3. Pharmacokinetic parameters used in vancomycin efficacy and toxicity assessments

The most common recommended pharmacokinetic parameters for vancomycin TDM are trough concentration, area under the curve of total daily dose (AUC24h) to minimum inhibitory concentration (MIC) ratio, and steady state plasma concentration (Css) for continuous infusion [13]. AUC refers to the total drug exposure to the administered dose in a defined time period. It has been suggested that AUC24h/MIC≥400, in microorganisms with minimum inhibitory concentration (MIC) of ≤ 1 mg/L, can be an important indicator of successful drug response [9]. Trough concentration assessment is the simplest method of vancomycin pharmacokinetic evaluation. It has been recommended that trough concentration of 15-20 µg/ml would be a suitable target concentration with promising drug efficacy and safety [3]. But previous studies revealed that in many patients, AUC24h/MIC values of ≥400 could be achieved with lower values of trough concentrations (<15 µg/ml) and these trough values could be associated with lower risk of vancomycin-induced nephrotoxicity [14, 15]. So, AUC calculation can be considered as the preferred method for vancomycin pharmacokinetic assessments. Another advantage of AUC calculation is the simplicity of vancomycin dosing based on AUC values according to the Eq. 1 [3].
\(Vancomycin\ dose=\frac{\text{Cl}}{\text{AUC}_{24h}}\) (Eq. 1)
Where vancomycin dose is in mg/day, Cl is drug clearance in L/h, and AUC24h is the area under the cure of total daily dose in mg.L/h.
Also, there are controversies regarding the intermittent or continuous infusion of vancomycin and previous studies failed to reach a superiority for either method. The most important advantages of continuous infusion over intermittent infusion are less variability in vancomycin plasma concentrations, less dependency on time and number of prepared blood samples, and lower incidence of AKI [12]. Results of a recent meta-analysis have demonstrated that although continuous infusion of vancomycin was accompanied by lower incidence of nephrotoxicity, there was no significant difference between continuous and intermittent infusion approaches in terms of clinical efficacy and mortality rate in the patients receiving vancomycin [16]. During intermittent infusion, trough concentration sampling should be done just before the next dose administration when steady state concentration (Css) is achieved, i.e., after 4 to 6 elimination half-lives (about 48 hours in normal kidney patients) [3] and can be used for the purpose of vancomycin TDM during continuous infusion approach. Upon intermittent vancomycin infusion, pharmacokinetic parameters such as k and Vd can be calculated through Eq. 2 and Eq. 3, using two level plasma sampling. To do so, one sample should be the first peak concentration (one hour after the end of infusion) and the other one can be drawn at an optional time during the interval dosing and before the next dose infusion.
\(C_{t}={(C}_{\max})e^{-k(t-t^{{}^{\prime}})}\) (Eq. 2)
Where Ct is plasma concentration at time t in mg/L, Cmax is the first peak concentration in mg/L, k is elimination constant in h-1, t is the time of second blood sampling in h, and t´ is the infusion time in h.
\(C_{\max}=\frac{K_{0}}{k\times V_{d}}(1-e^{-kt^{{}^{\prime}}})\) (Eq. 3)
Where Cmax is the first peak concentration in mg/L, K0 is the drug infusion rate in mg/h, k is elimination constant in h-1, Vd is volume of distribution in L, and t´ is the infusion time in h.
Then, steady state concentrations could be calculated using the aforementioned pharmacokinetic parameters according to Eq. 4 and Eq. 5 [3].
\(C_{\text{ss}}^{\max}=\frac{K_{0}(1-e^{-kt^{\prime}})}{k\times V_{d}(1-e^{-k\tau})}\)(Eq. 4)
\(C_{\text{ss}}^{\min}=C_{\text{ss}}^{\max}e^{-k(\tau-t^{{}^{\prime}})}\)(Eq. 5)
Where\(\text{\ C}_{\text{ss}}^{\max}\) and \(C_{\text{ss}}^{\min}\ \)are peak and trough concentrations at steady state, respectively in mg/L, K0 is drug infusion rate in mg/h, k is elimination constant in h-1, Vd is the volume of distribution in L, t´ is the infusion time in h, and τ is drug interval in h.
During the continuous infusion of vancomycin, steady state concentration could be calculated through the Eq. 6, in which vancomycin clearance is estimated from creatinine clearance through the Eq. 7 [17].
\(C_{\text{ss}}=\frac{K_{0}}{\text{Cl}}\) (Eq. 6)
\(Cl=0.04\left(\text{Cl}_{\text{cr}}\right)+0.22\) (Eq. 7)
Where \(C_{\text{ss}}\)steady state plasma concentration is in mg/L, K0 is infusion rate in mg/h, Cl is vancomycin clearance in L/h, and Clcr is creatinine clearance that is equal to the estimated glomerular filtration rate (eGFR) in L/h.
According to these formulas, target \(C_{\text{ss}}\) values of 20-30 µg/ml and AUC24h values of 400-700 mg.h/L can be achieved. In continuous infusion regimen, loading dose of 20 mg/kg accelerates the steady state concentration achievement. Afterwards, continuous infusion should be immediately initiated upon loading dose administration. According to Eq. 8, it was suggested that in continuous infusion approach, AUC24h could be calculated by one sample after steady state achievement [3].
\(\text{AUC}_{24h}=C_{\text{ss}}\times 24\) (Eq. 8)
Where AUC24h is the area under the curve of total daily dose and Css is the steady state vancomycin plasma concentration.

