INTRODUCTION

More than one-third of adults in the United States are obese and consequently at a significantly increased risk for heart disease, stroke, and type 2 diabetes.¹ In addition to these health implications, the physiologic changes from obesity also impact pharmacokinetic and pharmacodynamic properties of drugs. These changes can impact both efficacy and toxicity, especially in antimicrobials such as vancomcyin.²

Vancomycin is a tricyclic glycopeptide antibiotic commonly used as therapy for infections caused by Gram-positive organisms, most notably methicillin-resistant Staphylococcus aureus (MRSA).³Published adult dosing recommendations for vancomycin in the general population are 15 to 20 mg/kg per dose every 8 to 24 hours (based upon total body weight [TBW] and estimated renal function).⁴However, such recommendations may be inadequate in obese patients due to increases in vancomycin clearance and volume of distribution.⁵In addition, when applied to obese patients, the large single doses resulting from such weight-based recommendations increase the risk of dose-related toxicities.⁵

Variability in both practice and lack of evidence related to optimal initial dosing strategies ncomycin exist.⁵ For example, dosing based on TBW achieves target steady-state trough concentrations more frequently then when based on ideal body weight (IBW).² In contrast, one study5 demonstrated that use of adjusted body weight (ABW) provided the best predictor to serum concentrations, and another ⁶ recommended using 45 to 65 mg/kg/day based on IBW. ⁵, ⁶In addition to weight-based dosing, published dosing nomograms have also been extensively evaluated. ^{7 8}-⁹Their efficacy in achieving initial goal trough concentrations(10-20 mcg/mL) has been shown to range from 40-60% on the initial regimen,but the majority excluded patients weighing more than 120 kg or limited the maximum single dose to 2 gms.^{7 8}-⁹ Studies analyzing appropriate vancomycin dosing and monitoring in obese patients have reported variable success rates. In one, approximately 60% of initial vancomycin steady-state concentrations were subtherapeutic(<10 mcg/mL), leading to increased risk of resistance and treatment failure.⁸ Another concluded that obese patients most often reached target trough concentrations when given 20-30 mg/kg/day based on TBW.⁹

There has yet to be a consensus or guideline recommendations for dosing and monitoring in obese patients. At Duke University Hospital, a validated empiric dosing nomogram for patients weighing 50-100 kg has been in place since 2010. In order to fulfill an increasing and unmet need, an empiric vancomycin dosing nomogram was developed at Duke Regional Hospital (DRH) in 2016 targeting patients weighing 100 to 160 kg (see 1885-642X-pharmpract-16-03-1204-suppl.pdfAppendix). While we hypothesized this nomogram would provide appropriate initial vancomycin dosing guidelines in this population, it had not been previously evaluated. The purpose of our study was to evaluate this newly-implemented vancomycin dosing nomogram in achieving goal steady-state trough concentrations for obese adult patients.

METHODS

The primary objective of this single-center, retrospective cohort study was to describe the percentage of obese patients receiving initial vancomycin doses consistent with nomogram recommendations achieving targeted initial steady-state serum vancomycin concentrations. The secondary objectives were to describe the primary endpoint in subgroups based on patient weight and estimated creatinine clearance (CrCl). We also sought to describe the percentage of patients maintaining a target steady-state trough concentration, on a consistent regimen, for one subsequent level following an initial target steady-state trough concentration to assess the rate of accumulation. Lastly, patients were evaluated for vancomycin-related adverse effects, including new-onset kidney injury and Red Man syndrome.

This single-center, retrospective cohort study was approved by the Duke University Health System Institutional Review Board and conducted at DRH, a 369-bed community hospital in Durham, NC. Patients >18 years-old, admitted to a general medicine or surgery unit from December 1, 2015 to February 1, 2017 were included. Subjects who weighed >100 kg and had a BMI of >30 kg/m2 who received at least 2 scheduled vancomycin doses following the appropriate loading dose (per nomogram recommendations) were included if at least one steady-state trough vancomycin concentration (defined as following at least the third dose of the regimen and drawn within 2 hours of the next sequential dose) was measured. Patients were excluded for any of the following: renal dysfunction (defined as an estimated CrCl <10 mL/min), unstable renal function (defined as a change in serum creatinine (SCr) of 0.5 mg/dL or 50% reduction in estimated CrCl between initial dose and time of subsequent trough measurement), moderate to severe liver dysfunction at baseline (defined as aspartate aminotransferase or alanine aminotransferase levels >two times the upper limit of normal (ULN), or a total bilirubin level >two times the ULN), ascites, burns (>20% total body surface area), within 30 days of solid organ or hematopoietic stem cell transplantation, had cystic fibrosis, were patients in the critical care unit, or were pregnant.

