I'm sorry to break this to you, but your ICU patient’s valproate (VPA) level may be lying to you. Hidden within the total valproate concentration you checked is the concentration that matters – the free concentration, which is pharmacologically active and primarily responsible for the CNS exposure to VPA^1,2 – but without peeking under the hood and directly measuring that free concentration, it is nearly impossible to predict exactly how much VPA your patient is exposed to. Because of this, dose adjustments made based on total concentrations have the risk of widely missing the mark. While altered protein binding with valproic acid is a well-known phenomenon especially in the setting of drug interactions like co-therapy with phenytoin, the extent to which protein binding can be altered in ICU patients is criminally underrecognized. Here we'll talk a bit about how VPA’s pharmacokinetics are unique, what factors influence its binding to albumin, why those factors matter in the ICU, and what to do about it.

January 15th, 2024 Update - Check out the interactive Fraser Free VPA Estimation Calculator here!

As a side note – my literature search for this post turned up way more articles than I was expecting. Not all of them were particularly relevant and not all could be included, but if you would like the list of the 191 articles I found for this post, you can find them here.

Key Points

Valproic acid protein binding is substantially altered in critically ill patients and total VPA concentrations may be unreliable
Total and free VPA concentrations may be therapeutically discordant in up to 80% of ICU patients
Published prediction equations fail to satisfactorily account for the variability caused by critical illness and thus free VPA cannot be reliably predicted by available equations
Factors including hepatic dysfunction and hypoalbuminemia, renal dysfunction, co-therapy with phenytoin, elevation in free fatty acid, and the time of sampling may play a role in the variability seen with free VPA percentages
Elevated free VPA is associated with toxicity and should be directly measured in critically ill patients

Illustrative Case

A 65-year-old woman is admitted to the NeuroICU for status epilepticus. She has a past medical history of depression and a prior TBI complicated by post-traumatic epilepsy which had been controlled by levetiracetam. In the ED, she received lorazepam IV 4mg, levetiracetam IV 4500mg followed by 1500mg q12, and is ultimately intubated for airway protection and started on propofol for seizure control and vent tolerance. As the propofol is weaned the following day, she develops epileptiform discharges on EEG and is loaded with valproic acid IV 3000 mg. Her post-load valproic acid level was 91.4 mg/L and she started on 500 mg IV q8 (~20 mg/kg/day), experienced good response on EEG, and propofol was weaned off and vent weaning began. The following day, despite propofol being off and no seizure activity being captured on EEG, the patient’s mental status remained poor. Her total valproic acid level was 52 mg/L, ammonia was 18 umol/L, and her LFTs were notable for an albumin of 2.9 g/dL. Presuming she post-ictal and just needed time, she extubation was deferred. The following day, her mental status remained poor, but her free valproic acid concentration resulted from the reference laboratory at 32 mg/L, suggesting a toxic exposure to valproic acid. Her valproic acid dose was decreased and the following day the patient’s mental status began to improve and she was extubated later that day.

This case, which is informed by my practice (the details have been altered), is more common than we may think. Similar stories of discordant total and free levels are well-described in the literature^3–5 and problems with protein binding, especially in patients with hepatic and renal impairment, were reported before it was even approved in the United States.^6,7 Despite its unpredictable inter- and intraindividual pharmacokinetics, total concentrations remain the standard method of monitoring which can lead to situations like the one above in vulnerable populations.

Free Valproic Acid in Critically Ill Patients

How big of a deal is this discordance in ICU patients, you might ask? You would not be the first – multiple authors have evaluated discordance in measured VPA concentrations, and the effect is most profound in the critically ill. Discordance is defined as a difference in the interpretation of the free and total VPA level, which could mean a therapeutic total VPA but supratherapeutic free VPA, subtherapeutic total VPA but therapeutic free VPA, and so on. Discordance ranges from 29.7% to 87%, and the results are summarized below. Notably, while Gibbs et al reported hypoalbuminemia as a major driver of discordance in their general inpatient population, hypoalbuminemia has not seemed to explain the substantial discordance in other cohorts.

Author	Sample Size	Therapeutic Free VPA	Discordance %
Gibbs⁸	219 inpatients, 41 outpatients	4.8-17.3 mcg/mL	Inpatient: 63% Outpatient: 19%
Riker⁹	15 critically ill patients	5-17 mcg/mL	87%
Fisch¹⁰	104 patients in status epilepticus	5-10 mcg/mL	29.7%
Wallenburg¹¹	132 inpatients, 10% from ICUs	4-12 mcg/mL	32%
Drisaldi¹²	174 patients, 59.8% on the floor and 28.8% in the ICU	7-23 mcg/mL	43.1% (50% of ICU patients had discordance of >20% from expected free)
Brown¹³	256 critically ill patients	5-15 mcg/mL	70%

While differences in population and defined therapeutic ranges may explain the variation in discordance, the common conclusion is that discordance is common in ICU patients, and much of the discordance is driven by supratherapeutic free VPA levels even in the setting of subtherapeutic total VPA levels. Why might this be? Let’s dive in to what is unique about VPA and factors that may drive its maddeningly variable protein binding.

Valproic Acid

VPA is truly a medication with a can-do attitude. Good for seizures, status epilepticus, agitation, migraine, bipolar depression, and more, it can wear many hats in and out of the ICU. While versatile, it of course comes with its share of problems – multiple black box warnings (hepatotoxicity, risk in mitochondrial disease, teratogenicity, and pancreatitis) are only the start of the lengthy list of its potential adverse events which range from reasonably mild ADRs like weight gain and GI distress to fulminant hyperammonemic encephalopathy and severe thrombocytopenia. Paired with this is widely variable inter- and intraindividual pharmacokinetics and a narrow therapeutic index, and thus therapeutic drug monitoring is essentially mandated in all patients except for low dose use in low-risk patients.

VPA Pharmacokinetics – The Basics

If you’ve ever tried to look up the pharmacokinetics of VPA on a tertiary resource, you’ve likely noticed something – for each parameter, range of the provided value is quite wide. Absorption ranges from 80-100%, time to peak absorption ranges from 4 to 17 hours, volume of distribution ranges from 10-90 L, and half-life ranges from 9 to 19 hours. The factors contributing to this are numerous, including dosage form, age, body weight, food intake, the time the sample is taken, and how the actual PK models are built.¹⁴ Protein binding in particular, however, is the most variable at all – the reported range is often 10-20%, but its nonlinear, concentration dependent behavior influenced by a laundry list of factors make this far more complex. The relationship between total VPA and free VPA as well as total VPA and free fraction has been assessed in numerous populations using different dosage forms of VPA, methods of estimating free concentrations, and at various time points and no two studies seems to report the same relationship. The correlation between total VPA and free fraction has been estimated to be linear,¹⁵ exponential,¹⁶ or just plain poorly correlated,¹⁷ with the relationship between total VPA and measured free VPA following a similar pattern.^18–21 One concept remains true – VPA protein binding is a saturable process. As total VPA concentrations increase, the percent of that concentration that is unbound increases, leading free VPA concentrations to behave in a nonlinear fashion. The difference in relationship between total VPA and free VPA and free VPA percent is nicely depicted here:¹⁷

