A Vancomycin Dosage Regimen Developed for Obese Patients
A Vancomycin Dosage Regimen Developed for Obese Patients
The results of the study described here demonstrate that obese patients may benefit from a vancomycin dosing strategy different from that generally recommended for adult patients with normal renal function. When obese patients in our institution received vancomycin maintenance doses of 15 mg/kg every 8–12 hours, consistent with our original dosing protocol, above-target trough levels were noted in 55% of patients overall and 79% of ICU patients. Moreover, the mean trough in patients with undesirably high values was 29.1 μg/mL (Table 3). The original dosing protocol was consistent with recommendations subsequently published in the ASHP–IDSA–SIDP consensus statement, which recommends the maintenance administration of vancomycin 15–20 mg/kg (based on TBW) every 8–12 hours for most patients. Due to the frequent occurrence of elevated troughs, our revised institutional protocol recommends an empiric decrease in the maintenance dosage of vancomycin to either 10 mg/kg every 12 hours or 15 mg/kg every 24 hours (based on TBW). These dosing regimens were chosen because modeling indicated that they more consistently resulted in steady-state trough vancomycin concentrations of 10–20 μg/mL, as well as an AUC/MIC of ≥400 when the MIC is ≤1 μg/mL. Of note, a more stringent target trough concentration of 15–20 μg/mL was not used to derive the revised dosing protocol or to be the target trough in this study because a dosing regimen that yields an AUC/MIC of ≥400 despite a trough of 10–15 μg/mL is clinically acceptable. The results of our comparison study indicated a higher frequency of target trough attainment (59% versus 36%, p = 0.006) and a lower prevalence of above-target troughs (18% versus 55%, p < 0.001) with the revised versus original protocol.
Guidance for vancomycin dosing in an obese population is limited. Historically, many clinicians have assumed that in obese people, vancomycin CL increases proportionally with TBW. This assumption is logically based on the increase in organ size, blood volume, and the distribution of vancomycin into the excess extracellular fluid that accompanies excess adipose tissue. Most available data support the use of TBW for empirical vancomycin dosing on the basis of the strong linear correlation of vancomycin CL and TBW reported in multiple studies (although a more recent analysis published by Leong and associates showed that vancomycin CL was most closely associated with ABW). Notably, much of this research has also demonstrated that vancomycin V increases in obese patients but not to the same extent as CL. Based on these assumptions, if one obese patient and one normal-weight patient were to receive a similar vancomycin dose (i.e., calculated by milligram per kilogram of TBW), k in the obese patient should be greater, thus leading to a shorter t½ and a lower vancomycin trough concentration. In that context, the results of our study were intriguing. Our original protocol recommended a vancomycin dosage of 15 mg per kilogram of TBW per dose for all patients regardless of weight, and we observed that most obese patients with normal renal function had above-target troughs (>20 μg/mL).
A plausible explanation for the high frequency of above-target troughs documented in our study is suggested by two related factors. First, the significant variability and relative increase in vancomycin V in obese patients are often underappreciated. Vance-Bryan et al. suggested that the vancomycin pharmacokinetic parameter most influenced by weight is not CL but V; their research found that ABW and the percentage over lean body weight were both independent and significant predictors of V. Also, Ducharme et al. described significant variability in the vancomycin V, with a range of 0.58–1.17 L/kg of IBW; those investigators specifically commented that regardless of sex or age, the average vancomycin V was increased by 37% (from 0.65 L/kg of IBW in normal-weight patients to 0.89 L/kg of IBW in obese patients). Second, variability in V is often not even considered, either in the literature or in routine clinical practice. For example, in their attempt to determine the relationship between vancomycin CL and vancomycin trough concentrations, Leong et al. used a fixed V. Additionally, similar to most other institutions, our institution uses a dosing protocol that only accounts for body weight and estimated vancomycin CL and not V when determining an empirical vancomycin dosing regimen. Therefore, the variability of vancomycin pharmacokinetics certainly warrants continued therapeutic drug monitoring and appropriate dosage adjustments, especially in obese patients.
