Improved Glycaemic Control with Biphasic Insulin Aspart 30 in Type 2 Diabetes Patients Failing Oral Antidiabetic Drugs: PRESENT Study Results

Aims This paper presents the treatment outcomes for patients intiated on biphasic insulin aspart 30 (BIAsp 30) treatment: BIAsp 30-only, BIAsp 30 + sulphonylureas (SU), BIAsp 30 + biguanides (BI), BIAsp 30 + SU + BI, BIAsp 30 + alpha-glucosidase inhibitors (GI), and BIAsp 30 + BI + thiazolidinediones (TZD) after failing oral antidiabetic drugs (OADs) treatment. Methods This was a multi-national, multi-centre, six-month, prospective, open-labelled, uncontrolled, clinical experience evaluation study, with the exception of a three-month study in one country (China) (“all exclude China” and “China”). Initiation and discontinuation of BIAsp 30 treatment were entirely at the discretion of the attending physicians. Results Mean HbA1c, FPG and PPPG were significantly reduced from baseline at three and six months in all groups (P < 0.001). In “all exclude China”, reductions in mean HbA1c, FPG and PPPG at six months were as follows: BIAsp 30-only group (−2.12 ± 1.76% points; −4.82 ± 3.86 mmol/L; −6.89 ± 4.74 mmol/L), BIAsp 30 + BI group (−2.24 ± 1.77% points; −4.48 ± 3.68 mmol/L; −6.66 ± 4.55 mmol/L), BIAsp 30 + SU group (−1.95 ± 1.59% points; −3.98 ± 3.19 mmol/L; −6.25 ± 4.45 mmol/L) and BIAsp 30 + SU + BI group (−1.78 ± 1.20% points; −3.57 ± 2.78 mmol/L; −5.89 ± 3.98 mmol/L). The only serious adverse drug reaction was reported by the BIAsp 30-only group. In the “China” group, reductions in mean HbA1c, FPG and PPPG at three months were: BIAsp 30-only group (−2.16 ± 1.52% points; −3.34 ± 2.49 mmol/L; −6.29 ± 3.92 mmol/L), BIAsp 30 + BI group (−2.44 ± 1.52% points; −4.01 ± 2.50 mmol/L; −7.10 ± 3.96 mmol/L), BIAsp 30 + GI group (−2.33 ± 1.41% points; −4.34 ± 2.52 mmol/L; −7.97 ± 3.99 mmol/L) and BIAsp 30 + BI + TZD group (−1.21 ± 1.60% points; −3.50 ± 2.29 mmol/L; −5.97 ± 3.39 mmol/L). No serious ADR were reported in China. The most frequent hypoglycaemic episodes were diurnal and minor in nature. Conclusions BIAsp 30 treatment in a clinical setting improved glycaemic control in type 2 diabetes patients failing OADs.


Introduction
T he progressive nature of type 2 diabetes mellitus will require most patients to intensify their therapies in order to maintain good glycaemic control in the longer term [2]. Intensification of therapy has been shown to reduce the risk of many complications associated with type 2 diabetes [3]. There are various options for initiating insulin treatment in insulin-naïve patients. According to the recent consensus statement from the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD), patients experiencing poor glycaemic control with lifestyle interventions and oral antidiabetic drugs (OADs) should intensify their treatment with the addition of insulin [4]. For newly diagnosed patients with severely uncontrolled diabetes with catabolism, fasting plasma glucose (FPG) levels above 13.9 mmol/L, random glucose levels consistently above 16.7 mmol/L or HbA 1c above 10%, insulin therapy in combination with lifestyle change is the treatment of choice [4].
Insulin is effective in lowering hyperglycaemia, with potential beneficial effects on triglyceride and HDL cholesterol levels [4,5]. Its disadvantage remains its potential to induce weight gain and cause hypoglycaemia. Patients treated with a combination of insulin + OADs have benefited from the insulin-sparing effects of some OADs, which allow a lower dose of insulin to be used, thereby reducing the potential for hypoglycaemia [2,6]. A Cochrane review of 20 randomised controlled trials has shown that insulin + OAD combination therapy was associated with a 43% relative reduction in total daily insulin requirement, compared with insulin monotherapy [7]. The choice of glycaemic goals and the medications used to achieve them must be individualised for each patient, balancing the potential for lowering HbA 1c and anticipated long-term benefit with specific safety issues and other characteristics of regimens, including side effects, tolerability, patient burden, long-term adherence, cost and the effects of the medications on hypertension and dyslipidaemia [4].
While the efficacy and safety of BIAsp 30 in randomised controlled clinical trials have been established in the literature, data on its use in routine clinical practice situations remain limited. The Physicians' Routine Evaluation of Safety and Efficacy of NovoMix ® 30 Therapy (PRESENT) Study is the largest, multi-national, observational study on BIAsp 30 in real clinical settings completed to date. It aims to collect data complementary to the published clinical data on the safety and efficacy of BIAsp 30 use in type 2 diabetes. In this article, we present results from a subgroup poorly controlled (with HbA 1c Ն 7% at baseline) on OADs and who either added BIAsp 30 treatment to their existing OADs or transferred to BIAsp 30 monotherapy. The patient data were categorised according to the type of therapy received throughout the study, i.e., BIAsp 30 only or combination of BIAsp 30 with OADs (sulphonylureas or biguanides or sulphonylureas+biguanides).

