Investigation of Excipient Compatibility and Associated Degradations for the Formulation Development of a Small Molecule Pharmaceutical Compound

Introduction

Drug formulation starts with the selection of various suitable excipients for the desired or required dosage forms. Drug-excipient interactions can greatly impact pharmaceutical development on multiple aspects, including drug properties (eg, polymorphic forms), manufacturability and processes, drug product stability, and bioavailability, which in turn affects product efficacy and safety.^1-7 Excipient compatibility studies are therefore frequently carried out to assess the potential physical and chemical interactions between a drug and excipients, to predict potential compatibility issues and to provide the mechanistic understanding of the incompatibilities such as chemical degradations and active pharmaceutical ingredient (API) form changes. Results and knowledge from the excipient compatibility studies can be very useful in mitigating potential risks arising from the incompatibilities and guiding drug product development from selection of excipients to defining manufacturing processes and packaging configurations.^3,5,8

Excipient compatibility issues derive from either the direct interactions between a drug and excipients in the formulation or from the presence of reactive impurities in the excipients.^9-12 Common excipients used in solid dosage form development contain a variety of reactive impurities such as reducing sugars, peroxides, aldehydes, and trace metals. There have been extensive studies and reviews on chemical degradations caused by these impurities.^5,13-15

For the current study, the API was a small molecule compound with molecule weight less than 500 and pKa’s of 2.7 and 9.2. The API by itself had good long-term stability under normal storage conditions, but had significant compatibility issues when blended with commonly used immediate release solid dosage form excipients. Our studies revealed that the primary amine group with pKa of 9.2 in the structure was chiefly responsible for these issues. This article describes our effort in conducting excipient compatibility studies, and provides details on analytical characterization for mechanistic insights into these issues and development of mitigation strategies.

Compatibility Issue with Croscarmellose Sodium (ccS)

CCS is widely used in solid dosage formulations as a super disintegrant. CCS at 5%w/w was included in the blend for the excipient compatibility study. However, HPLC analysis results showed ~89% w/w recovery of the API for samples with 20% drug loading and ~70% recovery for samples with 5% drug loading. Further inspections of sample compositions and HPLC results showed that the sample without CCS at 20% drug loading had no recovery issue, suggesting CCS may be responsible for the low recovery. Additional HPLC analysis of the CCS-API (1:1, w/w) binary mixture and the API standard solution spiked into the placebo matrix confirmed that the low recovery issue was indeed caused by CCS. Since CCS is of ionic nature and the API is partially ionized in the HPLC diluent of methanol/water (40/60, v/v), it was hypothesized that the API conjugated with CCS through ionic interactions, forming the API-croscarmellose salt with reduced solubility. The low recovery suggested that this salt likely had precipitated out of the solutions; however, the precipitation could not easily be noticed in the initial analysis of the formulation blend samples because of other insoluble materials.

To verify this hypothesis, sodium chloride was dissolved in the HPLC diluent at up to 1M to enhance the ionic strength of the diluent. HPLC analysis of samples containing CCS and the API using the modified diluent showed no recovery issue, confirming that the enhanced ionic strength was able to break up the ionic interactions (“ion pairing”) between CCS and the API. For routine analysis of formulation samples containing CCS, the diluent for HPLC analysis is modified to 20mM NaCl in methanol/water (40/60, v/v) and has worked well in achieving the full recovery of the API with CCS in the matrix. Alternatively, the pH of the original methanol/water (40/60, v/v) diluent was adjusted to <2 with 0.1N hydrochloric acid (HCl). This modification was also capable of overcoming the low recovery issue due to CCS-API conjugation, mainly because of the enhanced API solubility under the acidic condition. Results of both approaches demonstrated that the API-CCS compatibility issue is not expected to have negative impact on in vivo dissolution of this immediate release drug product considering the low stomach pH and the bio-salts in the physiological system.

Additional experiments were carried out to illustrate whether the API-CCS conjugation occurs only in solution. API-CCS binary solid mixtures were prepared at ratios (% w/w) of 20:80, 50:50, and 80:20 by direct blending to investigate the potential API-CCS conjugation in solid state. These solid mixtures were analyzed by Raman spectroscopy along with the API and CCS. Figure 1 is the overlaid Raman spectra for these five samples, showing no indication of the API crystalline form change in the presence of CCS. This provided some assurance that form change is unlikely to occur in formulated drug products using CCS. It remains to be investigated whether conjugation of API-CCS could occur during the granulation process that could have any impact on API crystalline forms.Zoom In

Figure 1. Overlaid Raman Spectra for the API, CCS, and Three Mixtures

Excipient compatibility Study Design and outcome

Besides CCS, crospovidone and hydroxypropylcellulose (HPC) were also evaluated as disintegrants for the formulation.Table 1 summarizes all the ingredients and compositions of the blends for the excipient compatibility study. The API level was at 5% for all the blends.