The potential pitfalls of underappreciating the impact of a higher-than-expected V in obese patients are illustrated in Figure 1. According to our original institutional protocol, one obese and one normal-weight patient of the same sex who were similar in age and other characteristics and had the same estimated CLcr (e.g., 90 mL/min) would have both received a vancomycin regimen of 15 mg/kg of TBW every 8 or 12 hours. The data presented above support the assumption that vancomycin V is likely to be greater in the obese patient. Because both patients have the same CL, an elevated V in the obese patient would be accompanied by a reduced k and consequently an increase in the vancomycin trough level, as shown in Figure 1. One might argue that this hypothetical situation would not occur in clinical practice because obese patients have increased CL that would offset or even surpass the impact of the elevated V. However, in clinical settings, there may be some other reason for decreased renal function in the obese patient such that the impact of obesity on CL is offset. Because formal pharmacokinetic analyses were not performed in the analysis described here, we can only offer this scenario as a logical explanation for the high frequency of above-target vancomycin troughs in the revised-protocol cohort. However, this assumption is based on multiple previous analyses of postdistributive vancomycin levels in hospitalized obese patients that have been performed in our institutions; these analyses have repeatedly confirmed a larger-than-expected vancomycin V in the context of a CL that was close to the predicted value.
(Enlarge Image)
Figure 1.
Effect of an increased vancomycin volume of distribution (V) on the concentration–time profile after a given dose. Curve A represents a concentration–time profile resulting from predicted pharmacokinetic values. Curve B shows a concentration–time profile resulting from a higher-than-expected V. In the equation, k = elimination rate constant, CL = clearance, and Cminss = steady-state trough concentration. Adapted with permission from reference 7.
Despite the higher frequency of target vancomycin troughs among patients dosed according to the revised protocol, a below-target trough was noted in a clinically relevant higher percentage of those patients as well. This represents a clinical dilemma with regard to the potential risk of underdosing with the use of the revised protocol compared with the frequent above-target troughs observed using the original protocol. Although not specifically evaluated in our study, certain baseline patient conditions that may predict more rapid elimination of vancomycin and a low trough (<10 ÎĽg/mL) when using the revised protocol include younger age, an estimated CLcr of >150 mL/min, and a SCr of <0.8 mg/dL that cannot be explained by age or inactivity due to paralysis or critical illness.
Nephrotoxicity occurred in only 2.9% of cases included in our study. That low rate of nephrotoxicity is consistent with the rate of 5–7% described in a published study involving the relatively pure formulations of vancomycin used in clinical practice today but lower than the nephrotoxicity rate reported in a more recently published study. In a retrospective review of vancomycin use, Lodise et al. observed a nephrotoxicity rate of 22% in association with vancomycin troughs of >9.9 μg/mL during the initial 96 hours of treatment; of note, 26% of patients with a weight of >101 kg developed nephrotoxicity, and weights exceeding that threshold were associated with an increased risk of nephrotoxicity. We did not observe that phenomenon, and we also did not observe the findings of Lodise et al. that indicated an elevated nephrotoxicity rate (34.6%) in patients who received ≥4 g of vancomycin daily. In none of the 28 cases in our study involving total daily vancomycin doses of ≥4 g did the patient develop nephrotoxicity.
There are multiple possible explanations for the low rate of nephrotoxicity in our study cohort. First, vancomycin levels in the evaluated cases were not persistently elevated. In those cases in which an elevated trough was noted, the majority of patients either received a reduced dose or vancomycin was discontinued. Second, cases in which patients had a baseline CLcr of <60 mL/min were excluded from the study. Third, 51% of vancomycin use was indicated for skin and soft-tissue infections, only 22% of patients required an ICU stay, and the mean duration of vancomycin therapy was only 6.3 days; multiple analyses have shown that highly intensive, prolonged vancomycin therapy is associated with an increased risk of nephrotoxicity.
There were a number of limitations to our study. First, due to the retrospective nature of the investigation, the timing of blood collection for the first vancomycin trough in relation to the first vancomycin dose was not standardized. To account for this limitation, we evaluated the time at which samples were collected in relation to the predicted time to reach steady-state values (Table 2) and found no differences that would confound the outcomes of this study. Second, we did not evaluate the clinical outcomes in the evaluated cases due to the heterogeneity of the indications for and duration of vancomycin therapy. Third, our study held the potential for selection bias with regard to vancomycin dosing. Because the use of neither of the evaluated vancomycin dosing protocols nor any standardized dosing regimen was required in our institution, the decision to include patients in the original-protocol group who received maintenance dosages of 30–45 mg/kg/day after the protocol revision date might have been a confounding factor; however, as patient demographics and indications for vancomycin use were similar before and after the revision date, such confounding was unlikely. Finally, the two study groups were not evenly matched with respect to baseline weight; the revised-protocol group had a higher mean weight and a higher TBW:IBW ratio than the original-protocol group. Instead of devaluing the performance of the revised protocol, however, this weight difference further promotes the protocol's ability to achieve prompt target troughs in an obese population. It is logical to assume that we would have found even higher rates of above-target troughs in the original-protocol group if the patients had been more obese.
