Pancreas and Biliary Tract

Both endoscopic ultrasound–guided rendezvous (EUS-RV) and precut sphincterotomy are equally effective salvage techniques for patients with benign biliary obstruction and difficult bile duct cannulation, new data suggest.

Selective deep cannulation of the common bile duct remains the key rate-limiting step in successful endoscopic retrograde cholangiopancreatography (ERCP), especially in benign biliary disease.

In cases of difficult cannulation, the traditional fallback has been precut sphincterotomy. Recently, EUS-RV has emerged as an alternative. However, head-to-head comparisons of these salvage techniques in homogeneous patient populations have been lacking, until now. 

A team led by Arup Choudhury, MD, DM, with Department of Gastroenterology, Postgraduate Institute of Medical Education and Research, Chandigarh, India, compared the two salvage techniques in a single-center, randomized controlled trial of 100 patients with benign biliary disease and difficult bile duct cannulation.

There were 50 patients in each group. When one technique failed, patients were crossed over to the other technique.

The technical success rate for achieving deep biliary, the primary outcome measure, was similar with EUS-RV and precut sphincterotomy (92% and 90%, respectively; P = 1.00; relative risk [RR], 1.02), the authors reported in Annals of Internal Medicine.

Median procedure time was also comparable (10.1 minutes with EUS-RV and 9.75 minutes with precut sphincterotomy). As expected, radiation exposure was notably higher in the EUS-RV group (median, 200.2 vs 67.8 mGy).

There was no difference in overall complication rates (12% and 10%, respectively; RR, 1.20).

Five patients in each group (10%) developed post-ERCP pancreatitis (PEP); one patient in the EUS-RV had moderately severe pancreatitis, whereas the rest had mild pancreatitis.

In an exploratory analysis of the subcohort of 72 patients who did not have 1 or more inadvertent pancreatic duct cannulation, two (5.6%) patients in the precut sphincterotomy group had PEP, whereas none of the patients in the EUS-RV had PEP (RR, 0.21). The investigators caution that a larger, multicenter, randomized controlled trial would be required to evaluate the “probable benefit” of lower PEP in the EUS-RV approach.

None of the patients had bleeding or perforation, but two (4%) patients in the EUS-RV group had an infection after the intervention. One required repeated ERCP due to post procedure cholangitis, whereas the other developed lower respiratory tract infection with transient acute lung injury and sputum showing gram-negative organism. None of the patients required surgical intervention.

“Interestingly,” said the investigators, on crossover from one salvage technique to the other, all of the cases could be successfully cannulated, suggesting the two salvage techniques are “complementary to each other and can help achieve successful cannulation in all cases when used in any sequence.”

Summing up, they said it appears from this head-to-head comparison that both EUS-RV and precut sphincterotomy can be considered effective salvage techniques in expert centers with similar safety and success profiles.

Limitations included the single-center design with both procedures performed by expert operators. EUS-RV entailed additional cost of needle and use of a separate scope, and a cost-efficacy analysis was not done.

This study had no specific funding. Disclosures for the authors are available with the original article.

A version of this article appeared on Medscape.com.

Both endoscopic ultrasound–guided rendezvous (EUS-RV) and precut sphincterotomy are equally effective salvage techniques for patients with benign biliary obstruction and difficult bile duct cannulation, new data suggest.

Selective deep cannulation of the common bile duct remains the key rate-limiting step in successful endoscopic retrograde cholangiopancreatography (ERCP), especially in benign biliary disease.

In cases of difficult cannulation, the traditional fallback has been precut sphincterotomy. Recently, EUS-RV has emerged as an alternative. However, head-to-head comparisons of these salvage techniques in homogeneous patient populations have been lacking, until now. 

A team led by Arup Choudhury, MD, DM, with Department of Gastroenterology, Postgraduate Institute of Medical Education and Research, Chandigarh, India, compared the two salvage techniques in a single-center, randomized controlled trial of 100 patients with benign biliary disease and difficult bile duct cannulation.

There were 50 patients in each group. When one technique failed, patients were crossed over to the other technique.

The technical success rate for achieving deep biliary, the primary outcome measure, was similar with EUS-RV and precut sphincterotomy (92% and 90%, respectively; P = 1.00; relative risk [RR], 1.02), the authors reported in Annals of Internal Medicine.

Median procedure time was also comparable (10.1 minutes with EUS-RV and 9.75 minutes with precut sphincterotomy). As expected, radiation exposure was notably higher in the EUS-RV group (median, 200.2 vs 67.8 mGy).

There was no difference in overall complication rates (12% and 10%, respectively; RR, 1.20).

Five patients in each group (10%) developed post-ERCP pancreatitis (PEP); one patient in the EUS-RV had moderately severe pancreatitis, whereas the rest had mild pancreatitis.

In an exploratory analysis of the subcohort of 72 patients who did not have 1 or more inadvertent pancreatic duct cannulation, two (5.6%) patients in the precut sphincterotomy group had PEP, whereas none of the patients in the EUS-RV had PEP (RR, 0.21). The investigators caution that a larger, multicenter, randomized controlled trial would be required to evaluate the “probable benefit” of lower PEP in the EUS-RV approach.

None of the patients had bleeding or perforation, but two (4%) patients in the EUS-RV group had an infection after the intervention. One required repeated ERCP due to post procedure cholangitis, whereas the other developed lower respiratory tract infection with transient acute lung injury and sputum showing gram-negative organism. None of the patients required surgical intervention.

“Interestingly,” said the investigators, on crossover from one salvage technique to the other, all of the cases could be successfully cannulated, suggesting the two salvage techniques are “complementary to each other and can help achieve successful cannulation in all cases when used in any sequence.”

Summing up, they said it appears from this head-to-head comparison that both EUS-RV and precut sphincterotomy can be considered effective salvage techniques in expert centers with similar safety and success profiles.

Limitations included the single-center design with both procedures performed by expert operators. EUS-RV entailed additional cost of needle and use of a separate scope, and a cost-efficacy analysis was not done.

This study had no specific funding. Disclosures for the authors are available with the original article.

A version of this article appeared on Medscape.com.

Both endoscopic ultrasound–guided rendezvous (EUS-RV) and precut sphincterotomy are equally effective salvage techniques for patients with benign biliary obstruction and difficult bile duct cannulation, new data suggest.

Selective deep cannulation of the common bile duct remains the key rate-limiting step in successful endoscopic retrograde cholangiopancreatography (ERCP), especially in benign biliary disease.

In cases of difficult cannulation, the traditional fallback has been precut sphincterotomy. Recently, EUS-RV has emerged as an alternative. However, head-to-head comparisons of these salvage techniques in homogeneous patient populations have been lacking, until now. 

A team led by Arup Choudhury, MD, DM, with Department of Gastroenterology, Postgraduate Institute of Medical Education and Research, Chandigarh, India, compared the two salvage techniques in a single-center, randomized controlled trial of 100 patients with benign biliary disease and difficult bile duct cannulation.

There were 50 patients in each group. When one technique failed, patients were crossed over to the other technique.

The technical success rate for achieving deep biliary, the primary outcome measure, was similar with EUS-RV and precut sphincterotomy (92% and 90%, respectively; P = 1.00; relative risk [RR], 1.02), the authors reported in Annals of Internal Medicine.

Median procedure time was also comparable (10.1 minutes with EUS-RV and 9.75 minutes with precut sphincterotomy). As expected, radiation exposure was notably higher in the EUS-RV group (median, 200.2 vs 67.8 mGy).

There was no difference in overall complication rates (12% and 10%, respectively; RR, 1.20).

Five patients in each group (10%) developed post-ERCP pancreatitis (PEP); one patient in the EUS-RV had moderately severe pancreatitis, whereas the rest had mild pancreatitis.

In an exploratory analysis of the subcohort of 72 patients who did not have 1 or more inadvertent pancreatic duct cannulation, two (5.6%) patients in the precut sphincterotomy group had PEP, whereas none of the patients in the EUS-RV had PEP (RR, 0.21). The investigators caution that a larger, multicenter, randomized controlled trial would be required to evaluate the “probable benefit” of lower PEP in the EUS-RV approach.

None of the patients had bleeding or perforation, but two (4%) patients in the EUS-RV group had an infection after the intervention. One required repeated ERCP due to post procedure cholangitis, whereas the other developed lower respiratory tract infection with transient acute lung injury and sputum showing gram-negative organism. None of the patients required surgical intervention.

“Interestingly,” said the investigators, on crossover from one salvage technique to the other, all of the cases could be successfully cannulated, suggesting the two salvage techniques are “complementary to each other and can help achieve successful cannulation in all cases when used in any sequence.”

Summing up, they said it appears from this head-to-head comparison that both EUS-RV and precut sphincterotomy can be considered effective salvage techniques in expert centers with similar safety and success profiles.

Limitations included the single-center design with both procedures performed by expert operators. EUS-RV entailed additional cost of needle and use of a separate scope, and a cost-efficacy analysis was not done.

This study had no specific funding. Disclosures for the authors are available with the original article.

A version of this article appeared on Medscape.com.

The most common pathology affecting the pancreas is excess intra-pancreatic fat deposition (IPFD), often called fatty pancreas disease (FPD) — a disorder experienced by roughly one fifth of the world’s population. Although it is more common than type 2 diabetes, pancreatitis, and pancreatic cancer combined, it has remained relatively obscure.

By contrast, fatty liver — once called nonalcoholic fatty liver disease and recently renamed metabolic dysfunction–associated steatotic liver disease (MASLD) — is well-known.

“When it comes to diseases of the liver and pancreas, the liver is the big brother that has gotten all the attention, while the pancreas is the neglected little stepbrother that’s not sufficiently profiled in most medical textbooks and gets very little attention,” Max Petrov, MD, MPH, PhD, professor of pancreatology, University of Auckland, New Zealand, said in an interview. “The phenomenon of fatty pancreas has been observed for decades, but it is underappreciated and underrecognized.”

 

As early as 1926, fat depositions were identified during autopsies, but the condition remained relatively unknown, Mohammad Bilal, MD, associate professor of medicine-gastroenterology, University of Colorado Anschutz Medical Campus, Aurora, said in an interview. “Fortunately, FPD has recently been receiving more focus.”

Generally, healthy individuals have small amounts of fat in their pancreas. IPFD is defined as “the diffuse presence of fat in the pancreas, measured on a continuous scale,” and FPD refers to IPFD above the upper limit of normal. While there is no clear consensus as to what the normal range is, studies suggest it’s a pancreatic fat content ranging from 1.8% to 10.4%.

FPD’s “most important implication is that it can be a precursor for more challenging and burdensome diseases of the pancreas,” Petrov said.

Fatty changes in the pancreas affect both its endocrine and exocrine systems. FPD is associated with type 2 diabetes, the most common disease of the endocrine pancreas, as well as pancreatitis and pancreatic cancer, the most common diseases of the exocrine pancreas. It’s also implicated in the development of carotid atherosclerosis, pancreatic fistula following surgery, and exocrine pancreatic insufficiency (EPI).

 

A ‘Pandora’s Box’

Up to half of people with fatty pancreas are lean. The condition isn’t merely caused by an overflow of fat from the liver into the pancreas in people who consume more calories than they burn, Petrov said. Neither robust postmortem nor biopsy studies have found a statistically significant association between fatty deposition in the pancreas and liver fat.

Compared with the way people accumulate liver fat, the development of FPD is more complex, Petrov said.

“Hepatic fat is a relatively simple process: Lipid droplets accumulate in the hepatocytes; but, in the pancreas, there are several ways by which fat may accumulate,” he said.

One relates to the location of the pancreas within visceral, retroperitoneal fat, Petrov said. That fat can migrate and build up between pancreatic lobules.

Fat also can accumulate inside the lobes. This process can involve a buildup of fat droplets in acinar and stellate cells on the exocrine side and in the islets of Langerhans on the endocrine side. Additionally, when functional pancreatic cells die, particularly acinar cells, adult stem cells may replace them with adipocytes. Transformation of acinar cells into fat cells — a process called acinar-to-adipocyte transdifferentiation — also may be a way fat accumulates inside the lobes, Petrov said.

The accumulation of fat is a response to a wide array of insults to the pancreas over time. For example, obesity and metabolic syndrome lead to the accumulation of adipocytes and fat infiltration, whereas alcohol abuse and viral infections may lead to the death of acinar cells, which produce digestive enzymes.

Ultimately, the negative changes produced by excess fat in the pancreas are the origin of all common noninherited pancreatic diseases, bringing them under one umbrella, Petrov maintained. He dubbed this hypothesis PANcreatic Diseases Originating from intRapancreatic fAt (PANDORA).

The type of cells involved has implications for which disease may arise. For example, fat infiltration in stellate cells may promote pancreatic cancer, whereas its accumulation in the islets of Langerhans, which produce insulin and glucagon, is associated with type 2 diabetes.

The PANDORA hypothesis has eight foundational principles:

• Fatty pancreas is a key driver of pancreatic diseases in most people.
• Inflammation within the pancreatic microenvironment results from overwhelming lipotoxicity fueled by fatty pancreas.
• Aberrant communication between acinar cells involving lipid droplets drives acute pancreatitis.
• The pancreas responds to lipotoxicity with fibrosis and calcification — the hallmarks of chronic pancreatitis.
• Fat deposition affects signaling between stellate cells and other components of the microenvironment in ways that raise the risk for pancreatic cancer.
• The development of diabetes of the exocrine pancreas and EPI is affected by the presence of fatty pancreas.
• The higher risk for pancreatic disease in older adults is influenced by fatty pancreas.
• The multipronged nature of intrapancreatic fat deposition accounts for the common development of one pancreatic disease after another.

