ORIGINAL ARTICLE
Therapeutic Drug Monitoring of Pentobarbital: Experience at an Academic Medical Center Robert M. Humble, MS, Alexandra Ehlers, BS, MT(ASCP), Brittany L. Pakalniskis, MD, Cory Morris, BS, MT(ASCP), Denny Drees, BS, MT(ASCP), Jeff Kulhavy, BS, MT(ASCP), and Matthew D. Krasowski, MD, PhD
Background: Pentobarbital is used for management of intractable seizures and for reducing elevated intracranial pressure. Dosing of pentobarbital can be aided by therapeutic drug monitoring (TDM). There is no commercially available automated assay for measurement of pentobarbital serum/plasma concentrations; consequently, chromatography-based assays are often used.
Methods: Pentobarbital TDM was studied over a 14-year period at an academic medical center. 154 patients (94 adult, 60 pediatric) were identified who had pentobarbital levels ordered at least once during a hospital encounter. Chart review included patient diagnosis, indication for pentobarbital therapy, recent or concomitant medication with other barbiturates, patient disposition, organ donation, pentobarbital dosing changes, and neurosurgical procedures. Pentobarbital serum/plasma concentrations were determined on an automated clinical chemistry platform with a laboratory-developed test adapted from a urine barbiturates immunoassay.
Results: Chart review showed therapeutic use of pentobarbital generally consistent with previously published literature. The most common errors observed involved confusion in barbiturate names (eg, mix-up of pentobarbital and phenobarbital in test ordering or in provider notes) that seemed to have minimal impact on TDM effectiveness, with pentobarbital serum/plasma concentrations generally within target ranges. The laboratory-developed pentobarbital immunoassay showed cross-reactivity with phenobarbital and butalbital that was eliminated by alkaline and heat pretreatment. The immunoassay was linear to 20 mcg/mL and correlated closely with gas chromatography–mass spectrometry measurements at a reference laboratory.
Conclusions: Pentobarbital TDM can be performed by immunoassay on an automated clinical chemistry platform, providing an alternative to chromatography-based methods. Confusion in barbiturate names is common, especially pentobarbital and phenobarbital. Received for publication February 3, 2015; accepted April 18, 2015. From the Department of Pathology, University of Iowa Hospitals and Clinics, Iowa City, Iowa. The authors declare no conflict of interest. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.drug-monitoring.com). Correspondence: Matthew D. Krasowski, MD, PhD, Department of Pathology, University of Iowa Hospitals and Clinics, C-671 GH, 200 Hawkins Drive, Iowa City, IA 52242 (e-mail: [email protected]). Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Ther Drug Monit Volume 37, Number 6, December 2015
Key Words: drug monitoring, pentobarbital, barbiturate assay crossreactivity, organ donation, therapeutic drug monitoring errors (Ther Drug Monit 2015;37:783–791)
INTRODUCTION Pentobarbital is a barbiturate used for management of intractable seizures in status epilepticus and as an adjunct for reducing elevated intracranial pressure (ICP) in traumatic brain injury and other circumstances.1–6 In these clinical settings, pentobarbital may be administered continuously to induce a comatose state (“pentobarbital coma”). There is debate over the clinical effectiveness of pentobarbital in managing elevated ICP, especially given that pentobarbital can cause systemic hypotension that may offset the benefit of reducing ICP.7 Also, in the high doses of medication needed for pentobarbital coma, there is risk of other adverse effects or death. Dosing of pentobarbital can be aided by therapeutic drug monitoring (TDM) of serum/plasma concentrations in addition to following other end points such as burst suppression in electroencephalogram (EEG) recordings or frequency of elevated ICP episodes.1,5,8,9 Pentobarbital TDM can theoretically help in a variety of situations. First, in patients where there is poor control of elevated ICP or seizures despite standard dosing, TDM can determine if serum/plasma levels are subtherapeutic, suggesting that dosing should be increased. Second, TDM can also ascertain if loading doses have led to serum/plasma concentrations within the target range. Third, pentobarbital levels can be used for pharmacokinetic predictions of drug elimination (“washout”), if pentobarbital therapy is being discontinued in favor of other approaches. Lastly, in situations when the patient is not going to survive, pentobarbital serum/plasma concentrations need to be below a minimum threshold (typically 5 mcg/mL) to declare brain death. Pentobarbital is one of the drugs recognized to confound the diagnosis of brain death.10–12 In cases where organ donation is a possibility, the turnaround time of pentobarbital drug concentrations can impact the timing of brain death declaration and thus when organ recovery can proceed.10 A variety of analytical methods have been developed for measuring pentobarbital concentrations in serum or plasma. Pentobarbital concentrations may be determined by chromatography-based methods (with or without mass
783
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Humble et al
spectrometry), such as gas chromatography (GC),13,14 GC– mass spectrometry (GC/MS),15 high-performance liquid chromatography (HPLC),16 and HPLC–mass spectrometry (HPLC/MS).17 These methods can measure pentobarbital concentrations accurately even in the presence of other barbiturates or structurally similar molecules but require specialized instrumentation and technical expertise that may not be available in some clinical laboratories. In addition, these methods may not be conducive to rapid turnaround time for results. Pentobarbital assays that can run on automated clinical chemistry platforms may be useful to clinical laboratories as a substitute for chromatography-based methods. Yet, to our knowledge, there has not been a commercially available automated assay specifically marketed for the measurement of pentobarbital (as opposed to assays for measurement of total barbiturates18) in serum or plasma. As an alternative approach, multiple groups have adapted barbiturates immunoassays (typically marketed for drug-of-abuse screening of urine specimens) into laboratory-developed tests for pentobarbital drug monitoring in serum/plasma, although the most recent published reports are over 2 decades ago for assays marketed for older instrument platforms.19–22 Barbiturates drug-of-abuse screening assays often use secobarbital or a mixture of barbiturates as the target molecule(s) for assay antibody detection.23 These assays generally cross-react with a range of barbiturates, including amobarbital, butalbital, pentobarbital, phenobarbital, and thiopental. Thus, measurement of pentobarbital by such assays could be confounded if the patient has received other barbiturates. In the setting of pentobarbital administration for patients in the intensive care unit (ICU), phenobarbital is generally the most likely barbiturate to be coadministered, given the common use of phenobarbital in managing seizures and the declining therapeutic and nonmedical use of other barbiturates.24 The long serum half-life of phenobarbital may result in cross-reactivity with barbiturates assays days or even weeks after drug discontinuation. To remove potential interference, pretreatment of serum/plasma specimens under heat and alkaline conditions can degrade phenobarbital without affecting pentobarbital concentrations.22 If the medication regimen for a patient is unknown to the laboratory, specimens can first be screened with a phenobarbital drug-monitoring immunoassay (which in general does not cross-react with pentobarbital).25 If phenobarbital is present, then the sample can be treated to remove phenobarbital; if phenobarbital is not present, then this pretreatment is unnecessary. Alternatively, laboratories may choose to treat all specimens as if they contain phenobarbital. In the study, we review the use of pentobarbital TDM at an academic medical center over a 14-year period. We focused on how TDM impacted dosing adjustments and other clinical decisions such as declaration of brain death. In addition, we determined errors in TDM such as confusion between pentobarbital and phenobarbital in ordering or interpretation of drug levels. For the purposes of our study, we defined errors in TDM of pentobarbital to include when laboratory or patient care areas ordered testing on the wrong barbiturate (ie, phenobarbital was ordered but the patient was receiving pentobarbital) and when a provider note in the
784
Ther Drug Monit Volume 37, Number 6, December 2015
medical record referenced the wrong barbiturate (ie, note referenced phenobarbital when the patient was receiving pentobarbital). Many of these errors may have little overall impact on clinical management but represent opportunities for more serious problems. We also report a laboratorydeveloped assay that runs on a currently marketed clinical chemistry platform. To our knowledge, although there have been many studies evaluating the clinical effectiveness of pentobarbital coma therapy, there has been little investigation into the application of pentobarbital TDM.
METHODS AND MATERIALS Clinical Details Institutional Details The University of Iowa Hospitals and Clinics (UIHC) is a state academic medical center that serves as a tertiary and quaternary care center. The medical center includes cardiovascular, medical, surgical-neurologic, pediatric, and neonatal ICUs, along with a Level One Trauma Center. Retrospective analysis was performed over the time frame from January 1, 1999, to November 11, 2013, using an institutional review board–approved protocol. Chart review included patient diagnosis, indication for pentobarbital therapy, pentobarbital drug levels, recent or concomitant medication with other barbiturates (eg, butalbital, phenobarbital, thiopental), phenobarbital drug levels, patient disposition, organ donation (if the patient did not survive encounter), pentobarbital dosing changes, and mention of pentobarbital drug levels in clinical notes. The UIHC moved to a new electronic medical record (EMR) in May 2009. Most, but not all, historic data back to 1996 was uploaded into this new EMR. In both adult and pediatric populations, protocols for use of pentobarbital included continuous EEG monitoring, with burst suppression as the primary end point for pentobarbital dosing. The approximate burst suppression target range (indicated for seizure control) for pentobarbital serum/ plasma concentration is 4–20 mcg/mL.26 The approximate isoelectric target range (indicated for coma for metabolic or traumatic reasons) for pentobarbital serum/plasma concentration is 21–45 mcg/mL.9
Laboratory Details Analytical Methods Before June 2013, pentobarbital levels were determined on an Abbott Diagnostics (North Chicago, IL) TDx analyzer using a previously described method.22 Since June 2013, pentobarbital levels were determined using a laboratorydeveloped assay on Roche Diagnostics (Indianapolis, IN) Cobas c501 or c502 analyzers. The procedure for the c502 analyzer is as follows. The Cobas ONLINE TDM Phenobarbital and ONLINE DAT Barbiturates Plus reagents were obtained from Roche Diagnostics. Solutions of 1 M NaOH, 1 M HCl, and HPLCgrade deionized water were prepared in-house. Plasma or serum samples were analyzed on the Cobas 8000 c502 module (Roche Diagnostics). Samples were initially tested Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
for the presence of phenobarbital using the ONLINE TDM Phenobarbital reagent. For samples with $2.4 mcg/mL of phenobarbital, a pretreatment under basic conditions was required before analysis for pentobarbital. To 100 mL of sample in a 10 · 75-mm glass test tube, 100 mL of 1 M NaOH was added and vortexed to mix. Samples were loosely capped and placed in a heating block at 958C for 1 hour. After 1 hour, any condensation on the cap was returned to the tube and samples were removed from the heat block. To each tube, 1.8 mL of HPLC-grade water was added followed by 100 mL of 1 M HCl. Tubes were vortexed and transferred to racks for pentobarbital analysis on the Cobas c502 using modified settings for the ONLINE DAT Barbiturates Plus reagent. For samples with ,2.4 mcg/mL of phenobarbital, no pretreatment was required for pentobarbital analysis. To 100 mL of sample in a 10 · 75-mm glass test tube, 2 mL of HPLC-grade water was added and vortexed to mix. Samples were transferred to racks for analysis on the Cobas c502. The Cobas e502 settings were as follows: assay type, 2-point end; reaction time/assay points, 10/40–65; primary wavelength 505 nm; second wavelength, cancel; absorbance limit, 32,000; reaction direction, increase; sample volume, 2.5 mL; R1 reagent volume, 57 mL; R2 reagent volume, 57 mL; R3 reagent volume, 57 mL; calibration type, RCM (Rodbard), 6point; calibrator sample volume standards 1–4, 2.5 mL; and calibrator sample volume standards 5–6, 10 mL.
