Textural analysis on contrast-enhanced CT in pancreatic neuroendocrine neoplasms: association with WHO grade

Predicting histologic grades for pancreatic neuroendocrine tumors by radiologic image-based artificial intelligence: a systematic review and meta-analysis

Background

Accurate detection of the histological grade of pancreatic neuroendocrine tumors (PNETs) is important for patients' prognoses and treatment. Here, we investigated the performance of radiological image-based artificial intelligence (AI) models in predicting histological grades using meta-analysis.

Method

A systematic literature search was performed for studies published before September 2023. Study characteristics and diagnostic measures were extracted. Estimates were pooled using random-effects meta-analysis. Evaluation of risk of bias was performed by the QUADAS-2 tool.

Results

A total of 26 studies were included, 20 of which met the meta-analysis criteria. We found that the AI-based models had high area under the curve (AUC) values and showed moderate predictive value. The pooled distinguishing abilities between different grades of PNETs were 0.89 [0.84-0.90]. By performing subgroup analysis, we found that the radiomics feature-only models had a predictive value of 0.90 [0.87-0.92] with I2 = 89.91%, while the pooled AUC value of the combined group was 0.81 [0.77-0.84] with I2 = 41.54%. The validation group had a pooled AUC of 0.84 [0.81-0.87] without heterogenicity, whereas the validation-free group had high heterogenicity (I2 = 91.65%, P=0.000). The machine learning group had a pooled AUC of 0.83 [0.80-0.86] with I2 = 82.28%.

Conclusion

AI can be considered as a potential tool to detect histological PNETs grades. Sample diversity, lack of external validation, imaging modalities, inconsistent radiomics feature extraction across platforms, different modeling algorithms and software choices were sources of heterogeneity. Standardized imaging, transparent statistical methodologies for feature selection and model development are still needed in the future to achieve the transformation of radiomics results into clinical applications.

Systematic Review Registration

https://www.crd.york.ac.uk/prospero/, identifier CRD42022341852.

Neuroendocrine Carcinoma of the Urinary Bladder: CT Findings and Radiomics Signature

Background: We present a case series of Neuroendocrine Carcinoma of the Urinary Bladder (NECB) to analyse their radiologic appearance on CT, find a "Radiomic signature", and review the current literature. Methods: 14 CT cases of NECB were reviewed and compared with a control group of 42 patients with high-grade non-neuroendocrine bladder neoplasm for the following parameters: ring enhancement; implantation site; dimensions; density; margins; central necrosis; calcifications; number of lesions; wall thickness; depth of invasion in the soft tissue; invasion of fat tissue; invasion of adjacent organs; lymph-node involvement; abdominal organ metastasis. To extract radiomic features, volumes of interest of bladder lesions were manually delineated on the portal-venous phase. The radiomic features of the two groups were identified and compared. Results: Statistical differences among NECB and control group were found in the prevalence of male sex (100% vs. 69.0%), hydronephrosis (71.4% vs. 33.3%), mean density of the mass (51.01 ± 15.48 vs. 76.27 ± 22.26 HU); product of the maximum diameters on the axial plane (38.1 ± 59.3 vs. 14.44 ± 12.98 cm2) in the control group, trigonal region involvement (78.57% vs. 19.05%). About the radiomic features, Student's t-test showed significant correlation for the variables: "DependenceNonUniformity" (p: 0.048), "JointAverage" (p: 0.013), "LargeAreaLowGrayLevelEmphasis" (p: 0.014), "Maximum2DDiameterColumn" (p: 0.04), "Maximum 2DDiameterSlice" (p: 0.007), "MeanAbsoluteDeviation" (p: 0.021), "BoundingBoxA" (p: 0.022) and "CenterOfMassB" (p: 0.007). Conclusions: There is a typical pattern (male patient, large mass, trigonal area involvement) of NECB presentation on contrast-enhanced CT. Certain morphological characteristics and encouraging results about Radiomic features can help define the diagnosis.

Prognostic Assessment of Gastropancreatic Neuroendocrine Neoplasm: Prospects and Limits of Radiomics

Neuroendocrine neoplasms (NENs) are a group of lesions originating from cells of the diffuse neuroendocrine system. NENs may involve different sites, including the gastrointestinal tract (GEP-NENs). The incidence and prevalence of GEP-NENs has been constantly rising thanks to the increased diagnostic power of imaging and immuno-histochemistry. Despite the plethora of biochemical markers and imaging techniques, the prognosis and therapeutic choice in GEP-NENs still represents a challenge, mainly due to the great heterogeneity in terms of tumor lesions and clinical behavior. The concept that biomedical images contain information about tissue heterogeneity and pathological processes invisible to the human eye is now well established. From this substrate comes the idea of radiomics. Computational analysis has achieved promising results in several oncological settings, and the use of radiomics in different types of GEP-NENs is growing in the field of research, yet with conflicting results. The aim of this narrative review is to provide a comprehensive update on the role of radiomics on GEP-NEN management, focusing on the main clinical aspects analyzed by most existing reports: predicting tumor grade, distinguishing NET from other tumors, and prognosis assessment.

Prediction of Pathological Grades of Pancreatic Neuroendocrine Tumors Based on Dynamic Contrast-Enhanced Ultrasound Quantitative Analysis

Objective: To investigate whether the dynamic contrast-enhanced ultrasound (DCE-US) analysis and quantitative parameters could be helpful for predicting histopathologic grades of pancreatic neuroendocrine tumors (pNETs). Methods: This retrospective study conducted a comprehensive review of the CEUS database between March 2017 and November 2021 in Zhongshan Hospital, Fudan University. Ultrasound examinations were performed by an ACUSON Sequioa unit equipped with a 3.5 MHz 6C−1 convex array transducer, and an ACUSON OXANA2 unit equipped with a 3.5 MHz 5C−1 convex array transducer. SonoVue® (Bracco Inc., Milan, Italy) was used for all CEUS examinations. Time intensity curves (TICs) and quantitative parameters of DCE-US were created by Vuebox® software (Bracco, Italy). Inclusion criteria were: patients with histopathologically proved pNETs, patients who underwent pancreatic B-mode ultrasounds (BMUS) and CEUS scans one week before surgery or biopsy and had DCE-US imaging documented for more than 2 min, patients with solid or predominantly solid lesions and patients with definite diagnosis of histopathological grades of pNETs. Based on their prognosis, patients were categorized into two groups: pNETs G1/G2 group and pNETs G3/pNECs group. Results: A total of 42 patients who underwent surgery (n = 38) or biopsy (n = 4) and had histopathologically confirmed pNETs were included. According to the WHO 2019 criteria, all pNETs were classified into grade 1 (G1, n = 10), grade 2 (G2, n = 21), or grade 3 (G3)/pancreatic neuroendocrine carcinomas (pNECs) (n = 11), based on the Ki−67 proliferation index and the mitotic activity. The majority of the TICs (27/31) of pNETs G1/G2 were above or equal to those of pancreatic parenchyma in the arterial phase, but most (7/11) pNETs G3/pNECs had TICs below those of pancreatic parenchyma from arterial phase to late phase (p < 0.05). Among all the CEUS quantitative parameters of DCE-US, values of relative rise time (rPE), relative mean transit time (rmTT) and relative area under the curve (rAUC) were significantly higher in pNETs G1/G2 group than those in pNETs G3/pNECs group (p < 0.05). Taking an rPE below 1.09 as the optimal cut-off value, the sensitivity, specificity and accuracy for prediction of pNETs G3/pNECs from G1/G2 were 90.91% [58.70% to 99.80%], 67.64% [48.61% to 83.32%] and 85.78% [74.14% to 97.42%], respectively. Taking rAUC below 0.855 as the optimal cut-off value, the sensitivity, specificity and accuracy for prediction of pNETs G3/pNECs from G1/G2 were 90.91% [66.26% to 99.53%], 83.87% [67.37% to 92.91%] and 94.72% [88.30% to 100.00%], respectively. Conclusions: Dynamic contrast-enhanced ultrasound analysis might be helpful for predicting the pathological grades of pNETs. Among all quantitative parameters, rPE, rmTT and rAUC are potentially useful parameters for predicting G3/pNECs with aggressive behavior.