The idea that all common pancreatic diseases are the result of pathways emanating from FPD could “explain the bidirectional relationship between diabetes and pancreatitis or pancreatic cancer,” Petrov said.

 

Risk Factors, Symptoms, and Diagnosis

A variety of risk factors are involved in the accumulation of fat that may lead to pancreatic diseases, including aging, cholelithiasis, dyslipidemia, drugs/toxins (eg, steroids), genetic predisposition, iron overload, diet (eg, fatty foods, ultraprocessed foods), heavy alcohol use, overweight/obesity, pancreatic duct obstruction, tobacco use, viral infection (eg, hepatitis B, COVID-19), severe malnutrition, prediabetes, and dysglycemia.

Petrov described FPD as a “silent disease” that’s often asymptomatic, with its presence emerging as an incidental finding during abdominal ultrasonography for other reasons. However, patients may sometimes experience stomach pain or nausea if they have concurrent diseases of the pancreas, he said.

There are no currently available lab tests that can definitively detect the presence of FPD. Rather, the gold standard for a noninvasive diagnosis of FPD is MRI, with CT as the second-best choice, Petrov said.

In countries where advanced imaging is not available, a low-cost alternative might be a simple abdominal ultrasound, but it is not definitive, he said. “It’s operator-dependent and can be subjective.”

Some risk factors, such as derangements of glucose and lipid metabolism, especially in the presence of heavy alcohol use and a high-fat diet, can “be detected on lab tests,” Petrov said. “This, in combination with the abdominal ultrasound, might suggest the patients will benefit from deeper investigation, including MRI.”

Because the exocrine pancreas helps with digestion of fatty food, intralobular fatty deposits or replacement of pancreatic exocrine cells with adipose cells can lead to steatorrhea, Bilal said.

“Fat within the stool or oily diarrhea is a clue to the presence of FPD,” Bilal said.

Although this symptom isn’t unique to FPD and is found in other types of pancreatic conditions, its presence suggests that further investigation for FPD is warranted, he added.

 

Common-Sense Treatment Approaches

At present, there are no US Food and Drug Administration–approved treatments for FPD, Petrov said.

“What might be recommended is something along the lines of treatment of MASLD — appropriate diet and physical activity,” he said. Petrov hopes that as the disease entity garners more research attention, more clinical drug trials will be initiated, and new medications are found and approved.

Petrov suggested that there could be a “theoretical rationale” for the use of glucagon-like peptide 1 receptor agonists (GLP-1 RAs) as a treatment, given their effectiveness in multiple conditions, including MASLD, but no human trials have robustly shown specific benefits of these drugs for FPD.

Petrov added that, to date, 12 classes of drugs have been investigated for reducing IPFD: biguanides, sulfonylureas, GLP-1 RAs, thiazolidinediones, dipeptidyl peptidase–4 (DPP-4) inhibitors, sodium-glucose cotransporter 2 inhibitors, statins, fibrates, pancreatic lipase inhibitors, angiotensin II receptor blockers, somatostatin receptor agonists, and antioxidants.

Of these, most have shown promise in preclinical animal models. But only thiazolidinediones, GLP-1 RAs, DPP-4 inhibitors, and somatostatin receptor agonists have been investigated in randomized controlled trials in humans. The findings have been inconsistent, with the active treatment often not achieving statistically significant improvements.

“At this stage of our knowledge, we can’t recommend a specific pharmacotherapy,” Petrov said. But we can suggest dietary changes, such as saturated fat reduction, alcohol reduction, smoking cessation, reduction in consumption of ultraprocessed food, physical exercise, and addressing obesity and other drivers of metabolic disease.

Bilal, who is also a spokesperson for AGA, suggested that pancreatic enzyme replacement therapy, often used to treat pancreatic EPI, may treat some symptoms of FPD such as diarrhea.

Bariatric surgery has shown promise for FPD, in that it can decrease the patient’s body mass and potentially reduce the fat in the pancreas as well as it can improve metabolic diseases and hyperlipidemia. One study showed that it significantly decreased IPFD, fatty acid uptake, and blood flow, and these improvements were associated with more favorable glucose homeostasis and beta-cell function.

However, bariatric surgery is only appropriate for certain patients; is associated with potentially adverse sequelae including malnutrition, anemia, and digestive tract stenosis; and is currently not indicated for FPD.

Bilal advises clinicians to “keep an eye on FPD” if it’s detected incidentally and to screen patients more carefully for MASLD, metabolic disease, and diabetes.

“Although there are no consensus guidelines and recommendations for managing FPD at present, these common-sense approaches will benefit the patient’s overall health and hopefully will have a beneficial impact on pancreatic health as well,” he said.

Petrov reported no relevant financial relationships. Bilal reported being a consultant for Boston Scientific, Steris Endoscopy, and Cook Medical.

A version of this article first appeared on Medscape.com.

The most common pathology affecting the pancreas is excess intra-pancreatic fat deposition (IPFD), often called fatty pancreas disease (FPD) — a disorder experienced by roughly one fifth of the world’s population. Although it is more common than type 2 diabetes, pancreatitis, and pancreatic cancer combined, it has remained relatively obscure.

By contrast, fatty liver — once called nonalcoholic fatty liver disease and recently renamed metabolic dysfunction–associated steatotic liver disease (MASLD) — is well-known.

“When it comes to diseases of the liver and pancreas, the liver is the big brother that has gotten all the attention, while the pancreas is the neglected little stepbrother that’s not sufficiently profiled in most medical textbooks and gets very little attention,” Max Petrov, MD, MPH, PhD, professor of pancreatology, University of Auckland, New Zealand, said in an interview. “The phenomenon of fatty pancreas has been observed for decades, but it is underappreciated and underrecognized.”

 

As early as 1926, fat depositions were identified during autopsies, but the condition remained relatively unknown, Mohammad Bilal, MD, associate professor of medicine-gastroenterology, University of Colorado Anschutz Medical Campus, Aurora, said in an interview. “Fortunately, FPD has recently been receiving more focus.”

Generally, healthy individuals have small amounts of fat in their pancreas. IPFD is defined as “the diffuse presence of fat in the pancreas, measured on a continuous scale,” and FPD refers to IPFD above the upper limit of normal. While there is no clear consensus as to what the normal range is, studies suggest it’s a pancreatic fat content ranging from 1.8% to 10.4%.

FPD’s “most important implication is that it can be a precursor for more challenging and burdensome diseases of the pancreas,” Petrov said.

Fatty changes in the pancreas affect both its endocrine and exocrine systems. FPD is associated with type 2 diabetes, the most common disease of the endocrine pancreas, as well as pancreatitis and pancreatic cancer, the most common diseases of the exocrine pancreas. It’s also implicated in the development of carotid atherosclerosis, pancreatic fistula following surgery, and exocrine pancreatic insufficiency (EPI).

 

A ‘Pandora’s Box’

Up to half of people with fatty pancreas are lean. The condition isn’t merely caused by an overflow of fat from the liver into the pancreas in people who consume more calories than they burn, Petrov said. Neither robust postmortem nor biopsy studies have found a statistically significant association between fatty deposition in the pancreas and liver fat.

Compared with the way people accumulate liver fat, the development of FPD is more complex, Petrov said.

“Hepatic fat is a relatively simple process: Lipid droplets accumulate in the hepatocytes; but, in the pancreas, there are several ways by which fat may accumulate,” he said.

One relates to the location of the pancreas within visceral, retroperitoneal fat, Petrov said. That fat can migrate and build up between pancreatic lobules.

Fat also can accumulate inside the lobes. This process can involve a buildup of fat droplets in acinar and stellate cells on the exocrine side and in the islets of Langerhans on the endocrine side. Additionally, when functional pancreatic cells die, particularly acinar cells, adult stem cells may replace them with adipocytes. Transformation of acinar cells into fat cells — a process called acinar-to-adipocyte transdifferentiation — also may be a way fat accumulates inside the lobes, Petrov said.

The accumulation of fat is a response to a wide array of insults to the pancreas over time. For example, obesity and metabolic syndrome lead to the accumulation of adipocytes and fat infiltration, whereas alcohol abuse and viral infections may lead to the death of acinar cells, which produce digestive enzymes.

Ultimately, the negative changes produced by excess fat in the pancreas are the origin of all common noninherited pancreatic diseases, bringing them under one umbrella, Petrov maintained. He dubbed this hypothesis PANcreatic Diseases Originating from intRapancreatic fAt (PANDORA).

The type of cells involved has implications for which disease may arise. For example, fat infiltration in stellate cells may promote pancreatic cancer, whereas its accumulation in the islets of Langerhans, which produce insulin and glucagon, is associated with type 2 diabetes.

The PANDORA hypothesis has eight foundational principles:

• Fatty pancreas is a key driver of pancreatic diseases in most people.
• Inflammation within the pancreatic microenvironment results from overwhelming lipotoxicity fueled by fatty pancreas.
• Aberrant communication between acinar cells involving lipid droplets drives acute pancreatitis.
• The pancreas responds to lipotoxicity with fibrosis and calcification — the hallmarks of chronic pancreatitis.
• Fat deposition affects signaling between stellate cells and other components of the microenvironment in ways that raise the risk for pancreatic cancer.
• The development of diabetes of the exocrine pancreas and EPI is affected by the presence of fatty pancreas.
• The higher risk for pancreatic disease in older adults is influenced by fatty pancreas.
• The multipronged nature of intrapancreatic fat deposition accounts for the common development of one pancreatic disease after another.

The idea that all common pancreatic diseases are the result of pathways emanating from FPD could “explain the bidirectional relationship between diabetes and pancreatitis or pancreatic cancer,” Petrov said.

 

Risk Factors, Symptoms, and Diagnosis

A variety of risk factors are involved in the accumulation of fat that may lead to pancreatic diseases, including aging, cholelithiasis, dyslipidemia, drugs/toxins (eg, steroids), genetic predisposition, iron overload, diet (eg, fatty foods, ultraprocessed foods), heavy alcohol use, overweight/obesity, pancreatic duct obstruction, tobacco use, viral infection (eg, hepatitis B, COVID-19), severe malnutrition, prediabetes, and dysglycemia.

Petrov described FPD as a “silent disease” that’s often asymptomatic, with its presence emerging as an incidental finding during abdominal ultrasonography for other reasons. However, patients may sometimes experience stomach pain or nausea if they have concurrent diseases of the pancreas, he said.

There are no currently available lab tests that can definitively detect the presence of FPD. Rather, the gold standard for a noninvasive diagnosis of FPD is MRI, with CT as the second-best choice, Petrov said.

In countries where advanced imaging is not available, a low-cost alternative might be a simple abdominal ultrasound, but it is not definitive, he said. “It’s operator-dependent and can be subjective.”

Some risk factors, such as derangements of glucose and lipid metabolism, especially in the presence of heavy alcohol use and a high-fat diet, can “be detected on lab tests,” Petrov said. “This, in combination with the abdominal ultrasound, might suggest the patients will benefit from deeper investigation, including MRI.”

Because the exocrine pancreas helps with digestion of fatty food, intralobular fatty deposits or replacement of pancreatic exocrine cells with adipose cells can lead to steatorrhea, Bilal said.

“Fat within the stool or oily diarrhea is a clue to the presence of FPD,” Bilal said.

Although this symptom isn’t unique to FPD and is found in other types of pancreatic conditions, its presence suggests that further investigation for FPD is warranted, he added.

 

Common-Sense Treatment Approaches

At present, there are no US Food and Drug Administration–approved treatments for FPD, Petrov said.

“What might be recommended is something along the lines of treatment of MASLD — appropriate diet and physical activity,” he said. Petrov hopes that as the disease entity garners more research attention, more clinical drug trials will be initiated, and new medications are found and approved.

Petrov suggested that there could be a “theoretical rationale” for the use of glucagon-like peptide 1 receptor agonists (GLP-1 RAs) as a treatment, given their effectiveness in multiple conditions, including MASLD, but no human trials have robustly shown specific benefits of these drugs for FPD.

Petrov added that, to date, 12 classes of drugs have been investigated for reducing IPFD: biguanides, sulfonylureas, GLP-1 RAs, thiazolidinediones, dipeptidyl peptidase–4 (DPP-4) inhibitors, sodium-glucose cotransporter 2 inhibitors, statins, fibrates, pancreatic lipase inhibitors, angiotensin II receptor blockers, somatostatin receptor agonists, and antioxidants.

Of these, most have shown promise in preclinical animal models. But only thiazolidinediones, GLP-1 RAs, DPP-4 inhibitors, and somatostatin receptor agonists have been investigated in randomized controlled trials in humans. The findings have been inconsistent, with the active treatment often not achieving statistically significant improvements.

“At this stage of our knowledge, we can’t recommend a specific pharmacotherapy,” Petrov said. But we can suggest dietary changes, such as saturated fat reduction, alcohol reduction, smoking cessation, reduction in consumption of ultraprocessed food, physical exercise, and addressing obesity and other drivers of metabolic disease.

Bilal, who is also a spokesperson for AGA, suggested that pancreatic enzyme replacement therapy, often used to treat pancreatic EPI, may treat some symptoms of FPD such as diarrhea.