3.1. Trough concentration

Since many years ago, monitoring of the vancomycin trough concentration has been considered as an accurate, practical, and simple approach for vancomycin TDM purposes. Pros and cons of the trough-only vancomycin monitoring approach are summarized in Table 1. Sample preparation for trough concentration assessment should be done after steady-state concentration achievement. The suitable sampling time in patients with normal renal function can be scheduled after 48 hours of drug administration or before the forth dose. The exact time of sampling should be just before the next dose or up to 30 minutes prior to the next dose. Target trough concentration of 15-20 µg/ml was recommended in critically ill patients with severe Gram-positive infections [11]. Previous studies on vancomycin pharmacokinetics claimed that vancomycin trough concentration had a good correlation with AUC values, especially in adult patients with GFR≥100 ml/min. Also, it was maintained that in such patients, trough concentration of 15-20 µg/ml may result in AUC/MIC values of ≥400 µg.h/ml in microorganisms with MIC≤1 µg/ml [14]. Higher vancomycin trough concentrations (>20 µg/ml) are associated with vancomycin-induced nephrotoxicity. But not all patients with trough concentration of >20 µg/ml proceeded to nephrotoxicity. Nephrotoxicity occurred in about 25-40% of the patients with trough concentration of >20 µg/ml [14]. Although the target trough concentration of 15-20 µg/ml was suggested as an optimum concentration in vancomycin TDM assessments, it was reported that trough concentration of >12.1 µg/ml was significantly associated with an enhanced risk of nephrotoxicity occurrence [18]. Results of a population pharmacokinetic and vancomycin dose simulation study revealed that trough concentrations were highly varied among participating patients with different and/or same renal functions. So, it seems that in order to achieve a suitable clinical response and acceptable vancomycin efficacy with AUC values of 400 to 700 µg.h/ml, trough concentration of >15 µg/ml is not necessary in many patients and can induce nephrotoxicity with no further superior efficacy. Up to 60% of adult patients with trough concentration of <15 µg/ml could achieve target AUC24h/MIC target values of ≥400 µg.h/ml [19]. We can conclude that the preferred approach to vancomycin TDM and pharmacokinetic assessments could be AUC of intervals (AUCτ) calculation rather than trough-only monitoring approach [14]. As per the recent 2020 vancomycin guideline, AUC24h/MIC values of 400-600 µg.h/ml for severe MRSA infections would be a better alternative target to trough concentration of 15-20 µg/ml for vancomycin TDM purposes [19]. It was reported that AUC24h/MIC values of ≥400 µg.h/ml was associated with better clinical outcomes in septic patients and AUC24h/MIC values of ≤650 µg.h/ml was associated with lower risk of vancomycin induced AKI [19, 20]. Another drawback in trough-only monitoring approach could be the possible errors in sampling time. Results of a recent prospective study have revealed that fewer than half of the collected samples were within the normal range of trough concentration sampling times (10-12 hours post-dose) [21]. In general, trough-only monitoring approach with target concentration of 15-20 µg/ml has no longer been supported by recent infectious guidelines due to its lack of clinical efficacy and higher rate of vancomycin-induced nephrotoxicity [22]. According to a retrospective cohort study on vancomycin TDM, trough concentration-based dosing was accompanied by higher treatment failure rate and higher acute kidney injury occurrence in comparison to AUC-based dosing approach [22]. Trough-based vancomycin dose adjustment can be achieved through the Eq. 9 [23].
\(D_{2}=(\frac{C_{t2}}{C_{t1}})\times D_{1}\) (Eq. 9)
While \(D_{2}\ \)is the new dose in mg, \(C_{t2}\) is the target steady-state trough concentration in mg/L, \(C_{t1}\) is the current trough concentration in mg/L, and \(D_{1}\) is the previous dose in mg resulting in plasma trough concentration of \(C_{t1}\).

