3D printed personalized therapies for pediatric patients affected by adrenal insufficiency.

BACKGROUND: Adrenal insufficiency is usually diagnosed in children who will need lifelong hydrocortisone therapy. However, medicines for pediatrics, in terms of dosage and acceptability, are currently unavailable. RESEARCH DESIGN AND METHODS: Semi-solid extrusion (SSE) 3D printing (3DP) was utilized for manufacturing of personalized and chewable hydrocortisone formulations (printlets) for an upcoming clinical study in children at Vall d'Hebron University Hospital in Barcelona, Spain. The 3DP process was validated using a specific software for dynamic dose modulation. RESULTS: The printlets contained doses ranging from 1 to 6 mg hydrocortisone in three different flavor and color combinations to aid adherence among the pediatric patients. The pharma-ink (mixture of drugs and excipients) was assessed for its rheological behavior to ensure reproducibility of printlets through repeated printing cycles. The printlets showed immediate hydrocortisone release and were stable for 1 month of storage, adequate for prescribing instructions during the clinical trial. CONCLUSIONS: The results confirm the suitability and safety of the developed printlets for use in the clinical trial. The required technical information from The Spanish Medicines Agency for this clinical trial application was compiled to serve as guidelines for healthcare professionals seeking to apply for and conduct clinical trials on 3DP oral dosage forms.

Machine learning of Raman spectra predicts drug release from polysaccharide coatings for targeted colonic delivery.

Colonic drug delivery offers numerous pharmaceutical opportunities, including direct access to local therapeutic targets and drug bioavailability benefits arising from the colonic epithelium's reduced abundance of cytochrome P450 enzymes and particular efflux transporters. Current workflows for developing colonic drug delivery systems involve time-consuming, low throughput in vitro and in vivo screening methods, which hinder the identification of suitable enabling materials. Polysaccharides are useful materials for colonic targeting, as they can be utilised as dosage form coatings that are selectively digested by the colonic microbiota. However, polysaccharides are a heterogeneous family of molecules with varying suitability for this purpose. To address the need for high-throughput material selection tools for colonic drug delivery, we leveraged machine learning (ML) and publicly accessible experimental data to predict the release of the drug 5-aminosalicylic acid from polysaccharide-based coatings in simulated human, rat, and dog colonic environments. For the first time, Raman spectra alone were used to characterise polysaccharides for input as ML features. Models were validated on 8 unseen drug release profiles from new polysaccharide coatings, demonstrating the generalisability and reliability of the method. Further, model analysis facilitated an understanding of the chemical features that influence a polysaccharide's suitability for colonic drug delivery. This work represents a major step in employing spectral data for forecasting drug release from pharmaceutical formulations and marks a significant advancement in the field of colonic drug delivery. It offers a powerful tool for the efficient, sustainable, and successful development and pre-ranking of colon-targeted formulation coatings, paving the way for future more effective and targeted drug delivery strategies.

Cold Laser Sintering of Medicines: Toward Carbon Neutral Pharmaceutical Printing.

Selective laser sintering (SLS) is an emerging three-dimensional (3D) printing technology that uses a laser to fuse powder particles together, which allows the fabrication of personalized solid dosage forms. It possesses great potential for commercial use. However, a major drawback of SLS is the need to heat the powder bed while printing; this leads to high energy consumption (and hence a large carbon footprint), which may hinder its translation to industry. In this study, the concept of cold laser sintering (CLS) is introduced. In CLS, the aim is to sinter particles without heating the powder bed, where the energy from the laser, alone, is sufficient to fuse adjacent particles. The study demonstrated that a laser power above 1.8 W was sufficient to sinter both KollicoatIR and Eudragit L100-55-based formulations at room temperature. The cold sintering printing process was found to reduce carbon emissions by 99% compared to a commercial SLS printer. The CLS printed formulations possessed characteristics comparable to those made with conventional SLS printing, including a porous microstructure, fast disintegration time, and molecular dispersion of the drug. It was also possible to achieve higher drug loadings than was possible with conventional SLS printing. Increasing the laser power from 1.8 to 3.0 W increased the flexural strength of the printed formulations from 0.6 to 1.6 MPa, concomitantly increasing the disintegration time from 5 to over 300 s. CLS appears to offer a new route to laser-sintered pharmaceuticals that minimizes impact on the environment and is fit for purpose in Industry 5.0.

