In recent years, with the change of dietary structure, the number of obese people has gradually increased (Wang et al., 2020a). Physiological changes of obese patients are reflected in many aspects of drug metabolism (Cho et al., 2013). Therefore, knowledge of obesity-related pharmacokinetic (PK) changes is crucial for optimizing drug therapy in the obese population. Weight-based dosing is commonly used to several antimicrobial agents, such as aminoglycosides, vancomycin, amphotericin B, and polymyxins (Alobaid et al., 2016). However, for obese individuals, the use of weight-based dosing may result in drug overexposure and toxicity due to nonproportional increases in drug clearance with body weight (Mcleay et al., 2012), so an alternate dosing regimen may be needed. Some studies recommend using ideal body weight (IBW) or adjusted body weight (ABW) when calculating an antimicrobial dose in obese patients to balance efficacy and toxicity (Hites and Taccone, 2015).

Polymyxin B is a cationic polypeptide antibiotic used for the treatment of multidrug-resistant (MDR) Gram-negative bacterial infections. It exhibits concentration-dependent antibacterial activity and the AUC/MIC ratio correlates well with its efficacy (Lakota et al., 2018; Tsuji et al., 2019). Polymyxin B is generally dosed based on total body weight (TBW), but PK data in obesity is limited. Sandri et al firstly reported a population PK analysis of 23 patients weighing 41–110 kg and one patient weighing 250 kg (on continuous venovenous hemodialysis), and found clearance (CL) allometrically scaled with TBW^0.75 (Sandri et al., 2013). However, the case was rare and CL scaled just slightly with TBW (Pai, 2013). Furthermore, subsequent population PK studies revealed that TBW was not significantly associated with CL or volume of distribution (Vd) of polymyxin B, possibly due to nonproportional increases in drug clearance with TBW (Avedissian et al., 2018; Kubin et al., 2018; Manchandani et al., 2018; Miglis et al., 2018; Wang et al., 2020c; Wang et al., 2021; Yu et al., 2021). In addition, the dosage of colistin, also a polymyxin antibiotic, is based on IBW in obese patients (Lam and Athans, 2019). Therefore, experts suggested that TBW was not the appropriate polymyxin B dosing (Kassamali et al., 2015; Meng et al., 2017; Miglis et al., 2018).

From a toxicology perspective, the use of TBW-based dosing in obese patients can lead to polymyxin B overexposure and risk of acute kidney injury (AKI). In a retrospective study of 151 patients, Nelson et al. discovered that a daily dose ≥250 mg was an independent predictor of nephrotoxicity (OR 4.32, 95% CI 1.15–16.25, p = 0.03) in multivariate analysis (Nelson et al., 2015). A Meta-analysis also showed polymyxin dose had a significant relationship with the rate of nephrotoxicity (OR 1.89, 95% CI 1.40–2.55, p < 0.001) (Wagenlehner et al., 2021). These studies raised an important point about how to select polymyxin B dose for obese patients to maximize efficacy while balancing toxicity.

In view of the above, this study aimed to describe the population PK of polymyxin B for obese patients, as well as to design practical regimens for this population using Monte Carlo simulation.

As a weight-based dosing antibiotic with a narrow therapeutic window, polymyxin B regimen in obese patients is particularly concerning. The present study developed a two-compartment model to describe the population PK profile of polymyxin B in obese patients. The typical CL value was 2.86 L/h, which was slightly higher than previous reports (range, 1.87–2.63 L/h). In terms of V and V2 estimates, the values varied widely between studies (Avedissian et al., 2019). Compared with the PK parameters reported by our group and Sandri et al, the typical V value (11.24 L, Table 2) was higher than those of 6.22–6.98 L. In particular, the V2 value (39.70 L) was triple to quadrupled that reported previously (10.57–11.97 L) (Sandri et al., 2013; Wang et al., 2020c; Wang et al., 2021). Although polymyxin B is a weight-based drug, no effect of TBW, BMI, IBW, or ABW on polymyxin B PK parameters was observed in this study. This might be because the bodyweight range (75–125 kg) was narrow.

Obesity can cause related pathophysiological changes that could theoretically alter the PK characteristic of drugs, such as increased Vd, altered hepatic metabolism, renal CL, and protein binding. An understanding of how the Vd of a drug changes in obesity is critical because this parameter determines loading-dose selection (Hanley et al., 2010). Vd is affected by physicochemical property (lipophilicity or hydrophilicity), plasma protein binding rate, and tissue blood flow (Al-Dorzi et al., 2014; Polso et al., 2014). Hydrophilic drugs typically have a low Vd, poor tissue penetration, less deposition in fat, and are primarily excreted through the kidneys. Vd is little affected by obesity and is closely related to IBW or ABW (Al-Dorzi et al., 2014; Shashaty and Stapleton, 2014). However, polymyxin B, as a hydrophilic drug, has the low to intermediate Vd (about 6.3–47.2 L), intermediate to high protein binding (about 58–90%), and non-renal elimination pathways (Sandri et al., 2013; Abodakpi et al., 2015; Avedissian et al., 2019), which makes it difficult to predict the effect of obesity on polymyxin B Vd. In this study, both the V and V2 values of polymyxin B in obese patients were higher than those in non-obese patients as previously reported (Wang et al., 2020c; Wang et al., 2021). However, it has also been reported that the V and V2 values of polymyxin B in non-obese patients ranged from 6.3 to 78.2 L (Avedissian et al., 2019). More research will be required.

