^{1}

^{*}

^{1}

^{1}

^{2}

This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution license (http://creativecommons.org/licenses/by/3.0/).

A novel abuse deterrent, prolonged release tablet formulation of Hydrocodone for once-daily dosing has been developed, based on the novel proprietary Egalet® ADPREM technology. The tablet is an injection molded polymer system consisting of an erodible matrix in which the Active Pharmaceutical Ingredient (API), such as Hydrocodone, is dispersed. The matrix is partly covered with a water-impermeable, non-erodible shell which leaves both ends of the cylindrical tablet exposed to erosion by the gastrointestinal (GI) fluid.

Historically, Hydrocodone has been used primarily as cough medicine [

In recent years, Hydrocodone has become the most commonly prescribed opioid and is available in over 200 different products in the United States. Hydrocodone is prescribed as either antitussive or analgesic for treating moderate to moderately severe pain [

Pharmacodynamically Hydrocodone resembles codeine and morphine, being approximately equipotent to the latter [

Oral Hydrocodone is known from marketed combination products containing low doses of 5 to 10 mg Hydrocodone co-formulated with acetaminophen or ibuprofen. These products are, however, immediate release (IR) products which may be taken every three to six hours. In a clinical setting, this is not optimal for around the clock pain therapy.

A novel abuse deterrent, prolonged release tablet of Hydrocodone for once-daily dosing has been developed based on the novel proprietary Egalet® ADPREM technology (Abuse Deterrent Prolonged Release Erodible Matrix technology). The formulation is an injection molded polymer system consisting of an erodible matrix in which Active Pharmaceutical Ingredient (API) like Hydrocodone is dispersed. The matrix is partly covered with a water-impermeable, non-erodible shell which leaves both ends of the cylindrical tablet exposed to erosion by the gastrointestinal (GI) fluid. The tablet strength is defined by the initial API loaded into the tablet. The release rate is controlled by the well-defined, fixed size of the surface erosion area at both ends of the tablet. This allows a tightly controlled, prolonged release of the API which is only limited by the time of residence in the GI tract, the tablet geometry and composition. In addition, the tablet is designed to maintain its prolonged-release properties across a wide range of solvents, and is resistant to physical and chemical attempts to alter the slow-release of the API contained in the tablet.

The purpose of the current paper is to present how IVIVC were used: (i) to develop and optimize a new Hydrocodone Prolonged Release (PR) tablet based on the novel ADPREM technology; and (ii) to fix limits of dissolution in order to be bioequivalent between the tablets and batches during scale up, but also in case of post approval change [

Hydrocodone exhibits a high solubility in aqueous solvents and a good permeability. It is extensively metabolized in liver by CYP2D6 into Hydromorphone, a more potent opioid.

Three tablets (A to C) were developed based on Egalet® technology, all of them containing 20 mg of Hydrocodone as tartrate salt. The three tablets are based on one common qualitative and quantitative composition of excipients and a fix mass of Hydrocodone per tablet. They differ solely by the mass of the final tablet, corresponding to different diameters and lengths of the tablet (6, 7.5 or 9 mm, respectively) and adjusted by an increase of excipients' mass. The main characteristics of the tablets are presented in

The tablet batches were prepared according to Good Manufacturing Practice (GMP) by two component injection molding, employing an Arburg Allrounder 420 V 800-60/35 injection molding machine (Arburg, Greve, Denmark). The geometrical shape of the formulations was defined by a custom made stainless steel mold (ATZ form, Copenhagen, Denmark).

An IR tablet containing 10 mg of Hydrocodone in combination with 325 mg of Paracetamol (NORCO® 10/325 immediate-release tablet (Watson Pharma, lot 102663A) was also included in the clinical study as a reference and in order to perform deconvolution.

The dissolution profiles of the three test tablets were realized using a Pharmacopoeia media: phosphate buffer pH 6.8, using a USP Apparatus 2 paddle method (Vankel VK7025 coupled to a Varian Cary 50 UV-visible spectrophotometer). The dissolution volume was 1,000 mL. The paddle speed and temperature were set at 50 rpm and 37 °C, respectively. Samples were withdrawn at predefined time intervals (15 min) up to 600 minutes in a close loop system and were directly filtered with a 70 μm probe filter and subsequently measured by UV on line detection.