Fabrication of 3D printed mutable drug delivery devices: a comparative study of volumetric and digital light processing printing.

Mutable devices and dosage forms have the capacity to dynamically transform dimensionally, morphologically and mechanically upon exposure to non-mechanical external triggers. By leveraging these controllable transformations, these systems can be used as minimally invasive alternatives to implants and residence devices, foregoing the need for complex surgeries or endoscopies. 4D printing, the fabrication of 3D-printed structures that evolve their shape, properties, or functionality in response to stimuli over time, allows the production of such devices. This study explores the potential of volumetric printing, a novel vat photopolymerisation technology capable of ultra-rapid printing speeds, by comparing its performance against established digital light processing (DLP) printing in fabricating hydrogel-based drug-eluting devices. Six hydrogel formulations consisting of 2-(acryloyloxy)ethyl]trimethylammonium chloride solution, lithium phenyl-2,4,6-trimethylbenzoylphosphinate, varying molecular weights of the crosslinking monomer, poly(ethylene glycol) diacrylate, and paracetamol as a model drug were prepared for both vat photopolymerisation technologies. Comprehensive studies were conducted to investigate the swelling and water sorption profiles, drug release kinetics, and physicochemical properties of each formulation. Expandable drug-eluting 4D devices were successfully fabricated within 7.5 s using volumetric printing and were shown to display equivalent drug release kinetics to prints created using DLP printing, demonstrating drug release, swelling, and water sorption properties equivalent to or better than those of DLP-printed devices. The reported findings shed light on the advantages and limitations of each technology for creating these dynamic drug delivery systems and provides a direct comparison between the two technologies, while highlighting the promising potential of volumetric printing and further expanding the growing repertoire of pharmaceutical printing.

Exploring the interactions of JAK inhibitor and S1P receptor modulator drugs with the human gut microbiome: Implications for colonic drug delivery and inflammatory bowel disease.

The gut microbiota is a complex ecosystem, home to hundreds of bacterial species and a vast repository of enzymes capable of metabolising a wide range of pharmaceuticals. Several drugs have been shown to affect negatively the composition and function of the gut microbial ecosystem. Janus Kinase (JAK) inhibitors and Sphingosine-1-phosphate (S1P) receptor modulators are drugs recently approved for inflammatory bowel disease through an immediate release formulation and would potentially benefit from colonic targeted delivery to enhance the local drug concentration at the diseased site. However, their impact on the human gut microbiota and susceptibility to bacterial metabolism remain unexplored. With the use of calorimetric, optical density measurements, and metagenomics next-generation sequencing, we show that JAK inhibitors (tofacitinib citrate, baricitinib, filgotinib) have a minor impact on the composition of the human gut microbiota, while ozanimod exerts a significant antimicrobial effect, leading to a prevalence of the Enterococcus genus and a markedly different metabolic landscape when compared to the untreated microbiota. Moreover, ozanimod, unlike the JAK inhibitors, is the only drug subject to enzymatic degradation by the human gut microbiota sourced from six healthy donors. Overall, given the crucial role of the gut microbiome in health, screening assays to investigate the interaction of drugs with the microbiota should be encouraged for the pharmaceutical industry as a standard in the drug discovery and development process.

Smart laser Sintering: Deep Learning-Powered powder bed fusion 3D printing in precision medicine.

Medicines remain ineffective for over 50% of patients due to conventional mass production methods with fixed drug dosages. Three-dimensional (3D) printing, specifically selective laser sintering (SLS), offers a potential solution to this challenge, allowing the manufacturing of small, personalized batches of medication. Despite its simplicity and suitability for upscaling to large-scale production, SLS was not designed for pharmaceutical manufacturing and necessitates a time-consuming, trial-and-error adaptation process. In response, this study introduces a deep learning model trained on a variety of features to identify the best feature set to represent drugs and polymeric materials for the prediction of the printability of drug-loaded formulations using SLS. The proposed model demonstrates success by achieving 90% accuracy in predicting printability. Furthermore, explainability analysis unveils materials that facilitate SLS printability, offering invaluable insights for scientists to optimize SLS formulations, which can be expanded to other disciplines. This represents the first study in the field to develop an interpretable, uncertainty-optimized deep learning model for predicting the printability of drug-loaded formulations. This paves the way for accelerating formulation development, propelling us into a future of personalized medicine with unprecedented manufacturing precision.

