Comparative Analysis of Chemical Descriptors by Machine Learning Reveals Atomistic Insights into Solute-Lipid Interactions.

This study explores the research area of drug solubility in lipid excipients, an area persistently complex despite recent advancements in understanding and predicting solubility based on molecular structure. To this end, this research investigated novel descriptor sets, employing machine learning techniques to understand the determinants governing interactions between solutes and medium-chain triglycerides (MCTs). Quantitative structure-property relationships (QSPR) were constructed on an extended solubility data set comprising 182 experimental values of structurally diverse drug molecules, including both development and marketed drugs to extract meaningful property relationships. Four classes of molecular descriptors, ranging from traditional representations to complex geometrical descriptions, were assessed and compared in terms of their predictive accuracy and interpretability. These include two-dimensional (2D) and three-dimensional (3D) descriptors, Abraham solvation parameters, extended connectivity fingerprints (ECFPs), and the smooth overlap of atomic position (SOAP) descriptor. Through testing three distinct regularized regression algorithms alongside various preprocessing schemes, the SOAP descriptor enabled the construction of a superior performing model in terms of interpretability and accuracy. Its atom-centered characteristics allowed contributions to be estimated at the atomic level, thereby enabling the ranking of prevalent molecular motifs and their influence on drug solubility in MCTs. The performance on a separate test set demonstrated high predictive accuracy (RMSE = 0.50) for 2D and 3D, SOAP, and Abraham Solvation descriptors. The model trained on ECFP4 descriptors resulted in inferior predictive accuracy. Lastly, uncertainty estimations for each model were introduced to assess their applicability domains and provide information on where the models may extrapolate in chemical space and, thus, where more data may be necessary to refine a data-driven approach to predict solubility in MCTs. Overall, the presented approaches further enable computationally informed formulation development by introducing a novel in silico approach for rational drug development and prediction of dose loading in lipids.

Machine Estimation of Drug Melting Properties and Influence on Solubility Prediction.

There has been much recent interest in machine learning (ML) and molecular quantitative structure property relationships (QSPR). The present research evaluated modern ML-based methods implemented in commercial software (COSMOquick and Molecular Modeling Pro), compared to a classical group contribution approach (Joback and Reid method), to estimate melting points and enthalpy of fusion values. A broad data set of market compounds was gathered from the literature, together with new data measured by differential scanning calorimetry for drug candidates. The highest prediction accuracy was achieved by QSPR using stochastic gradient boosting. The model deviations were discussed, particularly the implications on thermodynamic solubility modeling, as this typically requires estimation of both melting point and enthalpy of fusion. The results suggested that despite considerable advancement in prediction accuracy, there are still limitations especially with complex drug candidates. It is recommended that in such cases, melting properties obtained in silico should be used carefully as input data for thermodynamic solubility modeling. Future research will show how the prediction limits of thermophysical drug properties can be further advanced by even larger data sets and other ML algorithms or also by using molecular simulations.

Unexpected Solubility Enhancement of Drug Bases in the Presence of a Dimethylaminoethyl Methacrylate Copolymer.

The methacrylate copolymer Eudragit EPO (EPO) has previously shown to greatly enhance solubilization of acidic drugs via ionic interactions and by multiple hydrophobic contacts with polymeric side chains. The latter type of interaction could also play a role for solubilization of other compounds than acids. The aim of this study was therefore to investigate the solubility of six poorly soluble bases in presence and absence of EPO by quantitative ultrapressure liquid chromatography with concomitant X-ray powder diffraction analysis of the solid state. For a better mechanistic understanding, spectra and diffusion data were obtained by 1H nuclear magnetic resonance (NMR) spectroscopy. Unexpected high solubility enhancement (up to 360-fold) was evidenced in the presence of EPO despite the fact that bases and polymer were both carrying positive charges. This exceptional and unexpected solubilization was not due to a change in the crystalline solid state. NMR spectra and measured diffusion coefficients indicated both strong drug-polymer interactions in the bulk solution, and diffusion data suggested conformational changes of the polymer in solution. Such conformational changes may have increased the accessibility and extent of hydrophobic contacts thereby leading to increased overall molecular interactions. These initially surprising solubilization results demonstrate that excipient selection should not be based solely on simple considerations of, for example, opposite charges of drug and excipient, but it requires a more refined molecular view. Different solution NMR techniques are especially promising tools to gain such mechanistic insights.

