Topical drugs are not developed by the same process as oral drugs. The process is more uncertain and contains gaps. This leads to a poor discharge of risks before going to the clinical phases. The topical drug development process is reviewed in the introduction of the thesis. In particular, past and current topical drug development practices are described and compared to the oral drug development process. The large risks taken during the topical drug development are pointed out. These risks are largely associated with a lack of pharmacokinetic's involvement prior to the drug candidate selection stage. Pharmacokinetics is considered after drug selection when it is often too late. Furthermore, the topical pharmacokinetic techniques available appear to be not suitable for three reasons: accessibility to the pharmacokinetic techniques, meaning of the data generated and reliability of these data. It concludes that the knowledge of target skin tissue concentration would be key for a more rational drug development process. To this end, the primary objective of this thesis is to define a way of measuring drug concentration in skin tissue after topical application that is reliable, effective and practical. A secondary objective is then from the knowledge of the skin tissue concentration, to develop a topical PharmacoKinetic/PharmacoDynamic model to predict likely efficacy for a topical drug candidate. First a direct skin tissue concentration approach is described that brings theoretical reliability into the pharmacokinetic data generated and improves throughput. However the pharmacokinetic data generated have limited use as total drug (bound + unbound) tissue concentration is measured while, pharmacodynamically, only the unbound fraction is of interest. An indirect skin tissue concentration determination is then proposed. It consists in predicting the in vivo unbound drug concentration in diseased skin tissues. Three steps are required: In the first step, the in vitro percutaneous flux is linked with the unbound drug concentration in the dermis. From there, the in vivo unbound drug concentration in all the skin tissues is defined using different physiological parameters. Finally, taking into account the effect of the skin disease on skin permeability and dermal capillary clearance, the in vivo unbound drug concentration in skin tissues in diseased skin is defined. The predicted concentration is therefore calculated from a constant (which is skin disease dependent) and from the in vitro percutaneous flux (which is an accessible and reliable experimental pharmacokinetic data). A PharmacoKinetic/PharmacoDynamic model is then built. This model delivers two types of information: -1- The "efficacy index" which is a prediction of efficacy for a drug candidate based on percutaneous flux and drug potency and -2- the "systemic safety index" which is an assessment of systemic exposure based on total systemic clearance and plasma protein binding. To check the validity of this new model, a validation exercise is run with the key eight topical drugs classes: NSAIDS, anaesthetics, retinoids, corticosteroids, vitamin D3 derivatives, antifungals, antibacterials for acne and immunomodulators. For seven out of the eight classes, the validation of the model is good. For the last class, the antibacterials for acne, the model underpredicts efficacy and it is suggested that the route of entry of antibacterial agents in acne occurs via the sebaceous duct as opposed to the more classic stratum corneum pathway. Finally, three pilot studies are conducted with the aim to improve the quality and relevance of the data generated with in vitro percutaneous flux studies as well as the access to this technique and throughput of this technique.

