Development of Predictive Statistical Model for Gaining Valuable Insights in Pharmaceutical Product Recalls.

The rapid progress in artificial intelligence (AI) has revolutionized problem-solving across various domains. The global challenge of pharmaceutical product recalls imposes the development of effective tools to control and reduce shortage of pharmaceutical products and help avoid such recalls. This study employs AI, specifically machine learning (MI), to analyze critical factors influencing formulation, manufacturing, and formulation complexity which could offer promising avenue for optimizing drug development processes. Utilizing FDAZilla and SafeRX tools, an open database model was constructed, and predictive statistical models were developed using Multivariate Analysis and the Least Absolute Shrinkage and Selection Operator (LASSO) Approach. The study focuses on key descriptors such as delivery route, dosage form, dose, BCS classification, solid-state and physicochemical properties, release type, half-life, and manufacturing complexity. Through statistical analysis, a data simplification process identifies critical descriptors, assigning risk numbers and computing a cumulative risk number to assess product complexity and recall likelihood. Partial Least Square Regression and the LASSO approach established quantitative relationships between key descriptors and cumulative risk numbers. Results have identified key descriptors; BCS Class I, dose number, release profile, and drug half-life influencing product recall risk. The LASSO model further confirms these identified descriptors with 71% accuracy. In conclusion, the study presents a holistic AI and machine learning approach for evaluating and forecasting pharmaceutical product recalls, underscoring the importance of descriptors, formulation complexity, and manufacturing processes in mitigating risks associated with product quality.

Assessing the Impact of Punch Geometries on Tablet Capping Using a Newly Proposed Capping Index.

PURPOSE: Increased tablet anisotropy could lead to increased tablet capping propensity. Tooling design variables such as cup depth could serve as a key player for inducing tablet anisotropy. METHODS: A new capping index (CI) consisting of the ratio of compact anisotropic index (CAI) and material anisotropic index (MAI) is proposed to evaluate tablet capping propensity as a function of punch cup depth. CAI is the ratio of axial to radial breaking force. MAI is the ratio of axial to radial Young's modulus. The impact of various punch cup depths [flat face, flat face beveled edge, flat face radius edge, standard concave, shallow concave, compound concave, deep concave, and extra deep concave] on the capping propensity of model acetaminophen tablets was studied. Tablets were manufactured at 50, 100, 200, 250, and 300 MPa compression pressure at 20 RPM on different cup depth tools using Natoli NP-RD30 tablet press. A partial least squares model (PLS) was computed to model the impact of the cup depth and compression parameters on the CI. RESULTS: The PLS model exhibited a positive correlation of increased cup depth to the capping index. The finite elemental analysis confirmed that a high capping tendency with increased cup depth is a direct result of non-uniform stress distribution across powder bed. CONCLUSIONS: Certainly, a proposed new capping index with multivariate statistical analysis gives guidance in selecting tool design and compression parameters for robust tablets.

Compression-decompression modulus (CDM) - an alternative/complementary approach to Heckel's analysis.

The novel modulus-based approach was developed to characterize the compression behavior of the materials and how it results into tablet mechanical strength (TMS) of the final tablet. The force-displacement profile for the model materials (Vivapur® 101, Starch 1500®, Emcompress®, and Tablettose® 100) was generated at different compression pressures (100, 150, and 200 MPa) and speeds (0.35, 0.55, and 0.75 m/s) using compaction emulator (Presster™). A generated continuous compression profile was evaluated with Heckel plot and the proposed material modulus method. The computed compression parameters were qualitatively and quantitatively correlated with TMS by principal component analysis and principal component regression, respectively. Compression modulus has negatively correlated, while decompression modulus is positively correlated to TMS. Proposed modulus descriptors are independent of particle density measurements required for the Heckel method and could overcome the limitations of the Heckel method to evaluate the decompression phase. Based on the outcome of the study, a two-dimensional compression and decompression modulus classification system (CDMCS) was proposed. The proposed CDMCS could be used to define critical material attributes in the early development stage or to understand reasons for tablet failure in the late development stage.

Correction to: Employing Multivariate Statistics and Latent Variable Models to Identify and Quantify Complex Relationships in Typical Compression Studies.