Prospective studies are needed to validate this lower vancomycin dosing regimen in obese patients in terms of actual clinical outcomes.
Discussion
The results of the study described here demonstrate that obese patients may benefit from a vancomycin dosing strategy different from that generally recommended for adult patients with normal renal function. When obese patients in our institution received vancomycin maintenance doses of 15 mg/kg every 8–12 hours, consistent with our original dosing protocol, above-target trough levels were noted in 55% of patients overall and 79% of ICU patients. Moreover, the mean trough in patients with undesirably high values was 29.1 μg/mL (Table 3). The original dosing protocol was consistent with recommendations subsequently published in the ASHP–IDSA–SIDP consensus statement, which recommends the maintenance administration of vancomycin 15–20 mg/kg (based on TBW) every 8–12 hours for most patients. Due to the frequent occurrence of elevated troughs, our revised institutional protocol recommends an empiric decrease in the maintenance dosage of vancomycin to either 10 mg/kg every 12 hours or 15 mg/kg every 24 hours (based on TBW). These dosing regimens were chosen because modeling indicated that they more consistently resulted in steady-state trough vancomycin concentrations of 10–20 μg/mL, as well as an AUC/MIC of ≥400 when the MIC is ≤1 μg/mL. Of note, a more stringent target trough concentration of 15–20 μg/mL was not used to derive the revised dosing protocol or to be the target trough in this study because a dosing regimen that yields an AUC/MIC of ≥400 despite a trough of 10–15 μg/mL is clinically acceptable. The results of our comparison study indicated a higher frequency of target trough attainment (59% versus 36%, p = 0.006) and a lower prevalence of above-target troughs (18% versus 55%, p < 0.001) with the revised versus original protocol.
Guidance for vancomycin dosing in an obese population is limited. Historically, many clinicians have assumed that in obese people, vancomycin CL increases proportionally with TBW. This assumption is logically based on the increase in organ size, blood volume, and the distribution of vancomycin into the excess extracellular fluid that accompanies excess adipose tissue. Most available data support the use of TBW for empirical vancomycin dosing on the basis of the strong linear correlation of vancomycin CL and TBW reported in multiple studies (although a more recent analysis published by Leong and associates showed that vancomycin CL was most closely associated with ABW). Notably, much of this research has also demonstrated that vancomycin V increases in obese patients but not to the same extent as CL. Based on these assumptions, if one obese patient and one normal-weight patient were to receive a similar vancomycin dose (i.e., calculated by milligram per kilogram of TBW), k in the obese patient should be greater, thus leading to a shorter t½ and a lower vancomycin trough concentration. In that context, the results of our study were intriguing. Our original protocol recommended a vancomycin dosage of 15 mg per kilogram of TBW per dose for all patients regardless of weight, and we observed that most obese patients with normal renal function had above-target troughs (>20 μg/mL).
A plausible explanation for the high frequency of above-target troughs documented in our study is suggested by two related factors. First, the significant variability and relative increase in vancomycin V in obese patients are often underappreciated. Vance-Bryan et al. suggested that the vancomycin pharmacokinetic parameter most influenced by weight is not CL but V; their research found that ABW and the percentage over lean body weight were both independent and significant predictors of V. Also, Ducharme et al. described significant variability in the vancomycin V, with a range of 0.58–1.17 L/kg of IBW; those investigators specifically commented that regardless of sex or age, the average vancomycin V was increased by 37% (from 0.65 L/kg of IBW in normal-weight patients to 0.89 L/kg of IBW in obese patients). Second, variability in V is often not even considered, either in the literature or in routine clinical practice. For example, in their attempt to determine the relationship between vancomycin CL and vancomycin trough concentrations, Leong et al. used a fixed V. Additionally, similar to most other institutions, our institution uses a dosing protocol that only accounts for body weight and estimated vancomycin CL and not V when determining an empirical vancomycin dosing regimen. Therefore, the variability of vancomycin pharmacokinetics certainly warrants continued therapeutic drug monitoring and appropriate dosage adjustments, especially in obese patients.
The potential pitfalls of underappreciating the impact of a higher-than-expected V in obese patients are illustrated in Figure 1. According to our original institutional protocol, one obese and one normal-weight patient of the same sex who were similar in age and other characteristics and had the same estimated CLcr (e.g., 90 mL/min) would have both received a vancomycin regimen of 15 mg/kg of TBW every 8 or 12 hours. The data presented above support the assumption that vancomycin V is likely to be greater in the obese patient. Because both patients have the same CL, an elevated V in the obese patient would be accompanied by a reduced k and consequently an increase in the vancomycin trough level, as shown in Figure 1. One might argue that this hypothetical situation would not occur in clinical practice because obese patients have increased CL that would offset or even surpass the impact of the elevated V. However, in clinical settings, there may be some other reason for decreased renal function in the obese patient such that the impact of obesity on CL is offset. Because formal pharmacokinetic analyses were not performed in the analysis described here, we can only offer this scenario as a logical explanation for the high frequency of above-target vancomycin troughs in the revised-protocol cohort. However, this assumption is based on multiple previous analyses of postdistributive vancomycin levels in hospitalized obese patients that have been performed in our institutions; these analyses have repeatedly confirmed a larger-than-expected vancomycin V in the context of a CL that was close to the predicted value.