3D printed dispersible efavirenz tablets: A strategy for nasogastric administration in children.

Enteral feeding tubes (EFTs) can be placed in children diagnosed with HIV which need nutritional support due to malnutrition. EFTs are the main route for medication administration in these patients, bringing up concerns about off label use of medicines, dose inaccuracy and tube clogging. Here we report for the first time the use of selective laser sintering (SLS) 3D printing to develop efavirenz (EFZ) dispersible printlets for patients with HIV that require EFT administration. Water soluble polymers Parteck® MXP and Kollidon® VA64 were used to obtain both 500 mg (P500 and K500) and 1000 mg printlets (P1000 and K1000) containing 200 mg of EFZ each. The use of SLS 3D printing obtained porous dosage forms with high drug content (20 % and 40 % w/w) and drug amorphization using both polymers. P500, K500 and K1000 printlets reached disintegration in under 230 s in 20 mL of water (25 ± 1 °C), whilst P1000 only partially disintegrated, possibly due to saturation of the polymer in the medium. As a result, the development of dispersible EFZ printlets using hydrophilic polymers can be explored as a potential strategy for drug delivery through EFTs in paediatrics with HIV, paving the way towards the exploration of more rapidly disintegrating polymers and excipients for SLS 3D printing.

Developing an innovative 3D printing platform for production of personalised medicines in a hospital for the OPERA clinical trial.

Breast cancer is the most frequently diagnosed cancer in women worldwide, and non-adherence to adjuvant hormonotherapy can negatively impact cancer recurrence and relapse. Non-adherence is associated with side effects of hormonotherapy. Pharmacological strategies to mitigate the side effects include coadministration of antidepressants, however patients remain non-adherent. The aim of this work was to develop medicines containing both hormonotherapy, tamoxifen (20 mg), along with anti-depressants, either venlafaxine (37.5 or 75 mg) or duloxetine (30 or 60 mg), to assess the acceptability and efficacy of this personalised approach for mitigating tamoxifen side effects in a clinical trial. A major criterion for the developed medicines was the production rate, specified at minimum 200 dosage units per hour to produce more than 40,000 units required for the clinical trial. A novel capsule filling approach enabled by the pharmaceutical 3D printer M3DIMAKER 2 was developed for this purpose. Firstly, semi-solid extrusion 3D printing enabled the filling of tamoxifen pharma-ink prepared according to French compounding regulation, followed by filling of commercial venlafaxine or duloxetine pellets enabled by the development of an innovative pellet dispensing printhead. The medicines were successfully developed and produced in the clinical pharmacy department of the cancer hospital Gustave Roussy, located in Paris, France. The developed medicines satisfied quality and production rate requirements and were stable for storage up to one year to cover the duration of the trial. This work demonstrates the feasibility of developing and producing combined tamoxifen medicines in a hospital setting through a pharmaceutical 3D printer to enable a clinical trial with a high medicines production rate requirement.

The colon targeting efficacies of mesalazine medications and their impacts on the gut microbiome.

