Saudi Consensus for Low-Premixed Insulin Analogues in Type 2 Diabetes: Consensus Report

We learned from the literature that premixed insulins are short-acting insulin or rapid-acting insulin analogue mixed with intermediate-acting insulin in a fixed ratio, addressing FBG and PPBG in one injection. There are two categories; high-mix and low-mix premixed insulins. We, a Saudi task force, gathered to develop an explicit, evidence-based consensus for the use of the low-mix premixed insulin for better glycemic control. The treatment with premixed aspart 30 was non-inferior to treatment with premixed


Introduction
According to the International Diabetes Federation (IDF) Diabetes Atlas, the worldwide prevalence of diabetes mellitus (DM) is expected to become 9.9% by the year 2045 with a total number of 629 Million [1]. It is associated with significant morbidity and mortality across the globe. Its prevalence is rising rapidly in Saudi Arabia [2]. To address both the fasting blood glucose (FBG) and post-prandial blood glucose (PPBG), the premix insulin formulations provide a combination of both rapid/short-acting and intermediate/long-acting insulins in a fixed ratio, in a single injection. There are two categories; high-mix and low-mix premixed insulins. We, a Saudi task force, gathered to develop an explicit, evidence-based consensus for the use of the low-mix premixed insulin for better glycemic control. This article has the recommendations of this expert panel.

Premixed Insulins
Premixed insulins are short-acting insulin or rapid-acting insulin analogue mixed with intermediate-acting insulin in a fixed ratio, addressing FBG and PPBG in one injection. They are listed hereunder in Table 1.
They have some advantages over the self-mixed insulin. These advantages include accurate dosage, efficacy, and patient convenience. All can be translated into increased compliance and better long-term control of DM [3] [4].

Pharmacokinetics of Insulin Analogues
The pharmacokinetics of medication includes several parameters: the absorption, the distribution, the metabolism, and the elimination. Insulin absorption is  Therefore, for the rapid-acting insulin analogues, the concentration of insulin should peak very quickly after the injection and be rapidly metabolized, and eliminated. The currently available rapid-acting insulins have duration of action of 3 -5 hours according to the injected dose. Meanwhile, the concentration of the long-acting insulin analogues needs to be stable and uniform. In addition, it should be noted that current basal insulin analogues have a waning insulin concentration that may not accommodate for the morning increase in the need for insulin to cover what is called the dawn phenomenon; the transitory increase in insulin resistance in the morning [5] [6] [7].
As for the premixed insulin analogues (biphasic), the perfect profile is the combination of both a rapid-onset and peak insulin action following the injection. That is to cover the postprandial needs, followed by a rapid return to a long plateau to cover the basal needs. Current formulations of premixed biphasic insulins are not ideal in this respect ( Figure 1) [5] [8] [9].
The rapid-acting insulin analogues' profiles simulate the normal physiological insulin secretion in response to meals better than the regular human insulin, with a faster onset and higher maximal concentration. However, the profiles of long-acting insulin analogues are not wholly flat and do not remain steady over the day. Premixed insulin analogues show more rapid absorption relative to premixed human insulin and have different peak concentrations in relation to the dose of the soluble rapid-acting component. The return to basal insulin levels is slow and uneven. This part of the biphasic insulin profile is known as the