Table 1. Composition of Excipient Compatibility Samples

Blends 1-4 were prepared by directly blending all solid ingredients together, while water at 20% blend weight was added to blends 5 and 6 to gain early information about impact of wet processing. Once prepared, all blend samples were aliquoted into small glass scintillation vials and stored at either 50°C/75% relative humidity (RH) in open containers or 60°C in closed containers. These storage conditions were harsher than regular long-term or accelerated DP storage conditions, but are typical for excipient compatibility studies. The humidity level in the 60°C oven was not controlled but estimated to be ~40%. Freshly prepared samples and samples regularly pulled from the storage chambers were tested by HPLC for purity and degradations, shown in Table 2.

Table 2. Total Degradation (%area) for Blends 1–6

Zoom In

It is clear from Table 2 that no blend was completely free from degradation for the study duration and conditions. Severe

degradations were observed for blends 1, 4, and 5, all containing stearic acid. Besides the chemical degradations, these three blends also showed physical appearance changes from white powder to dark brown sticky material. These physical appearance changes were attributed to stearic acid, which has a relatively low melting point of ~70°C and may have partially melted under the stressed conditions. The partial melting likely enhanced the mobility of the materials in the blends and accelerated the degradation. The propensity for stearic acid to facilitate API degradation is an important consideration when defining the formulation processes and conditions.

Blends 2 and 3 had much less degradation than the others, especially for samples stored in closed containers. As an example shown in Figure 2, different degradation profiles were observed for open vs. closed containers. For samples stored in closed containers at 60°C, one major degradation product along with several minor ones was detected at 22.9 min (RRT ~1.76) after 4 weeks. For samples stored in open containers at 50°C/75%RH, however, two other degradation products were generated at higher levels than RRT ~1.76. There were also more degradation products detected for the 50°C/75%RH samples, strongly suggesting the detrimental effect of high humidity on the stability of the formulations.Zoom In

Figure 2. Chromatograms for Blend 2 at 4 Weeks

HPLC conditions: reversed-phase amide column, 150 × 4.6 mm, 3.5 μm, 30°C; mobile phase A = 0.1% acetic acid in water, pH 4.5; mobile phase B = mobile phase A : acetonitrile (40:60); gradient profile: 3-25%B over 12.5 min then ramp up to 99%B from 12.5 min to 32 min, hold at 99% B for 3 min before column reconditioning; 1.2 mL/min; UV detection at 235 nm.

Another important aspect of the data was about the potential impact of the manufacturing process on the drug product. Compositions of blends 2 and 6 were very similar; however, blend 6 showed significantly more degradations than blend 2. Since blend 6 was prepared with water added, the results showed this API to be likely unsuitable for water-based wet granulation processes. This conclusion was supported by the aforementioned degradation under high humidity conditions.

Understanding the Chemical Degradation

High-resolution mass spectrometry (HRMS) and multiple-stage mass spectrometry (MSⁿ) were applied to elucidate the structures of the main degradants and facilitate the understanding of the degradation mechanisms. Based on the HRMS and MSⁿ results, degradant I was proposed to be an N-oxide, and degradants II and III are two different dimers of the API. Their formation is all related to the reactivity of the primary amine group of the API.

Both degradants I and III were observed as major products in the API forced degradation studies under basic (0.1N NaOH) and oxidative (0.3% hydrogen peroxide) conditions (data not shown). Of all excipients used for these compatibility studies, crospovidone and HPC are known to contain high levels of peroxides.¹³ It is therefore reasonable to conclude that degradants I and III were probably generated in the formulation blends through peroxide oxidation of the primary amine group. The micro environmental pH within the blends may have played an important role on this degradation as well. The API has high μg/mL solubility in the unbuffered water, potentially resulting in a weakly basic solution. Under high humidity conditions, the dissolved API would create a basic microenvironment for the oxidation to proceed. In order to minimize this degradation and control the formation of degradants I and III, crospovidone and HPC were replaced by croscarmellose Na as the disintegrant for further formulation development. In addition, the formulation/drug product will need protection from moisture, for example, by packaging with desiccant or in blister pack.

Degradant II appears to be formed via a different pathway. It was not observed in any of the API forced degradation studies. However, a separate spiking experiment helped shed light on its formation. When an API solution was spiked with a small amount of formaldehyde, the main impurity in the spiked sample solution was found to be same as degradant II based on HPLC retention time and MS results (data not shown). The API has an active primary amine group, and can readily react with formaldehyde to form a hemiaminal and then imine through elimination of water before coupling with another API molecule to form the dimer degradant. This amine-aldehyde reaction mechanism has been well understood and reported by others.^5,11,13 Residual formaldehyde is present in many excipients for solid dosage forms including those used in this formulation, and will continue to be a concern for this product that needs to be closed monitored.

