Optimising 3D printed medications for rare diseases: In-line mass uniformity testing in direct powder extrusion 3D printing

Optimising 3D printed medications for rare diseases: In-line mass uniformity testing in direct powder extrusion 3D printing Gloria Mora-Casta˜ no a,1 , Lucía Rodríguez-Pombo b,1 , Paola Carou-Senra b , Patricija Januskaite c , Carlos Rial d,e , Carlos Bendicho-Lavilla d,e , Maria L. Couce f , M´ onica Mill´ an-Jim´ enez a , Isidoro Caraballo a , Abdul W. Basit c,e,* , Carmen Alvarez-Lorenzo b , Alvaro Goyanes b,c,d,e,* a Department of Pharmacy and Pharmaceutical Technology, Faculty of Pharmacy, Universidad de Sevilla, Seville 41012, Spain b Departamento de Farmacología, Farmacia y Tecnología Farmac´ eutica, I+D Farma (GI-1645), Facultad de Farmacia, Instituto de Materiales (iMATUS) and Health Research Institute of Santiago de Compostela (IDIS), Universidade de Santiago de Compostela, Santiago de Compostela 15782, Spain c Department of Pharmaceutics, UCL School of Pharmacy, University College London, 29-39 Brunswick Square, London WC1N 1AX, United Kingdom d FABRX Ltd., Henwood House, Henwood, Ashford, Kent TN24 8DH, United Kingdom e FABRX Artificial Intelligence, Carretera de Escair´ on, 14, Currelos, O Savi˜ nao CP 27543, Spain f Servicio de Neonatología, Unidad de Diagn´ ostico y Tratamiento de Enfermedades Metab´ olicas Cong´ enitas, Health Research Institute of Santiago de Compostela (IDIS), Hospital Clínico Universitario de Santiago de Compostela, Universidad de Santiago de Compostela, IDIS, RICORS, CIBERER, MetabERN, Spain ARTICLE INFO Keywords: 3D printed pharmaceuticals Additive manufacturing Drug delivery systems, modified release formulations Rare metabolic disorders Pediatric precision treatments Pharma-inks ABSTRACT Biotinidase deficiency is a rare inherited disorder characterized by biotin metabolism issues, leading to neurological and cutaneous symptoms that can be alleviated through biotin administration. Three-dimensional (3D) printing (3DP) offers potential for personalized medicine production for rare diseases, due to its flexibility in designing dosage forms and controlling release profiles. For such point-of-care applications, rigorous quality control (QC) measures are essential to ensure precise dosing, optimal performance, and product safety, especially for low personalized doses in preclinical and clinical studies. In this work, we addressed QC challenges by integrating a precision balance into a direct powder extrusion pharmaceutical 3D printer (M3DIMAKER™) for real-time, in-line mass uniformity testing, a critical quality control step. Small and large capsule-shaped biotin printlets (3D printed tablets) for immediateand extended-release were printed. The integrated balance monitored and registered each printlet’s weight, identifying any deviations from acceptable limits. While all large printlet batches met mass uniformity criteria, some small printlet batches exhibited weight deviations. In vitro release studies showed large immediate-release printlets releasing 82% of biotin within 45 min, compared to 100% for small immediate-release printlets. For extended-release formulations, 35% of the drug was released from small printlets, whereas 24% was released from large printlets at the same time point. The integration of process analytical technology tools in 3DP shows promise in enhancing QC and scalability of personalized dosing at the point-of-care, demonstrating successful integration of a balance into a direct powder extrusion 3D printer for in-line mass uniformity testing across different sizes of capsule-shaped printlets. 1. Introduction Biotinidase deficiency is an inherited rare disease characterized by an autosomal recessive disorder of biotin metabolism (Yılmaz, 2024). With a prevalence of 1 in 60,000, untreated patients with biotinidase deficiency may present with neurological and/or cutaneous symptoms, including developmental delay, seizures, hypotonia, sensorineural hearing loss, optic atrophy, alopecia, and skin rashes (Tankeu, 2023; Cowan et al., 2010; Canda et al., 2020). Fortunately, these clinical features can be improved or prevented by administering personalized doses * Corresponding authors at: Department of Pharmaceutics, UCL School of Pharmacy, University College London, 29-39 Brunswick Square, London WC1N 1AX, United Kingdom (A.W. Basit); Departamento de Farmacología, Farmacia y Tecnología Farmac´ eutica, I+D Farma (GI-1645), Facultad de Farmacia, Instituto de Materiales (iMATUS) and Health Research Institute of Santiago de Compostela (IDIS), Universidade de Santiago de Compostela, Santiago de Compostela 15782, Spain (A. Goyanes). E-mail addresses: [email protected] (A.W. Basit), [email protected] (A. Goyanes). 