The existence of polymorphic forms of drug substances has implications for therapeutic performance, handling and storage. This study investigates the development of a novel approach to surface analysis of drug polymorphs, with the aim of extending the capabilities of this approach to perform real time analysis of polymorphic transformation during pharmaceutical product development. This was achieved here, using diffuse reflectance visible spectroscopy (DRVS) and the colour change which occurs when pH indicator dyes are deposited on the surface. The pH indicators used were phenol red (PR), thymol blue (TB) and methyl red (MR). Two polymorphs each of indomethacin (IMC), carbamazepine (CBZ), caffeine (CFN), sulfanilamide (SFN) and furosemide (FRS) were examined. The interaction of the adsorbed dye with each of the polymorphs showed different behaviour, manifested by different colours. An analysis of the crystal structures and the acid/base properties of the drug molecules provided a rationalisation for the different colours exhibited by the polymorphs‘ surfaces. The least stable form of each polymorphic pair studied showed more extensive interaction with the adsorbed dye molecules. Observed colour reveal underlying differences at a molecular level between the surfaces of pairs of polymorphs. The different colours exhibited by the indomethacin polymorphs were further examined using hygroscopicity studies, contact angle measurements and computer simulation. The contact angles of several liquids with the polymorph surface were measured in order to characterise the nature of the functional groups exposed on the surface of the polymorphs. The surface structure and external morphologies of polymorphs were predicted by molecular modelling using the attachment energy model. The predicted morphology was confirmed by scanning electron micrographs (SEM) and the miller index of the dominant face was confirmed by X-ray powder diffraction (XRD). Results revealed that although the surfaces of both polymorphs are largely hydrophobic, the metastable form- IMC-α has a greater number of polar functional groups on the surface. Further measurements were carried out using DRVS and adsorbed TB to study the kinetics of the solid-state transformation of SFN- to SFN-. The rate of transformation was followed at 128ºC by monitoring the ratio of the two DRVS bands at 454 nm and 604 nm. The kinetic data was analysed using sixteen solid-state kinetic models to obtain the best fit. The thermally induced polymorphic transformation of the SFN-β (particle size of ≥ 450μm) can be best described by the first order kinetic model (R2 = 0.992) with a rate constant, k of 2.43 x 102 s-1. The DRVS instrument used herein is not adapted for in situ studies; however, because of its non-destructive interaction with the sample and rapid data collection time of 5s per spectrum, it does offer considerable potential as a tool for real time monitoring of polymorphic transformation.

Conventional adult dosage forms such as tablets and capsules are often not suitable for the paediatric and geriatric population due to either swallowing difficulties or a requirement for tailored dosing to meet individual needs. Alternative oral formulations such as orally disintegrating tablets (ODTs) are available; however these usually require the incorporation of taste masking techniques. One approach to taste masking is to reduce contact between the bitter active pharmaceutical ingredient (API) and taste buds. This may be achieved by hindering release in the oral cavity using reverse enteric polymeric coatings. In vitro dissolution testing can be employed to elucidate taste masking capability by quantifying release of the API in simulated oral cavity conditions. This provides a robust analytical approach circumventing the expense and ethical challenges associated with human taste testing panels or animal testing. To achieve taste masking, drug release should be below the bitterness threshold concentration of the API. A vast array of dissolution methodologies has been employed in the evaluation of taste masked formulation performance in literature, with little agreement between approaches, and a lack of biorelevance. For optimal predictability, the dissolution test should be biorelevant and the dissolution media should mimic human saliva as closely as possible. Human saliva is thus a biological fluid of great importance in the field of dissolution testing. However, until now, no consensus has been reached on its key characteristics relevant to dissolution testing. As a result, it is difficult to select or develop an in vitro dissolution medium to best represent human saliva. In this thesis, for the first time, the pH, buffer capacity, surface tension, viscosity and flow rate of both unstimulated (US) and stimulated (SS) human saliva were investigated with a sufficient number of participants to generate statistically meaningful results (Chapter 3). This provides a platform of reference for future dissolution studies using simulated salivary fluids (SSFs).

