Podcast Summary: Endocrine Feedback Loop – EFL054

Episode Title: Renal Protection of Metformin
Release Date: October 31, 2024
Host: Dr. Chase Hendrickson (Vanderbilt University Medical Center)
Guests: Dr. Steve Whitland (University of Rochester), Dr. Silvio Inzuki (Yale)
Journal Article Reviewed: “Renal Protective Effect of Metformin in Type 2 Diabetes Patients” (forthcoming, JCEM)

Overview:

This episode of the Endocrine Feedback Loop podcast delves into new findings on metformin’s potential renal protective effects in type 2 diabetes. The podcast reviews a large, observational cohort study from Taiwan and critically appraises its design, results, and implications for clinical practice—offering nuance amid ongoing debate about optimal first-line diabetes therapies.

Key Discussion Points & Insights

1. Setting the Stage: Diabetes and Kidney Disease

• Diabetic kidney disease (DKD) is a leading cause of end-stage renal disease (ESRD), underscoring the need for prevention.
• Recent years have seen newer glucose-lowering drugs (SGLT2 inhibitors, GLP-1 agonists) take center stage for their renal and cardiovascular benefits, sometimes overshadowing metformin’s role.
• RCT Data on Metformin: Historically, randomized controlled trials (RCTs) have not shown clear renal benefits for metformin, but emerging observational data, especially from Asian cohorts, suggest possible positive effects.

Quote:

“Diabetic kidney disease is one of the most common complications of type 2 diabetes… ultimately, when end stage kidney disease is achieved, the only treatments obviously are renal replacement therapy with transplantation or with hemodialysis or peritoneal dialysis.”
— Dr. Silvio Inzuki [03:03]

2. Study Design: Retrospective Cohort Analysis

• Population: Over 300,000 patients with diabetes treated at Changung Memorial Hospital system (Taiwan) between 2006 and 2016.
  • Multiple rounds of exclusion and propensity score matching resulted in 13,000+ per group (metformin users vs non-users).
• Exposure: Use vs. non-use of metformin as initial therapy.
• Outcomes:
  • Doubling of serum creatinine (sustained ≥3 months)
  • eGFR <15 mL/min/1.73m² (sustained ≥3 months)
  • End-stage kidney disease (ESKD), as per ICD code.
• Statistical Methods:
  • Propensity score matching and multivariable Cox proportional hazards modeling.
  • Sensitivity analysis for unmeasured confounders.

Quote:

“I really, really like that. A lot of times in observational studies, you have to dig a little bit to really make sure you can understand which is the exposure, what’s the outcome, how this is working. The authors label this very clearly.”
— Dr. Chase Hendrickson [09:31]

3. Limitations in Observational Design & Residual Confounding

• Clinician Prescription Bias: Difficult to know all the reasons why a clinician started or withheld metformin.
• Channeling Bias: Patients may be prescribed to certain therapy due to unmeasured characteristics.
• Baseline eGFR: Though matched post-hoc, prescription patterns are influenced by kidney function—potential for residual confounding remains.

Quote:

“You can’t still get into the head of the clinician making that prescription…why did you prescribe metformin in this other patient?”
— Dr. Silvio Inzuki [18:08]

Quote:

“Propensity matching assumes you know all the variables that need to be matched and that’s not necessarily the case.”
— Dr. Steve Whitland [17:56]

4. Primary Results

• Renal Outcomes (favoring metformin):
  • Doubling of Creatinine: 1,038 (metformin) vs 1,186 (non) | HR 0.71 (95% CI: 0.65–0.77), p<0.0001
  • eGFR <15: 376 vs 439 | HR 0.61 (0.53–0.71), p<0.0001
  • ESKD: 248 vs 303 | HR 0.55 (0.47–0.66), p<0.0001
• Number Needed to Treat:
  • NNT to prevent doubling of creatinine: 89
  • NNT for eGFR <15: 208
  • NNT for ESKD: 239

Quote:

“I found interesting…the number needed to treat to prevent doubling creatinine was 89… end stage kidney disease was 239. Knowing how inexpensive metformin can be…that’s $229,440 per case of end stage kidney disease prevented, or $55,000 per annum…which I believe is cost effective.”
— Dr. Steve Whitland [28:53]

Critical Observations About Results:

• Absolute differences are small (dozens of patients), but with large populations, statistical significance is easy to obtain.
• Dropout Rate: By year 3, approximately 50% of the cohort is lost to follow-up; by year 5, about 75% gone. Raises questions about robustness over longer terms.
• Subgroup Gaps: Authors didn’t stratify effect by baseline eGFR, a major omission.

