Spectroscopy-based process analytical technology (PAT) has become the method of choice in the pharmaceutical industry for its non-destructive and rapid monitoring of critical quality attributes (CQA) of both intermediate and final drug products. When coupled with appropriate chemometric models, complex spectroscopic signals can be translated into meaningful process understanding. However, a significant amount of calibration data with corresponding reference values is often required to train traditional chemometric models, such as Partial Least Squares (PLS) regression, for optimum prediction accuracy. The scarcity of active pharmaceutical ingredients (APIs) and the time and effort required for collecting reference values limit the application of PAT at early drug product development stages. An alternative lean chemometric approach, iterative optimization technology (IOT), can be utilized in a calibration-free manner. Unlike the traditional regression approach in PLS, IOT formulates spectral interpretation as an optimization problem, using numerical solvers to estimate mixture compositions from known pure component spectra with pre-determined constraints. The consistency and reliability of the solver are critical for achieving accurate predictions and model robustness. In this study, a solver parameter, Lagrange Multiplier (LM), was studied along with principal component analysis (PCA) to understand solver performance and predictive behavior of IOT algorithm, particularly for the low-concentration components within the formulation. Furthermore, the integration of model diagnostics enhances confidence in predictive performance. Signal-to-noise ratio (SNR) was explored to characterize the predictive behavior of IOT, particularly in scenarios involving low-concentration components in formulation.

Understanding the initial water penetration dynamics in pharmaceutical tablets is critical-a precursor to drug bioavailability, yet the underlying mechanisms remain poorly understood. Using in-situ synchrotron X-ray-based micro-computed tomography (X-ray μCT), we visualized and quantified the microstructural changes in tablets composed of common excipients upon contact with a single water droplet. In pure microcrystalline cellulose (MCC PH-102 & PH-200), the hydrophobicity of the lubricant magnesium stearate (MgSt) was found to dictate water penetration, with 1% w/w MgSt significantly impeding penetration and promoting a more stable pore network compared to 0.25% w/w. Combining swelling MCC and rigid di-calcium phosphate (DCP) accelerated water penetration; here MCC-induced swelling generated synergistic pore expansion. Formulations containing the super disintegrant croscarmellose sodium (CCS) exhibited near instantaneous (<3 s) and rapid localized fragmentation, generating the highest porosity. Our findings reveal a complex interplay between excipient properties (swelling, rigidity, disruptive explosion), demonstrating that water penetration is governed by diverse and synergistic mechanisms. These crucial mechanistic insights are pivotal for the rational design of oral solid dosage forms with tailored formulations to improve oral bioavailability.

The performance of dry powder inhalers (DPIs) is governed by a complex interplay of deagglomeration, attachment, detachment, and dispersion. Here, we show that crystal morphology serves as a master variable that elucidates and redefines the mechanistic roles of process parameters in DPI systems. Through the deliberate engineering of four distinct crystal morphologies of fluticasone propionate (FP), we systematically decoupled the effects of blending time (15-120 min) and flow rate (40-80 L/min) on aerosolization. Analysis of the fine particle fraction (FPF) and particle deposition patterns allowed us to isolate the contributions of each parameter to the underlying aerosolization mechanisms. Our results reveal a sophisticated, morphology-dependent interplay. We found that for specific crystal habits, the role of blending time is not confined to modulating deagglomeration and attachment but extends to critically impacting detachment. Similarly, the influence of flow rate is not limited to detachment; for certain morphologies, it becomes a significant factor in driving deagglomeration. Ultimately, this work demonstrates that crystal morphology does not merely affect dispersion; it dictates the extent to which blending and flow rate control the entire cascade of aerosolization events. These findings establish a robust design principle for the rational engineering of inhalable crystals to achieve predictable and optimized therapeutic performance.

