Recommendations of Gentamicin Dose Based on Different Pharmacokinetic/Pharmacodynamic Targets for Intensive Care Adult Patients: A Redefining Approach

Introduction

Gentamicin is an aminoglycoside antibiotic which effectively treats serious Gram-negative infections.¹ In addition, global Gram-negative resistance rates are increasing and there is limited availability of antibiotics to treat the emergence of resistance.² An appropriate empirical dose of gentamicin can be suggested based on local epidemiological data and susceptibility patterns. Critical illness and severe infection lead to altered pathophysiology and gentamicin pharmacokinetics. Increased volume of distribution (Vd) is frequently reported in critically ill patients.^2,3 Furthermore, augmented renal clearance of gentamicin in patients with sepsis, severe trauma, undergone major surgeries, or use of inotropic agents also plays a role in an alteration of gentamicin pharmacokinetics.⁴ Therefore, appropriate dosing regimens of gentamicin in a timely manner, particularly during the acute phase of illness, are essential for successful therapeutic outcomes.

Despite the “concentration-dependent” bactericidal activity, the pharmacokinetic and pharmacodynamic (PK/PD) target of gentamicin therapy remains inconclusive in critically ill patients. The maximum gentamicin plasma concentration (C_max) to the minimum inhibitory concentration (MIC) ratio from 8–10 is advocated to justify the prediction of efficacy in critically ill patients for treatment of infections.^4,5 However, in vivo study reported that the 24-hour area under the concentration-time curve to MIC (AUC_24h/MIC) ratio around 30–50 was associated with bacteriostasis, and the ratios around 80–100 were associated with 1 and 2 log₁₀ bacterial reduction, respectively.^3,6 In addition, the AUC_24h/MIC ratio ≥110 significantly improved the clinical cure in patients with serious Gram-negative bacterial infections.^7,8 As a result, the AUC_24h/MIC ratio ≥110 has recently been proposed as an appropriate PK/PD target for severe infection.^6–9 From a safety perspective, AUC_24h >700 mg⋅h/L and the minimum gentamicin plasma concentration (C_min) >2 mg/L are also suggested as the indicator for nephrotoxicity.^9,10 To date, no studies have evaluated the achievement of gentamicin both for PK/PD targets for efficacy (C_max/MIC and AUC_24h/MIC ratios) and risk of nephrotoxicity (AUC_24h and C_min) in critically ill patients. The purpose of this is to define optimal once-daily gentamicin dosing regimens for efficacy and safety in critically ill patients during an acute illness (first 72 hours of infection).

Materials and Methods

Search Strategy, Study Selection and Data Extraction

The following Medical Subject Heading (MeSH) terms: “gentamicin”, “pharmacokinetics”, “critically ill”, “intensive care” and synonymous words were used to search in PubMed, EMBASE, SCOPUS, CINAHL and EBSCO. All original publications of gentamicin traditional pharmacokinetic studies in humans and in English that entered the databases by September 2022 were included. The searched studies were focused to extract individual or study-based pharmacokinetics data with mean and standard deviation. In addition, we also conducted a review on online available secondary references and included these if eligible. Two investigators (AMY and TD) independently found and evaluated studies for potential inclusion and exclusion.

Pharmacokinetic Model Development/Mathematical Pharmacokinetic Models

The traditional gentamicin pharmacokinetic studies in critically ill adult patients were included for evaluation. The studies were excluded if they included pregnant women, patients receiving extracorporeal membrane oxygenation, or renal replacement therapy. All selected published articles based on inclusion and exclusion criteria have undergone extensive scrutiny to gather the required pharmacokinetic parameters of gentamicin to generate the time versus concentration profiles (Figure S1). Data extracted from the selected publications of gentamicin pharmacokinetic parameters (Vd, CL, Ke) and total body weight were combined to calculate mean and standard deviation. In addition, the upper and lower range of each parameter were extracted and used in the MCS.

Furthermore, in a clinical setting gentamicin concentration best fit with the one-compartment pharmacokinetic model.⁵ Hence, in this study, a one-compartment model with first-order elimination was developed to predict gentamicin concentration for 72 hours of the initial antibiotic therapy. The gentamicin dosage was calculated based on total body weight. The equation used in the model to calculate gentamicin plasma concentrations in individual virtual patients is shown in equation 1. The AUC_24h for each day was calculated by using the trapezoidal formula for each virtual individual.

