Biomarker identifies precancerous pancreatic cysts

 

LAS VEGAS – In fluid derived from pancreatic cysts, methylated DNA markers predict the presence of high-grade dysplasia (HGD) or cancer, and could help physicians decide whether to surgically remove cysts – a procedure that often has serious complications.

If validated in larger studies, the biomarkers have the potential to supplant the Fukuoka criteria that is currently used. “The markers could cause a paradigm shift in how we approach these lesions in our clinical practice,” Shounak Majumder, MD, a fellow at the Mayo Clinic in Rochester, Minn., said in an interview.

Courtesy Dr. Lance Liotta Laboratory

Dr. Majumder presented the results at the annual meeting of the American College of Gastroenterology.

Less than 50% of cysts that are surgically resected turn out to be HGD or cancerous. “Having a cyst fluid marker could identify the patients that would benefit the most from surgery. If you’re going to go through a pancreatic resection, we’d rather give you the best chance of saying that we removed something that either has early cancer in it or will turn into cancer in the near future,” said Dr. Majumder.

The study looked at pancreatic cyst fluid from 83 cysts that had been surgically resected. The DNA samples were taken from the cyst fluid. Dr. Majumder believes that the cells shed from the cyst wall into the fluid. As a result, DNA from the fluid captures heterogeneity in the cyst more effectively than a biopsied sample.

The researchers performed methylation-specific PCR of the samples, normalizing the levels to beta-actin and age. Fourteen of the cysts were adenocarcinoma or HGD, and these were compared to cysts that had low-grade dysplasia or no dysplasia.

The researchers found five methylated DNA markers that distinguished cancer or HGD from controls with areas under the ROC curve of 0.90 or higher. The top two (BMP3, EMX1) detected 93% of cases (95% CI, 66%-100%) at a specificity of 90% (95% CI, 80%-96%). Applied to eight cysts with intermediate-grade dysplasia, the biomarkers would have identified three at 95% specificity.

By comparison, the Fukuoka guidelines have 56% sensitivity and 73% specificity.

A limitation to the technique is that DNA cannot be extracted from all samples. About 5%-10% of pancreatic fluid samples are unusable, according to Somashekar Krishna, MD, MPH, assistant professor of medicine at the Ohio State University Medical Center, who attended the session. Dr. Krishna is conducting research combining endomicroscopy with molecular markers.

“We should have a foolproof system where if one fails, the other kicks in, and we have an answer for every patient. My opinion is that endomicroscopy has to be combined with molecular studies. I think combined we’ll have an excellent diagnostic yield,” Dr. Krishna said in an interview.

Dr. Majumder and Dr. Krishna have declared no conflicts of interest.

 

 

LAS VEGAS – In fluid derived from pancreatic cysts, methylated DNA markers predict the presence of high-grade dysplasia (HGD) or cancer, and could help physicians decide whether to surgically remove cysts – a procedure that often has serious complications.

If validated in larger studies, the biomarkers have the potential to supplant the Fukuoka criteria that is currently used. “The markers could cause a paradigm shift in how we approach these lesions in our clinical practice,” Shounak Majumder, MD, a fellow at the Mayo Clinic in Rochester, Minn., said in an interview.

Courtesy Dr. Lance Liotta Laboratory

Dr. Majumder presented the results at the annual meeting of the American College of Gastroenterology.

Less than 50% of cysts that are surgically resected turn out to be HGD or cancerous. “Having a cyst fluid marker could identify the patients that would benefit the most from surgery. If you’re going to go through a pancreatic resection, we’d rather give you the best chance of saying that we removed something that either has early cancer in it or will turn into cancer in the near future,” said Dr. Majumder.

The study looked at pancreatic cyst fluid from 83 cysts that had been surgically resected. The DNA samples were taken from the cyst fluid. Dr. Majumder believes that the cells shed from the cyst wall into the fluid. As a result, DNA from the fluid captures heterogeneity in the cyst more effectively than a biopsied sample.