During the transmission process in publishing the article online, the equal (=) sign was replaced with "0" in Equations 1 to 5. The original article has been corrected.

Employing Multivariate Statistics and Latent Variable Models to Identify and Quantify Complex Relationships in Typical Compression Studies.

The effect of storage condition (% RH) on flufenamic acid:nicotinamide (FFA:NIC) cocrystal compressibility, compactibility, and tabletability profiles was not observed after visual evaluation or linear regression analysis. However, multivariate statistical analysis showed that storage condition had a significant effect on each compressional profile. Shapiro and Heckel equations were used to determine the compression parameters: porosity, Shapiro's compression parameter (f), densification factor (Da), plastic yield pressure (YPpl), and elastic yield pressure (YPel). Latent variable models such as exploratory factor analysis, principal component analysis, and principal component regression were employed to decode complex hidden main, interaction, and quadratic effects of % RH and the compression parameters on FFA:NIC tablet mechanical strength (TMS). Statistically significant correlations between f and Da, f and YPpl, and Da and YPel supported the idea that both rearrangement and fragmentation, and plastic deformation are important to FFA:NIC TMS. To the authors knowledge, this is the first time that simultaneously operating dual mechanisms of fragmentation and plastic deformation in low and midrange compression, and midrange plastic deformation have been identified and reported. A quantitative PCR model showed that f, Da, and YPel had statistically significant main effects along with a significant antagonist storage condition-porosity "conditional interaction effect". f exhibited a 2.35 times greater impact on TMS compared to Da. The model root-mean-square error at calibration and prediction stages were 0.04 MPa and 0.08 MPa, respectively. The R2 values at the calibration stage and at the prediction stage were 0.9005 and 0.7539, respectively. This research demonstrated the need for caution when interpreting the results of bivariate compression data because complex latent inter-relationships may be hidden from visual assessment and linear regression analysis, and result in false data interpretation as illustrated in this report.

Physical properties and solubility studies of Nifedipine-PEG 1450/HPMCAS-HF solid dispersions.

Low-order high-energy nifedipine (NIF) solid dispersions (SDs) were generated by melt solvent amorphization with polyethylene glycol (PEG) 1450 and hypromellose acetate succinate (HPMCAS-HF) to increase NIF solubility while achieving acceptable physical stability. HPMCAS-HF was used as a crystallization inhibitor. Individual formulation components, their physical mixtures (PMs), and SDs were characterized by differential scanning calorimetry, powder X-ray diffraction, and Fourier transform infrared spectroscopy (FTIR). NIF solubility and percent crystallinity (PC) were determined at the initial time and after 5 days stored at 25 °C and 60% RH. FTIR indicated that hydrogen bonding was involved with the amorphization process. FTIR showed that NIF:HPMCAS-HF intermolecular interactions were weaker than NIF:PEG 1450 interactions. NIF:PEG 1450 SD solubilities were significantly higher than their PM counterparts (p < 0.0001). The solubilities of NIF:PEG 1450:HPMCAS-HF SDs were significantly higher than their corresponding NIF:PEG 1450 SDs (p < 0.0001-0.043). All the SD solubilities showed a statistically significant decrease (p < 0.0001) after storage for 5 days. SDs PC were statistically lower than their comparable PMs (p < 0.0001). The PCs of SDs with HPMCAS-HF were significantly lower than SDs not containing only PEG 1450. All SDs exhibited a significant increase in PC (p < 0.0001-0.0089) on storage. Thermogravimetric analysis results showed that HPMCAS-HF bound water at higher temperatures than PEG 1450 (p < 0.0001-0.0039). HPMCAS-HF slowed the crystallization process of SDs, although it did not completely inhibit NIF crystal growth.

Computational Predictability of Microsponge Properties Using Different Multivariate Models.