(1)

Where D = gentamicin dose (mg/kg), TBW = total body weight (kg), Ke = elimination rate constant, t = infusion time (0.5 hour); T = dosing interval (24 hours), and CL= Clearance.

Monte Carlo Simulations

The mean, standard deviation (SD), upper and lower PK parameters for gentamicin obtained from the selected studies were used to build 10,000 virtual patients using the Crystal Ball, Oracle software.¹¹ Since the targeted population in the simulation was adults, total body weight (kg) was set at the range 40 to infinity in MCS analysis. Gentamicin plasma concentration versus time profiles from 0–72 hours were generated which were divided into first (0–24 h), second (24–48 h), and third day (48–72 h). The processes mentioned in the above section were repeated to assess the various doses regimen ranging from 5–10 mg/kg once-daily gentamicin.

Probability of Target Attainment Prediction

The probability of target attainment (PTA) was estimated employing PK/PD targets for efficacy, i.e., AUC_24h/MIC, and C_max/MIC with a set of distinctive MIC values. The gentamicin dosing efficacy and risk of nephrotoxicity were assessed for the first and third day by considering C_min >2 mg/L and AUC_24h >700 mg⋅h/L.^10,12 The optimal gentamicin dose recommended for critically ill patients was defined as the dose that achieves at least 90% of the efficacy target with the minimum risk of nephrotoxicity. MICs were set at 0.5, 1, 2, and 4 mg/L, representing a variety of susceptible pathogen inhibitions as determined by the United States Committee on Antimicrobial Susceptibility Testing (USCAST).¹³

Results

Gentamicin pharmacokinetic parameters selected from 21 studies were gathered for the simulation (Table S1).^14–34 This included critically ill patients admitted to medical, surgical, and traumatic care units. Among a total of 1215 patients, 90.5% of were confirmed for severe infections. Seven studies reported gentamicin PK parameters from the first dose, for which plasma levels during the acute phase (within 48–72 hours) were retrieved. Steady state PK parameters were obtained from another 2 studies. The pooled values (mean ± SD) of Vd, CL, and Ke of gentamicin during the acute phase were 0.33± 0.20 L/kg, 4.70 ± 2.89 L/h, and 0.18 ± 0.10 h⁻¹, respectively. Total body weight (70.8 ± 19.9 kg) was used as dosing weight in these critically ill patients. Additional parameters used in the model are presented in Table S2.

The PTAs for the gentamicin efficacy in severe infection (AUC_24h/MIC ≥110) are shown in Table 1. Based on AUC_24h/MIC ≥110, no gentamicin regimens were sufficient to produce the optimal efficacy in critically ill patients for MIC ≥1 mg/L. Gentamicin 7 mg/kg/day was adequate for MIC ≤0.5 mg/L. Figure 1 presents the PTAs for various AUC_24h/MIC targets. Gentamicin 5 and 7 mg/kg/day were the optimal doses for achieving the bacteriostatic (target AUC_24h/MIC > 50) for MIC 0.5 and 1 mg/L, respectively.

Table 1 Probability of Target Attainment (% PTA) for Gentamicin Regimens Achieving AUC24h/MIC ≥110 for MIC of 0.5, 1, 2, and 4 mg/L on Day 1 and Day 3 of Therapy in Critically Ill Patients

Figure 1 Distribution of achieving various AUC24h/MIC ratio using gentamicin regimens ranging from 5–10 mg/kg/day in critically ill patients. The dashed red line represents the optimal doses which achieved at least 90% of the probability of target attainment (PTA).

For the pathogens with MIC of 0.5 mg/L, the optimal gentamicin dose using C_max/MIC ≥8 and ≥10 was 5 mg/kg/day. However, when MIC increased to 1 mg/L, the optimal gentamicin doses were 8 and 10 mg/kg/day for the target C_max/MIC ≥ 8 and ≥ 10, respectively (see Table 2). No gentamicin regimens were sufficient to produce the optimal target of C_max/MIC in critically ill patients when MIC ≥2 mg/L.