Bariatric surgery has shown promise for FPD, in that it can decrease the patient’s body mass and potentially reduce the fat in the pancreas as well as it can improve metabolic diseases and hyperlipidemia. One study showed that it significantly decreased IPFD, fatty acid uptake, and blood flow, and these improvements were associated with more favorable glucose homeostasis and beta-cell function.

However, bariatric surgery is only appropriate for certain patients; is associated with potentially adverse sequelae including malnutrition, anemia, and digestive tract stenosis; and is currently not indicated for FPD.

Bilal advises clinicians to “keep an eye on FPD” if it’s detected incidentally and to screen patients more carefully for MASLD, metabolic disease, and diabetes.

“Although there are no consensus guidelines and recommendations for managing FPD at present, these common-sense approaches will benefit the patient’s overall health and hopefully will have a beneficial impact on pancreatic health as well,” he said.

Petrov reported no relevant financial relationships. Bilal reported being a consultant for Boston Scientific, Steris Endoscopy, and Cook Medical.

A version of this article first appeared on Medscape.com.

The most common pathology affecting the pancreas is excess intra-pancreatic fat deposition (IPFD), often called fatty pancreas disease (FPD) — a disorder experienced by roughly one fifth of the world’s population. Although it is more common than type 2 diabetes, pancreatitis, and pancreatic cancer combined, it has remained relatively obscure.

By contrast, fatty liver — once called nonalcoholic fatty liver disease and recently renamed metabolic dysfunction–associated steatotic liver disease (MASLD) — is well-known.

“When it comes to diseases of the liver and pancreas, the liver is the big brother that has gotten all the attention, while the pancreas is the neglected little stepbrother that’s not sufficiently profiled in most medical textbooks and gets very little attention,” Max Petrov, MD, MPH, PhD, professor of pancreatology, University of Auckland, New Zealand, said in an interview. “The phenomenon of fatty pancreas has been observed for decades, but it is underappreciated and underrecognized.”

 

As early as 1926, fat depositions were identified during autopsies, but the condition remained relatively unknown, Mohammad Bilal, MD, associate professor of medicine-gastroenterology, University of Colorado Anschutz Medical Campus, Aurora, said in an interview. “Fortunately, FPD has recently been receiving more focus.”

Generally, healthy individuals have small amounts of fat in their pancreas. IPFD is defined as “the diffuse presence of fat in the pancreas, measured on a continuous scale,” and FPD refers to IPFD above the upper limit of normal. While there is no clear consensus as to what the normal range is, studies suggest it’s a pancreatic fat content ranging from 1.8% to 10.4%.

FPD’s “most important implication is that it can be a precursor for more challenging and burdensome diseases of the pancreas,” Petrov said.

Fatty changes in the pancreas affect both its endocrine and exocrine systems. FPD is associated with type 2 diabetes, the most common disease of the endocrine pancreas, as well as pancreatitis and pancreatic cancer, the most common diseases of the exocrine pancreas. It’s also implicated in the development of carotid atherosclerosis, pancreatic fistula following surgery, and exocrine pancreatic insufficiency (EPI).

 

A ‘Pandora’s Box’

Up to half of people with fatty pancreas are lean. The condition isn’t merely caused by an overflow of fat from the liver into the pancreas in people who consume more calories than they burn, Petrov said. Neither robust postmortem nor biopsy studies have found a statistically significant association between fatty deposition in the pancreas and liver fat.

Compared with the way people accumulate liver fat, the development of FPD is more complex, Petrov said.

“Hepatic fat is a relatively simple process: Lipid droplets accumulate in the hepatocytes; but, in the pancreas, there are several ways by which fat may accumulate,” he said.

One relates to the location of the pancreas within visceral, retroperitoneal fat, Petrov said. That fat can migrate and build up between pancreatic lobules.

Fat also can accumulate inside the lobes. This process can involve a buildup of fat droplets in acinar and stellate cells on the exocrine side and in the islets of Langerhans on the endocrine side. Additionally, when functional pancreatic cells die, particularly acinar cells, adult stem cells may replace them with adipocytes. Transformation of acinar cells into fat cells — a process called acinar-to-adipocyte transdifferentiation — also may be a way fat accumulates inside the lobes, Petrov said.

The accumulation of fat is a response to a wide array of insults to the pancreas over time. For example, obesity and metabolic syndrome lead to the accumulation of adipocytes and fat infiltration, whereas alcohol abuse and viral infections may lead to the death of acinar cells, which produce digestive enzymes.

Ultimately, the negative changes produced by excess fat in the pancreas are the origin of all common noninherited pancreatic diseases, bringing them under one umbrella, Petrov maintained. He dubbed this hypothesis PANcreatic Diseases Originating from intRapancreatic fAt (PANDORA).

The type of cells involved has implications for which disease may arise. For example, fat infiltration in stellate cells may promote pancreatic cancer, whereas its accumulation in the islets of Langerhans, which produce insulin and glucagon, is associated with type 2 diabetes.

The PANDORA hypothesis has eight foundational principles:

• Fatty pancreas is a key driver of pancreatic diseases in most people.
• Inflammation within the pancreatic microenvironment results from overwhelming lipotoxicity fueled by fatty pancreas.
• Aberrant communication between acinar cells involving lipid droplets drives acute pancreatitis.
• The pancreas responds to lipotoxicity with fibrosis and calcification — the hallmarks of chronic pancreatitis.
• Fat deposition affects signaling between stellate cells and other components of the microenvironment in ways that raise the risk for pancreatic cancer.
• The development of diabetes of the exocrine pancreas and EPI is affected by the presence of fatty pancreas.
• The higher risk for pancreatic disease in older adults is influenced by fatty pancreas.
• The multipronged nature of intrapancreatic fat deposition accounts for the common development of one pancreatic disease after another.

The idea that all common pancreatic diseases are the result of pathways emanating from FPD could “explain the bidirectional relationship between diabetes and pancreatitis or pancreatic cancer,” Petrov said.

 

Risk Factors, Symptoms, and Diagnosis

A variety of risk factors are involved in the accumulation of fat that may lead to pancreatic diseases, including aging, cholelithiasis, dyslipidemia, drugs/toxins (eg, steroids), genetic predisposition, iron overload, diet (eg, fatty foods, ultraprocessed foods), heavy alcohol use, overweight/obesity, pancreatic duct obstruction, tobacco use, viral infection (eg, hepatitis B, COVID-19), severe malnutrition, prediabetes, and dysglycemia.

Petrov described FPD as a “silent disease” that’s often asymptomatic, with its presence emerging as an incidental finding during abdominal ultrasonography for other reasons. However, patients may sometimes experience stomach pain or nausea if they have concurrent diseases of the pancreas, he said.

There are no currently available lab tests that can definitively detect the presence of FPD. Rather, the gold standard for a noninvasive diagnosis of FPD is MRI, with CT as the second-best choice, Petrov said.

In countries where advanced imaging is not available, a low-cost alternative might be a simple abdominal ultrasound, but it is not definitive, he said. “It’s operator-dependent and can be subjective.”

Some risk factors, such as derangements of glucose and lipid metabolism, especially in the presence of heavy alcohol use and a high-fat diet, can “be detected on lab tests,” Petrov said. “This, in combination with the abdominal ultrasound, might suggest the patients will benefit from deeper investigation, including MRI.”

Because the exocrine pancreas helps with digestion of fatty food, intralobular fatty deposits or replacement of pancreatic exocrine cells with adipose cells can lead to steatorrhea, Bilal said.

“Fat within the stool or oily diarrhea is a clue to the presence of FPD,” Bilal said.

Although this symptom isn’t unique to FPD and is found in other types of pancreatic conditions, its presence suggests that further investigation for FPD is warranted, he added.

 

Common-Sense Treatment Approaches

At present, there are no US Food and Drug Administration–approved treatments for FPD, Petrov said.

“What might be recommended is something along the lines of treatment of MASLD — appropriate diet and physical activity,” he said. Petrov hopes that as the disease entity garners more research attention, more clinical drug trials will be initiated, and new medications are found and approved.

Petrov suggested that there could be a “theoretical rationale” for the use of glucagon-like peptide 1 receptor agonists (GLP-1 RAs) as a treatment, given their effectiveness in multiple conditions, including MASLD, but no human trials have robustly shown specific benefits of these drugs for FPD.

Petrov added that, to date, 12 classes of drugs have been investigated for reducing IPFD: biguanides, sulfonylureas, GLP-1 RAs, thiazolidinediones, dipeptidyl peptidase–4 (DPP-4) inhibitors, sodium-glucose cotransporter 2 inhibitors, statins, fibrates, pancreatic lipase inhibitors, angiotensin II receptor blockers, somatostatin receptor agonists, and antioxidants.

Of these, most have shown promise in preclinical animal models. But only thiazolidinediones, GLP-1 RAs, DPP-4 inhibitors, and somatostatin receptor agonists have been investigated in randomized controlled trials in humans. The findings have been inconsistent, with the active treatment often not achieving statistically significant improvements.

“At this stage of our knowledge, we can’t recommend a specific pharmacotherapy,” Petrov said. But we can suggest dietary changes, such as saturated fat reduction, alcohol reduction, smoking cessation, reduction in consumption of ultraprocessed food, physical exercise, and addressing obesity and other drivers of metabolic disease.

Bilal, who is also a spokesperson for AGA, suggested that pancreatic enzyme replacement therapy, often used to treat pancreatic EPI, may treat some symptoms of FPD such as diarrhea.

Bariatric surgery has shown promise for FPD, in that it can decrease the patient’s body mass and potentially reduce the fat in the pancreas as well as it can improve metabolic diseases and hyperlipidemia. One study showed that it significantly decreased IPFD, fatty acid uptake, and blood flow, and these improvements were associated with more favorable glucose homeostasis and beta-cell function.

However, bariatric surgery is only appropriate for certain patients; is associated with potentially adverse sequelae including malnutrition, anemia, and digestive tract stenosis; and is currently not indicated for FPD.

Bilal advises clinicians to “keep an eye on FPD” if it’s detected incidentally and to screen patients more carefully for MASLD, metabolic disease, and diabetes.

“Although there are no consensus guidelines and recommendations for managing FPD at present, these common-sense approaches will benefit the patient’s overall health and hopefully will have a beneficial impact on pancreatic health as well,” he said.

Petrov reported no relevant financial relationships. Bilal reported being a consultant for Boston Scientific, Steris Endoscopy, and Cook Medical.

A version of this article first appeared on Medscape.com.

Placing two cystic duct stents instead of one during endoscopic transpapillary gallbladder drainage (ETGBD) is associated with a lower rate of unplanned reintervention, according to a retrospective multicenter study.

These findings suggest that endoscopists should prioritize dual stent placement when feasible, and consider adding a second stent in patients who previously received a single stent, James D. Haddad, MD, of the University of Texas Southwestern, Dallas, and colleagues reported.

 

The American Gastroenterological Association (AGA) has recognized the role of endoscopic drainage in managing acute cholecystitis in high-risk patients, but specific guidance on optimal technique and follow-up remains unclear, the investigators wrote in Techniques and Innovations in Gastrointestinal Endoscopy.

“Despite accumulating data and increased interest in this technique, clear guidance on the ideal strategy for ETGBD is lacking,” Dr. Haddad and colleagues wrote. “For example, the optimal size, number, and follow-up of cystic duct stents for patients undergoing ETGBD has not been well established.”

To address this knowledge gap, the investigators analyzed data from 75 patients at five academic medical centers who had undergone ETGBD between June 2013 and October 2022. Patients were divided into two groups based on whether they received one or two cystic duct stents. 

The primary outcome was clinical success, defined as symptom resolution without requiring another drainage procedure. Secondary outcomes included technical success (defined as successful stent placement), along with rates of adverse events and unplanned reinterventions. 

Out of the 75 patients, 59 received a single stent, while 16 received dual stents. The median follow-up time was 407 days overall, with a longer follow-up in the single-stent group (433 days), compared with the double-stent group (118 days).

Clinical success was reported in 81.3% of cases, which technical success was achieved in 88.2% of cases. 

Patients who received two stents had significantly lower rates of unplanned reintervention, compared with those who received a single stent (0% vs 25.4%; P = .02). The median time to unplanned reintervention in the single-stent group was 210 days.

Use of a 7 French stent was strongly associated with placement of two stents (odd ratio [OR], 15.5; P = .01). Similarly, patients with a prior percutaneous cholecystostomy tube were significantly more likely to have two stents placed (OR, 10.8; P = .001).

Adverse event rates were uncommon and not statistically different between groups, with an overall rate of 6.7%. Post-endoscopic retrograde cholangiopancreatography pancreatitis was the most common adverse event, occurring in two patients in the single-stent group and one patient in the double-stent group. There were no reported cases of cystic duct or gallbladder perforation.

“In conclusion,” the investigators wrote, “ETGBD with dual transpapillary gallbladder stenting is associated with a lower rate of unplanned reinterventions, compared with that with single stenting, and has a low rate of adverse events. Endoscopists performing ETGBD should consider planned exchange of solitary transpapillary gallbladder stents or interval ERCP for reattempted placement of a second stent if placement of two stents is not possible at the index ERCP.”

The investigators disclosed relationships with Boston Scientific, Motus GI, and ConMed.







 

Placing two cystic duct stents instead of one during endoscopic transpapillary gallbladder drainage (ETGBD) is associated with a lower rate of unplanned reintervention, according to a retrospective multicenter study.

