Experts explore possibility, practical aspects of remission of type 2 diabetes

Roundtable Participants

Defining ‘remission’

Richard E. Pratley, MD: What’s a good definition of diabetes remission?

Roy Taylor, MD: The consensus panel [of which Taylor was a member] concluded that the term “remission” should be used to describe a person with type 2 diabetes who has achieved an HbA1c of less than 48 mmol/mol — 6.5% — measured at least 3 months after cessation of glucose-lowering pharmacotherapy. The panel worked hard to achieve a definition that was simple enough to be understood by anyone with diabetes as well as to allow comparison between studies. The particular points that were most debated were the time needed between measurements of HbA1c to establish that stable remission had been achieved, and also the vital matter of being off all hypoglycemic agents during this time.

Sangeeta R. Kashyap, MD: We also have to consider other definitions of remission when looking at other interventions. For example, in the STAMPEDE trial comparing bariatric surgery with intensive medical therapy for diabetes [of which Kashyap was an investigator], those who were able to maintain HbA1c below the diagnostic threshold for diabetes without the use of any diabetes medications a year after an intervention were considered to be in remission. We also considered relapse as those who were able to achieve that endpoint at a year but then later were not able to.

Ravi Retnakaran, MD, MS, FRCPC: The literature has been complicated by lack of consistency. People were using terms like “remission, “reversal” and “elimination of diabetes” but without standardization. This recent attempt to get consensus is an important step, so we can all at least be speaking the same language. The definition that’s been chosen balances the recognition that the disease state has moved back toward normality, but you have not necessarily cured this condition.

Reversing beta-cell dysfunction

Pratley: It seems we’re really talking about different types of remission — remission in the very early stages of diabetes, maybe within the first 1 or 2 years or so, when interventions might be different than for somebody who has had diabetes for a longer time, like the people in the STAMPEDE study. Ultimately, we’re talking about blood sugar control, in the absence of the need for pharmacologic therapy, but these things might represent also different physiologic states.

Dr. Retnakaran, can you describe the paradigm of early insulin treatment and glucose normalization, how that helps to restore glycemia, and what that means in terms of remission?

Retnakaran: It has been recognized for a long time that you can take people early in the course of diabetes, give them short-term intensive insulin therapy — in the range of 2 to 5 weeks, and not just basal insulin, but basal plus bolus or some kind of post-prandial control — and you can achieve what was termed at the time “remission,” the reflection that glucose levels normalize once you stop therapy.

To put that in context, meta-analysis has suggested that 3 months after you stop the insulin therapy, about 66% of people have been in remission. When you wait 6 months, about 59% of people have still been in remission, and when you wait about a year, 46% of people have been in remission. The glass-half-full is that you treated for only a short period, and a year later just under half the people can still be in remission. The glass-half-empty, though, is you’re losing the effect.

The key effect, at least for the insulin intervention, appears to be the improvement in beta-cell function. The people who are going to achieve remission have an improvement in beta-cell function that allows them to maintain relatively normal glucose levels after you stop the intervention. That’s helpful for understanding the pathophysiologic standpoint.

The clinical challenge, though, is understanding what element of the beta-cell function is improved. This varies with the duration of diabetes and according to the intervention. But certainly, with medical therapy as well as with lifestyle intervention, the earlier the intervention, the better, because earlier in the course of diabetes you are likely dealing with a greater reversible component to the beta-cell dysfunction than later in the course of diabetes. We believe that the intervention is helping to attenuate the impact of reversible beta-cell dysfunction, essentially reversing it. That’s how you achieve remission.

But there are two caveats: None of our conventional means of measuring beta-cell function can distinguish between reversible dysfunction and nonreversible dysfunction. And you lose the effect over time, at least with the insulin intervention. That says you have not necessarily prevented the ongoing deterioration of beta-cell function.

Pratley: Great summary. Let’s take the paradigm of bariatric surgery — slightly different definitions and a much different intervention, and still we have a lot of efficacy at reversal of diabetes. What do we know about the physiology that allows patients to reverse diabetes or go into remission?

Kashyap: Bariatric surgery is meant for patients who are overweight. From a physiological perspective, for several decades, we’ve known the effects of lipotoxicity on beta-cell dysfunction. Free fatty acids cause not only beta-cell dysfunction, but apoptosis of beta cells. With bariatric surgery, lipotoxicity is markedly reduced through weight loss.

