What Is the Best Strategy for Converting from Twice-Daily Divalproex to a Once-Daily Divalproex ER Regimen?

Ronald C. Reed; Sandeep Dutta

Disclosures

Clin Drug Invest. 2004;24(9) 

In This Article

Results

Table I shows the simulation results for predicted post-conversion VPA minimum concentration (Cmin), maximum concentration (Cmax) and the percentage of patients predicted to have plasma VPA Cmin >50 mg/L for three patient types (uninduced, induced and induced at higher divalproex doses) across four distinct conversion scenarios (immediate, delayed, stepwise and mixed conversion).

The mean plasma total VPA concentration-time profile following conversion from divalproex administered q12h to divalproex ER once a day is depicted in figure 1 (uninduced hypothetical patients) and in figure 2 and figure 3 (induced patient scenarios). The predicted variation from the mean plasma VPA concentration (± 1SD) is presented in panels a and b for each figure.

Mean plasma total valproic acid (VPA) concentration over time for a 625mg divalproex 12-hourly (q12h) regimen (solid black line) in the typical uninduced hypothetical patient converting to divalproex extended release (ER) 1500mg once daily (qd) using various conversion strategies. Variation around the mean (± 1SD) is depicted by error bars. Dotted horizontal lines at 50 and 100 mg/L refer to the usual therapeutic VPA concentration range. Panels b - d include the mean VPA (black through green solid lines) derived in panel a for comparison. (a) Immediate conversion (12 hours after last divalproex dose) to divalproex ER once daily (solid green line); (b) delayed conversion (24 hours after last divalproex dose) to divalproex ER once daily (dashed blue line); (c) stepwise conversion (750mg divalproex ER, morning and evening) to divalproex ER once daily (dashed red line); (d) mixed conversion (625mg divalproex morning + 750mg divalproex ER evening) to divalproex ER once daily (dashed pink line). Cmin = minimum VPA concentration.

Mean plasma total valproic acid (VPA) concentration over time for a typical induced patient, converting a 1250mg divalproex 12-hourly (q12h) regimen (solid black line) to divalproex extended release (ER) 3000mg once daily (qd) using various conversion strategies. Variation around the mean (± 1SD) is depicted by error bars. Dotted horizontal lines at 50 and 100 mg/L refer to the usual therapeutic VPA concentration range. Panels b - d include the mean VPA (black through green solid lines) derived in panel a for comparison. a) Immediate conversion to divalproex ER once daily (solid green line); (b) delayed conversion to divalproex ER once daily (dashed blue line); (c) stepwise conversion (1500mg divalproex ER, morning and evening) to divalproex ER once daily (dashed red line); (d) mixed conversion (1250mg divalproex morning + 1500mg divalproex ER evening) to divalproex ER once daily (dashed pink line). Cmin = minimum VPA concentration.

Mean plasma total valproic acid (VPA) concentration over time for an induced patient on a high divalproex dose, converting a 2000mg divalproex 12-hourly (q12h) regimen (solid black line) to divalproex extended release (ER) 4500mg once daily (qd) using various conversion strategies. Variation around the mean (± 1SD) is depicted by error bars. Dotted horizontal lines at 50 and 100 mg/L refer to the usual therapeutic VPA concentration range. Panels b - d include the mean VPA (black through green solid lines) derived in panel a for comparison. (a) Immediate conversion to divalproex ER once daily (solid green line); (b) delayed conversion to divalproex ER once daily (dashed blue line); (c) stepwise conversion (2250mg divalproex ER, morning and evening) to divalproex ER once daily (dashed red line); (d) mixed conversion (2000mg divalproex morning + 2250mg divalproex ER evening) to divalproex ER once daily (dashed pink line). Cmin = minimum VPA concentration.

Overall, visual inspection reveals the mean plasma VPA concentration-time profile following conversion to be smooth and nearly flat for the uninduced patient when using immediate conversion (solid black line changing to a solid green line, figure 1a), and is also smooth even with higher divalproex dosages used in patients on concomitant enzyme-inducing AEDs (solid black to green lines, figures 2a and 3a). Smoothness refers to VPA Cmax being reduced and Cmin being slightly increased, while average concentration (Cavg) remains similar for the divalproex ER once-daily regimens compared with the divalproex q12h regimens. This marked dampening in peak-trough (Cmax to Cmin) plasma VPA concentration swing with the divalproex ER once-daily dosage regimen, observed in all figures, was expected based on actual data previously obtained in healthy volunteers and patients with epilepsy.[16,17]

Computer simulations suggest that patients on a conventional divalproex q12h regimen can be directly and immediately converted (no greater than a 12-hour gap) to a divalproex ER once-daily regimen without a clinically significant perturbation in plasma total VPA concentration occurring in the first 48 hours of conversion (as depicted by the change from solid black to solid green lines in figures 1a, 2a and 3a). The mean VPA concentrations generally stay within the usual or targeted therapeutic range with this immediate, 'all-at-once' conversion strategy. It should be noted that the conventional, enteric-coated divalproex formulation is designed to produce coverage for only 12 hours in patients who are uninduced (a shorter duration of coverage is anticipated in divalproex-treated patients who are induced unless they are on much higher doses).