Numerous attempts have been made to derive equations to accurately predict this behavior, but few have made it to a successful external validation, and all suffer from large magnitude estimation errors (a sample of relevant studies are listed below).^{12,15,18,22–31}

Author	Population/N	Equation	Correlation	Notes
Otten¹⁵	6 pediatric-young adult outpatient epilepsy patients	%free = 0.12 * total VPA + 2.18	r = 0.81	Simple linear regression. Up to 81% error noted in one patient
Kodama¹⁸	9 healthy volunteers	Free VPA (umol/L) = ½ * (total VPA – 757 – 1/0.0281 + ((757 – total VPA + 1/0.0281)² + 4*total VPA/0.0281)^1/2)	Not reported	Derived from mathematical models of protein binding. 1 umol/L VPA = 7 mg/L
Kodama²²	7 healthy volunteers	Free VPA (umol/L) = ½ * (total VPA – 757 – 1/0.0281 + ((757 – total VPA + 1/0.0281)² + 4*total VPA/0.0281)^1/2)	r = 0.865	Loses accuracy at total VPA levels above 85 mg/L
Kodama²³	23 pediatric outpatients with epilepsy	Free VPA (umol/L) = ½ * (total VPA – 757 – 1/0.0281 + ((757 – total VPA + 1/0.0281)² + 4*total VPA/0.0281)^1/2)	r = 0.872	Tended to underestimate free VPA; lost accuracy at VPA levels above 80 mg/L
Hermida²⁴	53 adult patients on VPA	%free = 130.69 e^{-0.00496albumin (umol/L)}	r = -0.82	Up to 12.5 mg/mL prediction error when total VPA was greater than 60 mg/L. Equation from Parent et al.³²
Fisch¹⁰	106 patients with status epilepticus	%free = 130.69 e^{-0.00496albumin (umol/L)}	r not reported, but β 1.1	fVPA overestimated the measured is 30.4% of cases
Nasreddine²⁸	902 outpatients with epilepsy enrolled in VPA trial	Free VPA (mg/L) = (0.0016total VPA) + (0.012total VPA) + 0.4134	R² = 0.88	Albumin not available to include in model.
Dore	41 inpatients and outpatients on VPA	Free VPA (umol/L) = 103.667 + 0.362total VPA - 4.538Albumin (g/L)	R² = 0.855	Mean absolute prediction error of 3.4 mcg/mL; ICU patients not included
Giner³⁰	33 patients admitted to a tertiary hospital	Free VPA = 11.882 + 0.216total VPA – 4.722albumin	R²= 0.637	Prediction error ranged from -3.8 to 4.44 mg/L
Ishikawa³¹	75 patients on VPA, 51.5% admitted to hospital but 0% in ICU	Free VPA (mcg/mL) = total VPA x 0.377 * e^{0.001*Albumin (uM)}	r = 0.75 in an external data set	Average error of 7.2 mcg/mL

Notably, none of these equations have been validated in an ICU population.

Protein Binding – More than Just Albumin

Let’s briefly discuss the theoretical basis of protein binding. In school, I was taught protein binding largely as a binary concept – a drug is either bound to a plasma protein (most commonly albumin) or it was not. The reality is more complex, with plasma proteins acting as reservoirs of their own receptors competing for a drug’s attention, and drugs have different levels of affinity for different areas on protein. Drugs can bind to more than one site as well, and can compete with other medications and other substances within plasma for spots on the protein. This concept is outlined in a mathematical level of detail frankly beyond my comprehension by George Scatchard in 1949³³ which laid the foundation for the determination of medication protein binding behavior and in particular how to calculate the number of binding sites and the affinity of a medication to each binding site using what is called a Scatchard plot.

Early work with VPA and its protein binding revealed that VPA has two main binding sites on albumin with one predominating at therapeutic concentrations and the second becoming more relevant in toxicity.^7,34–36 A wide number of substances and medications alter the strength of VPA’s affinity for those binding sites, greatly affecting the proportion of drug that is unbound. Because of the breadth of the competition, VPA’s protein binding is also easily saturable and the proportion of unbound drug begins to increase at total concentrations as low as 50 mg/L from 5% unbound at low concentrations to over 40% unbound at concentrations above 150 mg/L.²¹ This was in outpatients with epilepsy – the curve is even more extreme in the critically ill, with reports of free fractions of 60-80% even at subtherapeutic total concentrations.^13,37,38 This wide fluctuation in protein binding then affects the intrinsic clearance of VPA, which is tightly linked to free concentrations and thus varies depending on dose as total concentrations increase, leading to nonlinear changes in serum concentrations when doses are changed.³⁹ A general rule of thumb is that dose changes based on total concentration should change the total concentration by 80% of the estimated linear change – for example, increasing a total daily dose of VPA from 1,000mg to 1,500mg would be expected to increase the total VPA concentration by 20%.

In this Scatchard plot, the affinity for VPA to its protein binding site is depicted on the Y axis and the number of albumin binding sites is depicted on the X axis. The affinity for albumin decreases as the concentration of phenytoin in solution increases (lines i-v)

This sort of protein variability likely sounds familiar – phenytoin is also heavily protein bound, dependent on albumin, can have altered binding in critical illness, and discordance between the total and free concentration of phenytoin is also common in critically ill patients.⁴⁰ Luckily, a number of correction equations for phenytoin protein binding exist to make interpretation of total concentrations simpler, and equations specifically designed for critically ill patients can improve the accuracy of estimates in ICU patients.⁴¹ The issue with VPA, however, is every correction equation, even if well-performing on in-sample data, simply fails to perform consistently in ICU patients. One study which evaluated the accuracy of a free VPA correction equation in 104 patients with status epilepticus found that estimates incorrectly estimated free VPA concentrations to be subtherapeutic, therapeutic, or supratherapeutic 39.8% of the time. The majority of mischaracterizations were overestimates, potentially putting patients at risk of treatment failure.¹⁰

What Drives VPA Protein Binding?

For phenytoin, hypoalbuminemia, critical illness, and renal dysfunction reliably predict altered protein binding and can be, with some confidence, used to estimate free phenytoin concentrations in the absence of direct measurement. What factors influence VPA’s altered protein binding, and why is estimation so difficult? VPA is a branched short-chain fatty acid which behaves in much the same way as other fatty acids when it comes to protein binding. This is relevant, as free fatty acids are among the most common molecules bound to albumin and changes nutrition, activity, or environment can significantly alter the kinetics of free fatty acid protein binding and potential alter drug protein binding.⁴² VPA likely binds to albumin at the same site as organic fatty acids like oleic acid and palmitic acid, as demonstrated by a linear increase in unbound VPA with increasing organic free fatty acids in serum samples.^36,43 VPA’s multiple metabolites are all free fatty acid derivatives as well and are unmeasured in clinical practice yet also contribute to VPA’s pharmacologic effect as well as net protein binding.⁴⁴ Here we will explore the many factors at play to help you determine who is at the biggest risk of discordant VPA concentrations.