The capabilities of principal component regression (PCR) and multiple linear regression (MLR) were evaluated to decipher and predict the impact of formulation and process parameters on the modeled metronidazole benzoate (MB)-ethyl cellulose (EC) microsponge (MBECM) properties. MBECM were prepared by a quasi-emulsion solvent diffusion method. A minimum experimentation was designed using Box-Behnken approach with one center point after initial screening experiments. Data was modeled by principal component analysis (PCA), PCR, and MLR. Two distinct groupings of developed MBECM was observed in initial qualitative PCA as a function of their respective formulation and processing parameters. Group A formulations with low dichloromethane, high PVA, and low stirring speed exhibited larger particle size, lower entrapment efficiency (EE), and lower actual drug content (ADC) than Group B formulations. Optimized quantitative PCR and MLR models demonstrated a linear dependence of particle size and quadratic dependence of EE and ADC on the studied formulation and process parameters. Interestingly, MLR models showed relatively better predictability of the selected MBECM formulation properties when compared with PCR. MBECM were amorphous in nature and spherical shaped. Carbopol® 940 NF based hydrogel of selected MBECM formulation exhibited a prolonged MB release than the commercial MB gel (Metrogyl®), showing no signs of necrosis in the goat mucosa. Thus, a properly designed minimum experimentation coupled with multivariate modeling generated a knowledge-rich target space, which enabled to understand and predict the performance of developed MBECM within a prescribed design space.

Factorial Design Based Multivariate Modeling and Optimization of Tunable Bioresponsive Arginine Grafted Poly(cystaminebis(acrylamide)-diaminohexane) Polymeric Matrix Based Nanocarriers.

Desired characteristics of nanocarriers are crucial to explore its therapeutic potential. This investigation aimed to develop tunable bioresponsive newly synthesized unique arginine grafted poly(cystaminebis(acrylamide)-diaminohexane) [ABP] polymeric matrix based nanocarriers by using L9 Taguchi factorial design, desirability function, and multivariate method. The selected formulation and process parameters were ABP concentration, acetone concentration, the volume ratio of acetone to ABP solution, and drug concentration. The measured nanocarrier characteristics were particle size, polydispersity index, zeta potential, and percentage drug loading. Experimental validation of nanocarrier characteristics computed from initially developed predictive model showed nonsignificant differences (p > 0.05). The multivariate modeling based optimized cationic nanocarrier formulation of <100 nm loaded with hydrophilic acetaminophen was readapted for a hydrophobic etoposide loading without significant changes (p > 0.05) except for improved loading percentage. This is the first study focusing on ABP polymeric matrix based nanocarrier development. Nanocarrier particle size was stable in PBS 7.4 for 48 h. The increase of zeta potential at lower pH 6.4, compared to the physiological pH, showed possible endosomal escape capability. The glutathione triggered release at the physiological conditions indicated the competence of cytosolic targeting delivery of the loaded drug from bioresponsive nanocarriers. In conclusion, this unique systematic approach provides rational evaluation and prediction of a tunable bioresponsive ABP based matrix nanocarrier, which was built on selected limited number of smart experimentation.

Comparative binder efficiency modeling of dry granulation binders using roller compaction.

Roller compaction parameters' impact on granules and tableting properties of coprocessed Avicel® DG [ADG], a physical mixture of the two components at the same composition present in ADG [PADCP], and microcrystalline cellulose and Kollidon® VA-64 Fine physical mixture [KVA64] was quantified by analysis of variance (ANOVA) and multivariate methods. Roller force, roller gap, and roller speed levels were selected for evaluation. A 33 full-factorial experimental design with three center points for roller force, roller gap, and roller speed was used. The response parameters studied were granule-to-fines (GF) ratio, compressibility index (CI), tablet thickness (TT), tablet friability (TF), tablet breaking force (TBF) and disintegration time (DT). A model acetaminophen tablet formulation was roller granulated and tableted at 10 kg scale. Principal component analysis of ADG and PADCP formulations were separated from KVA64 formulations, indicating different granule and tableting properties were binder dependent. This difference in binder performance was also confirmed by ANOVA. The ANOVA also showed that there were no statistical performance differences between coprocessed ADG and its comparable physical blend with the exception of TT. Principal component regression (PCR) analyses of ADG and PADCP revealed that these excipients exhibited a statistically significant negative effect on granules-to-fine (GF) ratio, TT, TBF, and DT. KVA64 demonstrated a positive effect on these parameters. The KVA64 physical mixture demonstrated an overall better performance and binding capability. This study strongly suggests that there is no performance advantage of coprocessed Avicel® DG when compared to a physical mixture of the two components at the same composition.