1 These authors contributed equally to this work. Contents lists available at ScienceDirect International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm https://doi.org/10.1016/j.ijpharm.2024.124964 Received 6 September 2024; Received in revised form 14 November 2024; Accepted 15 November 2024 International Journal of Pharmaceutics 668 (2025) 124964 Available online 16 November 2024 0378-5173/© 2024 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). of the water-soluble vitamin biotin. Treatment is tailored to the patient’s metabolic activity and therefore requires continuous adjustments: oral biotin doses of 5–10 mg/day for those with <10 % of average normal serum enzyme activity, and 2.5–10 mg/day for those with 10 % −30 % of average normal enzyme activity (Tankeu, 2023; Canda et al., 2020; Sayegh, 2020). Three-dimensional (3D) printing (3DP) is an innovative technology that combines computer-aided design and manufacturing to create an object through successive layers of material (Milliken, 2024; Cardoso and E.s., 2024). Being able to create objects with complex structures and geometries, this technique is well suited to meet the needs of developing personalized dosing for various patient groups, including paediatrics and geriatrics, polypharmacy patients, and individuals with rare diseases (Patel, 2024; Huanbutta, 2023; Fastø, 2019; Januskaite, 2020; Carou-Senra, 2023; Mazarura et al., 2022; Ahola, 2024; Funk, 2024). 3DP allows for the flexible production of small batches with customisations in dose, shape, drug release kinetics, composition, and the incorporation of multiple drugs into a single dosage form (Boniatti, 2021; Mora-Casta˜ no et al., 2023; Shojaie et al., 2023; Muhindo, 2023; Denis, 2024; Windolf, 2022). Pharmaceutical 3DP has already shown advantages over conventional formulations in the production of personalized medicines for bioequivalence (Lyousoufi, 2023) and clinical studies targeting specific populations and rare diseases (Goyanes, 2019a,b; Rodríguez-Pombo, 2024a,b; Liu, 2023). Among the various 3DP technologies, material extrusion is the most commonly used in the pharmaceutical sector, which includes semisolid extrusion (SSE), fused deposition modelling (FDM), and direct powder extrusion (DPE), depending on the type of drug-loaded ink (pharma-ink) used (Algahtani et al., 2018). SSE 3DP, which employs a gel or paste as the pharma-ink and operates at low temperatures, has been investigated extensively in bioprinting, personalized medicine, and novel dosage forms, such as chewable printlets (Carou-Senra, 2023; Wang, 2023a,b; Awad, 2023a,b; Chatzitaki, 2023; Utomo, 2023). Additionally, the use of disposable pre-filled syringes makes SSE closer to meeting the quality control (QC) requirements mandated by regulatory bodies (Vithani, 2019). FDM, in contrast, uses a drug-loaded thermoplastic filament as the pharma-ink, requiring a hot melt extruder (HME) for filament manufacture (Ghanizadeh Tabriz, 2023). However, this two-step thermal process can degrade the active pharmaceutical ingredient (API) and increase manufacturing time. The need for filaments with specific rheological and mechanical properties further limits its broader application. Therefore, it is necessary to evaluate the stability, quality, and behaviour of the filaments used in this technique (Bandari, 2021; MoraCasta˜ no, 2022; Oladeji, 2022; Xu, 2020; Okwuosa, 2021; Ayyoubi, 2023; Yang, 2023). DPE has emerged as an alternative to FDM, by directly printing powder or granulate blends as pharma-inks without filament preparation (Zheng, 2021; Aguilar-de-Leyva, 2023). The powder blend is fed directly into a hopper, transported to a heated nozzle by a single screw, and extruded in a layer-by-layer manner (Goyanes, 2019a,b). This process overcomes some limitations associated with FDM, as it allows for single-step production, reducing thermal stress, and is more efficient due to the small quantity of raw materials needed (Goyanes, 2019a,b; Rosch, 2023; Fanous, 2020). DPE also enables the formation of amorphous solid dispersions (ASDs) with high drug loading, known to enhance drug solubility (Boniatti, 2021; Goyanes, 2019a,b; Wang, 2023a,b; Pistone, 2023; Mora-Casta˜ no, 2024). Despite its advantages, DPE presents its own challenges in the production of pharmaceutical dosage forms, such as the influence of powder flow, rheological properties, and electrostatic forces. Poor powder can hinder continuous and homogeneous material feed through the screw, causing variability in dosing and limiting the uniformity of weight and API content in the final product (Boniatti, 2021; Rosch, 2023; Pistone, 2023). The implementation of 3DP in pharmaceuticals must address challenges related to Good Manufacturing Practice (GMP), and the quality of 3D printed products must be guaranteed through rigorous QC measures for optimal performance and safety (Muhindo, 2023; Rosch, 2023; Bendicho-Lavilla, 2024). Mass uniformity testing during the printing process is essential for evaluating reproducibility and ensuring accurate dosing in each pharmaceutical form, a crucial point in the QC of 3DP (Rosch, 2023; Deon, 2022). This becomes particularly important in the case of low personalized doses for rare diseases, where treatment customization and preclinical and clinical studies demand accurate weight determination to ensure efficacy and safety (Díaz-Torres, 2023; Johannesson, 2023). Process analytical technology (PAT) tools offer a solution for ensuring batch-wide mass