3.2. Peak concentration

Peak concentration is defined as the vancomycin plasma concentration drawn 1 hour after the end of the 1 hour-infusion period in order to pass the distribution phase [14]. There are controversies about the necessity of plasma peak concentration calculation for vancomycin TDM purposes [25]. Results of many population pharmacokinetic studies revealed that peak concentration was not associated with either vancomycin efficacy or vancomycin-induced nephrotoxicity [14]. However, peak concentration can be used as an essential point in AUC of interval calculation [14]. It was reported that using both trough and peak concentrations in AUC calculation could enhance the precision of assessments in comparison to trough-only consideration in AUC calculation [26]. Results of a recent Bayesian model-based population study have revealed that AUC estimation using peak and trough concentrations was worse than using trough-only approach. The same study claimed that using a peak concentration that drawn just after the end of the infusion period would be better in calculation of AUC values using peak and trough concentrations. So, it seems that peak concentration can be assessed just after the end of 1 hour-infusion in order to achieve better estimation in AUC calculation, especially in one-compartment models. Although the results of a recent pragmatic randomized controlled trial suggested that peak-trough-based TDM approach was significantly associated with higher therapeutic and clinical cure rate, compared to trough-only-based TDM approach, they failed to show a significant difference in all-cause mortality and vancomycin-induced nephrotoxicity between these two TDM approaches [23]. Peak-trough-based vancomycin dose adjustment could be achieved through the Eq. 10 and Eq. 11 [23] as well as by individualized calculation of pharmacokinetic parameters, as mentioned in the Introduction.
\(\tau=\frac{\operatorname{(ln}{C_{\text{peak}}-\ln C_{\text{trough}}})}{K_{e}}+t^{\prime}\)(Eq. 10)
\(Dose=C_{\text{peak}}\times K_{e}\times V_{d}(\frac{1-e^{-K_{e}\tau}}{1-e^{-K_{e}t^{{}^{\prime}}}})\)(Eq. 11)
Where τ is dosing interval in h, \(C_{\text{peak}}\) is steady-state peak concentration in mg/L, \(C_{\text{trough}}\) is steady-state trough concentration in mg/L, \(K_{e}\) is elimination constant in h-1, \(t^{{}^{\prime}}\) is infusion time in h,\(\text{\ \ V}_{d}\) is volume of distribution in L andDose is the new vancomycin dose in mg.