Inhalable Ipratropium Bromide Particle Engineering with Multicriteria Optimization.

Spray-dried ipratropium bromide (IPB) microspheres for oral inhalation were engineered using Quality by Design. The interrogation of material properties, process parameters, and critical product quality attributes interplay enabled rational product design. A 27-3 screening design exhibited the Maillard reaction between L-leucine (LL) and lactose at studied outlet temperatures (OT) >130°C. A response surface custom design was used in conjunction with multicriteria optimization to determine the operating design space to achieve inhalable microparticles. Statistically significant predictive models were developed for volume median diameter (p = 0.0001, adjusted R 2 = 0.9938), span (p = 0.0278, adjusted R 2 = 0.7912), yield (p = 0.0020, adjusted R 2 = 0.9320), and OT (p = 0.0082, adjusted R 2 = 0.8768). An independent verification batch confirmed the model's predictive capability. The prediction and actual values were in good agreement. Particle size and span were 3.32 ± 0.09 µm and 1.71 ± 0.18, which were 4.7 and 5.3% higher than the predicted values. The process yield was 50.3%, compared to the predicted value of 65.3%. The OT was 100°C versus the predicted value of 105°C. The label strength of IPB microparticles was 99.0 to 105.9% w/w suggesting that enrichment occurred during the spray-drying process. The present study can be utilized to initiate the design of the first commercial IPB dry powder inhaler.

Aggregate Elasticity, Crystal Structure, and Tableting Performance for p-Aminobenzoic Acid and a Series of Its Benzoate Esters.

The tableting performance for p-aminobenzoic acid (PABA) and a series of its benzoate esters with increasing alkyl chain length (methyl-, ethyl-, and n-butyl) was determined over a broad range of compaction pressures. The crystalline structure of methyl benzoate (Me-PABA) exhibits no slip systems and does not form viable compacts under any compaction pressure. The ethyl (Et-PABA) and n-butyl (Bu-PABA) esters each have a similar, corrugated-layer structure that displays a prominent slip plane and improves material plasticity at low compaction pressure. The compact tensile strength for Et-PABA is superior to that for Bu-PABA; however, neither material achieved a tensile strength greater than 2 MPa over the compression range studied. Complementary studies with powder Brillouin light scattering (BLS) show the maxima of the shear wave, acoustic frequency distribution red shift in an order consistent with both the observed tabletability and attachment energy calculations. Moreover, zero-porosity aggregate elastic moduli are determined for each material using the average acoustic frequency obtained from specific characteristics of the powder BLS spectra. The Young's moduli for Et- and Bu-PABA is significantly reduced relative to PABA and Me-PABA, and this reduction is further evident in tablet compressibility plots. PABA, however, is distinct with high elastic isotropy as interpreted from the narrow and well-defined powder BLS frequency distributions for both the shear and compressional acoustic modes. The acoustic isotropy is consistent with the quasi-isotropic distribution of hydrogen bonding for PABA. At low compaction pressure, PABA tablets display the lowest tensile strength of the series; however, above a compaction pressure of ca. 70 MPa PABA tablet tensile strength continues to increase while that for Et- and Bu-PABA plateaus. PABA displays lower plasticity relative to either ester, and this is consistent with its crystalline structure and high yield pressure determined from in-die Heckel analysis. Overall the complementary approach of using both structural and the acoustic inputs uniquely provided from powder BLS is anticipated to expand our comprehension of the structure-mechanics relationship and its role in tableting performance.

Physico-Mechanical Properties of Coprocessed Excipient MicroceLac® 100 by DM(3) Approach.

