LIVERPOOL — C-peptide testing a few years after a diagnosis of type 1 diabetes should be a routine procedure as it can help pinpoint occasional patients who have been misdiagnosed and actually have type 2 diabetes or a monogenic form of the disease, new research shows.
Researchers led by Mark Strachan, MD, from the Western General Hospital, Edinburgh, UK, used the C-peptide test in all patients in their clinic who had type 1 diabetes for about 3 years and showed that a small number had initially been wrongly diagnosed.
"So far, of the 757 patients with type 1 diabetes of at least 3 years' duration, 103 (13.7%) have a C-peptide level above 200 pmol/L," said Strachan, who reported the findings at the Diabetes UK Professional Conference last week.
"We have made a new diagnosis of monogenic or genetic diabetes in eight people and changed the diagnosis to type 2 diabetes in 27 other people. This has allowed us to make changes to treatment in many of these individuals. In 12 people to date we have actually been able to stop insulin therapy," he said in an interview with Medscape Medical News. Other people are awaiting a final diagnosis.
"Our diagnostic criteria for type 1 (and type 2) diabetes are imperfect. I firmly believe that anyone who starts on insulin...should have a C-peptide test [a few years after diagnosis] and this should become the norm," stressed Strachan.
C-peptide testing is a means of determining patients' endogenous insulin levels and was once routinely used in type 1 diabetes diagnoses until it was realized that there is often a 'honeymoon period' during which patients still have residual beta-cell activity, making the test outcome a little unreliable.
Asked to comment, Mark McCarthy, MB, BChir, MD, University of Oxford, UK, said: "The bottom line is that at the time of diagnosis of diabetes C-peptide isn't so useful because there's still residual endogenous insulin secretion in type 1 diabetes for a few years. That's when autoantibodies are more useful since most of those with type 1 diabetes will have autoantibody positivity to go along with the clinical picture."
But "Scroll forward a few years, and consider someone who has had diabetes for, say, 5 years. If they have type 1 diabetes, you would expect them to have lost all of their beta cells by then so their C-peptide should be undetectable.
"If you can detect it still at reasonable levels several years out, they almost certainly don't have type 1 but type 2 diabetes [or another form]."
"The main use here is as a way of going back to those with diabetes and doing a better job of classifying their major diabetes subtype," McCarthy added. It's not "entirely clear that will have a big impact on care...but this is a useful tool to contribute to personalized medicine in diabetes," he stressed.
A More Convenient, Cheaper Assay for C-Peptide
The C-peptide test has been considered fiddly, requiring special tubes and special care of the sample, but more convenient assays have been introduced, including the one used in the Edinburgh study (ARCHITECT immunoassay), which was developed by Timothy MacDonald, PhD, Exeter University, UK.
"If you transfer the blood on EDTA plasma it lasts for 24 hours. All samples that might be taken for HbA1c can [also] actually be used to measure C-peptide [now]," explained Andrew Hattersley, MD, CBE, also from the University of Exeter, who was presenting in Liverpool on behalf of MacDonald.
"The sample can be taken randomly in the nonfasting state. Data has shown the only thing that matters is that the patient shouldn't be hypoglycemic, nor do patients need to be taken off [their] insulin [treatment] because we have tested that too and it made minimal difference to the C-peptide measurement."
Strachan said the advantage of the improved blood test is that it is practical, "and it is done there and then in the clinic." Blood glucose was tested at the same time to ensure it was greater than 4 mmol/L (ie, the patient was not hypoglycemic).
Of those tested in their study, Strachan classified the findings according to the C-peptide level patients were found to be producing.
Fourteen of the more than 700 patients were found to produce over 900 pmol/L, and 13 of these patients were reclassified as having type 2 diabetes, with seven of these no longer needing insulin.
Nineteen patients had C-peptide between 600 to 900 pmol/L. Of these, nine had evidence of autoimmunity with a duration of diabetes between 6 to 13 years and the diagnosis remained as type 1 diabetes. Five were reclassified as having type 2 diabetes, one was diagnosed as monogenic diabetes, and one was diagnosed as mitochondrial diabetes.