Nuclear Medicine and Radiological Imaging of Pancreatic Neuroendocrine Neoplasms: A Multidisciplinary Update

Pancreatic neuroendocrine neoplasms (panNENs) are part of a large family of tumors arising from the neuroendocrine system. PanNENs show low-intermediate tumor grade and generally high somatostatin receptor (SSTR) expression. Therefore, panNENs benefit from functional imaging with ⁶⁸Ga-somatostatin analogues (SSA) for diagnosis, staging, and treatment choice in parallel with morphological imaging. This narrative review aims to present conventional imaging techniques and new perspectives in the management of panNENs, providing the clinicians with useful insight for clinical practice. The ⁶⁸Ga-SSA PET/CT is the most widely used in panNENs, not only fr diagnosis and staging purpose but also to characterize the biology of the tumor and its responsiveness to SSAs. On the contrary, the ¹⁸F-Fluordeoxiglucose (FDG) PET/CT is not employed systematically in all panNEN patients, being generally preferred in G2-G3, to predict aggressiveness and progression rate. The combination of ⁶⁸Ga-SSA PET/CT and ¹⁸F-FDG PET/CT can finally suggest the best therapeutic strategy. Other radiopharmaceuticals are ⁶⁸Ga-exendin-4 in case of insulinomas and ¹⁸F-dopamine (DOPA), which can be helpful in SSTR-negative tumors. New promising but still-under-investigation radiopharmaceuticals include radiolabeled SSTR antagonists and ¹⁸F-SSAs. Conventional imaging includes contrast enhanced CT and multiparametric MRI. There are now enriched by radiomics, a new non-invasive imaging approach, very promising to early predict tumor response or progression.

New frontiers in imaging including radiomics updates for pancreatic neuroendocrine neoplasms

OBJECTIVE

To illustrate the applications of various imaging tools including conventional MDCT, MRI including DWI, CT & MRI radiomics, FDG & DOTATATE PET-CT for diagnosis, staging, grading, prognostication, treatment planning and assessing treatment response in cases of pancreatic neuroendocrine neoplasms (PNENs).

BACKGROUND

Gastroenteropancreatic neuroendocrine neoplasms (GEP NENs) are very diverse clinically & biologically. Their treatment and prognosis depend on staging and primary site, as well as histological grading, the importance of which is also reflected in the recently updated WHO classification of GEP NENs. Grade 3 poorly differentiated neuroendocrine carcinomas (NECs) are aggressive & nearly always advanced at diagnosis with poor prognosis; whereas Grades-1 and 2 well-differentiated neuroendocrine tumors (NETs) can be quite indolent. Grade 3 well-differentiated NETs represent a new category of neoplasm with an intermediate prognosis. Importantly, the evidence suggest grade heterogeneity can occur within a given tumor and even grade progression can occur over time. Emerging evidence suggests that several non-invasive qualitative and quantitative imaging features on CT, dual-energy CT (DECT), MRI, PET and somatostatin receptor imaging with new tracers, as well as texture analysis, may be useful to grade, prognosticate, and accurately stage primary NENs. Imaging features may also help to inform choice of treatment and follow these neoplasms post-treatment.

CONCLUSION

GEP NENs treatment and prognosis depend on the stage as well as histological grade of the tumor. Traditional ways of imaging evaluation for diagnosis and staging does not yet yield sufficient information to replace operative and histological evaluation. Recognition of important qualitative imaging features together with quantitative features and advanced imaging tools including functional imaging with DWI MRI, DOTATATE PET/CT, texture analysis with radiomics and radiogenomic features appear promising for more accurate staging, tumor risk stratification, guiding management and assessing treatment response.

Update on quantitative radiomics of pancreatic tumors

Radiomics is a newer approach for analyzing radiological images obtained from conventional imaging modalities such as computed tomography, magnetic resonance imaging, endoscopic ultrasonography, and positron emission tomography. Radiomics involves extracting quantitative data from the images and assessing them to identify diagnostic or prognostic features such as tumor grade, resectability, tumor response to neoadjuvant therapy, and survival. The purpose of this review is to discuss the basic principles of radiomics and provide an overview of the current clinical applications of radiomics in the field of pancreatic tumors.

Artificial Intelligence Applied to Pancreatic Imaging: A Narrative Review

The diagnosis, evaluation, and treatment planning of pancreatic pathologies usually require the combined use of different imaging modalities, mainly, computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET). Artificial intelligence (AI) has the potential to transform the clinical practice of medical imaging and has been applied to various radiological techniques for different purposes, such as segmentation, lesion detection, characterization, risk stratification, or prediction of response to treatments. The aim of the present narrative review is to assess the available literature on the role of AI applied to pancreatic imaging. Up to now, the use of computer-aided diagnosis (CAD) and radiomics in pancreatic imaging has proven to be useful for both non-oncological and oncological purposes and represents a promising tool for personalized approaches to patients. Although great developments have occurred in recent years, it is important to address the obstacles that still need to be overcome before these technologies can be implemented into our clinical routine, mainly considering the heterogeneity among studies.

GEP-NET radiomics: a systematic review and radiomics quality score assessment

OBJECTIVE

The number of radiomics studies in gastroenteropancreatic neuroendocrine tumours (GEP-NETs) is rapidly increasing. This systematic review aims to provide an overview of the available evidence of radiomics for clinical outcome measures in GEP-NETs, to understand which applications hold the most promise and which areas lack evidence.

METHODS

PubMed, Embase, and Wiley/Cochrane Library databases were searched and a forward and backward reference check of the identified studies was executed. Inclusion criteria were (1) patients with GEP-NETs and (2) radiomics analysis on CT, MRI or PET. Two reviewers independently agreed on eligibility and assessed methodological quality with the radiomics quality score (RQS) and extracted outcome data.

RESULTS

In total, 1364 unique studies were identified and 45 were included for analysis. Most studies focused on GEP-NET grade and differential diagnosis of GEP-NETs from other neoplasms, while only a minority analysed treatment response or long-term outcomes. Several studies were able to predict tumour grade or to differentiate GEP-NETs from other lesions with a good performance (AUCs 0.74-0.96 and AUCs 0.80-0.99, respectively). Only one study developed a model to predict recurrence in pancreas NETs (AUC 0.77). The included studies reached a mean RQS of 18%.

CONCLUSION

Although radiomics for GEP-NETs is still a relatively new area, some promising models have been developed. Future research should focus on developing robust models for clinically relevant aims such as prediction of response or long-term outcome in GEP-NET, since evidence for these aims is still scarce.

KEY POINTS

• The majority of radiomics studies in gastroenteropancreatic neuroendocrine tumours is of low quality. • Most evidence for radiomics is available for the identification of tumour grade or differentiation of gastroenteropancreatic neuroendocrine tumours from other neoplasms. • Radiomics for the prediction of response or long-term outcome in gastroenteropancreatic neuroendocrine tumours warrants further research.

Advances in the imaging of gastroenteropancreatic neuroendocrine neoplasms

Gastroenteropancreatic neuroendocrine neoplasms comprise a heterogeneous group of tumors that differ in their pathogenesis, hormonal syndromes produced, biological behavior and consequently, in their requirement for and/or response to specific chemotherapeutic agents and molecular targeted therapies. Various imaging techniques are available for functional and morphological evaluation of these neoplasms and the selection of investigations performed in each patient should be customized to the clinical question. Also, with the increased availability of cross sectional imaging, these neoplasms are increasingly being detected incidentally in routine radiology practice. This article is a review of the various imaging modalities currently used in the evaluation of neuroendocrine neoplasms, along with a discussion of the role of advanced imaging techniques and a glimpse into the newer imaging horizons, mostly in the research stage.

