Retrospective Analysis of Oral Glucose Lowering Medications Effect on Risk of Hospitalization From Heart Failure
PubMed
Clinical Question
Are patients with type 2 diabetes (T2DM) at more of a risk of hospitalization from heart failure when newly initiated on treatment with a sulfonylurea versus a DPP-4 inhibitor or a thiazolidinedione (TZD)?
Bottom Line
When prescribing new oral glucose lowering medications to patients with T2DM, DPP-4 inhibitors (DPP-4i) have less of a risk of hospitalization due to heart failure as a primary diagnosis when compared to sulfonylureas. However the number of fatal hospitalizations due to heart failure was no different between DPP-4i, sulfonylureas, and TZD.
Major Points
Prior to this study, two other studies were performed to examine the effects of DPP-4i on vascular outcomes. Among these, the SAVOR study assessed saxagliptin (DPP-4i) versus placebo for hospitalization of patients due to heart failure. This study found that saxagliptin was associated with an increased risk of hospitalization over placebo. In contrast, the EXAMINE trial showed that in patients with recent acute coronary syndrome, alogliptin (DPP-4i) did not increase or decrease the risk of hospitalization from heart failure.
This study found that when compared to sulfonylureas, DPP-4i have a decreased risk of hospitalization from heart failure as a primary diagnosis as compared to SU or TZDs. It also showed that glitazones have a lower, but not significant incidence of hospitalization from heart failure (HHF) when compared to sulfonylureas, and a higher, but not significant incidence of HHF when compared to DPP-4i. None of the three classes of medication showed a difference in death occurring from HHF.
Guidelines
Design
- Retrospective analysis of claims data from the Nationwide OsMed Health-DB Database accounting for 32 Health Services of 16 Italian regions
- N=127,555
- DPP-4i (n=18,294)
- Sulfonylureas (n=92,463)
- TZD (n=16,798)
- Data Indexing Time Period: January 1st, 2010- December 31st, 2012
- Mean follow-up post index date: 2.6 years
- Analysis: Cox proportional hazard regression
- Propensity matching for sulfonylureas compared to other agents
- Primary outcome: Hospitalization for heart failure occurring after the first 6 months of therapy
Population
Inclusion Criteria
- Patients with T2DM who were newly prescribed a DPP-4i, sulfonylurea, or TZD for the first time +/- metformin
Exclusion Criteria
- Previous hospitalization for heart failure during the 12 months before the index date
- Insulin use during the 12 months before index date or during observation period
- Observation time shorter than 6 months after index date
Baseline Characteristics
There was not a significant difference between groups at baseline for any demographic.
- Mean age ± SD:
- DPP-4i: 67.0 ± 13.4
- Sulfonylureas: 62.3 ± 11.6
- TZD: 63.5 ± 13.2
- Sex (% male):
- DPP-4i: 56.3
- Sulfonylureas: 50.5
- TZD: 54.6
- Charlson Index:
- 1:
- DPP-4i: 58.4
- Sulfonylureas: 76.8
- TZD: 64.7
- 2-3
- DPP-4i: 37.7
- Sulfonylureas: 20.1
- TZD: 31.6
- ≳4:
- DPP-4i: 3.9
- Sulfonylureas: 3.1
- TZD: 3.7
- 1:
- Previous CV event (%):
- DPP-4i: 5.3
- Sulfonylureas: 4.4
- TZD: 4.6
- Previous glucose lowering medications (%):
- DPP-4i: 77.8
- Sulfonylureas: 24.1
- TZD: 54.8
- Other medications: diuretics 1.1%, beta blockers 2.2%, calcium channel blockers 1.5%, renin-angiotensin system blockers 15.3%, combination of blood pressure lowering drugs 31.8%, lipid lowering 26.3%, anti-inflammatory 20.0%, anti-chronic obstructive pulmonary disease 7.5%, antiplatelet 26.2%
Interventions
Not applicable.
Outcomes
Primary Outcomes
- Annual rate of hospitalization with a primary diagnosis of HF
- Propensity matched outcomes:
- DPP-4 inhibitors vs. Sulfonylureas: 0.70 (95% CI 0.52 - 0.94; P=0.018)
- TZD vs. Sulfonylureas: 0.82 (95% CI 0.63-1.06; P=0.126)
- Non-propensity matched outcomes:
- DPP-4 inhibitors vs. Sulfonylureas HR 0.78 (95% CI 0.62-0.97; P= 0.026)
- DPP-4 inhibitors vs. TZD HR 0.89 (95% CI 0.74-1.06; P=0.188)
Criticisms
- Using only the primary diagnosis of heart failure as the inclusion criteria may lead to missed events related to the oral medications, due to the order of coding.
- There are limitations that the study itself points out due to lacking data on BMI, glucose control, other complications due to progression of diabetes, and asymptomatic LVF. Not knowing this information makes it difficult to determine that HHF is a direct result of the medication.
- HHF may be due to a lack of adherence to the medication therapy, rather than a direct result of the medication itself.
Funding
- Funding was provided by the Italian Medicine Agency (government agency). Two authors received funding or lecture fees from Merck Sharp & Dome, AstraZeneca, Novartis, Boeringher Ingelheim, Sanofi, Mediolanum, or Takeda. The other author have no conflicts of interest.