Cross-Reactivity and Comparison Studies As discussed in the Introduction section, one challenge of using an immunoassay for pentobarbital TDM is the situation of a patient being administered one or more of the other barbiturates in addition to pentobarbital. The other barbiturate(s) can potentially cross-react with the immunoassay and influence the accuracy of pentobarbital concentration measurement. Independent of our chart review of pentobarbital TDM, we performed cross-reactivity studies of our pentobarbital immunoassay in the presence of other barbiturates as part of assay validation. Barbiturate standards were obtained from SigmaAldrich (St. Louis, MO; thiopental) or Cerilliant Corporation (Round Rock, TX; all other barbiturates). Cross-reactivity studies were performed with amobarbital (1–10 mcg/mL), butalbital (5–100 mcg/mL), methohexital (0.5–10 mcg/mL), phenobarbital (5–100 mcg/mL), secobarbital (1–10 mcg/ mL), and thiopental (5–100 mcg/mL), comparing results with and without heat/alkaline pretreatment. Concentrations tested covered the serum/plasma levels seen in pharmacokinetic or toxicologic studies. Pharmacokinetic studies of amobarbital using intravenous dosing have shown peak levels of less than 10 mcg/mL.27,28 Therapeutic levels of butalbital are 1–10 mcg/mL.29 Blood levels of butalbital in postmortem cases (not associated with butalbital overdose but as presumed therapeutic drug use) have shown concentrations up to 11 mcg/ mL.30 Pharmacokinetic studies of methohexital have shown peak levels of less than 3 mcg/mL.31 Similar concentrations have been seen in children receiving rectal methohexital.32 A pharmacokinetic study of rectally administered secobarbital in children have shown peak concentrations less than 3 mcg/ mL.33 Secobarbital toxicity is common when serum/plasma Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
TDM of Pentobarbital
concentrations exceed 6 mcg/mL; such concentrations may be seen in acute overdosage but would not be expected in controlled hospital settings.34 The plasma concentrations of thiopental associated with general anesthesia range from 11 to 69 mcg/mL.35,36 Comparison studies to chromatographic reference method were carried out using samples analyzed both by immunoassay (the UIHC) and by GC/MS method performed at a commercial reference laboratory (ARUP Laboratories, Salt Lake City, UT).37
RESULTS Clinical Details Characteristics of Patients for Whom Pentobarbital Drug Levels Were Ordered A total of 154 patients were identified who had pentobarbital drug levels ordered one or more times during a hospital encounter (all identified encounters were on inpatient units). This population was separated into adult (18 years or older, n = 94) and pediatric (younger than 18 years, n = 60) patients. The demographics are summarized in Table 1. There were a total of 7 patients (6 adult, 1 pediatric)
TABLE 1. Patient Demographics, Indications for Pentobarbital Therapy, and Clinical Outcomes for Patients on Whom Pentobarbital Levels Were Ordered for TDM*
No. patients, n (%) Male Female Median age, yrs Mean age 6 SD Indication for pentobarbital administration, n (%) Elevated intracranial pressure Seizure activity (primarily status epilepticus) Sedation No indication identifiable in chart Average total neurosurgical procedures per encounter† No. patients who did not survive hospital admission, n (%) Deceased patients who were organ donors, n (%)
Adult (18 Years and Older)
Pediatric (Younger Than 18 Years)
88 49 (55.7) 39 (44.3) 38.6 39.1 6 15.4
59 31 (52.5) 28 (47.5) 4.3 6.2 6 6.0
61 (69.3)
34 (57.6)
19 (21.6)
20 (33.9)
0 (0) 8 (9.1)
4 (6.8) 1 (1.7)
15.9
7.5
53 (60.2)
23 (39.0)
13 (14.8)
5 (8.5)
*Excluded are 7 patients who did not receive pentobarbital but had pentobarbital levels ordered erroneously. †Within each patient population, the average number of neurosurgical procedures undergone by patients during their encounter, determined by chart review.
785
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
Humble et al
who were not administered pentobarbital but had pentobarbital levels erroneously ordered. The mean number of pentobarbital levels per encounter was 3.8 and 4.4 in the adult and pediatric populations, respectively (Table 2). A distribution of the pentobarbital levels is summarized in histogram form in Figure 1. Among the adult patients who were administered pentobarbital and had TDM performed (n = 88; 356 total pentobarbital levels), the most common documented indications for pentobarbital were elevated ICP (n = 61) and seizure activity (n = 19). In some cases, there was no documented indication for pentobarbital administration in the medical record (n = 8), although these generally occurred in cases where the EMR data appeared incomplete (usually in cases before installation of new EMR in 2009 and may represent cases of historic data not uploaded into the new EMR). In the adult population, 65.9% of charts (n = 58) featured complete medication administration record (MAR) data. Of those with complete MAR data, 63.8% of patients (n = 37) received their pentobarbital drip continuously until pentobarbital was discontinued. Four adult encounters featured complete MAR, but the patient did not receive pentobarbital. Seventeen adult charts (29.3%, n = 17) featured discontinuation of the pentobarbital drip that was later restarted during the encounter. Some patients had pentobarbital stopped and restarted more than once; we observed complete hold of the pentobarbital drip and later reinitiation 28 times between these 17 patients. The time between discontinuation and reinitiation of the pentobarbital drip varied between occurrences from 30 minutes to nearly 28 days. The most common reason to reinitiate the pentobarbital drip was presence of seizure activity (53.6%). For 17.9% of the time pentobarbital was restarted because of poor ICP control. For 10.7% of the time,
TABLE 2. Application of Pentobarbital Drug Monitoring
Barbiturate(s) administered during admission, n (%) Pentobarbital alone Pentobarbital and phenobarbital Pentobarbital and thiopental Pentobarbital, phenobarbital, and thiopental Mean pentobarbital levels per encounter Pentobarbital levels measured to undetectable, n (%) Yes No Death before undetectable levels Only measured levels were undetectable
786
Adult (18 Years and Older)
Pediatric (Younger Than 18 Years)
78 (88.6) 8 (9.1)
19 (32.2) 26 (44.1)
2 (2.3)
12 (2.3)
0 (0)
2 (3.4)
3.8 6 3.2
4.4 6 4.0
48 (54.5) 10 (11.4) 25 (28.4)
23 (39.0) 18 (30.5) 15 (25.4)
5 (5.7)
3 (5.1)
FIGURE 1. Distribution of pentobarbital serum/plasma concentrations in the (A) pediatric and (B) adult patient populations. The data is from a total of 621 total measurements (256 pediatric and 356 adult). The far left bar in each graph captures all levels of 5 mcg/mL and less.
pentobarbital was stopped and later restarted to transport the patient. In 17.9% of instances, there did not mention a reason for the pentobarbital drip to be stopped and restarted. In the pediatric population who received pentobarbital and had TDM performed (n = 59; 265 total pentobarbital levels performed), the indications for pentobarbital therapy were elevated ICP (n = 34), seizure activity (n = 20), and sedation (n = 4). The indication for pentobarbital could not be identified in 1 case; as with similar cases in adults described above, this occurred before implementation of new EMR in Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
2009. In contrast to the adult population, pentobarbital was used for sedation in the absence of elevated ICP or seizure activity in 4 cases. In the pediatric population that had pentobarbital levels performed, 54.2% of charts featured complete MAR data (n = 32). Of those with complete MAR data, 78.1% of patients received their pentobarbital drip continuously until pentobarbital was discontinued (n = 25). In 3.1% of pediatric encounters, the chart featured complete MAR, but the patient did not receive pentobarbital. Six pediatric charts featured discontinuation of the pentobarbital drip that was later restarted during the encounter. Some patients had pentobarbital stopped and restarted more than once; we observed complete hold of the pentobarbital drip and later reinitiation 13 times between these 6 patients. The time between discontinuation and reinitiation of the pentobarbital drip varied between occurrences from 85 minutes to nearly 14 days. The most common reason to reinitiate the pentobarbital drip was presence of seizure activity (53.9%). In the pediatric population, there were no instances of restarting a held drip because of poor control of ICP. For 15.4% of the time, pentobarbital was stopped and later restarted to transport the patient. In 30.8% of instances, the chart did not mention a reason for the pentobarbital drip to be stopped and restarted. In the adult population, 39.8% of patients who received pentobarbital administration survived the inpatient encounter and were discharged (n = 35; Table 1). Of those deceased (n = 53), 24.5% became organ donors (n = 13). Among pediatric patients, 61.0% who received pentobarbital administration survived the inpatient encounter and were discharged (n = 36). Of those deceased, 21.7% became organ donors (n = 5).
Use of Pentobarbital TDM in Dosing Decisions In our study, although pentobarbital dosing changes were common, we found very few examples where a pentobarbital serum/plasma was cited in a physician note as the motivation to adjust pentobarbital dosage. This likely reflects not only primary involvement by pharmacists, but also the use of EEG burst suppression and other clinical end points in guiding pentobarbital dosing. There were just 2 instances of providers explicitly referencing and holding the pentobarbital drip in response to a serum/plasma pentobarbital level in the adult population. There was no explicit mention of the plasma/serum pentobarbital levels in the charts regarding decisions to hold or restart the drip in the pediatric population. Pentobarbital levels were followed until undetectable (,5.0 mcg/mL) in 54.5% (n = 48) of adult and 39.0% (n = 23) in pediatric patients (Table 2). In both adult and pediatric populations, death occurred before undetectable pentobarbital concentrations in approximately 25% of patients. There were only 30 serum/plasma concentrations that exceeded 50 mcg/mL (16 levels in 10 pediatric patients; 14 levels in 8 adult patients). These instances represented only 4.9% of the 612 pentobarbital levels performed (Fig. 1). Pentobarbital dosing was reduced in 8 of 10 pediatric patients with pentobarbital serum/plasma concentrations exceeding 50 mcg/mL. One pediatric patient died with a pentobarbital plasma concentration of 52.2 mcg/mL. This was a 16-year-old boy with intractable seizures despite Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
TDM of Pentobarbital
pentobarbital and multiple other antiepileptic medications. Pentobarbital dosing was reduced in 6 of 8 adult patients with pentobarbital serum/plasma concentrations exceeding 50 mcg/mL, while 2 patients died. Both adult patients who did not survive died of complications of severe head trauma from motor vehicle accident. Pentobarbital was used in both cases to manage elevated ICP.