Initially, we also saw reversal of hyperglycemia, within the first month of having bariatric surgery. We think it’s related to the influence of various incretin hormones and reduction in free fatty acids. Free fatty acids also improve insulin sensitivity, and they offload the beta cells. There was an immediate improvement in glucose sensitivity of the beta cells. Obviously, these effects were more potent with gastric bypass than the other forms of bariatric surgery, which are a little more restrictive in causing negative caloric balance.

Looking at the long term, patients with shorter duration of diabetes, who had better beta-cell function at baseline, who were not on insulin but were controlled on oral agents, and who had better control of their diabetes had better effects in terms of weight loss. Also, the nadir weight loss at the first year really determines long-term glycemic control. These findings reveal that beta cells are not dead, they might be dormant, but they’re able to be functional again, and that is why patients are able to be independent of insulin for up to 5 years. With gastric bypass, about 33% of patients were in remission at 3 years and about 26% at 5 years.

Taylor: There is no doubt about what is going on. First the Counterpoint study and then the Pancreas and DiRECT studies established this using state-of-the-art methods. Excess fat builds up in the liver, and then the pancreas, driven on by twin vicious cycles that cause type 2 diabetes. As Dr. Kashyap described, the pathophysiologic studies have shown that weight loss causes a rapid fall in the excess liver fat characteristic of type 2 diabetes in each organ. Weight loss leads to restoration of hepatic liver insulin sensitivity and hence normalization of fasting plasma glucose — FPG normalizes within 7 days in short-duration disease.

The decreased liver fat allows hepatic export of fat to decrease to a normal level, resulting in decrease in all ectopic sites, including intra-pancreatic fat. As pointed out by Dr. Retnakaran, this allows return of beta-cell function, re-differentiation of the beta cells and return of specialized function, because the endoplasmic reticulum stress can be reversed. Although there is not yet absolute proof that de-differentiation/re-differentiation is the only process involved, it can certainly explain the pathophysiologic observations.

Optimizing interventions

Pratley: So, we’re looking at two ends of the spectrum: very early diabetes where an intervention to decrease glucose levels reduces glucose toxicity and restores insulin secretion; and then probably at a much more advanced stage of diabetes, where a bigger intervention, bariatric surgery, has multiple effects, reducing insulin resistance and lipotoxicity and producing changes in the gut hormones that are useful for improving glycemic control.

Is there some middle ground where we can take patients with early diabetes and then perhaps do intensive glycemic control, but combine that with a weight-loss intervention?

Taylor: In pathophysiologic studies of short-duration type 2 diabetes — up to 4 years — we have shown that stopping oral hypoglycemic agents and commencing an 800 kcal per day diet achieves normal fasting glucose within 7 days.

Pratley: In our clinic, we’ve been playing with a combination of very intensive initial therapy for patients newly diagnosed with type 2 diabetes. We use a variety of techniques — insulin, a GLP-1 receptor agonist, sometimes in combination — combined in parallel with a lifestyle intervention focused on weight loss and increasing physical activity. It’s often the case that people are weaned off their medications after a month or 3 months. But my sense is that success has to do with the lifestyle intervention, making changes in intake, losing a little bit of weight, becoming more physically active. I’ve certainly had patients who have been able to maintain normal glycemia for a long time, but they tend to be the ones who are most successful with lifestyle changes. That would make sense also from what Dr. Kashyap said: The people who’ve had that initial weight loss are the ones who are most likely to benefit from diabetes remission with bariatric surgery.

Kashyap: Also, I think younger patients who are heavier would benefit the most from intensive metabolic control, that is, the combination of lifestyle aimed at weight loss and negative caloric balance, as well as pharmacotherapy to intensify glucose control. We really need to focus our efforts on the younger, heavier patients. Older patients have loss of beta cells, and we have less of an impact on those patients. But those patients are people we don’t want to ignore, and we want to consider surgical interventions for those who have more advanced disease.

Pratley: We have younger patients with type 2 diabetes, and the evolving literature suggests that they are at particularly high risk for developing complications of diabetes, not just because of their duration of diabetes, but in part because they have some gene set, perhaps, or combination of genes and lifestyle that has allowed them to express diabetes at this early age.