Patients were identified utilizing the Duke Enterprise Data Unified Content Explorer (DEDUCE). Separate admissions for the same patient were counted as individual cases. Data were collected using a Microsoft Access database and entry form. Patient demographics collected included gender, age, weight, height, BMI, and the presence of chronic kidney disease (CKD). Other data collected included vancomycin indication, vancomycin dosing regimens, and vancomycin serum trough concentrations, dates, and collection times. SCr and estimated CrCl at time of vancomycin initiation and trough concentration of maintenance regimen utilizing a modified Cockroft-Gault equation (removing weight and 72 from numerator and denominator, respectively).¹⁰ Of note, in patients >70 years old, a SCr below 1 mg/dL was rounded to 1 mg/dL to calculate CrCl. For initial loading doses, patients received 25 mg/kg TBW unless they had impaired renal function indicated by new-onset kidney injury or CKD Stage IV or worse. In this case, patients were loaded with 20 mg/kg TBW. However, we incorporated our institution’s policy of vancomycin dose capping at 2500 mg. For patients with therapeutic serum trough concentrations that were continued on the same regimen, SCr was collected again at the time of the next trough concentration. Lastly, presence of Red Man syndrome and new-onset kidney injury at the time of concentration collection (defined as an increase in SCr by 0.3 mg/dL or more within 48 hours, or an increase in SCr to 1.5 times baseline or more within the last 7 days, or urine output less than 0.5 mL/kg/h for 6 hours) was collected.¹¹The institutional nomogram was developed with the above in mind, utilizing traditional vancomycin pharmacokinetic calculations including the Matzke equation for the elimination rate constant. For patients receiving multiple courses of vancomycin during a single admission, only the first course was included in the study.¹²

RESULTS

Of 325 patients weighing over 100 kg and on vancomycin identified and screened, 85 (26.2%) met inclusion criteria. Patients were excluded for the following: doses were not consistent with nomogram recommendations (n=168), no trough concentration level (n=36), critical care unit status (n=28), BMI <30 kg/m2 (5), and weight <100 kg at time of vancomycin initiation (3). The study population was predominantly male with an average age of 60 years. Remaining subject demographics are summarized in Table 1. All subjects had an estimated CrCl > 30 mL/min and the mean CrCl was 81.3 mL/min. The majority of patients were in the 15-20 mcg/mL goal trough cohort and were receiving therapy for complicated skin and skin structure infections (SSTI).

Goal steady-state trough concentrations were reached in 42 patients (49.4%) with 27 (47.4%) in the 15-20 mcg/mL cohort and 15 (53.6%) in the 10-15 mcg/mL cohort. In the total population, 24.7% had subtherapeutic levels at steady state and 25.9% had supratherapeutic levels. There was also a similar distribution of subtherapeutic levels and supratherapeutic levels in each goal trough subgroup (Figure 1). Trough levels ranged from 6.1-30.9 mcg/mL. When this data was combined, 58 patients (68.2%) had levels that fell in the 10-20 mcg/mL range.

Figure 1. Number achieving trough concentrations based on target trough concentration goal. 

When divided into pre-specified subgroups based on goal trough concentrations, weight, and estimated CrCl (Table 2), the majority of patients fell into the 100-119 kg groups (n= 47, 55%). There were a limited number of patients >140 kg (n=13, 15%), and only 28 patients had an estimated CrCl <70 mL/min. 69% of the pre-specified subgroups containing at least one patient in the 15-20 mcg/mL goal cohort and 67% of the subgroups in the 10-15 mcg/mL cohort had mean trough concentrations at goal, respectively (Table 2). Notably, 16/21 (76%) of total patients with subtherapeutic trough concentrations had an estimated CrCl >70 mL/min. However, there were more patients in these subgroups and the majority still achieved goal trough concentrations (n=30, 52.6%). There was a noticeably higher rate of patients reaching initial supratherapeutic trough concentrations in the CrCl <70 mL/min subgroups compared to those with a CrCl >70 mL/min (35.7% vs. 19.3%).

Very few patients were continued on the same vancomycin regimen following the achievement of a target trough concentration long enough to check a second concentration (n=11, 26.2%). Of these 11 patients, 5 experienced accumulation to a supratherapeutic trough concentration on the subsequent level, with a mean (SD) time to next level of 2.9 (SD=1.2) days. However, 3 (60%) of these patients developed new-onset kidney injury between the first and second concentration drawn.

No patients had to have vancomycin discontinued due to adverse events. Five patients experienced new-onset kidney injury during treatment and one patient was reported to have Red Man syndrome which was noted to improve when the infusion was administered at a slower rate. No other drug-related adverse effects were reported.