CT-based radiomics for prediction of therapeutic response to Everolimus in metastatic neuroendocrine tumors

Aim

To test radiomic approach in patients with metastatic neuroendocrine tumors (NETs) treated with Everolimus, with the aim to predict progression-free survival (PFS) and death.

Materials and methods

Twenty-five patients with metastatic neuroendocrine tumors, 15/25 pancreatic (60%), 9/25 ileal (36%), 1/25 lung (4%), were retrospectively enrolled between August 2013 and December 2020. All patients underwent contrast-enhanced CT before starting Everolimus, histological diagnosis, tumor grading, PFS, overall survival (OS), death, and clinical data collected. Population was divided into two groups: responders (PFS ≤ 11 months) and non-responders (PFS > 11 months). 3D segmentation was performed on whole liver of naïve CT scans in arterial and venous phases, using a dedicated software (3DSlicer v4.10.2). A total of 107 radiomic features were extracted and compared between two groups (T test or Mann–Whitney), radiomics performance assessed with receiver operating characteristic curve, Kaplan–Meyer curves used for survival analysis, univariate and multivariate logistic regression performed to predict death, and interobserver variability assessed. All significant radiomic comparisons were validated by using a synthetic external cohort. P < 0.05 is considered significant.

Results

15/25 patients were classified as responders (median PFS 25 months and OS 29 months) and 10/25 as non-responders (median PFS 4.5 months and OS 23 months). Among radiomic parameters, Correlation and Imc1 showed significant differences between two groups (P < 0.05) with the best performance (internal cohort AUC 0.86–0.84, P < 0.0001; external cohort AUC 0.84–0.90; P < 0.0001). Correlation < 0.21 resulted correlated with death at Kaplan–Meyer analysis (P = 0.02). Univariate analysis showed three radiomic features independently correlated with death, and in multivariate analysis radiomic model showed good performance with AUC 0.87, sensitivity 100%, and specificity 66.7%. Three features achieved 0.77 ≤ ICC < 0.83 and one ICC = 0.92.

Conclusions

In patients affected by metastatic NETs eligible for Everolimus treatment, radiomics could be used as imaging biomarker able to predict PFS and death.

Current updates and future directions in diagnosis and management of gastroenteropancreatic neuroendocrine neoplasms

Gastroenteropancreatic neuroendocrine neoplasms are a heterogenous group of rare neoplasms that are increasingly being discovered, often incidentally, throughout the gastrointestinal tract with varying degrees of activity and malignant potential. Confusing nomenclature has added to the complexity of managing these lesions. The term carcinoid tumor and embryonic classification have been replaced with gastroenteropancreatic neuroendocrine neoplasm, which includes gastrointestinal neuroendocrine and pancreatic neuroendocrine neoplasms. A comprehensive multidisciplinary approach is important for clinicians to diagnose, stage and manage these lesions. While histological diagnosis is the gold standard, recent advancements in endoscopy, conventional imaging, functional imaging, and serum biomarkers complement histology for tailoring specific treatment options. In light of developing technology, our review sets out to characterize diagnostic and therapeutic advancements for managing gastroenteropancreatic neuroendocrine tumors, including innovations in radiolabeled peptide imaging, circulating biomarkers, and endoscopic treatment approaches adapted to different locations throughout the gastrointestinal system.

Utility of Quantitative Metrics from Dual-Layer Spectral-Detector CT for Differentiation of Pancreatic Neuroendocrine Tumor and Neuroendocrine Carcinoma

Background: The 2019 WHO classification separates neuroendocrine neoplasms (NENs) into neuroendocrine tumors (NET) and neuroendocrine carcinomas (NEC), which are considered to represent pathologically distinct entities warranting different management approaches. Dual-layer spectral-detector CT (DLCT) may aid their differentiation through specific material decomposition. Objective: To assess the utility of quantitative metrics derived from DLCT for the differentiation of pancreatic NET and NEC. Methods: This retrospective study included 104 patients (mean age 51±13 years; 53 women, 51 men) with pathologically confirmed NEN [89 NET, including 22 grade 1, 48 grade 2, and 19 grade 3 (G3); 15 NEC], who underwent multiphase DLCT within 15 days before biopsy or resection. Two radiologists independently placed ROIs to record tumor attenuation, iodine concentration (IC), and effective atomic number (Zeff) across phases, and also assessed qualitative features (composition, homogeneity, margins, calcifications, main pancreatic duct dilation, vascular invasion, lymphadenopathy). Interreader agreement was assessed. Mean values between readers were obtained for quantitative measures; consensus was reached for qualitative features. NET and NEC were compared using multivariable regression analysis and ROC analysis. Results: Interobserver agreement, expressed as intraclass correlation coefficients, ranged from 0.879 to 0.992 for quantitative metrics, and, expressed as kappa coefficients, from 0.763 to 0.823 for qualitative features. In multivariable analysis of qualitative and quantitative features, significant independent predictors of NEC (P<.05) were IC in portal venous phase (1.3 mg/mL in NEC vs 2.7 mg/mL in NET), Zeff in portal venous phase (8.1 vs 8.6), and attenuation in portal venous phase (78.2 vs 113.5 HU). AUC for predicting NEC was 0.897 for IC, 0.884 for Zeff, 0.921 for combination of IC and Zeff, and 0.855 for attenuation. Predicted probability based on combination of IC and Zeff achieved sensitivity of 93.3% and specificity of 80.9% for NEC. Significant independent predictors (P<.05) for differentiating G3 NET and NEC were IC (1.3 vs 2.0 mg/mL; AUC=0.789) and attenuation (78.2 vs 90.3 HU; AUC=0.647), both measured in portal venous phase. Conclusion: Incorporation of DLCT-metrics improves differentiation of NET and NEC compared with conventional CT attenuation and qualitative features. Clinical Impact: DLCT may help select patients with pancreatic NENs for platinum-based chemotherapies.

Retrospective Analysis of the Value of Enhanced CT Radiomics Analysis in the Differential Diagnosis Between Pancreatic Cancer and Chronic Pancreatitis

Purpose

To investigate the feasibility of enhanced computed tomography (CT) radiomics analysis to differentiate between pancreatic cancer (PC) and chronic pancreatitis.

Methods and materials

The CT images of 151 PCs and 24 chronic pancreatitis were retrospectively analyzed in the three-dimensional regions of interest on arterial phase (AP) and venous phase (VP) and segmented by MITK software. A multivariable logistic regression model was established based on the selected radiomics features. The radiomics score was calculated, and the nomogram was established. The discrimination of each model was analyzed by the receiver operating characteristic curve (ROC). Decision curve analysis (DCA) was used to evaluate clinical utility. The precision recall curve (PRC) was used to evaluate whether the model is affected by data imbalance. The Delong test was adopted to compare the diagnostic efficiency of each model.

Results

Significant differences were observed in the distribution of gender (P = 0.034), carbohydrate antigen 19-9 (P < 0.001), and carcinoembryonic antigen (P < 0.001) in patients with PC and chronic pancreatitis. The area under the ROC curve (AUC) value of AP multivariate regression model, VP multivariate regression model, AP combined with VP features model (Radiomics), clinical feature model, and radiomics combined with clinical feature model (COMB) was 0.905, 0.941, 0.941, 0.822, and 0.980, respectively. The sensitivity and specificity of the COMB model were 0.947 and 0.917, respectively. The results of DCA showed that the COMB model exhibited net clinical benefits and PRC shows that COMB model have good precision and recall (sensitivity).

Conclusion

The COMB model could be a potential tool to distinguish PC from chronic pancreatitis and aid in clinical decisions.