Pharmacodynamics
The pharmacodynamic insulin profile is alike to the pharmacokinetic profile in shape. For the rapid-acting insulin analogues, it is a rapid onset and offset, and for long-acting insulin analogues, it is a long, flat, and constant profile. In a clamp study, the blood glucose levels of the subjects are kept within a certain range by infusing IV glucose. The amount of infused glucose to maintain the blood levels depicts the amount of its uptake into the cells due to activation of insulin receptors. This is called the glucose infusion rate (GIR) [5] [15].
The kinetics and dynamics of rapid-acting insulin analogues (aspart and lisporo) are similar, as shown in Figure 2 [16].  [23]. In addition, these results are consistent with the differences between rapid-acting insulin analogues and human insulin [5].
Moreover, for those requiring higher doses of rapid-acting insulin, there is good evidence showing the relative increases in early PPBG utilization with the high-mix insulin analogues (e.g., premixed insulin aspart 50/50 and 70/30, and   [30]. Finally, whilst premixed insulin analogues have been shown to be associated with a lower incidence of nocturnal hypoglycemia compared with premixed human insulin and help in the intensification of insulin therapy, the protaminated/precipitated component does not provide the same duration of action or profile of physiological basal insulin replacement. In essence, premixed insulin is a mixture of rapid and intermediate-acting insulin, that is, carrying the same limitations as the individual components. Unfortunately, neither insulin glargine nor insulin detemir is suitable for mixing with other insulin analogues as this mixing substantially alters their pharmacokinetic properties. In summary, premixed insulin analogues do not have two distinct peaks. Once absorbed, the time of action corresponds to the rapid onset of action of the rapid-acting component, and the duration of action corresponds to the intermediate insulin component. The peak of action of the premixed insulin analogues is unimodal (one distinct peak), corresponding to the maximum effect of the rapid-acting component, and it is steep with a gradual decline. The rapid absorption is because of the rapid-acting component and this rapid pronounced onset of action signifies that these analogues can be utilized immediately before or following a meal [11]

Clinical Utility
It is not clear who could most benefit from the utilization of the premixed insulin analogues. However, based on data from one meta-analysis (Giugliano et al., 2011) [34] and the DURABLE study [35], it can be concluded that that certain people with DM could benefit more by starting an insulin treatment with this type of therapy. Those categories of patients who have HbA1c around 8.5% have higher elevations in PPBG are less willing to carry out several SMBG measurements throughout the day [35]. Whereas the low-mix 25 could be suitable for those patients who have not reached adequate PPBG levels [36] [37] [38], the high-mix 50/50 can enhance the premixed insulin therapy and is good for those who are already using premixed analogues [39].
The premixed insulin analogues 50/50 can also be an option for those using the basal insulin in addition to multiple doses of rapid-acting insulin but are not compliant with the regimen of therapy [40] [41]. In addition, these higher premixed insulins could be an initial choice of therapy for those needing higher daily doses of rapid-acting insulins or those who eat a high carbohydrate meal. [39] [42] [43]. International Journal of Clinical Medicine It is well established that people with T2DM who do not achieve glycemic targets despite the lifestyle modifications, and use of one or more oral antidiabetic drugs (OADs) should start insulin therapy [44] [45]. The type of insulin used in the initial phase depends on the insulin resistance degree and the quantity of consumed food at each meal [33] [46] [47]. Moreover, recent evidence suggests that the low-mix premixed insulin analogues 25 provides a similar glycemic control to that of insulin glargine plus lispro in insulin-naïve people with uncontrolled DM on OADs. Non-inferiority of the low-mix regimen 25 was shown in such category of patients [48].
Once the doctors have opted for one of the low-mix premixed insulin analogues, either the aspart 30/70 or lispro 25/75, insulin can be given at the start at a dose of 0.1 U/kg before breakfast. Then, the dose can be titrated or repeated before dinner on the following days, depending on individual needs [5]. This titration is based on the BG levels pre-meals. Insulin is to be adjusted every three days, according to the average BG levels at the respective times [49]. If HbA1c is not at targets at four months after the initiation of insulin despite an appropriate BG levels before meals, then PPBG should be evaluated [25]. For those patients already on the basal insulin therapy who may not achieve the HbA1c targets due to high PPBG levels despite good fasting levels, the premixed insulin analogues can be used. The total dose of the premixed insulin analogues per day could be given on a 1:1 ration (the same total daily dose). Half the dose is to be given before the breakfast and the other half before the dinner [49]. insulin aspart and 70% protaminated insulin aspart) and the premixed insulin lispro 25 (25% soluble insulin lispro and 75% neutral protamine lispro were used in a BID injection regimen in patients with T2DM. In addition, they assessed the patients' preference for pen devices. Glycemic control was assessed by measuring the HbA1c after three months of treatment [50].