uniformity and can be installed in-line, atline, off-line, or on-line (Seoane-Via˜ no, 2023). For example, a previous study integrated a balance into an SSE pharmaceutical 3D printer, for non-destructive and in-line weight uniformity testing of hydrocortisone printlets (Bendicho-Lavilla, 2024). The aim of this work was to implement, for the first time, a balance on a multi-printhead 3D printer to monitor the weight and reproducibility of medicine manufactured by DPE 3DP. Two different excipients (polyethylene oxide 100,000 and hydroxypropyl cellulose) were selected to manufacture capsule-shaped printlets for immediateand extended-release systems, respectively. Two printlet batch sizes, containing 2.5 and 10 mg of biotin, were manufactured for potential use in clinical and preclinical studies using the healthcare M3DIMAKER™ Studio software, which controlled both the 3D pharmaceutical printer and the in-line mass uniformity testing. Finally, physicochemical characterization techniques were used to evaluate the properties of the developed capsule-shaped printlets, assessing biotin content and release profiles for the different printlets. 2. Materials and methods 2.1. Materials Biotin (MW 244.31 g/mol, Acofarma, Barcelona, Spain) was used as the model drug. Polyethylene oxide 100,000 (PEO, 100,000 Da, SigmaAldrich, St. Louis, USA) and hydroxypropyl cellulose (HPC Klucel ELF, 40,000 Da, Ashland, Wilmington, USA) were the selected polymers to manufacture the pharma-inks. Pearlitol flash-mannitol (Roquette, Lestrem, France) was added to the immediate-release pharma-ink to improve printlet resolution and dissolution. Hydrochloric acid (37 %, Ph. Eur, Scharlau, Barcelona, Spain), sodium phosphate tribasic dodecahydrate (≥98.0 %, Honeywell Fluka, Buchs, Switzerland), sodium hydroxide (Ph. Eur. pellets, VWR International, Radnor, Pennsylvania, USA), and phosphate buffer solution (pH =3.9) were used for the preparation of the dissolution testing media. Acetonitrile (≥99.9 % v/v, HPLC grade, Merck, Darmstadt, Germany) and trifluoracetic acid (≥99.0 % v/v, HPLC grade, Fisher Scientific, Hampton, USA) were used as the mobile phase for the drug content assay. All materials were used as received. 2.2. Preparation of the pharma-ink The pharma-inks used for direct powder extrusion (DPE) were made by weighing out each component, containing 5 % w/w of the model drug, biotin. The raw materials were manually mixed using a mortar and Table 1 Composition of the pharma-inks for both immediateand extended-release systems. Pharma-ink HPC ELF (% w/w) PEO 100,000 (% w/w) Mannitol (% w/w) Biotin (% w/w) Immediaterelease (IR) −60 35 5 Extendedrelease (ER) 95 − − 5 G. Mora-Casta˜ no et al. International Journal of Pharmaceutics 668 (2025) 124964 2 pestle. Table 1 shows the composition of the pharma-inks developed in this work. 2.3. DPE 3D printing process validation The printing process was carried out using the M3DIMAKER™ 2 multi-printhead 3D printer (FABRX Ltd., London, UK) with the incorporated balance (Fig. 1). The build plate consisted of a glass platform covered by blue tape to aid in filament adhesion (3 M, Saint Paul, Minnesota, USA). This was then placed on the top of the balance to act as the printing platform. Initially, the capsule 3D model was selected from the software’s repository. The printing parameters were then selected using the M3DIMAKER Studio™ software. The next step was to choose the 3DP technology (in this case, DPE), and the drug loading of the pharma-ink (5 % w/w). Subsequently, based on the selected printing parameters, the capsule shape model (2.7 mm diameter, 8.6 mm length, Fig. 2) was scaled and printed in 4 varying sizes. The software sent instructions to the 3D printer to print a batch of 20 printlets arranged in 4 rows of 5 printlets each, with varying weights per row. The weight of each printlet was then automatically measured by the integrated balance inside the 3D printer. The software then used this information to establish a “dose library”, an algorithm that creates relationships between the size, weight, dose, and printing parameters. This “dose library” allows the user to select any dose in a simple manner, with the software adapting the size of the 3D model to manufacture the selected dose (RodríguezPombo, 2024a,b). 2.4. 3D printing process and mass uniformity testing The prepared pharma-inks were added to the hopper of a pharmaceutical multi-printhead 3D printer, M3DIMAKER™ 2 (FABRX Ltd., London, UK), using the direct single-screw powder extruder printhead, with a nozzle diameter of 0.4 mm. Two sizes were printed, small and large capsule-shaped printlets, to obtain the desired doses of biotin (2.5 mg and 10 mg), to be used for preclinical and clinical studies, respectively. The printing parameters generated using M3DIMAKER Studio™ software are highlighted in Table 2 for both biotin formulations. Batches of 10 capsule-shaped printlets were printed and weighed by the integrated balance for each size and formulation, and the software collected all the recorded weights of each batch. The integrated balance weighed each printlet immediately after it was printed as reported in previous work (Bendicho-Lavilla, 2024). The M3DIMAKER™ 2 multiprinthead 3D printer manufactured the printlets individually. After each printlet was complete, the printhead moved up and remained motionless for some seconds to allow the balance to stabilise. The balance reported the weight measurement for each printlet to the software, which displayed the value in the respective position of the printlet. The balance was then automatically tared and the next printlet was printed. 