Hepatic Dysfunction and Hypoalbuminemia

The impact of hepatic dysfunction on protein binding should be evident due to lower levels of albumin in most patients with cirrhosis. This was noted early in VPA’s use, with a study in 11 patients with cirrhosis demonstrating a near 20% difference in protein binding compared to health controls (free fraction of 29.3% in patients with alcoholic cirrhosis vs 11.3% in healthy controls).⁴⁵ Protein binding is directly correlated with albumin concentration even in non-cirrhotic patients as well, with an estimated 4-5% increase in free fraction with each 1 g/dL decrease in albumin, although this relationship had low accuracy for patients with total VPA concentrations above 80 mcg/mL.⁴⁶ This relationship has been confirmed multiple times in the hospitalized and ICU population, with multiple reports of extreme changes in protein binding with worsening albumin concentrations and at the population level each 1 g/dL decrease in albumin also correlating with a 4.6% increase in free concentration, even after adjusting for confounding.^{13,29,38,47,48}

Renal Dysfunction

Renal dysfunction, as another major cause of hypoalbuminemia, understandably also affects VPA protein binding. Early work by Gugler et al revealed that free VPA correlated well with increasing serum creatinine and BUN, suggesting that uremic compounds also compete with VPA for spots on albumin.⁷ Interestingly, renal dysfunction may be protective of several other albumin binding site interactions like those caused by NSAIDs (although why someone with CKD would be on an NSAID is another discussion), suggesting that there is a limited capacity for these interactions to occur when multiple competing factors are present.⁴⁹

Phenytoin

The effect of VPA on PHT levels

A discussion of VPA’s protein binding would be incomplete without a mention of the complex relationship between VPA and phenytoin. The impact of VPA on phenytoin concentrations was noted early in VPA’s use in clinical practice,⁵⁰ but the mechanism of protein binding displacement wasn’t noted for a few years.⁵¹ VPA and phenytoin share an albumin binding site, which is phenytoin’s only binding site while it is just one of two major binding sites for VPA. VPA’s association constant is higher than phenytoin for this site, which means that at VPA concentrations above 40 mg/L, VPA begins to actively displace phenytoin from its binding site, increasing its free concentration.³⁵

Estimated vs measured PHT
on VPA co-therapy

Because phenytoin has a low hepatic extraction ratio, the increased available free drug is cleared until a new equilibrium is reached, leading to a lower total phenytoin concentration but a consistent free concentration. On average, introduction of VPA will increase the free phenytoin fraction from 9.6% to 15.6%.⁵² This interaction can be modeled, with higher VPA concentrations leading to larger increases in phenytoin free fraction in a semi-logarithmic fashion.³⁴ In the therapeutic range of VPA, a regression equation of 0.0792 * total phenytoin + 0.000636 * total phenytoin * total VPA also demonstrated good accuracy in a sample of 33 patients on co-therapy.⁵³

The effect of PHT on VPA is less pronounced – a similar effect does occur at high phenytoin concentrations, leading to increases in free VPA fraction. At therapeutic concentrations of phenytoin, however, the effect is marginal, with only a limited impact on VPA protein binding.³⁵ Regardless, whenever a patient is initiated on each therapy together, especially in the ICU, monitoring free concentrations of both medications is paramount. While the effect of phenytoin on VPA is marginal under ideal circumstances, in critically ill patients and in low albumin states, the effect can be pronounced, with an increase in free VPA fraction from 24% at a total VPA level of 35 mcg/mL to 37.8% at 90 mcg/mL at a constant phenytoin concentration of 20 mcg/mL.²¹

The effect of PHT on VPA levels

Free Fatty Acids

Increasing oleic acid is directly
related to increasing free VPA

Perhaps the most underrecognized factor which impacts VPA protein binding is serum free fatty acid concentrations. As VPA is itself a fatty acid, it competes with physiologic serum fatty acids such as palmitic acid (the most abundant saturated fatty acid in mammals) for binding sites on albumin.³⁵ Early work found the effect to be quite significant in-vitro, with free fatty acids added to samples containing VPA increasing VPA free fraction in a concentration-dependent manner. Patel et al measured the free fraction of VPA at various concentrations of different free fatty acids and found that all four free fatty acids tested (stearic, palmitic, oleic, and linoleic acid) significantly increased free fraction, with linoleic acid more than doubling the free fraction at a concentration of 200 mcg/mL.³⁶ There was little described consequence to this in the clinical space until 1981 when Zimmerman et al reported on protein binding characteristics in 9 patients on VPA who were receiving Intralipid for free fatty acid deficiency. VPA protein binding was measured before and after Intralipid administration, with 6 of 8 patients who had adequate samples available demonstrating a median 18.9% increase in free fraction with a correlation coefficient of 0.80 for the relationship between change in free fatty acid concentration and change in free fraction. This was accompanied by an additional in-vitro study which confirmed a near linear relationship between oleic acid concentration and VPA free fraction.⁴³ A similar relationship has been noted even in patients not being directly infused free fatty acids – because fasting increases serum free fatty acid concentrations, Bowdle et al demonstrated significant variability in VPA protein binding in 6 healthy patients depending on whether patients were fasted or fed, with protein binding being on average 9% lower when patients were fasted. Additional analysis suggested free fatty acid alone accounted for at least 37% of the variation in VPA protein binding.⁵⁴ This phenomenon may be particularly relevant for laboratory personnel, as samples left at room temperature gradually have increasing concentrations of free fatty acids, and thus lower VPA protein binding.⁵⁵ This relationship may also exist in reverse, with VPA displacing palmitic acid albumin binding, and some authors suggesting this may contribute to VPA’s well-known side effect of weight gain.⁵⁶

While this may seem like it is only relevant to patients on TPN, a few key medications familiar to us in the ICU have high free fatty acid content which may carry a similar effect as Intralipid, namely propofol and clevidipine. Interestingly, Brown et al specifically evaluated receipt of propofol as a potential modulating factor in VPA protein binding and did not find that it was significantly associated with VPA protein binding after multivariable adjustment.¹³ Propofol exposure was a binary variable, however, and any exposure within 24 hours of the level being drawn was considered exposure, so this may not capture the nuanced and time-bound relationship between free fatty acid and VPA protein binding. Dose was also not considered, and it is likely that the extent of lipid exposure influences the effect on protein binding. Interestingly, VPA has also been shown to decrease propofol protein binding in vitro, however supraphysiologic VPA concentrations (1000 mg/L) were used to assess the response so it is questionable whether that is clinically relevant (protein binding was insignificantly changed at 100 mg/L except in hypoalbuminemic serum).⁵⁷

As a side note, lipid administration has interestingly been shown to be a potential therapy in ameliorating VPA-induced hyperammonemia, apparently through inhibiting the kidney from producing ammonia.⁵⁸

