Steady State Plasma Levels of Bupropion After Administration of 3x150 mg Extended Release Reference Tablets and Switching to 1x450 mg Extended Release 450ER Tablets

Key words

Bupropion, Pharmacokinetics, ER Formulation, Steady State, Simulations.

Introduction

Bupropion extended release tablets are administered for the treatment of major depressive disorders. Initially, patients suffering from depression are commonly treated with lower doses and the daily dose is increased until the desired therapeutic effect is observed. Bupropion hydrochloride was initially developed as an immediate release tablet that was administered three-times per day. It was later developed into a twice a day tablet (Wellbutrin SR®) followed by a once-daily tablet (Wellbutrin XL®) [3-4]. Patients receiving the once-per-day bupropion product may require high daily doses; Bupropion HCl is approved for a maximum daily dose of 450 mg. However, because the commercial product is only available as 150 and 300 mg tablets, patients requiring 450 mg have to take multiples of the available lower strength tablets

Bupropion hydrochloride 450 mg extended-release tablets (450ER), was developed to allow patients to take one tablet as opposed to multiple tablets, hence eliminating the risk of an accidental dosing errors. In a single dose, three-period, three-treatment, cross-over study, it was demonstrated that 450ER (1 x 450 mg tablet) was bioequivalent to REF (3 x 150 mg tablets) when dosed in the fasting state. However, when 450ER tablets were administered with a standard high fat/high calorie meal and compared to the fasting treatment, Cmax was bioequivalent but AUC increased slightly (25%) and didn’t meet the 80 – 125% bioequivalence (BE) confidence interval criteria (CI).

However, this increase of AUC after administration with food is routinely observed in bupropion extended-release products. REF shows a fed/fasting ratio of 110% (CI: 104-116%) for AUC while Cmax was lower (point estimate: 92 %). Similarly, Aplenzin® (Bupropion hydrobromide extended-release tablets containing the same dose of bupropion as 450ER tablets), show a ratio fed/fasting of 107.48% (CI: 98.20-117.64%) for Cmax and a ratio fed/fasting of 119.47% (CI: 113.23-126.06%) for AUCτ [5]. Both products are labeled indicating that they may be taken with no regard to food or meal timing.

It is known from the history of bupropion HCl [1-2] that high Cmax values may be linked to higher seizure risk. A reduction in the frequency of reported seizures occurred after the marketing transition from Bupropion immediate release to Bupropion twice a day and Bupropion extended release [3]. Thus, it was hypothesized that an increase in bupropion exposure with 450ER could result in an increase in the frequency of seizures. This concern was directed at those patients who would be switched from the current commercial bupropion 3x150mg to 450ER. PK modeling and simulation provide the context needed to compare these two oral formulations. Steady-state simulations may provide insight into the steady-state PK of bupropion in patients receiving REF tablets (3x150 mg) and who may be switched to the 450ER tablets (taken with food). One goal was to predict potential increases in steady-state Cmax or AUC values which might indicate that patients would experience a higher risk of developing seizures.

Materials and Methods

Clinical studies

Two single dose clinical studies form the basis of these simulations. The first study (Study 1) was a three-way cross over study in healthy volunteers comparing Forfivo XLTM 450mg (Bupropion hydrochloride extended-release tablets, 450ER, Edgemont Pharmaceutical LLC, manufactured by Pillar5 Pharma Inc. in Ontario Canada) under fasting and fed condition to Wellbutrin XL® 3 x 150 mg (REF, Glaxo Smith Kline Inc., Manufactured by DSM Pharmaceutical Inc. North Carolina, USA) under fasting condition only. This study was conducted on 36 healthy subjects. The second study (Study 2) was a four-way cross over comparing these same products both under fasting and fed conditions. This study was conducted on 20 healthy subjects, except the fed portion of REF which was conducted on 10 subjects. The protocols for both studies were reviewed and approved by investigational review boards. Bioanalyses for both studies were conducted using a validated LC/MS/MS procedure.

Steady State Simulations

Steady-state simulations were conducted to better understand the steady-state PK of bupropion in patients receiving REF (3x150 mg) and who may be switched to the 450ER tablets (taken with food). The goal was to predict potential increases in Cmax or AUCτ values which might indicate that patients would experience a higher risk of developing seizures.

Single dose of REF data generated in the four-way crossover showed a delayed Tmax and a lower Cmax. This lower exposure would support a safer product since the assumption is that seizures are associated with elevated Cmax. In keeping with this hypothesis, the publishedseizure rate for bupropion sustained-release formulations decreased to 0.1% at doses of up to 300 mg/day [4].

Steady-state plasma concentration versus time profiles were simulated with the Nonparametric Superposition (NPS) feature within WinNonlin® Pro (Pharsight Corporation, St. Louis, MO) using individual single dose profiles obtained from the study

2. Steady-state plasma concentration-time profiles were simulated for each individual subject.

In order to assess the effect of switching patients from REF 3x150mg to 450ER, steady-state plasma concentration versus time profiles (after 8 daily doses) were simulated for each individual subject. These simulations included additional points beyond the last 8th dose in order to provide the concentrations 288 hrs. Single dose profiles for 450ER were simulated using NPS to allow manual superposition of these single dose in order to provide(Day 9) with the steady-state profiles (Days 8 to 9).