Conclusions

Well-designed and executed excipient compatibility studies are the essential step in developing successful formulations and robust drug products. Excipient compatibility studies were conducted to support the immediate release solid dosage formulation development of a small molecule API. The studies encountered a number of challenges due to incompatibilities of the API and the commonly used excipients, and identified potential major formulation risks because of the extensive chemical degradation. Data from the excipient compatibility studies have laid a good foundation for further development for the selections of suitable excipients, formulation composition, and appropriate primary packaging for long-term shelf-life of the product.

Author Biographies

Dr. Jane Li is currently an analytical chemist in the Small Molecules Pharmaceutical Sciences Department of Genentech. She obtained her PhD in Analytical Chemistry from Iowa State University, and was a research investigator at GlaxoSmithKline before joining Genentech. She has 15 years of pharmaceutical industry experience in the analytical research and development area.

Hong Lin received his MS degree from the University of Illinois at Urbana Champaign. He is a senior research associate in the Small Molecule Analytical Chemistry and Quality Control Department at Genentech, Inc. He has over 10 years of analytical research and development experience in the pharmaceutical industry.

Dr. Christine Gu is currently a scientist in the Department of Small Molecule Pharmaceutical Science at Genentech. She is a subject matter expert in mass spectrometry. Before joining Genentech, she worked as a senior application scientist at Thermo Fisher Scientific. She holds a PhD degree in Toxicology at University of California, Riverside.

Priscilla Mantik received her BS degree in chemistry from the University of California, Los Angeles, in 1997. Currently, she works as a Senior Research Associate in the Small Molecule Pharmaceutics group at Genentech. She has taken the role of a lead formulator for several projects at different stages of development from preclinical to late stage, developing parenteral and oral solid dosage forms.

Stefanie Gee is currently a Research Associate in the Small Molecules Pharmaceutical Sciences Department of Genentech. She obtained her BS degree in Chemistry and Biochemistry from California Polytechnic State University, San Luis Obispo, CA. She has 7 years of pharmaceutical industry experience in the analytical research and development area.

Dr. Peter M. Yehl received his PhD in Analytical Chemistry from the University of Massachusetts (Amherst) in 1997 and has worked in the pharmaceutical industry for 17 years. His career started at Merck, where he worked in support of synthetic process development and scaling at the Rahway, NJ, site and later in Hoddesdon, England. He retransitioned to the US and to take a leadership position supporting early phase formulation development at Merck in West Point, PA, and later directed CMC Analytical Research and Development at PTC Therapeutics in South Plainfield, NJ. He relocated to the Bay area in 2011, taking on an analytical leadership role in Small Molecule Analytical Chemistry and QC at Genentech in South San Francisco, CA, in 2011.

References

Narang AS, Yamniuk AP, Zhang L, et al. J Pharm Pharmacol. 2012;64(4):553-565.
Gohel MC, Patel TM. J Therm Anal Calorim. 2013;111:2103-2108.
Jannin V, Rodier J, Musakhanian J. Int J Pharm. 2014;466(1-2):109-121.
Bharate SS, Bharate SB, Bajaj AN. J Excipients and Food Chem. 2010;1(3):3-26.
Narang AS, Desai D, Badawy S. Pharm Res. 2012;29(10):2660-2683.
Ravikiran P, Narang AS. Pharm Res. 2012;29(10):2639-2659.
Jackson K, Young D, Pant S. Pharm Sci Tech Today. 2000;3(10):336-345.
Badawy SF, Lin J, Gokhale M, et al. Int J Pharm. 2014;469:111-120.
Monajjemzadeh F, Hasssanzadeh D, Valizadeh H, et al. Eur J Pharm Biopharm. 2009;73(3):404-413.
L. Gu L, Strickley RG, Chi LH, Chowhan ZT. Pharm Res. 1990;7(4):379-383.
Baertschi SW, Alsante KM. Stress testing: the chemistry of drug degradation. In: Baertschi SW, ed. Pharmaceutical Stress Testing: Predicting Drug Degradation. London, England: Informa Healthcare; 2007.
Franolic JD, Lehr, GJ, Barry TL, Petzinger G. J Pharm Biomed Anal. 2001;26: 651-663.
Wu Y, Levons J, Narang AS, Raghavan K, Rao VM. AAPS PharmSciTech. 2011;12(4): 1248-1263.
Sims JL, Carrieira JA, Carrier DJ, et al. Pharm Dev Tech. 2003;8(2):119-126.
Wasylaschuk WR, Harmon PA, Wagner G, et al. J Pharm Sci. 2007;96(1):106-116