This guideline focuses on pancreatic solid tumors, primarily pancreatic cancer and pNETs, with ICD-11 codes C25.901 and D37.703, respectively. It addresses two major aspects: selection of imaging examination techniques and evaluation of imaging reports. Following evidence-based methodology, 20 recommendations were developed.
Diagnostic reporting standards
The diagnostic report for pancreatic solid tumors should include five key components6,7,33: tumor assessment, evaluation of surrounding vessels, regional lymph node assessment, invasion of adjacent organs and distant metastases, and other abnormal imaging findings within the scan range.
Should tumor size of pancreatic solid tumors be measured on CT or MRI?
Recommendation 2: Measurement of the maximum tumor diameter on either CT or MRI images is recommended.
Evidence quality: C; Recommendation strength: Weak
Tumor size determines the T stage, which is a strong prognostic factor for survival in various malignancies including pancreatic cancer.34,35 However, T staging is pathological, and preoperative prediction relies on imaging. Due to irregular tumor morphology, measuring the maximum diameter on CT or MRI often does not accurately reflect the true tumor size, potentially leading to over- or underestimation,36–41 which may affect treatment decisions.
Michallek et al.39 reported that CT tends to underestimate tumor size, while MRI correlates better with actual size; however, other studies found that although size measurements on CT or MRI may differ from gross pathology and potentially alter T staging, overall tumor staging is rarely affected.41
For pNETs, studies indicate high concordance between CT/MRI measurements and pathological specimens, with no significant difference between CT and MRI results.28,42
van Beek et al.43 concluded that preoperative CT and MRI neither overestimate nor underestimate pNET size, but MRI has advantages in consistency and reliability.
Currently, no consensus exists on the optimal modality or method for measuring pancreatic solid tumor size. This guideline recommends that either CT or MRI may be used clinically. For multifocal lesions, each lesion’s size should be measured; if the lesion is not clearly visualized (e.g., isoattenuating on CT or isointense on MRI), measurement is not feasible.
How should tumor size of pancreatic solid tumors be measured?
Recommendation 3: For PDAC, measure the maximum diameter on the largest cross-sectional image during the pancreatic parenchymal enhancement phase. For functional pNETs, measure the maximum diameter on the largest cross-sectional image during the arterial enhancement phase.
Evidence quality: B; Recommendation strength: Strong
According to the Response Evaluation Criteria in Solid Tumors and the structured reporting template for pancreatic cancer proposed by Al-Hawary et al.,33 tumor size should be measured as the maximum diameter on the largest cross-sectional tumor image.8
Cocquempot et al.40 found that tumor size measurement during the pancreatic parenchymal phase (∼40 seconds post-contrast injection) on CT is most accurate for PDAC, thus recommending measurement during this phase (Fig. 2).
pNETs are heterogeneous tumors; functional pNETs show early, marked, homogeneous arterial enhancement. Studies report higher sensitivity for detecting small functional pNETs in the arterial phase (83–88%) compared to the parenchymal phase (11–76%)44; therefore, measurement during the arterial phase is recommended.
Additionally, the European Neuroendocrine Tumor Society consensus guideline recommends that the late arterial phase suffices to evaluate arterial anatomy and its relationship with pNETs.5
In most cases, the late arterial phase allows pNET diagnosis, whereas the pancreatic parenchymal phase is optimal for PDAC diagnosis but less favorable for pNET characterization.
Is morphological assessment of the main pancreatic duct (MPD) obstruction site necessary?
Recommendation 4: Morphological assessment of the MPD obstruction site is necessary.