Pharmacists in Jordan are greatly positioned to play a key role in shaping the future direction of the country’s healthcare system, particularly in terms of improving patient and health outcomes. However, the available literature reports a number of problems affecting the Jordanian pharmacy sector across the professional practice, education and regulation spheres; issues which have been hindering the adequate provision of pharmaceutical care services, as well as the overall development of the profession. This study aims to examine the current status of the pharmacy profession in Jordan, notably in terms of the challenges and opportunities it faces, with a particular focus on the educational sector. To that end, a mixed methods approach was employed in which a series of studies were conducted. In the first study, semi-structured interviews and focus group sessions were carried out to identify and explore the main challenges facing the profession. Interview transcripts were thematically analysed, with eight principal ‘challenge areas’ being identified. These results were validated by focus group findings, and tailored recommendations were produced to address each challenge. In the second study, data from the Official National Register of pharmacists were obtained, collated and analysed, so as to establish workforce trends. Geographical, sectoral and gender distribution imbalances were identified; gaps in the current workforce intelligence were also highlighted and discussed. The third study employed a set of surveys to collect academic and institutional capacity information from pharmacy schools in Jordan. Surveys were completed using data triangulated from multiple sources (including official documents, schools websites and dean interviews). In the final study, a secondary qualitative content analysis of the interview transcripts was undertaken to identify gaps in pharmacy graduates’ skills and competencies. The Global Competency Framework for pharmacy was used as the starter coding framework. Sixty eight (68) behaviours from across the Framework’s four competency domains were identified by participants as lacking in graduates. As such, this thesis represents the first holistic investigation of the status of the pharmacy profession, workforce and education in Jordan, including the opportunities and challenges faced by the sector. This work’s original contribution to knowledge lies not only in the new baseline information produced, but also in the evidence-based guidance and recommendations presented to local stakeholders and researchers.

The popularity of dry powder inhalers (DPIs) to deliver drugs to the lungs is constantly increasing thanks to their advantages over nebulisers and pressurised metered dose inhalers (pMDIs), including the high stability of dry powders, avoidance of propellant gases and ease of use. DPIs generate dry powder aerosols that deposit on the lung mucosa upon inhalation. In order to achieve the desired therapeutic outcome, drug particles must first dissolve in the lung lining fluids, then diffuse across these fluids to reach the epithelium and be absorbed. The fate of inhaled particles once deposited on the lung surfaces has not been yet fully understood. However, the particle physicochemical properties are believed to play a role on their dissolution, interaction with lung lining fluids and permeability across the lung epithelium. The main aim of this doctoral thesis was a better understanding of the relationships between drug particle physicochemical properties and their fate in the lung tissue in terms of dissolution and drug absorption. Increased knowledge in this area would indeed assist the development of novel and more effective inhaled medications. The first objective was the development and validation of a simple and low cost deposition system to apply aerosolised dry powder particles in a narrow size range and a controlled dose to both epithelial and non-epithelial lung models (Calu-3 cells grown at the air-liquid interface (ALI) and airway mucus) for in-vitro studies. The deposition system consisted of a vacuum desiccator fitted with a PennCentury Dry Powder Insufflator™ – Model DP-4 without the needle but equipped with a PennCentury Air Pump™ – Model AP-1 (Penn-Century. Inc. Wyndmoor, PA). We demonstrated that it was able to homogeneously disperse different types of dry powders (micronised and spray dried), and consistently deliver controlled doses of drug in a narrow particle size range (3-5 µm). However, the system presented a major limitation as no real separation between respirable (< 10 μm) and non-respirable (>10 μm) particles could be achieved. The system was then exploited to investigate the effect of the formulation on drug absorption across Calu-3 cell layers. Salbutamol sulfate and indomethacin, respectively in class III (high solubility, low permeability) and II (low solubility and high permeability) of the Biopharmaceutical Classification System (BCS), were chosen as model dry powders. It was demonstrated that for both drugs, a dry powder formulation led to a faster absorption across Calu-3 layers than their solution counterpart. Indomethacin was more permeable than salbutamol in either case, proving that our system was capable of discriminating between drugs with different permeability profiles according to the BCS. Indomethacin low water solubility did not limit its absorption. Accordingly, the potential of novel indomethacin formulations produced by colleagues at University College London (UCL) as platforms to improve the absorption of poorly soluble drugs could not be appreciated. In the case of salbutamol, we attempted to gain a better understanding of its mechanism of absorption through the lung epithelium, particularly when delivered as a dry powder. The data showed that Organic Cation Transporters (OCT) are likely to contribute to salbutamol absorption when applied in solution, but no valid conclusions could be drawn when the drug was delivered as dry powder due to Calu-3 cell layers being disrupted during the course of the experiment. Finally, the role of mucus on salbutamol and indomethacin particle dissolution and drug absorption was investigated. A system consisting of a thin mucus layer coating Transwell® insert membranes was developed. Drug permeation across the mucus layer was monitored and compared with that across the Calu-3 cell layers. The rate of permeation of salbutamol sulfate and indomethacin across the three barriers investigated (clean Transwell® inserts, mucus layer and Calu-3 cell layers) followed an opposite order (clean insert > mucus layer > Calu-3 cell layer for salbutamol sulfate, Calu-3 cell layer > mucus layer > clean insert for indomethacin), demonstrating that the mucus was acting as a barrier in the case of salbutamol, but conversely promoted dissolution of indomethacin particles. A contribution to the clarification of the role of the mucus was made with the identification of some of the parameters that affect drug-mucus interaction: ionisation and lipophilicity. Solubility in water did not seem to have the same impact as for oral delivery. In this respect, we showed that the BCS, which only takes into account drug solubility and permeability, was a non-adequate description for the prediction of the behaviour of indomethacin in the lungs.

