Metabolic Alterations in Pancreatic Cancer Detected by In Vivo 1H-MR Spectroscopy: Correlation with Normal Pancreas, PET Metabolic Activity, Clinical Stages, and Survival Outcome

Objective: To compare the metabolites of in vivo 1H- MRS in pancreatic cancer with normal pancreas, and correlate these metabolites with Positron Emission Tomography (PET) metabolic activity, clinical stages, and survival outcomes. Methods: The prospective study included 58 patients (mean age 62.7 ± 12.1 years, range 34–81 years; 36 men, 22 women) with pathological proof of pancreatic adenocarcinoma, and all of them received 18F-fluorodeoxyglucose (FDG) PET/MRI before treatment. The single-voxel MRS with a point-resolved selective spectroscopy sequence was used to measure metabolites (creatine, Glx (glutamine and glutamate), N-acetylaspartate (NAA), and lipid) of pancreatic cancer and adjacent normal parenchyma, respectively. FDG-PET parameters included SUVmax, metabolic tumor volume (MTV), and total lesion glycolysis (TLG). Non-parametric tests were used to evaluate the differences of MRS metabolites between pancreatic cancer and those in normal pancreas, and their correlation with PET parameters and clinical stages. The correlation with progression-free survival (PFS) and overall survival (OS) was measured using the Kaplan–Meier and Cox proportional hazard models. Results: When compared with normal pancreas, the Glx, NAA, and lipid levels were significantly decreased in pancreatic cancer (all p < 0.05). Creatine, Glx, and lipid levels were all inversely correlated with both MTV (rho = −0.405~−0.454) and TLG (rho = −0.331~−0.441). For correlation with clinical stages, lower lipid levels were found in patients with T4 (vs. <T4, p = 0.038) and lower creatine levels were found in N1 (vs. N0, p = 0.019). Regarding survival outcomes, high TNM stage, low creatine, low Glx, and low lipid levels were associated with both poor PFS and OS (all p < 0.05). Additionally, creatine remained an independent factor for PFS and OS after adjusting for age, sex, tumor size, stages, and other metabolites levels. Conclusions: Decreased MRS metabolites in pancreatic cancer were associated with poor survival outcome, and may be used as prognostic image biomarkers for these patients.

Preoperative sarcopenia is associated with poor overall survival in pancreatic cancer patients following pancreaticoduodenectomy

OBJECTIVES

To analyze the effect of preoperative body composition on survival in patients with pancreatic cancer following pancreaticoduodenectomy (PD).

METHODS

Between October 2005 and August 2018, 116 patients (68 men, 48 women, mean age 66.2 ± 11.9 years) diagnosed with pancreatic adenocarcinoma following PD were retrospectively enrolled. The preoperative CT on vertebral level L3 was assessed for total abdominal muscle area (TAMA), visceral adipose tissue area (VAT), subcutaneous adipose tissue area (SAT), and mean skeletal muscle attenuation (SMD). The clinical data and pathological findings of tumors were collected. The impact of these factors on disease-free survival (DFS) and overall survival (OS) was evaluated by the Kaplan-Meier method and by univariable and multivariable Cox proportional hazards models.

RESULTS

The 3-year DFS and OS rates were 8% and 25%, respectively. Of 116 patients, 20 (17.2%), 3 (2.6%), and 46 (39.7%) patients were classified as having sarcopenia, sarcopenic obesity, and myosteatosis, respectively. The VAT-TAMA ratio (1.2 ± 0.7 vs 0.9 ± 0.5, p = 0.01) and the visceral to subcutaneous adipose tissue area ratio (1.3 ± 0.7 vs 0.9 ± 0.5, p = 0.04) were higher in sarcopenic patients than in the nonsarcopenic group. Preoperative sarcopenia and sarcopenic obesity were associated with shorter OS (p = 0.012 and p = 0.041, respectively), but not shorter DFS. Myosteatosis was neither associated with DFS nor OS. On multivariable analysis, sarcopenia was the only significant prognostic factor for OS (p = 0.039).

CONCLUSIONS

Preoperative sarcopenia assessed by CT is a poor prognostic factor for OS in pancreatic cancer patients after PD.

KEY POINTS

• Sarcopenia and sarcopenic obesity can be evaluated by abdominal CT on L3 level. • Patients with diabetes mellitus (DM) had lower sex-standardized subcutaneous adipose tissue area index and skeletal muscle density and higher visceral to subcutaneous adipose tissue area ratio than did those without DM. • Preoperative sarcopenia, sarcopenic obesity, and new-onset diabetes mellitus may predict poor overall survival in pancreatic cancer patients following pancreaticoduodenectomy.