At the other end of the spectrum, people who are older who develop some hyperglycemia may have a different sort of pathophysiology without necessarily that same risk for complications of diabetes. This speaks to some of the diabetes subtypes that are now being discussed. We have people who are obese and very insulin resistant, who are at increased risk for cardiovascular and chronic kidney disease. We have people who are more insulin deficient, and older patient population who has milder diabetes.

What do we know about personalizing these interventions?

Retnakaran: Our findings are similar: Someone who is younger and heavier — likely earlier in the course of the natural history of the beta cell — is posing a test for the beta cell by virtue of the resultant insulin resistance. The beta cell is failing, resulting in diabetes. The flip side is that because they are so early in the course of diabetes and posing a resistance challenge, they probably have more reversible beta cell dysfunction, which can benefit from the intervention.

We’ve spent a lot of time looking for predictors of response to the intervention. Despite our best efforts around BMI, waist circumference and every other measure of adiposity, we have not been able to clearly demonstrate this. Our best clinical predictor, at least for the medical interventions, remains duration of diabetes.

Click to enlarge

The other helpful predictor is glycemic response to the initial glucose-lowering intervention even in clinical care, which is usually metformin. We all have some ideas about the kind of glucose lowering we expect with metformin, but sometimes people who are early in the course of diabetes have a huge drop in HbA1c. From our experience, it looks like amplified glucose lowering because of the presence of reversible beta cell dysfunction. Hence, when we apply our glucose-lowering therapy, whether it be metformin or anything else, we’re not only getting the effect of the medication, but the beta-cell function improves because, we think, they had reversibility.

Benefits of diabetes remission

Pratley: What does reversal of diabetes means for the patient, for the natural history, for their risk for complications? Are we just looking at biochemical changes, or is there some tangible benefit that clinicians and patients could realize through this reversal?

Kashyap: We just published a paper on quality of life. A variety of quality of life metrics were improved after surgery. Some were related to being independent of medications; psychologically, that’s a huge boost. It’s so motivating for patients to know they can reverse their disease and that their efforts make a big difference in long-term outcomes. We tell them that these types of interventions, we hope, will add a few more years not only to life, but to quality of life.

Pratley: Because we know that bariatric surgery is associated with lower mortality, lower cardiovascular and cancer rates over the long term. The Swedish Obese Subjects study showed that as well as other studies. Has anybody ever looked at whether or not the remission of diabetes is one of those determinants of improved outcomes?

Retnakaran: We don’t have the evidence right now to say that remission itself is associated with long-term benefits in terms of mortality and complications. However, when thinking in terms of what we know about diabetes — the impact of cumulative glycemic exposure, plus or minus the other metabolic disturbances of diabetes, and the impact that has on risks for complications of diabetes — it’s reasonable right now to anticipate that the induction of remission for whatever period should ultimately be beneficial. We recognize that this will be hard to demonstrate in studies, so the absence of evidence right now of long-term benefit does not argue against our current efforts to pursue this type of therapy.

Pratley: We do need more evidence for the long-term utility of this sort of approach because that’s ultimately what changes our treatment paradigms and guidelines. For the moment, treatment guidelines recommend patients with diabetes and CVD go on a GLP-1 agonist or SGLT2 inhibitor, otherwise metformin followed by a variety of treatments. They say nothing about what you should do if people are able to maintain normal glycemia. I think that’s based on the lack of evidence and also a concern about perhaps therapeutic inertia, that if people improve their glycemia, they might get lost to follow-up and then not realize the long-term benefits.

Is remission something clinicians should talk about with their patients?

Retnakaran: This is a chronic progressive disorder. At the time of diagnosis, early in the course of diabetes, is a point when many of the interventions are going to have their greatest long-term benefit for changing the natural history. The opportunity for this type of intervention is in primary care. The medical community has accepted the idea that if you have a myocardial infarction, then time to revascularization is heart muscle; if you have a stroke, time is brain. I would argue that when you’re diagnosed with diabetes, time is beta-cell function.

Kashyap: Bariatric surgery is like CPR for a failing heart. We see patients who are medically refractory for a variety of reasons. For those patients to see diabetes remission over years is a terrific vacation from a horrible disease. We’ve seen complications occur overnight, especially with retinopathy, macular edema, worsening of renal function, albuminuria. It behooves us as clinicians to use our interventions and the data we have to empower our patients to strive for remission. Even if they’re partially in remission, if they’re able to improve their metabolic status, and they’re able to not require insulin.

