Diabetes therapy-focus on Asia: second-line therapy debate: insulin/secretagogues


  • Roy Eldor,

    Corresponding author
    1. Division of Diabetes and The Endocrinology Service, Department of Medicine, Hadassah- Hebrew University Medical Center, POB 12000, Jerusalem 91120, Israel
    2. Division of Diabetes, Department of Medicine, University of Texas, Health Science Center, San Antonio, USA.
    Search for more papers by this author
  • Itamar Raz

    1. Division of Diabetes and The Endocrinology Service, Department of Medicine, Hadassah- Hebrew University Medical Center, POB 12000, Jerusalem 91120, Israel
    Search for more papers by this author

*Correspondence to:
Roy Eldor, Division of Diabetes and The Endocrinology Service, Department of Medicine, Hadassah- Hebrew University Medical Center, POB 12000, Jerusalem 91120, Israel
E-mail: eldorroy@yahoo.com


The following review is based on the second part of a debate entitled ‘Diabetes therapy- focus on Asia: 2nd line therapy: GLP1/DPP4 inhibitors versus Secretagogue/insulin therapy’, which was held during the ‘1st Asia Pacific Congress on Controversies to Consensus in Diabetes, Obesity and Hypertension (CODHy)’, in Shanghai, China, 2011. As such we reviewed only insulin and secretagogue therapy despite the existence of other therapeutic options. The article aims to shed light on the circumstances most adequate for use of these as second-line agents, despite possible drawbacks. It is important to emphasize that regardless of it being a review of published evidence, it primarily represents the professional opinion of the writers.


The progressive nature of Type 2 Diabetes (T2DM) mandates necessary additions of antihyperglycaemic drugs to maintain normoglycaemia. In a recently published study of 1 799 T2DM patients followed for 2–5 years in a clinical practice setting, 42% needed the addition or substitution of metformin to another antihyperglycaemic agent (a metformin mono-therapy failure rate of 17% per year) [1]. After failure to adequately control glycaemia, of the initial antihyperglycaemic agent, how do we choose the second drug to be administered? To answer this question appropriately, we should contemplate the characteristics we seek in the next antihyperglycaemic drug we prescribe.

The long practiced gluco-centric paradigm is becoming less adequate. Merely reducing glycaemia, although disregarding possible harm is unacceptable. Large trials have questioned the validity and necessity of a drug regimen that is only focused on tight glycaemic control (ADOPT, VA-DT, ACCORD) [2–5]. Others have raised doubts regarding the safety and durability of antihyperglycaemic drugs [6–11]. Moreover, T2DM is a multifactorial, multi-organ disease. Our drugs should address the mechanism of the disease and attempt to treat the underlying pathogenetic processes rather than solely reduce blood glucose[12].

To facilitate this process, we have published a method of evaluating and comparing the therapeutic options available [13]. We ask seven questions regarding a drug’s attributes and give a score based on the best obtainable scientific evidence. This said, due to lack of published evidence; some of the answers given are only partially proven or based on assumption, clinical judgment and experience. All the answers are liable to change in accord with new evidence. The aim of this exercise is not to present a strict scoring system for antihyperglycaemic drugs, but to provoke what we believe is the correct intellectual process leading to an adequate, patient-centred choice of the most appropriate option.

The questions we pose are the following (Table 1):

Table 1.   Drug characteristics score
Possible score range
  1. PO, orally administered; SC, subcutaneous injection.

Hyperglycaemia control268 
Major side effects−2−101
CV disease02
Prolonged experience01
Glycaemic durability01
Method of deliveryPO = 1SC = −1
  • 1What is the relative antihyperglycaemic strength of the drug?
  • 2Does the drug cause weight gain or weight loss? Because T2DM is causally associated with obesity, and many patients with T2DM suffer from obesity, we regard the antihyperglycaemic drug’s effect on weight of utmost importance in our considerations for treatment [14–18]. This may be imperative in the Asian population where CV and diabetes increase significantly at lower BMI levels [19]. Furthermore, obesity and weight gain in general are associated with many, non vascular related illnesses (such as orthopaedic morbidities) that directly affect the patients’ quality of life and possibly longevity.
  • 3Does the drug have significant and common side effects (including a significantly increased risk of hypoglycaemia)?
  • 4Does the drug protect from cardiovascular disease? Does it predispose to cardiovascular disease? [20–23]
  • 5Do we have prolonged experience using the drug?
  • 6Is the drug beta cell protective, i.e. will prescribing it result in long-lasting glucose reductions (glycaemic durability)? [24,25]
  • 7What is the method of delivery of the drug, oral or injection? [26,27]

Considering these criteria, we will attempt to provide our view (based on published data and our clinical experience) regarding insulin and the secretagogues (with a focus on studies performed in Asia) and attribute an overall score to these in context of the available armamentarium of antihyperglycaemic drugs. We will then contemplate on the appropriate reasons and circumstances for choosing secretagogues or insulin as second-line therapy in patients with T2DM.

