Development of a simple liquid chromatography method for dissolution testing of low dose desogestrel and ethinylestradiol tablets

https://doi.org/10.1016/j.chroma.2020.461535Get rights and content

Highlights

  • Cyano stationary phase for isocratic separation of desogestrel and ethinylestradiol.

  • Large injection volume on a small column without overloading.

  • UV detection for both analytes versus fluorescence and UV in compendial monograph.

  • DOE for method robustness and establishing operational design region.

Abstract

Accurate quantitation of low dose, multi-active dissolution samples poses unique challenges in the pharmaceutical industry, often resulting in separate HPLC methods for each active or the use of multiple detectors for increased sensitivity. In this study, we report a fast, isocratic HPLC method utilizing only UV detection for dissolution testing of low dose desogestrel and ethinylestradiol tablets. Rapid separation is completed in 5 min using isocratic elution at a flow rate of 0.45 mL/min, with a column temperature at 30 °C, an injection volume of 50 μL and the detection wavelength at 200 nm. After extensive method development and optimization, the cyano stationary phase was used to overcome the large difference in hydrophobicity for desogestrel and ethinylestradiol, providing balanced retention for both analytes under isocratic elution. Chromatography modeling software was used to provide a rapid analysis of multiple columns and chromatography conditions. The optimized method boasts fast and efficient separation through use of a short, small I.D. column and a large injection volume of dissolution solution to achieve high sensitivity. The stable baseline from an isocratic separation allows low detection wavelengths to be used, resulting in accurate and precise quantitation of both desogestrel and ethinylestradiol. The method has been successfully validated for specificity, linearity, accuracy and precision in the range of 75 – 600 ng/mL for desogestrel and 10 – 80 ng/mL for ethinylestradiol using both HPLC and UHPLC systems. The method robustness was characterized using a design of experiment approach, and the operational design region of the method was established.

Introduction

Desogestrel (DES) is a type of progestin that is used in hormonal contraception for women. It can be used alone or in combination with an estrogen such as ethinylestradiol (EE) for contraceptive medication [1], [2], [3], [4], [5]. The structures of desogestrel and ethinylestradiol are shown in Fig. 1.

Due to coagulation caused by ethinylestradiol and other estrogens and the increasing risk of thromboembolic complications, dosages of estrogens have decreased since the inception of contraceptives [1], [6], [7], [8]. For example, in the combinatorial tablet developed at our company, the doses of desogestrel and ethinylestradiol are only 0.15 mg and 0.02 mg per tablet, respectively [2], [9], [10], [11].

To understand the in vivo drug release behavior, dissolution testing is performed at release and throughout stability. During dissolution testing, one tablet is added to 500 mL or 900 mL of dissolution media with appropriate agitation to release the active pharmaceutical ingredient into the dissolution media, which is then analyzed using an HPLC method [11]. Sample solutions are pulled at multiple timepoints, commonly 15, 30, 45, and 60 min, to study the rate of dissolution of each compound [11].

There are two major challenges to develop an HPLC method to accurately quantitate desogestrel and ethinylestradiol in a dissolution solution. First, the concentrations of desogestrel and ethinylestradiol in the dissolution solution (500 mL per compendial method) are 0.30 and 0.04 µg/mL, respectively, at full release, making it difficult to quantitate both accurately and precisely. Furthermore, full release is not expected at early timepoints, making accurate and precise quantitation at even lower concentrations a necessity [12]. Second, due to the variability of individual tablets and the nature of dissolution testing, each sample has to be tested with 6 replicates at multiple time points [10]. Because of this, isocratic HPLC separation is preferred as it does not require column re-equilibration, and thus increases throughput. However, isocratic separation of desogestrel and ethinylestradiol is difficult due to their different hydrophobicities (logP values of 5.40 for desogestrel and 3.67 for ethinylestradiol, calculated using ACD/Labs Percepta GALAS algorithm (Toronto, Canada)) as shown in Fig. 1. As a control for column screening, we also include desogestrel impurity B in this study (also shown in Fig. 1, logP value of 4.97 calculated using ACD/Labs Percepta GALAS algorithm).

To overcome these challenges, the method described in the current compendial monograph employs an isocratic separation but with two detection techniques, UV for desogestrel and fluorescence for ethinylestradiol [7], [11]. Fluorescence detection is used in the compendial method because ethinylestradiol elutes near the solvent front (k < 2.0) and is susceptible to interference from matrix peaks also eluting near the void (excipients, dissolution media, etc.) when UV detection is used [13]. Although fluorescence detection offers high selectivity and sensitivity, it requires the use of a costly additional detector on top of the universally standard LC configuration that typically uses only one type, UV detection. Furthermore, the requirement of two detection methods increases the complexity of the method and the burden on supply sites to maintain this instrumentation and analysis long-term.

Herein, we report the development and optimization of an isocratic HPLC-UV method for fast, accurate, and precise quantitation of both ethinylestradiol and desogestrel in dissolution samples of Mercilon tablets. Method validation including accuracy, linearity, precision, specificity, and robustness will be discussed.

Section snippets

Reagents and materials

Methanol and acetonitrile (optima grade) and sodium dodecyl sulfate (SDS) were purchased from Fisher Scientific (Waltham, MA, USA). Ethinylestradiol and desogestrel reference standards, desogestrel impurity B, and Mercilon tablets were provided by Merck & Co., Inc. (Kenilworth, NJ, USA).

List of HPLC Columns: The following columns were used for initial screening: 1) Zorbax SB C18 (3.5 µm, 4.6 × 150 mm), 2) Zorbax CN (3.5 µm, 4.6 × 150 mm), 3) Zorbax Phenyl (3.5 µm, 4.6 × 150 mm), 4) Ascentis

Column screening to minimize relative retention of desogestrel versus ethinylestradiol

As shown in Fig. 1, desogestrel and ethinylestradiol have very different hydrophobicities, with logP values of 5.40 for desogestrel and 3.67 for ethinylestradiol, respectively. If these compounds are separated on a reversed phase column where the retention mechanism is dominantly hydrophobic interactions, their retention will be very different. This will result in either too much retention of desogestrel or too little retention of ethinylestradiol using an isocratic separation. To achieve

Conclusions

A simple HPLC method with UV absorbance detection was successfully developed and validated for the quantitation of desogestrel and ethinylestradiol in dissolution samples for desogestrel and ethinylestradiol tablets. Balanced retention of ethinylestradiol and desogestrel was obtained using a cyano stationary phase, which enables isocratic separation to improve the throughput. The sensitivity was improved by the combination of a large injection volume and a small dimension column. The flow rate,

CRediT authorship contribution statement

Luis A. Jimenez: Investigation, Validation, Writing - original draft, Writing - review & editing. Orane White: Investigation, Validation. Eric Luther: Investigation. Christine L. Kirkpatrick: Investigation, Writing - original draft, Writing - review & editing. Jinjian Zheng: Conceptualization, Methodology, Writing - original draft, Writing - review & editing. Xiaoyi Gong: Supervision. Xihui Liang: Supervision. Panos Diamandopoulos: Supervision. Leah Buhler: Project administration. Mark D.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to thank Grace Chen and Michael Haider from Merck & Co., Inc., West Point, PA, USA and Merck & Co., Inc., Rahway, NJ, USA, respectively, for critical review of manuscript and other colleagues for sharing of instrumentation and invaluable discussions.

References (23)

View full text