Evidence quality: A; Recommendation strength: Strong
Early pancreatic cancer has a better prognosis but remains challenging to diagnose.45
A Japanese multicenter study identified isolated MPD stricture as an imaging feature suggestive of early pancreatic cancer, providing diagnostic clues.46
With disease progression, MPD may show abrupt cutoff and upstream ductal dilation,45,47–50 especially in isoattenuating/signal pancreatic cancers.51
Meta-analyses indicate that multiple MPD strictures, duct-penetrating signs, and absence of upstream MPD dilation are important imaging features differentiating autoimmune pancreatitis from pancreatic cancer, with MPD cutoff being the most specific sign for pancreatic cancer.52,53
Hypovascular pNETs typically do not cause MPD cutoff or upstream dilation, aiding differentiation from pancreatic cancer.54,55 However, large pNETs or pancreatic neuroendocrine carcinomas (pNECs) may compress or invade the MPD, causing abrupt cutoff and upstream dilation.56,57
Therefore, evaluating MPD obstruction morphology is critical for early lesion detection and differential diagnosis among pancreatic tumors.58–60
This guideline recommends that imaging reports clearly state whether MPD obstruction is present; if so, specify the location and characterize the obstruction as stricture or cutoff.
How should MPD diameter be measured?
Recommendation 5: Measure the MPD diameter on magnetic resonance cholangiopancreatography (MRCP), T2-weighted imaging (T2WI), or contrast-enhanced CT during the pancreatic parenchymal or portal venous phase, selecting the plane perpendicular to the MPD long axis.
Evidence quality: C; Recommendation strength: Weak
MRI/MRCP detection of pancreatic ductal changes is equivalent or slightly superior to CT,61 though some studies suggest CT curved planar reconstructions can match MRCP quality.62
No consensus exists on MPD measurement methodology. Based on current literature, this guideline recommends measuring the maximum MPD diameter on 2D MRCP, T2WI, or contrast-enhanced CT during the pancreatic parenchymal or portal venous phase, selecting the optimal plane perpendicular to the MPD long axis.63–65
How is MPD dilation defined?
Recommendation 6: MPD dilation is defined as a maximum diameter >3 mm.
Evidence quality: C; Recommendation strength: Weak
Pancreatic solid tumors can compress or invade the pancreatic duct, causing varying degrees of upstream ductal dilation.
The upper limit of normal pancreatic duct diameter remains debated. A 1976 ERCP study of 35 patients without pancreatic disease suggested an upper normal MPD diameter of 3 mm.66
Other studies define dilation as MPD diameters ≥3 mm in the pancreatic head, ≥2 mm in the body, and ≥1 mm in the tail.67,68
The 2018 International Consensus Guidelines on Chronic Pancreatitis cross-sectional imaging diagnosis and severity scoring (including MRI and CT) also recommend >3 mm as the threshold for dilation.69
A recent population-based cross-sectional study published in Gut found that applying traditional reference values led to up to 11% of healthy volunteers being diagnosed with MPD dilation, prompting unnecessary further testing. It proposed new age-adjusted upper limits for asymptomatic individuals with normal liver function and lipase levels65: 3 mm for those <65 years old and 4 mm for those ≥65 years old.
However, whether to update routine MPD reference values remains controversial without consensus.
Therefore, this guideline recommends retaining the definition of MPD dilation as maximum diameter >3 mm in imaging reports.33,70
Is morphological assessment of the common bile duct (CBD) obstruction site necessary?
Recommendation 7: Morphological assessment of the CBD obstruction site is necessary.
Evidence quality: A; Recommendation strength: Strong
The CBD obstruction site usually corresponds to the lesion location; studying its morphology aids lesion characterization.
Inflammatory CBD strictures typically show concentric narrowing with smooth walls and a “rat-tail” appearance without abrupt ductal cutoff, whereas malignant strictures often exhibit abrupt cutoff or eccentric “moth-eaten” narrowing.71
In periampullary malignancies, up to 50–80% of high-risk patients presenting with clinical or biochemical jaundice and/or imaging-detected masses exhibit biliary and pancreatic duct dilation (double duct sign).67,72,73
For pNETs, large tumors or pNECs may compress or invade the CBD, causing biliary dilation.74
In non-jaundiced patients, meta-analyses suggest that incidental CBD dilation is mostly due to benign causes (e.g., chronic pancreatitis, biliary stones), with periampullary tumors accounting for only ∼5%.75
Thus, in jaundiced patients, CBD obstruction with dilation warrants high suspicion for periampullary malignancy.
This guideline recommends that imaging reports specify whether CBD obstruction is present; if so, detail the obstruction location (proximal to pancreatic head, pancreatic head segment, periampullary region) and characterize the obstruction as stricture (concentric or eccentric) or cutoff.