The integration of Process Analytical Technology (PAT) into pharmaceutical process development has become a critical focus since the issuance of FDA PAT guidance in 2004. While PAT applications for blend uniformity (BU) and content uniformity (CU) are well-established, their utilization in early-phase development in an end-to-end fashion is seldomly reported. This study presents a novel case of deploying end-to-end PAT capabilities, combined with material-sparing chemometric approaches, to mitigate the impact of coarse API particle size distribution on BU and CU. Near-infrared spectroscopy (NIRS) data were collected across three interfaces-bin-blender, tablet press feedframe, and tablets-to generate high-density real-time process data. Chemometric modeling via Classical Least Squares (CLS) was employed to translate spectra into actionable concentration data without the need for additional API calibration samples. Furthermore, this study provides a comparative analysis of CLS based upon measured versus estimated pure component spectra, highlighting the advantages and limitations of each approach. The findings underscore the potential of material-sparing PAT methods to enhance process understanding and robustness in early-phase development. The results also set the stage for broader adoption of PAT and phase-appropriate method development, fostering innovation in pharmaceutical research and development.

In low-dose tablet formulations, achieving content uniformity is of utmost importance. Conventional bulk characterization methods, such as blend uniformity testing, lack the ability to capture the structural and compositional characteristics of API within tablets. To fill this gap, we employed a correlative imaging and analysis workflow that integrates synchrotron X-ray micro-computed tomography (SyncCT), mosaic field-of-view scanning electron microscopy (mSEM), and energy-dispersive X-ray spectroscopy (EDX) methods, combined with AI-enabled image segmentation algorithms. This workflow was utilized to investigate the effect of ordered mixing, introduced by conical screening milling (comilling), on API distribution in two real-world tablets. Specifically, two tablets with identical formulations were prepared; one incorporating comilling and the other without it. The tablets were then subjected to the imaging workflow. The results showed that the comilling process significantly improved the uniformity of API distribution in the tablet through ordered mixing, as evidenced by mSEM-EDX and SyncCT imaging. Furthermore, the API uniformity was achieved without altering the microstructural characteristics, as evidenced by consistent pore size distribution and tortuosity values between the comilled and non-comilled tablets. Overall, this study demonstrated the potential of advanced imaging and AI-based image analysis in facilitating formulation development for low-dose tablets.

Objectives:Inavolisib is a potent selective inhibitor of the Class I PI3K alpha isoform, and it is currently being evaluated for the treatment of patients with PIK3CA-mutated, HR+, HER2−, locally advanced or metastatic breast cancer. The in vitro data suggest that inavolisib may exhibit modest CYP3A4 time-dependent inhibition (TDI) and induction. Therefore, our objective was to develop a PBPK model of inavolisib to assess its clinical DDI risk. Methods:The PBPK model for inavolisib was first developed and verified using in vitro results, clinical PK data, and human ADME data obtained following both IV and PO administration of inavolisib in a mass balance study. The key parameters driving the perpetrator DDI risk include the CYP3A4 TDI (Kinact and KI) and induction (Emax and EC50), which were determined from experiments. The inavolisib PBPK model was further validated using independent inavolisib PK data from Phase I monotherapy. The CYP3A4 interaction component of the inavolisib PBPK model was further validated using palbociclib PK data from the Phase I combination arm (inavolisib+palbociclib+letrozole combination); Of note, palbociclib is a CYP3A4 substrate (fm,CYP3A4 > 50%). The validated PBPK model was then applied to predict the effects of inavolisib on midazolam to evaluate the magnitude of perpetrator DDI risk from inavolisib. Sensitivity analyses of different perpetrator scenarios and different hepatocyte donors were also performed. Results:The model simulations successfully recapitulated the observed inavolisib PK data, with the majority of observed data within the 90% prediction interval and the overall shape of the PK profile well captured, suggesting successful model verification. All prediction/observation (P/O) ratios of inavolisib exposure metrics (AUC, Cmax, Tmax) are within 2-fold. The PK profile of palbociclib when co-administered with inavolisib was also well-captured, indicating that the CYP3A4 interaction component of the inavolisib PBPK model was validated. For DDI prediction, the midazolam AUC and Cmax ratios (with vs. without inavolisib) were predicted using sensitivity analysis containing three scenarios: CYP3A4 TDI only, CYP3A4 induction only, and combined effect of CYP3A4 TDI and induction. When considering the combined effect of both CYP3A4 inhibition and induction, the predicted geometric mean ratio (range) of the midazolam AUC and Cmax were 0.92 (0.74 - 1.05) and 0.93 (0.80 - 1.03), respectively. The range of the prediction was generated using different hepatocyte donors.Conclusions:We have successfully developed and verified a PBPK model for inavolisib that is able to describe all inavolisib PK data and inavolisib-palbociclib interaction accurately. Using this model, inavolisib at the therapeutic dose was predicted to only have a modest effect on the PK of sensitive CYP3A4 substrates in all scenarios simulated, suggesting a low likelihood of CYP3A4-mediated DDI.Citations: N/A