As shown in Table I , for patients who immediately change to a once-daily divalproex ER regimen 12 hours after the last divalproex q12h dose, Cmin values >50 mg/L are expected to be maintained for 90% of both uninduced and induced patients on typical divalproex doses, and in 99% of induced patients on the higher (4500mg) once-daily divalproex ER dose. The predicted upswing, in the mean or typical patient, from VPA Cmin to Cmax values within the first 48 hours after the immediate conversion strategy is low, amounting to only 12-22 mg/L, most likely because of the slow, controlled release with subsequent apparent zero-order absorption rate documented for this ER formulation.[18]

The alternative conversion strategies utilised (as represented in panels b-d of figures 1-3, superimposed on the mean VPA concentration-time profile predicted for immediate divalproex q12h to once-daily divalproex ER conversion [data from panel a] for comparison) produced slightly to markedly greater perturbation in plasma total VPA concentrations in the first 48 hours after conversion, as compared with the immediate conversion strategy. The most notable and exaggerated perturbation predicted using alternative conversion strategies was the decline in VPA concentration likely to occur in patients subjected to a delayed conversion strategy, i.e. those who are instructed to wait a full 24 hours after their last dose in a divalproex q12h regimen to convert to divalproex ER once daily (figures 1b, 2b and 3b, and Table I ). As expected for a conventional formulation designed to provide coverage for only 12 hours, each scenario shows that delayed conversion (waiting 24 hours to convert divalproex to divalproex ER) produces marked, potentially clinically significant falls in plasma VPA concentration. For the induced patient converted to divalproex ER 3000mg once daily via a delayed conversion strategy, the mean plasma VPA Cmin value is predicted to fall to 43 mg/L ( Table I ); the Cmin of an induced patient who is 1SD away from the mean or typical induced patient may be as low as ~25 mg/L (figure 2b). The corresponding variability for Cmin in the induced patient on a divalproex q12h regimen undergoing delayed conversion dose is high (43% coefficient of variation).

Another method to examine the impact of the decline in plasma total VPA concentrations secondary to various conversion strategies is to predict the percentage of patients who have total VPA Cmin maintained at a reasonable clinical value, i.e. >50 mg/L. These values for each patient-dosage scenario across four conversion strategies are listed in Table I . While VPA Cmin values >50 mg/L are predicted to occur in at least 90% of uninduced and induced patients undergoing immediate conversion, the percentage of patients predicted to maintain VPA Cmin values >50 mg/L is above 82% for those following a stepwise or mixed conversion, but is predicted to be remarkably less for patients instructed to follow a delayed conversion strategy. For those undergoing a delayed conversion strategy, only 52% of the uninduced patients, a mere 33% of the induced patients converting from a divalproex 1250mg q12h dose, and 56% of induced patients converting from the higher divalproex 2000mg q12h dose are expected to maintain VPA Cmin values >50 mg/L.

For the delayed conversion strategy, the VPA Cmin to Cmax upswing after conversion (or amount and/or rate of VPA concentration change) appears to be much greater than that observed visually with other conversion strategies (figures 1b, 2b and 3b). Whether this rate and/or extent of change in total VPA concentrations could result in a greater risk of clinical toxicity in some patients is not clear. In the mixed conversion regimen, the predicted mean Cmax value achieved, 129 mg/L, was highest for induced patients taking larger doses (figure 3d), but the VPA Cmax attained during this conversion strategy was actually no different from the steady-state Cmax observed with divalproex q12h. Assuming some adverse effects of valproate are related to peak total VPA concentrations (such as tremor or nausea), the risk of clinical toxicity in patients with a mixed conversion strategy seems to be equivalent to a regular regimen of divalproex q12h. Overall, it appears that the risk of generating high plasma VPA Cmax values, with a corresponding risk of clinical toxicity, is low with any conversion strategy tested in our study.

It should be noted that all alternative conversion strategies would eventually produce an average steady-state plasma total VPA concentration similar to that achieved with an immediate, all-at-once conversion (simulation data not shown here). We have elected to show the perturbations in steady-state total VPA concentrations that would be predicted to occur in the important first 48 hours after conversion, rather than examine the time period beyond that, when perturbations in plasma VPA concentrations would be expected to be dampening.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....