Gentamicin in Infective Endocarditis: friend, foe, or just folklore?

t’s good to be back! After a couple of busy weeks on service, it feels like a relief to return to sharing ideas in the ID world. This week, I’m part of the antibiotic stewardship team, and as part of the rotation we’re encouraged to write an educational piece for others in training. I wanted to tackle a classic yet controversial question: the role of gentamicin in endocarditis. Is it really needed? What about prosthetic valve endocarditis (PVE)? And are there still situations where it makes sense? Let’s walk through the evidence together.

Why Gentamicin?

Gentamicin is an aminoglycoside antibiotic that inhibits protein synthesis by binding to the 30S ribosomal subunit. Historically, it was used as an adjunctive agent alongside beta-lactams or glycopeptides for serious staphylococcal infections, including infective endocarditis (IE).

The rationale came from animal models in the 1970s–80s showing that gentamicin, when combined with penicillin or cephalosporins, had a synergistic effect, killing bacteria faster than beta-lactams alone and sterilizing vegetations more quickly. This concept was carried over into clinical practice, and for decades, “gentamicin synergy” became standard teaching.

But animal models don’t always translate into patient benefit. And as we’ve gained more data, the harms of aminoglycosides have become clearer.

The Case Against Gentamicin: Toxicity

The problem is toxicity. In IE, renal failure is the second most common complication after heart failure.

• The EURO-ENDO registry reported acute renal failure in nearly 18% of patients with IE.

• Single-center studies suggest 40–69% of patients develop some degree of renal dysfunction, with 6% requiring dialysis, which carries a 40% mortality risk.

Causes of kidney injury include glomerulonephritis from immune complexes, septic emboli, hemodynamic instability, and drug toxicity (especially aminoglycosides, vancomycin, and certain beta-lactams). Adding gentamicin to this mix is like pouring fuel on the fire.

Organism-Specific Evidence

Staphylococcus spp.

Native Valve Endocarditis (NVE):
The early belief was that gentamicin helped clear bacteremia faster. Several small RCTs (Abrams, Korzeniowski, Ribera, Fortun, Fowler) put that to the test: beta-lactam ± gentamicin. A meta-analysis showed:

• No reduction in relapse (RR 0.86; 95% CI 0.30–2.47)

• Higher risk of renal failure (RR 2.12; 95% CI 0.93–4.81)

• Higher mortality (RR 2.89; 95% CI 1.1–7.6)

In the largest study, 22% of patients receiving gentamicin developed reduced creatinine clearance vs 8% without (P = .005).
Conclusion: gentamicin should not be used for native valve staphylococcal IE.

Prosthetic Valve Endocarditis (PVE):
Traditionally, staphylococcal PVE has been treated with a triple regimen: a backbone (oxacillin or vancomycin) + rifampin + gentamicin. This was based on in vitro synergy data and the theoretical need for more aggressive regimens in prosthetic infections.

But observational data have been underwhelming. Studies by Ryder, Ramos-Martinez, and Drinkovic showed no survival benefit with triple therapy and suggested worse toxicity. Some analyses even hinted that rifampin-based dual therapy without gentamicin was safer and equally effective.

Viridans Group Streptococci (VGS) and Streptococcus gallolyticus

Penicillin-Susceptible Isolates:

• Classic regimens used 4 weeks of penicillin ± gentamicin.

• A 51-patient RCT tested 2 weeks of ceftriaxone + gentamicin vs 4 weeks of ceftriaxone. Cure rates were similar — but crucially, there was no arm of ceftriaxone 2 weeks alone, leaving it unclear whether gentamicin was necessary.

• In the OPAT era, the appeal of shortening therapy with gentamicin is outweighed by the risk of nephrotoxicity, especially in older patients.

Penicillin-Intermediate Isolates (MIC 0.25–0.5 µg/mL):

• Guidelines recommend adding gentamicin, citing concerns about borderline beta-lactam activity.

• But evidence is extremely limited. In practice, ceftriaxone or high-dose penicillin alone often works, and if gentamicin is used, it should be limited to 2 weeks.

Penicillin-Resistant Isolates (MIC ≥0.5 µg/mL):

• Rare in the U.S. but more common elsewhere.

• Therapy often involves ceftriaxone ± gentamicin, vancomycin, or linezolid.

• Evidence is sparse, and therapy must be individualized.

Nutritionally Variant Streptococci (Granulicatella, Abiotrophia, Gemella)

• Granulicatella, Gemella, and Abiotrophia species are uncommon but clinically significant causes of infective endocarditis (IE). You can check my post on strep viridans endocarditis risk!

  These organisms, historically grouped as nutritionally variant streptococci (NVS), are well-known for their variable and often unpredictable susceptibilities to beta-lactams, as well as their association with high relapse rates and complicated courses compared to classic viridans group streptococci.

  A study by Chesdachai et al. highlighted these differences, reporting higher rates of IE with Granulicatella (57.9%) compared to Abiotrophia (26.3%) and Gemella (15.8%). Interestingly, the rates of penicillin non-susceptibility were strikingly high for Abiotrophia (66.7%) and Granulicatella (53.7%), underscoring the challenges of relying on penicillin monotherapy for these infections.