Can we estimate the critical micelle concentration of amphiphilic drug bases from molecular connectivity indices?

Self-aggregation of drugs is since many years an important topic in the pharmaceutical sciences. Drugs can aggregate similar to surfactants and thereby exhibit a critical micelle concentration (CMC). The present work focused on amphiphilic drug bases and it was aimed to predict log(CMC) based on chemical structure alone. A dataset of 35 compounds was gathered mostly form the literature and complemented with own measurements based on ultrasonic resonator technology. The hydrophilic-lipophilic balance (HLB) values of the protonated bases were calculated and provided a range of 22.9-27.4. Based on a hypothesis from surfactant sciences, it was tried to predict log(CMC) with connectivity and shape indices as well as molecular dipole moment. A fairly good model was obtained using the Randix index (RI), 3 D Wiener number (WN) and molecular dipole moment (DM) (R2 = 0.824). Interestingly, a simple linear regression of log(CMC) with the Randic index alone, resulted in an acceptable model (R2 = 0.755). The present work should help with early identification of drug bases that exhibit surfactant-like behavior and an estimation of log(CMC) values is proposed. An improved understanding of drug aggregation and prediction of log(CMC) helps to better cope with physical consequences like, for example, "anomalous" drug solubility in drug discovery and development.

A Systematic Study of Molecular Interactions of Anionic Drugs with a Dimethylaminoethyl Methacrylate Copolymer Regarding Solubility Enhancement.

The methacrylate-copolymer Eudragit EPO (EPO) has raised interest in solubility enhancement of anionic drugs. Effects on aqueous drug solubility at rather low polymer concentrations are barely known despite their importance upon dissolution and dilution of oral dosage forms. We provide evidence for substantial enhancement (factor 4-230) of aqueous solubility of poorly water-soluble anionic drugs induced by low (0.1-5% (w/w)) concentration of EPO for a panel of seven acidic crystalline drugs. Diffusion data (determined by 1H nuclear magnetic resonance spectroscopy) indicate that the solubility increasing effect monitored by quantitative ultraperformance liquid chromatography was caused primarily by molecular API polymer interactions in the bulk liquid phase. Residual solid API remained unaltered as tested by X-ray powder diffraction. The solubility enhancement (SE) revealed a significant rank correlation (rSpearman = -0.83) with rDiffAPI, where SE and rDiffAPI are defined ratios of solubility and diffusion coefficient in the presence and absence of EPO. SE decreased in the order of indomethacin, mefenamic acid, warfarin, piroxicam, furosemide, bezafibrate, and tolbutamide. The solubilizing effect was attributed to both ionic and hydrophobic interactions between drugs and EPO. The excellent solubilizing properties of EPO are highly promising for pharmaceutical development, and the data set provides first steps toward an understanding of drug-excipient interaction mechanisms.

Theoretical Considerations of the Prigogine-Defay Ratio with Regard to the Glass-Forming Ability of Drugs from Undercooled Melts.

Drug behavior in undercooled melts is highly important for pharmaceutics with regard to amorphous solid dispersions, and therefore, categories were recently introduced that differentiate glass formers (GFs) from other drugs that are nonglass formers (nGFs). The present study is based on the assumption that molecular properties relevant for the so-called Prigogine-Defay (PD) ratio would be indicative of a drug's glass-forming ability. The PD ratio depends in theory on the entropy of fusion and molar volume. Experimental data were gathered from a broad set of pharmaceutical compounds (n = 54) using differential scanning calorimetry. The obtained entropy of fusion and molar volume were indeed found to significantly discriminate GFs from nGFs. In a next step, the entropy of fusion was predicted by different in silico methods. A first group contribution method provided rather unreliable estimates for the entropy of fusion, while an alternative in silico approach seemed more promising for drug categorization. Thus, a significant discrimination model employed molar volume, a so-called effective hydrogen bond number, and effective number of torsional bonds (or torsional units) to categorize GFs and nGFs (p ≤ 0.0000). The results led to new insights into drug vitrification and to practical rules of thumb. The latter may serve as guidance in pharmaceutical profiling and early formulation development with respect to amorphous drug formulations.