Secretagogues (sulphonylureas, glinides)

Until recently, the sulphonylureas have been considered the ‘classic’ first or second-line therapy in diabetes. This was largely attributed to their potent and rapid hypoglycaemic effect [25]. Nevertheless, this effect is short-lived, with eventual increase in the HbA1c to pre-treatment levels after a limited period of time [25]. Similarly, their short acting counterparts, the glinides (nateglinide) have been shown to have no effect on the rate of deterioration from the state of impaired glucose tolerance to diabetes[28]. As the deterioration of glycaemic control from IGT to T2DM and after disease onset is largely attributed to beta cell failure [24], this characteristic of the sulphonylurea group implies a possible negative effect on long-term beta cell function and survival. Indeed, several in vitro studies have suggested a pro-apoptotic effect of sulphonylureas and glinides on beta cell survival [29–32].

Sulphonylurea treatment is associated with significant weight gain and hypoglycaemic events [33]. No study to date has clearly implicated sulphonylureas or glinides as having a protective or negative cardiovascular effect [34,35]. Nevertheless, doubts have arisen regarding a specific inhibitory effect of ‘older’ sulphonylureas (glybenclamide) on the cardio-protective process of ischemic preconditioning [36]. This was accompanied by several studies showing a possible association between sulphonylurea use and poor cardiovascular outcomes [37–39]. On the contrary, many other studies did not find any association, especially with the new and more pancreatic selective agents (such as gliclazide) [4,40,41]. Recently, a 5 year course of nateglinide treatment has been shown to have no effect on cardiovascular outcomes in persons with impaired glucose tolerance, previous cardiovascular disease or cardiovascular risk factors [28].

Among the different secretagogues, the shorter acting glinides may offer better glycaemic control with reduced hypoglycaemia, when compared to the longer acting sulphonylureas. In a trial including 88 elderly diabetic patients, a 51% reduction in risk of hypoglycaemia, accompanied by a greater reduction in HbA1c (−0.77 versus−0.36) was observed with repaglinide versus glibenclamide [42]. However, taken together, the secretagogue group receives a relatively low score when compared to other modes of treatment (Table 2). Still, its acute antihyperglycaemic effect cannot be overlooked, and may be beneficial in appropriate circumstances.

Table 2.   Comparison of different therapeutic options using the drug characteristics score
 LifestyleMetforminGLP-1 AgonistsSulphonylureaGlinidesBasal insulin analogueShort acting insulin analogue
Hyperglycaemia control6888888
Major Side effects11−1−2−100
CV disease2200000
Prolonged experience2101111
Glycaemic Durability (beta cell survival)1010011
Method of delivery11−111−1−1


Insulin is by far the most potent antihyperglycaemic agent available, achieving up to threefold reduction in HbA1c when compared to all other antihyperglycaemics[43]. This can be partly attributed to insulin being a ‘dose-less’ drug; the actual dose is gradually increased until the patient achieves adequate glycaemic control [44]. Moreover, insulin is often the only drug that will treat glycaemia when others have failed (due to progressive beta cell failure) [45].

Insulin therapy is usually associated with some weight gain and hypoglycaemic events [33,46,47]. Part of this may be attributed to better glycaemic control and a reduction in urinary caloric losses as a result of glucosuria [48]. Furthermore, insulin’s effect on weight differs between the available analogues, with the least weight gain observed with the new long-acting basal analogues (insulin glargine and detemir), as compared to short acting insulin or NPH-insulin [48–52].

Unlike secretagogues, insulin does not lose its efficacy with time, both due to its more versatile dose and a suggested beta cell protective effect. Several studies in newly diagnosed type 2 diabetes patients have suggested a beta cell protective effect to insulin therapy. In a Swedish multicentre randomized clinical trial, patients with recently diagnosed type 2 diabetes were treated with two daily injections of premixed 30% soluble and 70% NPH-insulin or glibenclamide. After 1 year and despite a similar HbA1c, glucagon-stimulated C-peptide response was significantly increased in the insulin-treated group. After 2 years, the HbA1c significantly increased in the glibenclamide group, but did not change in the insulin group [53].