2.5. Characterization of the capsule-shaped printlets The dimensions (length and diameter) of the printlets (n =10) in each batch were measured using a digital calliper (Mercer Type 130, G&M Tools and Equipment Ltd, Carlisle, UK). 2.5.1. X-ray powder diffraction (XRPD) analysis The diffractograms of pure powder samples (biotin, HPC ELF, PEO 100,000, and mannitol), the pharma-ink, and the printlets were obtained in a D8 Advance diffractometer (Bruker, Billerica, MA, USA) equipped with a sealed X-ray tube ((CuK α 1 (λ =1.5406 Å)) and a LYNXEYE-type detector. The intensity and voltage applied during the assay were 40 mA and 40 kV, respectively. The diffractograms were obtained in the 2θ angular range of 3 – 50◦with a step of 0.04◦and a counting time of 4 s per step. Samples were deposited on an oriented Se (511) plate to avoid scattering noise and were spun to minimize the effect of the preferential orientation. 2.5.2. Thermal analysis Differential scanning calorimetry (DSC) was utilized to understand the thermal behaviour of the pure excipients (HPC ELF, PEO 100,000, and mannitol) and drug (biotin), the pharma-inks, and printlets. Thermograms were obtained using a DSC Q100 (TA Instruments, New Castle, DE, USA) with a refrigerated cooling accessory at a heating rate of 10 ◦C/min. The temperature range was 0 – 300 ◦C and nitrogen was used as the purge gas (flow rate =50 mL/min). All measurements were performed using non-hermetic aluminium pans, in which 2 – 5 mg of sample was accurately weighed using the precision balance (0.0001 mg) of TGA55 Discovery series equipment (TA Instruments, New Castle, DE, Fig. 1. Image of the balance-3D printer M3DIMAKER™ 2 system. G. Mora-Casta˜ no et al. International Journal of Pharmaceutics 668 (2025) 124964 3 USA). Thermal Gravimetry Analysis (TGA) was used to investigate the thermal stability of all formulation components. The thermal behaviour of the pure HPC Klucel ELF and biotin, the pharma-inks, and the printlet were evaluated. TGA measurements were carried out using a TGA55 Discovery series (TA Instruments, New Castle, DE, USA) with a refrigerated cooling accessory at a heating rate of 10 ◦C/min. The temperature range was 25–350 ◦C and nitrogen was used as the purge gas, at a flow rate of 50 mL/min. All experiments were performed using non-hermetic aluminium pans, in which 3 – 9 mg of sample was accurately weighed using the precision balance (0.0001 mg) of the same equipment. All data from DSC and TGA experiments was collected with the TA Advantage software for Q series (version 2.8) and processed using TA Instruments Universal Analysis 2000 (version 4.5.0.5). 2.5.3. Fourier-transform infrared spectroscopy (FTIR) The attenuated total reflectance Fourier transform infrared (ATRFTIR) spectra of pure excipients and biotin, the pharma-inks, and printlets were collected using a Varian 670 FTIR spectrometer (Varian Inc., Palo Alto, USA). All samples were scanned between 4000 and 400 cm −1 at a resolution of 4 cm −1 for 32 scans. 2.5.4. Scanning electron microscopy (SEM) imaging SEM imaging was performed to visualize the printlet surface and the internal structure after cutting. Both IR and ER printlets were cut in half and placed on metal supports, before being analysed using a Zeiss EVO LS15 scanning electron microscope equipped with a backscattered electron detector (Oberkochen, Germany), operating at 20 kV. 2.6. Drug loading determination Samples of the different printlets (large and small) were placed in a beaker (50 mL for small printlets and 100 mL for large printlets) with Milli-Q® water and were subjected to magnetic stirring overnight. Sample solutions were then filtered through 0.22 µm filters (Millipore Ltd., Dublin, Ireland) and the concentration of biotin was determined with a HPLC-UV LC-1100 Series HPLC system (Agilent Technologies, Santa Clara, USA). The assay entailed injecting 30 µL samples for analysis using a mobile phase of acetonitrile and water (25:75 v/v), containing 0.05 % v/v trifluoro acetic acid through a Symmetry 5 µm C18 column, 4.6 mm ×250 mm (Waters, Milford, Massachusetts), maintained at 30 ◦C. The mobile phase was pumped at a flow rate of 1.0 mL/ min and the eluent was screened at a wavelength of 210 nm. The retention time was 3.8 min, and the concentration range was 0.50–120 µg/mL. 2.7. In vitro release studies Biotin release profiles of the large capsule-shaped printlets were evaluated using an SR8-Plus Dissolution Test Station (Hanson Research, Chatsworth, CA, USA) with USP-II apparatus. The biotin release profiles of the small capsule-shaped printlets were evaluated using an ERWEKA DT 126 