Study Design and Treatment
This was a multi-national, multi-centre, sixmonth, prospective, open-labelled, uncontrolled, clinical experience evaluation study. The objective of this observational study was to evaluate the efficacy and safety of using BIAsp 30, as a monotherapy or in combination with OADs, for type 2 diabetes management in routine clinical practice. The requirement for ethics committee approval or patient informed consent for this study was according to local regulations. In two participating countries, ethics committee approval was required but patient informed consent was not required. In another two participating countries, both ethics committee approval and patient informed consent were required. For the other countries, both ethics committee and patient informed consent approvals were not obtained as these were not required. As this was an observational study, the initiation and discontinuation of BIAsp 30 treatment were entirely at the discretion of the attending physicians and no intervention was added to the physician's routine practice. Except for patients in one country, all patients purchased their own biphasic insulin aspart 30 as in routine clinical practice. No special investigational procedures outside clinical practice were planned.

Patient Inclusion and Exclusion Criteria
As this was an observational study, the only criteria were that patients had type 2 diabetes mellitus, were inadequately controlled on their current therapy and were prescribed BIAsp 30 as monotherapy or in combination with OADs in accordance with the approved labelling. In this paper, we present data from insulin-naïve patients with a baseline HbA 1c of Ն7% who were previously treated only with OADs.

Participating Countries
This study was planned for 15 countries [18] However, data from Sri Lanka were excluded from this paper because data were collected only at baseline. The data presented here are from the remaining 14 countries, with data from China presented separately as data were collected for 3 months only.

Data Collection and Study Endpoints
The efficacy endpoints were the changes in HbA 1c , FPG and postprandial plasma glucose (PPPG) at the end of treatment from baseline. The safety endpoints were the occurrence of hypoglycaemic episodes and adverse drug reactions (ADRs). Patient data were collected at baseline, three months and six months using standardised forms. These were patient demography (at baseline), weight, duration of diabetes, current diabetes therapy, HbA 1c , FPG and PPPG measurements, number of hypoglycaemic episodes and ADRs. Hypoglycaemic episodes and ADRs were based on patient recollection and clinical records from three months prior to baseline visit and from the last visit for three-and six-month datapoints. Major episodes were defined as those where the patient was unable to perform self-treatment, i.e., glucose had to be administered to the patient by another person. Daytime (diurnal) episodes occurred between 06:00 to 00:00.

Statistical Analyses
The patients were categorised into groups according to the therapies received constantly during the study: BIAsp 30-only, BIAsp 30 + sulphonylureas (SU), BIAsp 30 + biguanides (BI), BIAsp 30 + SU + BI, BIAsp 30 + alpha glucosidase inhibitors (GI), and BIAsp 30 + BI + thiazolidinediones (TZD). These BIAsp 30 + OAD combinations had the largest sample sizes in the study population. Patients who had changes in therapy during the study period were not included in this subanalysis. Other combinations were omitted because their sample sizes were too small for meaningful analysis. The safety analysis set consisted of enrolled patients with baseline data. Hence, patients who withdrew from the study but who have baseline data are included in the analysis and contribute with whatever data that is available. Baseline demographic information, diabetes therapy, and efficacy and safety outcomes were presented as descriptive statistics (%, mean Ϯ S.D. and 95% confidence interval). Changes in HbA 1c , FPG and PPPG from baseline were analysed using the paired t-test. Baseline and demographic variables (patients' country, gender, weight, body mass index [BMI], ethnicity, age, duration of diabetes, HbA 1c at baseline, total daily dose of BIAsp 30 per body weight [bw], previous OAD treatment) and current treatment were fitted into a generalised linear model to determine the factors that affected the change in HbA 1c at six months from baseline. Hypoglycaemic episodes and ADRs were presented according to category and severity using summary statistics and event rates. All the statistical analyses were performed using SAS ® version 9.1.3 (SAS Institute, NC, USA). Table 1 shows the disposition of subjects by the type of therapy during the study. The safety population in the BIAsp 30-only group (N = 2,507 and 2,464) was the largest compared with the other BIAsp 30 + OAD combination in both the "all exclude China" and "China" groups (Table 1). In "all exclude China", the majority of patients completed the six-month study (88% to 91% across the groups), while a small proportion continued the study until three months (9% to 13% across the groups). Baseline BMI, HbA 1c , FPG and PPPG were lowest in the BIAsp 30 + SU + BI group (Table 2a). The most common prior OAD treatments among the groups were SU + BI and SU-only (Table 3a). In the "China" group, baseline BMI, HbA 1c , and FPG were lowest in the BIAsp 30-only group while PPPG was low in both the BIAsp 30-only and BIAsp 30 + BI + TZD groups (Table 2b). The most common prior OAD treatments among the groups were SU + BI and BI-only (Table 3b). Since the dosage of SU was not analysed due to the many types available, the dosage of OADs used during the study will not be discussed.