Successful treatment of ulcerative colitis (UC) is highly dependent on several parameters, including dosing regimen and the ability to deliver drugs to the disease site. In this study two strategies for delivering mesalazine (5-aminosalicylic acid, 5-ASA) to the colon were compared in an advanced in vitro model of the human gastrointestinal (GI) tract, the SHIME® system. Herein, a prodrug strategy employing bacteria-mediated drug release (sulfasalazine, Azulfidine®) was evaluated alongside a formulation strategy that utilised pH and bacteria-mediated release (5-ASA, Octasa® 1600 mg). SHIME® experiments were performed simulating both the GI physiology and colonic microbiota under healthy and inflammatory bowel disease (IBD) conditions, to study the impact of the disease state and ileal pH variability on colonic 5-ASA delivery. In addition, the effects of the products on the colonic microbiome were investigated by monitoring bacterial growth and metabolites. Results demonstrated that both the prodrug and formulation approaches resulted in a similar percentage of 5-ASA recovery under healthy conditions. On the contrary, during experiments simulating the GI physiology and microbiome of IBD patients (the target population) the formulation strategy resulted in a higher proportion of 5-ASA delivery to the colonic region as compared to the prodrug approach (P < 0.0001). Interestingly, the two products had distinct effects on the synthesis of key bacterial metabolites, such as lactate and short chain fatty acids, which varied according to disease state and ileal pH variability. Further, both 5-ASA and sulfasalazine significantly reduced the growth of the faecal microbiota sourced from six healthy humans. The findings support that the approach selected for colonic drug delivery could significantly influence the effectiveness of UC treatment, and highlight that drugs licensed for UC may differentially impact the growth and functioning of the colonic microbiota.

Paediatric clinical study of 3D printed personalised medicines for rare metabolic disorders.

Rare diseases are infrequent, but together they affect up to 6-10 % of the world's population, mainly children. Patients require precise doses and strict adherence to avoid metabolic or cardiac failure in some cases, which cannot be addressed in a reliable way using pharmaceutical compounding. 3D printing (3DP) is a disruptive technology that allows the real-time personalization of the dose and the modulation of the dosage form to adapt the medicine to the therapeutic needs of each patient. 3D printed chewable medicines containing amino acids (citrulline, isoleucine, valine, and isoleucine and valine combinations) were prepared in a hospital setting, and the efficacy and acceptability were evaluated in comparison to conventional compounded medicines in six children. The inclusion of new flavours (lemon, vanilla and peach) to obtain more information on patient preferences and the implementation of a mobile app to obtain patient feedback in real-time was also used. The 3D printed medicines controlled amino acid levels within target levels as well as the conventional medicines. The deviation of citrulline levels was narrower and closer within the target concentration with the chewable formulations. According to participants' responses, the chewable formulations were well accepted and can improve adherence and quality of life. For the first time, 3DP enabled two actives to be combined in the same formulation, reducing the number of administrations. This study demonstrated the benefits of preparing 3D printed personalized treatments for children diagnosed with rare metabolic disorders using a novel technology in real clinical practice.

Can in vitro/in silico tools improve colonic concentration estimations for oral extended-release formulations? A case study with upadacitinib.

Upadacitinib, classified as a highly soluble drug, is commercially marketed as RINVOQ®, a modified-release formulation incorporating hydroxypropyl methylcellulose as a matrix system to target extended release throughout the gastrointestinal (GI) tract. Our study aimed to explore how drug release will occur throughout the GI tract using a plethora of in vitro and in silico tools. We built a Physiologically-Based Pharmacokinetic (PBPK) model in GastroPlus™ to predict the systemic concentrations of the drug when administered using in vitro dissolution profiles as input to drive luminal dissolution. A series of in vitro dissolution experiments were gathered using the USP Apparatus I, III and IV in presence of biorelevant media, simulating both fasted and fed state conditions. A key outcome from the current study was to establish an in vitro-in vivo correlation (IVIVC) between (i) the dissolution profiles obtained from the USP I, III and IV methods and (ii) the fraction absorbed of drug as deconvoluted from the plasma concentration-time profile of the drug. When linking the fraction dissolved as measured in the USP IV model, a Level A IVIVC was established. Moreover, when using the different dissolution profiles as input for PBPK modeling, it was also observed that predictions for plasma Cmax and AUC were most accurate for USP IV compared to the other models (based on predicted versus observed ratios). Furthermore, the PBPK model has the utility to extract the predicted concentrations at the level of the colon which can be of utmost interest when working with specific in vitro assays.

Poly(D,l-lactide-co-glycolide) particles are metabolised by the gut microbiome and elevate short chain fatty acids.