These findings suggest that endoscopists should prioritize dual stent placement when feasible, and consider adding a second stent in patients who previously received a single stent, James D. Haddad, MD, of the University of Texas Southwestern, Dallas, and colleagues reported.

 

The American Gastroenterological Association (AGA) has recognized the role of endoscopic drainage in managing acute cholecystitis in high-risk patients, but specific guidance on optimal technique and follow-up remains unclear, the investigators wrote in Techniques and Innovations in Gastrointestinal Endoscopy.

“Despite accumulating data and increased interest in this technique, clear guidance on the ideal strategy for ETGBD is lacking,” Dr. Haddad and colleagues wrote. “For example, the optimal size, number, and follow-up of cystic duct stents for patients undergoing ETGBD has not been well established.”

To address this knowledge gap, the investigators analyzed data from 75 patients at five academic medical centers who had undergone ETGBD between June 2013 and October 2022. Patients were divided into two groups based on whether they received one or two cystic duct stents. 

The primary outcome was clinical success, defined as symptom resolution without requiring another drainage procedure. Secondary outcomes included technical success (defined as successful stent placement), along with rates of adverse events and unplanned reinterventions. 

Out of the 75 patients, 59 received a single stent, while 16 received dual stents. The median follow-up time was 407 days overall, with a longer follow-up in the single-stent group (433 days), compared with the double-stent group (118 days).

Clinical success was reported in 81.3% of cases, which technical success was achieved in 88.2% of cases. 

Patients who received two stents had significantly lower rates of unplanned reintervention, compared with those who received a single stent (0% vs 25.4%; P = .02). The median time to unplanned reintervention in the single-stent group was 210 days.

Use of a 7 French stent was strongly associated with placement of two stents (odd ratio [OR], 15.5; P = .01). Similarly, patients with a prior percutaneous cholecystostomy tube were significantly more likely to have two stents placed (OR, 10.8; P = .001).

Adverse event rates were uncommon and not statistically different between groups, with an overall rate of 6.7%. Post-endoscopic retrograde cholangiopancreatography pancreatitis was the most common adverse event, occurring in two patients in the single-stent group and one patient in the double-stent group. There were no reported cases of cystic duct or gallbladder perforation.

“In conclusion,” the investigators wrote, “ETGBD with dual transpapillary gallbladder stenting is associated with a lower rate of unplanned reinterventions, compared with that with single stenting, and has a low rate of adverse events. Endoscopists performing ETGBD should consider planned exchange of solitary transpapillary gallbladder stents or interval ERCP for reattempted placement of a second stent if placement of two stents is not possible at the index ERCP.”

The investigators disclosed relationships with Boston Scientific, Motus GI, and ConMed.







 

Placing two cystic duct stents instead of one during endoscopic transpapillary gallbladder drainage (ETGBD) is associated with a lower rate of unplanned reintervention, according to a retrospective multicenter study.

These findings suggest that endoscopists should prioritize dual stent placement when feasible, and consider adding a second stent in patients who previously received a single stent, James D. Haddad, MD, of the University of Texas Southwestern, Dallas, and colleagues reported.

 

The American Gastroenterological Association (AGA) has recognized the role of endoscopic drainage in managing acute cholecystitis in high-risk patients, but specific guidance on optimal technique and follow-up remains unclear, the investigators wrote in Techniques and Innovations in Gastrointestinal Endoscopy.

“Despite accumulating data and increased interest in this technique, clear guidance on the ideal strategy for ETGBD is lacking,” Dr. Haddad and colleagues wrote. “For example, the optimal size, number, and follow-up of cystic duct stents for patients undergoing ETGBD has not been well established.”

To address this knowledge gap, the investigators analyzed data from 75 patients at five academic medical centers who had undergone ETGBD between June 2013 and October 2022. Patients were divided into two groups based on whether they received one or two cystic duct stents. 

The primary outcome was clinical success, defined as symptom resolution without requiring another drainage procedure. Secondary outcomes included technical success (defined as successful stent placement), along with rates of adverse events and unplanned reinterventions. 

Out of the 75 patients, 59 received a single stent, while 16 received dual stents. The median follow-up time was 407 days overall, with a longer follow-up in the single-stent group (433 days), compared with the double-stent group (118 days).

Clinical success was reported in 81.3% of cases, which technical success was achieved in 88.2% of cases. 

Patients who received two stents had significantly lower rates of unplanned reintervention, compared with those who received a single stent (0% vs 25.4%; P = .02). The median time to unplanned reintervention in the single-stent group was 210 days.

Use of a 7 French stent was strongly associated with placement of two stents (odd ratio [OR], 15.5; P = .01). Similarly, patients with a prior percutaneous cholecystostomy tube were significantly more likely to have two stents placed (OR, 10.8; P = .001).

Adverse event rates were uncommon and not statistically different between groups, with an overall rate of 6.7%. Post-endoscopic retrograde cholangiopancreatography pancreatitis was the most common adverse event, occurring in two patients in the single-stent group and one patient in the double-stent group. There were no reported cases of cystic duct or gallbladder perforation.

“In conclusion,” the investigators wrote, “ETGBD with dual transpapillary gallbladder stenting is associated with a lower rate of unplanned reinterventions, compared with that with single stenting, and has a low rate of adverse events. Endoscopists performing ETGBD should consider planned exchange of solitary transpapillary gallbladder stents or interval ERCP for reattempted placement of a second stent if placement of two stents is not possible at the index ERCP.”

The investigators disclosed relationships with Boston Scientific, Motus GI, and ConMed.







 

Clinicians should be aware of how to manage certain gastrointestinal (GI) and liver conditions associated with pregnancy, such as hyperemesis gravidarum, intrahepatic cholestasis of pregnancy, and acute fatty liver of pregnancy, according to a clinical practice update (CPU) from the American Gastroenterological Association.

Notably, procedures, medications, or other interventions intended to improve maternal health shouldn’t be withheld solely because the patient is pregnant, the authors wrote. Instead, treatments should be personalized based on a risk-benefit assessment.  

 

“Pregnancy causes significant physiological changes that can affect the GI tract and liver function. Some common conditions — such as nausea, vomiting, gastroesophageal reflux disease (GERD), and constipation — may be exacerbated, and underlying GI or liver diseases can behave differently during pregnancy,” said lead author Shivangi Kothari, MD, associate professor of medicine and associate director of endoscopy at the University of Rochester Medical Center and Strong Memorial Hospital, both in Rochester, New York.  

“These conditions can pose significant risks to both the mother and fetus, and their management requires a specialized, updated approach,” she said. “This clinical practice update stresses the need for coordinated, multidisciplinary care among obstetricians, gastroenterologists, hepatologists, and maternal-and-fetal medicine experts to ensure optimal outcomes, particularly in complex or high-risk cases.”  

The update was published online in Gastroenterology.  

 

Pregnancy-Related Concerns

The best path to optimal outcomes is to start early, the authors wrote. Before pregnancy, patients should consider preconception and contraceptive care counseling with a multidisciplinary team that can address GI and liver issues, especially among reproductive-age people who want to become pregnant.

Once pregnant, though, patients shouldn’t be deterred from receiving procedures, medications, or interventions just because they’re pregnant, the authors wrote. Instead, taking an individual approach will help clinicians decide what to do based on the risks and benefits.  

At the beginning of pregnancy, early treatment of nausea and vomiting can reduce progression to hyperemesis gravidarum, the authors wrote. Stepwise treatment can include vitamin B6, doxylamine, hydration, and adequate nutrition, followed by ondansetron, metoclopramide, promethazine, and intravenous glucocorticoids in moderate to severe cases.  

Constipation may also pose a problem because of hormonal, physiological, and medication-related changes. Treatment options can include dietary fiber, lactulose, and polyethylene glycol-based laxatives.  

Patients with certain conditions — such as complex inflammatory bowel disease (IBD), advanced cirrhosis, or liver transplant — should work with a multidisciplinary team to coordinate birth, preferably in a tertiary care center, the authors wrote.  

For patients with IBD, clinical remission helps to improve pregnancy outcomes, including before conception, during pregnancy, and throughout the postpartum period. Biologic agents should be used during pregnancy and postpartum, though methotrexate, thalidomide, and ozanimod should be stopped at least 6 months before conception.  

For patients with chronic hepatitis B, serum hepatitis B virus DNA and liver biochemical levels should be tested. Patients with a serum level > 200,000 IU/mL during the third trimester should be considered for treatment with tenofovir disoproxil fumarate.  

For patients on immunosuppressive therapy for chronic liver diseases or after liver transplantation, therapy should continue at the lowest effective dose. However, mycophenolate mofetil shouldn’t be administered during pregnancy.  

Intrahepatic cholestasis of pregnancy may be diagnosed during the second or third trimester based on pruritus and a serum bile acid level > 10 μmol/L. Treatment should include oral ursodeoxycholic acid, with a total daily dose of 10-15 mg/kg.  

Other pregnancy-related liver diseases — such as pre-eclampsia; hemolysis, elevated liver enzymes, and low platelets syndrome; and acute fatty liver of pregnancy — require careful birth planning and evaluation for possible liver transplantation. For certain high-risk patients, daily aspirin should start at week 12 of gestation.  

In addition, elective endoscopic procedures should wait until after birth, and nonemergent but necessary procedures should be performed during the second trimester. Patients with cirrhosis should undergo evaluation for esophageal varices, and upper endoscopy should happen during the second trimester to guide beta-blocker therapy or endoscopic variceal litigation.  

Endoscopic retrograde cholangiopancreatography can be performed for urgent indications, such as choledocholithiasis, cholangitis, and some gallstone pancreatitis cases, ideally during the second trimester.  

Cholecystectomy is considered safe during pregnancy, with a laparoscopic approach as the standard of care regardless of trimester, though the second trimester is ideal. 

Pregnancy-Related Updates in Practice

Ultimately, clinicians should familiarize themselves with the best practice advice to feel comfortable when counseling and managing pregnancy-related concerns, especially high-risk patients, said Eugenia Shmidt, MD, assistant professor of gastroenterology, hepatology, and nutrition, and founder of the IBD Preconception and Pregnancy Planning Clinic at the University of Minnesota, Minneapolis.

“Half of all patients with GI and liver disease are women, and oftentimes, they don’t have appropriate guidance regarding reproductive health in the context of their disease,” she said. “There exists a very large knowledge gap in this area, particularly because most clinical trials exclude pregnant people.”  

Most importantly, the advice statements can guide practitioners on how to help pregnant patients make informed reproductive decisions, she added.  

“This CPU makes it clear that preconception counseling and multidisciplinary care are key in optimizing reproductive health, regardless of the underlying GI or liver disease,” Shmidt said. “GI practitioners should be counseling women well in advance of pregnancy and recruiting all relevant stakeholders as early as possible, even prior to conception. This way, pregnancy care is not reactive, but instead proactive.”  

The authors received no specific funding for this update. Kothari and Shmidt reported no relevant disclosures.

A version of this article appeared on Medscape.com.

Clinicians should be aware of how to manage certain gastrointestinal (GI) and liver conditions associated with pregnancy, such as hyperemesis gravidarum, intrahepatic cholestasis of pregnancy, and acute fatty liver of pregnancy, according to a clinical practice update (CPU) from the American Gastroenterological Association.

Notably, procedures, medications, or other interventions intended to improve maternal health shouldn’t be withheld solely because the patient is pregnant, the authors wrote. Instead, treatments should be personalized based on a risk-benefit assessment.  

 

“Pregnancy causes significant physiological changes that can affect the GI tract and liver function. Some common conditions — such as nausea, vomiting, gastroesophageal reflux disease (GERD), and constipation — may be exacerbated, and underlying GI or liver diseases can behave differently during pregnancy,” said lead author Shivangi Kothari, MD, associate professor of medicine and associate director of endoscopy at the University of Rochester Medical Center and Strong Memorial Hospital, both in Rochester, New York.  

“These conditions can pose significant risks to both the mother and fetus, and their management requires a specialized, updated approach,” she said. “This clinical practice update stresses the need for coordinated, multidisciplinary care among obstetricians, gastroenterologists, hepatologists, and maternal-and-fetal medicine experts to ensure optimal outcomes, particularly in complex or high-risk cases.”  

The update was published online in Gastroenterology.  

 

Pregnancy-Related Concerns

The best path to optimal outcomes is to start early, the authors wrote. Before pregnancy, patients should consider preconception and contraceptive care counseling with a multidisciplinary team that can address GI and liver issues, especially among reproductive-age people who want to become pregnant.

Once pregnant, though, patients shouldn’t be deterred from receiving procedures, medications, or interventions just because they’re pregnant, the authors wrote. Instead, taking an individual approach will help clinicians decide what to do based on the risks and benefits.  

At the beginning of pregnancy, early treatment of nausea and vomiting can reduce progression to hyperemesis gravidarum, the authors wrote. Stepwise treatment can include vitamin B6, doxylamine, hydration, and adequate nutrition, followed by ondansetron, metoclopramide, promethazine, and intravenous glucocorticoids in moderate to severe cases.  

Constipation may also pose a problem because of hormonal, physiological, and medication-related changes. Treatment options can include dietary fiber, lactulose, and polyethylene glycol-based laxatives.  

Patients with certain conditions — such as complex inflammatory bowel disease (IBD), advanced cirrhosis, or liver transplant — should work with a multidisciplinary team to coordinate birth, preferably in a tertiary care center, the authors wrote.  