Table 2 Probability of Target Attainment (% PTA) for Gentamicin Regimens Achieving Cmax/MIC Ratio of ≥ 8, and ≥ 10 for MIC of 0.5, 1, 2, and 4 Mg/L on Day 1 of Therapy in Critically Ill Patients

Table 3 presents the risk of nephrotoxicity, focusing on either AUC_24h >700 mg⋅h/L, or C_min >2 mg/L. In this study, the probability of developing nephrotoxicity depended on gentamicin dose and duration of therapy. Higher incidence of nephrotoxicity was recorded with criteria of minimum plasma gentamicin level (C_min >2 mg/L), compared with the AUC_24h>700 mg⋅h/L.

Table 3 The Probability of Developing Nephrotoxicity Predicted by AUC24h >700 mg*h/L and Cmin > 2 mg/L on Day 1 and Day 3 of Gentamicin Therapy in Critically Ill Patients

Discussion

Gentamicin is known for its concentration-dependent activity and can be predicted by C_max and C_min for the efficacy and risk of nephrotoxicity or ototoxicity. Hence once-daily dosing is more favorable for clinical practice. During therapeutic drug monitoring, individualized therapy using a Bayesian prediction method avoids the problem of inter- or intra-patient variability (IIV) in patients. IIV leads to low or high gentamicin exposure in clinical settings. Therefore, physicians and clinical pharmacists are always working together to improve the effectiveness and reduce the risk of once-daily gentamicin dosing in hospitals throughout the world.^9,10,12 Our study employs the MCS method to create virtual populations of critically ill patients. This method was examined for various “what if” scenarios for optimal dosing and risk of nephrotoxicity. This technique generates PK/PD data to assess antibacterial dosing regimens and determine PTA. Previously, several studies have used the MCS to establish optimal antibiotic doses for vulnerable patients, proving its efficacy.^35–37 Gathered published pharmacokinetic parameters showed enhanced Vd and fluctuation of clearance due to aggressive fluid resuscitation, vasodilation and capillary leakage, systematic inflammatory response syndrome in acute sepsis. Changes in the Vd (range from 0.27–0.83 L/kg) parameter have been reported in critically ill patients admitted in intensive care units.³⁸ To date, PK/PD targets for gentamicin mostly depend on the plasma concentration and severity of the infection.⁵ In vitro studies data suggest that C_max/MIC ~8–10 is sufficient for all pathogens, even Enterobacteriaceae.³⁹ In clinical settings, C_max/MIC ratio ~8–10 predicts a cure rate of 90% in Gram-negative infected patients.^40,41 A more aggressive PK/PD target of C_max/MIC ≥10 should be considered in infections caused by the high burden resistant pathogens.^42,43 The AUC_24h/MIC was also reported to be a good predictor of the efficacy in ODDG.^6,9

In vivo study proved that there is 1–2 log₁₀ of killing for K. pneumonia when AUC_24h/MIC ~80 and 100, respectively.^3,6,39 Smith et al⁸ reported that the AUC_24h/MIC ratio ≥110 significantly improved the clinical cure in patients with serious Gram-negative bacterial infections. Higher AUC_24h/MIC of ≥150 and ≥175 of gentamicin monotherapy was also associated with fever and leukocytosis resolution in sepsis patients, respectively.⁴⁰ Therefore, in this study, we utilized both C_max/MIC and AUC_24h/MIC for predicting the efficacy of gentamicin in critically ill patients.

For the highly susceptible pathogens with MIC ≤0.5 mg/L, our study revealed that gentamicin 5 mg/kg/day was the optimal dose with C_max/MIC ≥8–10. When the MIC increased to 1 mg/L, a recommended dose should be 8 mg/kg/day to attain PTA. However, for pathogens with MIC of 2 mg/L, no studied dose achieved the optimal target (>90% PTA). Even with the maximum studied dose of 10 mg/kg/day, the target attainment for C_max/MIC of 8 or 10 remained lower than 80% (76.2% and 66.9%, respectively). In similar fashion Gonçalves-Pereira et al⁴⁴ reported that 65.5% of critically ill patients receiving gentamicin 7.4 mg/kg/day were able to reach the target C_max/MIC of 8 for the treatment of bacterial infections with MIC of 2 mg/L. Rea et al⁴⁵ also reported that a gentamicin dosing regimen of 7 mg/kg in critically ill patients is inadequate and only fulfills 20% of the C_max/MIC >10 for MIC of 2 mg/L.