Coadministration of Pentobarbital With Other Barbiturates Table 2 summarizes how often pentobarbital was used alone and together with one or more barbiturate(s) in the same inpatient encounter. As noted in the Methods section, all specimens to be tested for pentobarbital are first analyzed for phenobarbital as per our laboratory procedure. If phenobarbital is not detected, turnaround time is roughly 2 hours; if phenobarbital is detected, the specimen must be heated in an alkaline solution, increasing the turnaround time by approximately 1 hour. In the study, coadministration of pentobarbital and phenobarbital occurred in 9.1% (n = 8) of adult cases and 47.5% of pediatric cases (n = 28). Of these 28 pediatric cases, 2 additionally had thiopental administration as well; however, no cases involving thiopental were seen after 2010. There were 2 cases where pentobarbital levels were ordered in patients receiving other barbiturates but not pentobarbital. Both occurred in adult patients: one was administered phenobarbital, and the other was administered phenobarbital and thiopental. Among patients not receiving pentobarbital who had pentobarbital levels ordered, only 2 patients had apparent pentobarbital concentrations exceeding 5.0 mcg/mL. Both cases involved suicidal overdose in adults. One of the patients tested positive for barbiturates in their urine and had a history of polysubstance abuse. This patient had an apparent serum pentobarbital concentration of 6.4 mcg/mL. Confirmation testing was not performed for barbiturate identification. The second patient overdosed on an unknown amount of medication containing butalbital. This patient had an apparent serum pentobarbital concentration of 9.1 mcg/mL. Neither of these cases would have had heat and alkaline treatment of the specimen given that no phenobarbital was detected. As noted earlier, several of our patients also received thiopental. Two adult patients received thiopental at the UIHC in the ICUs; both had issues with elevated ICP. Fourteen pediatric patients received thiopental at the UIHC; all had issues with elevated ICP. Most pediatric patients received thiopental in the pediatric intensive care unit (n = 10). One adult patient received thiopental at an outside hospital before transfer to the UIHC. No thiopental was administered in the operating room in both the adult and pediatric populations.
Errors in TDM of Pentobarbital As summarized in Table 3, chart review revealed errors mainly involving confusion in barbiturate names. There were instances where a pentobarbital level was ordered by laboratory staff in error and later cancelled (n = 8 in adults, n = 8 in pediatrics). There were instances of a pentobarbital level ordered in error by the patient care area and similarly cancelled (n = 5 in adults, n = 4 pediatrics). In addition, there
787
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
Humble et al
TABLE 3. Errors in Pentobarbital Drug Monitoring and Medical Record Documentation Adult (18 Years and Older) Order error by laboratory staff with subsequent cancellation of order Order error by patient care area with subsequent cancellation of order Provider note in medical record references the wrong barbiturate Order error not otherwise specified (without cancellation of order)*
Pediatric (Younger Than 18 Years)
8
8
5
4
212
43
29
7
*Includes examples such as phenobarbital levels performed in patients receiving pentobarbital but not phenobarbital.
were occurrences of ordered phenobarbital levels when a patient was not receiving phenobarbital but was receiving pentobarbital; these were not cancelled (n = 29 adults, n = 7 pediatrics). Finally, there were many instances of provider notes referencing the wrong barbiturate in patients receiving pentobarbital administration (eg, phenobarbital referenced when the patient was receiving pentobarbital; n = 212 occurrences between 42 adult patients and n = 43 occurrences between 13 pediatric patients).
Laboratory Details Characteristics of the Pentobarbital Immunoassay The Roche Cobas ONLINE DAT Barbiturates Plus reagent was adapted for determining serum/plasma concentrations of pentobarbital. The assay was linear from 2 to 20 mcg/mL (Fig. 2); at higher pentobarbital concentrations, the immunoassay underestimated the true concentration. Cross-reactivity studies were performed with amobarbital, butalbital, methohexital, phenobarbital, secobarbital, and thiopental, comparing results with and without heat/alkaline pretreatment (Table 4). Amobarbital (1–10 mcg/mL) and methohexital (0.5–10 mcg/mL) showed no cross-reactivity. Without heat/alkaline pretreatment, butalbital showed crossreactivity at 10 mcg/mL and above. Heat and alkaline pretreatment removed cross-reactivity except at 100 mcg/mL, a very high concentration well above the therapeutic range.29 Without heat/alkaline pretreatment, phenobarbital showed cross-reactivity at 25 mcg/mL and above; heat and alkaline pretreatment eliminated the cross-reactivity at concentrations up to 100 mcg/mL. Without heat/alkaline pretreatment, secobarbital showed cross-reactivity at 5 and 10 mcg/mL. After heat and alkaline pretreatment, cross-reactivity was still seen at 10 mcg/mL, a concentration above that seen with therapeutic use of secobarbital and in the range associated with toxicity from acute overdosage.33,34 Thiopental showed
788
FIGURE 2. Characteristics of the pentobarbital immunoassay. Linearity studies of the immunoassay show linear relationship of measured versus spiked concentration from 2 to approximately 20 mcg/mL. At spiked concentrations higher than 20 mcg/mL, the immunoassay underestimates the true concentration. The dashed line indicates a 1:1 relationship. The solid line is from linear regression.
cross-reactivity only at the very highest concentration tested (100 mcg/mL). This is above the concentrations associated with general anesthesia.35,36 Comparison studies of the immunoassay with the older TDx laboratory-developed test are shown in the supplemental data (see Figure S1, Supplemental Digital Content 1, http://links.lww.com/TDM/A116). A comparison was also performed with a reference GC/MS method, performed by a commercial reference laboratory (see Figure S2, Supplemental Digital Content 1, http://links.lww.com/ TDM/A116). These studies included analysis with and without heat/alkaline pretreatment before immunoassay measurement and also a separate immunoassay measurement for phenobarbital [5 specimens shown to contain phenobarbital and pentobarbital are highlighted by circles in Supplemental Digital Content 1 (see Figure S2, http://links.lww. com/TDM/A116)]. Two specimens containing phenobarbital (69.3 mcg/mL phenobarbital in a specimen with 100 mcg/mL pentobarbital by GC/MS and 34.5 mcg/mL phenobarbital in a specimen with 29 mcg/mL pentobarbital by GC/MS) showed marked differences when comparing with or without heat/alkaline pretreatment. Three other specimens contained phenobarbital at concentrations ranging from 15.7 to 24.1 mg/mL and showed less dramatic differences with and without heat/alkaline pretreatment (see Figure S2, Supplemental Digital Content 1, http://links.lww. com/TDM/A116).
DISCUSSION In our retrospective review at an academic medical center, pentobarbital was used primarily for management of elevated ICP or seizures in both the adult and pediatric Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
TDM of Pentobarbital
populations. This is consistent with published literature on therapeutic use of pentobarbital.1–6 Consistent with the severity of the clinical conditions (eg, head trauma, status epilepticus) that often lead to pentobarbital administration, 59.6% of adult and 40.0% of pediatric patients did not survive the hospital admission. In this study, pentobarbital serum/plasma concentrations were generally within the approximate target range of 21–45 mcg/mL cited in the literature for pentobarbital coma.9 Only 4.9% of levels exceeded 50 mcg/mL. Levels at this higher range either led to dosage reductions or were found in patients who were administered pentobarbital as one of multiple therapies for severe seizures or elevated ICP due to head trauma. Preanalytical and postanalytical errors have been identified as a major issue in the clinical practice of TDM of a variety of drugs.38–41 Our study revealed that confusion in barbiturate names (eg, pentobarbital versus phenobarbital) is a challenge in pentobarbital therapy and drug monitoring, with mistakes in ordering of correct tests and also in documentation in the medical record. To our knowledge, errors in barbiturate drug level testing and provider notes did not result in incorrect drug administration. It is possible the issue of referencing the wrong barbiturate in provider notes is due to problems with dictation in some cases. But overall, a contributing factor may be that pentobarbital is not
a commonly used medication and also has a similar name to the more widely used phenobarbital. Pentobarbital is one of the drugs recognized to confound declaration of brain death.10–12 Thus, pentobarbital serum/plasma concentrations need to be below a minimum threshold (typically 5 mcg/mL) to declare brain death. In this study, we observed patients receiving pentobarbital with pentobarbital serum/plasma concentrations measured to undetectable levels after pentobarbital discontinuation in a significant numbers of patients. A quick turnaround time for pentobarbital levels allows for more rapid declaration of brain death in cases where organ donation is a possibility and thus the potential speed of organ recovery. This study presents a laboratory-developed pentobarbital immunoassay suitable for TDM that can be run on an automated clinical chemistry analyzer. Similar to previously reported laboratory-developed immunoassays on older instrument platforms, the current immunoassay does cross-react with some other barbiturates.38–41 This interference can be nullified by heat and alkaline pretreatment. If a laboratory was to treat all specimens as if they contained phenobarbital (by heat/alkaline pretreatment), turnaround time would increase by approximately 1 hour. This study showed that another barbiturate (most often phenobarbital and less commonly thiopental) was used together with pentobarbital in 8.5% of adult cases and 46.6% of pediatric cases (n = 28). In both the adult and pediatric populations, we see no evidence that thiopental interfered with the pentobarbital immunoassay. This is consistent with our findings in the cross-reactivity studies. The assay was linear from 2 to 20 mcg/mL. Given that pentobarbital serum/plasma concentrations may exceed 20 mcg/mL (279 of 621, 44.9%, levels in our retrospective study), dilutions will be required in some cases. However, this immunoassay may provide more rapid turnaround time and be more cost-effective than a dedicated chromatography-based assay. Further, the availability of an immunoassay may allow laboratories to perform pentobarbital levels in-house instead of referring to external reference laboratory. In our study, there is no evidence to suggest that background interference or cross-reaction between barbiturates yielded falsely elevated results in patients administered pentobarbital. Over the course of our retrospective analysis, we only detected 2 patients suspected of having crossreactivity from another barbiturate resulting in apparent pentobarbital concentrations exceeding 5.0 mcg/mL. Neither of these cases received pentobarbital. One of these cases involved an overdose of butalbital. The other was of unknown drug overdose associated with positive screen for urine barbiturates; butalbital would be one possibility as a crossreactant in this case. Given that patients administered pentobarbital tend to be in inpatient settings (often ICUs), administration of additional barbiturates other than phenobarbital or thiopental would be very unlikely. Butalbital was not used as a hospital medication for any patient in our study receiving pentobarbital. Butalbital cross-reactivity is practically eliminated by heat and alkaline pretreatment. Thus, performing this pretreatment for all specimens before pentobarbital analysis would greatly reduce the possibility of butalbital interference even in butalbital overdose cases.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