The purpose of this study is to present a correlative microscopy-tomography approach in conjunction with machine learning-based image segmentation techniques, with the goal of enabling quantitative structural and compositional elucidation of real-world pharmaceutical tablets. Specifically, the approach involves three sequential steps: 1) user-oriented tablet constituent identification and characterization using correlative mosaic field-of-view SEM and energy dispersive X-ray spectroscopy techniques, 2) phase contrast synchrotron X-ray micro-computed tomography (SyncCT) characterization of a large, representative volume of the tablet, and 3) constituent segmentation and quantification of the imaging data through user-guided, iterative supervised machine learning and deep learning. This approach was implemented on a real-world tablet containing 15

This study investigated the added value of combining both near-infrared (NIR) and Raman spectroscopy into a single NIRaman Combi Fiber Probe for in-line blend potency determination in the feed frame of a rotary tablet press. A five-component platform formulation was used, containing acetylsalicylic acid as the Active Pharmaceutical Ingredient (API). Calibration models for the determination of 1 and 5%w/w label claim tablets were developed using NIR and Raman spectra of powder blends ranging from 0.75 to 1.25%w/w and 3.75 to 6.25%w/w API, respectively. Step-change experiments with deliberate 10% deviation steps from the label claims were performed, from which the collected spectra were used for model validation. For model development and validation, low-level data fusion was explored through concatenation of preprocessed NIR and Raman spectra. Mid-level data fusion was also evaluated, based on extracted features of the preprocessed data. Herewith, score vectors were extracted by transforming preprocessed spectra through Principal Component Analysis, followed by critical feature selection through Elastic Net Regression. Partial Least Squares regression was applied to regress singular, low-level or mid-level fused data versus blend potency. It could be concluded that irrespective of the data fusion technique, an increase in Step-Change Sensitivity (SCS) and decrease in Root Mean Squared Error (RMSE) was observed when predicting the 5%w/w step-change experiment. For the prediction of the 1%w/w step-change experiment, no added benefit with regard to SCS and RMSE was observed due to the addition of the noisy NIR spectra.

Discharge of powder from a hopper or bin is a common operation in solid dosage form manufacture. Powder flow obstruction during hopper/bin discharge, such as arching or ratholing, remains an outstanding risk and cannot be reliably diagnosed using the existing flow function coefficient-based method. In this study, we showed that the major principal stress (σ1) at the bin outlet is required for an accurate prediction of powder flow obstruction risks. We noted that powder is susceptible to flow obstruction when the unconfined yield strength exceeds the stress facilitating powder failure. We presented a complete model to calculate the stress conditions and subsequently predict flow obstruction risks in hopper/bin discharge based on this criterion. The method was experimentally verified by hopper/bin discharge experiments encompassing 10 powder blends and 2 equipment systems. Importantly, we showed that the active stress state assumption should be employed for the powder flow obstruction prediction because σ1 is high and powder is more susceptible to flow obstruction. Prediction under the passive stress state can lead to significant under-estimation of flow obstruction risks. Therefore, the hopper design protocol, which assumes the passive stress state in arching prediction, should not be indiscriminately used toward pharmaceutical powder flow applications.

The purpose of the study is introduce a two-phase flow model to simulate water penetration into pharmaceutical tablets. This model was built by integrating Darcy's law with the continuity principle, on the premise that water penetration was driven by capillary actions. Notably, this model concerned both the ingress of water (wetting phase) and simultaneous displacement of air (non-wetting phase). Due to the interference of the two fluids, the relative permeability and capillary pressure vary during water penetration. Evolution of these parameters was incorporated in the model. Calibration of the model by water penetration experiments of the microcrystalline cellulose (MCC) tablet yielded an average pore radius of 42 nm. This derived result was corroborated by FIB-SEM analysis revealing the presence of extensive microporosity within MCC particles with an average radius of ∼30 nm. Further validation was achieved through close resemblance between the simulated and experimental water penetration profiles of MCC tablets possessing different porosities. Overall, this study underscored the advantage of the two-phase flow model over single-phase flow models, by capturing the dependence of permeability and capillary pressure on water saturation. Therefore it holds promise for an enhanced description of water penetration into tablets.