Are All the Low-Mix Premixed Insulin Analogues the Same?
A total of 151 patients with T2DM were screened. Inclusion required insulin treatment for the past six months, total daily insulin dose <1.80 IU/kg, age more than 18 years, HbA1c <12%, and body mass index <35 kg/m 2 . Patients had to be eligible for BID mixed insulin treatment and be able and willing to perform SMBG. Previous therapy with insulin analogues or the use of OADs within the last month was not allowed. Furthermore, those cases with abnormal renal, hepatic, or cardiac functions were excluded. Also, other exclusion criteria included severe uncontrolled hypertension, allergy to the trial medications, pregnancy, or drug or alcohol abuse. Nine patients did not fulfill the inclusion or exclusion criteria, and five decided to withdraw before randomization. Randomization was International Journal of Clinical Medicine carried out using an interactive voice response system. The investigator telephoned the system and obtained the treatment sequence from the underlying randomization codes, thereby ascertaining unbiased treatment allocation in the open-label trial design [50].

Efficacy Assessments
Glycemic control was assessed by the measurement of HbA1c after 12 weeks of treatment. In addition, short-term glycemic control was assessed by SMBG measurements. OneTouch® Profile® BG meter (LifeScan, Inc., Milpitas, California) were used. Patients were asked to perform a seven-point BG profile on any day in the week prior to the start and end of treatment periods [50].

Safety Assessments
Adverse events (AEs) and hypoglycemic episodes were recorded throughout the trial. The investigator classified the AEs as serious or non-serious and assessed the severity of the events. AEs were considered serious if they resulted in death, a life-threatening experience, inpatient hospitalization or prolongation of hospitalization, persistent or significant disability/incapacity, or congenital anomaly/birth defect. All AEs not fulfilling the above definition were classified as non-serious. AE severity was assessed as mild, moderate, or severe. Relation to the trial product was assessed by the investigator as probable, possible, unlikely, or impossible to assess. Hypoglycemic episodes were classified as major (requiring third-party assistance), minor (BG reading <2.8 mmol/L with or without symptoms of hypoglycemia), or symptoms of hypoglycemia (not confirmed by BG reading). Vital signs (pulse and blood pressure, sitting) were recorded at the beginning and end of each treatment period. Measurements were taken after patients had been resting in a chair for 5 minutes [50].

Preference for Pen Device
Patient satisfaction with the pen devices was assessed using two device-specific questionnaires: 1 for the FlexPen and 1 for the Humalog Pen, differing by the device names only. The questionnaires included 16 questions aimed at assessing features of the pens that were related to ease of use, utility and convenience, and confidence in correct diabetes management. In addition, patients were asked to state whether they had experienced any problems using the pens and if so, they were asked to describe them. Each questionnaire was completed at the end of the treatment period in which the patients had used the particular device. The questionnaires were completed by the patients at the clinic before the clinical consultation. While completing the questionnaires, patients were left alone, but with access to a staff member if questions arose. If uncertain about specific questions, the patient was asked to read the question again and answer in the most appropriate way. The patients placed the questionnaire in an envelope and sealed it. At the end of the second treatment period, patients rated the importance of the International Journal of Clinical Medicine features assessed in the device-specific questionnaires. Furthermore, after having used both pen devices, patients' overall pen preference was evaluated using a comparative questionnaire. The patients were asked which pen they found easiest to use and which they would prefer to continue using [50].

Glycemic Control
A total of 132 patients were exposed to Asp 30 and/or Lisp25 and had evaluable efficacy data, thus fulfilling the criteria of the ITT population. Treatment with Asp 30 was non-inferior to treatment with Lisp25 (Table 2). No carryover effects between treatments were indicated. Overall, a mean reduction in HbA1c of ~0.5% from baseline to the end of the second treatment period was observed.
The BG levels at each of the seven-time points of the BG profiles were comparable ( Table 2). The BG analyses were made without adjustment for baseline Table 2. Glycemic control with biphasic insulin aspart 30 (30% soluble insulin aspart and 70% protaminated insulin aspart [Asp30]) and biphasic insulin lispro 25 (25% soluble insulin lispro and 75% neutral protamine lispro [lisp25]) as assessed by HbA1c and BG levels after 12 weeks of BID treatment [50]. Asp

Safety
The safety population included all 133 exposed patients.