PURPOSE: To determine the effect of relative humidity (RH) and hydroxypropyl methylcellulose (HPMC) on the physico-mechanical properties of coprocessed MacroceLac(®) 100 using 'DM(3)' approach. METHODS: Effects of RH and 5% w/w HPMC on MacroceLac(®) 100 Compressibility Index (CI) and tablet mechanical strength (TMS) were evaluated by 'DM(3)'. The 'DM(3)' approach evaluates material properties by combining 'design of experiments', material's 'macroscopic' properties, 'molecular' properties, and 'multivariate analysis' tools. A 4X4 full-factorial experimental design was used to study the relationship of MacroceLac(®) 100 molecular properties (moisture content, dehydration, crystallization, fusion enthalpy, and moisture uptake) and macroscopic particle size and shape on CI and TMS. A physical binary mixture (PBM) of similar composition to MacroceLac(®) 100 was also evaluated. Multivariate analysis of variance (MANOVA), principle component analysis, and partial least squares (PLS) were used to analyze the data. RESULTS: MANOVA CI ranking was: PBM-HPMC > PBM > MicroceLac(®)100 > MicroceLac(®)100-HPMC (p < 0.0001). MANOVA showed PBM's and PBM-HPMC's TMS values were lower than MicroceLac(®)100 and MicroceLac(®)100-HPMC (p < 0.0001). PLS showed that % RH, HPMC, and several molecular properties significantly affected CI and TMS. CONCLUSIONS: Significant MicroceLac(®)100 changes occurred with % RH exposure affecting performance attributes. HPMC physical addition did not prevent molecular or macroscopic matrix changes.

Characterization of critical physical and mechanical properties of freeze-dried grape powder for development of a clinical patient delivery system.

Grapes are hypothesized to be a "food medicine." Freeze-dried grape powder (FDGP) is being used to test clinical activity for a variety of applications and a reproducible and reliable delivery system was required. The FDGP was characterized using traditional physico-chemical methods to generate the data needed to identify its primary liability, i.e. moisture sorption. Above a threshold level of moisture content (~25% w/w, at RT), the material becomes both difficult to handle and exhibits significant degradation of several potentially clinically important chemical components (catechin, epicatechin, resveratrol). A moisture sorption isotherm was then used to tie the threshold to the exposure relative humidity above which this occurs. Kinetic uptake studies were used to estimate the maximum safe exposure time at a given humidity (a square root time dependence of moisture uptake was observed). Armed with this knowledge, a FDGP compact coated with a compression coat [100% bees wax or combinations of carnauba wax (70%) with HPC (30%) or Avicel(®) PH 102 (30%) or lactose monohydrate (30%)] was developed that will insure the shelf life of the material without the need for special handling for approximately more than 3 months.

Optimized L-SNEDDS and spray-dried S-SNEDDS using a linked QbD-DM3 rational design for model compound ketoprofen.

A QbD-DM3 strategy was used to design ketoprofen (KTF) optimized liquid (L-SNEDDS) and solid self-nanoemulsifying drug delivery systems (S-SNEDDS). Principal component analysis was used to identify the optimized L-SNEDDS containing Capmul® MCM NF, 10 % w/w; Kolliphor® ELP, 60 % w/w; and propylene glycol, 30 % w/w. The S-SNEDDS was manufactured by spray-drying a feed dispersion prepared by dissolving the optimized KTF-loaded L-SNEDDS in an ethanol-Aerosil® 200 dispersion. A Box Behnken design was employed to evaluate the effect of drug concentration (DC), Aerosil® 200 concentration (AC) and feed rate (FR) on maximizing percent yield (PY) and loading efficiency (LE). The optimal levels of DC, AC, and FR were 19.9 % w/w, 30.0 % w/w, and 15.0 %, respectively. The optimized S-SNEDDS was amorphous, and its dissolution showed a 2.37-fold increase in drug release compared to KTF in 0.1 HCl. An optimized independent spray-dried S-SNEDDS verification batch showed that the predicted and observed PY and LE were 70.49 % and 92.49 %, and 70.02 % and 91.27 %, respectively. The optimized L-SNEDDS and S-SNEDDS also met their quality target product profile criteria for globule size <100 nm, polydispersity index < 0.400, emulsification time < 30 s, and KTF L-SNEDDS solubility 100-fold greater than its water solubility.

Negative porosity issue in the Heckel analysis: A possible solution.