Nomogram for preoperative estimation of histologic grade in gastrointestinal neuroendocrine tumors

BACKGROUND

The treatment strategies and prognosis for gastroenteropancreatic neuroendocrine tumors were associated with tumor grade. Preoperative predictive grading could be of great benefit in the selection of treatment options for patients. However, there is still a lack of effective non-invasive strategies to detect gastrointestinal neuroendocrine tumors (GI-NETs) grading preoperatively.

METHODS

The data on 147 consecutive GI-NETs patients was retrospectively collected from January 1, 2012, to December 31, 2019. Logistic regression was used to construct a predictive model of gastrointestinal neuroendocrine tumor grading using preoperative laboratory and imaging parameters.The validity of the model was assessed by area under the receiver operating characteristic curve (AUC), calibration curve, and decision curve analysis (DCA).

RESULTS

The factors associated with GI-NETs grading were age, tumor size, lymph nodes, neuron-specific enolase (NSE), hemoglobin (HGB) and sex, and two models were constructed by logistic regression for prediction. Combining these 6 factors, the nomogram was constructed for model 1 to distinguish between G3 and G1/2, achieving a good AUC of 0.921 (95% CI: 0.884-0.965), and the sensitivity, specificity, accuracy were 0.9167, 0.8256, 0.8630, respectively. The model 2 was to distinguish between G1 and G2/3, and the variables were age, tumor size, lymph nodes, NSE, with an AUC of 0.847 (95% CI: 0.799-0.915), and the sensitivity, specificity, accuracy were 0.7882, 0.8710, 0.8231, respectively. Two online web servers were established on the basis of the proposed nomogram to facilitate clinical use. Both models showed an excellent calibration curve through 1000 times bootstrapped dataset and the clinical usefulness were confirmed using decision curve analysis.

CONCLUSION

The model served as a valuable non-invasive tool for differentiating between different grades of GI-NETs, personalizing the calculation which can lead to a rational treatment choice.

Enhanced computed tomography features predict pancreatic neuroendocrine neoplasm with Ki-67 index less than 5

BACKGROUND

Several studies have suggested that patients with pancreatic neuroendocrine neoplasm (pNEN) with the Ki-67 index of < 5% are more likely to show better prognosis after clinical intervention. Moreover, the Ki-67 index at 5% has also been suggested as a potential threshold by the 2016 European Neuroendocrine Tumor Society guidelines.

OBJECTIVE

Based on preoperative enhanced computed tomography (CT), this study aimed to investigate imaging characteristics eligible to discriminate the ≤ 5% Ki-67 group from the > 5% Ki-67 group of patients with nonmetastatic pNEN.

METHODS

Patients with pathologically diagnosed pNEN and preoperative multiphase CT were enrolled. Their Ki-67 index was calculated and grouped according to the 5% cutoff value. The following CT imaging characteristics and some serum biomarkers were assessed between the two groups: the diameter, location, tumor margin, calcification, pancreatic atrophy, distal pancreatic duct dilation, vessel involvement, and enhancement pattern characteristics of both arterial phase (AP) and portal vein phase (PVP).

RESULTS

A total of 142 patients with pNEN were enrolled in this study, comprising 104 in the low (Ki-67, 1%-5%) and 38 in the high index group (Ki-67, >5%). Alpha fetoprotein and cancer antigen 125 were significantly different between the two groups (P-values, 0.030 and 0.049, respectively). The diameter (P < 0.0001), margin (P = 0.003), distal main ductal dilation (P = 0.021), vessel involvement (P = 0.002), AP hypoenhancement (P < 0.0001), PVP hypoenhancement (P = 0.003), AP ratio (P = 0.0001), and PVP ratio (P = 0.0003) were significantly different between the low and high index groups. The area under the curve of the multivariate logistic regression model was 0.853.

CONCLUSION

Nonmetastatic pNENs with larger diameter, ill-defined margin, distal main ductal dilation, and tumor hypoenhancement in AP in preoperative enhanced CT tend to have a Ki-67 index of > 5%.The results of this study provide an alternative method to clinicians to decide whether surgery is appropriate.

Current update on imaging for pancreatic neuroendocrine neoplasms

Pancreatic neuroendocrine neoplasms (panNEN) are a heterogeneous group of tumors with differing pathological, genetic, and clinical features. Based on clinical findings, they may be categorized into functioning and nonfunctioning tumors. Adoption of the 2017 World Health Organization classification system, particularly its differentiation between grade 3, well-differentiated pancreatic neuroendocrine tumors (panNET) and grade 3, poorly-differentiated pancreatic neuroendocrine carcinomas (panNEC) has emphasized the role imaging plays in characterizing these lesions. Endoscopic ultrasound can help obtain biopsy specimen and assess tumor margins and local spread. Enhancement patterns on computed tomography (CT) and magnetic resonance imaging (MRI) may be used to classify panNEN. Contrast enhanced MRI and diffusion-weighted imaging have been reported to be useful for characterization of panNEN and quantifying metastatic burden. Current and emerging radiotracers have broadened the utility of functional imaging in evaluating panNEN. Fluorine-18 fluorodeoxyglucose positron emission tomography (PET)/CT and somatostatin receptor imaging such as Gallium-68 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid–octreotate PET/CT may be useful for improved identification of panNEN in comparison to anatomic modalities. These new techniques can also play a direct role in optimizing the selection of treatment for individuals and predicting tumor response based on somatostatin receptor expression. In addition, emerging methods of radiomics such as texture analysis may be a potential tool for staging and outcome prediction in panNEN, however further investigation is required before clinical implementation.

Multimodality Imaging for the Staging of Pancreatic Cancer

Imaging plays a key role in the diagnosis, staging, and follow-up of pancreatic ductal adenocarcinoma. The pancreatic protocol dual-phase multidetector computed tomography scan is the imaging modality of choice. A computed tomography scan is highly accurate for pancreatic tumor detection, assessment of resectability, and detection of metastatic disease. This article reviews key principles of the acquisition, interpretation, and reporting of pancreatic ductal adenocarcinoma imaging with computed tomography scanning and highlights potential roles for newer and supplemental imaging technologies. We discuss the importance of structured interpretation and reporting for providing the most complete and accurate assessment of tumor stage and resectability.

Imaging of Pancreatic Neuroendocrine Neoplasms

Pancreatic neuroendocrine neoplasms (panNENs) represent the second most common pancreatic tumors. They are a heterogeneous group of neoplasms with varying clinical expression and biological behavior, from indolent to aggressive ones. PanNENs can be functioning or non-functioning in accordance with their ability or not to produce metabolically active hormones. They are histopathologically classified according to the 2017 World Health Organization (WHO) classification system. Although the final diagnosis of neuroendocrine tumor relies on histologic examination of biopsy or surgical specimens, both morphologic and functional imaging are crucial for patient care. Morphologic imaging with ultrasonography (US), computed tomography (CT) and magnetic resonance imaging (MRI) is used for initial evaluation and staging of disease, as well as surveillance and therapy monitoring. Functional imaging techniques with somatostatin receptor scintigraphy (SRS) and positron emission tomography (PET) are used for functional and metabolic assessment that is helpful for therapy management and post-therapeutic re-staging. This article reviews the morphological and functional imaging modalities now available and the imaging features of panNENs. Finally, future imaging challenges, such as radiomics analysis, are illustrated.

Computed tomography-based radiomics approach in pancreatic tumors characterization

Radiomics (or texture analysis) is a new imaging analysis technique that allows calculating the distribution of texture features of pixel and voxel values depend on the type of ROI (3D or 2D), their relationships in the image. Depending on the software, up to several thousand texture elements can be obtained. Radiomics opens up wide opportunities for differential diagnosis and prognosis of pancreatic neoplasias. The aim of this review was to highlight the main diagnostic advantages of texture analysis in different pancreatic tumors. The review describes the diagnostic performance of radiomics in different pancreatic tumor types, application methods, and problems. Texture analysis in PDAC is able to predict tumor grade and associates with lymphovascular invasion and postoperative margin status. In pancreatic neuroendocrine tumors, texture features strongly correlate with differentiation grade and allows distinguishing it from the intrapancreatic accessory spleen. In pancreatic cystic lesions, radiomics is able to accurately differentiate MCN from SCN and distinguish clinically insignificant lesions from IPMNs with advanced neoplasia. In conclusion, the use of the CT radiomics approach provides a higher diagnostic performance of CT imaging in pancreatic tumors differentiation and prognosis. Future studies should be carried out to improve accuracy and facilitate radiomics workflow in pancreatic imaging.

