Patients with type 2 diabetes whose condition deteriorates rapidly soon after diagnosis may have asymptomatic pancreatic cancer, say European investigators.
In a study of more than 550,000 diabetes patients, researchers found that patients who received glucagonlike peptide-1 (GLP-1) receptor agonists, or incretin mimetics, were at significantly increased risk of developing pancreatic cancer.
However, the researchers observed that the increased risk diminished rapidly after diagnosis. Given that they also found that the risk for pancreatic cancer was markedly increased after starting insulin therapy, they suggested that "reverse causation" may be in play, with asymptomatic pancreatic cancer initially causing diabetes before progressing to a symptomatic stage.
The new findings were presented here at the inaugural meeting of the European CanCer Organisation (ECCO) Congress 2017.
Alice Koechlin, from the International Prevention Research Institute (IPRI) in Lyon, France, said in a statement: "Doctors and their diabetic patients should be aware that the onset of diabetes or rapidly deteriorating diabetes could be the first sign of hidden pancreatic cancer, and steps should be taken to investigate it."
She cautioned that investigating the possibility of undiagnosed pancreatic cancer is "difficult," as there is currently no good, noninvasive method for detecting asymptomatic pancreatic cancer.
"We hope that our results will encourage the search for blood markers indicating the presence of pancreatic cancer, which could guide decisions to perform a confirmation examination like endoscopy," she said.
Peter Naredi, MD, PhD, chair of the Congress, president of ECCO, and professor of surgery at the Sahlgrenska Academy, University of Gothenburg, Sweden, also emphasized the need for better ways to detect pancreatic cancer early on, because so few patients are diagnosed at a curable stage.
This study "opens up the possibility to combine the diagnosis of an associated disease, type 2 diabetes, with blood biomarkers," he said, adding: "It is a step in the right direction if we can increase the proportion of early-diagnosed pancreatic cancers."
Complex Interaction With Diabetes
The study, which was led by Philippe Autier, MD, also from the IPRI, was undertaken to investigate the complex interaction between type 2 diabetes and pancreatic cancer and to study the safety of GLP-1 receptor agonists in patients with diabetes.
These drugs act as incretin mimetics and reduce diabetic hyperglycemia via modifying the release of insulin by the pancreas. It has been posited that such therapies could promote the development of pancreatic cancer.
The European Medicines Agency therefore requested a postauthorization safety study of GLP-1 receptor agonists. For this, the IPRI collaborated with the Agence InterMutualiste and the Belgian Cancer Register, which has access to complete social security data for 11 million people in Belgium, and the University of Milano, Bicocca, which has access to data on 10 million people in Lombardy, Italy
The researchers collated information on all patients with diabetes who had a first prescription of a noninsulin, nonincretin antiabetic drug (NIAD), such as metformin, or an incretin mimetic from 2008 onward and followed them until the end of 2013 in Belgium and the end of 2012 in Lombardy.
The Belgian cohort included 368,654 patients, of whom 22,982 were receiving incretin therapy; the Lombardy cohort comprised 190,371 individuals, of whom 10,310 were treated with incretins.
The mean age of the patients who received incretin mimetics was between 56 years and 58 years; the mean age for those taking NIADs was between 61 years and 65 years. The vast majority of patients were not receiving insulin therapy.
Multivariate analysis indicated that the risk for pancreatic cancer was significantly increased among patients receiving incretin therapy compared with those given NIADs, at an adjusted hazard ratio of 2.12 for the Belgian cohort, 2.17 for the Lombardy cohort, and 2.14 overall.
When the researchers analyzed the relationship between pancreatic cancer and incretin use by duration of exposure, they found, however, that the risk of developing cancer decreased over time, from a peak 2.3-fold increased risk at 3 to 6 months, a 2.0-fold increased risk at 12 months, and a 1.7-fold increased risk after the first year.
They examined the risk for pancreatic cancer associated with insulin prescription during follow-up, which they regarded as an indication of worsening diabetes. The adjusted hazard ratio for developing pancreatic cancer among patients prescribed insulin vs those not given insulin was 6.61 in the Belgian cohort, 7.46 in the Lombardy cohort, and 6.89 overall.
Following the reverse causation hypothesis, the team therefore suggested that occult pancreatic cancer may cause disturbances in pancreatic function, which lead to glucose metabolism disorders.
These, in turn, may lead to the development of type 2 diabetes, which then rapidly progresses to the point at which insulin therapy is necessary. It is at this stage that the pancreatic cancer often becomes symptomatic and, thus, detectable.
This theory, Koechlin said, is supported by observations from previous studies, such as a recent meta-analysis of 88 studies that demonstrated a decrease in the risk for pancreatic adenoma carcinoma with increasing duration of diabetes.
Koechlin concluded: "The last question is: Should these patients diagnosed with diabetes be screened for pancreatic cancer?
"For now, no such test exists, but we could imagine in the future something like this: If the patient is diagnosed with diabetes and his/her diabetes degrades rapidly, there could be...some tests to identify markers for pancreatic cancer. Then, the pancreatic cancer could be diagnosed at an earlier stage, and with more chance of curing the disease," she added.
During the discussion that followed the presentation, session co-chair Franco Roviello, MD, professor of surgical oncology, University of Siena and Instituto Toscano Tumori, Siena, Italy, described the study as "very, very interesting."
He wondered, given the high hazard ratios, whether it would be beneficial for patients to undergo CT scanning at the time of diagnosis to determine whether they had pancreatic cancer.
Koechlin said she believes that it is important to "identify a sequence of events that would lead to higher risk and then give CT scans," adding: "I don't think we can prescribe this to all newly diagnosed patients."
New tests for pancreatic cancer are under development, including a test based on metaolomics, as recently reported by Medscape Medical News.
This study was part of the postauthorization safety study (PASS) of GLP-1 requested by the European Medicines Agency. The protocol was developed through interaction between iPRI and the Committee for Medicinal Products for Human Use of the European Medicines Agency. The PASS was funded by Sanofi. The authors have disclosed no relevant financial relationships.
European CanCer Organisation (ECCO) Congress 2017. Abstract 540, presented January 30, 2017.