Quote:

“I would have loved to have seen in the subgroup Figure…I did not see that in that table or figure…I think it would have been very telling.”
— Dr. Silvio Inzuki [35:44]

5. Mechanistic Theories & Plausibility

• Possible Renal Mechanisms:
  • Decreases in hepatic glucose production, improved insulin sensitivity (known pathways).
  • Animal data: reduced fibrosis, decreased tubular injury.
  • Other proposed: anti-inflammatory, immunomodulatory, mitochondrial effects, microbiome shifts.
• Uncertainty: Authors and guests emphasize that true mechanisms remain speculative; direct kidney benefit remains unproven.

Quote:

“Whenever we don’t know about mechanisms, we throw around words like anti-inflammatory and reduction in oxidative stress…Who knows if there’s a purported beneficial kidney effect. Who knows what the mechanism is there?”
— Dr. Silvio Inzuki [34:27]

6. Context within Recent Guidelines & Practice

• Changing Guidelines:
  • Increasingly nuanced; not always defaulting to metformin as first line, particularly in patients with established kidney disease or cardiovascular disease, where SGLT2 or GLP-1 agents may be prioritized.
• Practicality:
  • Most patients end up on dual (metformin + newer agent) therapy.
  • Metformin remains inexpensive and widely accessible.

Quote:

“Most of our patients are going to do well on two medications. And the pairing of metformin with an SGLT2 or a GLP1 has become very standard.”
— Dr. Silvio Inzuki [07:49]

7. Cautions & Conclusions

• Major Limitation:
  • Observational design: can’t prove causation, only association.
  • Unmeasured confounding (physician prescribing bias, uncollected patient variables).
  • Short average follow-up (~3 years).
• Strengths:
  • Propensity matching and thorough statistical modeling increase confidence, though residual confounding cannot be eliminated.
  • Large, real-world patient sample; reflects broad clinical populations.
• Clinical Impact:
  • Evidence does not suggest harm—metformin remains reasonable as an initial therapy, especially when cost is a concern.
  • No evidence here justifies removing metformin; some support for keeping it in the drug regimen.

Quote:

“There’s nothing in this study that would further disqualify Metformin from its lofty cost effective benefit as a first line drug for type 2 diabetes. I don’t know that there’s data here that makes it look like an SGLT2 inhibitor, which almost by accident was found to be a super kidney drug.”
— Dr. Steve Whitland [40:03]

• Comparison with Newer Agents:
  • SGLT2 inhibitors and GLP-1 agonists are costly; their absolute risk reductions for major outcomes are not substantially higher than metformin, especially when used as add-on therapy.

Quote:

“No one’s ever shown that an SGLT2 inhibitor or GLP1 agonist is better than metformin…if they were compared head to head with metformin in a monotherapy trial, I’m not certain that they would win…so more reason to leave the metformin alone and leave it on board.”
— Dr. Silvio Inzuki [44:22]

Memorable Quotes & Timestamps

• On DKD as a major complication:

  “Diabetic kidney disease is one of the most common complications…one of the more feared complications on behalf of our patients…”
  — Dr. Silvio Inzuki [03:03]

• On uncertainty of results due to design:

  “Residual confounding in all of these studies…an important point to think about.”
  — Dr. Chase Hendrickson [18:27]

• On cost-effectiveness:

  “…a year of metformin is $240. GLP1s and SGLT2s are more than 10 times that.”
  — Dr. Steve Whitland [43:56]

Notable Segment Timestamps

• Background & Introduction: 00:00–09:31
• Study Design & Methods: 09:31–21:41
• Statistical Analysis: 21:41–24:16
• Core Results & Outcome Numbers: 24:16–29:17
• Discussion and Mechanisms: 29:17–37:03
• Critical Appraisal of Subgroup Analyses: 35:44–39:30
• Clinical Implications & Wrap-up: 39:30–45:52

Conclusion

The podcast provides a nuanced, critical engagement with new observational data suggesting metformin may confer “renal protection” in type 2 diabetes. However, limitations inherent to retrospective design, high attrition rates, lack of granular subgroup analysis, and persistent confounding prevent definitive conclusions. Despite these caveats, the study does reinforce the appropriateness of metformin as a cost-effective, safe first-line therapy—particularly in combination regimens—while highlighting the ongoing need for more robust comparative data involving modern diabetes agents.

Recommendation:
Clinicians should remain confident using metformin, especially given its cost-effectiveness and safety profile, but should not yet consider it equivalent to—or a replacement for—SGLT2 inhibitors in nephroprotection without further definitive evidence.