BIAsp 30 Exposure During Study
BIAsp 30 dosage per body weight at three and six months increased from baseline in all the groups (  group, at all visits, the mean dosage was highest in the BIAsp 30 + BI + TZD group (Table 4b). In all the groups, the majority of patients followed a twice-daily injection regimen of BIAsp 30.
In the "China" group, the number of hypoglycaemic episodes at three months were as follows: 1,237 episodes in the BIAsp-only group, 279 Generally the most frequent hypoglycaemic episodes were diurnal and minor in nature ( Figure 2). The rate of nocturnal episodes increased slightly from baseline in the BIAsp 30 + BI + TZD group.

Discussion
Epidemiological analysis of data from the UKPDS study has shown that for every 1% point reduction in HbA 1c , the risks of microvascular complications and myocardial infarction are reduced by 35% and 18%, respectively [3,19,20]. The mean HbA 1c in the insulin monotherapy and combination therapy groups in this study were all reduced by approximately 2% points. This observation suggests that  the reductions in HbA 1c achieved by patients receiving BIAsp could be associated with decreased risks of diabetes-related complications but this will need to be formally tested.
The improvements in HbA 1c in this study were consistent with the observations of a 16-week randomised controlled trial of BIAsp 30 treatment in patients with type 2 diabetes [16]. In that study, the mean HbA 1c was reduced by 1.6% points in patients treated with insulin monotherapy and by 1.7% points in patients treated with insulin plus metformin.
The glucose-lowering effectiveness of individual therapies and combinations is dependent not only on the intrinsic characteristics of the intervention, but also on the baseline glycaemia, duration of diabetes, previous therapy and other factors [4]: many of these were not controlled in the present study and thus comparisons in effectiveness across the different intervention groups is not meaningful. However, it is noteworthy that diabetes control was improved in all the groups studied.

Treatment Regimen
Although all the glucose parameters showed significant improvements at the end of the study, a substantial proportion of patients did not achieve the ADA recommended HbA 1c target of less than 7%. This was probably because the study was not conducted with a trial-to-target regimen and the insulin dosage was not optimised. The eventual dose requirements for patients with type 2 diabetes around the world is expected to be 50 to 100 U per day [21], which is much higher than the average doses for our patients.

Limitations
This study was observational in nature and hence had its inherent limitations. Patients were not randomised to different therapies, nor was the dose of insulin or other anti-diabetic agents controlled. Thus, it is not possible to compare the responses across the groups in a meaningful manner. The method of data collection for hypoglycaemic episodes and ADRs was based on patient recollection, which could have resulted in under-reporting. Further, blood glucose measurements were not recorded during hypoglycaemic episodes that occurred prior to the start of the study. The study was conducted over a short period and was inadequate for capturing long-term trends and observations. Therefore, the results should be taken with some caution. However, the improvement in measure of diabetes control in this large number of patients enrolled in this study are concordant with findings from randomised controlled trials. Further, this study was carried out in a clinical setting, which would be a closer reflection of the "real-world" experience, as compared to highly selected patients clinical trials.

Conclusions
The initiation of BIAsp 30 treatment in a clinical setting, either in as a monotherapy or in combination with OADs, was observed to improve glycaemic control in patients with type 2 diabetes who were poorly controlled on oral agents. This improvement was accompanied with a low occurrence of major and nocturnal hypoglycaemic episodes.