For patients with IBD, clinical remission helps to improve pregnancy outcomes, including before conception, during pregnancy, and throughout the postpartum period. Biologic agents should be used during pregnancy and postpartum, though methotrexate, thalidomide, and ozanimod should be stopped at least 6 months before conception.  

For patients with chronic hepatitis B, serum hepatitis B virus DNA and liver biochemical levels should be tested. Patients with a serum level > 200,000 IU/mL during the third trimester should be considered for treatment with tenofovir disoproxil fumarate.  

For patients on immunosuppressive therapy for chronic liver diseases or after liver transplantation, therapy should continue at the lowest effective dose. However, mycophenolate mofetil shouldn’t be administered during pregnancy.  

Intrahepatic cholestasis of pregnancy may be diagnosed during the second or third trimester based on pruritus and a serum bile acid level > 10 μmol/L. Treatment should include oral ursodeoxycholic acid, with a total daily dose of 10-15 mg/kg.  

Other pregnancy-related liver diseases — such as pre-eclampsia; hemolysis, elevated liver enzymes, and low platelets syndrome; and acute fatty liver of pregnancy — require careful birth planning and evaluation for possible liver transplantation. For certain high-risk patients, daily aspirin should start at week 12 of gestation.  

In addition, elective endoscopic procedures should wait until after birth, and nonemergent but necessary procedures should be performed during the second trimester. Patients with cirrhosis should undergo evaluation for esophageal varices, and upper endoscopy should happen during the second trimester to guide beta-blocker therapy or endoscopic variceal litigation.  

Endoscopic retrograde cholangiopancreatography can be performed for urgent indications, such as choledocholithiasis, cholangitis, and some gallstone pancreatitis cases, ideally during the second trimester.  

Cholecystectomy is considered safe during pregnancy, with a laparoscopic approach as the standard of care regardless of trimester, though the second trimester is ideal. 

Pregnancy-Related Updates in Practice

Ultimately, clinicians should familiarize themselves with the best practice advice to feel comfortable when counseling and managing pregnancy-related concerns, especially high-risk patients, said Eugenia Shmidt, MD, assistant professor of gastroenterology, hepatology, and nutrition, and founder of the IBD Preconception and Pregnancy Planning Clinic at the University of Minnesota, Minneapolis.

“Half of all patients with GI and liver disease are women, and oftentimes, they don’t have appropriate guidance regarding reproductive health in the context of their disease,” she said. “There exists a very large knowledge gap in this area, particularly because most clinical trials exclude pregnant people.”  

Most importantly, the advice statements can guide practitioners on how to help pregnant patients make informed reproductive decisions, she added.  

“This CPU makes it clear that preconception counseling and multidisciplinary care are key in optimizing reproductive health, regardless of the underlying GI or liver disease,” Shmidt said. “GI practitioners should be counseling women well in advance of pregnancy and recruiting all relevant stakeholders as early as possible, even prior to conception. This way, pregnancy care is not reactive, but instead proactive.”  

The authors received no specific funding for this update. Kothari and Shmidt reported no relevant disclosures.

A version of this article appeared on Medscape.com.

Clinicians should be aware of how to manage certain gastrointestinal (GI) and liver conditions associated with pregnancy, such as hyperemesis gravidarum, intrahepatic cholestasis of pregnancy, and acute fatty liver of pregnancy, according to a clinical practice update (CPU) from the American Gastroenterological Association.

Notably, procedures, medications, or other interventions intended to improve maternal health shouldn’t be withheld solely because the patient is pregnant, the authors wrote. Instead, treatments should be personalized based on a risk-benefit assessment.  

 

“Pregnancy causes significant physiological changes that can affect the GI tract and liver function. Some common conditions — such as nausea, vomiting, gastroesophageal reflux disease (GERD), and constipation — may be exacerbated, and underlying GI or liver diseases can behave differently during pregnancy,” said lead author Shivangi Kothari, MD, associate professor of medicine and associate director of endoscopy at the University of Rochester Medical Center and Strong Memorial Hospital, both in Rochester, New York.  

“These conditions can pose significant risks to both the mother and fetus, and their management requires a specialized, updated approach,” she said. “This clinical practice update stresses the need for coordinated, multidisciplinary care among obstetricians, gastroenterologists, hepatologists, and maternal-and-fetal medicine experts to ensure optimal outcomes, particularly in complex or high-risk cases.”  

The update was published online in Gastroenterology.  

 

Pregnancy-Related Concerns

The best path to optimal outcomes is to start early, the authors wrote. Before pregnancy, patients should consider preconception and contraceptive care counseling with a multidisciplinary team that can address GI and liver issues, especially among reproductive-age people who want to become pregnant.

Once pregnant, though, patients shouldn’t be deterred from receiving procedures, medications, or interventions just because they’re pregnant, the authors wrote. Instead, taking an individual approach will help clinicians decide what to do based on the risks and benefits.  

At the beginning of pregnancy, early treatment of nausea and vomiting can reduce progression to hyperemesis gravidarum, the authors wrote. Stepwise treatment can include vitamin B6, doxylamine, hydration, and adequate nutrition, followed by ondansetron, metoclopramide, promethazine, and intravenous glucocorticoids in moderate to severe cases.  

Constipation may also pose a problem because of hormonal, physiological, and medication-related changes. Treatment options can include dietary fiber, lactulose, and polyethylene glycol-based laxatives.  

Patients with certain conditions — such as complex inflammatory bowel disease (IBD), advanced cirrhosis, or liver transplant — should work with a multidisciplinary team to coordinate birth, preferably in a tertiary care center, the authors wrote.  

For patients with IBD, clinical remission helps to improve pregnancy outcomes, including before conception, during pregnancy, and throughout the postpartum period. Biologic agents should be used during pregnancy and postpartum, though methotrexate, thalidomide, and ozanimod should be stopped at least 6 months before conception.  

For patients with chronic hepatitis B, serum hepatitis B virus DNA and liver biochemical levels should be tested. Patients with a serum level > 200,000 IU/mL during the third trimester should be considered for treatment with tenofovir disoproxil fumarate.  

For patients on immunosuppressive therapy for chronic liver diseases or after liver transplantation, therapy should continue at the lowest effective dose. However, mycophenolate mofetil shouldn’t be administered during pregnancy.  

Intrahepatic cholestasis of pregnancy may be diagnosed during the second or third trimester based on pruritus and a serum bile acid level > 10 μmol/L. Treatment should include oral ursodeoxycholic acid, with a total daily dose of 10-15 mg/kg.  

Other pregnancy-related liver diseases — such as pre-eclampsia; hemolysis, elevated liver enzymes, and low platelets syndrome; and acute fatty liver of pregnancy — require careful birth planning and evaluation for possible liver transplantation. For certain high-risk patients, daily aspirin should start at week 12 of gestation.  

In addition, elective endoscopic procedures should wait until after birth, and nonemergent but necessary procedures should be performed during the second trimester. Patients with cirrhosis should undergo evaluation for esophageal varices, and upper endoscopy should happen during the second trimester to guide beta-blocker therapy or endoscopic variceal litigation.  

Endoscopic retrograde cholangiopancreatography can be performed for urgent indications, such as choledocholithiasis, cholangitis, and some gallstone pancreatitis cases, ideally during the second trimester.  

Cholecystectomy is considered safe during pregnancy, with a laparoscopic approach as the standard of care regardless of trimester, though the second trimester is ideal. 

Pregnancy-Related Updates in Practice

Ultimately, clinicians should familiarize themselves with the best practice advice to feel comfortable when counseling and managing pregnancy-related concerns, especially high-risk patients, said Eugenia Shmidt, MD, assistant professor of gastroenterology, hepatology, and nutrition, and founder of the IBD Preconception and Pregnancy Planning Clinic at the University of Minnesota, Minneapolis.

“Half of all patients with GI and liver disease are women, and oftentimes, they don’t have appropriate guidance regarding reproductive health in the context of their disease,” she said. “There exists a very large knowledge gap in this area, particularly because most clinical trials exclude pregnant people.”  

Most importantly, the advice statements can guide practitioners on how to help pregnant patients make informed reproductive decisions, she added.  

“This CPU makes it clear that preconception counseling and multidisciplinary care are key in optimizing reproductive health, regardless of the underlying GI or liver disease,” Shmidt said. “GI practitioners should be counseling women well in advance of pregnancy and recruiting all relevant stakeholders as early as possible, even prior to conception. This way, pregnancy care is not reactive, but instead proactive.”  

The authors received no specific funding for this update. Kothari and Shmidt reported no relevant disclosures.

A version of this article appeared on Medscape.com.

Dear colleagues,

As gastroenterologists and endoscopists, we spend significant time preventing and diagnosing GI malignancies. While colorectal and esophageal cancer and their precursor lesions are well known to us, our approach to rarer malignancies is less well defined.

For instance, is it worthwhile screening for pancreatic cancer, and, if so, how should this be done? Likewise, diagnosing cholangiocarcinoma is challenging; how best should one evaluate for this in higher risk populations, such as primary sclerosing cholangitis? And what about the costs, financial and otherwise, associated with screening?

An Approach to Pancreatic Cancer Screening

BY DANIEL A. BERNSTEIN, MD, AND DARSHAN KOTHARI, MD

Pancreatic cancer carries a dismal prognosis, now accounting for the third-most cancer-related mortality in the United States. A small proportion of patients are diagnosed at a local stage of disease, with over half found to have metastatic disease at presentation. Given the low overall incidence and lifetime risk in the general population, population-based screening is not justified.

About 10% of cases of pancreas cancer are associated with germ-line mutations and/or with a strong family history of pancreatic cancer. Several academic societies and expert committees now recommend regular screening for pancreatic cancer in patients who are considered high-risk individuals, as they carry a fivefold relative risk for pancreatic cancer. Moreover, studies suggest that screening has the potential to identify early-stage resectable disease and decrease mortality in this patient population.

Duke University

Patients who benefit from pancreatic cancer screening are those who carry an increased lifetime risk (in excess of 5%) of pancreatic cancer. High-risk individuals include those with germ-line mutations and/or those with a family history of pancreatic cancer in first-degree relatives. Consensus guidelines by the International Cancer of the Pancreas Screening Consortium and the American Society for Gastrointestinal Endoscopy provide medical centers with detailed recommendations on who and when to start screening.

High-risk individuals fall into three categories:

• Patients with high-risk germline mutations including: familial atypical multiple mole melanoma syndrome (CDKN2A), hereditary breast and ovarian cancer syndromes (BRCA1, BRCA2, and PALB2), Peutz-Jeghers syndrome (STK11), and hereditary pancreatitis (PRSS1 and SPINK1)
• Patients with low- to moderate-risk germ-line mutations with at least one first-degree relative with pancreatic cancer: Lynch Syndrome (particularly MLH1 mutation), ataxia-telangiectasia (ATM), or Li-Fraumeni syndrome (p53)
• Patients with one first-degree relative with pancreatic cancer who in turn has one first-degree relative with pancreatic cancer (eg, a patient’s mother and maternal aunt or a patient’s father and patient’s sister)

Consistent with established guidelines, we recommend screening for high-risk patients beginning at age 50, or 10 years before the youngest age at which pancreas cancer was diagnosed in an affected relative. Screening is recommended earlier in patients with particularly high risk: at age 40 for patients with CDKN2A and STKI11 mutations and age 40 for patients with PRSS1 mutation or 20 years after the first attack of acute pancreatitis. For patients with a strong family history of pancreas cancer, we recommend comprehensive evaluation by a certified genetic counselor at a high-volume cancer center.

Duke University

In practice, patients at our institution who are identified as high risk based on the above criteria are referred for an initial consultation at our pancreas center. In most cases, this should occur no sooner than 5 years prior to the recommended starting age for screening. All patients who are identified as high risk should be screened annually for diabetes given the growing evidence base supporting an association between new-onset diabetes and pancreatic cancer.

After an initial visit and discussion of the risks and benefits of screening, most screening protocols start with a baseline endoscopic ultrasound (EUS) and contrast-enhanced magnetic resonance abdomen with magnetic resonance cholangiopancreatography (MRI/MRCP), which will be repeated annually or sooner as the clinical condition warrants. A sooner-interval EUS should be considered for patients already undergoing screening who are newly found to have diabetes.

At our institution, we start with an in-person clinic evaluation followed by EUS. Thereafter, patients undergo MRI/MRCP (synchronized with a same-day clinic visit) alternating with EUS every 6 months to ensure patients are seen twice a year, though there is no specific data to support this approach. Non-diabetics also undergo yearly diabetes screening which will trigger an EUS if patients become diabetic.

We engage in shared decision-making with our high-risk individuals undergoing pancreatic cancer screening and at each visit we review their concurrent medical conditions and suitability to continue screening. We consider discontinuing screening after age 75, at the onset of any life-limiting illness, or after a discussion of risks and benefits if comorbidities lead to a substantial deterioration in a patient’s overall health status.

While a growing body of evidence exists to support the application of pancreatic cancer screening in high-risk individuals, this preventive service remains underutilized. Recent analysis of the screening cohort at our institution showed a demographically homogeneous group of mostly highly educated, high-income White females. These findings are consistent with the patient cohorts described in other pancreatic cancer screening programs and represent only a fraction of people who would qualify for pancreatic cancer screening.