Pharmaceutical drug dosage forms are critical for ensuring the effective and safe delivery of active pharmaceutical ingredients to patients. However, traditional formulation development often relies on extensive lab and animal experimentation, which can be time-consuming and costly. This manuscript presents a generative artificial intelligence method that creates digital versions of drug products from images of exemplar products. This approach employs an image generator guided by critical quality attributes, such as particle size and drug loading, to create realistic digital product variations that can be analyzed and optimized digitally. This paper shows how this method was validated through two case studies: one for the determination of the amount of material that will create a percolating network in an oral tablet product and another for the optimization of drug distribution in a long-acting HIV inhibitor implant. The results demonstrate that the generative AI method accurately predicts a percolation threshold of 4.2% weight of microcrystalline cellulose and generates implant formulations with controlled drug loading and particle size distributions. Comparisons with real samples reveal that the synthesized structures exhibit comparable particle size distributions and transport properties in release media. Pharmaceutical drug dosage forms are traditionally determined through extensive physical experimentation. Here, the authors present a generative AI method that creates digital drug products from images, matching and improving critical quality attributes such as particle size and drug loading.

PURPOSE:We aim to present a refined thin-film model describing the drug particle dissolution considering radial diffusion in spherical boundary layer, and to demonstrate the ability of the model to describe the dissolution behavior of bulk drug powders.METHODS:The dissolution model introduced in this study was refined from a radial diffusion-based model previously published by our laboratory (So et al. in Pharm Res. 39:907-17, 2022). The refined model was created to simulate the dissolution of bulk powders, and to account for the evolution of particle size and diffusion layer thickness during dissolution. In vitro dissolution testing, using fractionated hydrochlorothiazide powders, was employed to assess the performance of the model.RESULTS:Overall, there was a good agreement between the experimental dissolution data and the predicted dissolution profiles using the proposed model across all size fractions of hydrochlorothiazide. The model over-predicted the dissolution rate when the particles became smaller. Notably, the classic Nernst-Brunner formalism led to an under-estimation of the dissolution rate. Additionally, calculation based on the equivalent particle size derived from the specific surface area substantially over-predicted the dissolution rate.CONCLUSION:The study demonstrated the potential of the radial diffusion-based model to describe dissolution of drug powders. In contrast, the classic Nernst-Brunner equation could under-estimate drug dissolution rate, largely due to the underlying assumption of translational diffusion. Moreover, the study indicated that not all surfaces on a drug particle contribute to dissolution. Therefore, relying on the experimentally-determined specific surface area for predicting drug dissolution is not advisable.

The incorporation of a counterion into an amorphous solid dispersion (ASD) has been proven to be an attractive strategy to improve the drug dissolution rate. In this work, the generality of enhancing the dissolution rates of free acid ASDs by incorporating sodium hydroxide (NaOH) was studied by surface-area-normalized dissolution. A set of diverse drug molecules, two common polymer carriers (copovidone or PVPVA and hydroxypropyl methylcellulose acetate succinate or HPMCAS), and two sample preparation methods (rotary evaporation and spray drying) were investigated. When PVPVA was used as the polymer carrier for the drugs in this study, enhancements of dissolution rates from 7 to 78 times were observed by the incorporation of NaOH into the ASDs at a 1:1 molar ratio with respect to the drug. The drugs having lower amorphous solubilities showed greater enhancement ratios, providing a promising path to improve the drug release performance from their ASDs. Samples generated by rotary evaporation and spray drying demonstrated comparable dissolution rates and enhancements when NaOH was added, establishing a theoretical foundation to bridge the ASD dissolution performance for samples prepared by different solvent-removal processes. In the comparison of polymer carriers, when HPMCAS was applied in the selected system (indomethacin ASD), a dissolution rate enhancement of 2.7 times by the incorporated NaOH was observed, significantly lower than the enhancement of 53 times from the PVPVA-based ASD. This was attributed to the combination of a lower dissolution rate of HPMCAS and the competition for NaOH between IMC and HPMCAS. By studying the generality of enhancing ASD dissolution rates by the incorporation of counterions, this study provides valuable insights into further improving drug release from ASD formulations of poorly water-soluble drugs.