  Further in vitro susceptibility testing supports these concerns. Tuohy et al. reported the following rates of susceptibility for Granulicatella and Abiotrophia, respectively:

  • Penicillin: 55% vs. 8%

  • Amoxicillin: 81% vs. 92%

  • Ceftriaxone: 63% vs. 83%

  • Meropenem: 96% vs. 100%

  Notably, both species demonstrated 100% susceptibility to clindamycin, levofloxacin, and vancomycin, which reinforces vancomycin’s role as a reliable therapeutic option when beta-lactam resistance is encountered. Another study further confirmed these trends, finding that only 11% of A. defectiva isolates were penicillin susceptible compared to 39% of G. adiacens and 100% of G. elegans. Taken together, these data suggest that A. defectiva is generally less susceptible to penicillin, whereas G. adiacens demonstrates relatively lower susceptibility to ceftriaxone. Across all studies, however, vancomycin consistently appears highly active.

  When it comes to clinical outcomes, the evidence is limited but points to a challenging disease course. Several case series — including one with 73 patients, a prospective cohort with 48 cases, and smaller series of 12 and 10 patients — consistently demonstrated similar outcomes with monotherapy versus combination therapy with aminoglycosides (PMID: 25461651). Importantly, relapse rates were not clearly improved by adding gentamicin, despite its historical use to overcome tolerance and improve bactericidal activity.

  As a result, there remains equipoise in the field: some clinicians continue to favor adding an aminoglycoside (usually gentamicin) to a beta-lactam when treating these infections, while others omit it, citing questionable benefit and high nephrotoxicity risk and treat with Vancomycin. The most recent WikiGuidelines recommendations reflect this uncertainty. If aminoglycosides are used, it is crucial to:

  • Confirm in vitro susceptibility,

  • Establish a careful monitoring plan for renal function

  • Engage in shared decision-making with patients

  • Strongly consider limiting gentamicin to 2 weeks, borrowing from experience in enterococcal IE.

  In practice, these infections often require prolonged antibiotic therapy and frequently surgical intervention, given the organisms’ ability to cause destructive valve disease and recurrent infection.

Enterococci

Ampicillin-Sensitive Enterococci

• Why enterococci are tricky:
  Enterococci are intrinsically resistant to many antibiotics. Unlike streptococci, they don’t reliably die with beta-lactams alone — they’re tolerant, meaning antibiotics stop them from growing but don’t kill them. In the 1940s–50s, early penicillin monotherapy for enterococcal IE often failed. But those regimens used low doses and short courses (often 10–14 days), which we now know were inadequate for deep-seated infections. Still, this cemented the idea that penicillin “doesn’t work” alone.

• Birth of the synergy concept:
  In vitro experiments showed that when penicillin (or later ampicillin) was combined with streptomycin or gentamicin, the combo had synergy — killing enterococci more effectively than either drug alone. This became the dogma: for E. faecalis endocarditis, you must use a cell wall agent (penicillin, ampicillin, or vancomycin) plus an aminoglycoside.

• Ampicillin + Gentamicin:
  For decades, the “standard” was ampicillin (or penicillin) plus gentamicin for 4–6 weeks. Cure rates were good in younger patients, but renal toxicity was frequent, especially in older adults. As life expectancy increased, more patients with endocarditis were older and had comorbidities — and aminoglycoside nephrotoxicity became a serious problem.

• Ampicillin + Ceftriaxone:
  In the early 2000s, Spanish investigators (Gavaldà et al.) introduced the idea of dual beta-lactam therapy: ampicillin plus ceftriaxone. The rationale: ceftriaxone inhibits a different penicillin-binding protein than ampicillin, and the combination mimics the “synergy” effect without an aminoglycoside.

  • In vitro: bactericidal activity comparable to ampicillin + gentamicin.

  • Clinical trials: A pivotal 2013 RCT (Fernández-Hidalgo et al.) showed ampicillin + ceftriaxone was non-inferior to ampicillin + gentamicin for cure rates — but with dramatically less nephrotoxicity (3% vs 25%).

  • Real-world practice: Today, many centers (especially in Europe and the U.S.) favor ampicillin + ceftriaxone for most E. faecalis endocarditis. Gentamicin is rarely used unless ceftriaxone resistance is suspected or drug access is limited.

Ampicillin-resistant enterococci:
For E. faecium or the rare intsances where E. faecalis is (often ampicillin-resistant, vancomycin-resistant), the situation is harder. Traditional options included vancomycin + gentamicin, but toxicity is a major barrier. Now, linezolid or daptomycin are usually favored, sometimes in combination with another agent (ampicillin, ceftaroline, or fosfomycin) for synergy. This will be a topic for another post.

Bartonella

• French studies in the 1990s reported excellent outcomes with doxycycline + gentamicin, especially for B. quintana.

• But newer systematic reviews show no clear benefit of adding gentamicin.

• Since Bartonella IE often presents with baseline renal failure, aminoglycosides are especially dangerous.

• Alternatives (doxycycline + rifampin, fluoroquinolones, beta-lactams) seem just as effective.

• See my prior post on bartonellosis!

Brucella

• Brucella IE is rare but severe. Case reports and a small observational study suggested aminoglycoside-containing regimens reduce mortality.

• Current practice often uses doxycycline + rifampin + aminoglycoside, though this is based on expert opinion or retrospective data.

Final Thoughts

The role of aminoglycosides in endocarditis has been steadily declining. For most organisms, the evidence shows no improvement in outcomes but a clear increase in toxicity.

There remain a few niche scenarios — Brucella, selected Bartonella, or resistant streptococci — where gentamicin is still used, but even here the supporting data are weak.

We may be witnessing the end of the aminoglycoside era in IE. It’s a reminder of how tradition in medicine can persist long after the data have shifted — and how stewardship pushes us to constantly re-examine old practices.

Musical Coda