DISCUSSION

The results of our study found that our nomogram achieved target trough concentrations nearly 50% of the time. Prior attempts to utilize nomograms to provide initial dosing recommendations for vancomycin in obese patients have been met with variable success. One protocol employed a 20 mg/kg loading dose followed by 10 mg/kg/dose (based on TBW) every 12-24 hours in morbidly obese adults (BMI >40).⁸This dose was chosen based on previous findings that demonstrated a high rate of supratherapeutic concentrations with higher doses.⁸With this decreased dose, initial goal trough concentrations were achieved in 35.4% of patients, while subtherapeutic troughs occurred in 56.3% and supratherapeutic troughs in only 8.3% of patients.⁸Another recent retrospective study concluded that obese (BMI 30-40) and morbidly obese (BMI >40) patients most often reached target trough concentrations when given 20-30 mg/kg/day based on TBW.⁹However, this study had limitations which included a high rate of subtherapeutic trough concentrations (48%) and no loading doses were given. ⁹

Compared to the aformentioned studies and another by Morrill et al, which utilized a similar dosing strategy and yielded 48% subtherapeutic initial trough levels, our study had a more even distribution of non-therapeutic trough concentrations.^{7 8}-⁹Approximately 25% of patients had subtherapeutic trough levels with no level being lower than 6 mcg/mL, while another 25% of patients had supratherapeutic levels with only one level being greater than 30 mcg/mL (30.9). While we had a slightly higher rate of new-onset kidney injury during therapy compared to the previous trials, all patients experiencing kidney injury were on concomitant nephrotoxic medications including piperacillin-tazobactam, thiazide diuretics, and intravenous acyclovir.⁸, ⁹,¹³

The results of this study fall within the range of results in previous studies evaluating vancomycin dosing nomograms, achieving goal steady-state trough concentrations nearly 50% of the time.⁷,^{14 15}-¹⁶Unlike the majority of previous studies analyzing vancomycin nomograms, this study only included obese patients weighing at least 100 kg with no maximum weight ^.4-7,14-16When looking at the limited previous literature on vancomycin dosing in obese patients, our nomogram appears to be safe and similarly effective. Notable studies analyzing vancomycin dosing in obese patients have utilized protocols or nomograms that have based dosing on simplified mg/kg calculations paired with estimated renal function for determining frequency.⁸,⁹,¹³Our nomogram was developed utilizing traditional pharmacokinetic calculations for each subgroup using TBW for volume of distribution calculations and normalized CrCl which ultimately leads to a lower estimation of drug clearance in these patients. Utilizing this method of dosing, we predicted that our patients would receive large enough doses without experiencing toxic levels as a result of too frequent dosing.

This was also the first study to our knowledge to collect data on vancomycin accumulation in the real-world obese patient population. While our data is limited to 11 patients who were continued on their original therapeutic regimen long enough to receive a second trough level, it does reveal a concern for drug accumulation in this population. Nearly half (45%) of these patients experienced a subsequent supratherapeutic level following an initial therapeutic trough concentration and no change in dosing regimen. It is important to note that 3 of these patients had significant increases in SCr levels near the time of the follow-up level. Further studies are needed in this area to assess vancomycin adjustments in these patients to avoid potentially toxic accumulation.

Our study was not without limitations. Though our nomogram was designed using common calculations utilized in clinical practice, there are potential limitations with the pharmacokinetics of using the standard Vd, Matzke equation, and SCr rounding in the obese population.¹² However, there is no current consensus on the best method. AUC-based monitoring has also shown promising data, but until more implementable evidence exists, many institutions will continue traditional vancomycin dosing.¹⁷With no active or historical comparator, we were only able to report descriptive statistics limiting ability to show any association with patient specific factors and vancomycin concentrations. We were also limited to a small sample size. Although over 300 patients were screened for inclusion, pharmacists were not required to utilize the nomogram during the evaluation period which led to many exclusions. We also excluded patients in the critical care unit per the institution’s pharmacokinetic policy which limits extrapolation to these patients. This limited sample size and utilization also inhibited our ability to truly evaluate the effectiveness of our nomogram in patients with poor CrCl and those weighing over 140 kg. Lastly, we did not evaluate clinical outcomes of the patients.

CONCLUSIONS

Overall achievement of initial steady-state vancomycin serum concentrations in our study of obese patients (approximately 50%) was consistent with the use of published vancomycin dosing nomograms. Notably, our study had an even distribution of non-therapeutic trough concentrations (25% subtherapeutic and 25% supratherapeutic). Our study also added evidence for the risk vancomycin accumulation in continued dosing in this patient population. Future plans should include identifying patient-specific factors associated with non-therapeutic trough levels in the obese patient population and developing accurate pharmacokinetic models for this population.