light series (ERWEKA GmbH, Langen, Germany) with USP-II apparatus (mounted with mini paddles). The speed of both paddles was set at 100 rpm with a temperature of 37±0.5 ◦C. For the in vitro immediate-release studies, the large printlets were dropped in 1000 mL of 0.1 M HCl (pH =1.2) for 2 h to simulate human gastric conditions for clinical studies. The small printlets were dropped in 100 mL of phosphate buffer pH 3.9 for 2 h to simulate rat gastric conditions for preclinical studies. For the in vitro extended-release studies, the large printlets were dropped in 750 mL of 0.1 M HCl for 2 h to simulate gastric conditions. After 2 h, 250 mL of trisodium phosphate solution (0.2 M) was added into each vessel and the pH was adjusted to 6.8 using 1 M NaOH or 1 M HCl solutions to simulate human intestinal conditions. The small printlets were dropped in 100 mL of phosphate buffer pH 3.9 for 2 h to simulate rat gastric conditions, followed by adjustment of the pH to 6.5 with 5 M NaOH, simulating rat intestinal conditions. 5 mL of sample was manually withdrawn from each vessel and immediately replaced with fresh medium, filtered through 0.22 μ m filters (Millipore Ltd., Dublin, Ireland), and analysed using HPLC to determine the amount of drug released (method described previously). Tests were conducted in triplicate under sink conditions. Data were reported throughout as mean ±standard deviation (n =3). 3. Results and discussion 3.1. 3D printing process and in-line mass uniformity testing Two different sizes for both IR and ER printlets were selected to evaluate, for the first time, the feasibility of an integrated balance and multi-printhead 3D printer system to carry out the in-line mass uniformity assay for DPE 3DP. The multi-printhead 3D printer system (Fig. 1) Fig. 2. Images of the design of the capsule-shaped printlet model (2.7 mm diameter, 8.6 mm length). Table 2 Printing parameters used during the printing process for both IR and ER printlets. Printing parameter IR small size printlet ER small size printlet IR large size printlet ER large size printlet Fill pattern Rectilinear Rectilinear Rectilinear Rectilinear Infill (%) 50 50 50 50 Raft and brim No No No No Extrusion speed (mm/s) 10.5 5.25 16.5 8.25 Travel speed (mm/s) 10 10 10 10 First layer height (mm) 0.30 0.15 0.30 0.15 Layer height (mm) 0.30 0.30 0.30 0.30 Extrusion temperature (◦C) 100 170 100 170 G. Mora-Casta˜ no et al. International Journal of Pharmaceutics 668 (2025) 124964 4 was insulated using two polycarbonate cover doors to mitigate any weight measurement fluctuations resulting from air movement. Additionally, the balance was placed beneath the printing platform, which was fitted with an anti-vibration plate to prevent weight variations due to vibrations produced during the printing process. The M3DIMAKER Studio™ software was configured to control both the 3D printer and the in-line mass uniformity test. First, a validation process was conducted to optimise the printing parameters for each pharma-ink, allowing the software to create a dose library that linked the 3D scale models to different weights and doses. The initial capsule-shaped 3D model (Fig. 2) was then scaled to produce four different sizes. The mean weight per row (n =5) was recorded, and the software established correlations between the model size, weight, dose and printing parameters. This validation step allows users to select any weight or dose within the validated range, as M3DIMAKER Studio™ will automatically select the corresponding file from the dose library. The software-controlled balance was successfully integrated into the pharmaceutical multi-printhead 3D printer, enabling the preparation of up to three dosage forms simultaneously and conducting in-line mass uniformity testing in the same process. This study extends our previous work related to the in-line mass uniformity test for SSE 3DP by allowing up to three DPE printheads to be used at the same time and evaluating a different 3DP technology (Bendicho-Lavilla, 2024). The capsule shape was chosen for its favourable swallowability and ease of administration in both preclinical and clinical studies (Goyanes, 2017). However, this shape poses a challenge for DPE, because the first layers are composed of a single line that supports the entire structure. This may cause the object to move during printing due to the instability of the structure if the pharma-ink does not adhere well to the printing platform. HPC was selected for extended-release printlets due to its good mechanical strength, stability across a wide pH range, and its established use as a matrix in controlled-release drug delivery systems (Rowe et al., 2009). HPC was also chosen since its favourable printability has been successfully demonstrated in other studies with DPE 3DP (Zheng, 2021; S´ anchez-Guirales, 2021; Mendibil, 2021). PEO 100,000 was selected for immediate-release printlets due to its high water solubility and prior use in extrusion based 3DP to produce orodispersible films and immediaterelease tablets (Jennotte, 2023; Racaniello, 2023; Cho, 2020). Each size and formulation batch (n =10) was