Pancreatic parenchymal atrophy, abbreviated PPA, is defined on computed tomography as a narrowing of the pancreatic tissue below a line connecting the cephalic and caudal margins of the lesion. It is not a single disease but the end point of many different insults. Free radicals generated by alcohol damage acinar cells, the workhorses of the pancreas, triggering inflammation and fibrosis; alcohol and its metabolites also directly injure these cells, promoting lipid accumulation and cell loss. Alcohol and other substances raise protein concentrations in pancreatic juice, forming plugs that obstruct the ducts. Acute pancreatitis causes acinar cell necrosis, which heals as scar tissue. In animal models, both obstruction of pancreatic duct flow and ischemia of the parenchyma produced atrophy, and impaired blood flow drove the gland to replace itself with fat. Precancerous lesions called pancreatic intraepithelial neoplasia, or PanIN, can block the flow of juice and set off the same cascade of fibrosis and shrinkage.

The review classifies PPA into three morphological types with distinct clinical meanings. Focal PPA, or FPPA, appears as localized or segmental wasting of tissue adjacent to a duct narrowing, or sometimes without any visible narrowing at all. Upstream PPA, or UPPA, describes global atrophy of the tail-side pancreas downstream of a main duct stricture. Diffuse PPA, or DPPA, involves moderate to severe replacement of the entire gland by fat and is linked to metabolic factors and chronic inflammation rather than to early malignancy. The authors excluded DPPA from their surgical analysis and focused on the focal and upstream forms, because these are the patterns most tightly associated with early pancreatic cancer. FPPA is frequently confused with the generalized atrophy that follows pancreatitis or with pancreatic steatosis, and clinicians must integrate multimodal imaging, changes over time, and the patient’s medical history to tell them apart.

Under the microscope, PPA is a dynamic imbalance between the regeneration and loss of pancreatic epithelial cells. Acinar cells make up more than 90 percent of the pancreatic parenchyma, and their loss is the defining feature. The vacant space is filled by adipocytes and fibroblasts, which is why the authors describe atrophy and fat replacement as two sides of the same coin. Pathologists have traditionally defined pancreatic fat replacement as replacement of more than 25 percent of the parenchyma by adipocytes, though recent reports have proposed a lower threshold of 10 percent. Even in severe cases, the lobular architecture of the gland is generally preserved, and in extreme fat replacement only the ducts and islets remain embedded in adipose tissue. The literature offers a confusing array of names for this process, including pancreatic lipomatosis, fatty infiltration, lipomatous pseudohypertrophy, and pancreatic steatosis, and no consensus definition yet exists.

The epidemiology is striking. Pancreatic fat replacement affects an estimated 16 percent of the global population, most of whom never develop symptoms. The average fat content of the pancreas in people in their eighties has been reported at 35 percent, showing that the process advances with age. Diffuse fat replacement is associated with acute and chronic pancreatitis, pancreatic tumors, diabetes, and arteriosclerosis, and with the rare genetic disorder Shwachman–Diamond syndrome, which produces primary lipomatous pseudohypertrophy alongside aplastic anemia and exocrine dysfunction. Localized fat replacement, however, is a different story: it is frequently linked to PanIN and to intraductal papillary mucinous neoplasms. High-grade PanIN is often invisible as a mass on scans; what reveals it is precisely this indirect signature of localized fat replacement, which can prompt further investigation with pancreatic duct cytology or endoscopic ultrasound-guided biopsy.

The evidence connecting atrophy to early cancer is compelling. In a recent large-scale study, suspicious findings of pancreatic cancer were visible on prediagnostic CT scans in 47.8 percent of patients, and the most characteristic finding was progressive focal parenchymal atrophy appearing roughly 2.7 years before diagnosis, followed by progressive main pancreatic duct dilatation about 1.1 years before diagnosis. Focal atrophy occurred more often in the body and tail of the pancreas than in the head. A systematic review found that the pooled prevalence of intrapancreatic fat deposition in people with pancreatic cancer or premalignant lesions was 52 percent, and the presence of such lesions significantly increased the risk of fat deposition. Focal pancreatic steatosis on CT may even indicate pancreatic carcinoma in situ, the earliest stage of the disease, when abnormal cells are still confined to the duct lining.