A pharmacokinetic (PK) study based on a 4-arm, single dose, randomized cross-over design comparing the three test tablets to the reference formulations, NORCO®, was performed on 25 fasted subjects. The PK plasma samples were collected at predefined time intervals from 0 to 42 hours and measured by a validated HPLC-MS method. Classical bioavailability parameters were calculated: Maximum observed concentration and time to obtain it (Cmax and Tmax) and extent absorbed: area under the curve (AUC). As mentioned in the guidelines [

The absorption kinetics were calculated using a deconvolution technique using the IR reference tablet as response function according to the method described by Langenbucher [

The deconvolution technique was applied independently of any modelization of the absorption. Deconvolution allows isolating the input (« absorption ») function by a numerical algorithm as a function of the observed concentration for the studied tablet and for the IR reference tablet. In the current case this input function reflected the in vivo release observed after administration of the PR test tablets. The simulations of the curves from the theoretic input were performed using convolution [

The clinical study was designed to support a level A correlation. Level A correlation is “a point-to-point relationship between

Model predictability was estimated internally by comparison of prediction errors on pharmacokinetic parameters used to establish bioavailability (BA) and bioequivalence (BE): Cmax, Tmax and AUC, derived from mean observed and predicted

All calculations were done using Microsoft Excel. As the purpose of this paper is to demonstrate the feasibility to use IVIVC for the current tablets, all the calculations were performed on mean curves calculated based on all subjects (mean plasma time-concentration and mean of absorption kinetics) and on mean of the dissolution.

The result of dissolution analysis of the three test formulations is presented in

Based on the IR formulation, the deconvolution was assessed and provided the results presented in

The bioequivalence results on AUC and Cmax are presented in

The analysis of the data indicated that for the AUC, the ratio between the IR and PR tablets corrected by the dose were of 1.00, 1.0. and 0.93 for tablet formulations A, B and C respectively, and the coefficient of variation (CV) of the analysis of variance (ANOVA) was 7.4%. For Cmax, the ratios corrected by the dose were 0.49, 0.40 and 0.30 for tablet formulations A, B and C, respectively, and the CV of the ANOVA was 14%. The results indicate that the AUC of all formulations are bioequivalent but a difference could be observed on Cmax. The coefficients of variations, which were good estimates of the intra subject variability, were low, indicating that the IVIVC could be used to predict minimal variation between tablet formulations with a good discriminative power.

All tablet formulations were used to establish IVIVC; the resulting

The data show that a small lag time exists

Based on this IVIVC and on

The predictability was good and in accordance with the FDA recommendation (5) with a mean error of −0.32% and −6.63% on Cmax and AUCinf, respectively, no case being greater than + or −10%.

A common challenge of all pharmaceutical companies is to develop new products as fast as possible to cover unmet medical needs, and to ensure, at the same time, safety and efficacy. Many strategies exist and amongst them,

The IVIVC based on the current study could be used as the predictability was correct: (i) to optimize the tablet formulation; but also (ii) to help for regulatory purposes, like fixing dissolution limits, scale up and post approval changes. The findings are coherent with the biopharmaceutical class of Hydrocodone—Class 1 according to Biopharmaceutical Classification System BCS (high solubility, high permeability, [

One application of IVIVC is to predict bioavailability and to set dissolution limits. Based on the result of the BE study for the main parameters of interest, Cmax and AUC up to infinity, the residual error variance was extracted from the ANOVA and estimated to be 14 and 7.4% respectively (coefficient of variation). Using this value, in formulation B for example, the limits to have Cmax and AUC within the bioequivalence limits were estimated by

To establish the limits, the absorption was modelized according to a multi zero order absorption (

This modelization is necessary in order to calculate the increase or decrease in absorption rate resulting in a modification of AUC and Cmax within the limits of bioequivalence. A similar mechanism of absorption was used. Based on the results of the modelization of the absorption the mean error from the model was of −0.97% confirming the good accuracy of the model. These findings are in accordance with a scintigraphic study published employing an Egalet® matrix system loaded with caffeine [

Those findings confirmed the observed difference in release rate between

However, as the API cannot be absorbed more than 100%, the upper limit would reach 100% earlier. The lower limits for Cmax and AUC were of importance here to determine the lower dissolution limits. This slower dissolution would impact the rate of absorption but also the quantity released and absorbed as a part of the drug could be expelled in the feces and not absorbed as the release would be too slow. In our simulation, we fixed arbitrarily the maximum transit time and release limits to 24 h. The impact of a longer transit time (36 h) was investigated and led only to a marginal increase of AUC of less than 5% for the lower limit (slow tablet). The dissolution limits calculated based on this principle are presented in