Neuroendocrine Carcinomas of the Digestive Tract: What Is New?

Neuroendocrine carcinomas (NEC) are rare tumors with a rising incidence. They show poorly differentiated morphology with a high proliferation rate (Ki-67 index). They frequently arise in the lung (small and large-cell lung cancer) but rarely from the gastrointestinal tract. Due to their rarity, very little is known about digestive NEC and few studies have been conducted. Therefore, most of therapeutic recommendations are issued from work on small-cell lung cancers (SCLC). Recent improvement in pathology and imaging has allowed for better detection and classification of high-grade NEN. The 2019 World Health Organization (WHO) classification has described a new entity of well-differentiated grade 3 neuroendocrine tumors (NET G-3), with better prognosis, that should be managed separately from NEC. NEC are aggressive neoplasms often diagnosed at a metastatic state. In the localized setting, surgery can be performed in selected patients followed by adjuvant platinum-based chemotherapy. Concurrent chemoradiotherapy is also an option for NEC of the lung, rectum, and esophagus. In metastatic NEC, chemotherapy is administered with a classic combination of platinum salts and etoposide in the first-line setting. Peptide receptor radionuclide therapy (PRRT) has shown positive results in high-grade NEN populations and immunotherapy trials are still ongoing. Available therapies have improved the overall survival of NEC but there is still an urgent need for improvement. This narrative review sums up the current data on digestive NEC while exploring future directions for their management.

Target Heterogeneity in Oncology: The Best Predictor for Differential Response to Radioligand Therapy in Neuroendocrine Tumors and Prostate Cancer

Simple Summary

In the era of precision medicine, novel targets have emerged on the surface of cancer cells, which have been exploited for the purpose of radioligand therapy. However, there have been variations in the way these receptors are expressed, especially in prostate cancers and neuroendocrine tumors. This variable expression of receptors across the grades of cancers led to the concept of ‘target heterogeneity’, which has not just impacted therapeutic decisions but also their outcomes. Radiopharmaceuticals targeting receptors need to be used when there are specific indicators—either clinical, radiological, or at molecular level—warranting their use. In addition, response to these radioligands can be assessed using different techniques, whereby we can prognosticate further outcomes. We shall also discuss, in this review, the conventional as well as novel approaches of detecting heterogeneity in prostate cancers and neuroendocrine tumors.

Abstract

Tumor or target heterogeneity (TH) implies presence of variable cellular populations having different genomic characteristics within the same tumor, or in different tumor sites of the same patient. The challenge is to identify this heterogeneity, as it has emerged as the most common cause of ‘treatment resistance’, to current therapeutic agents. We have focused our discussion on ‘Prostate Cancer’ and ‘Neuroendocrine Tumors’, and looked at the established methods for demonstrating heterogeneity, each with its advantages and drawbacks. Also, the available theranostic radiotracers targeting PSMA and somatostatin receptors combined with targeted systemic agents, have been described. Lu-177 labeled PSMA and DOTATATE are the ‘standard of care’ radionuclide therapeutic tracers for management of progressive treatment-resistant prostate cancer and NET. These approved therapies have shown reasonable benefit in treatment outcome, with improvement in quality of life parameters. Various biomarkers and predictors of response to radionuclide therapies targeting TH which are currently available and those which can be explored have been elaborated in details. Imaging-based features using artificial intelligence (AI) need to be developed to further predict the presence of TH. Also, novel theranostic tools binding to newer targets on surface of cancer cell should be explored to overcome the treatment resistance to current treatment regimens.

Quantitative Computed Tomography Image Analysis to Predict Pancreatic Neuroendocrine Tumor Grade

PURPOSE

The therapeutic management of pancreatic neuroendocrine tumors (PanNETs) is based on pathological tumor grade assessment. A noninvasive imaging method to grade tumors would facilitate treatment selection. This study evaluated the ability of quantitative image analysis derived from computed tomography (CT) images to predict PanNET grade.

METHODS

Institutional database was queried for resected PanNET (2000-2017) with a preoperative arterial phase CT scan. Radiomic features were extracted from the primary tumor on the CT scan using quantitative image analysis, and qualitative radiographic descriptors were assessed by two radiologists. Significant features were identified by univariable analysis and used to build multivariable models to predict PanNET grade.

RESULTS

Overall, 150 patients were included. The performance of models based on qualitative radiographic descriptors varied between the two radiologists (reader 1: sensitivity, 33%; specificity, 66%; negative predictive value [NPV], 63%; and positive predictive value [PPV], 37%; reader 2: sensitivity, 45%; specificity, 70%; NPV, 72%; and PPV, 47%). The model based on radiomics had a better performance predicting the tumor grade with a sensitivity of 54%, a specificity of 80%, an NPV of 81%, and a PPV of 54%. The inclusion of radiomics in the radiographic descriptor models improved both the radiologists' performance.

CONCLUSION

CT quantitative image analysis of PanNETs helps predict tumor grade from routinely acquired scans and should be investigated in future prospective studies.

Radiomics in Oncology, Part 1: Technical Principles and Gastrointestinal Application in CT and MRI

Simple Summary

Part I is an overview aimed to investigate some technical principles and the main fields of radiomic application in gastrointestinal oncologic imaging (CT and MRI) with a focus on diagnosis, prediction prognosis, and assessment of response to therapy in gastrointestinal cancers, describing mostly the results for each pre-eminent tumor. In particular, this paper provides a general description of the main radiomic drawbacks and future challenges, which limit radiomic application in clinical setting as routine. Further investigations need to standardize and validate the Radiomics as a helpful tool in management of oncologic patients. In that context, Radiomics has been playing a relevant role and could be considered as a future imaging landscape.

Abstract

Radiomics has been playing a pivotal role in oncological translational imaging, particularly in cancer diagnosis, prediction prognosis, and therapy response assessment. Recently, promising results were achieved in management of cancer patients by extracting mineable high-dimensional data from medical images, supporting clinicians in decision-making process in the new era of target therapy and personalized medicine. Radiomics could provide quantitative data, extracted from medical images, that could reflect microenvironmental tumor heterogeneity, which might be a useful information for treatment tailoring. Thus, it could be helpful to overcome the main limitations of traditional tumor biopsy, often affected by bias in tumor sampling, lack of repeatability and possible procedure complications. This quantitative approach has been widely investigated as a non-invasive and an objective imaging biomarker in cancer patients; however, it is not applied as a clinical routine due to several limitations related to lack of standardization and validation of images acquisition protocols, features segmentation, extraction, processing, and data analysis. This field is in continuous evolution in each type of cancer, and results support the idea that in the future Radiomics might be a reliable application in oncologic imaging. The first part of this review aimed to describe some radiomic technical principles and clinical applications to gastrointestinal oncologic imaging (CT and MRI) with a focus on diagnosis, prediction prognosis, and assessment of response to therapy.

