Ruchit Trivedi BioDuro
Up to 70 percent of active pharmaceutical ingredients (APIs) in today’s drug pipeline are poorly soluble, which complicates delivery and results in poor bioavailability. Organic solvents, such as dimethyl sulfoxide (DMSO), are toxic and incompatible with clinical use; however, researchers often use them in early drug discovery and preclinical studies to assess their lead candidates. The challenge of poor solubility is often not properly addressed until late in the development process, which can create potential delays for clinical studies.
To improve a drug’s kinetic solubility, formulators commonly use amorphous dispersions (amorphous API in a polymer matrix). Techniques to convert a crystalline API to amorphous form include spray dried dispersion (SDD) and hot melt extrusion (HME). Determining the right polymer to keep an API in a shelf-stable, non-crystalline form is traditionally a lengthy evaluation process that requires gram quantities of API. The ability to address solubility issues early can save costs and time and may rescue potentially life-saving compounds during development. This article describes a streamlined approach for quickly finding the best formulation while minimizing the amount of API required. The approach uses small-scale studies that test multiple excipient matrices in parallel and only require milligram quantities of API. These screenings prove to be a good indication of SDD or HME results. Once the screenings identify promising excipient matrices, they are scaled up to make spray dried dispersions and/or extrudates and combined with other solubilizing techniques to make the API into a usable dosage form to be studied in animals.
This approach integrates formulation and manufacturing with discovery, so parallel in vivo drug metabolism and pharmacokinetics (DMPK) studies can quickly give insights to the bioavailability of the candidate API formulations. This increases the reliability of the process and improves efficiency because it helps to enable better early decisions about whether to develop an API.
Case study
The following case study demonstrates this comprehensive approach for generating solid dosage forms for poorly soluble compounds. In this case study, a client had an oral antibacterial API candidate with very poor solubility and needed to develop an immediate-release tablet for Phase 1 clinical trials. By evaluating a broad set of solubilization approaches and correlating those formulations with in vivo and in vitro data, BioDuro developed an optimal and clinically relevant formulation, which enabled the program to move forward into clinical trials.
The API had a strong crystal structure and a very high melting temperature of 262°C. The first challenge was to determine the best solvent system to use, with the aim of achieving 10 mg/ml solubility. The API was extremely hydrophobic, with solubility of only 8 μg/ml in water, and was difficult to dissolve even in organic solvents. As shown in Table 1, the maximum solubility achieved was 5 mg/ml, using a 3-component solvent system containing tetrahydrofuran (THF), ethanol, and water (85:10:5) and requiring a specific order of addition for the solvent components.
The second challenge was stability. Even after making the API amorphous, we learned that it would immediately start to crystallize again and lose solubility in any formulation with more than 25 percent API. Creating a usable drug formulation with a stable shelf life required the right ratio of API and excipients to keep the API amorphous for at least two years. We used two small-scale screening methods—micro-evaporation and differential scanning calorimetry—to efficiently determine the best amorphous dispersion formulations for the API.
Micro-evaporation screening method for SDD polymer selection
We started with micro-evaporation studies to identify the polymer matrix that best enhanced solubility. Micro-evaporative dispersions present an innovative and efficient solution to overcoming low API solubility. This method requires only milligrams of API, and allows parallel evaluation of multiple approaches, including different excipients, proportions, and concentrations. Micro-evaporation is a small-scale screen that is a good predictor of SDD formulation. Results from micro-evaporation studies then can be leveraged into downstream development of the most suitable formulation to put on a spray dryer.
The protocol for the micro-evaporation screen starts with preparing API-polymer mixtures in solvent and drying them in a speedvac concentrator. We resuspended the dried films in pH 6.8 phosphate buffer by vortexing and transferred test samples to 96-well plates to determine dissolution performance by UV absorbance. Figure 1 shows this workflow and the results for API alone, compared to three candidate matrices: 25:75 API:HPMCAS-MF, 25:75 API:Kollidon VA64, and 25:75 API:Kollidon VA64 with 1 percent TPGS.
All three matrices showed multi-fold solubility improvement over the API alone. Kollidon VA64 is a better polymer backbone for the matrix compared to HPMCAS-MF, and addition of the surfactant TPGS further improved API solubility. We then used the polymers with the best dissolution performance for SDD.
Differential scanning calorimetry screening method for HME polymer selection
For APIs with thermal stability above 200°C, HME is a cost-effective method for making amorphous disper- sions. To identify the ideal polymer for the HME technique, we used a different screen. Differential scanning calorimetry (DSC) with a heat-cool-heat cycle allowed us to determine the heat capacity and miscibility of API-polymer mixtures, using only minimal API.
