Divalproex to Divalproex Extended Release Conversion

Sandeep Dutta; Ronald C. Reed


Clin Drug Invest. 2004;24(9) 

In This Article

Divalproex Extended Release (ER) versus Divalproex Bioavailability Assessments

The bioavailability of divalproex ER relative to divalproex has been evaluated in five multiple-dose studies.[11,12,13,15,16] The study designs, pharmacokinetic parameters and equivalence assessments are presented in Table I , Table II and Table III . The mean plasma VPA concentration-time profiles are shown in Figure 1. Studies 1, 2 and 3 evaluated mg/mg total daily dose conversion from divalproex to divalproex ER. These studies revealed that divalproex ER bioavailability was about 11% lower than that of divalproex.

Mean plasma valproic acid concentration vs time profiles comparing equal and unequal total daily dose comparisons. ER = extended release; qd = once daily; qxh = every x hours.

Since the bioavailability of divalproex ER relative to divalproex was approximately 0.89, in order to achieve equivalent AUC while switching a patient from divalproex to divalproex ER, the divalproex ER total daily dose may have to be increased by approximately 12% (i.e. 1.0/0.89 = 1.12). Since divalproex ER dosage strengths (250 and 500mg) are not 12% higher than divalproex dosage strengths (125, 250 and 500mg), the bioavailability of 8-20% higher divalproex ER daily doses was compared with corresponding divalproex total daily doses in studies 4 and 5 ( Table I , Table II and Table III ). In studies 4 and 5, 8-20% higher divalproex ER daily doses were found to be equivalent, based on AUC, to corresponding divalproex total daily doses. A meta-analysis of data from all five studies confirmed the divalproex ER/divalproex relative bioavailability to be 0.89.[17]

Divalproex is available as 125, 250 and 500mg tablets. Therefore, the possible divalproex total daily dosage regimens are in increments of 125mg. Table IV shows that all divalproex dosage regimens, 625 mg/day and above, can be converted to a divalproex ER regimen with 8-20% higher doses, except for the 750 and 1000 mg/day divalproex regimens. Table IV also presents the predicted average increment in AUC, Cmax and Cmin that can be expected while converting patients from a divalproex regimen to the next available 8-20% higher divalproex ER regimen. The predicted increments were calculated by multiplying the divalproex ER/divalproex dose ratio with the divalproex ER/divalproex dose-normalised AUC, Cmax and Cmin ratios of 0.89, 0.79 and 0.96 (from meta-analysis),[17] respectively. The divalproex to divalproex ER dose conversion guidelines can be simplified as shown in Table V .

For the 750 and 1000mg total daily divalproex regimens, 33% and 25% dose increments while converting from divalproex to the next available higher divalproex ER regimen are expected to result in 19% and 11% increments, respectively, in AUC on average. More importantly, these dose increments are not expected to substantially increase the Cmax; a 5% increase for the 750mg total daily divalproex dose to 1000mg divalproex ER conversion and 1% decrease for the 1000mg total daily divalproex dose to 1250mg divalproex ER are expected on average.

The divalproex ER formulation was designed to produce a 'smoother' concentration-time profile with lower Cmax and higher Cmin, and consequently lower fluctuation compared with the conventional divalproex formulation. Therefore, the consistently lower divalproex ER/divalproex Cmax ratio from different studies is not surprising.

A major concern while increasing the dose in a patient is the potential adverse events that may be precipitated as a result of increased exposure as measured by AUC and Cmax. These five bioavailability studies were powered to establish pharmacokinetic equivalence for AUC and not to detect differences in safety or efficacy parameters between the divalproex and divalproex ER formulations. The lower Cmax values observed for the divalproex ER formulation does not necessarily indicate that divalproex ER will have a better adverse-effect profile than the divalproex formulation. In general, for drugs that have direct relationships between concentration and adverse events, greater numbers of adverse effects are observed with higher Cmax values. Although sustained-release VPA formulations have been purported to reduce adverse effects,[4] in the case of valproate therapy, it appears that not all adverse events are related to Cmax. However, for adverse drug effects that are likely to be Cmax related, the lower Cmax of the 8-20% higher daily dose divalproex ER formulation is likely to be beneficial to the patient.