How should CBD diameter be measured?
Recommendation 8: Measure the CBD diameter on MRCP, T2WI, or contrast-enhanced CT images, selecting the largest cross-sectional area perpendicular to the CBD long axis.
Evidence quality: C; Recommendation strength: Weak
A randomized controlled trial demonstrated high correlation between MRI sequences and ultrasound measurements of extrahepatic bile duct diameter, confirming MRI as a reliable method.76
CT offers high resolution and multiplanar reconstruction; studies suggest using coronal and oblique sagittal CT reconstructions to select the maximal cross-section perpendicular to the CBD long axis for measurement.77,78
No universally accepted CBD measurement method exists. Based on current evidence, this guideline recommends measuring CBD diameter on MRCP, T2WI, or contrast-enhanced CT images, selecting the maximal cross-section perpendicular to the CBD long axis.65
How is CBD dilation defined?
Recommendation 9: CBD dilation is defined as maximum diameter >8 mm with gallbladder present, and >10 mm post-cholecystectomy.
Evidence quality: B; Recommendation strength: Strong
Normal CBD diameter is controversial, with studies reporting a range of 4–8 mm.79
CBD diameter correlates with age, increasing in healthy individuals as age advances,80–82 necessitating age-adjusted reference values.
Post-cholecystectomy compensatory CBD dilation is recognized; some studies define dilation as >10 mm in this population,83,84 though upper limits require further validation.65
A recent Gut publication showed that applying traditional reference values led to 18.2% of healthy volunteers being classified as having CBD dilation, causing unnecessary investigations. It proposed new age-adjusted upper limits for asymptomatic individuals with normal liver function and lipase: 8 mm for those <65 years old and 11 mm for those ≥65 years old.65
However, updating routine CBD reference values remains debated without consensus.
Currently, most CT and MRI studies continue to define CBD dilation as >8 mm with gallbladder present and >10 mm post-cholecystectomy.79,83,84
How is PPA Defined and what is its clinical significance?
Recommendation 10: PPA refers to reduced pancreatic volume and is classified as PPPA and UPPA. PPA is an important indirect imaging sign for diagnosing early and advanced pancreatic cancer.
Evidence quality: B; Recommendation strength: Strong
Early pancreatic cancer has significantly better prognosis than advanced disease,85 but early-stage tumors (including carcinoma in situ) or isoattenuating pancreatic cancers are often difficult to detect on conventional imaging, requiring indirect imaging signs for diagnosis.
PPA is the most important indirect imaging sign for diagnosing such challenging pancreatic cancers.50,51
PPA denotes pancreatic volume loss, with specific classification and quantification methods proposed.86–88
Yamao et al.86 classified PPA into PPPA and UPPA: PPPA corresponds to atrophic changes in the pancreatic parenchyma at the MPD stricture site; UPPA refers to atrophy of the upstream pancreatic parenchyma caused by MPD stricture (Fig. 3).73
For pancreatic cancers without clear MPD stricture on CT, Koiwai et al.89 defined pancreatic body width ≤10 mm as another form of PPA.
Recent studies suggest PPPA may result from branch duct obstruction or closure caused by pancreatic intraepithelial neoplasia or small pancreatic cancers, leading to local fibrosis or fatty replacement due to impaired pancreatic juice drainage, making PPPA a key early imaging marker.86–91
Advanced pancreatic cancer often causes MPD stricture or obstruction with upstream fibrosis leading to UPPA.50
PPPA and UPPA correlate with pancreatic cancer development and progression.
This guideline recommends that imaging reports describe the presence or absence of PPPA and UPPA.
How should PPA be accurately assessed?
Recommendation 11: PPA should be primarily assessed on CT/MRI contrast-enhanced parenchymal phase images. PPPA measurement criteria include: (1) partial pancreatic parenchymal indentation with parenchymal edge to MPD wall distance ≤4 mm; (2) PPPA length of 10–25 mm; (3) upstream pancreatic parenchymal width >6 mm. UPPA criteria include parenchymal edge to MPD distances at the stricture and upstream sites both ≤4 mm. For pancreatic cancers without clear MPD stricture, pancreatic body width ≤10 mm defines PPA.