The production of short chain fatty acids (SCFAs) by the colonic microbiome has numerous benefits for human health, including maintenance of epithelial barrier function, suppression of colitis, and protection against carcinogenesis. Despite the therapeutic potential, there is currently no optimal approach for elevating the colonic microbiome's synthesis of SCFAs. In this study, poly(D,l-lactide-co-glycolide) (PLGA) was investigated for this application, as it was hypothesised that the colonic microbiota would metabolise PLGA to its lactate monomers, which would promote the resident microbiota's synthesis of SCFAs. Two grades of spray dried PLGA, alongside a lactate bolus control, were screened in an advanced model of the human colon, known as the M-SHIME® system. Whilst the high molecular weight (Mw) grade of PLGA was stable in the presence of the microbiota sourced from three healthy humans, the low Mw PLGA (PLGA 2) was found to be metabolised. This microbial degradation led to sustained release of lactate over 48 h and increased concentrations of the SCFAs propionate and butyrate. Further, microbial synthesis of harmful ammonium was significantly reduced compared to untreated controls. Interestingly, both types of PLGA were found to influence the composition of the luminal and mucosal microbiota in a donor-specific manner. An in vitro model of an inflamed colonic epithelium also showed the polymer to affect the expression of pro- and anti-inflammatory markers, such as interleukins 8 and 10. The findings of this study reveal PLGA's sensitivity to enzymatic metabolism in the gut, which could be harnessed for therapeutic elevation of colonic SCFAs.

Customizable orodispersible films: Inkjet printing and data matrix encoding for personalized hydrocortisone dosing.

The aim of this study was to exploit the versatility of inkjet printing to develop flexible doses of drug-loaded orodispersible films that encoded information in a data matrix pattern, and to introduce a specialised data matrix-generator software specifically focused on the healthcare sector. Pharma-inks (drug-loaded inks) containing hydrocortisone (HC) were developed and characterised based on their rheological properties and drug content. Different strategies were investigated to improve HC solubility: formation of ß-cyclodextrin complexes, Soluplus® based micelles, and the use of co-solvent systems. The software automatically adapted the data matrix size and identified the number of layers for printing. HC content deposited in each film layer was measured, and it was found that the proportion of co-solvent used directly affected the drug solubility and simultaneously played a role in the modification of the viscosity and surface tension of the inks. The formation of ß-cyclodextrin complexes improved the drug quantity deposited in each layer. On the contrary, micelle-based inks were not suitable for printing. Orodispersible films containing flexible and low doses of personalised HC were successfully prepared, and the development of a code generator software oriented to medical use provided an additional, innovative, and revolutionary advantage to personalised medicine safety and accessibility.

Do probiotic interventions improve female unexplained infertility? A critical commentary.

Disruption of women's gut and cervicovaginal microbiota has been associated with multiple gynaecological diseases such as endometriosis, polycystic ovary syndrome, non-cyclic pelvic pain and infertility. Female infertility affects 12.6% of women worldwide; its aetiology is complex and multifactorial and can be underpinned by uterine pathologies, systemic diseases and age. In addition, a new perspective has emerged on the role of the gut and vaginal microbiomes in reproductive health. Research shows that the administration of precisely selected probiotics, often in combination with prior antibiotic treatment, may facilitate the restoration of symbiotic microbiota to increase successful conception and assisted reproductive technology outcomes. However, clarity on this issue from fuller research is currently hampered by a lack of consistency and harmonization in clinical studies: various lactobacilli and bifidobacteria species have been delivered through both the oral and vaginal routes, in different dosages, for different treatment durations. This commentary explores the intricate relationship between the microbiota in the cervicovaginal area and gut of women, exploring their potential contribution to infertility. It highlights ongoing research on the use of probiotic formulations in improving pregnancy outcomes, critically examining the divergent findings in these studies, which complicate a conclusive assessment of the efficacy of these interventions.

The role of artificial intelligence in generating original scientific research.