789
TABLE 4. Cross-Reactivity Studies of the Pentobarbital Immunoassay
Amobarbital
Butalbital
Methohexital
Phenobarbital
Secobarbital
Thiopental
Apparent Pentobarbital Concentration by Immunoassay (mcg/mL)*
Spiked Concentration (mcg/mL)
No Alkaline and Heat Treatment
Alkaline and Heat Treated
1 5 10 5 10 30 50 100 0.5 2.5 5 10 5 25 100 1 5 10 5 10 25 50 100
,5.0 ,5.0 ,5.0 ,5.0 7.3 22.3 30.8 69.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 12.1 23.5 ,5.0 6.6 17.3 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0
,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 11.5 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 13.0 ,5.0 ,5.0 ,5.0 ,5.0 5.0
*Values in bold type were 5.0 mcg/mL or greater.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
Humble et al
All specimens to be analyzed for pentobarbital are screened for phenobarbital as part of the pentobarbital protocol. The cross-reactivity studies demonstrate that thiopental and phenobarbital do not interfere with the pentobarbital immunoassay after heat/alkaline pretreatment. In addition, our laboratory-developed immunoassay runs on an automated clinical chemistry platform and requires significantly less technical staff labor time and expertise than chromatographybased assays. The availability of an immunoassay also more readily allows for assay runs across all times of the day. In contrast, for hospital-based laboratories, chromatographybased assays may be especially difficult to support during times of lower staffing, such as evenings, nights, weekends, and holidays. If a clinical laboratory seeks to adapt the pentobarbital immunoassay, it is essential to perform thorough validation studies including cross-reactivity analysis. Future studies could investigate the impact of turnaround time of pentobarbital drug level on organ donation rates. Our study supplements the existing literature in several ways. Observation of pentobarbital TDM over a 14-year period at an academic medical center allows us to observe long-term trends in use of pentobarbital TDM in both adult and pediatric populations. We shed light on common preanalytical and postanalytical errors that can serve as a means to improve on existing patient care practices and use of EMRs. Our study presents an alternative methodology to measure pentobarbital concentrations in serum/plasma that accounts for the presence of other barbiturates, has a faster turnaround time than other methodologies, and reduces the labor burden of the clinical laboratory by use of an automated clinical chemistry platform. There are some shortcomings of our study. The EMR data was incomplete for some patients, particularly for cases before the implementation of a new EMR system in 2009. In addition, our study involved data from a single institution; other institutions may have different protocols in place for TDM of pentobarbital. Nevertheless, the results provide the largest set of data published to date on the clinical practice of pentobarbital TDM. ACKNOWLEDGMENTS The authors are grateful for helpful discussions with Douglas Morgan and Ellen Nickel (Department of Pharmaceutical Care, University of Iowa Hospitals and Clinics) on the clinical use of pentobarbital in the pediatric and adult populations, respectively. The authors also thank Jon Maakestad (Department of Pathology, University of Iowa Hospitals and Clinics) for assistance in preparing barbiturate concentrations for cross-reactivity testing. REFERENCES 1. Mellion SA, Bennett KS, Ellsworth GL, et al. High-dose barbiturates for refractory intracranial hypertension in children with severe traumatic brain injury. Pediatr Crit Care Med. 2013;14:239–247. 2. Mansour N, deSouza RM, Sikorski C, et al. Role of barbiturate coma in the management of focally induced, severe cerebral edema in children. J Neurosurg Pediatr. 2013;12:37–43. 3. Marshall GT, James RF, Landman MP, et al. Pentobarbital coma for refractory intra-cranial hypertension after severe traumatic brain injury:
790
4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
14. 15.
16. 17.
18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.
mortality predictions and one-year outcomes in 55 patients. J Trauma. 2010;69:275–283. Abend NS, Dlugos DJ. Treatment of refractory status epilepticus: literature review and a proposed protocol. Pediatr Neurol. 2008;38:377–390. Censullo JL, Sebastian S. Pentobarbital sodium coma for refractory intracranial hypertension. J Neurosci Nurs. 2003;35:252–262. Kinoshita H, Nakagawa E, Iwasaki Y, et al. Pentobarbital therapy for status epilepticus in children: timing of tapering. Pediatr Neurol. 1995; 13:164–168. Roberts I, Sydenham E. Barbiturates for acute traumatic brain injury. Cochrane Database Syst Rev. 2012;12:CD000033. Bader MK, Arbour R, Palmer S. Refractory increased intracranial pressure in severe traumatic brain injury: barbiturate coma and bispectral index monitoring. AACN Clin Issues. 2005;16:526–541. Heinemeyer G, Roots I, Dennhardt R. Monitoring of pentobarbital plasma levels in critical care patients suffering from increased intracranial pressure. Ther Drug Monit. 1986;8:145–150. Meinitzer A, Kalcher K, Gartner G, et al. Drugs and brain death diagnostics: determination of drugs capable of inducing EEG zero line. Clin Chem Lab Med. 2008;46:1732–1738. Wijdicks EF. The diagnosis of brain death. N Engl J Med. 2001;344: 1215–1221. Wenck DJ. Drugs and brain death testing. Crit Care Med. 2000;28:601. Lehane DP, Menyharth P, Lum G, et al. Therapeutic drug monitoring: measurements of antiepileptic and barbiturate drug levels in blood by gas chromatography with nitrogen—selective detector. Ann Clin Lab Sci. 1976;6:404–410. Wallace JE, Hall LR, Harris SC. Determination of pentobarbital and certain other barbiturates by capillary gas-liquid chromatography. J Anal Toxicol. 1983;7:178–180. Johnson LL, Garg U. Quantitation of amobarbital, butalbital, pentobarbital, phenobarbital, and secobarbital in urine, serum, and plasma using gas chromatography-mass spectrometry (GC-MS). Methods Mol Biol. 2010;603:65–74. Kelner M, Bailey DN. Reversed-phase liquid-chromatographic simultaneous analysis for thiopental and pentobarbital in serum. Clin Chem. 1983;29:1097–1100. Spell JC, Srinivasan K, Stewart JT, et al. Supercritical fluid extraction and negative ion electrospray liquid chromatography tandem mass spectrometry analysis of phenobarbital, butalbital, pentobarbital and thiopental in human serum. Rapid Commun Mass Spectrom. 1998;12:890–894. Cannon RD, Wong SH, Gock SB, et al. Comparison of the serum barbiturate fluorescence polarization immunoassay by the COBAS INTEGRA to a GC/MS method. Ther Drug Monit. 1999;21:553–558. Wellington P, Rutkowski RB. Rapid estimation of plasma pentobarbital levels by an enzyme immunoassay for barbiturates (EMIT). Ther Drug Monit. 1982;4:319–322. Pape BE, Cary PL, Clay LC, et al. Pentobarbital quantitation using EMIT serum barbiturate assay reagents: application to monitoring of high-dose pentobarbital therapy. Ther Drug Monit. 1983;5:467–471. Li PK, Lee JT, Schreiber RM. Rapid quantification of pentobarbital in serum by fluorescence polarization immunoassay. Clin Chem. 1984;30:307–308. Earl R, Sobeski L, Timko D, et al. Pentobarbital quantification in the presence of phenobarbital by fluorescence polarization immunoassay. Clin Chem. 1991;37:1774–1777. Krasowski MD, Siam MG, Iyer M, et al. Chemoinformatic methods for predicting interference in drug of abuse/toxicology immunoassays. Clin Chem. 2009;55:1203–1213. Coupey SM. Barbiturates. Pediatr Rev. 1997;18:260–265. Krasowski MD, Siam MG, Iyer M, et al. Molecular similarity methods for predicting cross-reactivity with therapeutic drug monitoring immunoassays. Ther Drug Monit. 2009;31:337–344. Lowenstein DH, Aminoff MJ, Simon RP. Barbiturate anesthesia in the treatment of status epilepticus: clinical experience with 14 patients. Neurology. 1988;38:395–400. Mawer GE, Miller NE, Turnberg LA. Metabolism of amylobarbitone in patients with chronic liver disease. Br J Pharmacol. 1972;44:549–560. Balasubramaniam K, Lucas SB, Mawer GE, et al. The kinetics of amylobarbitone metabolism in healthy men and women. Br J Pharmacol. 1970;39:564–572. Moffat AC, ed. Clarke’s Isolation and Identification of Drugs. 2nd ed. London, United Kingdom: The Pharmaceutical Press; 1986.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
TDM of Pentobarbital
30. Lewis RJ, Johnson RD, Southern TL, et al. Distribution of butalbital in postmortem tissues and fluids from non-overdose cases. J Anal Toxicol. 2003;27:145–148. 31. Breimer DD. Pharmacokinetics of methohexitone following intravenous infusion in humans. Br J Anaesth. 1976;48:643–649. 32. Liu LM, Gaudreault P, Friedman PA, et al. Methohexital plasma concentrations in children following rectal administration. Anesthesiology. 1985;62:567–570. 33. Levine HL, Cohen ME, Duffner PK, et al. Rectal absorption and disposition of secobarbital in epileptic children. Pediatr Pharmacol (New York). 1982;2:33–38. 34. Greenblatt DJ, Allen MD, Harmatz JS, et al. Overdosage with pentobarbital and secobarbital: assessment of factors related to outcome. J Clin Pharmacol. 1979;19:758–768. 35. Loft S, Jensen V, Rorsgaard S. Influence of moderate alcohol intake on wakening plasma thiopental concentration. Acta Anaesthesiol Scand. 1983;27:266–269.
36. Stanski DR, Hudson RJ, Homer TD, et al. Pharmacodynamic modeling of thiopental anesthesia. J Pharmacokinet Biopharm. 1984;12:223–240. 37. Pentobarbital serum/plasma assay. Salt Lake City, UT: ARUP Laboratories. Available at: http://ltd.aruplab.com/Tests/Pub/2011549. Accessed April 13, 2015. 38. Kontny NE, Boos J, Wurthwein G, et al. Minimization of the preanalytical error in pharmacokinetic analyses and therapeutic drug monitoring: focus on IV drug administration. Ther Drug Monit. 2012;34:460–466. 39. Kontny NE, Hempel G, Boos J, et al. Minimization of the preanalytical error in plasma samples for pharmacokinetic analyses and therapeutic drug monitoring–using doxorubicin as an example. Ther Drug Monit. 2011;33:766–771. 40. Schiff GD, Klass D, Peterson J, et al. Linking laboratory and pharmacy: opportunities for reducing errors and improving care. Arch Intern Med. 2003;163:893–900. 41. Sieradzan R, Fuller AV. A multidisciplinary approach to enhance documentation of antibiotic serum sampling. Hosp Pharm. 1995;30:872, 874, 876–877.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
791
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Therapeutic Drug Monitoring of Pentobarbital: Experience at an Academic Medical Center Robert M. Humble, MS, Alexandra Ehlers, BS, MT(ASCP), Brittany L. Pakalniskis, MD, Cory Morris, BS, MT(ASCP), Denny Drees, BS, MT(ASCP), Jeff Kulhavy, BS, MT(ASCP), and Matthew D. Krasowski, MD, PhD
Background: Pentobarbital is used for management of intractable seizures and for reducing elevated intracranial pressure. Dosing of pentobarbital can be aided by therapeutic drug monitoring (TDM). There is no commercially available automated assay for measurement of pentobarbital serum/plasma concentrations; consequently, chromatography-based assays are often used.