Artificial intelligence in pancreatic disease

In recent years, the application of artificial intelligence (AI) in radiology has been growing rapidly, fueled by the availability of large datasets, advances in computing power, and newly developed algorithms. Progress in AI applied to medical imaging analyses has transformed these images into quantitative data, termed radiomics. When combined with patients’ clinical data, these models, when developed by machine learning, have the potential to improve diagnostic, prognostic, and predictive accuracy. Currently, limited literature is available on the use of radiomics for pancreatic disease. Here, we will review recent studies in the application of AI in a variety of pancreatic diseases, mainly involving lesion detection, tumor characterization, tumor grading, response, and prognosis evaluation. Finally, we will also discuss the challenges and prospects in the field of radiomics for pancreatic disease.

AB004. S1-3. Biliary imaging in diagnosis and staging

Preoperative staging is the prime concern when evaluating the possibility of surgical resectability in cholangiocarcinoma. Currently, multidetector computed tomography (MDCT) is the essential primary step to assess the local extension of the tumor and it should be undertaken prior to biliary drainage to prevent modifications of the bile duct wall thickness or enhancement by a drainage catheter mimicking the tumor extension. Tumor invasion of the liver parenchyma, hepatic vein and inferior vena cava (IVC), or intrahepatic metastasis can be assessed by MDCT and magnetic resonance imaging (MRI). In addition, lymph node metastasis is evaluated mainly by MDCT, and fluorodeoxyglucose PET scan is sometimes informative to assess distant and/or lymph node metastasis. In terms of recent advances in imaging techniques, MDCT and 3D CT angiography have replaced conventional angiography to assess the degree of vascular involvement and to delineate the vascular anatomy in each individual cholangiocarcinoma case. The characteristic MR findings include hypointensity on T1-weighted images, hyperintensity on T2-weighted images, the presence or absence of a central scar, ductal dilatation, satellite nodules, portal vein invasion, and lobar atrophy. Magnetic resonance cholangiopancreatography (MRCP) is useful for a comprehensive understanding of the biliary tree as a whole; however, it is still unable to diagnose the intricate local anatomy of the separated intrahepatic segmental ducts and to design an appropriate operative procedure in patients with Bismuth type III or IV cholangiocarcinoma. MRI and MRCP remain mandatory to define resectability of perihilar cholangiocarcinomas. A recent study found gadoxetic acid–enhanced 3-T MR imaging provides high per-lesion sensitivity and PPV for preoperative malignant liver lesion detection overall, although sensitivity for 0.2–0.5 cm malignant lesions is poor. The recently introduced simultaneous PET/MRI combines the metabolic imaging capabilities of PET and the unmatched soft tissue contrast of MRI and appears to be a promising modality for oncologic imaging. Simultaneous acquisition provides better spatial registration compared to the sequential acquisition of PET and CT. PET/MRI is expected not only to improve the diagnostic potential but also holds promise for better delineation of the target volume in staging and surgical planning.

Multiparametric PET/MR imaging biomarkers are associated with overall survival in patients with pancreatic cancer

PURPOSE

To correlate the overall survival (OS) with the imaging biomarkers of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), diffusion-weighted imaging (DWI), magnetic resonance spectroscopy, and glucose metabolic activity derived from integrated fluorine 18 fluorodeoxyglucose positron emission tomography (¹⁸F-FDG PET)/MRI in patients with pancreatic cancer.

METHODS

This prospective study was approved by the institutional review board and informed consent was obtained from all participants. Sixty-three consecutive patients (mean age, 62.7 ± 12 y; men/women, 40/23) with pancreatic cancer underwent PET/MRI before treatment. The imaging biomarkers were comprised of DCE-MRI parameters (peak, IAUC ₆₀ , K ^trans , k _ep , v _e ), the minimum apparent diffusion coefficient (ADC_min), choline level, standardized uptake values, metabolic tumor volume, and total lesion glycolysis (TLG) of the tumors. The relationships between these imaging biomarkers with OS were evaluated with the Kaplan-Meier and Cox proportional hazard models.

RESULTS

Seventeen (27%) patients received curative surgery, with the median follow-up duration being 638 days. Univariate analysis showed that patients at a low TNM stage (≦3, P = 0.041), high peak (P = 0.006), high ADC_min (P = 0.002) and low TLG (P = 0.01) had better OS. Moreover, high TLG/peak ratio was associated with poor OS (P = 0.016). Multivariate analysis indicated that ADC_min (P = 0.011) and TLG/peak ratio (P = 0.006) were independent predictors of OS after adjustment for age, gender, tumor size, and TNM stage. The TLG/peak ratio was an independent predictor of OS in a subgroup of patients who did not receive curative surgery (P = 0.013).

CONCLUSION

The flow-metabolism mismatch reflected by the TLG/peak ratio may better predict OS than other imaging biomarkers from PET/MRI in pancreatic cancer patients.