Digestive Well-Differentiated Grade 3 Neuroendocrine Tumors: Current Management and Future Directions

Digestive well-differentiated grade 3 neuroendocrine tumors (NET G-3) have been clearly defined since the 2017 World Health Organization classification. They are still a rare category lacking specific data and standardized management. Their distinction from other types of neuroendocrine neoplasms (NEN) not only lies in morphology but also in genotype, aggressiveness, functional imaging uptake, and treatment response. Most of the available data comes from pancreatic series, which is the most frequent tumor site for this entity. In the non-metastatic setting, surgical resection is recommended, irrespective of grade and tumor site. For metastatic NET G-3, chemotherapy is the main first-line treatment with temozolomide-based regimen showing more efficacy than platinum-based regimen, especially when Ki-67 index <55%. Targeted therapies, such as sunitinib and everolimus, have also shown some positive therapeutic efficacy in small samples of patients. Functional imaging plays a key role for detection but also treatment selection. In the second or further-line setting, peptide receptor radionuclide therapy has shown promising response rates in high-grade NEN. Finally, immunotherapy is currently investigated as a new therapeutic approach with trials still ongoing. More data will come with future work now focusing on this specific subgroup. The aim of this review is to summarize the current data on digestive NET G-3 and explore future directions for their management.

A [68Ga]Ga-DOTANOC PET/CT Radiomic Model for Non-Invasive Prediction of Tumour Grade in Pancreatic Neuroendocrine Tumours

Predicting grade 1 (G1) and 2 (G2) primary pancreatic neuroendocrine tumour (panNET) is crucial to foresee panNET clinical behaviour. Fifty-one patients with G1-G2 primary panNET demonstrated by pre-surgical [68Ga]Ga-DOTANOC PET/CT and diagnostic conventional imaging were grouped according to the tumour grade assessment method: histology on the whole excised primary lesion (HS) or biopsy (BS). First-order and second-order radiomic features (RFs) were computed from SUV maps for the whole tumour volume on HS. The RFs showing the lowest p-values and the highest area under the curve (AUC) were selected. Three radiomic models were assessed: A (trained on HS, validated on BS), B (trained on BS, validated on HS), and C (using the cross-validation on the whole dataset). The second-order normalized homogeneity and entropy was the most effective RFs couple predicting G2 and G1. The best performance was achieved by model A (test AUC = 0.90, sensitivity = 0.88, specificity = 0.89), followed by model C (median test AUC = 0.87, sensitivity = 0.83, specificity = 0.82). Model B performed worse. Using HS to train a radiomic model leads to the best prediction, although a “hybrid” (HS+BS) population performs better than biopsy-only. The non-invasive prediction of panNET grading may be especially useful in lesions not amenable to biopsy while [68Ga]Ga-DOTANOC heterogeneity might recommend FDG PET/CT.

Radiomics in pancreatic neuroendocrine tumors: methodological issues and clinical significance

PURPOSE

To present the state-of-art of radiomics in the context of pancreatic neuroendocrine tumors (PanNETs), with a focus on the methodological and technical approaches used, to support the search of guidelines for optimal applications. Furthermore, an up-to-date overview of the current clinical applications of radiomics in the field of PanNETs is provided.

METHODS

Original articles were searched on PubMed and Science Direct with specific keywords. Evaluations of the selected studies have been focused mainly on (i) the general radiomic workflow and the assessment of radiomic features robustness/reproducibility, as well as on the major clinical applications and investigations accomplished so far with radiomics in the field of PanNETs: (ii) grade prediction, (iii) differential diagnosis from other neoplasms, (iv) assessment of tumor behavior and aggressiveness, and (v) treatment response prediction.

RESULTS

Thirty-one articles involving PanNETs radiomic-related objectives were selected. In regard to the grade differentiation task, yielded AUCs are currently in the range of 0.7-0.9. For differential diagnosis, the majority of studies are still focused on the preliminary identification of discriminative radiomic features. Limited information is known on the prediction of tumors aggressiveness and of treatment response.

CONCLUSIONS

Radiomics is recently expanding in the setting of PanNETs. From the analysis of the published data, it is emerging how, prior to clinical application, further validations are necessary and methodological implementations require optimization. Nevertheless, this new discipline might have the potential in assisting the current urgent need of improving the management strategies in PanNETs patients.

Application of Radiomics Analysis Based on CT Combined With Machine Learning in Diagnostic of Pancreatic Neuroendocrine Tumors Patient's Pathological Grades

Purpose

To evaluate the value of multiple machine learning methods in classifying pathological grades (G1,G2, and G3), and to provide the best machine learning method for the identification of pathological grades of pancreatic neuroendocrine tumors (PNETs) based on radiomics.

Materials and Methods

A retrospective study was conducted on 82 patients with Pancreatic Neuroendocrine tumors. All patients had definite pathological diagnosis and grading results. Using Lifex software to extract the radiomics features from CT images manually. The sensitivity, specificity, area under the curve (AUC) and accuracy were used to evaluate the performance of the classification model.

Result

Our analysis shows that the CT based radiomics features combined with multi algorithm machine learning method has a strong ability to identify the pathological grades of pancreatic neuroendocrine tumors. DC + AdaBoost, DC + GBDT, and Xgboost+RF were very valuable for the differential diagnosis of three pathological grades of PNET. They showed a strong ability to identify the pathological grade of pancreatic neuroendocrine tumors. The validation set AUC of DC + AdaBoost is 0.82 (G1 vs G2), 0.70 (G2 vs G3), and 0.85 (G1 vs G3), respectively.

Conclusion

In conclusion, based on enhanced CT radiomics features could differentiate between different pathological grades of pancreatic neuroendocrine tumors. Feature selection method Distance Correlation + classifier method Adaptive Boosting show a good application prospect.

Applications of artificial intelligence in pancreatic and biliary diseases

The application of artificial intelligence (AI) in medicine has increased rapidly with respect to tasks including disease detection/diagnosis, risk stratification, and prognosis prediction. With recent advances in computing power and algorithms, AI has shown promise in taking advantage of vast electronic health data and imaging studies to supplement clinicians. Machine learning and deep learning are the most widely used AI methodologies for medical research and have been applied in pancreatobiliary diseases for which diagnosis and treatment selection are often complicated and require joint consideration of data from multiple sources. The aim of this review is to provide a concise introduction of the major AI methodologies and the current landscape of AI research in pancreatobiliary diseases.

Tumor Heterogenity in Gastro-Entero-Pancreatic Neuroendocrine Neoplasia

Owing to the rarity and the biological and clinical heterogeneity of gastroenteropancreatic neuroendocrine neoplasia (GEP NEN), the management of these patients may be challenging for physicians. This review highlights the specific features of GEP NEN with particular attention on the role of Ki67 heterogeneity, the potential prognostic role of novel radiological techniques, and the clinical usefulness of functional imaging, including 68Ga-DOTA-SST PET/CT and 18F-FDG PET/CT. Understanding these specific features may help to plan proper and tailored follow-up programs and therapeutic approaches.

Correlation between radiologic features on contrast-enhanced CT and pathological tumor grades in pancreatic neuroendocrine neoplasms

Contrast-enhanced computed tomography (CT) contributes to the increasing detection of pancreatic neuroendocrine neoplasms (PNENs). Nevertheless, its value for differentiating pathological tumor grades is not well recognized. In this report, we have conducted a retrospective study on the relationship between the 2017 World Health Organization (WHO) classification and CT imaging features in 94 patients. Most of the investigated features eventually provided statistically significant indicators for discerning PNENs G3 from PNENs G1/G2, including tumor size, shape, margin, heterogeneity, intratumoral blood vessels, vascular invasion, enhancement pattern in both contrast phases, enhancement degree in both phases, tumor-to-pancreas contrast ratio in both phases, common bile duct dilatation, lymph node metastases, and liver metastases. Ill-defined tumor margin was an independent predictor for PNENs G3 with the highest area under the curve (AUC) of 0.906 in the multivariable logistic regression and receiver operating characteristic curve analysis. The portal enhancement ratio (PER) was shown the highest AUC of 0.855 in terms of quantitative features. Our data suggest that the traditional contrast-enhanced CT still plays a vital role in differentiation of tumor grades and heterogeneity analysis prior to treatment.