A parameterization of compaction simulator generated dynamic compression profile with a few grams of powder provides important information about the material deformation and compact elasticity. The Heckel equation is by far the most popular choice among pharmaceutical scientists for such parametrization. A general approach of Heckel analysis uses pycnometric powder density (ρP0) for relative density calculation. However, as 'in-die' tablet bulk density at applied compression pressure (ρBP) becomes greater than or equal to the measured ρP0, the general approach typically poses a negative porosity challenge at high compression pressure regions. It is only theoretically possible to have a tablet with zero or negative porosity. Negative porosity may be detected during 'in-die' compression analysis, but it will not exist after ejection of the tablet in practical aspect. Thus, the present work proposes a new approach to using pycnometric tablet density (ρPP) in the relative density calculations of Heckel analysis. This ρPP may be a better representation of actual tablet particle volume, as it is composed of non-accessible intra-particulate pores, which are broken under applied compression pressure. A new approach showed its immunity for Heckel high-pressure negative porosity. It enables the utilization of the compression and decompression phases of dynamic compression profiles to evaluate macroscopic compaction performance. The proposed approach was validated with a reported modified Heckel approach. The Heckel parameters computed with both methodologies for microcrystalline cellulose and lactose were not statistically different. However, a modified Heckel approach was unable to compute Heckel parameters of poorly compacting starch unlike the new approach. A modified Heckel approach became invalid during starch compaction at low compression pressures (below 400 MPa), where starch was forming weaker but still intact tablets. Certainly, a complete Heckel profiling with a new approach could save time and costs in an early development stage for designing and screening scientifically based lead prototype formulations.

Predictive Performance Comparison of Computed Linear and Quadratic Multivariate Models for In-Situ UV Fiber Optics Tablet Dissolution Testing.

A present investigation aimed for multivariate modeling as a solution to resolve inaccuracy in dissolution testing experienced in the use of in-situ UV fiber optics dissolution systems (FODS) due to signal saturation problems. This problem is specifically encountered with high absorbance of moderate to high dose formulations. A high absorbance not only impede a real-time assessment but can also result in inaccurate dissolution profiles. Full spectra (F) and low absorbance regions (L) were employed to develop linear and quadratic (Q) partial least squares (PLS) and principal component regression (PCR) models. The conventional dissolution of atenolol, ibuprofen, and metformin HCl immediate-release (IR) tablets followed by HPLC analysis was used as a reference method to gauge multivariate models' performance in the 'built-in' Opt-Diss model. The linear multivariate modeling outputs resulted in accurate dissolution profiles, despite the potentially high UV signal saturation at later time points. Conversely, the 'built-in' Opt-Diss model and multivariate quadratic models failed to predict dissolution profiles accurately. The current studies show a good agreement in the predictions across both low absorbance region and full spectra, demonstrating the multivariate models' robust predictability. Overall, linear PLS and PCR models showed statistically similar results, which demonstrated their applicative flexibility for using FODS despite signal saturation and provides a unique alternative to traditional and labor-intensive UV or HPLC dissolution testing.

Comparative evaluation of the powder and compression properties of various grades and brands of microcrystalline cellulose by multivariate methods.

The present work challenges a newly developed approach to tablet formulation development by using chemically identical materials (grades and brands of microcrystalline cellulose). Tablet properties with respect to process and formulation parameters (e.g. compression speed, added lubricant and Emcompress fractions) were evaluated by 2(3)-factorial designs. Tablets of constant true volume were prepared on a compaction simulator at constant pressure (approx. 100 MPa). The highly repeatable and accurate force-displacement data obtained was evaluated by simple 'in-die' Heckel method and work descriptors. Relationships and interactions between formulation, process and tablet parameters were identified and quantified by multivariate analysis techniques; principal component analysis (PCA) and partial least square regressions (PLS). The method proved to be able to distinguish between different grades of MCC and even between two different brands of the same grade (Avicel PH 101 and Vivapur 101). One example of interaction was studied in more detail by mixed level design: The interaction effect of lubricant and Emcompress on elastic recovery of Avicel PH 102 was demonstrated to be complex and non-linear using the development tool under investigation.