Our simulation showed that when the MIC is 1 mg/L, the optimal gentamicin dose is 8 mg/kg/day to achieve the C_max/MIC target >8. However, gentamicin 10 mg/kg/day should be considered if the target Cmax/MIC is >10. The French guideline also recommended a maximum gentamicin dose of 8 mg/kg/day for general critically ill and surgical patients with trauma, if the MIC is not more than 0.5 mg/L.⁴⁶ However, Allou et al⁴⁷ reported only 30% of ICU patients reached a C_max of ≥16 mg/L with gentamicin 8 mg/kg/day. The study in critically ill patients with severe sepsis also achieved a C_max target of ≥30 mg/L in 59% of patients on the first day of therapy.⁴⁸

While the C_max/MIC target is commonly used in clinical practice due to its simplicity, Roger et al⁴⁹ provided strong evidence that the C_max/MIC >8 is not a reliable indicator of PK/PD target attainment and clinical outcomes based on efficacy and multivariable analysis. AUC_24h/MIC has the advantage of representing cumulative exposure throughout the dosing period and being less affected by variations in sampling times for drug concentration.⁴⁹ The optimal AUC_24h/MIC target may vary depending on the severity of the infection and the patient population, ranging from 30–50 for non-critically ill patients with lower and uncomplicated urinary tract infections (UTIs) or those receiving combination therapy to 80–100 for critically ill patients with non-UTI infections or those receiving gentamicin monotherapy.^3,6 He et al⁵⁰ proposed using an AUC_24h/MIC ≥100 to guide gentamicin dosing in critically ill patients, particularly for infections with MIC ≤1 mg/L. They suggested a starting gentamicin dose of 7 mg/kg/day.⁵⁰

In vivo study revealed that the AUC_24h/MIC ~30–50 was associated with bacteriostatis; therefore, the AUC_24h/MIC ratio >50 should be adequate for prevention of infections, especially in surgical or trauma injury patients.⁶ These patients commonly received gentamicin in combination with cephalosporins and adequate wound irrigation and debridement for prevention of skin and soft tissue infection.⁵¹ In our study, when the MIC was <1 mg/L, the lowest gentamicin dose that attained the AUC_24h/MIC >50 for more than 90% was 7 mg/kg/day. Similarly, the USCAST¹³ recommended a dose of 7 mg/kg/day, which achieved a high PTA (99.8% and 89.5%) with an AUC_24h/MIC ≥30.7 based on MIC values of 1 and 2 mg/L, respectively. However, higher AUC_24h/MIC targets of ≥80.3 resulted in lower PTA, 58.8% and 2.1% for MIC values of 1 and 2 mg/L, respectively.¹³

The AUC_24h/MIC ratio ≥110 is currently accepted to be beneficial for treating severe or serious infections.⁸ Our study revealed that the optimal gentamicin dose for the AUC_24h/MIC ≥110 target was 7 mg/kg/day for pathogens with MIC ≤0.5 mg/L. However, when the MIC increased to 1 mg/L, approximately 56% of patients reached this target with the dose of 7 mg/kg/day. After increasing the gentamicin dose to 10 mg/kg/day, the AUC_24h/MIC target stayed around 80%. The gentamicin dose of 10 mg/kg/day was the only regimen that achieved >90% target of AUC_24h/MIC ≥150 for pathogens with MIC ≤0.5 mg/L, but it simultaneously increased the risk of nephrotoxicity. Therefore, gentamicin should not be recommended as monotherapy for pathogens with MIC >2 mg/L when considering AUC_24h/MIC >110 in critically ill patients.