Similarly and even more striking results were shown in two trials conducted in Asia. In a study done in Taiwan, newly diagnosed T2DM patients with severe hyperglycaemia were hospitalized and received intensive insulin therapy for 2 weeks. After release from the hospital, the patients were randomized to receive standard oral antihyperglycaemic therapy with glibenclamide and/or metformin or two injections of NPH-insulin/day for 6 months. After 6 months insulin was stopped and all patients received oral antihyperglycaemic therapy. In the insulin-treated group, the HbA1c was significantly lower, both at the end of 6 months and after 1 year. Moreover, beta cell function significantly improved [54].

The second study was a multicentre, randomized trial conducted in China [55]. In the trial, uncontrolled patients with T2DM were randomly assigned to continuous or dosed insulin therapy (using insulin pumps or multiple daily injections) or oral antihyperglycaemic agents to achieve rapid normoglycaemia. When normoglycaemia was maintained for 2 weeks, all therapy was stopped and the patients were consulted to adhere to a healthy lifestyle. Patients were followed for the length of their treatment free period thereafter. Initially, more patients in the insulin arms achieved normoglycaemia, and in a shorter time. Among those who could enter the trial and stop all medications, previously insulin-treated patients were significantly more likely to be drug free and adequately controlled at 1 year of follow-up. This was accompanied by improved beta cell function as assessed by Homa B and acute insulin response.

An important aspect of insulin therapy is its mode of administration. The initial stages of treatment demand greater effort on behalf of the patient and his diabetes team. Patients should be educated in self injecting, dose titration and management of side effects and complications. More than other therapeutic options, initiating insulin therapy demands a significant level of involvement and availability of the treating physicians, nurses, dieticians and diabetes educators [56,57].

When comparing insulin to available therapeutic options, it is evident that the basal analogues are more adequate as initial therapeutic additions than rapid acting analogues (Table 2). Furthermore, the effectiveness of the drug over any other antihyperglycaemic preparation, and the emerging evidence regarding a possible long-term effect when used early in the course of the disease, present insulin as a valid therapeutic option in select patients.

Conclusions and recommendations

Adequately treating the diabetic patient is both an art and a science. Tailoring the therapeutic regimen that will both achieve normoglycaemia and yet fit the patients’ characteristics, needs and lifestyle requires profound knowledge of the attributes of the drug and a holistic acquaintance with the patient. Successfully incorporating sulphonylurea or insulin as second-line therapies may be beneficial to some and harmful to others. We believe that based on their short-lived, yet robust antihyperglycaemic effect, sulphonylureas can be successfully incorporated into therapy for a limited time to rapidly normalize glycaemia. Furthermore, they can be offered to the well controlled patient on an intermittent basis to help overcome events of overeating or eating food products with high glycaemic indices (such as patients going to weddings, holiday feasts, etc.). When initiating sulphonylurea/glinide treatment, weight gain and secretagogue failure to adequately control glycaemia, despite disease progression, should be anticipated, and a limited therapeutic course should be considered. One may only speculate, whether future combination therapy with (possibly) beta cell protective antihyperglycaemic therapies, (perhaps Glitazones or GLP-1 agonists), will prevent secretagogue failure [32].

Insulin, can be offered to the recently diagnosed patient and to the poorly controlled patient (high HbA1C despite metformin therapy) as a therapeutic choice to achieve rapid normoglycaemia and long-term glycaemic control. The specific regime should be tailored to the patient, although we lean towards recommending initial therapy with basal type insulin, because this method seems to be effective in the Asian population [54,58]. In this setting, attempts to stop/replace insulin with oral drugs after a period of normoglycaemia (at least 2 weeks) should be considered. As with secretagogue therapy, weight gain should be anticipated and measures to prevent it, used.

Overall, although posing significant side effects at times and necessitating a more intense follow-up regime, we believe that secretagogue and insulin therapy should be considered as valid second-line therapeutic options in select patients with T2DM. Prudent follow-up of patients while anticipating therapeutic outcomes will result in enhanced patients’ safety and facilitate achieving and maintaining normoglycaemia. Successfully ‘tailoring’ these drugs to the right patient at the appropriate situation poses a challenge to the physician and highlights the ‘Art and Science’ of practicing medicine.

Conflicts of interest

IR is on the Advisory Board for Novo Nordisk, Astra Zeneca, BMS, MSD and Eli Lilly; consultant for AstraZeneca/BMS, J&J and Eli Lilly (Andromeda, Heal-or, Insuline, Trans-pharma, Teva- Israeli firms); on the Speaker’s Bureau for Eli Lilly, Novo Nordisk, AstraZeneca, Roche and J&J. RE has no conflicts of interests to declare for this paper.