Evidence quality: C; Recommendation strength: Weak
Contrast-enhanced CT or MRI parenchymal phase images provide optimal tumor-to-pancreas contrast and clear venous visualization, making them ideal for thickness measurements.92
Current PPA measurement methods are mostly linear, but no consensus exists on absolute thresholds to quantify pancreatic atrophy, especially since parenchymal thickness decreases with age and age-adjusted standards are lacking.
Sandini et al.93 calculated the MPD-to-pancreatic thickness ratio at the widest MPD diameter, finding that a ratio ≥3.5 indicates UPPA and predicts postoperative prognosis.
Nakahodo et al.87,88 defined PPPA as the product of maximal and minimal diameters of the pancreatic indentation exceeding 50 mm2 or a distance >5 mm under the line connecting the edges of the indentation.
Yamao et al.86 proposed PPPA criteria on CT images: (1) parenchymal edge to MPD distance ≤4 mm at the stricture site, with upstream parenchymal edge to MPD distance >6 mm; (2) PPPA length 10–25 mm; (3) upstream parenchymal width >6 mm (Fig. 4a). UPPA is defined as parenchymal edge to MPD distances ≤4 mm at both stricture and upstream sites (Fig. 4b).73
For pancreatic cancers without clear MPD stricture on CT, Koiwai et al.89 defined pancreatic body width ≤10 mm as PPA.
This guideline recommends adopting these criteria for PPA measurement.
Is assessment of obstructive acute pancreatitis necessary?
Recommendation 12: Assessment for the presence of obstructive acute pancreatitis is necessary.
Evidence quality: B; Recommendation strength: Strong
Explanation: Alcohol and gallstones are the most important etiologies of acute pancreatitis. Pancreatic cancer is a relatively uncommon cause, with an incidence of 0.9% to 3.6%.94 Mujica et al.95 hypothesized possible mechanisms by which pancreatic cancer induces acute pancreatitis, including mechanical obstruction of the pancreatic duct, ischemia caused by malignant tumor cells obstructing blood vessels, and direct activation of pancreatic enzymes by tumor tissue. The presence of obstructive pancreatitis is assessed based on suspicious masses, main pancreatic duct interruption with upstream ductal dilation, pancreatic enlargement upstream, blurring of peripancreatic fat planes, and stranding edema.96,97 Studies have found that 59% of pancreatic cancer cases were initially misdiagnosed as acute pancreatitis due to inflammatory changes masking underlying masses or secondary signs.98 Tummala et al.99 studied 218 patients with acute pancreatitis undergoing EUS-FNA and identified 38 cases of pancreatic cancer diagnosed promptly after the first episode of acute pancreatitis, with a resection rate of 39%, thereby improving patient survival. Therefore, in patients with acute pancreatitis after excluding common causes such as alcohol and gallstones, careful evaluation of pancreatic imaging is warranted to identify potential underlying malignancy,50,99,100 and this should be explicitly stated in imaging reports (Fig. 5).
Is evaluation of pseudocysts/retention cysts necessary?
Recommendation 13: Evaluation for the presence of pseudocysts or retention cysts is necessary.
Evidence quality: B; Recommendation strength: Strong
Explanation: Pancreatic cancer originates from ductal epithelial cells and can obstruct the pancreatic duct, causing upstream retention of pancreatic juice and cystic dilation, forming retention cysts lined by ductal epithelium on pathology.101 When intraductal pressure increases or the pancreatic duct ruptures, pseudocysts may form; these can also develop secondary to obstructive acute pancreatitis. Pseudocysts consist of fluid collections surrounded by non-epithelial tissue within or adjacent to the pancreas.102 Differentiation between retention cysts and pseudocysts requires pathological examination; radiologically, they are often difficult to distinguish, both typically presenting as unilocular cysts with variably thick walls, commonly located in the pancreatic body or tail without obvious mural nodules.101 Studies have identified pancreatic cysts ≥5 mm as independent predictors for pancreatic cancer development.103 In patients with elevated CA19-9 and retention cysts, vigilance for small or isoattenuating pancreatic cancers nearby is warranted.50,104 This guideline recommends that when retention or pseudocysts are detected, they should be clearly reported with measurements of size and description of location (Fig. 6).
