Summary: "Early Pancreatic Cancer: Clinical Implications, Workup, and Imaging Findings with Histopathologic Correlation for Personalized Surveillance"

What Is "Early" Pancreatic Cancer?

• Precursor lesions with high-grade dysplasia (stage 0 / Tis - carcinoma in situ)
• Invasive carcinomas ≤2.0 cm confined to the pancreas (stage I)

Clinical Implications

• Pancreatic cancer is projected to become the 2nd leading cause of cancer death in the US by 2040.

• 50% of cases are still diagnosed at metastatic stage.

• Stage T1 pancreatic cancer can progress to stage T4 in just over 1 year, making the detection window extremely narrow.
• 5-year survival for stage IA disease has improved, reaching 83.7% in 2012.
• Tumor size alone is insufficient for prognosis in small tumors; extrapancreatic extension significantly worsens outcomes.

Histopathology & Precursor Lesions

• KRAS mutations occur earliest (nearly universal, even in low-grade lesions)
• CDKN2A mutations appear relatively early
• TP53 and SMAD4 mutations emerge later, mostly in high-grade/invasive lesions
• An alternative "catastrophic" model involving chromothripsis may explain the sometimes abrupt clinical course

• Accounts for 85-90% of PDAC origins
• Low-grade PanIN: lifetime progression risk ~1.3-1.5% over 18-35 years
• High-grade PanIN: progresses to PDAC in ~6.8-12.3 years
• Causes lobulocentric atrophy via chronic pancreatitis-like changes (fibrosis, loss of acinar structure)
• No pathognomonic imaging features exist for high-grade PanIN specifically

• 3 types: branch duct (BD), main duct (MD), mixed
• High-grade IPMN progresses to PDAC in ~3.7 years
• 10-year cumulative malignancy risk for BD-IPMN reaches 5.9%
• The "field effect" hypothesis suggests the entire pancreas harboring IPMNs may have elevated malignancy susceptibility - so concomitant PDAC can develop independently of known cysts

Imaging Findings - Two Categories

Direct Findings (tumor itself):

• On CT: PDAC is typically hypoattenuating in the pancreatic phase; however, 10-17% are isoattenuating (harder to detect, more common in small cancers). Delayed phase may reveal these as hyperattenuating due to stromal fibrosis.
• On MRI: hypointense on T1WI, hyperintense on DWI, variable T2 signal. T1WI has higher sensitivity than DWI for small cancers; combining both improves confidence.
• EUS sensitivity for small pancreatic cancers: 94.4%, surpassing CT and MRI. Particularly useful when CT is indeterminate (sensitivity 85%, specificity 58%).

Indirect Findings (secondary parenchymal/ductal changes) - Key for Early Detection:

• Three morphologic patterns: slimness (long-segment thinning), cave-in (unilateral depression), and slit (wedge-shaped)
• Progressive FPPA is suspicious for malignancy; if atrophy transitions to focal enlargement, this strongly suggests early PDAC
• Thin-section imaging and multiplanar reformations parallel to the pancreatic axis improve detection

Combining Findings:

IPMN Management (2024 Kyoto Guidelines)

• Obstructive jaundice with pancreatic head cyst
• Enhancing mural nodule ≥5 mm or solid component
• MPD ≥10 mm
• Positive/suspicious cytology

• Acute pancreatitis, elevated CA 19-9, new-onset diabetes, cyst ≥30 mm, small enhancing nodule (<5 mm), thickened/enhancing cyst wall, MPD 5-10 mm, abrupt caliber change with upstream atrophy, cyst growth ≥2.5 mm/year

Role of Imaging Modalities

Diagnostic Workup

1. CA 19-9: Useful when elevated in context, but poor sensitivity/specificity for screening; ~5-10% of individuals are Lewis antigen-negative and cannot synthesize CA 19-9
2. EUS-FNA/biopsy: Gold standard for visible lesions; sensitivity 85-92%, specificity 96-98%; not effective for microscopic PanIN
3. ERP-based cytology (SPACE): Repetitive pancreatic juice sampling via nasopancreatic drainage tube; useful for stage 0 confirmation; mostly Japanese data; high procedural risk
4. Liquid biopsy (ctDNA, circulating tumor RNA, extracellular vesicles): Emerging; KRAS mutations and DNA methylation are key targets; combining with CA 19-9 improves accuracy; still developmental for early-stage disease

Proposed Diagnostic Algorithm (Figure 13)

• MPD stricture: High priority - if true precancerous lesion, progression to invasive cancer occurs within ~1 year. Evaluate with EUS; if new/progressive, proceed to ERP-based cytology.
• FPPA: More complex - may show subtle changes for years before a ~1-year progressive phase. Follow-up intervals should be individualized based on suspicion level and risk factors.
• Surgery without biopsy may be appropriate in select patients with high clinical suspicion and clearly resectable disease (especially stage 0).
• Multidisciplinary discussion is essential to balance overtreatment risk against delayed diagnosis.