Based on the dissolution limits and on the modelization of the absorption, the

The rapid release formulation A and the slower C are presented in

It can be observed that these two formulations A and C, corresponding to a faster and slower release, did not fall within the limits calculated based on formulation B results that confirm the absence of bioequivalence observed for Cmax (

Based on the results from the BE study, the dissolution limits to ensure BE of the Cmax and AUC of any tablet formulation

Based on the dissolution, it could be anticipated that new tablet formulations would be bioequivalent with the existing formula. The calculation of the 90% confidence interval (CI) gave the following results 0.89–0.97 and 1.01–1.06 for Cmax and AUC infinity, respectively. Thus the development of this new tablet formulation could simply have a definite target dissolution profile for the formulators, at a very low risk of BE study failure. Alternatively, the developed tablet could even be submitted as a new drug application based on a surrogate dissolution test only, as presented in

Another example on the possible usage of these limits is for use in the submission dossier, in accordance with the FDA note for guidance, for the scale up with a scaling factor greater than 10. As dissolution limits were calculated and ensure bioequivalence on the main PK parameters, if the dissolution observed in the full scale batch would be within the limits, no BE study would have to be carried out to support the scaling up. Similarly, any modification in the tablet within a predefined range for the excipients could be supported solely by dissolution as a BE surrogate.

A similar exercise could be performed for formulation A and C and for any formulations in between as a similar release mechanism and time scaling were observed for all three formulations.

The work presented in this paper emphasizes the importance of dissolution as a prediction technique for development and optimization of tablet formulation, but also as a quality control tool. The IVIVC was initiated and validated for three formulations. A good internal predictability was observed for all tablet formulations. The validated IVIVC could be used to optimize the formulation and to achieve a desired profile. In addition, this technique could help to establish the dissolution limits associated with a certainty of bioequivalence. Based on this validated level A, IVIVC dissolution could be used as surrogate of bioequivalence.

Egalet® ADPREM tablets.

Plasma concentration up to 24 hours post dose of the three test formulations (

Input kinetics of Hydrocodone after administration of the 3 test tablet formulations based on deconvolution using IR data,

Dissolution

Levy's plot for time scaling.

IVIVC established on all formulations using a non-linear time scaling common to all formulations.

Simulated mean curves obtained based on IVIVC (lines)

Modelization of a multi zero order absorption according to Equation 3.

Dissolution limits to ensure bioequivalence to formulation B.

Dissolution on a target formulation derived from formulation B to ensure decrease of Cmax maintains a plateau.

Main formulation characteristics.

Strength mg | 20 | 20 | 20 |

Length mm | 6 | 7.5 | 9 |

Short diameter mm | 4.7 | 4.4 | 4.4 |

Long diameter mm | 9.4 | 8.6 | 8.3 |

Average tablet mass mg | 240 | 257 | 314 |

Result of bioequivalence study (corrected by the dose).

A 6 mm | Reference | Ratio | 0.49 | 1.00 |

90% CI | 0.45–0.52 | 0.96–1.04 | ||

B 7.5 mm | Ratio | 0.40 | 1.03 | |

90% CI | 0.37–0.43 | 0.99–1.07 | ||

C 9 mm | Ratio | 0.30 | 0.93 | |

90% CI | 0.28–0.32 | 0.89–0.97 | ||

A 6 mm | B | Ratio | 1.20 | 0.97 |

90% CI | 1.12–1.30 | 0.93–1.01 | ||

C 9 mm | Ratio | 0.75 | 0.90 | |

90% CI | 0.70–0.81 | 0.87–0.94 |

Predictability based on AUC and Cmax of the mean curve.

Cmax pmol/L | AUC 0-42 pmol.h/L | AUC inf pmol.h/L | Cmax pmol/L | AUC 0-42 pmol.h/L | AUC inf pmol.h/L | Cmax % | AUC 0-42 % | AUC inf % | |

A | 70332 | 1016005 | 1021836 | 70449 | 1052679 | 1064817 | -0.17 | -3.48 | -4.04 |

B | 56421 | 1007388 | 1019908 | 60393 | 1083974 | 1107549 | -6.58 | -7.07 | -7.91 |

C | 48839 | 902842 | 918345 | 46166 | 960346 | 997534 | 5.79 | -5.99 | -7.94 |

Mean predictability | -0.32 | -5.51 | -6.63 |