successfully printed using the M3DIMAKER™ 2 (Fig. 3). The pharma-inks flowed smoothly through the single-screw DPE system and adhered well to the build plate without the need for heating the printing platform. Moreover, the printlets were easily removable from the build plate, retaining their shape and leaving no residual material. All printlets were individually weighed in-line and reported by the M3DIMAKER Studio™ software during the printing process (Fig. 4). Additionally, dimensional (diameter and length) measurements of the capsule-shaped printlets for each batch were taken (Table 3). The surface area/volume (SA/V) ratio was also calculated. According to the European Pharmacopeia (Ph. Eur.) (Europe, C.o. 2.9.5, 2024), no more than 2 of the 20 individual masses in a batch should deviate from the average mass by more than ±10 % for small printlets or ±7.5 % for large printlets, with none deviating by more than twice these percentages. The individual weights of each printlet from each batch are shown in Fig. 5. As observed in Fig. 5B and D, all large printlet batches met the accepted mass uniformity limits, according to the criteria that no more than 2 of the 20 individual masses in a batch may deviate from the average mass by more than ±10 % for small printlets or ±7.5 % for large printlets, with none deviating by more than twice these percentages. However, some of the small printlets (Fig. 5A and C) fell outside the accepted weight limits, rendering them non-compliant with the mass uniformity criteria. This is the case for the ER small printlet batch, where four printlets deviated from the average theoretical weight. This is consistent with the higher coefficient of variation (CV) in weights of the small printlets, which was double that of the large printlets. This could be attributed to the smaller dimensions of the small capsule-shaped printlets and the nozzle diameter used (Bendicho-Lavilla, 2024). Using a smaller nozzle diameter could potentially resolve this issue. It should be highlighted that the narrowest accepted weight range applies to the small printlet batches, indicating that even minor deviations in weight are detected as out of limits by the M3DIMAKER Studio™ software. These findings highlight the potential of integrating a balance as a weight control system within the DPE 3D printer. Using the M3DIMAKER Studio™ software, weight deviations from the average can be identified in real-time during the 3DP process, which can accelerate the traditionally time-intensive QC assessments essential for solid dosage forms. In a related study, Bendicho-Lavilla et al. successfully integrated an online analytical balance into an SSE pharmaceutical printer, automating mass uniformity testing (Bendicho-Lavilla, 2024). Building on this, we aimed to further advance QC by implementing an analytical balance in a multi-printhead DPE pharmaceutical printer. This multi-printhead capability allows simultaneous use of multiple DPE printheads, with each printlet weighed in real-time. Fig. 3. Images of the capsule-shaped IR printlets: (A) small and large size; (B) complete large batch printed onto the printing platform. Images of the capsule-shaped ER printlets: (C) small and large size; (D) complete large batch printed onto the printing platform. Scale is in cm. G. Mora-Casta˜ no et al. International Journal of Pharmaceutics 668 (2025) 124964 5 The present work offers significant scientific contributions by addressing the technical challenges of integrating a multi-printhead DPE printer with an analytical balance. Unlike SSE, which is easier to pair with a balance due to gentler operating conditions, such as minimal vibrations, no surface contact, and no fan requirement, DPE presents a more demanding setup. Specifically, DPE requires smaller layer heights, leading to frequent contact with the balance surface and additional vibrations. Additionally, the fan on the DPE printhead must be intermittently turned off during balance taring and weighing, while avoiding prolonged downtime to prevent material clumping in the hopper. This study represents a meaningful advance in QC for medicines produced using DPE technology. 3.2. Characterization of the capsule-shaped printlets 3.2.1. Thermal behaviour and XRPD analysis DSC and XRPD analyses were performed to determine the solid-state structure of biotin in the pharma-inks before printing and in the final printlets. Pure biotin exhibited a melting point of 232 ◦C in the first heating cycle, as observed in Fig. 6A. The absence of this endothermic peak in the DSC thermograms of the pharma-inks and printlets suggests that biotin may be amorphous and solubilised within the 3D printed capsule-shaped printlets during the excipient melting process. Alternatively, biotin may have remained crystalline and was only solubilized during the DSC assay. These findings are in conjunction with the XRPD results (Fig. 7), which showed crystalline peaks of biotin, in which no heat was applied during the analysis, suggesting solubilisation of biotin within the excipients during the heating in the DSC pan. PEO and mannitol exhibited endothermic