Aspirin and NSAIDs

Like free fatty acids, aspirin and NSAIDs and their relationship to VPA protein binding was frequently published on in the 1980s and 90s but do not seem to come up in clinical practice often. Case reports of VPA toxicity due to even low-dose aspirin use appear sporadically in the literature,^4,59,60 but it does not flag in all EMR systems (it doesn’t in mine, although you can find it in Lexicomp and Micromedex). The same can be said about NSAIDs, with reports of both naproxen⁶¹ and ibuprofen⁶² supporting significantly increased free VPA after initiation of NSAIDs. While less well-explored in the in-vitro space, Fleitman et al did investigate the effects of several medications on VPA albumin binding, including salicylic acid, and found that salicylic acid more than doubled the VPA free fraction (15% to 35%) at a total VPA concentration of 128.5 mcg/mL and increased the free fraction 15x (2% to 30%) at a total VPA concentration of 40.2 mcg/mL.⁶³ This has been confirmed in multiple patient populations, including in the serum of patients with liver disease,⁶⁴ children,⁶⁵ and renal failure, although the interaction may be less significant in uremic plasma.⁶⁶

Time Itself?

Variation in VPA levels during
continuous infusion

There is just one more variable left which significantly influences VPA’s protein binding to discuss – time itself. Like several other medications (most classically described with tacrolimus⁶⁷), VPA’s pharmacokinetic parameters overall vary over the course of a 24-hour period. Loiseau et al first investigated this in two arms – one arm had serial plasma samples taken after BID oral administration, and the second arm had serial measurements during a continuous infusion at a constant rate. The oral administration arms understandably had significant variation throughout the day (peaks and troughs), but the variation was substantially higher than would be expected for normal inter-dose variation (average of 112% variation). This seemed to be driven by continued concentration decline even after the PM dose, suggestive of delayed absorption likely from an interaction with food. To remove the effect of absorption, serial levels in patients on a continuous infusion were also measured, which similarly showed variation out of proportion to lab sample error, with an average variation of 27%. The authors conclude that effects of protein binding variation may explain changes in clearance, and that a consistent morning sampling time is likely ideal to limit the effect of diurnal variation.⁶⁸ The conjecture that diurnal changes in protein binding may drive variation has been confirmed in several studies, with Marty et al reporting about a 100% variation in inter-dose protein binding,⁶⁹ Patel et al reporting 30-60% variation across the dosing interval,⁷⁰ and Patients et al reporting about 79% variability in protein binding across the dosing interval.⁷¹ While much of the variation can be explained by the concentration-dependent protein binding (i.e. at peak concentrations more drug will be free because concentrations are higher⁷²), two other explanations may be at play. First, changes in free fatty acid concentration throughout the day likely impacts the changes in protein binding to a large extent. Patel et al demonstrated a striking correlation between free fatty acid concentrations which varied with respect to meal timing and fasting and free fraction. Early in the morning after patients had been fasted for the night, free fraction was nearly double the average free fraction during the day after patients had been fed, consistent with prior findings of the influence of free fatty acids on protein binding.⁷⁰ Second, clearance of free VPA appears to be higher by about 10-15% in the evening, which has been demonstrated both with intermittent dosing⁷² and continuous infusion (in rats).⁷³ While the pattern of clearance doesn’t perfectly fit a circadian pattern and likely more so reflects oscillations in competing substances caused by diet, the overall takeaway is that picking a consistent time to check steady state troughs is ideal to avoid catching interdose variation.⁷⁴

Example of differences in peaks and troughs between morning and evening doses of VPA

Relationship between changes in free fatty acid concentrations throughout the day and changes in VPA protein binding throughout the day

Clinical Implications

Naturally, the solution to this conundrum is to directly measure the free VPA concentration in all critically ill patients. There are a few problems with this approach, however – first and most obvious is that measuring free valproic acid is technically challenging and requires experience with ultrafiltration,⁷⁵ is subject to large magnitude measurement error if samples are handled incorrectly,⁷⁶ and is a send-out laboratory at many hospitals which impairs its ability to be used for real-time decision making. Second, while the therapeutic range of 50-100 mg/L of total VPA is reasonably well-established (worth noting that this range was determined essentially by a good guess and has just sort of stuck since 1975 when Loiseau et al said “The therapeutic range seems to 50-100 mg/L” in their discussion),^77–81 the therapeutic range of free VPA remains largely unknown. Different reference laboratories list different normal ranges, from 5-25 mcg/mL at Mayo,⁸² 4.8-17.3 mcg/mL at Quest,⁸³ 7-23 mcg/mL at ARUP,⁸⁴ and 6-22 mcg/mL at Lab Corp.⁸⁵

While the laboratories are not explicit in how these ranges are derived, they are presumably from internal data on what the free concentrations tend to be in patients with therapeutic total concentrations. The empirical evidence does not provide much additional insight – only a handful of studies have attempted to specifically quantify the relationship between free VPA and therapeutic effect with inconsistent results given varying methods and patient populations. Yu measured total and free VPA concentrations in 18 pediatric patients with epilepsy and found an impressive free VPA range of 5-58 mcg/mL although the relationship between free concentrations and seizure control was not reported.¹⁹ Farrell et al attempted to establish the relationship between both total and free VPA concentrations and seizure control in 61 outpatient pediatric patients with epilepsy. Over 75% of patients achieved seizure control in a surprisingly narrow free VPA range of 1.25-3.77 mcg/mL.⁸⁶ Kilpatrick et al attempted to correlate VPA concentrations and seizure control in a population of 70 outpatient adults with epilepsy and failed to detect a population-level correlation with either total or free VPA but did find that at the individual level, 75% of patients with changes in seizure frequency during the study period had total levels that correlated with seizure changes and 67% had free concentrations that correlated with response. The authors also used a free VPA of 2.8-7.1 mcg/mL as the definition of therapeutic but do not explain where this was derived from.⁸⁷ The concept of individual-level therapeutic ranges is also supported by Nakashima et al, who used popPK methods to individually determine trough goals in 77 outpatients with epilepsy.⁸⁸ These data largely raise more questions than they answer, and are effectively useless when considering the use of VPA in the ICU population.