Pharmacokinetic parameters were derived for each subject’s simulated profile using noncompartmental modeling in WinNonlin Pro.

Treatment Assumptions

For the switch assessment (switching from REF 3x150mg to 450ER on Day 9), it was assumed that REF steady-state was achieved by the 8th daily dose. A visual assessment of Panel A in Figure 1A shows that steady-state was achieved by Days 3 or 4 validating the assumption of steady-state achievement by Day 8.

---

---

---

---

Biostatistical Evaluation

To assess the equivalence of treatments, particularly after switching from REF 3x150 mg fasting to 450ER fed, 90% confidence intervals of the geometric means of the individual ratios of the simulated Cmax and AUCτ were calculated using the BE Wizard in WinNonlin Pro.

Results and Discussion

Figure 1A illustrates the mean concentration-time profiles for 450ER (fasting), 450ER (fed) and REF (fed).The two fasting treatments in this figure were nearly superimposable. This figure confirms that bupropion steady-state is achieved before Day 8. Because the extrapolated portions of the concentration-time curves are critical to the superposition approach, the fits were assessed by comparing the half lives derived from the nonparametric simulations with those determined from the noncompartmental analyses of the single dose data. The graphics provided in Figure 2 demonstrate excellent agreement between these half lives and confirm that the superposition approach provides reasonable estimates of steady-state concentrations

The mean (SD) graphs of the simulated steady-state concentration time profiles for each of the three treatments are provided as Figure 1B. This figure illustrates that REF provides tighter control of plasma concentrations with lower peak to trough fluctuations. Summary statistics for the simulated steady-state Cmax, Tmax and AUC values are provided in Table 1. Additionally Table 1 provides the corresponding 90% confidence intervals of the ln-transformed individual ratios of Cmax and AUCτ for 450ER (fed and fasting) relative to REF (fasting).

---

---

The point estimates and confidence intervals from the steady-state simulations were similar to thosereported in the single dose study where 450ER(fasting) was equivalent to REF (fasting).The Test/Ref point estimates for steady state AUC and Cmax were 96.63% and 94.58%, respectively. The Test/Ref point estimates from the single dose BE study were 94.94% and 96.51%, respectively and confidence intervals fell within 80.00% to 125.00%. For the 450ER (fed) to REF (fasting) comparison, steady-state Cmax values were equivalent, but AUC was high, similar to the single dose study. The Test Fed/Ref point estimates for steady state AUC and Cmax were 122.37% and 101.19%, respectively; the point estimates from the single dose BE study were 124.89% and 101.41%, respectively. The confidence intervals for simulated steady-state AUC and Cmax were 113.94% to 131.41% and 92.02% to 111.28%, respectively. The similarity in the statistics serves to validate these simulations. The point estimates and confidence intervals from this simulation, as listed in Table 2, predict that both

---

---

Summary Statistics for Simulated Bupropion PK Parameters for REF after 8 Days of dosing and for 450ER after dosing on Day 9, and Ln transformed data. Ratio (% Ref ) and 90% confidence intervals for AUClast and Cmax

Cmax and AUC were bioequivalent.That is, on the initial day when a patient is switched from REF (3x150 mg, while fasting) to 450ER (450 mg with food),the predicted Cmax and AUC values will be bio -equivalent. The Test/Ref point estimates for AUC and Cmax were 101.01% and 94.62%, respectively; the confidence intervals for AUC and Cmax were 90.36% to 112.90% and 82.86% to 108.05%, respectively.

Conclusion

The results show that the 90% confidence intervals for ln-transformed AUCtau and Cmax values on Day 9 (after switching to 450ER) relative to day 8 (steady-state REF before switching) are within the 80 – 125 % bioequivalence acceptance range. This predicts that switching from a steady-state dosing regimen of 3x150 mg REF tablets to 450ER taken with food will not result in significantly increased AUC or Cmax levels. These simulations support the equivalence of 450ER and REF and that a switch from REF in patients who take the product in the fasting state to 450ER if taken with food (the worst case scenario) does not significantly increase bupropion exposure and therefore, does not expose patients to elevated risk of seizures. When patients switch from REF to 450ER fasting, the two products would also be bioequivalent as demonstrated by the single-dose study.

Although the professional labeling for REF states that food did not affect the Cmax or AUC of bupropion, data generated in the single dose study demonstrated that there is a small food effect that is similar in these two products (REF and 450ER). Observed Cmax and AUC values are provided in Table 3. Figure 3 illustrates the mean bupropion concentrations obtained after administration of REF when fasting or after a standard FDA meal. The Test/Ref (Fed/Fasting) point estimates for REF bupropion AUC and Cmax were 125.30% and 92.08%, respectively. In the most recent 450ER study (Study 1), the Fed/Fasting

---

---

---

---

bupropion point estimates for AUC and Cmax were 124.89% and 101.41%, respectively. Thus, the extent of absorption for both 450ER and REF increases by 25% when the products are dosed with food, relative to the fasting state. Given the safety experience with REF, this increase in exposure is not clinically relevant.