Previous studies reported that the risk of gentamicin-induced nephrotoxicity depended on dose, long term use, and concurrent administration of other nephrotoxic agents. It was also provoked by alteration in pathophysiological conditions of individual patients.^52,53 Nicolau et al⁵⁴ found that out of 2184 patients, only 1.2% of patients experienced nephrotoxicity (rise in serum creatinine of >0.5 mg/L) after a median duration of 7 days of therapy. The risk of nephrotoxicity is significantly lower when C_min <2 mg/L.¹⁰ However, C_min <1 mg/L was recommended in clinical practice to minimize nephrotoxicity.⁹ The AUC_24h >700 mg⋅h/L was also a good predictor of nephrotoxicity in patients receiving gentamicin once-daily dosing for at least 72 hours.¹² However, in this study, we selected two parameters i.e., AUC_24h >700 mg⋅h/L or C_min >2 mg/L to predict gentamicin-induced nephrotoxicity. Using the target of AUC_24h >700 mg⋅h/L, we found a minimal risk of nephrotoxicity. Our simulation revealed the C_min >1 mg/L ranged from 19–37.9% after three days of gentamicin 5–10 mg/kg/day therapy in critically ill patients. For C_min >2 mg/L, the risk of nephrotoxicity ranged from 5.5–18.4% for 5–10 mg/kg/day of gentamicin dosing regimen on day 3. Therefore, gentamicin once-daily dosing for short duration (for 3 days) showed a minimal risk of nephrotoxicity. Routine monitoring of plasma gentamicin concentrations for nephrotoxicity may not be necessary within the first 3 days after gentamicin initiation.

Gentamicin dosing should be based on epidemiological or geographic antimicrobial susceptibility testing and clinical effectiveness on various MICs.^55–57 Considering both efficacy targets of C_max/MIC ~8–10 and AUC_24h/MIC ≥110 and the risk of nephrotoxicity, the optimal gentamicin dose of 8 mg/kg/day would be sufficient to initiate therapy in critically ill patients. More than 90% of critically ill patients receiving this dose could achieve both targets for treating pathogens with MIC <0.5 mg/L. For pathogens with MIC of 1 mg/L, the %PTA after 3 days of therapy ranged between ~70–90%. Our gentamicin dosing recommendation of 8 mg/kg/day is consistent with the recommendation by the French guideline.⁴⁷ However, gentamicin monotherapy should not be recommended for the treatment of pathogens with MIC ≥2 mg/L.

To our knowledge, this is the first study employing both AUC_24h/MIC and C_max/MIC ratios, as the targets for efficacy of gentamicin in critically ill patients. In this study, we focused on the acute phase of illness, i.e., within 3 days of gentamicin initiation, where the culture and sensitivity results would be available thereafter. If the culture results reveal pathogens with the MIC of gentamicin >2 mg/L, then alternative antibiotics or gentamicin use in combination with other antibiotics should be initiated to manage the infection. There is a lack of consensus in the scientific community regarding a single PK/PD target for both efficacy and nephrotoxicity. Bland et al⁶ have already emphasized the significance of utilizing accurate and effective PK/PD targets. Gentamicin continues to be widely used as an empirical antibiotic and resistance patterns are also changing drastically in significance. Hence, particularly in tropical countries, there is a need for optimal dosing information. This study serves as a valuable resource and guidance for clinicians.

There are a few limitations to this study. First, we gathered the pharmacokinetic data from 21 earlier published studies representing critically ill patients; therefore, inter- or intra-patient variability in pathophysiological conditions could not be excluded in a heterogeneous population. Second, total body weight was used for the prediction; therefore, the recommended optimal gentamicin doses cannot be applicable to the ideal body weight or adjusted body weight for dosing. Third, our result may not apply if the patient has significant changes in pharmacokinetic parameters within the first 72 hours. Our study predicted gentamicin concentrations based on the pharmacokinetic parameters during the acute phase of illness. If clinicians plan to continue treatment for more than 3 days, efficacy and toxicity should be closely monitored.

Conclusion

This is the first study targeting both AUC_24h/MIC and C_max/MIC ratios as targets for efficacy of gentamicin in critically ill patients. The initial gentamicin dose of 8 mg/kg/day was adequate for empirical or documented therapy against a susceptible pathogen, MIC <1 mg/L since it showed both efficacy targets with minimum risk of nephrotoxicity. However, for a documented therapy against a pathogen with MIC ≥2 mg/L, gentamicin monotherapy should not be recommended. While using gentamicin, clinicians should be cautious and maintain close monitoring for therapeutic outcomes and risk of nephrotoxicity. Furthermore, clinical studies are essential to confirm our recommendation.