Study and Computational Modeling of Fatty Acid Effects on Drug Solubility in Lipid-Based Systems.

Lipid-based systems have many advantages in formulation of poorly water-soluble drugs but issues of a limited solvent capacity are often encountered in development. One of the possible solubilization approaches of especially basic drugs could be the addition of fatty acids to oils but currently, a systematic study is lacking. Therefore, the present work investigated apparently neutral and basic drugs in medium chain triglycerides (MCT) alone and with added either caproic acid (C6), caprylic acid (C8), capric acid (C10) or oleic acid (C18:1) at different levels (5 - 20%, w/w). A miniaturized solubility assay was used together with X-ray diffraction to analyze the residual solid and finally, solubility data were modeled using the conductor-like screening model for real solvents (COSMO-RS). Some drug bases had an MCT solubility of only a few mg/ml or less but addition of fatty acids provided in some formulations exceptional drug loading of up to about 20% (w/w). The solubility changes were in general more pronounced the shorter the chain length was and the longest oleic acid even displayed a negative effect in mixtures of celecoxib and fenofibrate. The COSMO-RS prediction accuracy was highly specific for the given compounds with root mean square errors (RMSE) ranging from an excellent 0.07 to a highest value of 1.12. The latter was obtained with the strongest model base pimozide for which a new solid form was found in some samples. In conclusion, targeting specific molecular interactions with the solute combined with mechanistic modeling provides new tools to advance lipid-based drug delivery.

Multiorgan Crystal Deposition of an Amphoteric Drug in Rats Due to Lysosomal Accumulation and Conversion to a Poorly Soluble Hydrochloride Salt.

Poor solubility of drug candidates mainly affects bioavailability, but poor solubility of drugs and metabolites can also lead to precipitation within tissues, particularly when high doses are tested. RO0728617 is an amphoteric compound bearing basic and acidic moieties that has previously demonstrated good solubility at physiological pH but underwent widespread crystal deposition in multiple tissues in rat toxicity studies. The aim of our investigation was to better characterize these findings and their underlying mechanism(s), and to identify possible screening methods in the drug development process. Main microscopic features observed in rat RO0728617 toxicity studies were extensive infiltrates of crystal-containing macrophages in multiple organs. Matrix-assisted laser desorption/ionization Fourier transform ion cyclotron resonance mass spectrometry revealed that these crystals contained the orally administered parent compound, and locality was confirmed to be intracytoplasmic and partly intralysosomal by electron microscopic examination. Crystal formation was explained by lysosomal accumulation of the compound followed by precipitation of the hydrochloride salt under physiological conditions in the lysosomes, which have a lower pH and higher chloride concentration in comparison to the cytosol. This study demonstrates that risk of drug precipitation can be assessed by comparing the estimated lysosomal drug concentration at a given dose with the solubility of the compound at lysosomal conditions.

Partial Solvation Parameters of Drugs as a New Thermodynamic Tool for Pharmaceutics.

Partial solvation parameters (PSP) have much in common with the Hansen solubility parameter or with a linear solvation energy relationship (LSER), but there are advantages based on the sound thermodynamic basis. It is, therefore, surprising that PSP has so far not been harnessed in pharmaceutics for the selection of excipients or property estimation of formulations and their components. This work introduces PSP calculation for drugs, where the raw data were obtained from inverse gas chromatography. It was shown that only a few probe gases were needed to get reasonable estimates of the drug PSPs. Interestingly, an alternative calculation of LSER parameters in silico did not reflect the experimentally obtained activity coefficients for all probe gases as well, which was attributed to the complexity of the drug structures. The experimental PSPs were proven to be helpful in predicting drug solubility in various solvents and the PSP framework allowed calculation of the different surface energy contributions. A specific benefit of PSP is that parameters can be readily converted to either classical solubility or LSER parameters. Therefore, PSP is not just about a new definition of solvatochromic parameters, but the underlying thermodynamics provides a unified approach, which holds much promise for broad applications in pharmaceutics.