Detecting these changes requires a deliberate imaging strategy. The authors identify contrast-enhanced CT with three-directional imaging as the ideal first-line modality, with MRI, magnetic resonance cholangiopancreatography, and endoscopic ultrasound serving as sensitive tools for mapping the distribution of atrophic and fatty lesions. Endoscopic ultrasound offers sufficient resolution to detect changes of carcinoma in situ, which is often accompanied by chronic pancreatitis and fatty infiltration in the background parenchyma. Intraductal ultrasonography can simultaneously visualize lesions inside the duct and the surrounding anatomy to guide resection. Cytological evaluation of pancreatic juice obtained by endoscopic retrograde cholangiopancreatography can confirm the presence of cancer cells, but the sensitivity of this technique is limited and the risk of post-procedure pancreatitis is real, so the authors recommend reserving it for selected patients in whom the result would change management.

The surgical implications are the heart of the review. In one key study, patients with focal or upstream atrophy showed significantly longer intraductal lateral cancer extension than those without atrophy, with a median extension of 20.0 millimeters versus 5.0 millimeters. Two patients in that series had positive resection margins despite wide resection of the atrophic area, and three developed recurrence in the remnant pancreas. Because cancer can spread along the duct more than twice the length of the visible atrophy, the authors recommend that surgeons choose the transection line carefully and rely on frozen-section histopathology to guide additional resection. Total pancreatectomy should be avoided when feasible, but additional pancreatectomy is justified for positive margins or for areas of chronic pancreatitis and fat infiltration. For lesions in the head or tail, pancreatoduodenectomy or distal pancreatectomy with appropriate margins carries a low risk of positive margins; for body lesions, an extended pancreatoduodenectomy is technically easier to extend intraoperatively.

Not every atrophic pancreas demands radical surgery. When no tumor is detectable and carcinoma in situ is suspected, limited procedures such as middle pancreatectomy or spleen-preserving distal pancreatectomy are considered acceptable, particularly since minimally invasive techniques have improved their safety. However, the authors issue a sharp warning: clinically T1 pancreatic cancers, even those smaller than 10 millimeters, can already show regional lymph node metastasis or invasion of the periarterial neural plexuses, so limited resection without lymph node dissection must be avoided once invasive cancer is suspected. Patients with carcinoma in situ face a real risk of recurrence in the remnant pancreas and should be monitored every six months for more than five years, with repeated pancreatectomy considered if recurrence appears. Prognosis falls steeply with invasion: the predicted five-year overall survival is 76 percent for carcinoma in situ, 36 percent for tumors under 1 centimeter, and just 17 percent for tumors of 1.1 to 2.0 centimeters.

The review’s authors acknowledge that major questions remain unresolved. Most studies of PPA are retrospective and rely on CT definitions, and no research has yet determined whether atrophy is a consequence of cancer or a contributor to carcinogenesis itself. The evidence linking the length of intraductal cancer extension to the choice of treatment is scarce, and nationwide studies are needed. Still, the practical message is clear and potentially transformative: a focal shadow of a shrinking pancreas on a routine scan, appearing years before any mass, may be the first and only chance to catch pancreatic cancer at its most curable stage. For clinicians, recognizing that pattern, and knowing how to operate on it, could mean the difference between a 76 percent five-year survival and a 17 percent one.

Subject of Research: Pancreatic parenchymal atrophy as an early imaging marker of pancreatic ductal adenocarcinoma and its surgical management

Article Title: Pancreatic Atrophy: A Narrative Review and Surgical Interpretation

Article References: Fujino, R., Okada, K.-I., Masugi, Y., Iwasaki, E., & Mori, M. (2026). Pancreatic Atrophy: A Narrative Review and Surgical Interpretation. Annals of Gastroenterological Surgery, 10(5), 1419-1427. https://doi.org/10.1002/ags3.70230

Image Credits: AI Generated

DOI: 10.1002/ags3.70230

Keywords: pancreatic cancer, pancreatic parenchymal atrophy, pancreatic ductal adenocarcinoma, computed tomography, pancreatic fat replacement, PanIN, carcinoma in situ, pancreatectomy, endoscopic ultrasound, surgical margins, early detection, gastrointestinal surgery