Methods: Pentobarbital TDM was studied over a 14-year period at an academic medical center. 154 patients (94 adult, 60 pediatric) were identified who had pentobarbital levels ordered at least once during a hospital encounter. Chart review included patient diagnosis, indication for pentobarbital therapy, recent or concomitant medication with other barbiturates, patient disposition, organ donation, pentobarbital dosing changes, and neurosurgical procedures. Pentobarbital serum/plasma concentrations were determined on an automated clinical chemistry platform with a laboratory-developed test adapted from a urine barbiturates immunoassay.
Results: Chart review showed therapeutic use of pentobarbital generally consistent with previously published literature. The most common errors observed involved confusion in barbiturate names (eg, mix-up of pentobarbital and phenobarbital in test ordering or in provider notes) that seemed to have minimal impact on TDM effectiveness, with pentobarbital serum/plasma concentrations generally within target ranges. The laboratory-developed pentobarbital immunoassay showed cross-reactivity with phenobarbital and butalbital that was eliminated by alkaline and heat pretreatment. The immunoassay was linear to 20 mcg/mL and correlated closely with gas chromatography–mass spectrometry measurements at a reference laboratory.
Conclusions: Pentobarbital TDM can be performed by immunoassay on an automated clinical chemistry platform, providing an alternative to chromatography-based methods. Confusion in barbiturate names is common, especially pentobarbital and phenobarbital. Received for publication February 3, 2015; accepted April 18, 2015. From the Department of Pathology, University of Iowa Hospitals and Clinics, Iowa City, Iowa. The authors declare no conflict of interest. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (www.drug-monitoring.com). Correspondence: Matthew D. Krasowski, MD, PhD, Department of Pathology, University of Iowa Hospitals and Clinics, C-671 GH, 200 Hawkins Drive, Iowa City, IA 52242 (e-mail: [email protected]). Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Ther Drug Monit Volume 37, Number 6, December 2015
Key Words: drug monitoring, pentobarbital, barbiturate assay crossreactivity, organ donation, therapeutic drug monitoring errors (Ther Drug Monit 2015;37:783–791)
INTRODUCTION Pentobarbital is a barbiturate used for management of intractable seizures in status epilepticus and as an adjunct for reducing elevated intracranial pressure (ICP) in traumatic brain injury and other circumstances.1–6 In these clinical settings, pentobarbital may be administered continuously to induce a comatose state (“pentobarbital coma”). There is debate over the clinical effectiveness of pentobarbital in managing elevated ICP, especially given that pentobarbital can cause systemic hypotension that may offset the benefit of reducing ICP.7 Also, in the high doses of medication needed for pentobarbital coma, there is risk of other adverse effects or death. Dosing of pentobarbital can be aided by therapeutic drug monitoring (TDM) of serum/plasma concentrations in addition to following other end points such as burst suppression in electroencephalogram (EEG) recordings or frequency of elevated ICP episodes.1,5,8,9 Pentobarbital TDM can theoretically help in a variety of situations. First, in patients where there is poor control of elevated ICP or seizures despite standard dosing, TDM can determine if serum/plasma levels are subtherapeutic, suggesting that dosing should be increased. Second, TDM can also ascertain if loading doses have led to serum/plasma concentrations within the target range. Third, pentobarbital levels can be used for pharmacokinetic predictions of drug elimination (“washout”), if pentobarbital therapy is being discontinued in favor of other approaches. Lastly, in situations when the patient is not going to survive, pentobarbital serum/plasma concentrations need to be below a minimum threshold (typically 5 mcg/mL) to declare brain death. Pentobarbital is one of the drugs recognized to confound the diagnosis of brain death.10–12 In cases where organ donation is a possibility, the turnaround time of pentobarbital drug concentrations can impact the timing of brain death declaration and thus when organ recovery can proceed.10 A variety of analytical methods have been developed for measuring pentobarbital concentrations in serum or plasma. Pentobarbital concentrations may be determined by chromatography-based methods (with or without mass
783
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Humble et al
spectrometry), such as gas chromatography (GC),13,14 GC– mass spectrometry (GC/MS),15 high-performance liquid chromatography (HPLC),16 and HPLC–mass spectrometry (HPLC/MS).17 These methods can measure pentobarbital concentrations accurately even in the presence of other barbiturates or structurally similar molecules but require specialized instrumentation and technical expertise that may not be available in some clinical laboratories. In addition, these methods may not be conducive to rapid turnaround time for results. Pentobarbital assays that can run on automated clinical chemistry platforms may be useful to clinical laboratories as a substitute for chromatography-based methods. Yet, to our knowledge, there has not been a commercially available automated assay specifically marketed for the measurement of pentobarbital (as opposed to assays for measurement of total barbiturates18) in serum or plasma. As an alternative approach, multiple groups have adapted barbiturates immunoassays (typically marketed for drug-of-abuse screening of urine specimens) into laboratory-developed tests for pentobarbital drug monitoring in serum/plasma, although the most recent published reports are over 2 decades ago for assays marketed for older instrument platforms.19–22 Barbiturates drug-of-abuse screening assays often use secobarbital or a mixture of barbiturates as the target molecule(s) for assay antibody detection.23 These assays generally cross-react with a range of barbiturates, including amobarbital, butalbital, pentobarbital, phenobarbital, and thiopental. Thus, measurement of pentobarbital by such assays could be confounded if the patient has received other barbiturates. In the setting of pentobarbital administration for patients in the intensive care unit (ICU), phenobarbital is generally the most likely barbiturate to be coadministered, given the common use of phenobarbital in managing seizures and the declining therapeutic and nonmedical use of other barbiturates.24 The long serum half-life of phenobarbital may result in cross-reactivity with barbiturates assays days or even weeks after drug discontinuation. To remove potential interference, pretreatment of serum/plasma specimens under heat and alkaline conditions can degrade phenobarbital without affecting pentobarbital concentrations.22 If the medication regimen for a patient is unknown to the laboratory, specimens can first be screened with a phenobarbital drug-monitoring immunoassay (which in general does not cross-react with pentobarbital).25 If phenobarbital is present, then the sample can be treated to remove phenobarbital; if phenobarbital is not present, then this pretreatment is unnecessary. Alternatively, laboratories may choose to treat all specimens as if they contain phenobarbital. In the study, we review the use of pentobarbital TDM at an academic medical center over a 14-year period. We focused on how TDM impacted dosing adjustments and other clinical decisions such as declaration of brain death. In addition, we determined errors in TDM such as confusion between pentobarbital and phenobarbital in ordering or interpretation of drug levels. For the purposes of our study, we defined errors in TDM of pentobarbital to include when laboratory or patient care areas ordered testing on the wrong barbiturate (ie, phenobarbital was ordered but the patient was receiving pentobarbital) and when a provider note in the
784
Ther Drug Monit Volume 37, Number 6, December 2015
medical record referenced the wrong barbiturate (ie, note referenced phenobarbital when the patient was receiving pentobarbital). Many of these errors may have little overall impact on clinical management but represent opportunities for more serious problems. We also report a laboratorydeveloped assay that runs on a currently marketed clinical chemistry platform. To our knowledge, although there have been many studies evaluating the clinical effectiveness of pentobarbital coma therapy, there has been little investigation into the application of pentobarbital TDM.
METHODS AND MATERIALS Clinical Details Institutional Details The University of Iowa Hospitals and Clinics (UIHC) is a state academic medical center that serves as a tertiary and quaternary care center. The medical center includes cardiovascular, medical, surgical-neurologic, pediatric, and neonatal ICUs, along with a Level One Trauma Center. Retrospective analysis was performed over the time frame from January 1, 1999, to November 11, 2013, using an institutional review board–approved protocol. Chart review included patient diagnosis, indication for pentobarbital therapy, pentobarbital drug levels, recent or concomitant medication with other barbiturates (eg, butalbital, phenobarbital, thiopental), phenobarbital drug levels, patient disposition, organ donation (if the patient did not survive encounter), pentobarbital dosing changes, and mention of pentobarbital drug levels in clinical notes. The UIHC moved to a new electronic medical record (EMR) in May 2009. Most, but not all, historic data back to 1996 was uploaded into this new EMR. In both adult and pediatric populations, protocols for use of pentobarbital included continuous EEG monitoring, with burst suppression as the primary end point for pentobarbital dosing. The approximate burst suppression target range (indicated for seizure control) for pentobarbital serum/ plasma concentration is 4–20 mcg/mL.26 The approximate isoelectric target range (indicated for coma for metabolic or traumatic reasons) for pentobarbital serum/plasma concentration is 21–45 mcg/mL.9
Laboratory Details Analytical Methods Before June 2013, pentobarbital levels were determined on an Abbott Diagnostics (North Chicago, IL) TDx analyzer using a previously described method.22 Since June 2013, pentobarbital levels were determined using a laboratorydeveloped assay on Roche Diagnostics (Indianapolis, IN) Cobas c501 or c502 analyzers. The procedure for the c502 analyzer is as follows. The Cobas ONLINE TDM Phenobarbital and ONLINE DAT Barbiturates Plus reagents were obtained from Roche Diagnostics. Solutions of 1 M NaOH, 1 M HCl, and HPLCgrade deionized water were prepared in-house. Plasma or serum samples were analyzed on the Cobas 8000 c502 module (Roche Diagnostics). Samples were initially tested Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
for the presence of phenobarbital using the ONLINE TDM Phenobarbital reagent. For samples with $2.4 mcg/mL of phenobarbital, a pretreatment under basic conditions was required before analysis for pentobarbital. To 100 mL of sample in a 10 · 75-mm glass test tube, 100 mL of 1 M NaOH was added and vortexed to mix. Samples were loosely capped and placed in a heating block at 958C for 1 hour. After 1 hour, any condensation on the cap was returned to the tube and samples were removed from the heat block. To each tube, 1.8 mL of HPLC-grade water was added followed by 100 mL of 1 M HCl. Tubes were vortexed and transferred to racks for pentobarbital analysis on the Cobas c502 using modified settings for the ONLINE DAT Barbiturates Plus reagent. For samples with ,2.4 mcg/mL of phenobarbital, no pretreatment was required for pentobarbital analysis. To 100 mL of sample in a 10 · 75-mm glass test tube, 2 mL of HPLC-grade water was added and vortexed to mix. Samples were transferred to racks for analysis on the Cobas c502. The Cobas e502 settings were as follows: assay type, 2-point end; reaction time/assay points, 10/40–65; primary wavelength 505 nm; second wavelength, cancel; absorbance limit, 32,000; reaction direction, increase; sample volume, 2.5 mL; R1 reagent volume, 57 mL; R2 reagent volume, 57 mL; R3 reagent volume, 57 mL; calibration type, RCM (Rodbard), 6point; calibrator sample volume standards 1–4, 2.5 mL; and calibrator sample volume standards 5–6, 10 mL.