The other angle to evaluate free VPA ranges from is toxicity, where there is some more concrete evidence for how to best use the levels. Beyond anecdotal reports of altered mental status with elevated free concentrations (free concentrations of 34.7 mcg/mL³ and 37.8 mcg/mL⁸⁹ in two illustrative case reports and as low as 15 mcg/mL⁵ in another), larger population studies have also shown a correlation with free concentrations and several key adverse events. Itoh et al assessed the association with VPA dose (mg/kg), total VPA, and free VPA with the incidence of hyperammonemia and found free VPA (slightly) outperformed dose and total VPA in terms of linear correlation with plasma ammonia in 19 outpatients with epilepsy. CART analysis was used to identify breakpoints for the incidence of hyperammonemia (plasma ammonia > 60 micromol/L) which supported a relatively low free VPA cutoff of 5.05 mcg/mL as predictive of the adverse event.⁹⁰ Brown et al also evaluated the incidence of specific adverse events in relation to free VPA concentrations in 256 critically ill patients on valproic acid. While stepwise increases in free VPA were not significantly associated with increasing incidence of hepatotoxicity, thrombocytopenia, or hyperammonemia, the model estimate favored an increase risk of hyperammonemia when free VPA concentrations were above 15 mcg/mL (OR 1.91, 95% CI 0.81-4.49). While not found by Brown and colleagues, Nasreddine et al also evaluated the relationship between free VPA and thrombocytopenia in 264 patients who had enrolled in an outpatient valproic acid monotherapy trial and revealed a linear relationship between free VPA and platelet count, with a 9.8% increase in the odds of thrombocytopenia (platelet count <100,000) per 1 mcg/mL increase in free VPA (OR 1.098, 95% CI 1.077-1.118). A breakpoint of 19.95 mcg/mL was determined from an ROC curve as having 87% sensitivity and 81% specificity of predicting the incidence of thrombocytopenia.⁹¹

Platelet count declines as free VPA trough increases

What do I do with this information?

The bottom line here is VPA protein binding is highly variable and often discordant in ICU patients. While it is not unusual for protein binding characteristics of medications to be altered in critically ill patients, the degree to which VPA’s kinetics vary between (and within) patients is unusual. Despite being on the market for almost 50 years, there is still so much to learn about VPA and what drives its unique pharmacokinetic properties. Even its total therapeutic range of 50-100 mg/L was derived mostly from a good guess and seizure response, especially in status epilepticus, has not universally been associated with this range.⁹² Nevertheless, elevated VPA levels, especially elevated free VPA levels, are strongly linked with increased risk of serious toxicity and therapeutic drug monitoring is universal standard of care for patients treated for seizures with VPA. In critically ill patients especially, however, the free concentration of VPA does not follow expected patterns and cannot be reliably predicted from total concentration alone and should be directly measured in patients at risk for discordance. While an ideal free VPA therapeutic range has not been clearly defined, aiming to keep free VPA levels between 5-15 mg/L is likely prudent, and levels above 20-25 mg/L should prompt dose reduction and evaluation for toxicity. Direct measurement of free VPA requires specialized skills and equipment and thus is not readily available at most institutions, but hopefully increased awareness of free VPA importance and research on the clinical consequences of discordance will increase availability of the test.

Andrew Webb, PharmD, BCCCP

Clinical Pharmacist, Neurocritical Care

Massachusetts General Hospital

Ajwebb@mgh.harvard.edu

@AJWPharm

References

1. Rapeport W, Mendelow A, French G, et al. Plasma protein-binding and CSF concentrations of valproic acid in man following acute oral dosing. British Journal of Clinical Pharmacology. 1983;16(4):365-369. doi:10.1111/j.1365-2125.1983.tb02179.x

2. Wieser HG. Comparison of valproate concentrations in human plasma, CSF and brain tissue after administration of different formulations of valproate or valpromide. Epilepsy Research. 1991;9(2):154-159. doi:10.1016/0920-1211(91)90028-E

3. Smits JEMP, Wallenburg E, Spanje A van, Luin M van, Marijnissen RM. Valproate Intoxication in a Patient With Blood Valproate Levels Within Therapeutic Range. J Clin Psychiatry. 2017;78(4):22075. doi:10.4088/JCP.15cr10147

4. de Leon J, Kiesel JL, Fleming MW, Strobl B. Valproic Acid Toxicity Associated With Low Dose of Aspirin and Low Total Valproic Acid Levels: A Case Report. Journal of Clinical Psychopharmacology. 2009;29(5):509. doi:10.1097/JCP.0b013e3181b4b07c

5. Mayerhoff DI, Nurenberg J, Shah S, Schleifer SJ. Neurotoxicity Associated With Free Valproic Acid. AJP. 2005;162(4):810-810. doi:10.1176/appi.ajp.162.4.810

6. Gugler R, Schell A, Eichelbaum M, Fröscher W, Schulz HU. Disposition of valproic acid in man. Eur J Clin Pharmacol. 1977;12(2):125-132. doi:10.1007/BF00645133

7. Gugler R, Mueller G. Plasma protein binding of valproic acid in healthy subjects and in patients with renal disease. British Journal of Clinical Pharmacology. 1978;5(5):441-446. doi:10.1111/j.1365-2125.1978.tb01652.x

8. Gibbs HG, Zimmerman DE, Shermock KM, Clarke W, Mirski MA, Lewin JJ III. Comparison of free fraction serum valproic acid concentrations between inpatients and outpatients. American Journal of Health-System Pharmacy. 2015;72(2):121-126. doi:10.2146/ajhp140191

9. Riker RR, Gagnon DJ, Hatton C, et al. Valproate Protein Binding Is Highly Variable in ICU Patients and Not Predicted by Total Serum Concentrations: A Case Series and Literature Review. Pharmacotherapy. 2017;37(4):500-508. doi:10.1002/phar.1912

10. Fisch U, Baumann SM, Semmlack S, Marsch S, Rüegg S, Sutter R. Accuracy of Calculated Free Valproate Levels in Adult Patients With Status Epilepticus. Neurology. 2021;96(1):e102-e110. doi:10.1212/WNL.0000000000011000

11. Wallenburg E, Klok B, de Jong K, et al. Monitoring Protein-Unbound Valproic Acid Serum Concentrations in Clinical Practice. Therapeutic Drug Monitoring. 2017;39(3):269. doi:10.1097/FTD.0000000000000405

12. Drisaldi A, Weeda E, Neyens R, et al. Accuracy of Valproic Acid Concentration Correction Based on Serum Albumin. Neurocrit Care. 2019;30(2):301-306. doi:10.1007/s12028-018-0627-4

13. Brown CS, Liu J, Riker RR, et al. Evaluation of Free Valproate Concentration in Critically Ill Patients. Critical Care Explorations. 2022;4(9):e0746. doi:10.1097/CCE.0000000000000746

14. Methaneethorn J. A systematic review of population pharmacokinetics of valproic acid. British Journal of Clinical Pharmacology. 2018;84(5):816-834. doi:10.1111/bcp.13510

15. Otten N, Hall K, Irvine-Meek J, Leroux M, Budnik D, Seshia S. Free Valproic Acid: Steady-State Pharmacokinetics in Patients with Intractable Epilepsy. Canadian Journal of Neurological Sciences. 1984;11(4):457-460. doi:10.1017/S031716710004600X

16. Bellver MJG, Sánchez MJG, Gonzalez ACA, Buelga DS, Dominguez–Gil A. Plasma protein binding kinetics of valproic acid over a broad dosage range: therapeutic implications. Journal of Clinical Pharmacy and Therapeutics. 1993;18(3):191-197. doi:10.1111/j.1365-2710.1993.tb00612.x