(Enlarge Image)
Figure 1.
Effect of an increased vancomycin volume of distribution (V) on the concentration–time profile after a given dose. Curve A represents a concentration–time profile resulting from predicted pharmacokinetic values. Curve B shows a concentration–time profile resulting from a higher-than-expected V. In the equation, k = elimination rate constant, CL = clearance, and Cminss = steady-state trough concentration. Adapted with permission from reference 7.
Despite the higher frequency of target vancomycin troughs among patients dosed according to the revised protocol, a below-target trough was noted in a clinically relevant higher percentage of those patients as well. This represents a clinical dilemma with regard to the potential risk of underdosing with the use of the revised protocol compared with the frequent above-target troughs observed using the original protocol. Although not specifically evaluated in our study, certain baseline patient conditions that may predict more rapid elimination of vancomycin and a low trough (<10 ÎĽg/mL) when using the revised protocol include younger age, an estimated CLcr of >150 mL/min, and a SCr of <0.8 mg/dL that cannot be explained by age or inactivity due to paralysis or critical illness.
Nephrotoxicity occurred in only 2.9% of cases included in our study. That low rate of nephrotoxicity is consistent with the rate of 5–7% described in a published study involving the relatively pure formulations of vancomycin used in clinical practice today but lower than the nephrotoxicity rate reported in a more recently published study. In a retrospective review of vancomycin use, Lodise et al. observed a nephrotoxicity rate of 22% in association with vancomycin troughs of >9.9 μg/mL during the initial 96 hours of treatment; of note, 26% of patients with a weight of >101 kg developed nephrotoxicity, and weights exceeding that threshold were associated with an increased risk of nephrotoxicity. We did not observe that phenomenon, and we also did not observe the findings of Lodise et al. that indicated an elevated nephrotoxicity rate (34.6%) in patients who received ≥4 g of vancomycin daily. In none of the 28 cases in our study involving total daily vancomycin doses of ≥4 g did the patient develop nephrotoxicity.
There are multiple possible explanations for the low rate of nephrotoxicity in our study cohort. First, vancomycin levels in the evaluated cases were not persistently elevated. In those cases in which an elevated trough was noted, the majority of patients either received a reduced dose or vancomycin was discontinued. Second, cases in which patients had a baseline CLcr of <60 mL/min were excluded from the study. Third, 51% of vancomycin use was indicated for skin and soft-tissue infections, only 22% of patients required an ICU stay, and the mean duration of vancomycin therapy was only 6.3 days; multiple analyses have shown that highly intensive, prolonged vancomycin therapy is associated with an increased risk of nephrotoxicity.
There were a number of limitations to our study. First, due to the retrospective nature of the investigation, the timing of blood collection for the first vancomycin trough in relation to the first vancomycin dose was not standardized. To account for this limitation, we evaluated the time at which samples were collected in relation to the predicted time to reach steady-state values (Table 2) and found no differences that would confound the outcomes of this study. Second, we did not evaluate the clinical outcomes in the evaluated cases due to the heterogeneity of the indications for and duration of vancomycin therapy. Third, our study held the potential for selection bias with regard to vancomycin dosing. Because the use of neither of the evaluated vancomycin dosing protocols nor any standardized dosing regimen was required in our institution, the decision to include patients in the original-protocol group who received maintenance dosages of 30–45 mg/kg/day after the protocol revision date might have been a confounding factor; however, as patient demographics and indications for vancomycin use were similar before and after the revision date, such confounding was unlikely. Finally, the two study groups were not evenly matched with respect to baseline weight; the revised-protocol group had a higher mean weight and a higher TBW:IBW ratio than the original-protocol group. Instead of devaluing the performance of the revised protocol, however, this weight difference further promotes the protocol's ability to achieve prompt target troughs in an obese population. It is logical to assume that we would have found even higher rates of above-target troughs in the original-protocol group if the patients had been more obese.
Prospective studies are needed to validate this lower vancomycin dosing regimen in obese patients in terms of actual clinical outcomes.