Cross-Reactivity and Comparison Studies As discussed in the Introduction section, one challenge of using an immunoassay for pentobarbital TDM is the situation of a patient being administered one or more of the other barbiturates in addition to pentobarbital. The other barbiturate(s) can potentially cross-react with the immunoassay and influence the accuracy of pentobarbital concentration measurement. Independent of our chart review of pentobarbital TDM, we performed cross-reactivity studies of our pentobarbital immunoassay in the presence of other barbiturates as part of assay validation. Barbiturate standards were obtained from SigmaAldrich (St. Louis, MO; thiopental) or Cerilliant Corporation (Round Rock, TX; all other barbiturates). Cross-reactivity studies were performed with amobarbital (1–10 mcg/mL), butalbital (5–100 mcg/mL), methohexital (0.5–10 mcg/mL), phenobarbital (5–100 mcg/mL), secobarbital (1–10 mcg/ mL), and thiopental (5–100 mcg/mL), comparing results with and without heat/alkaline pretreatment. Concentrations tested covered the serum/plasma levels seen in pharmacokinetic or toxicologic studies. Pharmacokinetic studies of amobarbital using intravenous dosing have shown peak levels of less than 10 mcg/mL.27,28 Therapeutic levels of butalbital are 1–10 mcg/mL.29 Blood levels of butalbital in postmortem cases (not associated with butalbital overdose but as presumed therapeutic drug use) have shown concentrations up to 11 mcg/ mL.30 Pharmacokinetic studies of methohexital have shown peak levels of less than 3 mcg/mL.31 Similar concentrations have been seen in children receiving rectal methohexital.32 A pharmacokinetic study of rectally administered secobarbital in children have shown peak concentrations less than 3 mcg/ mL.33 Secobarbital toxicity is common when serum/plasma Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
TDM of Pentobarbital
concentrations exceed 6 mcg/mL; such concentrations may be seen in acute overdosage but would not be expected in controlled hospital settings.34 The plasma concentrations of thiopental associated with general anesthesia range from 11 to 69 mcg/mL.35,36 Comparison studies to chromatographic reference method were carried out using samples analyzed both by immunoassay (the UIHC) and by GC/MS method performed at a commercial reference laboratory (ARUP Laboratories, Salt Lake City, UT).37
RESULTS Clinical Details Characteristics of Patients for Whom Pentobarbital Drug Levels Were Ordered A total of 154 patients were identified who had pentobarbital drug levels ordered one or more times during a hospital encounter (all identified encounters were on inpatient units). This population was separated into adult (18 years or older, n = 94) and pediatric (younger than 18 years, n = 60) patients. The demographics are summarized in Table 1. There were a total of 7 patients (6 adult, 1 pediatric)
TABLE 1. Patient Demographics, Indications for Pentobarbital Therapy, and Clinical Outcomes for Patients on Whom Pentobarbital Levels Were Ordered for TDM*
No. patients, n (%) Male Female Median age, yrs Mean age 6 SD Indication for pentobarbital administration, n (%) Elevated intracranial pressure Seizure activity (primarily status epilepticus) Sedation No indication identifiable in chart Average total neurosurgical procedures per encounter† No. patients who did not survive hospital admission, n (%) Deceased patients who were organ donors, n (%)
Adult (18 Years and Older)
Pediatric (Younger Than 18 Years)
88 49 (55.7) 39 (44.3) 38.6 39.1 6 15.4
59 31 (52.5) 28 (47.5) 4.3 6.2 6 6.0
61 (69.3)
34 (57.6)
19 (21.6)
20 (33.9)
0 (0) 8 (9.1)
4 (6.8) 1 (1.7)
15.9
7.5
53 (60.2)
23 (39.0)
13 (14.8)
5 (8.5)
*Excluded are 7 patients who did not receive pentobarbital but had pentobarbital levels ordered erroneously. †Within each patient population, the average number of neurosurgical procedures undergone by patients during their encounter, determined by chart review.
785
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
Humble et al
who were not administered pentobarbital but had pentobarbital levels erroneously ordered. The mean number of pentobarbital levels per encounter was 3.8 and 4.4 in the adult and pediatric populations, respectively (Table 2). A distribution of the pentobarbital levels is summarized in histogram form in Figure 1. Among the adult patients who were administered pentobarbital and had TDM performed (n = 88; 356 total pentobarbital levels), the most common documented indications for pentobarbital were elevated ICP (n = 61) and seizure activity (n = 19). In some cases, there was no documented indication for pentobarbital administration in the medical record (n = 8), although these generally occurred in cases where the EMR data appeared incomplete (usually in cases before installation of new EMR in 2009 and may represent cases of historic data not uploaded into the new EMR). In the adult population, 65.9% of charts (n = 58) featured complete medication administration record (MAR) data. Of those with complete MAR data, 63.8% of patients (n = 37) received their pentobarbital drip continuously until pentobarbital was discontinued. Four adult encounters featured complete MAR, but the patient did not receive pentobarbital. Seventeen adult charts (29.3%, n = 17) featured discontinuation of the pentobarbital drip that was later restarted during the encounter. Some patients had pentobarbital stopped and restarted more than once; we observed complete hold of the pentobarbital drip and later reinitiation 28 times between these 17 patients. The time between discontinuation and reinitiation of the pentobarbital drip varied between occurrences from 30 minutes to nearly 28 days. The most common reason to reinitiate the pentobarbital drip was presence of seizure activity (53.6%). For 17.9% of the time pentobarbital was restarted because of poor ICP control. For 10.7% of the time,
TABLE 2. Application of Pentobarbital Drug Monitoring
Barbiturate(s) administered during admission, n (%) Pentobarbital alone Pentobarbital and phenobarbital Pentobarbital and thiopental Pentobarbital, phenobarbital, and thiopental Mean pentobarbital levels per encounter Pentobarbital levels measured to undetectable, n (%) Yes No Death before undetectable levels Only measured levels were undetectable
786
Adult (18 Years and Older)
Pediatric (Younger Than 18 Years)
78 (88.6) 8 (9.1)
19 (32.2) 26 (44.1)
2 (2.3)
12 (2.3)
0 (0)
2 (3.4)
3.8 6 3.2
4.4 6 4.0
48 (54.5) 10 (11.4) 25 (28.4)
23 (39.0) 18 (30.5) 15 (25.4)
5 (5.7)
3 (5.1)
FIGURE 1. Distribution of pentobarbital serum/plasma concentrations in the (A) pediatric and (B) adult patient populations. The data is from a total of 621 total measurements (256 pediatric and 356 adult). The far left bar in each graph captures all levels of 5 mcg/mL and less.
pentobarbital was stopped and later restarted to transport the patient. In 17.9% of instances, there did not mention a reason for the pentobarbital drip to be stopped and restarted. In the pediatric population who received pentobarbital and had TDM performed (n = 59; 265 total pentobarbital levels performed), the indications for pentobarbital therapy were elevated ICP (n = 34), seizure activity (n = 20), and sedation (n = 4). The indication for pentobarbital could not be identified in 1 case; as with similar cases in adults described above, this occurred before implementation of new EMR in Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
2009. In contrast to the adult population, pentobarbital was used for sedation in the absence of elevated ICP or seizure activity in 4 cases. In the pediatric population that had pentobarbital levels performed, 54.2% of charts featured complete MAR data (n = 32). Of those with complete MAR data, 78.1% of patients received their pentobarbital drip continuously until pentobarbital was discontinued (n = 25). In 3.1% of pediatric encounters, the chart featured complete MAR, but the patient did not receive pentobarbital. Six pediatric charts featured discontinuation of the pentobarbital drip that was later restarted during the encounter. Some patients had pentobarbital stopped and restarted more than once; we observed complete hold of the pentobarbital drip and later reinitiation 13 times between these 6 patients. The time between discontinuation and reinitiation of the pentobarbital drip varied between occurrences from 85 minutes to nearly 14 days. The most common reason to reinitiate the pentobarbital drip was presence of seizure activity (53.9%). In the pediatric population, there were no instances of restarting a held drip because of poor control of ICP. For 15.4% of the time, pentobarbital was stopped and later restarted to transport the patient. In 30.8% of instances, the chart did not mention a reason for the pentobarbital drip to be stopped and restarted. In the adult population, 39.8% of patients who received pentobarbital administration survived the inpatient encounter and were discharged (n = 35; Table 1). Of those deceased (n = 53), 24.5% became organ donors (n = 13). Among pediatric patients, 61.0% who received pentobarbital administration survived the inpatient encounter and were discharged (n = 36). Of those deceased, 21.7% became organ donors (n = 5).
Use of Pentobarbital TDM in Dosing Decisions In our study, although pentobarbital dosing changes were common, we found very few examples where a pentobarbital serum/plasma was cited in a physician note as the motivation to adjust pentobarbital dosage. This likely reflects not only primary involvement by pharmacists, but also the use of EEG burst suppression and other clinical end points in guiding pentobarbital dosing. There were just 2 instances of providers explicitly referencing and holding the pentobarbital drip in response to a serum/plasma pentobarbital level in the adult population. There was no explicit mention of the plasma/serum pentobarbital levels in the charts regarding decisions to hold or restart the drip in the pediatric population. Pentobarbital levels were followed until undetectable (,5.0 mcg/mL) in 54.5% (n = 48) of adult and 39.0% (n = 23) in pediatric patients (Table 2). In both adult and pediatric populations, death occurred before undetectable pentobarbital concentrations in approximately 25% of patients. There were only 30 serum/plasma concentrations that exceeded 50 mcg/mL (16 levels in 10 pediatric patients; 14 levels in 8 adult patients). These instances represented only 4.9% of the 612 pentobarbital levels performed (Fig. 1). Pentobarbital dosing was reduced in 8 of 10 pediatric patients with pentobarbital serum/plasma concentrations exceeding 50 mcg/mL. One pediatric patient died with a pentobarbital plasma concentration of 52.2 mcg/mL. This was a 16-year-old boy with intractable seizures despite Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
TDM of Pentobarbital
pentobarbital and multiple other antiepileptic medications. Pentobarbital dosing was reduced in 6 of 8 adult patients with pentobarbital serum/plasma concentrations exceeding 50 mcg/mL, while 2 patients died. Both adult patients who did not survive died of complications of severe head trauma from motor vehicle accident. Pentobarbital was used in both cases to manage elevated ICP.