3.3. AUC

Based on recent reports on vancomycin dosing, AUC24hcould be the preferred approach to TDM purposes [27]. Pros and cons of the AUC-based vancomycin monitoring approach are summarized in Table 2. AUC24h calculation can be done, based on Bayesian software programs using a trough-only sampling approach or peak-trough sampling approach, while the latter results in higher accuracy in AUC estimation [12]. Vancomycin dose adjustment and AUC calculation, based on available Bayesian software programs including Adult and Pediatric Kinetics (APK), BestDose, DoseMe, InsightRx, and Precise PK can be considered as an alternative approach to practical uses of clinicians and pharmacists for the purpose of vancomycin TDM and dose-optimization. Such available soft wares are simple, flexible, and user friendly that can be used by pharmacists and clinicians in the field of vancomycin TDM [28]. Results of a recent review article on the evaluation of the accuracy and efficacy of such Bayesian tools have revealed that similar AUC estimation could be achieved through this approach in comparison to pharmacokinetic equations using two-point blood sampling assay for TDM purposes [29], but further larger meta-analysis and systematic review studies are required, especially for patients with altered pharmacokinetics to assess their accuracy and clinical efficacy in comparison to previous approaches such as AUC calculation using trapezoidal method and individualized pharmacokinetic parameters calculation using at least two vancomycin plasma concentration.
The recommended target value of ≥400 µg.h/ml with MIC value of <1 µg/ml as well as a cut-off point of ≈600 µg.h/ml should be considered to avoid vancomycin-induced AKI occurrence [30]. It was reported that although there was a significant correlation between trough concentration and AUC24h, it was moderate (R2 of 0.51). Results of a recent population pharmacokinetic study has revealed that AUC values could vary about 30-folds in the patients with different renal functions, lending support to the importance of vancomycin TDM and individualized pharmacotherapy to avoid vancomycin-induced nephrotoxicity in over-dose patients and prevent clinical response failure in under-dose individuals. Also, the studies indicated that trough-only monitoring approach could not be an accurate and suitable surrogate of AUC calculation since the significant correlation was not obvious [14, 24]. It was suggested that an AUC24h threshold value of 700 µg.h/ml should be considered to avoid vancomycin-induced nephrotoxicity. AUC24h values of >700 µg.h/ml were significantly associated with higher incidence of vancomycin-induced nephrotoxicity [14]. Results of a retrospective pharmacokinetic study on American population revealed that patients with AUC24h≥297 µg.h/ml had more than 2.7-fold improvement in clinical response in comparison to those with lower AUC24h values. Also, it was reported that patients with AUC24h≥710 µg.h/ml had more than 7-folds higher risk of nephrotoxicity occurrence due to vancomycin over-exposure [31]. In a recent prospective study, among the participants, 19% had therapeutic trough concentration while 70% of them had therapeutic AUC values. Also, the results of this study revealed that 31% of the patients with AUC≥400 µg.h/ml had trough concentration of <10 µg/ml with 68% of whom were with trough concentration of <15 µg/ml, suggesting that AUC rather than the vancomycin trough concentration can be considered as a suitable pharmacokinetic parameter, in order to obtain enough clinical efficacy with lower incidence of nephrotoxicity. The acceptable AUC targets can be achieved with lower plasma trough concentrations [21]. Results of a retrospective pharmacokinetic study in Japanese population revealed that AUC-guided vancomycin TDM (target AUC>400 µg.h/ml), compared to trough-guided TDM (target trough concentration of 15-20 µg/ml), could be associated with lower risk of nephrotoxicity occurrence [32, 33]. Overall, according to the reports, AUC-guided, Bayesian estimation dosing of vancomycin was accompanied by lower incidence of vancomycin-induced nephrotoxicity, shorter duration of antibiotic therapy, fewer blood samples, less vancomycin exposure, and less over-dose occurrence with cost-effectiveness. So, it seems reasonable to shift from trough-only-guided dosing approach to AUC-guided dosing approach for vancomycin TDM in referral hospitals in order to maintain the therapeutic window [21, 34]. Besides the many advantages mentioned about the use of AUC24h/MIC target concentration of 400-600 µg.h/ml for MRSA infections, yet there are some drawbacks that should be taken into accounts. First, the target AUC24h/MIC value of 400-600 µg.h/ml does not contribute to other Gram-positive microorganisms that are less virulent than MRSA, such as Methicillin-resistant coagulase negative Staphylococcus aureus . Also, it seems that the recommended concentration of 400-600 µg.h/ml is suitable for sepsis, pneumonia, and endocarditis while other severe infections such as meningitis and osteomyelitis may require different AUC target values. Meanwhile, a recent meta-analysis has revealed that AUC24h/MIC target concentration of >400 µg.h/ml is not associated with reduced morbidity and mortality in severe cases of MRSA infection [19].

3.4. Vancomycin clearance (ClV)

Vancomycin clearance (ClV) is considered as a pharmacokinetic parameter in the prediction of vancomycin efficacy and toxicity. Results of a previous observational study on vancomycin administration revealed that ClV was correlated with creatinine clearance (calculated via Cockcroft-Gault equation), serum creatinine, gender, age, weight, and neutropenia. There was a correlation with R2 of 0.5 between ClVand creatinine clearance, so it seems that ClV should not be considered as a suitable predictor in vancomycin clinical pharmacokinetic assessments. Results of this study revealed that creatinine clearance had a good correlation with 24h-urine creatinine (with R2 of 0.8-0.9). It seems that creatinine clearance calculation using 24h-urine creatinine assessment can promote the correlation between ClV and creatinine clearance in vancomycin TDM. In general, it can be suggested that ClV, due to its high prediction errors, can not serve as a suitable and practical clinical pharmacokinetic parameter for TDM purposes [35].

3.5. Elimination constant (k)

Elimination constant (k) is an indicator of renal function during the administration of a hydrophilic drug such as vancomycin with almost complete renal excretion. So, the higher the k values, the better kidney function is predictable. While in patients who progress to AKI due to vancomycin exposure, lower k values and higher t ½ amounts are expected. In cases with normal renal function with t ½ of about 4-6 hours, k values of 0.115-0.173 h-1 are acceptable and the values lower than the mentioned values can be considered as an alternative pharmacokinetic parameter for early detection of vancomycin associated nephrotoxicity.