peaks at 65 ◦C and 170 ◦C, respectively. As expected, the immediate-release pharma-ink and printlet showed both endothermic peaks attributed to PEO and mannitol. To assess biotin stability at the printing temperature, TGA was performed. The highest printing temperature used in this work was for the ER pharma-ink (170 ◦C), and according to the TGA results (Fig. 6B), pure biotin had a degradation onset of 250 ◦C, well above the maximum printing temperature. The observed weight loss in the pharma-ink before printing and in the printlet could be attributed to some residual moisture loss. Therefore, the printing temperatures used in this work (100 ◦C and 170 ◦C for IR and ER pharma-inks, respectively) did not lead to any biotin degradation, which was later confirmed with HPLC analysis. As shown in the x-ray diffractograms (Fig. 7), for both ER and IR pharma-inks, the crystallinity peaks of biotin can be observed. Peaks at 16◦and 17◦could be observed in printlet IR and the intensity was higher compared to pharma-ink IR. A peak at 22◦is observed in both the pharma-ink ER and printlet ER. These findings suggest that part of the biotin remains in a crystalline state within the printlets. Although crystalline biotin was not observed in the DSC analysis, it was confirmed by the x-ray diffractograms. Together, the DSC and XRPD results indicate that a portion of the biotin may be solubilized in the polymer matrix, while some remains in the crystalline form. In addition, PEO constituted the major excipient in the IR pharma-ink (60 % w/w), which may overlap with other biotin peaks in the IR printlet diffractogram, potentially obscuring some of the biotin signals. 3.2.2. FTIR analysis FTIR was performed to elucidate any possible chemical interactions between the drug and the excipients during the printing process (Fig. 8). The spectrum suggests that there were no chemical interactions between the drug and the excipients. All the bands present in the pharmaink and printlets can be attributed to the presence of HPC Klucel ELF and PEO due to the higher percentage in the formulation (95 % w/w for HPC Klucel ELF in ER samples and 60 % w/w for PEO in IR samples). It is possible that the drug peaks have overlapped with the characteristic bands of the HPC Klucel ELF and PEO. 3.2.3. SEM imaging The images obtained from SEM microscopy allowed observation of the surface and the cross-sections of the printlets (Fig. 9). Fig. 9B and D give visual representations of the different layers in the Fig. 4. Screenshot of the M3DIMAKER Studio™ software, showing the weights obtained in-line during the 3DP process using the integrated balance for (A) small IR printlets and (B) large IR printlets. Printlets within the accepted limits for the mass uniformity assay are shown in green. Printlets outside of the accepted limits would be shown in red. Table 3 Dimensions and weight results of each printed batch. Results are shown as mean ±standard deviation (n =10). CV is referred to as the Coefficient of Variation. Batch Diameter (mm) Length (mm) SA/V ratio (mm −1 ) Integrated balance weight (mg) CV integrated balance weight (%) IR small 3.11±0.06 7.97±0.17 1.62±0.03 57.00±4.00 6.21 IR large 4.60±0.10 12.89±0.10 1.07±0.02 197.00±7.00 3.74 ER small 2.71±0.07 8.71±0.13 1.79±0.04 62.00±6.00 10.04 ER large 4.51±0.11 13.93±0.09 1.08±0.02 194.00±11.00 5.87 G. Mora-Casta˜ no et al. International Journal of Pharmaceutics 668 (2025) 124964 6 printlets. In Fig. 9D, the layers of printlet ER are less defined and this can be attributed to the first printed layer which adheres to the printing platform and is squashed under the weight of the printlet. As depicted in Fig. 9A and C, the cross-sections of both printlets are homogeneous and smooth, but some white points can be observed. Those points are likely attributed to biotin, as indicated by XRPD results, which suggested the presence of biotin in its crystalline state. 3.3. Drug loading in the printlets The drug loading of the different batches was close to the theoretical drug loading of 5 % w/w (Table 4). The drug loading of the ER printlets was slightly lower than the theoretical value. This reduction in drug loading may be attributed to incomplete extraction before HPLC quantification. This ensures that there was no degradation during the printing process, as expected by the TGA results. 3.4. In vitro release studies Biotin release from ER and IR printlets is shown in Fig. 10. It was observed that the printlets eroded during the dissolution studies and the drug was progressively released from the dosage forms as they eroded. It was observed that the release from the small printlets was faster than the large printlets due to the dosage form size, independent of the composition. Each dosage form has a different surface area to volume ratio (SA/V) because of the different dimensions (Table 3), resulting in a slightly different drug release rate. For the IR printlets, 82 % and 100 % of the biotin was released within 45 min for the large and small printlets, respectively (Fig. 10A). According to the Ph. Eur., conventional IR dosage forms should release at least 