Using C-Peptide: New Understanding of Type 1 and Type 2 Diabetes
Seventy patients had C-peptide levels between 200 to 600 pmol/L, and of these, 24 were negative for antibodies. Of this latter group, nine were reclassified as having type 2 diabetes and four were found to have different genetic types of diabetes.
McCarthy noted that "autoantibody testing is less helpful" a number of years into a diagnosis of type 1 diabetes, "since those levels fall off after a few years."
If a patient is 10 years into type 1 diabetes "and still has a good C-peptide level, then they don't have type 1 diabetes," he stressed.
Strachan said they also showed in their study that some people who presented with diabetic ketoacidosis actually had type 2 diabetes.
"Traditionally, diabetic ketoacidosis is seen in type 1 diabetes. We've identified four people who presented with clear-cut ketoacidosis treated with insulin who actually have type 2 diabetes. These patients have all come off insulin now."
"[Using] C-peptide has given us a new understanding of type 1 and type 2 diabetes," explained Strachan.
Substantial Savings if New Diagnosis Is Made and Insulin Stopped
Strachan also pointed out that the C-peptide test is very inexpensive, costing between £6 to £10 per test.
"Although there are some additional costs in terms of genetic and antibody tests, when you look at it alongside...associated savings, it is an extremely cost-effective intervention."
He noted that a triple antibody test in Scotland costs around £26, a monogenic screen is £114, and an Exeter Type 1 diabetes genetic risk score is £60.
The savings become apparent when a patient can come off insulin, he noted.
"We know insulin therapy costs around £800 per year, while insulin pumptherapy is around £2700 per year, so getting even one person off insulin can result in a substantial cost saving," he emphasized.
Strachan has submitted a business case to NHS Scotland to roll out C-peptide testing on a national basis, which has been widely supported by stakeholders for type 1 diabetes there. If funding is approved, the test could become routine — in Scotland at least.
How to Test for Diabetes in an Ideal World?
McCarthy said, in an ideal world, "What you'd...do is a combination of genetic risk score, a C-peptide, and autoantibody tests," at diagnosis.
"If the patient has autoantibodies (and a positive type 1 diabetes genetic risk score, if available) but still has normal glucose levels, C-peptide testing can be used to monitor how close the patient is to needing insulin. If it's low, you can tell the patient they'll probably need insulin sooner or later. It's probably best to get them started on it sooner, rather than later, to avoid diabetic ketoacidosis."
"The bottom line is C-peptide testing is clinically useful, but additional testing may be needed in new-onset patients," he emphasized.
"But if at 5 to 6 years a patient diagnosed with type 1 diabetes still has positive C-peptide, you have to question the diagnosis."
Meanwhile, Colin Dayan, MD, of Cardiff University, told Medscape Medical News that the C-peptide test could also be useful for GPs monitoring their patients with diabetes.
"Every day GPs are seeing patients with a high BMI at around 30 years of age with diabetes type unknown.
"This is further complicated with ethnic minority patients who get diabetes at a younger age and at a lower BMI. The overlap is a real smudge and it is important to clarify this. C-peptide is a good way of reclassifying people several years in. It is cheap and...reliable," he concluded.
McCarthy has served as a director, officer, partner, employee, advisor, consultant, or trustee for Merck, Pfizer, Eli Lilly, and Novo Nordisk. He has received research grants from Pfizer, Merck, Takeda, Servier, Sanofi Aventis, Roche, AstraZeneca, Novo Nordisk, Eli Lilly, Janssen, and Boehringer Ingelheim and an income in an amount equal to or greater than $250 from Merck, Pfizer, and Lilly.
Diabetes UK Professional Conference (DUPC) 2019. Presented March 7, 2019. Abstract 8 (P97)
Follow Lisa Nainggolan on Twitter: @lisanainggolan1. For more diabetes and endocrinology news, follow us on Twitter and on Facebook.
Medscape Medical News © 2019 WebMD, LLC