Preferred imaging modality for peripancreatic vascular assessment
Recommendation 14: Contrast-enhanced pancreatic CT is the preferred modality for evaluating peripancreatic vessels.
Evidence quality: A; Recommendation strength: Strong
Explanation: In the absence of distant metastases, resectability assessment of pancreatic cancer primarily depends on evaluation of tumor–vascular relationships. According to multiple studies including the NCCN, contrast-enhanced pancreatic CT is the preferred imaging modality for staging and resectability determination of pancreatic cancer.8,20,33 Although MRI has demonstrated comparable sensitivity and specificity to CT for assessing vascular involvement in pancreatic cancer,24,25,105,106 CT is more widely utilized due to lower cost and greater availability.25 Additionally, studies indicate that multidetector CT (MDCT) with three-dimensional reconstruction significantly improves accuracy in assessing vascular invasion compared to MDCT without 3D reconstruction (100%; 95% CI, 91–100% vs. 79%; 95% CI, 64–89%).25 Evaluation requires separate assessment of arterial and venous peripancreatic vessels and their branches.33,107
Key points in peripancreatic vascular assessment
Recommendation 15: Imaging reports should evaluate the degree of tumor contact with peripancreatic arteries and veins (including major branches), presence of vascular deformation, vascular variants, and venous tumor thrombus.
Evidence quality: A; Recommendation strength: Strong
Explanation: In 2014, Al-Hawary et al.33 proposed a CT-based vascular involvement classification system based on the 2013 NCCN guidelines, standardizing imaging reporting for pancreatic cancer vascular assessment; this system was adopted by the 2023 NCCN guidelines.8 Peripancreatic arteries include the celiac axis, superior mesenteric artery, common hepatic artery, and abdominal aorta. Imaging evaluation should clearly describe the tumor–vessel relationship, specifying the contact arc angle (≤180° or >180°) and whether luminal narrowing or deformation is present. When hazy or stranding increased density is observed at the tumor–artery interface, the contact relationship and contact arc angle should be explicitly reported.8,33 Arterial variants and their relationship to the tumor should also be documented (Figs. 7 and 8).
Peripancreatic veins include the portal vein, superior mesenteric vein (SMV), and inferior vena cava. Venous imaging assessment parallels that of arteries, with additional description of the relationship between the first-order SMV branches and the tumor (contact arc angle ≤180° or >180°), presence of tumor thrombus or bland thrombus, local luminal narrowing, irregular or teardrop-shaped deformation, and signs of portal hypertension with collateral circulation (Figs. 9 and 10).30
Imaging criteria for resectability assessment
Recommendation 16: Imaging-based resectability criteria are stratified by tumor location and the degree of tumor contact with arteries and veins into resectable, borderline resectable, and locally advanced categories.
Evidence quality: A; Recommendation strength: Strong
Explanation: Imaging assessment of resectability primarily relies on tumor location and the extent of tumor contact with surrounding vessels. This guideline adopts the 2022 NCCN criteria for resectability,8 categorizing non-metastatic pancreatic cancer into three groups: resectable, borderline resectable, and locally advanced (Table 2).