The incorporation of counterions into amorphous solid dispersions (ASDs) has been proven to be effective for improving the dissolution rates of ionizable drugs in ASDs. In this work, the effect of dissolution buffer pH and concentration on the dissolution rate of indomethacin-copovidone 40:60 (IMC-PVPVA, w/w) ASD with or without incorporated sodium hydroxide (NaOH) was studied by surface area-normalized dissolution to provide further mechanistic understanding of this phenomenon. Buffer pH from 4.7 to 7.2 and concentration from 20 to 100 mM at pH 5.5 were investigated. As the buffer pH decreased, the IMC dissolution rate from both ASDs decreased. Compared to IMC-PVPVA ASD, the dissolution rate decrease from IMCNa-PVPVA ASD was more resistant to the decrease of buffer pH. In contrast, while buffer concentration had a negligible impact on the IMC dissolution rate from IMC-PVPVA ASD, the increase of buffer concentration significantly reduced the IMC dissolution rate from IMCNa-PVPVA ASD. Surrogate evaluation of microenvironment pH modification by the dissolution of IMCNa-PVPVA ASD demonstrated the successful elevation of buffer microenvironment pH and the suppression of such pH elevation by the increase of buffer concentration. These results are consistent with the hypothesis that the dissolution rate enhancement by the incorporation of counterions originates from the enhanced drug solubility by ionization and the modification of diffusion layer pH in favor of drug dissolution. At the studied drug loading (∼40%), relatively congruent release between IMC and PVPVA was observed when IMC was ionized in ASD or in solution, highlighting the importance of studying the ionization effect on the congruent release of ASDs, especially when drug ionization is expected in vivo. Overall, this work further supports the application of incorporating counterions into ASDs for improving the dissolution rates of ionizable drugs when enabling formulation development is needed.

We aim to perform a systematic study of the time consolidation effect, i.e. the reduction of powder flowability resulting from at-rest storage, on a diverse array of pharmaceutical powders under different stress, humidity, and length of time. A ring shear cell-based methodology was employed. An instantaneous flow function was obtained, followed immediately by at-rest consolidation at precisely controlled humidity, stress, and duration. The consolidated powder was then subjected to shear-cell measurement. The difference in flowability between the immediate and consolidated specimens were attributed to the time consolidation effect. Among the six excipients tested, three exhibited time consolidation at varying extents. Citric acid and starch underwent time consolidation only at high relative humidity (RH = 75

Assessment and understanding of changes in particle size of active pharmaceutical ingredients (API) and excipients as a function of solid dosage form processing is an important but under-investigated area that can impact drug product quality. In this study, X-ray microscopy (XRM) was investigated as a method for determining the in situ particle size distribution of API agglomerates and an excipient at different processing stages in tablet manufacturing. An artificial intelligence (AI)–facilitated XRM image analysis tool was applied for quantitative analysis of thousands of individual particles, both of the API and the major filler component of the formulation, microcrystalline cellulose (MCC). Domain size distributions for API and MCC were generated along with the calculation of the porosity of each respective component. The API domain size distributions correlated with laser diffraction measurements and sieve analysis of the API, formulation blend, and granulation. The XRM analysis demonstrated that attrition of the API agglomerates occurred secondary to the granulation stage. These results were corroborated by particle size distribution and sieve potency data which showed generation of an API fines fraction. Additionally, changes in the XRM-calculated size distribution of MCC particles in subsequent processing steps were rationalized based on the known plastic deformation mechanism of MCC. The XRM data indicated that size distribution of the primary MCC particles, which make up the larger functional MCC agglomerates, is conserved across the stages of processing. The results indicate that XRM can be successfully applied as a direct, non-invasive method to track API and excipient particle properties and microstructure for in-process control samples and in the final solid dosage form. The XRM and AI image analysis methodology provides a data-rich way to interrogate the impact of processing stresses on API and excipients for enhanced process understanding and utilization for Quality by Design (QbD).