Key Takeaways for Radiologists

1. Indirect imaging findings (FPPA, FFPE, MPD stricture, side-branch dilatation) often precede tumor visualization by 1-5 years - recognizing them is the key opportunity for early intervention.
2. Multiple concurrent indirect findings substantially increase confidence in malignancy.
3. Progressive changes over serial imaging are more meaningful than single time-point findings.
4. The entire pancreas - not just known cysts - must be evaluated in patients with IPMNs.
5. Radial clustering of microcysts or dilated side branches around a focal point is a characteristic early PDAC pattern.
6. EUS should be pursued whenever cross-sectional imaging is indeterminate for a small lesion.

FAPI PET/CT for Early Pancreatic Cancer Detection: Recent Clinical Evidence

Why FAPI Over FDG?

1. Meta-Analysis (Highest Evidence)

2. Landmark Observational Trial

• Sensitivity 100%, PPV 96.3% per patient for histopathology-validated lesions
• Head-to-head per-lesion detection vs. FDG vs. CE-CT: 68Ga-FAPI 84.7% / FDG 46.5% / CT 52.9%
• SUVmax significantly correlated with histopathologic FAP expression (Spearman r = 0.48, P = 0.04)
• Minor/major management changes in 8.4% of patients
• Substantial interreader agreement (Fleiss κ = 0.77)

3. Key Head-to-Head Comparisons

• FAPI-04 identified all 62 patients; FDG missed 1
• Primary tumor SUVmax: FAPI 10.63 vs. FDG 2.87 (P < 0.0001) - nearly 4x higher uptake
• Lymph node metastasis SUVmax: 2.90 vs. 1.43 (P < 0.0001)
• Liver metastasis SUVmax: 6.11 vs. 3.10 (P = 0.002)
• 14 patients upstaged by FAPI; only 1 downstaged
• N and M staging significantly superior with FAPI (P = 0.002 and P = 0.008)

• FAPI-74 sensitivity for primary pancreatic tumors: 100% vs. 86% for FDG
• Superior detection of local recurrence (92% vs. 56%, P = 0.021), lymph nodes (71% vs. 40%, P < 0.001), and bone/visceral metastases (98% vs. 47%, P < 0.001)
• Upstaged 25% of treatment-naïve patients

4. Phase 2 Trial with Histopathologic Validation

• 155 participants enrolled; median age 62 years
• Patient-based PPV for detecting FAP-positive tumors: 90% (95% CI 84-95%)
• Region-based PPV: 92% (85-96%)
• Confirmed safety - only 5/90 adverse events possibly related to the tracer; none serious
• Scans acquired at a median of just 11 minutes post-injection (fast protocol feasibility)
• Conclusion: confirms [⁶⁸Ga]Ga-FAPI-46 as a valid imaging biomarker for FAP-expressing tumors

5. Preventing Futile Surgery

• 31% of patients (5/16) had CT-occult metastases detected by ⁶⁸Ga-FAPI PET/CT, preventing unnecessary surgery
• SUVmax of the primary tumor correlated with lymph node ratio in resected specimens (P = 0.04) - suggesting FAPI SUV may be a surrogate for tumor biology/aggressiveness
• This is directly relevant to the question of "early" vs. locally advanced cancer: FAPI catches hidden M1 disease that standard staging misses

6. Novel Applications: Dynamic Imaging & Digital Biopsy

• Used dynamic FAPI-46 PET with voxel-level kinetic analysis ("digital biopsy") in 47 patients with morphologically unclear pancreatic lesions
• Key finding: malignant voxels show slower FAPI clearance than non-malignant pathology - this kinetic signature differentiates PDAC from inflammatory disease
• Could overcome the false-positive problem that limits FAPI's specificity in early detection settings

• Repeatability coefficients: SUVmean 23.7%, SUVpeak 23.9%, SUVmax 29.8%
• ICCs all >0.98 - excellent repeatability
• Establishes the measurement foundation needed for FAPI to be used as a response biomarker in clinical trials

7. Active Phase 3 Trials (2026)

• Evaluating [¹⁸F]FAPI-74 PET/CT specifically for detecting metastatic disease in PDAC
• 18 sites, 200 subjects planned over 24 months
• Primary endpoints: sensitivity and specificity for M1 detection
• Uses ¹⁸F-labeling (longer half-life than ⁶⁸Ga, enabling wider distribution)

Summary & Limitations