Table 22023 NCCN clinical practice guidelines for pancreatic cancer (Version 2) resectability criteria
| Resectability status | Arteries | Veins |
|---|
| Resectable | Solid tumor with clear fat planes around the celiac axis, superior mesenteric artery, and common hepatic artery | Solid tumor with clear fat planes around the superior mesenteric vein and portal vein; solid tumor contact with SMV or PV ≤180° with smooth contours |
| Borderline Resectable | [Pancreatic head or groove region:] ① Solid tumor contact with common hepatic artery but not involving celiac axis or hepatic artery branches; ② Solid tumor contact with superior mesenteric artery ≤180°; ③ Presence of arterial variants (e.g., accessory right hepatic artery, replaced right hepatic artery, replaced common hepatic artery, or accessory/replaced arterial origins) with tumor contact and extent should be specified; [Pancreatic body or tail:] Solid tumor contact with celiac axis ≤180° | ① Solid tumor contact with SMV or PV >180°, or tumor contact ≤180° with irregular venous contour or venous thrombus, with suitable proximal and distal veins for safe and complete resection and reconstruction; ② Solid tumor contact with inferior vena cava |
| Locally Advanced | [Pancreatic head or groove region:] Solid tumor contact with SMA or celiac axis >180°; [Pancreatic body or tail:] ① Tumor invasion of SMA or celiac axis >180°; ② Tumor involvement of both celiac axis and abdominal aorta | Due to tumor invasion or thrombosis (tumor or bland thrombus) of SMV or PV, reconstruction is not feasible |
Optimal imaging modality for regional lymph node assessment
Recommendation 17: Preoperative imaging prediction of lymph node metastasis remains challenging and requires further validation through large-scale, multicenter studies to determine the optimal imaging modality.
Evidence quality: C; Recommendation strength: Weak
Explanation: A meta-analysis including 157 cases of pancreatic and ampullary malignancies concluded that CT has low diagnostic accuracy for extraregional lymph node metastasis, with pooled sensitivity and positive predictive value of only 25% and 28%, respectively; thus, CT alone is not recommended when extraregional lymph node metastasis is suspected.108 Another meta-analysis found that although CT has a high positive predictive value (81%) for assessing pancreatic cancer resectability, it exhibits a high false-positive rate for regional lymph node metastasis.109 Despite MRI’s superior soft tissue resolution and studies suggesting that diffusion-weighted imaging (DWI) and intravoxel incoherent motion sequences can effectively identify metastatic lymph nodes in pancreatic cancer,110–112 Adham et al.113 recently reported limited diagnostic value and poor interobserver agreement (kappa = 0.257) for MRI-based lymph node assessment in pancreatic cancer. A meta-analysis found no significant difference between PET/CT and conventional CT in diagnosing regional lymph node metastasis in pancreatic cancer.114 In a prospective study comparing PET/CT, CT, and MRI, Kauhanen et al.115 found PET/CT sensitivity for metastatic regional lymph nodes was low and comparable to CT and MRI (all approximately 30%). Collectively, these findings indicate that no single conventional imaging modality—whether CT, MRI, or PET—provides adequate accuracy for predicting lymph node metastasis in pancreatic cancer. Consequently, preoperative imaging assessment of lymph node status remains a significant clinical challenge that requires further investigation through large-scale, multicenter studies to identify optimal imaging strategies.
Imaging criteria for suspicious lymph nodes
Recommendation 18: Lymph nodes exhibiting a short-axis diameter >10 mm, heterogeneous density/signal, heterogeneous enhancement, internal necrosis, confluence, indistinct margins, or diffusion restriction on MRI, especially when multiple imaging features coexist, strongly suggest lymph node metastasis.
Evidence quality: C; Recommendation strength: Weak
Explanation: Assessing lymph node involvement is crucial for cancer staging, treatment planning, and prognosis. However, morphological criteria for lymph node evaluation based on CT and MRI remain inconsistent. Although lymph node size is widely used as an indicator, its reliability for assessing lymph node metastasis is limited, and measurement standards are not uniform. Combining morphological features of lymph nodes can be somewhat helpful, but currently this approach is restricted to specific diseases and anatomical sites. Studies on regional lymph node morphology have shown that a short-axis diameter >10 mm is not a reliable parameter for evaluating lymph node metastasis in pancreatic cancer patients,116,117 whereas in pNETs, short-axis size is a reliable predictor of lymph node metastasis.118
The Node-RADS (Lymph Node Reporting and Data System), proposed in 2021, addresses these limitations by incorporating lymph node size (short-axis >10 mm) and morphology (texture, margin, shape) into a scoring system.119 A PubMed search for “Node-RADS” reveals its current application primarily in assessing lymph node metastasis in colorectal, gastric, bladder, lung, and prostate cancers; no studies have yet validated its use in pancreatic cancer lymph node assessment, necessitating further high-quality research to establish its utility in this context.