Coadministration of Pentobarbital With Other Barbiturates Table 2 summarizes how often pentobarbital was used alone and together with one or more barbiturate(s) in the same inpatient encounter. As noted in the Methods section, all specimens to be tested for pentobarbital are first analyzed for phenobarbital as per our laboratory procedure. If phenobarbital is not detected, turnaround time is roughly 2 hours; if phenobarbital is detected, the specimen must be heated in an alkaline solution, increasing the turnaround time by approximately 1 hour. In the study, coadministration of pentobarbital and phenobarbital occurred in 9.1% (n = 8) of adult cases and 47.5% of pediatric cases (n = 28). Of these 28 pediatric cases, 2 additionally had thiopental administration as well; however, no cases involving thiopental were seen after 2010. There were 2 cases where pentobarbital levels were ordered in patients receiving other barbiturates but not pentobarbital. Both occurred in adult patients: one was administered phenobarbital, and the other was administered phenobarbital and thiopental. Among patients not receiving pentobarbital who had pentobarbital levels ordered, only 2 patients had apparent pentobarbital concentrations exceeding 5.0 mcg/mL. Both cases involved suicidal overdose in adults. One of the patients tested positive for barbiturates in their urine and had a history of polysubstance abuse. This patient had an apparent serum pentobarbital concentration of 6.4 mcg/mL. Confirmation testing was not performed for barbiturate identification. The second patient overdosed on an unknown amount of medication containing butalbital. This patient had an apparent serum pentobarbital concentration of 9.1 mcg/mL. Neither of these cases would have had heat and alkaline treatment of the specimen given that no phenobarbital was detected. As noted earlier, several of our patients also received thiopental. Two adult patients received thiopental at the UIHC in the ICUs; both had issues with elevated ICP. Fourteen pediatric patients received thiopental at the UIHC; all had issues with elevated ICP. Most pediatric patients received thiopental in the pediatric intensive care unit (n = 10). One adult patient received thiopental at an outside hospital before transfer to the UIHC. No thiopental was administered in the operating room in both the adult and pediatric populations.
Errors in TDM of Pentobarbital As summarized in Table 3, chart review revealed errors mainly involving confusion in barbiturate names. There were instances where a pentobarbital level was ordered by laboratory staff in error and later cancelled (n = 8 in adults, n = 8 in pediatrics). There were instances of a pentobarbital level ordered in error by the patient care area and similarly cancelled (n = 5 in adults, n = 4 pediatrics). In addition, there
787
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
Humble et al
TABLE 3. Errors in Pentobarbital Drug Monitoring and Medical Record Documentation Adult (18 Years and Older) Order error by laboratory staff with subsequent cancellation of order Order error by patient care area with subsequent cancellation of order Provider note in medical record references the wrong barbiturate Order error not otherwise specified (without cancellation of order)*
Pediatric (Younger Than 18 Years)
8
8
5
4
212
43
29
7
*Includes examples such as phenobarbital levels performed in patients receiving pentobarbital but not phenobarbital.
were occurrences of ordered phenobarbital levels when a patient was not receiving phenobarbital but was receiving pentobarbital; these were not cancelled (n = 29 adults, n = 7 pediatrics). Finally, there were many instances of provider notes referencing the wrong barbiturate in patients receiving pentobarbital administration (eg, phenobarbital referenced when the patient was receiving pentobarbital; n = 212 occurrences between 42 adult patients and n = 43 occurrences between 13 pediatric patients).
Laboratory Details Characteristics of the Pentobarbital Immunoassay The Roche Cobas ONLINE DAT Barbiturates Plus reagent was adapted for determining serum/plasma concentrations of pentobarbital. The assay was linear from 2 to 20 mcg/mL (Fig. 2); at higher pentobarbital concentrations, the immunoassay underestimated the true concentration. Cross-reactivity studies were performed with amobarbital, butalbital, methohexital, phenobarbital, secobarbital, and thiopental, comparing results with and without heat/alkaline pretreatment (Table 4). Amobarbital (1–10 mcg/mL) and methohexital (0.5–10 mcg/mL) showed no cross-reactivity. Without heat/alkaline pretreatment, butalbital showed crossreactivity at 10 mcg/mL and above. Heat and alkaline pretreatment removed cross-reactivity except at 100 mcg/mL, a very high concentration well above the therapeutic range.29 Without heat/alkaline pretreatment, phenobarbital showed cross-reactivity at 25 mcg/mL and above; heat and alkaline pretreatment eliminated the cross-reactivity at concentrations up to 100 mcg/mL. Without heat/alkaline pretreatment, secobarbital showed cross-reactivity at 5 and 10 mcg/mL. After heat and alkaline pretreatment, cross-reactivity was still seen at 10 mcg/mL, a concentration above that seen with therapeutic use of secobarbital and in the range associated with toxicity from acute overdosage.33,34 Thiopental showed
788
FIGURE 2. Characteristics of the pentobarbital immunoassay. Linearity studies of the immunoassay show linear relationship of measured versus spiked concentration from 2 to approximately 20 mcg/mL. At spiked concentrations higher than 20 mcg/mL, the immunoassay underestimates the true concentration. The dashed line indicates a 1:1 relationship. The solid line is from linear regression.
cross-reactivity only at the very highest concentration tested (100 mcg/mL). This is above the concentrations associated with general anesthesia.35,36 Comparison studies of the immunoassay with the older TDx laboratory-developed test are shown in the supplemental data (see Figure S1, Supplemental Digital Content 1, http://links.lww.com/TDM/A116). A comparison was also performed with a reference GC/MS method, performed by a commercial reference laboratory (see Figure S2, Supplemental Digital Content 1, http://links.lww.com/ TDM/A116). These studies included analysis with and without heat/alkaline pretreatment before immunoassay measurement and also a separate immunoassay measurement for phenobarbital [5 specimens shown to contain phenobarbital and pentobarbital are highlighted by circles in Supplemental Digital Content 1 (see Figure S2, http://links.lww. com/TDM/A116)]. Two specimens containing phenobarbital (69.3 mcg/mL phenobarbital in a specimen with 100 mcg/mL pentobarbital by GC/MS and 34.5 mcg/mL phenobarbital in a specimen with 29 mcg/mL pentobarbital by GC/MS) showed marked differences when comparing with or without heat/alkaline pretreatment. Three other specimens contained phenobarbital at concentrations ranging from 15.7 to 24.1 mg/mL and showed less dramatic differences with and without heat/alkaline pretreatment (see Figure S2, Supplemental Digital Content 1, http://links.lww. com/TDM/A116).
DISCUSSION In our retrospective review at an academic medical center, pentobarbital was used primarily for management of elevated ICP or seizures in both the adult and pediatric Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
TDM of Pentobarbital
populations. This is consistent with published literature on therapeutic use of pentobarbital.1–6 Consistent with the severity of the clinical conditions (eg, head trauma, status epilepticus) that often lead to pentobarbital administration, 59.6% of adult and 40.0% of pediatric patients did not survive the hospital admission. In this study, pentobarbital serum/plasma concentrations were generally within the approximate target range of 21–45 mcg/mL cited in the literature for pentobarbital coma.9 Only 4.9% of levels exceeded 50 mcg/mL. Levels at this higher range either led to dosage reductions or were found in patients who were administered pentobarbital as one of multiple therapies for severe seizures or elevated ICP due to head trauma. Preanalytical and postanalytical errors have been identified as a major issue in the clinical practice of TDM of a variety of drugs.38–41 Our study revealed that confusion in barbiturate names (eg, pentobarbital versus phenobarbital) is a challenge in pentobarbital therapy and drug monitoring, with mistakes in ordering of correct tests and also in documentation in the medical record. To our knowledge, errors in barbiturate drug level testing and provider notes did not result in incorrect drug administration. It is possible the issue of referencing the wrong barbiturate in provider notes is due to problems with dictation in some cases. But overall, a contributing factor may be that pentobarbital is not
a commonly used medication and also has a similar name to the more widely used phenobarbital. Pentobarbital is one of the drugs recognized to confound declaration of brain death.10–12 Thus, pentobarbital serum/plasma concentrations need to be below a minimum threshold (typically 5 mcg/mL) to declare brain death. In this study, we observed patients receiving pentobarbital with pentobarbital serum/plasma concentrations measured to undetectable levels after pentobarbital discontinuation in a significant numbers of patients. A quick turnaround time for pentobarbital levels allows for more rapid declaration of brain death in cases where organ donation is a possibility and thus the potential speed of organ recovery. This study presents a laboratory-developed pentobarbital immunoassay suitable for TDM that can be run on an automated clinical chemistry analyzer. Similar to previously reported laboratory-developed immunoassays on older instrument platforms, the current immunoassay does cross-react with some other barbiturates.38–41 This interference can be nullified by heat and alkaline pretreatment. If a laboratory was to treat all specimens as if they contained phenobarbital (by heat/alkaline pretreatment), turnaround time would increase by approximately 1 hour. This study showed that another barbiturate (most often phenobarbital and less commonly thiopental) was used together with pentobarbital in 8.5% of adult cases and 46.6% of pediatric cases (n = 28). In both the adult and pediatric populations, we see no evidence that thiopental interfered with the pentobarbital immunoassay. This is consistent with our findings in the cross-reactivity studies. The assay was linear from 2 to 20 mcg/mL. Given that pentobarbital serum/plasma concentrations may exceed 20 mcg/mL (279 of 621, 44.9%, levels in our retrospective study), dilutions will be required in some cases. However, this immunoassay may provide more rapid turnaround time and be more cost-effective than a dedicated chromatography-based assay. Further, the availability of an immunoassay may allow laboratories to perform pentobarbital levels in-house instead of referring to external reference laboratory. In our study, there is no evidence to suggest that background interference or cross-reaction between barbiturates yielded falsely elevated results in patients administered pentobarbital. Over the course of our retrospective analysis, we only detected 2 patients suspected of having crossreactivity from another barbiturate resulting in apparent pentobarbital concentrations exceeding 5.0 mcg/mL. Neither of these cases received pentobarbital. One of these cases involved an overdose of butalbital. The other was of unknown drug overdose associated with positive screen for urine barbiturates; butalbital would be one possibility as a crossreactant in this case. Given that patients administered pentobarbital tend to be in inpatient settings (often ICUs), administration of additional barbiturates other than phenobarbital or thiopental would be very unlikely. Butalbital was not used as a hospital medication for any patient in our study receiving pentobarbital. Butalbital cross-reactivity is practically eliminated by heat and alkaline pretreatment. Thus, performing this pretreatment for all specimens before pentobarbital analysis would greatly reduce the possibility of butalbital interference even in butalbital overdose cases.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