Patients' Preference for Pen Device
The device-specific questionnaires assessed features of the pens related to their ease of use, utility and convenience, and to patients' confidence in the correct management of their diabetes (e.g., confidence in setting, controlling, and injecting the correct dose). Analyses of the questionnaire results were performed on the safety population; however, owing to incomplete answers, the exact International Journal of Clinical Medicine number of patients in each crossover analysis was lower (between 111 and 118).
Patients were consistently more satisfied with the FlexPen than with the Humalog Pen, with statistically significantly higher scores given to the FlexPen for all 16-device features assessed in the device-specific questionnaires (  Users were asked to rate the importance of the individual feature of the pen to ensure that the questions dealt with features were relevant for the patients. All were of great significance, as the rating "very/ rather important" was given by 80% to 99% of patients, depending on the actual feature. For example, 96.0% of patients believed that it was "very/rather important" that the dose scale was easy to read (question 1). Regarding this feature, 99.2% of patients answered that they found the dose scale of the FlexPen "very/fairly easy to read," whereas 68.5% gave this assessment to the Humalog Pen [50].
When asked if they had experienced any problems using the pens, 32.4% answered that they had experienced problems with the Humalog Pen, and 9.0% had experienced problems with the FlexPen (P < 0.001). The most frequently reported problems with the FlexPen were various kinds of injection difficulties (8 patients). For the Humalog Pen, this type of problem was also the most frequently reported (23 patients). In addition, several patients found it difficult to read the dose scale of the Humalog Pen (nine patients), and concern was expressed as to whether the correct dose had been injected using this pen (seven patients) [50].
Patients were asked about their pen preference in the comparative questionnaire, completed at the end of the second treatment period (Figure 3). For over-

Summary
The results of this study showed that treatment with premixed aspart 30 was non-inferior to treatment with premixed insulin lispro 25. In addition, Selfmonitored blood glucose (SMBG) levels were comparable. Safety profiles were similar between both treatments, as was the incidence of hypoglycemic episodes [50].
Moreover, the switch between both products of the low-mix family can be carried out without any problem. The US FDA stated that, in emergency conditions, one insulin mix product might be substituted for another on a unitper-unit basis [51].  [52].
In another simulated injection study by Ignaut et al. (2009) showed patients' preference for Humalog KwikPen to the vial and syringe and FlexPen. Humalog KwikPen was found to be easy-to-use, easy-to-hold while injecting, and easy-to-press the injection button. Accurate dose measurements were reported with Humalog KwikPen more than with vial and syringe. Feature comparison of Humalog KwikPen to vial and syringe found the pen device preferable to vial and syringe in all attributes compared. A pen device may offer greater patient acceptance of insulin therapy and improved treatment satisfaction. These findings may be clinically beneficial because health gains may be obtained through increased adherence and enhanced diabetes self-management. The advantages of pen devices used in this study should be considered by health care providers in consultation with patients regarding diabetes treatment options [53]. A third study was a randomized, open-label, two-period, eight-sequence crossover study in insulin pen-naïve patients with DM was carried out by Clark et al. (2010) to determine the patient ease-of-use and preference for the Humalog KwikPenversus the next generation FlexPen. The results of this study showed that among pen-naïve patients with DM who had a preference, the majority preferred the insulin lispro pen over the insulin aspart pen with regard to ease-of-use [54].

Recommendations
• Both products of the low-mix premixed insulin analogues aspart 30/70 and premixed insulin lispro 25/75 have comparable efficacy and safety as shown from the medical literature. Therefore, we can change from one to another safely as demonstrated by the US FDA statement. • In addition, the ergonomic features of KwikPen's design and function may offer important advantages for the user during insulin administration.