Currently, most studies consider the following imaging features suspicious for lymph node metastasis: short-axis diameter >10 mm, round shape, heterogeneous density/signal, heterogeneous enhancement, presence of necrosis, node confluence, indistinct margins, and restricted diffusion on MRI. The coexistence of multiple such imaging signs strongly suggests lymph node metastasis (Fig. 11).111,120,121
Should the Specific location of suspicious lymph nodes be described?
Recommendation 19: The specific location of suspicious lymph nodes should be described.
Evidence quality: A; Recommendation strength: Strong
Explanation: The Japan Pancreas Society established a lymph node station and grouping system based on the anatomical lymphatic drainage around the pancreas under physiological conditions, and the likelihood of positive lymph nodes in pancreatic cancer. Peripancreatic lymph nodes are divided into 3 stations and 18 groups,122 each group accompanying specific peripancreatic vessels with distinct distribution patterns. Pancreatic head cancers commonly metastasize to groups 6, 8, 13, 14, and 17, whereas pancreatic body and tail cancers frequently involve groups 8, 9, 10, 11, and 18.122,123
Group 13 is the sentinel lymph node for pancreatic head cancer, while groups 9 and 11 serve as sentinel nodes for body and tail cancers.123,124
Liu et al.125 analyzed lymph node metastasis patterns in 132 pancreatic head cancer specimens following pancreaticoduodenectomy, finding that lymphatic spread is not skip metastasis but follows a sequential pathway from group 13 to 14 and then 16.
Moreover, metastatic lymph nodes at different locations correlate with prognosis. For example, metastasis to the hepatoduodenal ligament group (group 12) is an independent adverse prognostic factor126; involvement of the superior mesenteric artery group (group 14) is significantly associated with shorter disease-free survival126,127; para-aortic group (group 16) metastasis is an independent poor prognostic factor128; and splenic artery group (group 11) metastasis is an independent adverse prognostic factor for overall survival in pancreatic body and tail cancer patients.129
Therefore, this guideline recommends that diagnostic imaging reports specify the location of suspicious lymph nodes. When analyzing images, careful observation along lymph node distribution pathways relative to the pancreatic tumor location can aid in diagnosing suspected metastatic lymph nodes.
Preferred imaging modality for detecting hepatic and peritoneal metastases
Recommendation 20: MRI with DWI is the preferred modality for diagnosing hepatic and peritoneal metastases.
Evidence quality: A; Recommendation strength: Strong
Explanation: Approximately 13–23% of pancreatic cancer patients are found intraoperatively to have hepatic or peritoneal metastases, rendering the tumor unresectable.130 Although multiphasic helical CT with volumetric scanning has improved detection rates of metastatic lesions, contrast-enhanced CT sensitivity for hepatic metastases remains suboptimal (38–76%).131 MRI with DWI demonstrates high sensitivity (86–97%) for detecting focal hepatic lesions, including subcentimeter metastases (sensitivity 60–91%).132–134 Meta-analyses indicate that MRI outperforms CT in overall diagnostic accuracy for hepatic metastases in pancreatic cancer, with pooled sensitivity higher for MRI than CT (85% vs. 75%), while pooled specificity is comparable (98% vs. 94%).135 Another meta-analysis similarly found comparable specificity between CT and MRI (94% vs. 96%) but superior sensitivity for MRI (83% vs. 45%) in diagnosing hepatic metastases.136 Peritoneal metastases from pancreatic cancer typically manifest as miliary nodules on the omentum and mesentery, irregular peritoneal thickening, and small-volume ascites, all suggestive of peritoneal carcinomatosis or implantation (Fig. 12).137 CT sensitivity for peritoneal metastases depends on lesion size; Archer et al.138 reported sensitivity of only 25% for lesions <0.5 cm, increasing to 90% for lesions >5 cm. MRI, including DWI sequences, is a sensitive imaging modality for detecting peritoneal metastases.139 Studies suggest that combining high b-value (b = 800 s/mm2) DWI with conventional MRI improves preoperative detection sensitivity (from 0.58 to 0.85) and accuracy (from 0.67 to 0.85) for peritoneal tumors compared to conventional MRI alone.140