SGLT2 Inhibitors
SGLT2 inhibitors (SGLT2i) are oral glucose-lowering agents that block sodium-glucose cotransporter-2 in the renal proximal tubule, causing urinary glucose excretion regardless of insulin. Their primary clinical significance in the tirzepatide context is not glycaemic control — where they are modest — but rather their dedicated powered cardiovascular and kidney outcomes trials, which give them a distinct and evidence-based role in T2D management that tirzepatide cannot currently replicate. KDIGO 2022 recommends SGLT2i with Grade 1A evidence as second-line therapy (after metformin) in T2D+CKD, ahead of GLP-1 RAs.
Key agents: empagliflozin (Jardiance, Boehringer/Lilly), dapagliflozin (Farxiga/Forxiga, AstraZeneca), canagliflozin (Invokana, J&J).
Ontology SGLT2 Inhibitors [relates] Tirzepatide SGLT2 Inhibitors [relates] KDIGO 2022 CKD Diabetes Guideline SGLT2 Inhibitors [relates] Dedicated CKD Outcome Trial
Kidney Outcomes Trials
CREDENCE (Canagliflozin, NEJM 2019)
- Population: T2D+CKD (eGFR 30–90), n=4,401, median follow-up 2.62 years (stopped early for efficacy)
- Primary endpoint (ESRD, doubling creatinine, renal/CV death): HR 0.70 (95% CI 0.59–0.82) — 30% relative reduction
- Event rates: 43.2 vs 61.2 per 1000 patient-years
- CV secondary: MACE HR 0.80 (P=0.01); HF hospitalisation HR 0.61 (P<0.001)
- ESRD alone: HR 0.68 (95% CI 0.54–0.86)
- First dedicated kidney outcomes trial for any antidiabetic drug in T2D+CKD
DAPA-CKD (Dapagliflozin, NEJM 2020)
- Population: CKD eGFR 25–75 (mean 43.1), n=4,304, median 2.4 years; importantly 32% did not have T2D — establishing kidney protection independent of glucose lowering
- Primary endpoint (≥50% eGFR decline, ESRD, renal/CV death): HR 0.61 (95% CI 0.51–0.72) — 39% reduction; strongest effect of the three trials
- Kidney-specific endpoint alone: HR 0.56
- CV death/HF hospitalisation: HR 0.71
- Benefit consistent in T2D and non-T2D subgroups — confirms haemodynamic/non-glycaemic mechanism
EMPA-KIDNEY (Empagliflozin, NEJM 2023)
- Population: CKD eGFR 20–90, n=6,609, median 2.0 years (stopped early); 54% without diabetes; 35% had baseline eGFR <30
- Primary endpoint (kidney progression or CV death): HR 0.72 (95% CI 0.64–0.82) — 28% reduction; event rates 13.1% vs 16.9%
- Largest of the three; confirms benefit extends to eGFR 20–30 range (where CREDENCE excluded patients)
- All-cause hospitalisation significantly lower
Ontology SGLT2 Inhibitors [evidence_from] Dedicated CKD Outcome Trial SGLT2 Inhibitors [relates] Tirzepatide in Advanced CKD
Cardiovascular Outcomes Trial
EMPA-REG OUTCOME (Empagliflozin, NEJM 2015)
- Population: T2D+established ASCVD, n=7,020, mean age 63, median 3.1 years; mean baseline eGFR 74 (26% had eGFR 30–60)
- 3-point MACE: HR 0.86 (95% CI 0.74–0.99, P=0.04) — placebo-controlled superiority
- CV death: HR 0.62 (P<0.001) — 38% reduction (3.7% vs 5.9%)
- HF hospitalisation: HR 0.65 (P=0.002) — 35% reduction (2.7% vs 4.1%)
- All-cause mortality: 32% relative reduction (5.7% vs 8.3%)
- First placebo-controlled MACE superiority result for any antidiabetic drug; established SGLT2i as cardioprotective
Companion trials: CANVAS (canagliflozin, HR 0.86 MACE, 2017), DECLARE-TIMI 58 (dapagliflozin, HR 0.93 MACE non-inferior, 2019, established HF benefit even without established ASCVD).
Ontology SGLT2 Inhibitors [relates] Cardiovascular Risk SGLT2 Inhibitors [relates] Heart Failure Outcomes
The eGFR Threshold: A Critical Nuance
SGLT2i have a counter-intuitive eGFR profile that clinicians must understand:
| eGFR range | Glucose-lowering effect | Kidney/CV protection |
|---|---|---|
| ≥60 | Full (HbA1c −0.6 to −0.9%) | Full |
| 45–59 | Partial (HbA1c −0.3 to −0.5%) | Full |
| 30–44 | Minimal glycaemic benefit | Full — do not stop |
| 20–29 | No glycaemic benefit | Maintained — continue for CKD/CV |
| <20 / dialysis | No benefit | Insufficient data |
Clinical implication: SGLT2i should not be discontinued when eGFR drops below 45 in a patient using them for kidney or cardiovascular protection (not for glucose lowering). KDIGO 2022 and ADA/EASD 2022 consensus both support continuation to eGFR ≥20. This is the opposite of the historical approach of stopping “renal drugs” at low eGFR.
The mechanism at low eGFR is haemodynamic and metabolic — not glycosuric: SGLT2i reduce intraglomerular pressure (beneficial eGFR “dip” at initiation, analogous to RAAS agents), reduce tubular oxygen demand, and modulate mitochondrial energetics in the myocardium.
SGLT2i vs Tirzepatide: Complementary, Not Competing
| Domain | SGLT2 Inhibitors | Tirzepatide |
|---|---|---|
| HbA1c reduction | −0.6 to −0.9% | −1.87 to −2.58% |
| Weight loss | −2 to −3 kg | −7 to −13 kg |
| Dedicated kidney RCT | Yes — CREDENCE/DAPA-CKD/EMPA-KIDNEY | No — post-hoc only |
| Placebo-controlled MACE superiority | Yes — EMPA-REG OUTCOME | No — active-controlled non-inferiority |
| HF hospitalisation reduction | Yes — HR 0.61–0.65 (placebo-controlled) | Post-hoc; SUMMIT is HFpEF |
| Works below eGFR 45 | Yes (for CKD/CV, not glucose) | Yes (not renally cleared) |
| Route | Oral (daily) | Subcutaneous injection (weekly) |
| Guideline line (T2D+CKD) | 2nd line (KDIGO 1A) | 3rd line (after metformin+SGLT2i) |
The KDIGO 2022 algorithm explicitly positions SGLT2i before GLP-1 RAs for T2D+CKD. When SGLT2i are used, tirzepatide is the preferred intensification agent — the two drugs have minimal pharmacodynamic overlap and are expected to be additive. ADA/EASD 2022 consensus supports combination use.
Ontology SGLT2 Inhibitors [relates] ADA Standards of Care 2023-2024 SGLT2 Inhibitors [relates] ADA EASD 2022 Consensus Report
Connections
- KDIGO 2022 CKD Diabetes Guideline — recommended_by (Grade 1A, 2nd line T2D+CKD)
- Dedicated CKD Outcome Trial — evidence_gap tirzepatide; evidence_confirmed SGLT2i
- Tirzepatide in Advanced CKD — relates (SGLT2i have eGFR 20–30 data; tirzepatide has eGFR <30 excluded)
- Cardiovascular Risk — relates (EMPA-REG is placebo-controlled MACE superiority)
- Heart Failure Outcomes — relates (HF hospitalisation benefit confirmed)
- ADA Standards of Care 2023-2024 — recommended_by
- ADA EASD 2022 Consensus Report — recommended_by