New prediction methods for solubility parameters based on molecular sigma profiles using pharmaceutical materials.

Solubility parameters have been applied extensively in the chemical and pharmaceutical sciences. Particularly attractive is calculation of solubility parameters based on chemical structure and recently, new in silico methods have been proposed. Thus, screening charge densities of molecular surfaces (i.e. so-called σ-profiles) are used by the conductor-like screening model for real solvents (COSMO-RS) and can be employed in a quantitative structure property relationship (QSPR) to predict solubility parameters. In the current study, it was aimed to compare both in silico methods with an experimental dataset of pharmaceutical compounds, which was complemented with own measurements by inverse gas chromatography. An initial evaluation of the total solubility parameters of reference solvents resulted in excellent predictions (observed versus predicted values) with R2 of 0.855 (COSMO-RS) and 0.945 (QSPR). The subsequent main study of pharmaceutical compounds exhibited R2 values of 0.701 (COSMO-RS) and 0.717 (QSPR). The comparatively lower prediction was to some extent due to the solid state of pharmaceuticals with known conceptual limitations of the solubility parameter and possible experimental bias. Total solubility parameters were also estimated by classical group contribution methods, which had comparatively lower prediction power. Therefore, the new in silico methods are highly promising for pharmaceutical applications.

The quest for exceptional drug solubilization in diluted surfactant solutions and consideration of residual solid state.

Solubility screening in different surfactant solutions is an important part of pharmaceutical profiling. A particular interest is in low surfactant concentrations that mimic the dilution of an oral dosage form. Despite of intensive previous research on solubilization in micelles, there is only limited data available at low surfactant concentrations and generally missing is a physical state analysis of the residual solid. The present work therefore studied 13 model drugs in 6 different oral surfactant solutions (0.5%, w/w) by concomitant X-ray diffraction (XRPD) analysis to consider effects on solvent-mediated phase transformations. A particular aspect was potential occurrence of exceptionally high drug solubilization. As a result, general solubilization correlations were observed especially between surfactants that share chemical similarity. Exceptional solubility enhancement of several hundred-fold was evidenced in case of sodium dodecyl sulfate solutions with dipyridamole and progesterone. Furthermore, carbamazepine and testosterone showed surfactant-type dependent hydrate formation. The present results are of practical relevance for an optimization of surfactant screenings in preformulation and early development and provide a basis for mechanistic modeling of surfactant effects on solubilization and solid state modifications.

Interactions of dimethylaminoethyl methacrylate copolymer with non-acidic drugs demonstrated high solubilization in vitro and pronounced sustained release in vivo.

Recent work demonstrated remarkable solubilization effects of methacrylate-copolymer Eudragit EPO (EPO) not only with acidic drugs but interestingly also with poorly soluble basic compounds. The current work studied EPO-mediated solubilization effects first in vitro using felodipine (FLP) and tamoxifen (TMX) as model compounds. EPO-containing solutions were subsequently compared in a rat pharmacokinetic study against reference solutions and suspensions. Surprisingly, solution formulations with EPO did not result in an increased relative oral bioavailability. Exposure was reduced for both drugs and plasma-profiles of the EPO solutions showed a delayed and lower maximum plasma concentration compared to the reference formulations. This sustained in vivo release was likely due to combined effects of strong drug-polymer interactions and pH-dependent precipitation of the polymer in the rat intestine. Remarkable was that in vitro drug-polymer coprecipitates did not reveal crystalline drug by polarized light microscopy. Thus, such a formulation approach provides a rather simple opportunity to modify drug release in vivo. However, this may be rather an approach for preclinical formulations, if high peak-to-trough ratios of plasma levels are problematic regarding adverse effects related to Cmax or if plasma concentrations drop too fast below required pharmacological concentrations.