A survey of patients undergoing screening at our institution identified cost, travel, and time associated with pancreatic cancer screening to be frequent challenges to participation. Further studies are needed to fully explore the barriers and psychological burden of pancreas cancer screening in high-risk individuals, and to identify ways to enrich the cohort of patients undergoing screening. This may involve novel methods to identify family members of patients with a new diagnosis of pancreas cancer and increasing health literacy around pancreatic cancer screening among patients and providers.

Pancreatic cancer screening has the potential to identify early-stage disease in patients who are at high risk because of germ-line mutations and/or family history. We recommend that patients engage in pancreatic cancer screening at high-volume centers with well-supported oncology, genetics, and research infrastructure.

Dr. Bernstein is a gastroenterology fellow at Duke University School of Medicine, Durham, North Carolina. Dr. Kothari is an associate professor of medicine in gastroenterology and hepatology at Duke University School of Medicine.

Screening for Cholangiocarcinoma

BY JUDAH KUPFERMAN, MD, AND APARNA GOEL, MD

Cholangiocarcinoma is a rare but aggressive cancer of the bile ducts that poses many diagnostic challenges. Approximately 3% of gastrointestinal cancers are attributed to cholangiocarcinoma, and while the annual incidence of disease in the United States is about 1.26 per 100,000 people, the incidence of intrahepatic disease has been rising considerably.^1,2 Screening for cholangiocarcinoma is reserved for high-risk individuals — such as those with primary sclerosing cholangitis (PSC), secondary sclerosing cholangitis (SSC), and biliary tract disorders such as choledochal cysts or Caroli’s disease. The goal is to balance the benefits of early diagnosis with the costs and risks associated with screening, particularly given the limitations of available tools like MRI with cholangiopancreatography (MRCP), which has a sensitivity of 70%-85%. In general, we recommend annual cholangiocarcinoma screening for high-risk individuals with MRI and MRCP as well as with cancer antigen (CA) 19-9. .

Stanford School of Medicine

Screening in Patients with Primary Sclerosing Cholangitis

The lifetime risk of cholangiocarcinoma in patients with PSC is 10%-15% with an annual risk of 0.5%-1.5%. In our experience, this is often the most feared complication for PSC patients, even more so than the risk of liver transplantation. We recommend annual MRI with MRCP in addition to CA 19-9 for patients with PSC in the first decade of their diagnosis, as most cancers are diagnosed during this period. If a patient’s imaging has remained stable for over a decade and there is minimal hepatic fibrosis, we discuss the option of reducing screening frequency to every 2 years to minimize costs and exposure to MRI contrast risks.

If MRI reveals a concerning new large duct stricture, we will evaluate this with an endoscopic retrograde cholangiopancreatography (ERCP), as differentiating benign and malignant strictures is quite challenging with MRI. We generally recommend ERCP with brush cytology and fluorescence in situ hybridization to improve diagnostic yield. Depending on imaging findings and location of the new large duct stricture, we may consider cholangioscopy during ERCP for direct visualization of the bile duct and directed tissue biopsies. Unfortunately, even in young, asymptomatic patients who undergo regular screening, cholangiocarcinoma is frequently diagnosed at an advanced stage.
 

Screening in Patients with Secondary Sclerosing Cholangitis

Patients with SSC may develop cholangiocarcinoma because of chronic inflammatory and fibrotic processes, such as IgG4-associated cholangiopathy, sarcoidosis, ischemic cholangiopathy, cystic fibrosis, recurrent pyogenic cholangitis, severe sepsis (as recently seen from SARS-CoV-2), surgical complications, or other etiologies. When the condition is reversible, such as with IgG4-associated cholangiopathy, cancer screening may not be necessary. However, when irreversible damage occurs, the cancer risk increases, though it varies by disease type and severity. In most cases, we recommend routine screening for cholangiocarcinoma with MRI and CA 19-9 in this population.

Stanford School of Medicine

Screening in Patients with Biliary Tract Disorders

Biliary tract disorders such as choledochal cysts and Caroli’s disease also harbor an increased risk of cholangiocarcinoma. Choledochal cysts are congenital cystic dilations of the bile duct that have a 10%-30% lifetime risk of malignant transformation to cholangiocarcinoma. Surgical intervention to remove the cyst is often recommended because of this high risk. However, some patients may be unable or unwilling to undergo this surgery or they may have residual cysts. We recommend ongoing screening with MRI and CA 19-9 for these patients. Similarly, Caroli’s disease is a congenital disease associated with intrahepatic and extrahepatic bile duct cysts and associated with a 5%-15% lifetime risk of cholangiocarcinoma. MRI with MRCP and CA 19-9 should be performed routinely for patients with Caroli’s disease and syndrome.

Risks and Challenges in Cholangiocarcinoma Screening

While MRI with MRCP is the gold standard for cholangiocarcinoma screening, its limitations must be carefully considered. One growing concern is the potential for gadolinium retention in the brain, bones, or skin following repeated MRI scans. Though the long-term effects of gadolinium retention are not fully understood, we factor this into screening decisions, particularly for younger patients who may undergo decades of regular imaging.

MRI is not always feasible for certain patients, including those with metal implants, on hemodialysis, or with severe allergic reactions. In such cases, CT or ultrasound may serve as alternatives, though with lower sensitivity for detecting cholangiocarcinoma. Additionally, claustrophobia during MRI can be addressed with sedation, but this underscores the importance of shared decision-making.

From our perspective, cholangiocarcinoma screening in high-risk patients is crucial but not without challenges. Our current screening methods, while essential, are far from perfect, often missing early cancers or leading to unnecessary interventions. Because of these limitations, the window for treatment of localized disease can easily be missed. In our practice, we tailor screening strategies to each patient’s specific needs, weighing the potential benefits against the risks, costs, and the inherent uncertainty of early detection tools. We believe it is essential to involve patients in this decision-making process to provide a balanced, individualized approach that considers both clinical evidence and the personal preferences of each person.

Dr. Kupferman is a gastroenterology fellow at Stanford University School of Medicine in California. Dr. Goel is a transplant hepatologist and a clinical associate professor in gastroenterology & hepatology at Stanford.

References

1. Vithayathil M and Khan SA. J Hepatol. 2022 Dec. doi: 10.1016/j.jhep.2022.07.022.

2. Patel N and Benipal B. Cureus. 2019 Jan. doi: 10.7759/cureus.3962.

Dear colleagues,

As gastroenterologists and endoscopists, we spend significant time preventing and diagnosing GI malignancies. While colorectal and esophageal cancer and their precursor lesions are well known to us, our approach to rarer malignancies is less well defined.

For instance, is it worthwhile screening for pancreatic cancer, and, if so, how should this be done? Likewise, diagnosing cholangiocarcinoma is challenging; how best should one evaluate for this in higher risk populations, such as primary sclerosing cholangitis? And what about the costs, financial and otherwise, associated with screening?

An Approach to Pancreatic Cancer Screening

BY DANIEL A. BERNSTEIN, MD, AND DARSHAN KOTHARI, MD

Pancreatic cancer carries a dismal prognosis, now accounting for the third-most cancer-related mortality in the United States. A small proportion of patients are diagnosed at a local stage of disease, with over half found to have metastatic disease at presentation. Given the low overall incidence and lifetime risk in the general population, population-based screening is not justified.

About 10% of cases of pancreas cancer are associated with germ-line mutations and/or with a strong family history of pancreatic cancer. Several academic societies and expert committees now recommend regular screening for pancreatic cancer in patients who are considered high-risk individuals, as they carry a fivefold relative risk for pancreatic cancer. Moreover, studies suggest that screening has the potential to identify early-stage resectable disease and decrease mortality in this patient population.

Duke University

Patients who benefit from pancreatic cancer screening are those who carry an increased lifetime risk (in excess of 5%) of pancreatic cancer. High-risk individuals include those with germ-line mutations and/or those with a family history of pancreatic cancer in first-degree relatives. Consensus guidelines by the International Cancer of the Pancreas Screening Consortium and the American Society for Gastrointestinal Endoscopy provide medical centers with detailed recommendations on who and when to start screening.

High-risk individuals fall into three categories:

• Patients with high-risk germline mutations including: familial atypical multiple mole melanoma syndrome (CDKN2A), hereditary breast and ovarian cancer syndromes (BRCA1, BRCA2, and PALB2), Peutz-Jeghers syndrome (STK11), and hereditary pancreatitis (PRSS1 and SPINK1)
• Patients with low- to moderate-risk germ-line mutations with at least one first-degree relative with pancreatic cancer: Lynch Syndrome (particularly MLH1 mutation), ataxia-telangiectasia (ATM), or Li-Fraumeni syndrome (p53)
• Patients with one first-degree relative with pancreatic cancer who in turn has one first-degree relative with pancreatic cancer (eg, a patient’s mother and maternal aunt or a patient’s father and patient’s sister)

Consistent with established guidelines, we recommend screening for high-risk patients beginning at age 50, or 10 years before the youngest age at which pancreas cancer was diagnosed in an affected relative. Screening is recommended earlier in patients with particularly high risk: at age 40 for patients with CDKN2A and STKI11 mutations and age 40 for patients with PRSS1 mutation or 20 years after the first attack of acute pancreatitis. For patients with a strong family history of pancreas cancer, we recommend comprehensive evaluation by a certified genetic counselor at a high-volume cancer center.

Duke University

In practice, patients at our institution who are identified as high risk based on the above criteria are referred for an initial consultation at our pancreas center. In most cases, this should occur no sooner than 5 years prior to the recommended starting age for screening. All patients who are identified as high risk should be screened annually for diabetes given the growing evidence base supporting an association between new-onset diabetes and pancreatic cancer.

After an initial visit and discussion of the risks and benefits of screening, most screening protocols start with a baseline endoscopic ultrasound (EUS) and contrast-enhanced magnetic resonance abdomen with magnetic resonance cholangiopancreatography (MRI/MRCP), which will be repeated annually or sooner as the clinical condition warrants. A sooner-interval EUS should be considered for patients already undergoing screening who are newly found to have diabetes.

At our institution, we start with an in-person clinic evaluation followed by EUS. Thereafter, patients undergo MRI/MRCP (synchronized with a same-day clinic visit) alternating with EUS every 6 months to ensure patients are seen twice a year, though there is no specific data to support this approach. Non-diabetics also undergo yearly diabetes screening which will trigger an EUS if patients become diabetic.

We engage in shared decision-making with our high-risk individuals undergoing pancreatic cancer screening and at each visit we review their concurrent medical conditions and suitability to continue screening. We consider discontinuing screening after age 75, at the onset of any life-limiting illness, or after a discussion of risks and benefits if comorbidities lead to a substantial deterioration in a patient’s overall health status.

While a growing body of evidence exists to support the application of pancreatic cancer screening in high-risk individuals, this preventive service remains underutilized. Recent analysis of the screening cohort at our institution showed a demographically homogeneous group of mostly highly educated, high-income White females. These findings are consistent with the patient cohorts described in other pancreatic cancer screening programs and represent only a fraction of people who would qualify for pancreatic cancer screening.

A survey of patients undergoing screening at our institution identified cost, travel, and time associated with pancreatic cancer screening to be frequent challenges to participation. Further studies are needed to fully explore the barriers and psychological burden of pancreas cancer screening in high-risk individuals, and to identify ways to enrich the cohort of patients undergoing screening. This may involve novel methods to identify family members of patients with a new diagnosis of pancreas cancer and increasing health literacy around pancreatic cancer screening among patients and providers.

Pancreatic cancer screening has the potential to identify early-stage disease in patients who are at high risk because of germ-line mutations and/or family history. We recommend that patients engage in pancreatic cancer screening at high-volume centers with well-supported oncology, genetics, and research infrastructure.

Dr. Bernstein is a gastroenterology fellow at Duke University School of Medicine, Durham, North Carolina. Dr. Kothari is an associate professor of medicine in gastroenterology and hepatology at Duke University School of Medicine.

Screening for Cholangiocarcinoma

BY JUDAH KUPFERMAN, MD, AND APARNA GOEL, MD

Cholangiocarcinoma is a rare but aggressive cancer of the bile ducts that poses many diagnostic challenges. Approximately 3% of gastrointestinal cancers are attributed to cholangiocarcinoma, and while the annual incidence of disease in the United States is about 1.26 per 100,000 people, the incidence of intrahepatic disease has been rising considerably.^1,2 Screening for cholangiocarcinoma is reserved for high-risk individuals — such as those with primary sclerosing cholangitis (PSC), secondary sclerosing cholangitis (SSC), and biliary tract disorders such as choledochal cysts or Caroli’s disease. The goal is to balance the benefits of early diagnosis with the costs and risks associated with screening, particularly given the limitations of available tools like MRI with cholangiopancreatography (MRCP), which has a sensitivity of 70%-85%. In general, we recommend annual cholangiocarcinoma screening for high-risk individuals with MRI and MRCP as well as with cancer antigen (CA) 19-9. .

Stanford School of Medicine

Screening in Patients with Primary Sclerosing Cholangitis

The lifetime risk of cholangiocarcinoma in patients with PSC is 10%-15% with an annual risk of 0.5%-1.5%. In our experience, this is often the most feared complication for PSC patients, even more so than the risk of liver transplantation. We recommend annual MRI with MRCP in addition to CA 19-9 for patients with PSC in the first decade of their diagnosis, as most cancers are diagnosed during this period. If a patient’s imaging has remained stable for over a decade and there is minimal hepatic fibrosis, we discuss the option of reducing screening frequency to every 2 years to minimize costs and exposure to MRI contrast risks.

