, 9.63) vs. 0.74 (0.65, 0.94)), tumor to spleen ratio of 68Ga-DOTA-TATE (T/S G, 2.36 (0.97, 3.70) vs. 0.25 (0.23, 0.38)), tumor to mediastinum ratio of 68Ga-DOTA-TATE (T/M G, 104.41 (34.03, 206.52) vs. 16.00 (12.87, 21.46)), SUV Gmax/minimum apparent diffusion coeffecient (ADC min, 55.14 (22.50, 96.37) vs. 6.76 (4.39, 12.76)) and SUV Gmean/ADC min (34.57 (13.47, 55.13) vs. 3.57 (2.46, 6.81)), and the differences were statistically significant ( U=28.00, 25.00, 32.00, 198.00, 54.00, 31.00, 28.00, 19.00, 10.00, 56.00 and 44.00, all P<0.01). The area under the curve (AUC) and diagnostic accuracy of dual-probe PET/MR imaging in the diagnosis of PNEN and non-PNEN were 0.941 and 96.6%, respectively. The AUC and diagnostic accuracy of model Y 1 in the diagnosis of PNEN and non-PNEN were 0.959 and 96.6%, respectively. There was no significant difference in AUC between model Y 1 and dual-probe PET/MR imaging in PNEN diagnosis ( P>0.05), however combining model Y 1 could improve the accuracy of PNEN diagnosis (100.0%). Compared with those of PNET G1 group, the parameters of G2 Group were higher, which included the maximum diameter of tumor (2.69 cm (2.08 cm, 5.00 cm) vs. 1.50 cm (1.20 cm, 2.50 cm)), metabolic tumor volume (MTV, 7.56 mL (4.45 mL, 53.57 mL) vs. 2.16 mL (1.22 mL, 5.48 mL)), total lesion glycolysis (TLG, 22.24 (11.95, 189.85) vs. 3.81 (2.11, 18.67)), tumor to background ratio of 18F-FDG (TBR F, 2.94 (2.00, 3.96) vs. 1.48 (1.29, 3.72)), tumor to liver ratio of 18F-FDG (T/L F, 2.32 (1.35, 2.98) vs. 1.08 (0.90, 2.17)) and SSR-expressing tumor volume (SRETV, 8.00 (3.06, 40.00) vs. 1.91 (0.95, 4.88)), and the differences were statistically significant ( U=66.00、66.00、77.00、93.00、90.00、65.50, all P<0.05). The maximum diameter of tumor was the best single parameter for the differential diagnosis of PNET G2 and G1, AUC was 0.796 and the cutoff value was 1.90 cm. The model Y 2, which combined the maximum diameter of tumor and TBR G had an AUC of 0.835 for the differential diagnosis of PNET G2 and G1. There was no significant difference in AUC between the maximum diameter of tumor and model Y 2 ( P>0.05). However the combination of the maximum diameter of tumor and model Y 2 could improve the accuracy of differential diagnosis of PNET G2 and G1 (94.87%). Conclusion:The combination of multi-parameter of 68Ga-DOTA-TATE and dual-probe 18F-FDG PET/MR imaging can improve the diagnostic and grading accuracy of PNEN, which may be helpful in the selection of clinical treatment for patients.