Glass-forming ability of compounds in marketed amorphous drug products.

This note is about the glass-forming ability (GFA) of drugs marketed as amorphous solid dispersions or as pure amorphous compounds. A thermoanalytical method was complemented with an in silico study, which made use of molecular properties that were identified earlier as being relevant for GFA. Thus, molar volume together with effective numbers of torsional bonds and hydrogen bonding were used to map drugs that are as amorphous products on the market either as solid dispersion of without co-processed carrier as amorphous drug in a solid dosage form. Differential scanning calorimetry experiments showed that most compounds were stable glass formers (GFs) (class III) followed by so-called unstable GFs (class II) and finally, only vemurafenib was found in class I with increased crystallization propensity. The in silico results, however showed that all drugs were either clearly in the chemical space expected for GFs or they were borderline to the region that holds for high crystallization tendency. Interestingly, the pure amorphous compounds scattered in a very confined region of the molecular predictors. These findings can guide amorphous product development of future drug candidates. Based on the compound location in the given chemical space, amorphous formulation opportunities can be balanced against the risks of physical instability upon storage.

Miniaturized X-ray powder diffraction assay (MixRay) for quantitative kinetic analysis of solvent-mediated phase transformations in pharmaceutics.

Many pharmaceutical compounds exhibit polymorphism, which may result in solvent-mediated phase transformations. Since the polymorphic form has an essential influence on physicochemical characteristics such as solubility or dissolution rate, it is crucial to know the exact polymorphic composition of a drug throughout pharmaceutical development. This study addressed the need to perform quantitative X-ray analysis of polymorphic mixtures on a 96-well scale (MixRay). A calibration of polymorphic mixtures (anhydrate and hydrate) was performed with three model drugs, caffeine, piroxicam, and testosterone, and linear correlations were obtained for all compounds. The MixRay approach for piroxicam was applied to a solubility and residual solid screening assay (SORESOS) to quantify the amount of hydrate and anhydrate corresponding to kinetic bulk concentrations. Changes in these drug concentrations correlated well with the kinetic changes in the residual solid. The influence of excipients on the solid state and kinetic concentrations of piroxicam was also studied. Excipients strongly affected polymorphic transformation kinetics of piroxicam and concentrations after 24h depended on the excipient used. The new calibration X-ray method combined with bulk concentration analysis provides a valuable tool for both pharmaceutical profiling and early formulation development.

Miniaturized screening of polymers for amorphous drug stabilization (SPADS): rapid assessment of solid dispersion systems.

PURPOSE: Development of a novel, rapid, miniaturized approach to identify amorphous solid dispersions with maximum supersaturation and solid state stability. METHOD: Three different miniaturized assays are combined in a 2-step decision process to assess the supersaturation potential and drug-polymer miscibility and stability of amorphous compositions. Step 1: SPADS dissolution assay. Drug dissolution is determined in 96-well plates to detect systems that generate and maintain supersaturation. Promising combinations graduate to step 2. Step 2: SPADS interaction and SPADS imaging assays. FTIR microspectroscopy is used to study intermolecular interactions. Atomic force microscopy is applied to analyze molecular homogeneity and stability. Based on the screening results, selected drug-polymer combinations were also prepared by spray-drying and characterized by classical dissolution tests and a 6-month physical stability study. RESULTS: From the 7 different polymers and 4 drug loads tested, EUDRAGIT E PO at a drug load of 20% performed best for the model drug CETP(2). The classical dissolution and stability tests confirmed the results from the miniaturized assays. CONCLUSION: The results demonstrate that the SPADS approach is a useful de-risking tool allowing the rapid, rational, time- and cost-effective identification of polymers and drug loads with appropriate dual function in supersaturation performance and amorphous drug stabilization.

Application of a statistical method to the absorption of a new model drug from micellar and lipid formulations--evaluation of qualitative excipient effects.