If MRI reveals a concerning new large duct stricture, we will evaluate this with an endoscopic retrograde cholangiopancreatography (ERCP), as differentiating benign and malignant strictures is quite challenging with MRI. We generally recommend ERCP with brush cytology and fluorescence in situ hybridization to improve diagnostic yield. Depending on imaging findings and location of the new large duct stricture, we may consider cholangioscopy during ERCP for direct visualization of the bile duct and directed tissue biopsies. Unfortunately, even in young, asymptomatic patients who undergo regular screening, cholangiocarcinoma is frequently diagnosed at an advanced stage.
 

Screening in Patients with Secondary Sclerosing Cholangitis

Patients with SSC may develop cholangiocarcinoma because of chronic inflammatory and fibrotic processes, such as IgG4-associated cholangiopathy, sarcoidosis, ischemic cholangiopathy, cystic fibrosis, recurrent pyogenic cholangitis, severe sepsis (as recently seen from SARS-CoV-2), surgical complications, or other etiologies. When the condition is reversible, such as with IgG4-associated cholangiopathy, cancer screening may not be necessary. However, when irreversible damage occurs, the cancer risk increases, though it varies by disease type and severity. In most cases, we recommend routine screening for cholangiocarcinoma with MRI and CA 19-9 in this population.

Stanford School of Medicine

Screening in Patients with Biliary Tract Disorders

Biliary tract disorders such as choledochal cysts and Caroli’s disease also harbor an increased risk of cholangiocarcinoma. Choledochal cysts are congenital cystic dilations of the bile duct that have a 10%-30% lifetime risk of malignant transformation to cholangiocarcinoma. Surgical intervention to remove the cyst is often recommended because of this high risk. However, some patients may be unable or unwilling to undergo this surgery or they may have residual cysts. We recommend ongoing screening with MRI and CA 19-9 for these patients. Similarly, Caroli’s disease is a congenital disease associated with intrahepatic and extrahepatic bile duct cysts and associated with a 5%-15% lifetime risk of cholangiocarcinoma. MRI with MRCP and CA 19-9 should be performed routinely for patients with Caroli’s disease and syndrome.

Risks and Challenges in Cholangiocarcinoma Screening

While MRI with MRCP is the gold standard for cholangiocarcinoma screening, its limitations must be carefully considered. One growing concern is the potential for gadolinium retention in the brain, bones, or skin following repeated MRI scans. Though the long-term effects of gadolinium retention are not fully understood, we factor this into screening decisions, particularly for younger patients who may undergo decades of regular imaging.

MRI is not always feasible for certain patients, including those with metal implants, on hemodialysis, or with severe allergic reactions. In such cases, CT or ultrasound may serve as alternatives, though with lower sensitivity for detecting cholangiocarcinoma. Additionally, claustrophobia during MRI can be addressed with sedation, but this underscores the importance of shared decision-making.

From our perspective, cholangiocarcinoma screening in high-risk patients is crucial but not without challenges. Our current screening methods, while essential, are far from perfect, often missing early cancers or leading to unnecessary interventions. Because of these limitations, the window for treatment of localized disease can easily be missed. In our practice, we tailor screening strategies to each patient’s specific needs, weighing the potential benefits against the risks, costs, and the inherent uncertainty of early detection tools. We believe it is essential to involve patients in this decision-making process to provide a balanced, individualized approach that considers both clinical evidence and the personal preferences of each person.

Dr. Kupferman is a gastroenterology fellow at Stanford University School of Medicine in California. Dr. Goel is a transplant hepatologist and a clinical associate professor in gastroenterology & hepatology at Stanford.

References

1. Vithayathil M and Khan SA. J Hepatol. 2022 Dec. doi: 10.1016/j.jhep.2022.07.022.

2. Patel N and Benipal B. Cureus. 2019 Jan. doi: 10.7759/cureus.3962.

Dear colleagues,

As gastroenterologists and endoscopists, we spend significant time preventing and diagnosing GI malignancies. While colorectal and esophageal cancer and their precursor lesions are well known to us, our approach to rarer malignancies is less well defined.

For instance, is it worthwhile screening for pancreatic cancer, and, if so, how should this be done? Likewise, diagnosing cholangiocarcinoma is challenging; how best should one evaluate for this in higher risk populations, such as primary sclerosing cholangitis? And what about the costs, financial and otherwise, associated with screening?

An Approach to Pancreatic Cancer Screening

BY DANIEL A. BERNSTEIN, MD, AND DARSHAN KOTHARI, MD

Pancreatic cancer carries a dismal prognosis, now accounting for the third-most cancer-related mortality in the United States. A small proportion of patients are diagnosed at a local stage of disease, with over half found to have metastatic disease at presentation. Given the low overall incidence and lifetime risk in the general population, population-based screening is not justified.

About 10% of cases of pancreas cancer are associated with germ-line mutations and/or with a strong family history of pancreatic cancer. Several academic societies and expert committees now recommend regular screening for pancreatic cancer in patients who are considered high-risk individuals, as they carry a fivefold relative risk for pancreatic cancer. Moreover, studies suggest that screening has the potential to identify early-stage resectable disease and decrease mortality in this patient population.

Duke University

Patients who benefit from pancreatic cancer screening are those who carry an increased lifetime risk (in excess of 5%) of pancreatic cancer. High-risk individuals include those with germ-line mutations and/or those with a family history of pancreatic cancer in first-degree relatives. Consensus guidelines by the International Cancer of the Pancreas Screening Consortium and the American Society for Gastrointestinal Endoscopy provide medical centers with detailed recommendations on who and when to start screening.

High-risk individuals fall into three categories:

• Patients with high-risk germline mutations including: familial atypical multiple mole melanoma syndrome (CDKN2A), hereditary breast and ovarian cancer syndromes (BRCA1, BRCA2, and PALB2), Peutz-Jeghers syndrome (STK11), and hereditary pancreatitis (PRSS1 and SPINK1)
• Patients with low- to moderate-risk germ-line mutations with at least one first-degree relative with pancreatic cancer: Lynch Syndrome (particularly MLH1 mutation), ataxia-telangiectasia (ATM), or Li-Fraumeni syndrome (p53)
• Patients with one first-degree relative with pancreatic cancer who in turn has one first-degree relative with pancreatic cancer (eg, a patient’s mother and maternal aunt or a patient’s father and patient’s sister)

Consistent with established guidelines, we recommend screening for high-risk patients beginning at age 50, or 10 years before the youngest age at which pancreas cancer was diagnosed in an affected relative. Screening is recommended earlier in patients with particularly high risk: at age 40 for patients with CDKN2A and STKI11 mutations and age 40 for patients with PRSS1 mutation or 20 years after the first attack of acute pancreatitis. For patients with a strong family history of pancreas cancer, we recommend comprehensive evaluation by a certified genetic counselor at a high-volume cancer center.

Duke University

In practice, patients at our institution who are identified as high risk based on the above criteria are referred for an initial consultation at our pancreas center. In most cases, this should occur no sooner than 5 years prior to the recommended starting age for screening. All patients who are identified as high risk should be screened annually for diabetes given the growing evidence base supporting an association between new-onset diabetes and pancreatic cancer.

After an initial visit and discussion of the risks and benefits of screening, most screening protocols start with a baseline endoscopic ultrasound (EUS) and contrast-enhanced magnetic resonance abdomen with magnetic resonance cholangiopancreatography (MRI/MRCP), which will be repeated annually or sooner as the clinical condition warrants. A sooner-interval EUS should be considered for patients already undergoing screening who are newly found to have diabetes.

At our institution, we start with an in-person clinic evaluation followed by EUS. Thereafter, patients undergo MRI/MRCP (synchronized with a same-day clinic visit) alternating with EUS every 6 months to ensure patients are seen twice a year, though there is no specific data to support this approach. Non-diabetics also undergo yearly diabetes screening which will trigger an EUS if patients become diabetic.

We engage in shared decision-making with our high-risk individuals undergoing pancreatic cancer screening and at each visit we review their concurrent medical conditions and suitability to continue screening. We consider discontinuing screening after age 75, at the onset of any life-limiting illness, or after a discussion of risks and benefits if comorbidities lead to a substantial deterioration in a patient’s overall health status.

While a growing body of evidence exists to support the application of pancreatic cancer screening in high-risk individuals, this preventive service remains underutilized. Recent analysis of the screening cohort at our institution showed a demographically homogeneous group of mostly highly educated, high-income White females. These findings are consistent with the patient cohorts described in other pancreatic cancer screening programs and represent only a fraction of people who would qualify for pancreatic cancer screening.

A survey of patients undergoing screening at our institution identified cost, travel, and time associated with pancreatic cancer screening to be frequent challenges to participation. Further studies are needed to fully explore the barriers and psychological burden of pancreas cancer screening in high-risk individuals, and to identify ways to enrich the cohort of patients undergoing screening. This may involve novel methods to identify family members of patients with a new diagnosis of pancreas cancer and increasing health literacy around pancreatic cancer screening among patients and providers.

Pancreatic cancer screening has the potential to identify early-stage disease in patients who are at high risk because of germ-line mutations and/or family history. We recommend that patients engage in pancreatic cancer screening at high-volume centers with well-supported oncology, genetics, and research infrastructure.

Dr. Bernstein is a gastroenterology fellow at Duke University School of Medicine, Durham, North Carolina. Dr. Kothari is an associate professor of medicine in gastroenterology and hepatology at Duke University School of Medicine.

Screening for Cholangiocarcinoma

BY JUDAH KUPFERMAN, MD, AND APARNA GOEL, MD

Cholangiocarcinoma is a rare but aggressive cancer of the bile ducts that poses many diagnostic challenges. Approximately 3% of gastrointestinal cancers are attributed to cholangiocarcinoma, and while the annual incidence of disease in the United States is about 1.26 per 100,000 people, the incidence of intrahepatic disease has been rising considerably.^1,2 Screening for cholangiocarcinoma is reserved for high-risk individuals — such as those with primary sclerosing cholangitis (PSC), secondary sclerosing cholangitis (SSC), and biliary tract disorders such as choledochal cysts or Caroli’s disease. The goal is to balance the benefits of early diagnosis with the costs and risks associated with screening, particularly given the limitations of available tools like MRI with cholangiopancreatography (MRCP), which has a sensitivity of 70%-85%. In general, we recommend annual cholangiocarcinoma screening for high-risk individuals with MRI and MRCP as well as with cancer antigen (CA) 19-9. .

Stanford School of Medicine

Screening in Patients with Primary Sclerosing Cholangitis

The lifetime risk of cholangiocarcinoma in patients with PSC is 10%-15% with an annual risk of 0.5%-1.5%. In our experience, this is often the most feared complication for PSC patients, even more so than the risk of liver transplantation. We recommend annual MRI with MRCP in addition to CA 19-9 for patients with PSC in the first decade of their diagnosis, as most cancers are diagnosed during this period. If a patient’s imaging has remained stable for over a decade and there is minimal hepatic fibrosis, we discuss the option of reducing screening frequency to every 2 years to minimize costs and exposure to MRI contrast risks.

If MRI reveals a concerning new large duct stricture, we will evaluate this with an endoscopic retrograde cholangiopancreatography (ERCP), as differentiating benign and malignant strictures is quite challenging with MRI. We generally recommend ERCP with brush cytology and fluorescence in situ hybridization to improve diagnostic yield. Depending on imaging findings and location of the new large duct stricture, we may consider cholangioscopy during ERCP for direct visualization of the bile duct and directed tissue biopsies. Unfortunately, even in young, asymptomatic patients who undergo regular screening, cholangiocarcinoma is frequently diagnosed at an advanced stage.
 

Screening in Patients with Secondary Sclerosing Cholangitis

Patients with SSC may develop cholangiocarcinoma because of chronic inflammatory and fibrotic processes, such as IgG4-associated cholangiopathy, sarcoidosis, ischemic cholangiopathy, cystic fibrosis, recurrent pyogenic cholangitis, severe sepsis (as recently seen from SARS-CoV-2), surgical complications, or other etiologies. When the condition is reversible, such as with IgG4-associated cholangiopathy, cancer screening may not be necessary. However, when irreversible damage occurs, the cancer risk increases, though it varies by disease type and severity. In most cases, we recommend routine screening for cholangiocarcinoma with MRI and CA 19-9 in this population.

Stanford School of Medicine

Screening in Patients with Biliary Tract Disorders

Biliary tract disorders such as choledochal cysts and Caroli’s disease also harbor an increased risk of cholangiocarcinoma. Choledochal cysts are congenital cystic dilations of the bile duct that have a 10%-30% lifetime risk of malignant transformation to cholangiocarcinoma. Surgical intervention to remove the cyst is often recommended because of this high risk. However, some patients may be unable or unwilling to undergo this surgery or they may have residual cysts. We recommend ongoing screening with MRI and CA 19-9 for these patients. Similarly, Caroli’s disease is a congenital disease associated with intrahepatic and extrahepatic bile duct cysts and associated with a 5%-15% lifetime risk of cholangiocarcinoma. MRI with MRCP and CA 19-9 should be performed routinely for patients with Caroli’s disease and syndrome.

Risks and Challenges in Cholangiocarcinoma Screening

While MRI with MRCP is the gold standard for cholangiocarcinoma screening, its limitations must be carefully considered. One growing concern is the potential for gadolinium retention in the brain, bones, or skin following repeated MRI scans. Though the long-term effects of gadolinium retention are not fully understood, we factor this into screening decisions, particularly for younger patients who may undergo decades of regular imaging.