Pratley: So, the timeline for early diagnosis is a critical period. We learned from the U.K. Prospective Diabetes Study (UKPDS) and patients with newly diagnosed type 2 diabetes that more intensive glycemic control in this early phase was associated with long-term benefits in terms of microvascular complications, as well as potentially macrovascular disease. It makes sense then that by reversing hyperglycemia and inducing this remission, we’re prolonging that phase of early glycemic control, metabolic compensation, that should yield long-term benefits.

For my own part, the ADA recommendations, if you think about it, advocate personalizing therapy. They advocate achieving HbA1c targets of 6.5% or less, if somebody can do that safely without hyperglycemia and without excessive medication burden. We know we can do that with diabetes remission. The recommendations also refer to this sort of continuous quality improvement cycle where you’re reassessing patients adjusting therapies. That’s the perfect paradigm for dealing with patients early on in the course of the disease. Make lifestyle changes maybe with intensive initial therapy, then assess and back off if their blood sugars normalize and they don’t have other indications for being on medications, such as CVD or chronic kidney disease that would perhaps benefit from a trial off medications. You’re going to reassess them in 3 months or 6 months and, if necessary, readjust their therapy. That’s not inconsistent with our current diabetes recommendations.

Remission in practice

Pratley: What are your closing comments about diabetes remission for clinicians who see patients with diabetes every day?

Taylor: Type 2 diabetes is a terrible burden for anyone, with risk to eyesight, to limb and to life — especially in younger patients. Because weight loss decreases excess hepatic lipoprotein production as the root cause of both reversal of type 2 diabetes and dramatic halving of 10-year CV risk, it is the ideal response to diagnosis of type 2 diabetes. All patients should receive an explanation that type 2 diabetes is usually able to be reversed to normal by substantial weight loss. The pros and cons can be described. Only those wanting to attempt remission should be encouraged to firstly discuss this with family and then decide whether they wish to achieve this.

Kashyap: Look at the characteristics of the diabetes. What is the phenotype of diabetes based on the patient’s age, demographics, etc? Individualize care but be aggressive from the onset. Start intensive metabolic control when someone has metabolic syndrome. Let’s not wait until they have diabetes. We are underutilizing these interventions. With the number of overweight patients in the U.S., only less than 2% are on anti-obesity medications, which are FDA approved for weight loss, and less than 2% of eligible patients are receiving bariatric surgery. Let’s give attention to weight and the harmful effects of obesity on diabetes.

Retnakaran: Consider the benefits of remission or attempting it as early as possible in the course of diabetes. First, the natural history of declining beta-cell function is starting long before the diagnosis of diabetes. Prediabetes is when you should be thinking about this, and even before prediabetes for individuals in whom you could identify such risk. For instance, women with gestational diabetes will be a population with declining beta-cell function over time. Second, until we know that we have stopped the deterioration of beta-cell function over time, we always have to think of the individual with remission as being at risk for relapse. We still need to be thinking about the possibility of complications emerging over time.

• References:
• Aminian A, et al. Ann Surg. 2021;doi:10.1097/SLA.0000000000005003.
• Kramer CK, et al. Lancet Diabetes Endocrinol. 2013;doi:10.1016/S2213-8587(13)70006-8.
• Leslie WS, et al. Diabetologia. 2021;doi:10.1007/s00125-021-05471.
• Lim EL, et al. Diabetologia. 2011;doi:10.1007/s00125-011-2204-7.
• Riddle MC, et al. Diabetologia. 2021;doi:10.1007/s00125-021-05542-z.
• Schauer PR, et al. N Engl J Med. 2017;doi:10.1056/NEJMoa1600869.
• Taylor R, et al. Cell Metab. 2018;doi:10.1016/j.cmet.2018.07.003.

• For more information:
• Sangeeta R. Kashyap, MD, can be reached at kashyas@ccf.org.
• Richard E. Pratley, MD, can be reached at richard.pratley.md@AdventHealth.com. Twitter: @RpratleyMD.
• Ravi Retnakaran, MD, MSc, FRCPC, can be reached at ravi.retnakaran@sinaihealthsystem.ca.
• Roy Taylor, MD, FRCP, can be reached at roy.taylor@newcastle.ac.uk.