Artificial intelligence (AI) is a revolutionary technology that is finding wide application across numerous sectors. Large language models (LLMs) are an emerging subset technology of AI and have been developed to communicate using human languages. At their core, LLMs are trained with vast amounts of information extracted from the internet, including text and images. Their ability to create human-like, expert text in almost any subject means they are increasingly being used as an aid to presentation, particularly in scientific writing. However, we wondered whether LLMs could go further, generating original scientific research and preparing the results for publication. We taskedGPT-4, an LLM, to write an original pharmaceutics manuscript, on a topic that is itself novel. It was able to conceive a research hypothesis, define an experimental protocol, produce photo-realistic images of 3D printed tablets, generate believable analytical data from a range of instruments and write a convincing publication-ready manuscript with evidence of critical interpretation. The model achieved all this is less than 1 h. Moreover, the generated data were multi-modal in nature, including thermal analyses, vibrational spectroscopy and dissolution testing, demonstrating multi-disciplinary expertise in the LLM. One area in which the model failed, however, was in referencing to the literature. Since the generated experimental results appeared believable though, we suggest that LLMs could certainly play a role in scientific research but with human input, interpretation and data validation. We discuss the potential benefits and current bottlenecks for realising this ambition here.

Visualizing disintegration of 3D printed tablets in humans using MRI and comparison with in vitro data.

Three-dimensional (3D) printing is revolutionising the way that medicines are manufactured today, paving the way towards more personalised medicine. However, there is limited in vivo data on 3D printed dosage forms, and no studies to date have been performed investigating the intestinal behaviour of these drug products in humans, hindering the complete translation of 3D printed medications into clinical practice. Furthermore, it is unknown whether conventional in vitro release tests can accurately predict the in vivo performance of 3D printed formulations in humans. In this study, selective laser sintering (SLS) 3D printing technology has been used to produce two placebo torus-shaped tablets (printlets) using different laser scanning speeds. The printlets were administered to 6 human volunteers, and in vivo disintegration times were assessed using magnetic resonance imaging (MRI). In vitro disintegration tests were performed using a standard USP disintegration apparatus, as well as an alternative method based on the use of reduced media volume and minimal agitation. Printlets fabricated at a laser scanning speed of 90 mm/s exhibited an average in vitro disintegration time of 7.2 ± 1 min (measured using the USP apparatus) and 25.5 ± 4.1 min (measured using the alternative method). In contrast, printlets manufactured at a higher laser scanning speed of 130 mm/s had an in vitro disintegration time of 2.8 ± 0.8 min (USP apparatus) and 18.8 ± 1.9 min (alternative method). When tested in humans, printlets fabricated at a laser scanning speed of 90 mm/s showed an average disintegration time of 17.3 ± 7.2 min, while those manufactured at a laser scanning speed of 130 mm/s exhibited a shorter disintegration time of 12.7 ± 6.8 min. Although the disintegration times obtained using the alternative method more closely resembled those obtained in vivo, no clear correlation was observed between the in vitro and in vivo disintegration times, highlighting the need to develop better in vitro methodology for 3D printed drug products.

Electroactive Polymers for On-Demand Drug Release.

Conductive materials have played a significant role in advancing society into the digital era. Such materials are able to harness the power of electricity and are used to control many aspects of daily life. Conductive polymers (CPs) are an emerging group of polymers that possess metal-like conductivity yet retain desirable polymeric features, such as processability, mechanical properties, and biodegradability. Upon receiving an electrical stimulus, CPs can be tailored to achieve a number of responses, such as harvesting energy and stimulating tissue growth. The recent FDA approval of a CP-based material for a medical device has invigorated their research in healthcare. In drug delivery, CPs can act as electrical switches, drug release is achieved at a flick of a switch, thereby providing unprecedented control over drug release. In this review, recent developments in CP as electroactive polymers for voltage-stimuli responsive drug delivery systems are evaluated. The review demonstrates the distinct drug release profiles achieved by electroactive formulations, and both the precision and ease of stimuli response. This level of dynamism promises to yield "smart medicines" and warrants further research. The review concludes by providing an outlook on electroactive formulations in drug delivery and highlighting their integral roles in healthcare IoT.

Inkjet Printing of Pharmaceuticals.

