Gentamicin in urosepsis: interpreting levels when renal function is reduced
A 70-year-old man (68 kg, 172 cm) is admitted with urosepsis. Blood and urine cultures grow Escherichia coli susceptible to gentamicin (MIC 1 mg/L), ceftriaxone (MIC 1 mg/L or less) and piperacillin-tazobactam; the isolate is ciprofloxacin-resistant and a phenotypic ESBL screen is negative. Serum creatinine has been stable at 2.0 mg/dL (177 umol/L) for 48 hours and urine output is 1.2 L/day. He has received gentamicin 340 mg IV over 30 minutes every 24 hours for three doses. On day 3, a peak level of 11.2 mg/L was drawn 30 minutes after the end of the infusion and a trough of 2.8 mg/L was drawn 30 minutes before the fourth dose. He has been afebrile for 24 hours, mean arterial pressure is 82 mmHg, and he is receiving no other nephrotoxins. The team asks whether the current regimen can be continued, and the microbiology laboratory has confirmed the ceftriaxone susceptibility.
1. Calculate Cockcroft-Gault creatinine clearance and, from the two concentrations, estimate the elimination rate constant, half-life and volume of distribution. Judge whether the peak and trough are at target and state the main safety concern.
Read the worked answer
Creatinine clearance (Cockcroft-Gault, actual body weight):
CrCl = [(140 - 70) x 68] / (72 x 2.0) = 4760 / 144 = 33 mL/min.
Elimination rate constant (levels 23 hours apart):
ke = ln(11.2 / 2.8) / 23 = 1.386 / 23 = 0.060 h^-1, so t1/2 = 0.693 / 0.060 = 11.5 hours.
Volume of distribution: back-extrapolate the peak to the end of the infusion: C0 = 11.2 x e^(0.060 x 0.5) = 11.5 mg/L; Vd = 340 / 11.5 = 29.6 L (0.44 L/kg), above the usual 0.25-0.35 L/kg, consistent with sepsis.
Interpretation: peak 11.2 mg/L against MIC 1 mg/L gives Cmax/MIC about 11, which is adequate for a gram-negative infection. The trough of 2.8 mg/L exceeds the conventional target of less than 1 mg/L, so the patient is accumulating drug. The 24-hour AUC is dose/(Vd x ke) = 340/(29.6 x 0.060) = about 190 mg x h/L, well above the 70-120 mg x h/L efficacy range and into the exposure range linked with nephrotoxicity. With a half-life of 11.5 hours, a 24-hour interval cannot yield an acceptable trough: about 40 hours (ln 11.2 / 0.060) are needed for the concentration to fall to 1 mg/L. The safety concern is dose-related acute kidney injury and, with prolonged use, vestibulocochlear toxicity.
2. The team wants to keep gentamicin as directed therapy. Propose a revised dose and interval that keeps the trough below 1 mg/L while maintaining Cmax/MIC of at least 8, and state the monitoring plan.
Read the worked answer
Two reasonable options: keep 340 mg but extend to every 48 hours, or reduce to 4 mg/kg (272 mg) every 36 hours. The first is simpler and safer at this level of renal function.
With 340 mg every 48 hours, predicted steady-state predose concentration = 11.5 x [1/(1 - e^(-0.060 x 48))] x e^(-0.060 x 48) = 11.5 x 1.06 x 0.056 = about 0.7 mg/L, with a predicted peak of about 12 mg/L, giving Cmax/MIC of about 12. Both targets are met.
Monitoring plan:
- Recheck a peak (30 minutes after end of infusion) and trough (within 30 minutes before the next dose) after the second dose on the new schedule.
- Document sampling times precisely; a level drawn too early reads falsely high and leads to unnecessary dose cuts.
- Daily serum creatinine and urine output; avoid NSAIDs, iodinated contrast and other nephrotoxins.
- Ask about tinnitus and vertigo; consider audiometry if treatment exceeds 7 days.
- Extended-interval dosing assumes no neutropenia, no endocarditis and no pregnancy; it is not the same as synergy dosing.
3. Give the antimicrobial stewardship argument for stopping gentamicin altogether, and state what regimen you would use instead, including duration and follow-up.
Read the worked answer
Gentamicin adds nothing here. The organism is fully susceptible to ceftriaxone, the ESBL screen is negative, the source is urinary and drainable, and the patient is responding clinically. Aminoglycosides are not recommended as monotherapy for urosepsis, and adding one to a beta-lactam has not improved outcomes in this setting, while a trough of 2.8 mg/L carries a real risk of acute kidney injury and ototoxicity. Continuing it because it was started empirically is the definition of unjustified polypharmacy.
Recommendation: stop gentamicin and give ceftriaxone 2 g IV once daily (1 g is also acceptable). Treat for a total of 7 days of effective therapy, counted from source control (or from the first afebrile day with a controlled source). Reassess at 48-72 hours. Oral switch is only possible if an oral agent is susceptible - ciprofloxacin is resistant here, so check trimethoprim-sulfamethoxazole susceptibility if early discharge is planned, otherwise complete IV therapy.
Follow-up: repeat blood cultures only if fever recurs; monitor creatinine for 48 hours after stopping the aminoglycoside, since aminoglycoside-associated kidney injury can be delayed. If the isolate had been an ESBL producer, or if the patient failed to improve, meropenem rather than ceftriaxone would be the choice.