789
TABLE 4. Cross-Reactivity Studies of the Pentobarbital Immunoassay
Amobarbital
Butalbital
Methohexital
Phenobarbital
Secobarbital
Thiopental
Apparent Pentobarbital Concentration by Immunoassay (mcg/mL)*
Spiked Concentration (mcg/mL)
No Alkaline and Heat Treatment
Alkaline and Heat Treated
1 5 10 5 10 30 50 100 0.5 2.5 5 10 5 25 100 1 5 10 5 10 25 50 100
,5.0 ,5.0 ,5.0 ,5.0 7.3 22.3 30.8 69.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 12.1 23.5 ,5.0 6.6 17.3 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0
,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 11.5 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 ,5.0 13.0 ,5.0 ,5.0 ,5.0 ,5.0 5.0
*Values in bold type were 5.0 mcg/mL or greater.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
Humble et al
All specimens to be analyzed for pentobarbital are screened for phenobarbital as part of the pentobarbital protocol. The cross-reactivity studies demonstrate that thiopental and phenobarbital do not interfere with the pentobarbital immunoassay after heat/alkaline pretreatment. In addition, our laboratory-developed immunoassay runs on an automated clinical chemistry platform and requires significantly less technical staff labor time and expertise than chromatographybased assays. The availability of an immunoassay also more readily allows for assay runs across all times of the day. In contrast, for hospital-based laboratories, chromatographybased assays may be especially difficult to support during times of lower staffing, such as evenings, nights, weekends, and holidays. If a clinical laboratory seeks to adapt the pentobarbital immunoassay, it is essential to perform thorough validation studies including cross-reactivity analysis. Future studies could investigate the impact of turnaround time of pentobarbital drug level on organ donation rates. Our study supplements the existing literature in several ways. Observation of pentobarbital TDM over a 14-year period at an academic medical center allows us to observe long-term trends in use of pentobarbital TDM in both adult and pediatric populations. We shed light on common preanalytical and postanalytical errors that can serve as a means to improve on existing patient care practices and use of EMRs. Our study presents an alternative methodology to measure pentobarbital concentrations in serum/plasma that accounts for the presence of other barbiturates, has a faster turnaround time than other methodologies, and reduces the labor burden of the clinical laboratory by use of an automated clinical chemistry platform. There are some shortcomings of our study. The EMR data was incomplete for some patients, particularly for cases before the implementation of a new EMR system in 2009. In addition, our study involved data from a single institution; other institutions may have different protocols in place for TDM of pentobarbital. Nevertheless, the results provide the largest set of data published to date on the clinical practice of pentobarbital TDM. ACKNOWLEDGMENTS The authors are grateful for helpful discussions with Douglas Morgan and Ellen Nickel (Department of Pharmaceutical Care, University of Iowa Hospitals and Clinics) on the clinical use of pentobarbital in the pediatric and adult populations, respectively. The authors also thank Jon Maakestad (Department of Pathology, University of Iowa Hospitals and Clinics) for assistance in preparing barbiturate concentrations for cross-reactivity testing. REFERENCES 1. Mellion SA, Bennett KS, Ellsworth GL, et al. High-dose barbiturates for refractory intracranial hypertension in children with severe traumatic brain injury. Pediatr Crit Care Med. 2013;14:239–247. 2. Mansour N, deSouza RM, Sikorski C, et al. Role of barbiturate coma in the management of focally induced, severe cerebral edema in children. J Neurosurg Pediatr. 2013;12:37–43. 3. Marshall GT, James RF, Landman MP, et al. Pentobarbital coma for refractory intra-cranial hypertension after severe traumatic brain injury:
790
4. 5. 6. 7. 8. 9. 10. 11. 12. 13.
14. 15.
16. 17.
18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.
mortality predictions and one-year outcomes in 55 patients. J Trauma. 2010;69:275–283. Abend NS, Dlugos DJ. Treatment of refractory status epilepticus: literature review and a proposed protocol. Pediatr Neurol. 2008;38:377–390. Censullo JL, Sebastian S. Pentobarbital sodium coma for refractory intracranial hypertension. J Neurosci Nurs. 2003;35:252–262. Kinoshita H, Nakagawa E, Iwasaki Y, et al. Pentobarbital therapy for status epilepticus in children: timing of tapering. Pediatr Neurol. 1995; 13:164–168. Roberts I, Sydenham E. Barbiturates for acute traumatic brain injury. Cochrane Database Syst Rev. 2012;12:CD000033. Bader MK, Arbour R, Palmer S. Refractory increased intracranial pressure in severe traumatic brain injury: barbiturate coma and bispectral index monitoring. AACN Clin Issues. 2005;16:526–541. Heinemeyer G, Roots I, Dennhardt R. Monitoring of pentobarbital plasma levels in critical care patients suffering from increased intracranial pressure. Ther Drug Monit. 1986;8:145–150. Meinitzer A, Kalcher K, Gartner G, et al. Drugs and brain death diagnostics: determination of drugs capable of inducing EEG zero line. Clin Chem Lab Med. 2008;46:1732–1738. Wijdicks EF. The diagnosis of brain death. N Engl J Med. 2001;344: 1215–1221. Wenck DJ. Drugs and brain death testing. Crit Care Med. 2000;28:601. Lehane DP, Menyharth P, Lum G, et al. Therapeutic drug monitoring: measurements of antiepileptic and barbiturate drug levels in blood by gas chromatography with nitrogen—selective detector. Ann Clin Lab Sci. 1976;6:404–410. Wallace JE, Hall LR, Harris SC. Determination of pentobarbital and certain other barbiturates by capillary gas-liquid chromatography. J Anal Toxicol. 1983;7:178–180. Johnson LL, Garg U. Quantitation of amobarbital, butalbital, pentobarbital, phenobarbital, and secobarbital in urine, serum, and plasma using gas chromatography-mass spectrometry (GC-MS). Methods Mol Biol. 2010;603:65–74. Kelner M, Bailey DN. Reversed-phase liquid-chromatographic simultaneous analysis for thiopental and pentobarbital in serum. Clin Chem. 1983;29:1097–1100. Spell JC, Srinivasan K, Stewart JT, et al. Supercritical fluid extraction and negative ion electrospray liquid chromatography tandem mass spectrometry analysis of phenobarbital, butalbital, pentobarbital and thiopental in human serum. Rapid Commun Mass Spectrom. 1998;12:890–894. Cannon RD, Wong SH, Gock SB, et al. Comparison of the serum barbiturate fluorescence polarization immunoassay by the COBAS INTEGRA to a GC/MS method. Ther Drug Monit. 1999;21:553–558. Wellington P, Rutkowski RB. Rapid estimation of plasma pentobarbital levels by an enzyme immunoassay for barbiturates (EMIT). Ther Drug Monit. 1982;4:319–322. Pape BE, Cary PL, Clay LC, et al. Pentobarbital quantitation using EMIT serum barbiturate assay reagents: application to monitoring of high-dose pentobarbital therapy. Ther Drug Monit. 1983;5:467–471. Li PK, Lee JT, Schreiber RM. Rapid quantification of pentobarbital in serum by fluorescence polarization immunoassay. Clin Chem. 1984;30:307–308. Earl R, Sobeski L, Timko D, et al. Pentobarbital quantification in the presence of phenobarbital by fluorescence polarization immunoassay. Clin Chem. 1991;37:1774–1777. Krasowski MD, Siam MG, Iyer M, et al. Chemoinformatic methods for predicting interference in drug of abuse/toxicology immunoassays. Clin Chem. 2009;55:1203–1213. Coupey SM. Barbiturates. Pediatr Rev. 1997;18:260–265. Krasowski MD, Siam MG, Iyer M, et al. Molecular similarity methods for predicting cross-reactivity with therapeutic drug monitoring immunoassays. Ther Drug Monit. 2009;31:337–344. Lowenstein DH, Aminoff MJ, Simon RP. Barbiturate anesthesia in the treatment of status epilepticus: clinical experience with 14 patients. Neurology. 1988;38:395–400. Mawer GE, Miller NE, Turnberg LA. Metabolism of amylobarbitone in patients with chronic liver disease. Br J Pharmacol. 1972;44:549–560. Balasubramaniam K, Lucas SB, Mawer GE, et al. The kinetics of amylobarbitone metabolism in healthy men and women. Br J Pharmacol. 1970;39:564–572. Moffat AC, ed. Clarke’s Isolation and Identification of Drugs. 2nd ed. London, United Kingdom: The Pharmaceutical Press; 1986.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Ther Drug Monit Volume 37, Number 6, December 2015
TDM of Pentobarbital
30. Lewis RJ, Johnson RD, Southern TL, et al. Distribution of butalbital in postmortem tissues and fluids from non-overdose cases. J Anal Toxicol. 2003;27:145–148. 31. Breimer DD. Pharmacokinetics of methohexitone following intravenous infusion in humans. Br J Anaesth. 1976;48:643–649. 32. Liu LM, Gaudreault P, Friedman PA, et al. Methohexital plasma concentrations in children following rectal administration. Anesthesiology. 1985;62:567–570. 33. Levine HL, Cohen ME, Duffner PK, et al. Rectal absorption and disposition of secobarbital in epileptic children. Pediatr Pharmacol (New York). 1982;2:33–38. 34. Greenblatt DJ, Allen MD, Harmatz JS, et al. Overdosage with pentobarbital and secobarbital: assessment of factors related to outcome. J Clin Pharmacol. 1979;19:758–768. 35. Loft S, Jensen V, Rorsgaard S. Influence of moderate alcohol intake on wakening plasma thiopental concentration. Acta Anaesthesiol Scand. 1983;27:266–269.
36. Stanski DR, Hudson RJ, Homer TD, et al. Pharmacodynamic modeling of thiopental anesthesia. J Pharmacokinet Biopharm. 1984;12:223–240. 37. Pentobarbital serum/plasma assay. Salt Lake City, UT: ARUP Laboratories. Available at: http://ltd.aruplab.com/Tests/Pub/2011549. Accessed April 13, 2015. 38. Kontny NE, Boos J, Wurthwein G, et al. Minimization of the preanalytical error in pharmacokinetic analyses and therapeutic drug monitoring: focus on IV drug administration. Ther Drug Monit. 2012;34:460–466. 39. Kontny NE, Hempel G, Boos J, et al. Minimization of the preanalytical error in plasma samples for pharmacokinetic analyses and therapeutic drug monitoring–using doxorubicin as an example. Ther Drug Monit. 2011;33:766–771. 40. Schiff GD, Klass D, Peterson J, et al. Linking laboratory and pharmacy: opportunities for reducing errors and improving care. Arch Intern Med. 2003;163:893–900. 41. Sieradzan R, Fuller AV. A multidisciplinary approach to enhance documentation of antibiotic serum sampling. Hosp Pharm. 1995;30:872, 874, 876–877.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
791
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