17. Dutta S, Faught E, Limdi NA. Valproate protein binding following rapid intravenous administration of high doses of valproic acid in patients with epilepsy. Journal of Clinical Pharmacy and Therapeutics. 2007;32(4):365-371. doi:10.1111/j.1365-2710.2007.00831.x

18. Kodama Y, Tsutsumi K, Teraoka I, Fujii I, Takeyama M. Effect of Unbound Clearance on Binding Parameters of Valproic Acid to Serum Proteins. The Journal of Clinical Pharmacology. 1993;33(2):130-135. doi:10.1002/j.1552-4604.1993.tb03932.x

19. Yu HY. Clinical Implications of Serum Protein Binding in Epileptic Children During Sodium Valproate Maintenance Therapy. Therapeutic Drug Monitoring. 1984;6(4):414. doi:10.1097/00007691-198412000-00006

20. Cloyd JC, Dutta S, Cao G, Walch JK, Collins SD, Granneman GR. Valproate unbound fraction and distribution volume following rapid infusions in patients with epilepsy. Epilepsy Research. 2003;53(1):19-27. doi:10.1016/S0920-1211(02)00251-6

21. Cramer JA, Mattson RH. Valproic Acid: In Vitro Plasma Protein Binding and Interaction with Phenytoin. Therapeutic Drug Monitoring. 1979;1(1):105. doi:10.1097/00007691-197901000-00011

22. Kodama Y, Kuranari M, Tsutsumi K, Kimoto H, Fujii I, Takeyama M. Prediction of Unbound Serum Valproic Acid Concentration by Using In Vivo Binding Parameters. Therapeutic Drug Monitoring. 1992;14(5):349. doi:10.1097/00007691-199210000-00001

23. Kodama Y, Kuranari M, Kodama H, Ashikari Y, Fujii I, Takevama M. EVALUATION OF BINDING EQUATION METHOD FOR UNBOUND SERUM CONCENTRATION PREDICTION OF VALPROIC ACID IN POLYTHERAPY PEDIATRIC PA TIENTS WITH EPILEPSY. American Journal of Therapeutics. 1995;2(2):106. doi:10.1097/00045391-199502000-00005

24. Hermida J, Tutor JC. A Theoretical Method for Normalizing Total Serum Valproic Acid Concentration in Hypoalbuminemic Patients. Journal of Pharmacological Sciences. 2005;97(4):489-493. doi:10.1254/jphs.FPE04007X

25. Ueshima S, Aiba T, Makita T, et al. Characterization of non-linear relationship between total and unbound serum concentrations of valproic acid in epileptic children. Journal of Clinical Pharmacy and Therapeutics. 2008;33(1):31-38. doi:10.1111/j.1365-2710.2008.00885.x

26. Li Z, Gao W, Liu G, Chen W. Prediction of Serum-Free and Cerebrospinal Fluid Valproic Acid Levels in Patients With Hypoalbuminemia After Craniotomy. Therapeutic Drug Monitoring. 2020;42(4):610. doi:10.1097/FTD.0000000000000749

27. Xu S, Chen Y, Zhao M, Zhao L. Development of a Simple and Rapid Method to Measure Free Fraction of Valproic Acid in Plasma Using Ultrafiltration and Ultra High Performance Liquid Chromatography–Mass Spectroscopy: Application to Therapeutic Drug Monitoring. Therapeutic Drug Monitoring. 2017;39(5):575. doi:10.1097/FTD.0000000000000431

28. Nasreddine W, Dirani M, Atweh S, Makki A, Beydoun A. Determinants of free serum valproate concentration: A prospective study in patients on divalproex sodium monotherapy. Seizure. 2018;59:24-27. doi:10.1016/j.seizure.2018.04.012

29. Doré M, San Juan AE, Frenette AJ, Williamson D. Clinical Importance of Monitoring Unbound Valproic Acid Concentration in Patients with Hypoalbuminemia. Pharmacotherapy: The Journal of Human Pharmacology and Drug Therapy. 2017;37(8):900-907. doi:10.1002/phar.1965

30. Giner SC, Medall MDB, Piqueres RF. Design and validation of a predictive equation to estimate unbound valproic acid concentration. Eur J Hosp Pharm. Published online November 7, 2021. doi:10.1136/ejhpharm-2021-003092

31. Ishikawa M, Uchida M, Asakawa T, et al. A novel method for predicting the unbound valproic acid concentration. Drug Metabolism and Pharmacokinetics. 2023;50:100503. doi:10.1016/j.dmpk.2023.100503

32. Parent X, Marzullo C, Gutbub AM. [Valproic acid: a simple method for the estimation of free serum concentration]. Ann Biol Clin (Paris). 1993;51(6):649-650.

33. Scatchard G. The Attractions of Proteins for Small Molecules and Ions. Annals of the New York Academy of Sciences. 1949;51(4):660-672. doi:10.1111/j.1749-6632.1949.tb27297.x

34. Kober A, Olsson Y, Sjöholm I. Binding of Drugs to Human Serum Albumin: XIV. The Theoretical Basis for the Interaction between Phenytoin and Valproate1. Mol Pharmacol. 1980;18(2):237-242.

35. Monks A, Boobis S, Wadsworth J, Richens A. Plasma protein binding interaction between phenytoin and valproic acid in vitro. British Journal of Clinical Pharmacology. 1978;6(6):487-492. doi:10.1111/j.1365-2125.1978.tb00871.x

36. Patel IH, Levy RH. Valproic Acid Binding to Human Serum Albumin and Determination of Free Fraction in the Presence of Anticonvulsants and Free Fatty Acids. Epilepsia. 1979;20(1):85-90. doi:10.1111/j.1528-1157.1979.tb04779.x

37. Hatton C, Riker RR, Gagnon DJ, May T, Seder DB, Fraser GL. Free serum valproate concentration more reliable than total concentration in critically ill patients. Resuscitation. 2016;105:e15-e16. doi:10.1016/j.resuscitation.2016.05.027

38. Maat MM de, Leeuwen HJ van, Edelbroek PM. High Unbound Fraction of Valproic Acid in a Hypoalbuminemic Critically III Patient on Renal Replacement Therapy. Ann Pharmacother. 2011;45(3):420-420. doi:10.1345/aph.1P30a

39. Bowdle TA, Patel IH, Levy RH, Wilensky AJ. Valproic acid dosage and plasma protein binding and clearance. Clinical Pharmacology & Therapeutics. 1980;28(4):486-492. doi:10.1038/clpt.1980.192

40. Buckley MS, Reeves BA, Barletta JF, Bikin DS. Correlation of Free and Total Phenytoin Serum Concentrations in Critically Ill Patients. Ann Pharmacother. 2016;50(4):276-281. doi:10.1177/1060028015627468

41. Barra ME, Phillips KM, Chung DY, Rosenthal ES. A Novel Correction Equation Avoids High-Magnitude Errors in Interpreting Therapeutic Drug Monitoring of Phenytoin Among Critically Ill Patients. Ther Drug Monit. 2020;42(4):617-625. doi:10.1097/FTD.0000000000000739