MRI is not always feasible for certain patients, including those with metal implants, on hemodialysis, or with severe allergic reactions. In such cases, CT or ultrasound may serve as alternatives, though with lower sensitivity for detecting cholangiocarcinoma. Additionally, claustrophobia during MRI can be addressed with sedation, but this underscores the importance of shared decision-making.

From our perspective, cholangiocarcinoma screening in high-risk patients is crucial but not without challenges. Our current screening methods, while essential, are far from perfect, often missing early cancers or leading to unnecessary interventions. Because of these limitations, the window for treatment of localized disease can easily be missed. In our practice, we tailor screening strategies to each patient’s specific needs, weighing the potential benefits against the risks, costs, and the inherent uncertainty of early detection tools. We believe it is essential to involve patients in this decision-making process to provide a balanced, individualized approach that considers both clinical evidence and the personal preferences of each person.

Dr. Kupferman is a gastroenterology fellow at Stanford University School of Medicine in California. Dr. Goel is a transplant hepatologist and a clinical associate professor in gastroenterology & hepatology at Stanford.

References

1. Vithayathil M and Khan SA. J Hepatol. 2022 Dec. doi: 10.1016/j.jhep.2022.07.022.

2. Patel N and Benipal B. Cureus. 2019 Jan. doi: 10.7759/cureus.3962.

PHILADELPHIA — Researchers have developed an artificial intelligence (AI) tool capable of detecting and differentiating cystic and solid pancreatic lesions during endoscopic ultrasound (EUS) with high accuracy.

This was a transatlantic collaborative effort involving researchers in Portugal, Spain, the United States, and Brazil, and the AI tool “works on different platforms and different devices,” Miguel Mascarenhas, MD, PhD, with Centro Hospitalar Universitário de São João, Porto, Portugal, said in a presentation at the annual meeting of the American College of Gastroenterology.

Mascarenhas noted that pancreatic cystic lesions (PCLs) are a common incidental finding during imaging and are differentiated by whether they’re mucinous PCLs (M-PCLs) or non-mucinous PCLs (NM-PCLs). The malignancy risk is almost exclusive of PCL with a mucinous phenotype.

Pancreatic solid lesions are also prevalent, and differentiation is challenging. Pancreatic ductal adenocarcinoma (P-DAC) is the most common pancreatic solid lesion and has a poor prognosis because of late-stage disease at diagnosis. Pancreatic neuroendocrine tumors (P-NETs) are less common but have malignant potential.

EUS is the “gold standard” for pancreatic lesion evaluation, but its diagnostic accuracy is suboptimal, particularly for lesions < 10 mm, Mascarenhas noted.

With an eye toward improving diagnostic accuracy, he and colleagues developed a convolutional neural network for detecting and differentiating cystic (M-PCL and NM-PCL) and solid (P-DAC and P-NET) pancreatic lesions.

They leveraged data from 378 EUS exams with 126,000 still images — 19,528 M-PCL, 8175 NM-PCL, 64,286 P-DAC, 29,153 P-NET, and 4858 normal pancreas images.

The AI tool demonstrated 99.1% accuracy for identifying normal pancreatic tissue, and it showed 99% and 99.8% accuracy for M-PCL and NM-PCL, respectively.

For pancreatic solid lesions, P-DAC and P-NET were distinguished with 94% accuracy, with 98.7% and 83.6% sensitivity for P-DAC and P-NET, respectively.
 

Real-Time Validation Next

“AI is delivering promising results throughout medicine, but particularly in gastroenterology, which is one of the most fertile areas of AI research. This comes mostly from the deployment of deep-learning models, most of them convolutional neural networks, which are highly efficient for image analysis,” Mascarenhas told attendees.

This is the “first worldwide convolutional neural network” capable of detecting and differentiating both cystic and solid pancreatic lesions. The use of a large dataset from four centers in two continents helps minimize the impact of demographic bias, Mascarenhas added.

The study is based on still images, not full videos, he noted. As a next step, the team is conducting a multicenter study focused on real-time clinical validation of the model during EUS procedures.

“AI has the potential to improve the diagnostic accuracy of endoscopic ultrasound. We’re just on the tip of the iceberg. There is enormous potential to harness AI, and we welcome all the groups that might want to join our research,” Mascarenhas said.

 

Brennan Spiegel, MD, MSHS, AGAF, director of Health Services Research at Cedars-Sinai Medical Center, Los Angeles, who wasn’t involved in the study, is optimistic about emerging applications for AI.

“AI holds incredible promise in gastroenterology, especially for diagnosing complex pancreatic lesions where early, accurate differentiation can be lifesaving,” Spiegel said in an interview.

“This study’s high accuracy across diverse datasets is encouraging; however, as a retrospective analysis, it leaves the real-time clinical impact still to be proven. Prospective studies will be essential to confirm AI’s role in enhancing our diagnostic capabilities,” Spiegel cautioned.

“More generally, AI is rapidly transforming gastroenterology by enhancing our ability to detect, differentiate, and monitor conditions with unprecedented precision. From improving early cancer detection to guiding complex diagnostic procedures, AI stands to become an invaluable tool that complements clinical expertise. As we refine these technologies, the potential for AI to elevate both diagnostic accuracy and patient outcomes in GI is truly remarkable,” Spiegel said.

The study had no specific funding. Mascarenhas and Spiegel have declared no conflicts of interest.

A version of this article appeared on Medscape.com.

PHILADELPHIA — Researchers have developed an artificial intelligence (AI) tool capable of detecting and differentiating cystic and solid pancreatic lesions during endoscopic ultrasound (EUS) with high accuracy.

This was a transatlantic collaborative effort involving researchers in Portugal, Spain, the United States, and Brazil, and the AI tool “works on different platforms and different devices,” Miguel Mascarenhas, MD, PhD, with Centro Hospitalar Universitário de São João, Porto, Portugal, said in a presentation at the annual meeting of the American College of Gastroenterology.

Mascarenhas noted that pancreatic cystic lesions (PCLs) are a common incidental finding during imaging and are differentiated by whether they’re mucinous PCLs (M-PCLs) or non-mucinous PCLs (NM-PCLs). The malignancy risk is almost exclusive of PCL with a mucinous phenotype.

Pancreatic solid lesions are also prevalent, and differentiation is challenging. Pancreatic ductal adenocarcinoma (P-DAC) is the most common pancreatic solid lesion and has a poor prognosis because of late-stage disease at diagnosis. Pancreatic neuroendocrine tumors (P-NETs) are less common but have malignant potential.

EUS is the “gold standard” for pancreatic lesion evaluation, but its diagnostic accuracy is suboptimal, particularly for lesions < 10 mm, Mascarenhas noted.

With an eye toward improving diagnostic accuracy, he and colleagues developed a convolutional neural network for detecting and differentiating cystic (M-PCL and NM-PCL) and solid (P-DAC and P-NET) pancreatic lesions.

They leveraged data from 378 EUS exams with 126,000 still images — 19,528 M-PCL, 8175 NM-PCL, 64,286 P-DAC, 29,153 P-NET, and 4858 normal pancreas images.

The AI tool demonstrated 99.1% accuracy for identifying normal pancreatic tissue, and it showed 99% and 99.8% accuracy for M-PCL and NM-PCL, respectively.

For pancreatic solid lesions, P-DAC and P-NET were distinguished with 94% accuracy, with 98.7% and 83.6% sensitivity for P-DAC and P-NET, respectively.
 

Real-Time Validation Next

“AI is delivering promising results throughout medicine, but particularly in gastroenterology, which is one of the most fertile areas of AI research. This comes mostly from the deployment of deep-learning models, most of them convolutional neural networks, which are highly efficient for image analysis,” Mascarenhas told attendees.

This is the “first worldwide convolutional neural network” capable of detecting and differentiating both cystic and solid pancreatic lesions. The use of a large dataset from four centers in two continents helps minimize the impact of demographic bias, Mascarenhas added.

The study is based on still images, not full videos, he noted. As a next step, the team is conducting a multicenter study focused on real-time clinical validation of the model during EUS procedures.

“AI has the potential to improve the diagnostic accuracy of endoscopic ultrasound. We’re just on the tip of the iceberg. There is enormous potential to harness AI, and we welcome all the groups that might want to join our research,” Mascarenhas said.

 

Brennan Spiegel, MD, MSHS, AGAF, director of Health Services Research at Cedars-Sinai Medical Center, Los Angeles, who wasn’t involved in the study, is optimistic about emerging applications for AI.

“AI holds incredible promise in gastroenterology, especially for diagnosing complex pancreatic lesions where early, accurate differentiation can be lifesaving,” Spiegel said in an interview.

“This study’s high accuracy across diverse datasets is encouraging; however, as a retrospective analysis, it leaves the real-time clinical impact still to be proven. Prospective studies will be essential to confirm AI’s role in enhancing our diagnostic capabilities,” Spiegel cautioned.

“More generally, AI is rapidly transforming gastroenterology by enhancing our ability to detect, differentiate, and monitor conditions with unprecedented precision. From improving early cancer detection to guiding complex diagnostic procedures, AI stands to become an invaluable tool that complements clinical expertise. As we refine these technologies, the potential for AI to elevate both diagnostic accuracy and patient outcomes in GI is truly remarkable,” Spiegel said.

The study had no specific funding. Mascarenhas and Spiegel have declared no conflicts of interest.

A version of this article appeared on Medscape.com.

PHILADELPHIA — Researchers have developed an artificial intelligence (AI) tool capable of detecting and differentiating cystic and solid pancreatic lesions during endoscopic ultrasound (EUS) with high accuracy.

This was a transatlantic collaborative effort involving researchers in Portugal, Spain, the United States, and Brazil, and the AI tool “works on different platforms and different devices,” Miguel Mascarenhas, MD, PhD, with Centro Hospitalar Universitário de São João, Porto, Portugal, said in a presentation at the annual meeting of the American College of Gastroenterology.

Mascarenhas noted that pancreatic cystic lesions (PCLs) are a common incidental finding during imaging and are differentiated by whether they’re mucinous PCLs (M-PCLs) or non-mucinous PCLs (NM-PCLs). The malignancy risk is almost exclusive of PCL with a mucinous phenotype.

Pancreatic solid lesions are also prevalent, and differentiation is challenging. Pancreatic ductal adenocarcinoma (P-DAC) is the most common pancreatic solid lesion and has a poor prognosis because of late-stage disease at diagnosis. Pancreatic neuroendocrine tumors (P-NETs) are less common but have malignant potential.

EUS is the “gold standard” for pancreatic lesion evaluation, but its diagnostic accuracy is suboptimal, particularly for lesions < 10 mm, Mascarenhas noted.

With an eye toward improving diagnostic accuracy, he and colleagues developed a convolutional neural network for detecting and differentiating cystic (M-PCL and NM-PCL) and solid (P-DAC and P-NET) pancreatic lesions.

They leveraged data from 378 EUS exams with 126,000 still images — 19,528 M-PCL, 8175 NM-PCL, 64,286 P-DAC, 29,153 P-NET, and 4858 normal pancreas images.

The AI tool demonstrated 99.1% accuracy for identifying normal pancreatic tissue, and it showed 99% and 99.8% accuracy for M-PCL and NM-PCL, respectively.

For pancreatic solid lesions, P-DAC and P-NET were distinguished with 94% accuracy, with 98.7% and 83.6% sensitivity for P-DAC and P-NET, respectively.
 

Real-Time Validation Next

“AI is delivering promising results throughout medicine, but particularly in gastroenterology, which is one of the most fertile areas of AI research. This comes mostly from the deployment of deep-learning models, most of them convolutional neural networks, which are highly efficient for image analysis,” Mascarenhas told attendees.

This is the “first worldwide convolutional neural network” capable of detecting and differentiating both cystic and solid pancreatic lesions. The use of a large dataset from four centers in two continents helps minimize the impact of demographic bias, Mascarenhas added.

The study is based on still images, not full videos, he noted. As a next step, the team is conducting a multicenter study focused on real-time clinical validation of the model during EUS procedures.

“AI has the potential to improve the diagnostic accuracy of endoscopic ultrasound. We’re just on the tip of the iceberg. There is enormous potential to harness AI, and we welcome all the groups that might want to join our research,” Mascarenhas said.

 

Brennan Spiegel, MD, MSHS, AGAF, director of Health Services Research at Cedars-Sinai Medical Center, Los Angeles, who wasn’t involved in the study, is optimistic about emerging applications for AI.

“AI holds incredible promise in gastroenterology, especially for diagnosing complex pancreatic lesions where early, accurate differentiation can be lifesaving,” Spiegel said in an interview.

“This study’s high accuracy across diverse datasets is encouraging; however, as a retrospective analysis, it leaves the real-time clinical impact still to be proven. Prospective studies will be essential to confirm AI’s role in enhancing our diagnostic capabilities,” Spiegel cautioned.

“More generally, AI is rapidly transforming gastroenterology by enhancing our ability to detect, differentiate, and monitor conditions with unprecedented precision. From improving early cancer detection to guiding complex diagnostic procedures, AI stands to become an invaluable tool that complements clinical expertise. As we refine these technologies, the potential for AI to elevate both diagnostic accuracy and patient outcomes in GI is truly remarkable,” Spiegel said.

The study had no specific funding. Mascarenhas and Spiegel have declared no conflicts of interest.

A version of this article appeared on Medscape.com.