Pancreas image mining: a systematic review of radiomics

OBJECTIVES

To systematically review published studies on the use of radiomics of the pancreas.

METHODS

The search was conducted in the MEDLINE database. Human studies that investigated the applications of radiomics in diseases of the pancreas were included. The radiomics quality score was calculated for each included study.

RESULTS

A total of 72 studies encompassing 8863 participants were included. Of them, 66 investigated focal pancreatic lesions (pancreatic cancer, precancerous lesions, or benign lesions); 4, pancreatitis; and 2, diabetes mellitus. The principal applications of radiomics were differential diagnosis between various types of focal pancreatic lesions (n = 19), classification of pancreatic diseases (n = 23), and prediction of prognosis or treatment response (n = 30). Second-order texture features were most useful for the purpose of differential diagnosis of diseases of the pancreas (with 100% of studies investigating them found a statistically significant feature), whereas filtered image features were most useful for the purpose of classification of diseases of the pancreas and prediction of diseases of the pancreas (with 100% of studies investigating them found a statistically significant feature). The median radiomics quality score of the included studies was 28%, with the interquartile range of 22% to 36%. The radiomics quality score was significantly correlated with the number of extracted radiomics features (r = 0.52, p < 0.001) and the study sample size (r = 0.34, p = 0.003).

CONCLUSIONS

Radiomics of the pancreas holds promise as a quantitative imaging biomarker of both focal pancreatic lesions and diffuse changes of the pancreas. The usefulness of radiomics features may vary depending on the purpose of their application. Standardisation of image acquisition protocols and image pre-processing is warranted prior to considering the use of radiomics of the pancreas in routine clinical practice.

KEY POINTS

• Methodologically sound studies on radiomics of the pancreas are characterised by a large sample size and a large number of extracted features. • Optimisation of the radiomics pipeline will increase the clinical utility of mineable pancreas imaging data. • Radiomics of the pancreas is a promising personalised medicine tool in diseases of the pancreas.

CT and MRI of pancreatic tumors: an update in the era of radiomics

Radiomics is a relatively new approach for image analysis. As a part of radiomics, texture analysis, which consists in extracting a great amount of quantitative data from original images, can be used to identify specific features that can help determining the actual nature of a pancreatic lesion and providing other information such as resectability, tumor grade, tumor response to neoadjuvant therapy or survival after surgery. In this review, the basic of radiomics, recent developments and the results of texture analysis using computed tomography and magnetic resonance imaging in the field of pancreatic tumors are presented. Future applications of radiomics, such as artificial intelligence, are discussed.

Pancreatic Cancer Imaging: A New Look at an Old Problem

Computed tomography is the most commonly used imaging modality to detect and stage pancreatic cancer. Previous advances in pancreatic cancer imaging have focused on optimizing image acquisition parameters and reporting standards. However, current state-of-the-art imaging approaches still misdiagnose some potentially curable pancreatic cancers and do not provide prognostic information or inform optimal management strategies beyond stage. Several recent developments in pancreatic cancer imaging, including artificial intelligence and advanced visualization techniques, are rapidly changing the field. The purpose of this article is to review how these recent advances have the potential to revolutionize pancreatic cancer imaging.

Anatomic and Functional Imaging of Neuroendocrine Tumors

OPINION STATEMENT

Neuroendocrine tumors (NETs) can occur in a wide variety of organs and display a spectrum of pathologic behavior. Accurate and effective imaging is paramount to the diagnosis, staging, therapy, and surveillance of patients with NET. There have been continuous advancements in the imaging of NET which includes anatomic and functional techniques.

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.

Liver Tumor Burden in Pancreatic Neuroendocrine Tumors: CT Features and Texture Analysis in the Prediction of Tumor Grade and 18F-FDG Uptake

Pancreatic neuroendocrine tumors (p-NETs) are a rare group of neoplasms that often present with liver metastases. Histological characteristics, metabolic behavior, and liver tumor burden (LTB) are important prognostic factors. In this study, the usefulness of texture analysis of liver metastases in evaluating the biological aggressiveness of p-NETs was assessed. Fifty-six patients with liver metastases from p-NET were retrospectively enrolled. Qualitative and quantitative CT features of LTB were evaluated. Histogram-derived parameters of liver metastases were calculated and correlated with the tumor grade (G) and ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) standardized uptake value (SUV). Arterial relative enhancement was inversely related with G (−0.37, p = 0.006). Different metastatic spread patterns of LTB were not associated with histological grade. Arterial_entropy was significantly correlated to G (−0.368, p = 0.038) and to Ki67 percentage (−0.421, p = 0.018). The ROC curve for the Arterial_entropy reported an area under the curve (AUC) of 0.736 (95% confidence interval 0.545–0.928, p = 0.035) in the identification of G1–2 tumors. Arterial_uniformity values were correlated to G (0.346, p = 0.005) and Ki67 levels (0.383, p = 0.033). Arterial_entropy values were directly correlated with the SUV (0.449, p = 0.047) which was inversely correlated with Arterial_uniformity (−0.499, p = 0.025). Skewness and kurtosis reported no significant correlations. In conclusion, histogram-derived parameters may predict adverse histological features and metabolic behavior of p-NET liver metastases.

Comparison between neuroendocrine carcinomas and well-differentiated neuroendocrine tumors of the pancreas using dynamic enhanced CT

OBJECTIVES

To identify CT features distinguishing neuroendocrine carcinomas (NECs) of pancreas from well-differentiated neuroendocrine tumors (NETs) according to the World Health Organization 2017 and 2019 classification systems.

METHODS

This retrospective study included 69 patients with pathologically confirmed pancreatic neuroendocrine neoplasms who underwent dynamic CT (17, 17, 18, and 17 patients for well-differentiated grade 1, 2, 3 NET and NEC, respectively). CT was used to perform qualitative analysis (component, homogeneity, calcification, peripancreatic infiltration, main pancreatic ductal dilatation, bile duct dilatation, intraductal extension, and vascular invasion) and quantitative analysis (interface between tumor and parenchyma [delta], arterial enhancement ratio [AER], portal enhancement ratio [PER], and dynamic enhancement pattern). Uni- and multivariate logistic regression analyses were performed to identify features indicating NEC. Optimal cutoff values for enhancement ratios were determined.

RESULTS

NECs demonstrated significantly higher frequencies of main pancreatic ductal dilatation, bile duct dilatation, vascular invasion, and significantly lower delta (i.e., lower conspicuity), AER, and PER than well-differentiated NET (p < 0.05). On multivariate analysis, PER was the only independent factor selected by the model for differentiation of NEC from well-differentiated NET (odds ratio, < 0.001; 95% confidence interval [CI], < 0.001-0.012). PER < 0.8 showed the sensitivity of 94.1% (95% CI, 71.3-99.9) and the specificity of 88.5% (95% CI, 76.6-95.6). When three significant CT features were combined, the sensitivity and specificity for diagnosing NEC were 88.2% and 88.5%, respectively.

CONCLUSIONS

Tumor-parenchyma enhancement ratio in portal phase is a useful CT feature to distinguish NECs from well-differentiated NETs. Combining qualitative and quantitative CT features may aid in achieving good diagnostic accuracy in the differentiation between NEC and well-differentiated NET.

KEY POINTS

• Neuroendocrine carcinoma of the pancreas should be distinguished from well-differentiated neuroendocrine tumor in line with the revised grading and staging system. • Neuroendocrine carcinoma of the pancreas can be differentiated from well-differentiated neuroendocrine tumor on dynamic CT based on assessment of the portal enhancement ratio, arterial enhancement ratio, tumor conspicuity, dilatation of the main pancreatic duct or bile duct, and vascular invasion. • Tumor-parenchyma enhancement ratio in portal phase of dynamic CT is a useful feature, which may help to distinguish neuroendocrine carcinoma from well-differentiated neuroendocrine tumor of the pancreas.