The researchers performed methylation-specific PCR of the samples, normalizing the levels to beta-actin and age. Fourteen of the cysts were adenocarcinoma or HGD, and these were compared to cysts that had low-grade dysplasia or no dysplasia.

The researchers found five methylated DNA markers that distinguished cancer or HGD from controls with areas under the ROC curve of 0.90 or higher. The top two (BMP3, EMX1) detected 93% of cases (95% CI, 66%-100%) at a specificity of 90% (95% CI, 80%-96%). Applied to eight cysts with intermediate-grade dysplasia, the biomarkers would have identified three at 95% specificity.

By comparison, the Fukuoka guidelines have 56% sensitivity and 73% specificity.

A limitation to the technique is that DNA cannot be extracted from all samples. About 5%-10% of pancreatic fluid samples are unusable, according to Somashekar Krishna, MD, MPH, assistant professor of medicine at the Ohio State University Medical Center, who attended the session. Dr. Krishna is conducting research combining endomicroscopy with molecular markers.

“We should have a foolproof system where if one fails, the other kicks in, and we have an answer for every patient. My opinion is that endomicroscopy has to be combined with molecular studies. I think combined we’ll have an excellent diagnostic yield,” Dr. Krishna said in an interview.

Dr. Majumder and Dr. Krishna have declared no conflicts of interest.

 

 

LAS VEGAS – In fluid derived from pancreatic cysts, methylated DNA markers predict the presence of high-grade dysplasia (HGD) or cancer, and could help physicians decide whether to surgically remove cysts – a procedure that often has serious complications.

If validated in larger studies, the biomarkers have the potential to supplant the Fukuoka criteria that is currently used. “The markers could cause a paradigm shift in how we approach these lesions in our clinical practice,” Shounak Majumder, MD, a fellow at the Mayo Clinic in Rochester, Minn., said in an interview.

Courtesy Dr. Lance Liotta Laboratory

Dr. Majumder presented the results at the annual meeting of the American College of Gastroenterology.

Less than 50% of cysts that are surgically resected turn out to be HGD or cancerous. “Having a cyst fluid marker could identify the patients that would benefit the most from surgery. If you’re going to go through a pancreatic resection, we’d rather give you the best chance of saying that we removed something that either has early cancer in it or will turn into cancer in the near future,” said Dr. Majumder.

The study looked at pancreatic cyst fluid from 83 cysts that had been surgically resected. The DNA samples were taken from the cyst fluid. Dr. Majumder believes that the cells shed from the cyst wall into the fluid. As a result, DNA from the fluid captures heterogeneity in the cyst more effectively than a biopsied sample.

The researchers performed methylation-specific PCR of the samples, normalizing the levels to beta-actin and age. Fourteen of the cysts were adenocarcinoma or HGD, and these were compared to cysts that had low-grade dysplasia or no dysplasia.

The researchers found five methylated DNA markers that distinguished cancer or HGD from controls with areas under the ROC curve of 0.90 or higher. The top two (BMP3, EMX1) detected 93% of cases (95% CI, 66%-100%) at a specificity of 90% (95% CI, 80%-96%). Applied to eight cysts with intermediate-grade dysplasia, the biomarkers would have identified three at 95% specificity.

By comparison, the Fukuoka guidelines have 56% sensitivity and 73% specificity.

A limitation to the technique is that DNA cannot be extracted from all samples. About 5%-10% of pancreatic fluid samples are unusable, according to Somashekar Krishna, MD, MPH, assistant professor of medicine at the Ohio State University Medical Center, who attended the session. Dr. Krishna is conducting research combining endomicroscopy with molecular markers.

“We should have a foolproof system where if one fails, the other kicks in, and we have an answer for every patient. My opinion is that endomicroscopy has to be combined with molecular studies. I think combined we’ll have an excellent diagnostic yield,” Dr. Krishna said in an interview.

Dr. Majumder and Dr. Krishna have declared no conflicts of interest.