75 % of the active substance within a specified time, typically 45 min or less (Europe, E.C.o. 5.17.1, 2024). Therefore, it was demonstrated that PEO-based printlets met the criteria for IR dosage forms, regardless of the biotin’s solid-state. As shown in the XRPD results, part of the biotin remained in the crystalline form; however this did not affect Fig. 5. In-line registered weights by the integrated balance during the printing process: (A) small IR printlet batch, (B) large IR printlet batch, (C) small ER printlet batch, and (D) large ER printlet batch (n =10). The red lines show the average weight of each batch (57 mg and 197 mg for the small and large size IR printlets, respectively; and 62 mg and 194 mg for small and large size ER printlets, respectively). The green dashed lines represent the accepted percentage of deviation of the average mass (±10 % and ±7.5 % of deviation for small and large printlets, respectively), percentages are different depending on average mass according to the corresponding Ph.Eur monograph (Europe, C.o. 2.9.5. Uniformity of mass os single-dose preparations. European Pharmacopoeia, 2024). The orange dashed lines represent twice the accepted percentage deviation from the average mass. G. Mora-Casta˜ no et al. International Journal of Pharmaceutics 668 (2025) 124964 7 the dissolution of the IR formulation, as the printlets complied with Ph. Eur. Specifications for IR dosage forms (at least 75 % of the drug released within 45 min). For the ER formulations, 35 % and 24 % of the biotin was released at 45 min for the small and large printlets, respectively (Fig. 10B). At 120 min, 100 % of the biotin was released from the small printlets, while 91 % was released from the large printlets by 210 min. Based on the obtained results, the HPC printlets complied with the ER dosage form requirements. At 120 min, the pH was modified for the ER printlets. As observed in Fig. 10B, the drug release rate for the large printlets was not affected by the pH increase, with 59 %, 78 %, 83 % and 91 % of the biotin released at 120, 150, 180, and 210 min, respectively. Therefore, the pH change did not impact the extended-release profile. In contrast, the small printlets released 100 % of the biotin at 120 min, so the pH change was not significant. Fig. 6. (A) DSC thermograms of pure biotin and excipients, pharma-inks, and printlets; (B) TGA thermograms of pure biotin and HPC Klucel ELF, pharma-ink before printing and printlet ER. Fig. 7. X-ray powder diffractograms of pure biotin and excipients, pharma-inks before printing and printlets. G. Mora-Casta˜ no et al. International Journal of Pharmaceutics 668 (2025) 124964 8 The balance-3D printer system accelerated, in terms of time and human resources, the mass uniformity assay which is mandatory for solid dosage forms according to various pharmacopoeias. This work successfully demonstrated the feasibility of a point-of-care manufacturing model, where pharma-inks are prepared by a GMPcompliant industry partner, and QC is ensured through an in-line PAT tool, specifically the integrated balance. Both the 3DP process and the printlet weight measurements were controlled by a single software system, facilitating increased throughput. Further integration of additional PAT tools could enable a more comprehensive analysis of printed drug products, ensuring they meet necessary Critical Quality Attributes (CQAs). For example, the integration of a pressure sensor into an SSE 3D printer enabled the monitoring of the printing process whilst simultaneously characterizing the rheological properties of the pharma-inks (Díaz-Torres, 2022). Additionally, an in-line near-infrared (NIR) spectroscopy system served as a PAT tool to quantify the drug load of the printlets (Seoane-Via˜ no, 2023). 4. Conclusions This study investigated, for the first time, the implementation of an in-line and software-controlled analytical balance within a pharmaceutical multi-printhead DPE printer. This represents a significant advancement in QC for the pharmaceutical DPE 3D printing process, particularly for low dose drug delivery systems, treatment personalization, and preclinical and clinical studies for rare diseases. Moreover, two different pharma-inks based on PEO (immediate-release excipient) and HPC (extended-release excipient) were developed for printing capsuleshaped printlets using DPE for the first time. Automated mass uniformity testing was conducted on different batches (n =10) of DPE 3D printed biotin formulations in two different sizes. The specialised integrated software successfully registered the weight of each printlet and detected any deviations from the acceptable limits. The mass uniformity results indicated that the printing process was more reliable for manufacturing large capsule-shaped printlets compared to small ones. However, further studies and improvements are necessary to ensure consistent DPE printing for small capsule-shaped printlets as well. Fig. 8. FTIR spectra of the excipients, pure biotin, pharma-inks, and printlets. G. Mora-Casta˜ no et al. International Journal of Pharmaceutics 668 (2025) 124964 9