The scope of the present article is to study formulation parameters of micellar and of lipid delivery systems on the exposure of a new drug compound A in Wistar rats. A statistical analysis is to be performed a posteriori from a data set of all rat studies that were conducted during the preclinical development of the drug. Several formulations were evaluated mainly in view of sufficient exposure in toxicological studies. Because of the low solubility and high lipophilicity of compound A, the preclinical formulation development focused on micellar solutions and different kinds of lipid drug delivery systems. Candidate formulations were first tested for their dilution in artificial intestinal fluids before they were evaluated in the rat. A partial least square model was applied to the entire pharmacokinetic data set, and the type of delivery system, as well as excipients, were investigated in view of effects on the area under the plasma level curve. The results showed that self-emulsifying systems and in particular self-microemulsifying drug delivery systems were most effective in pushing the exposure of compound A. Another significant factor was the dose. A data subset showed nonlinearity in the pharmacokinetics with respect to the dose. However, the most important findings of the multivariate data analysis were overall effects of excipients on the exposure. These effects are considered as a sum of several influences so that the underlying mechanism is essentially complex and is not fully understood. Cremophor and lecithin exhibited a positive effect, whereas TPGS containing systems reached only below average exposure. No significant effect was observed with polysorbate 80 or Solutol HS. The model indicated the favorable use of a cosurfactant, in particular Capmul MCM. Similarly the use of a cosolvent showed a positive coefficient and ethanol was here best in class. No marked effects were observed for the oil selection, but a tendency toward below average exposure was displayed when long-chain triglycerides were in the formulation. The a posteriori analysis of the pharmacokinetic data using multivariate statistical models was very helpful to clarify effects of drug delivery systems as well as of general effects of excipients. Guidance was provided for the formulator, but further studies are needed to better understand the complex effects on a mechanistic level.

Miniaturized assay for solubility and residual solid screening (SORESOS) in early drug development.

PURPOSE: The aim was to develop a miniaturized method for solubility and residual solid screening of drug compounds in aqueous and non-aqueous vehicles in early drug development. METHODS: Different crystal modifications of caffeine, carbamazepine, and piroxicam were added into 96-well filter plates and solubility was determined in 100 microl of 17 pharmaceutical vehicles. After filtration, drug concentration was determined by Ultra Performance Liquid Chromatography (UPLC). Residual solid drug in the filter plates was analyzed by high-throughput (HT) transmission X-ray Powder Diffraction (XRPD). RESULTS: HT XRPD analysis revealed solid form conversions of all compounds during solubility determination, e.g., formation of hydrates in aqueous vehicles (caffeine, carbamazepine, piroxicam) or conversion of a metastable crystal form to the stable form (caffeine). Drug solubility was strongly dependent on the crystal modifications formed during the solubility assay. CONCLUSIONS: The new assay allows the simultaneous, small scale screening of drug solubility in various pharmaceutical vehicles and identification of changes in solid form. It is useful for the identification of formulations and formulation options in non-clinical and clinical development.

Drug-excipient compatibility testing using a high-throughput approach and statistical design.

The aim of our research was to develop a miniaturized high throughput drug-excipient compatibility test. Experiments were planned and evaluated using statistical experimental design. Binary mixtures of a drug, acetylsalicylic acid, or fluoxetine hydrochloride, and of excipients commonly used in solid dosage forms were prepared at a ratio of approximately 1:100 in 96-well microtiter plates. Samples were exposed to different temperature (40 degrees C/ 50 degrees C) and humidity (10%/75%) for different time (1 week/4 weeks), and chemical drug degradation was analyzed using a fast gradient high pressure liquid chromatography (HPLC). Categorical statistical design was applied to identify the effects and interactions of time, temperature, humidity, and excipient on drug degradation. Acetylsalicylic acid was least stable in the presence of magnesium stearate, dibasic calcium phosphate, or sodium starch glycolate. Fluoxetine hydrochloride exhibited a marked degradation only with lactose. Factor-interaction plots revealed that the relative humidity had the strongest effect on the drug excipient blends tested. In conclusion, the developed technique enables fast drug-excipient compatibility testing and identification of interactions. Since only 0.1 mg of drug is needed per data point, fast rational preselection of the pharmaceutical additives can be performed early in solid dosage form development.