With regard to CL, it is the primary determinant when designing a maintenance dose. CL is largely controlled by hepatic and renal physiology (Hanley et al., 2010). As polymyxin B is primarily eliminated by non-renal pathways, obesity is proposed to have only a minor effect on polymyxin B CL, which was observed in this study. On the other hand, many obese patients have co-morbidities such as arterial hypertension, and diabetes (Hites and Taccone, 2015). Therefore, renal insufficiency is common in this population, which may influence polymyxin B CL. In this study, CrCL values varied largely (range from 21.35 ml/min to 239.99 ml/min); however, no effect of CrCL on polymyxin B PK parameter was observed during modeling.

Simulations of likely polymyxin B AUC_ss,24h in the obese patients were conducted for various dosing strategies (Figure 4). For five fixed regimens, the dose of 50 mg or 75 mg q12h was obviously too low and the dose of 125 mg or 150 q12h was too high for the obese population. The dose of 100 mg q12h had the high probability of achieving target exposure (AUC_ss,24h of 50–100 mg h/L) with the low probability of achieving an exposure higher than the proposed target window (AUC_ss,24h > 100 mg h/L). As to bodyweight-based regimens, TBW-based regimens at 120 kg resulted in a high probability of toxicity, IBW-based regimens at 51 kg were associated with a low likelihood of achieving therapeutic window, and all ABW-based regimens had a high probability of achieving target exposure. At present, TBW adjustment is the recommended dosing strategy used in package insert and medication guides (Li et al., 2019; Tsuji et al., 2019). Nevertheless, there is growing scientific evidence that TBW may not or imperceptibly alter polymyxin B PK parameters (Kubin et al., 2018; Manchandani et al., 2018; Miglis et al., 2018; Wang et al., 2020c; Wang et al., 2021; Yu et al., 2021). Manchandani et al. studied 35 patients with a wide range of body weight (36–112 kg) and found the best-fit volume of distribution could not be predicted by TBW (Manchandani et al., 2018). Miglis et al. found that when ABW-based regimens (52, 70, and 85 kg) were associated with a lower probability of toxicity than that obtained in TBW (50, 75, and 110 kg)-based simulations (Miglis et al., 2018). This result was consistent with our research.

In addition to the PTA analysis (Table 3), when MIC values ≤0.5 mg/L, all regimens with a loading dose could achieve a ≥90% PTA, except 100 mg loading dose plus 50 mg q12h fixed. This finding was in agreement with those reports that a loading dose of up to 2.5 mg/kg may be required to achieve PK/PD targets (Avedissian et al., 2018; Miglis et al., 2018). At a MIC of 1.0 mg/L, only a fixed dose of 150 mg q12h and two TBW-based regimens for patients weighing 120 kg had the PTAs over 90%. However, under these regimens, 47.7%–64.5% of patients would have the AUC_ss,24h greater than 100 mg h/L based on Monte Carlo simulations (Table 3). Given that polymyxin B may need to be used to combat extensively drug-resistant organisms that possess a polymyxin MIC of 1.0 mg/L, clinicians should handle the situation with extreme caution and implement TDM. With the current EUCAST MIC breakpoints of 2.0 mg/L (Pogue et al., 2020; Satlin et al., 2020), no regimen could achieve PTA ≥90%. Recently, for insufficient data, particularly clinical PK/PD, CLSI has eliminated the “susceptible” interpretive category for polymyxins, whereas EUCAST maintained the MIC breakpoints for Enterobacterales, Acinetobacter spp., and P. aeruginosa (Satlin et al., 2020). Taken together, the ABW-based regimens but with a daily dose <250 mg would be the optimum regimen for the obese population when MIC values ≤0.5 mg/L.

This analysis has some limitations. First, because only two morbidly obese patients (BMI ≥40) were found during the screening process, the study participants only represented a panel of obese subjects (Table 1). Second, half of blood samples were obtained as part of clinical care for TDM, and therefore, both rich and sparse sampling schedules that differed between patients were included. Third, during the hospitalization of different patients, the polymyxin susceptibility testing methods have changed several times. As the high MW and low diffusing capacity, the very major error for polymyxin susceptibility was relatively high by different antimicrobial susceptibility test methods (Pfennigwerth et al., 2019). Therefore, specific MIC distributions were unavailable. Lastly, this study was not designed to evaluate the clinical efficacy and toxicity of polymyxin B therapy in the obese population.

In conclusion, this study firstly developed a two-compartment population PK model of polymyxin B for obese patients. The results suggested that TBW, ABW, or IBW had no effect on PK parameters. Monte Carlo simulation indicated that ABW-based regimens could achieve adequate exposure with a lower risk of toxicity when compared to TBW-based regimens and the fixed dose of 125 mg or 150 mg would have a high probability of toxicity. PTAs results revealed that TBW, ABW, and IBW-based regimens had comparable PTA values. Therefore, for the obese population, ABW-based regimens but with a daily dose <250 mg would be the optimal regimen to improve polymyxin B therapeutic efficacy and reduce the incidence of nephrotoxicity with the MIC ≤0.5 mg/L.