Important differences in the durability of glycaemic response among second-line treatment options when added to metformin in type 2 diabetes: a retrospective cohort study.

IMPORTANCE: There is limited information about the durability of glycaemic control when different oral glucose-lowering therapies (GLTs) are used as add-on treatments to metformin (MET) in patients with type 2 diabetes mellitus (T2DM). OBJECTIVE: To compare time to treatment failure between different classes of oral GLT when used as second line (add-on) treatments to MET monotherapy at HbA1c ≥ 7.5%. DESIGN, SETTING AND PARTICIPANTS: A retrospective cohort study on 20,070 patients who were newly treated with a sulphonylurea (SU), dipeptidyl-peptidase-4 (DPP-4) inhibitor or thiazolidinedione (TZD) following MET therapy failure (2007-2014). Patients' data was sourced from UK General Practices via The Health Improvement Network (THIN) database. The risk of dual therapy failure was compared between three treatment groups: MET + SU (reference group, n = 15,508), MET + DPP-4 inhibitor (n = 3,080) and MET + TZD (n = 1,482). Follow-up was until treatment substitution or intensification with a 3rd GLT, or for up to 5 years (totalling 46,430 person-years). Propensity score weighting and Cox proportional hazard regression analyses were employed. MAIN OUTCOMES AND MEASURES: Risk of dual therapy failure was compared between treatment groups while adjusting for baseline covariates. RESULTS: Unadjusted survival analysis showed the incidence of dual therapy failure at 1 year was 15% with SU, 23% with DPP-4 inhibitor and 8% with TZD. Corresponding failure rates at 2 years were 26, 38 and 12%, respectively. Adjusted multivariate models showed that, compared to the SU group, adding a DPP-4 inhibitor was associated with an increased risk of treatment failure (adjusted hazard ratio, aHR, 1.58; 95% CI: 1.48-1.68), while adding a TZD was associated with a reduced hazard (aHR, 0.45; 95% CI: 0.41-0.50). Baseline parameters associated with an increased hazard of intensification included HbA1c, diabetes duration, gender, smoking status and the use of statins. CONCLUSIONS AND RELEVANCE: In routine clinical practice, adding a DPP-4 inhibitor to MET is associated with an increased, earlier requirement for treatment intensification compared to adding an SU or TZD. Adding a TZD to MET resulted in the most durable glycaemic response. Key messages The Agency for Healthcare Research and Quality has suggested that the durability of glycaemic response after treatment intensification is best investigated using well-designed long-term observational studies. In routine clinical practice, among patients with T2DM receiving a second line glucose lowering treatment as add-on to MET, the addition of a Thiazolidinediones is associated with the most durable glycaemic response, followed by a Sulfonylurea and then a DPP-4 inhibitor. Factors associated with earlier dual therapy failure included concomitant use of statin therapy, being female, a smoker, those with longer diabetes duration and higher baseline HbA1c levels. The addition of a Thiazolidinediones was associated with significant weight gain (1.8 kg, p < 0.001), while add-on DPP-4 inhibitor produced a significant weight reduction (-1.8 kg, p < 0.001). A very small reduction in body weight was observed with the SU (-0.2 kg, p < 0.001).