Inkjet printing (IJP) is an additive manufacturing process that selectively deposits ink materials, layer-by-layer, to create 3D objects or 2D patterns with precise control over their structure and composition. This technology has emerged as an attractive and versatile approach to address the ever-evolving demands of personalized medicine in the healthcare industry. Although originally developed for nonhealthcare applications, IJP harnesses the potential of pharma-inks, which are meticulously formulated inks containing drugs and pharmaceutical excipients. Delving into the formulation and components of pharma-inks, the key to precise and adaptable material deposition enabled by IJP is unraveled. The review extends its focus to substrate materials, including paper, films, foams, lenses, and 3D-printed materials, showcasing their diverse advantages, while exploring a wide spectrum of therapeutic applications. Additionally, the potential benefits of hardware and software improvements, along with artificial intelligence integration, are discussed to enhance IJP's precision and efficiency. Embracing these advancements, IJP holds immense potential to reshape traditional medicine manufacturing processes, ushering in an era of medical precision. However, further exploration and optimization are needed to fully utilize IJP's healthcare capabilities. As researchers push the boundaries of IJP, the vision of patient-specific treatment is on the horizon of becoming a tangible reality.

Virtually Possible: Enhancing Quality Control of 3D-Printed Medicines with Machine Vision Trained on Photorealistic Images.

Three-dimensional (3D) printing is an advanced pharmaceutical manufacturing technology, and concerted efforts are underway to establish its applicability to various industries. However, for any technology to achieve widespread adoption, robustness and reliability are critical factors. Machine vision (MV), a subset of artificial intelligence (AI), has emerged as a powerful tool to replace human inspection with unprecedented speed and accuracy. Previous studies have demonstrated the potential of MV in pharmaceutical processes. However, training models using real images proves to be both costly and time consuming. In this study, we present an alternative approach, where synthetic images were used to train models to classify the quality of dosage forms. We generated 200 photorealistic virtual images that replicated 3D-printed dosage forms, where seven machine learning techniques (MLTs) were used to perform image classification. By exploring various MV pipelines, including image resizing and transformation, we achieved remarkable classification accuracies of 80.8%, 74.3%, and 75.5% for capsules, tablets, and films, respectively, for classifying stereolithography (SLA)-printed dosage forms. Additionally, we subjected the MLTs to rigorous stress tests, evaluating their scalability to classify over 3000 images and their ability to handle irrelevant images, where accuracies of 66.5% (capsules), 72.0% (tablets), and 70.9% (films) were obtained. Moreover, model confidence was also measured, and Brier scores ranged from 0.20 to 0.40. Our results demonstrate promising proof of concept that virtual images exhibit great potential for image classification of SLA-printed dosage forms. By using photorealistic virtual images, which are faster and cheaper to generate, we pave the way for accelerated, reliable, and sustainable AI model development to enhance the quality control of 3D-printed medicines.

Supramolecular chemistry enables vat photopolymerization 3D printing of novel water-soluble tablets.

Vat photopolymerization has garnered interest from pharmaceutical researchers for the fabrication of personalised medicines, especially for drugs that require high precision dosing or are heat labile. However, the 3D printed structures created thus far have been insoluble, limiting printable dosage forms to sustained-release systems or drug-eluting medical devices which do not require dissolution of the printed matrix. Resins that produce water-soluble structures will enable more versatile drug release profiles and expand potential applications. To achieve this, instead of employing cross-linking chemistry to fabricate matrices, supramolecular chemistry may be used to impart dynamic interaction between polymer chains. In this study, water-soluble drug-loaded printlets (3D printed tablets) are fabricated via digital light processing (DLP) 3DP for the first time. Six formulations with varying ratios of an electrolyte acrylate monomer, [2-(acryloyloxy)ethyl]trimethylammonium chloride (TMAEA), and a co-monomer, 1-vinyl-2-pyrrolidone (NVP), were prepared to produce paracetamol-loaded printlets. 1H NMR spectroscopy analysis confirmed the integration of TMAEA and NVP in the polymer, and residual TMAEA monomers were found to be present only in trace amounts (0.71 - 1.37 %w/w). The apparent molecular mass of the photopolymerised polymer was found to exceed 300,000 Da with hydrodynamic radii of 15 - 20 nm, estimated based on 1H DOSY NMR measurements The loaded paracetamol was completely released from the printlets between 45 minutes to 5 hours. In vivo single-dose acute toxicity studies in rats suggest that the printlets did not cause any tissue damage. The findings reported in this study represent a significant step towards the adoption of vat photopolymerization-based 3DP to produce personalised medicines.