42. Spector AA, Santos EC, Ashbrook JD, Fletcher JE. Influence of Free Fatty Acid Concentration on Drug Binding to Plasma Albumin*. Annals of the New York Academy of Sciences. 1973;226(1):247-258. doi:10.1111/j.1749-6632.1973.tb20486.x

43. Zimmerman CL, Patel IH, Levy RH, Edwards D, Nelson SD, Hutchinson M. Protein Binding of Valproic Acid in the Presence of Elevated Free Fatty Acids in Patient and Normal Human Serum. Epilepsia. 1981;22(1):11-17. doi:10.1111/j.1528-1157.1981.tb04328.x

44. Semmes RLO, Shen DD. Nonlinear Binding of Valproic Acid (VPA) and E-Δ2-Valproic Acid to Rat Plasma Proteins. Pharm Res. 1990;7(5):461-467. doi:10.1023/A:1015804413818

45. Klotz U, Rapp T, Müller WA. Disposition of valproic acid in patients with liver disease. Eur J Clin Pharmacol. 1978;13(1):55-60. doi:10.1007/BF00606683

46. Sadahiro N, Kodama S, Matsui T, Komatsu M, Matsuo T. Effect of serum albumin on free fractions of phenobarbital and valproic acid in patients with convulsive seizures. Brain and Development. 1985;7(4):377-384. doi:10.1016/S0387-7604(85)80134-0

47. Lagneau F, Perbet S, Delefosse D, Wernet A, Stocco J, Marty J. Drugs pharmacokinetics in ICU patients: consequences of hypoalbuminemia upon drugs monitoring and dosing scheme. Intensive Care Med. 2004;30(6):1247-1247. doi:10.1007/s00134-004-2313-6

48. Haroldson JA, Kramer LE, Wolff DL, Lake KD. Elevated Free Fractions of Valproic Acid in a Heart Transplant Patient with Hypoalbuminemia. Ann Pharmacother. 2000;34(2):183-187. doi:10.1345/aph.19147

49. Dasgupta A, Volk A. Displacement of Valproic Acid and Carbamazepine from Protein Binding in Normal and Uremic Sera by Tolmetin, Ibuprofen, and Naproxen: Presence of Inhibitor in Uremic Serum That Blocks Valproic Acid-Naproxen Interactions. Therapeutic Drug Monitoring. 1996;18(3):284. doi:10.1097/00007691-199606000-00011

50. Clinical and Pharmacological Aspects of Sodium Valproate (Epilim) in the Treatment of Epilepsy: Proceedings of a Symposium Held at Nottingham University 23/24 September, 1975. MCS Consultants for Reckitt-Labaz; 1976.

51. Patsalos PN, Lascelles PT. Valproate may lower serum-phenytoin. Lancet. 1977;1(8001):50-51. doi:10.1016/s0140-6736(77)91693-2

52. Perucca E, Hebdige S, Frigo GM, Gatti G, Lecchini S, Crema A. Interaction between phenytoin and valproic acid: Plasma protein binding and metabolic effects. Clinical Pharmacology & Therapeutics. 1980;28(6):779-789. doi:10.1038/clpt.1980.235

53. May TW, Rambeck B, Nothbaum N. Nomogram for the Prediction of Unbound Phenytoin Concentrations in Patients on a Combined Treatment of Phenytoin and Valproic Acid. ENE. 1991;31(1):57-60. doi:10.1159/000116647

54. Bowdle TA, Patel IH, Levy RH, Wilensky AJ. The influence of free fatty acids on valproic acid plasma protein binding during fasting in normal humans. European Journal of Clinical Pharmacology. 1982;23(4):343-347. doi:10.1007/BF00613618

55. Albani F, Riva R, Procaccianti G, Baruzzi A, Perucca E. Free Fraction of Valproic Acid: In Vitro Time-Dependent Increase and Correlation with Free Fatty Acid Concentration in Human Plasma and Serum. Epilepsia. 1983;24(1):65-73. doi:10.1111/j.1528-1157.1983.tb04867.x

56. Vorum H, Gram L, Honoré B. Valproate and palmitate binding to serum albumin in valproate-treated patients. relation to obesity. Epilepsy Research. 1993;16(1):55-64. doi:10.1016/0920-1211(93)90040-E

57. Ishii-Maruhama M, Higuchi H, Nakanou M, et al. In vitro changes in the proportion of protein-unbound-free propofol induced by valproate. J Anesth. 2018;32(5):688-693. doi:10.1007/s00540-018-2540-6

58. Warter JM, Marescaux C, Hirsch E, et al. Decrease of valproate-induced hyperammonemia in normal subjects by lipid ingestion. Journal of the Neurological Sciences. 1985;69(3):285-290. doi:10.1016/0022-510X(85)90140-6

59. Sandson NB, Marcucci C, Bourke DL, Smith-Lamacchia R. An Interaction Between Aspirin and Valproate: The Relevance of Plasma Protein Displacement Drug-Drug Interactions. AJP. 2006;163(11):1891-1896. doi:10.1176/ajp.2006.163.11.1891

60. Goulden KJ, Dooley JM, Camfield PR, Fraser AD. Clinical valproate toxicity induced by acetylsalicylic acid. Neurology. 1987;37(8):1392-1392. doi:10.1212/WNL.37.8.1392

61. Grimaldi R, Lecchini S, Crema F, Perucca E. In vivo plasma protein binding interaction between valproic acid and naproxen. European Journal of Drug Metabolism and Pharmacokinetics. 1984;9(4):359-363. doi:10.1007/BF03189686

62. Lana F, Martí-Bonany J, Fuster J, de Leon J. Reduction in serum concentration of valproic acid secondary to the intake of ibuprofen as an example of valproic acid auto-induction metabolism. Actas Esp Psiquiatr. 2016;44(4):136-144.

63. Fleitman JS, Bruni J, Perrin JH, Wilder BJ. Albumin-Binding Interactions of Sodium Valproate. The Journal of Clinical Pharmacology. 1980;20(8-9):514-517. doi:10.1002/j.1552-4604.1980.tb02544.x

64. Urien S, Albengres E, Tillement JP. Serum protein binding of valproic acid in healthy subjects and in patients with liver disease. Int J Clin Pharmacol Ther Toxicol. 1981;19(7):319-325.

65. Orr JM, Abbott FS, Farrell K, Ferguson S, Sheppard I, Godolphin W. Interaction between valproic acid and aspirin in epileptic children: Serum protein binding and metabolic effects. Clinical Pharmacology & Therapeutics. 1982;31(5):642-649. doi:10.1038/clpt.1982.89

66. Dasgupta A, Jacques M. Reduced in vitro displacement of valproic acid from protein binding by salicylate in uremic sera compared with normal sera. Role of uremic compounds. Am J Clin Pathol. 1994;101(3):349-353.