Based on our predictions, no major increase in VPA Cmax is expected for the conversion from 750 or 1000mg total daily divalproex dose to the next available higher once-daily divalproex ER regimen. However, for the 500mg divalproex to 750mg divalproex ER conversion, a 50% dose increment, an increment in AUC of 34% and Cmax of 19% can be expected on average. In this particular case, the clinician should consider the necessity of conversion from the conventional to the divalproex ER formulation, and the risk/benefit of a small reduction in bioavailability versus up to a 50% increase in daily dose that would be expected from such a manoeuvre, particularly if the switch were to be performed to decrease adverse events. Therefore, if a conversion to a once-daily regimen is deemed necessary, a cautious approach of first converting the patient to the next higher total daily divalproex dose regimen (625mg for the 500mg, 875mg for the 750mg, and 1125mg for the 1000mg divalproex regimens) and evaluation of safety of the higher divalproex regimen might be considered by the clinician. If there are no safety issues, these patients may be subsequently converted to the next higher available divalproex ER regimen.

The effect of food on divalproex ER/divalproex relative bioavailability was minimal. For extended-release formulations, the effect of food can be of great importance, since food-related changes in gastrointestinal conditions (e.g. motility, pH) might affect the extended-release properties of the formulation, possibly leading to dose dumping. Previous studies have demonstrated that food does not affect the extent of absorption (AUC) of VPA from conventional divalproex tablets; only time to Cmax is delayed.[18] The effect of food on VPA pharmacokinetics following divalproex ER administration can be divided into two categories: the effect of food on extent of absorption (AUC) and the effect of food on bioavailability of divalproex ER relative to other VPA products. The effect of food on the extent of absorption following divalproex ER administration is <10%.[11]

If food has an effect on extent of absorption of VPA from divalproex ER, but not divalproex, food may have an effect on divalproex ER/divalproex relative bioavailability. An optimal study design for evaluating the effect of food on relative bioavailability is a comparison of divalproex ER versus divalproex under fasting conditions, as well as under nonfasting conditions. This would entail a 4-way randomised, crossover study where each formulation is administered under fasting and nonfasting conditions to each participant. Such a study was not performed. Although not as rigorous an assessment, the effect of food on relative bioavailability can be inferred by performing a meta-analysis of the available bioavailability data.

To evaluate the effect of food on relative bioavailability (F, AUC ratio) data from different subsets of bioavailability studies with the two formulations administered under different meal conditions were analysed. Relative bioavailability was estimated for the three studies (studies 1, 4 and 5) that had the two formulations administered under fasting conditions and for the two studies (studies 2 and 3) that had the two formulations administered under nonfasting conditions.

The point estimate (95% CI) from fasting divalproex ER versus fasting divalproex comparison was 0.895 (0.870, 0.920), similar to the estimate of 0.870 (0.741, 1.021) from nonfasting divalproex ER versus nonfasting divalproex comparison. Excluding study 3 data (where a true nonfasting condition was not used, as both divalproex ER and divalproex [morning dose only] were given immediately prior to a low-calorie breakfast) provides a nonfasting divalproex ER/nonfasting divalproex AUC ratio of 0.944 (0.898, 0.993), which again is not substantially different from the above values. Therefore, it might be concluded that the effect of food on divalproex ER/divalproex relative bioavailability is minimal.

The frequency of divalproex administration has no effect on divalproex ER/divalproex AUC ratios or the peak-trough fluctuation in plasma VPA concentrations. The effect of divalproex dose frequencies (i.e. q6h vs q8h vs q12h) on the AUC ratio and peak-trough fluctuation was evaluated by comparing data from different studies.

In theory, the divalproex regimen dose frequency would not be expected to have any effect on the AUC ratio, as the total daily dose and the elimination clearance, which are independent of dose frequency, are the only determinants of the AUC ratio. The point estimates for the AUC ratios ( Table III ) for divalproex ER once daily versus divalproex q6h (study 2), q8h (study 5) and q12h (studies 1 and 4) were 0.944, 0.873, 0.891 and 0.905, respectively. These AUC ratios were, in general, similar across different divalproex regimens with an average value of approximately 0.90.

Based on first principles, the degree of peak-trough fluctuation (DFL) in plasma VPA concentrations for the divalproex regimens would be expected to be lowest for the q6h regimen and highest for the q12h regimen for any given total daily dose. For an extended-release formulation given once daily to offer any advantage in terms of patient convenience, better safety profile due to reduced peak concentrations and/or better or equivalent efficacy due to increased Ctrough values, the extended-release formulation would be expected to have, at a minimum, lower peak-trough fluctuation than the conventional formulation given once daily or q12h. Divalproex ER not only meets this minimum requirement, but also produces about 20% less peak-trough fluctuation than q8h and q6h divalproex regimens as is evident from the DFL values presented in Table II . For the less rigorous comparison with q12h divalproex regimens, once-daily divalproex ER produces up to 50% less peak-trough fluctuation in plasma VPA concentrations. Therefore, the peak-trough fluctuations were consistently lower for the divalproex ER regimens, irrespective of the frequency of the divalproex regimens.