Purpose The purpose of the study is to present a mathematical model capable of describing drug particle dissolution in 3-dimensional (3D) space, and to provide experimental model verification. Through this study, we also aim to elaborate limitations of the classic, 1D-based Nernst-Brunner formalism in dissolution modeling. Methods The 3D dissolution model was derived by treating the dissolution of a spherical particle as a diffusion-driven process, and by solving Fick's 2(nd) law of diffusion in spherical coordinates using numerical methods. The resulting model was experimentally verified through analyzing the dissolution behavior of single succinic acid particles in un-stirred water droplet under polarized light microscopy, in combination with image segmentation techniques. Results A set of working equations was developed to describe drug particle dissolution in 3D space. The predicted dissolution time and profile are in good agreement with the experimental results. The model clearly shows that the concentration gradient within the diffusion layer, in realistic 3D condition, must not be a constant value as implicated in the Nernst-Brunner formalism. The actual concentration profile is a hyperbola, and the concentration gradient at the surface of the particle can be significantly higher than the classic 1D-based dissolution model. Conclusion The study demonstrates that the classic, 1D-based dissolution models may lead to significant under-estimation of drug dissolution rates. In contrast, modeling dissolution in 3D space yields more reliable results. This study merits further development of comprehensive 3D drug dissolution models, by considering polydispersed particle ensemble and imposing the changes of diffusion layer thickness during dissolution.

With the advent of continuous direct compression (CDC) process, it becomes increasingly desirable to characterize inherent powder blend heterogeneity at a small batch scale for a robust and CDC-amenable formulation. To accomplish this goal, a near infrared spectroscopy (NIRS)-based characterization approach was developed and implemented on multiple direct compression (DC) blends in this study, with the intended purpose of complementing existing formulation development tools and enabling to build an early CMC data package for late-phased process analytical technology (PAT) method development. Three fumaric acid DC blends, designed to harbor varied degrees of inherent blend heterogeneity, were employed. Near infrared spectral data were collected on a kg-scale batch blender via both time- and angle-based triggering modes. The time-triggered data were used to investigate the blending heterogeneity with respect to rotation angles, while the angle-triggered data were used to provide blending variability characterization and compare against off-line HPLC-based results. The time-triggered data revealed that the greatest blend variability was observed between revolutions, while the blending variability within a single revolution stayed relatively low with respect to rotation angles. This confirmed earlier literature findings that the bottom layer of powder blends tends to move with the blender within each revolution, and the most intense powder mixing takes place across revolutions. This also indicates the use of blending speed and the number of co-adds are not able to increase sampling volume to improve signal-to-noise ratio under a tumble-bin blender as what were typically done in a feedframe application. The angle-triggered data showed that there is a consistent trend between NIRS and HPLC-based methods on characterizing blend heterogeneity across the blends at a given sample size. This study contributes to establishing NIRS as a potential characterization approach for inherent powder blend heterogeneity for early R&D. It also highlights the promise of continuous characterization of inherent powder blend heterogeneity from gram scale to mini-batch CDC scale.

The purpose of the study is to build a "virtual roller compactor" as a predictive tool to assess the roll force (RF)-maximum pressure (Pmax) and RF-ribbon density relationship for pharmaceutical roller compaction. We provided a theoretical basis to demonstrate that, there exists a critical nip angle for a pharmaceutical powder, beyond which the RF-Pmax relationship is insensitive to wall friction angle or effective angle of internal friction. We showed that for most pharmaceutical roller compaction, the critical nip angle is lower than 17 degree, and can be exceeded via wall friction elevation, using rolls with non-smooth surface. Under this condition, the original Johanson model can be substantially simplified to a single equation requiring only one material property (compressibility). By performing manufacturing-scale roller compaction using materials with diverse compressibility, we showed that the simplified, friction angle-free model performed similar to the original Johanson model. It can predict the RF-Pmax and RF-ribbon density relationship well after applying a correction factor. The predictive tool, in the form of a user-friendly graphical user interface, was created based on the simplified model. The tool was adopted for in-house, bench-scale formulation development and scale-up because of its ease-of-use, good predicting capability, and very low material demand.