API and polymer were melted and mixed in the DSC pan, then cooled. The API-polymer mixture is then melted again to measure heat capacity. The matrix with the lowest heat capacity has the lowest residual crystallinity and the most miscibility, which indicates that the API-polymer will stay as a homogeneous dispersion. Table 2 shows the melt peaks of the API with or without different matrices in this heat-cool-heat DSC analysis. 25:75 API:Kollidon VA64 and 25:10:65 API:TPGS:Kollidon VA64 showed the lowest heat capacity and the best miscibility, making them the best candidate amorphous dispersions for HME with this API.
We then scaled up select formulations that performed well in our screens and made amorphous dispersions using SDD and HME for in vitro and in vivo assessments.
Assessment of formulations in vitro
To test for kinetic solubility enhancement of each candidate formulation, we performed a non-sink dissolution study. We made super-saturated mixtures of API-polymer powder in 1 milliliter of fasted-state simulated intestinal fluid (FaSSIF) and measured how long each formulation could stay in solution. Non-sink dissolution can give you an idea of a formulation’s pharmacokinetics because better kinetic solubility, as measured here, should allow for increased intestinal drug absorption in vivo.
As shown in Figure 2, the results indicated that 25:10:65 API:TPGS:Kollidon VA64 SDD and 25:10:65 API:TPGS:Kollidon VA64 HME showed improved solubility compared to the API by itself. The SDD showed a 3-fold higher increase in dissolution than the extrusion, while the extrusion still showed 30-fold better dissolution than the API alone. Spray drying resulted in higher solubility for about 10 minutes before crystallization in solution.
To characterize and confirm the amorphous structure of these API formulations, we used powder X-ray crystallography and DSC. As the X-ray diffraction patterns in Figure 3 show, the API is crystalline, while spray drying with polymers HPMCAS-L or Kollidon VA64 results in amorphous dispersions. As shown in Figure 4, DSC of these formulations showed no clear melting point, which is additional evidence of amorphous structure. DSC also provides the glass transition temperature, an indication of how stable the amorphous dispersion is. Pharmacokinetic assessment of formulations in vivo
In collaboration with our discovery colleagues, we rapidly generated pharmacokinetic data that correlated well with the non-sink dissolution results. Figure 5 shows the study design comparing the bioavailability of formulations using various solubility enhancement techniques: SDD, HME, and micro-evaporation. Organic solvent (PEG) and nano-suspensions were included as controls to show maximum bioavailability, but neither method is suitable for clinical use.
The results presented in Table 3 show that the SDD using polymer Kollidon VA64 with TPGS gave similar performance as the organic solvent. HME with the same excipients did not perform as well. However, the micro-evaporation tubes showed similar results to the SDD, underscoring how this screening method can provide an affordable option for early pharmacokinetic studies.
Case study conclusions
Success of an insoluble API depends on swift assessment of solubility challenges and rapid development of an enabling formulation. Using our solubility enhancement expertise, we were able to identify the best solubilization technology and the best formulation to scale up to produce clinical materials. Our initial screening studies only used milligram quantities of API to rapidly guide polymer selection.
Using solubilization techniques, we were able to formulate poorly soluble API and generate in vivo data that was representative of the final clinical formulation. We used this early data for the final GMP formulation—spray drying API mixed with Kollidon VA64 and TPGS surfactant, using the THF-ethanol-water mixture as spray drying solvent—and were able to scale up the process and successfully manufacture tablets for clinical studies.
Ruchit Trivedi, PhD, is director of formulation at BioDuro, a contract research, development, and manufacturing organization based in San Diego, CA (858 529 6600, www.bioduro. com). He holds a bachelor’s degree in pharmacy from North Gujarat University, India, and a doctorate in pharmaceutical sciences from the University of Colorado. Prior to joining BioDuro, he held positions at Array Biopharma as a formulation scientist developing amorphous-dispersion-based solid oral formulations and at Banner Life Sciences (formerly Banner Pharmacaps) developing abuse-deterrent formulations.
BioDuro’s approach for generating solid dosage forms for poorly soluble compounds, called Solution Engine, has led to a 100 percent success rate for solving difficult formulations. We do this by integrating API information from clients, micro-evaporative screening for optimal matrix selection, using our deep experience in amorphous dispersion manufacturing, and in vivo expertise of our in-house DMPK discovery team. All these elements are part of a single workflow—the BioDuro Solution Engine—which both accelerates timelines and rescues programs by developing clinically acceptable solid oral dosage forms.