Complementary role of computed tomography texture analysis for differentiation of pancreatic ductal adenocarcinoma from pancreatic neuroendocrine tumors in the portal-venous enhancement phase

Purpose

To assess the role of CT-texture analysis (CTTA) for differentiation of pancreatic ductal adenocarcinoma (PDAC) from pancreatic neuroendocrine neoplasm (PNEN) in the portal-venous phase as compared with visual assessment and tumor-to-pancreas attenuation ratios.

Methods

53 patients (66.1 ± 8.6y) with PDAC and 42 patients (65.5 ± 12.2y) with PNEN who underwent contrast-enhanced CT for primary staging were evaluated. Volumes of interests (VOIs) were set in the tumor tissue at the portal-venous phase excluding adjacent structures. Based on pyradiomics library, 92 textural features were extracted including 1st, 2nd, and higher order features, and then compared between PNEN and PDAC. The visual assessment classified tumors into hypo-, iso-, or hyperdense to pancreas parenchyma or into homogeneous/heterogeneous. Additionally, attenuation ratios between the tumors and the non-involved pancreas were calculated.

Results

8/92 (8.6%) highly significant (p < 0.005) discriminatory textural features between PDAC and PNEN were identified including the 1st order features “median,” “total energy,” “energy,” “10th percentile,” “90th percentile,” “minimum,” “maximum,” and the 2nd order feature “Gray-Level co-occurrence Matrix (GLCM) Informational Measure of Correlation (Imc2).” In PNEN, the higher order feature “GLSZM Small Area High Gray-Level Emphasis” proved significantly higher in G1 compared to G2/3 tumors (p < 0.05). The tumor/parenchyma ratios as well as the visual assessment into hypo-/iso-/hyperdense or homogeneous/heterogeneous did not significantly differ between PDAC and PNEN.

Conclusions

Our data indicate that CTTA is a feasible tool for differentiation of PNEN from PDAC and also of G1 from G2/3 PNEN in the portal-venous phase. Visual assessment and tumor-to-parenchyma ratios were not useful for discrimination.

Electronic supplementary material

The online version of this article (10.1007/s00261-020-02406-9) contains supplementary material, which is available to authorized users.

Feasibility of using computed tomography texture analysis parameters as imaging biomarkers for predicting risk grade of gastrointestinal stromal tumors: comparison with visual inspection

PURPOSE

To evaluate the feasibility of using computed tomography texture analysis (CTTA) parameters for predicting malignant risk grade and mitosis index of gastrointestinal stromal tumors (GISTs), compared with visual inspection.

METHOD AND MATERIALS

CTTA was performed on portal phase CT images of 145 surgically confirmed GISTs (mean size: 42.9 ± 37.5 mm), using TexRAD software. Mean, standard deviation, entropy, mean of positive pixels (MPP), skewness, and kurtosis of CTTA parameters, on spatial scaling factor (SSF), 2-6 were compared by risk grade, mitosis rate, and the presence or absence of necrosis on visual inspection. CTTA parameters were correlated with risk grade. Diagnostic performance was evaluated with receiver operating characteristic curve analysis. Enhancement pattern, necrosis, heterogeneity, calcification, growth pattern, and mucosal ulceration were subjectively evaluated by two observers.

RESULTS

Three to four parameters at different scales were significantly different according to the risk grade, mitosis rate, and the presence or absence of necrosis (p < 0.041). MPP at fine or medium scale (r = - 0.547 to - 393) and kurtosis at coarse scale (r = 0.424-0.454) correlated significantly with risk grade (p < 0.001). HG-GIST was best differentiated from LG-GIST by MPP at SSF 2 (AUC, 0.782), and kurtosis at SSF 4 (AUC, 0.779) (all p < 0.001). CT features predictive of HG-GIST were density lower than or equal to that of the erector spinae muscles on enhanced images (OR 2.1; p = 0.037; AUC, 0.59), necrosis (OR, 6.1; p < 0.001; AUC, 0.70), heterogeneity (OR, 4.3; p < 0.001; AUC, 0.67), and mucosal ulceration (OR, 3.3; p = 0.002; AUC, 0.62).

CONCLUSION

Using TexRAD, MPP and kurtosis are feasible in predicting risk grade and mitosis index of GISTs. CTTA demonstrated meaningful accuracy in preoperative risk stratification of GISTs.

Differentiation Between G1 and G2/G3 Phyllodes Tumors of Breast Using Mammography and Mammographic Texture Analysis

Purpose: To determine the potential of mammography (MG) and mammographic texture analysis in differentiation between Grade 1 (G1) and Grade 2/ Grade 3 (G2/G3) phyllodes tumors (PTs) of breast.

Materials and methods: A total of 80 female patients with histologically proven PTs were included in this study. 45 subjects who underwent pretreatment MG from 2010 to 2017 were retrospectively analyzed, including 14 PTs G1 and 31 PTs G2/G3. Tumor size, shape, margin, density, homogeneity, presence of fat, or calcifications, a halo-sign as well as some indirect manifestations were evaluated. Texture analysis features were performed using commercial software. Receiver operating characteristic curve (ROC) was used to determine the sensitivity and specificity of prediction.

Results: G2/G3 PTs showed a larger size (>4.0 cm) compared to PTs G1 (64.52 vs. 28.57%, p = 0.025). A strong lobulation or multinodular confluent was more common in G2/G3 PTs compared to PTs G1 (64.52 vs. 14.29%, p = 0.004). Significant differences were also observed in tumors' growth speed and clinical manifestations (p = 0.007, 0.022, respectively). Ten texture features showed significant differences between the two groups (p < 0.05), Correlation_AllDirection_offset7_SD and ClusterProminence_AllDirection_offset7_SD were independent risk factors. The area under the curve (AUC) of imaging-based diagnosis, texture analysis-based diagnosis and the combination of the two approaches were 0.805, 0.730, and 0.843 (90.3% sensitivity and 85.7% specificity).

Conclusions: Texture analysis has great potential to improve the diagnostic efficacy of MG in differentiating PTs G1 from PTs G2/G3.

Differentiation of chronic mass-forming pancreatitis from pancreatic ductal adenocarcinoma using contrast-enhanced computed tomography

PURPOSE

Both chronic mass-forming pancreatitis (CMFP) and pancreatic ductal adenocarcinoma (PDAC) are focal pancreatic lesions and share very similar clinical symptoms and imaging performance. There is great clinical value in preoperative differentiation of those two lesions. The purpose of this study was to investigate the value of computed tomography (CT) features in discriminating CMFP from PDAC.

PATIENTS AND METHODS

Forty-seven patients with pathologically confirmed PDAC and 21 patients with CMFP were included in this study. Demographic and CT features, including tumor location, size, margin, pancreatic or bile duct dilatation, vascular invasion, cystic necrosis, pancreatic atrophy, calcification, and tumor contrast enhancement, were retrospectively analyzed and compared. Multivariate logistic regression analyses were adopted to identify relevant CT imaging features to discriminate CMFP from PDAC.

RESULTS

There were significant differences between CMFP and PDAC with respect to main pancreatic duct dilatation, vascular invasion, cystic necrosis, pancreatic atrophy, calcification, and tumor contrast enhancement. Delayed contrast enhancement (>70.5 Hounsfield units) showed high sensitivity and specificity of 84.2% and 84.7%. The areas under the curve (AUCs) of the predicting models based on qualitative and quantitative variables were 0.770 (95% CI: 0.660-0.880) and 0.943 (95% CI: 0.888-0.999), respectively. When all significant variables were used in combination to build a predicting model, the AUC was 0.969 (95% CI: 0.930-1.000) with 84.2% sensitivity and 94.7% specificity.

CONCLUSION

Main pancreatic duct dilatation, vascular invasion, cystic necrosis, pancreatic atrophy, calcification, tumor size, and tumor contrast enhancement were shown to be useful CT imaging features in discriminating CMFP from PDAC.