67. Min DI, Chen HY, Fabrega A, et al. CIRCADIAN VARIATION OF TACROLIMUS DISPOSITION IN LIVER ALLOGRAFT RECIPIENTS1. Transplantation. 1996;62(8):1190. doi:10.1097/00007890-199610270-00031

68. Loiseau P, Cenraud B, Levy RH, et al. Diurnal Variations in Steady-state Plasma Concentrations of Valproic Acid in Epileptic Patients. Clin Pharmacokinet. 1982;7(6):544-552. doi:10.2165/00003088-198207060-00004

69. Marty J, Kilpatrick C, Moulds R. Intra-dose variation in plasma protein binding of sodium valproate in epileptic patients. British Journal of Clinical Pharmacology. 1982;14(3):399-404. doi:10.1111/j.1365-2125.1982.tb01998.x

70. Patel IH, Venkataramanan R, Levy RH, Viswanathan CT, Ojemann LM. Diurnal Oscillations in Plasma Protein Binding of Valproic Acid. Epilepsia. 1982;23(3):283-290. doi:10.1111/j.1528-1157.1982.tb06193.x

71. Patients E, Riva R, Albani F, et al. Diurnal Fluctuations in Free and Total Plasma Concentrations of Valproic Acid at Steady State in. Therapeutic Drug Monitoring. 1983;5(2):191. doi:10.1097/00007691-198306000-00007

72. Bauer LA, Davis R, Wilensky A, Raisys V, Levy RH. Valproic acid clearance: Unbound fraction and diurnal variation in young and elderly adults. Clinical Pharmacology & Therapeutics. 1985;37(6):697-700. doi:10.1038/clpt.1985.116

73. Arens TL, Pollack GM. Nonstationary disposition of valproic acid during prolonged intravenous infusion: contributions of unbound clearance and protein binding. Biopharmaceutics & Drug Disposition. 2001;22(6):243-249. doi:10.1002/bdd.259

74. Roman E, Ponniah P, Lambert J, Buchanan N. Free sodium valproate monitoring. British Journal of Clinical Pharmacology. 1982;13(3):452-455. doi:10.1111/j.1365-2125.1982.tb01403.x

75. Wu X, Li H, Dong W, et al. Determination of Free Valproic Acid Concentration in 569 Clinical Samples by LC-MS/MS After Hollow Fiber Centrifugal Ultrafiltration Treatment. Ther Drug Monit. 2021;43(6):789-796. doi:10.1097/FTD.0000000000000903

76. Wang D, Champion-Lyons E, Neyens R, Bohm N, Caddell B, Babic N. The Effect of Sample Handling on Free Valproic Acid Levels. The Journal of Applied Laboratory Medicine. 2021;6(3):645-653. doi:10.1093/jalm/jfaa181

77. Loiseau P, Brachet A, Henry P. Concentration of Dipropylacetate in Plasma. Epilepsia. 1975;16(4):609-615. doi:10.1111/j.1528-1157.1975.tb04743.x

78. Barnes SE, Bower BD. Sodium Valproate in the Treatment of Intractable Childhood Epilepsy. Developmental Medicine & Child Neurology. 1975;17(2):175-181. doi:10.1111/j.1469-8749.1975.tb03469.x

79. Gugler R, von Unruh GE. Clinical Pharmacokinetics of Valproic Acid. Clin Pharmacokinet. 1980;5(1):67-83. doi:10.2165/00003088-198005010-00002

80. Chadwick DW. Concentration-Effect Relationships of Valproic Acid. Clin Pharmacokinet. 1985;10(2):155-163. doi:10.2165/00003088-198510020-00003

81. Beydoun A, Sackellares JC, Shu V. Safety and Efficacy of Divalproex Sodium Monotherapy in Partial Epilepsy: A Double-blind, Concentration-response Design Clinical Trial. Neurology. 1997;48(1):182-188. doi:10.1212/WNL.48.1.182

82. VALPF - Overview: Valproic Acid, Free, Serum. Accessed May 21, 2023. https://www.mayocliniclabs.com/test-catalog/overview/37068#Clinical-and-Interpretive

83. Valproic Acid, Free | Test Detail | Quest Diagnostics. Accessed May 21, 2023. https://testdirectory.questdiagnostics.com/test/test-detail/3314/valproic-acid-free?p=r&q=Valproic%20Acid,%20Free&cc=MASTER

84. Valproic Acid, Free and Total | ARUP Laboratories Test Directory. Accessed May 21, 2023. https://ltd.aruplab.com/Tests/Pub/0099310

85. 070789: Valproic Acid, Free, Serum or Plasma | Labcorp. Accessed May 21, 2023. https://www.labcorp.com/tests/070789/valproic-acid-free-serum-or-plasma

86. Farrell K, Abbott FS, Orr JM, Applegarth DA, Jan JE, Wong PK. Free and Total Serum Valproate Concentrations: Their Relationship to Seizure Control, Liver Enzymes and Plasma Ammonia in Children. Canadian Journal of Neurological Sciences. 1986;13(3):252-255. doi:10.1017/S0317167100036374

87. Kilpatrick CJ, Bury RW, Fullinfaw RO, Moulds RF. Plasma concentrations of unbound valproate and the management of epilepsy. Aust N Z J Med. 1987;17(6):574-579. doi:10.1111/j.1445-5994.1987.tb01259.x

88. Nakashima H, Oniki K, Nishimura M, et al. Determination of the Optimal Concentration of Valproic Acid in Patients with Epilepsy: A Population Pharmacokinetic-Pharmacodynamic Analysis. PLOS ONE. 2015;10(10):e0141266. doi:10.1371/journal.pone.0141266

89. Dautzenberg G, Nederlof M, Beekman A, Egberts T, Heerdink ER. Severe Cognitive Impairment Associated With a High Free But Therapeutic Total Concentration of Valproic Acid Due to Hypoalbuminemia in an Older Patient With Bipolar Disorder. Journal of Clinical Psychopharmacology. 2018;38(3):265. doi:10.1097/JCP.0000000000000872

90. Itoh H, Suzuki Y, Fujisaki K, Sato Y, Takeyama M. Correlation between Plasma Ammonia Level and Serum Trough Concentration of Free Valproic Acid in Patients with Epilepsy. Biological and Pharmaceutical Bulletin. 2012;35(6):971-974. doi:10.1248/bpb.35.971

91. Nasreddine W, Atweh SF, Beydoun AA, Dirani M, Nawfal O, Beydoun A. Predicting the occurrence of thrombocytopenia from free valproate levels: A prospective study. Seizure. 2022;94:33-38. doi:10.1016/j.seizure.2021.11.018

92. Vijiala S, André P, Buclin T, Decosterd LA, Rossetti AO, Novy J. Valproate in status epilepticus: Correlation between loading dose, serum levels, and clinical response. European Journal of Neurology. 2022;29(9):2607-2611. doi:10.1111/ene.15441

Search This Blog

NeuroWiseRx

Unbound and Uncharted - Therapeutic Drug Monitoring of Free Valproic Acid in ICU Patients