The importance of localised delivery of chemotherapeutic drugs for cancer treatment and specifically solid tumours has been widely reported. In this study, the anticancer drug N4-myristoyl gemcitabine (a lipophilic form of gemcitabine) was formulated as organogel to achieve a localised depot delivery. Thus, the first goal of this study was to evaluate the suitability of the oragnogel for intartumoural injection and this attained by investigating the thermostability and elasticity of the organogel. Further to this, the second goal was to slow the release of N4-myristoyl gemcitabine from the organogel. Accomplishment of these two goals will guarantee a better efficacy of cancer treatment by obtaining direct contact of the organogel containing the N4-myristoyl gemcitabine with the cancerous cells. The studies herein selected the 12-hydroxystearic acid (12-HSA) as the gelator and using 2 types of solvents the liquid part of the organogel. The first type of solvent was a series of oils which were soybean oil (SO), medium chain triglyceride (MCT), glyceryl tributyrate (TGB) and glyceryl triacetate (GTA) whilst, the second type of solvent was propylene glycol (PG). Initially thermal stability was screened using table top rheology and DSC from 0.5% to 5% w/w 12-HSA in different oils. Also to test the mechanical strength of the organogels, amplitude sweep, frequency sweep, time dependant recovery and creep and recovery tests were executed to differentiate between the organogels. The best organogels were the 5% w/w 12-HSA in SO and MCT due to their highest thermal stability, denser scaffolds, thixotropic behaviour and were the least compliant. The same experiments were utilised to evaluate the selected range of 0.5% to 14% w/w 12-HSA in PG. 14% w/w 12-HSA in PG was selected again due to its higher thermal stability, thixotropic behaviour and was less compliant compared to other concentrations of 12-HSA in PG. Drug release from the selected organogels was then carried out. The cumulative percentage released from 0.5% and 0.3% w/w N4-myristoyl gemcitabine in 5% w/w 12-HSA/MCT organogels as a solid organogel was 18.95% and 26.62% after 30 days whilst for the organogel liquefied with N-methyl pyrrolidone (NMP), the cumulative percentage released was 35.02% and 34.37% within the same frame time. Further to this, a sample and separate release method was used to study the liquefied form of the 5% w/w 12-HSA/MCT. Also, this method revealed that the 5% w/w 12-HSA/MCT organogels gave a slow release of N4-myristoyl gemcitabine and 56.18% and 70.07% was released from the 0.5% and 0.3% w/w selected organogels respectively within 30 days. For the 14% w/w 12-HSA in PG organogel, the cumulative percentage released for 0.5% and 0.3% w/w N4-myristoyl gemcitabine in 14% w/w 12-HSA/PG organogels was 26% and 40% respectively after 30 days. To conclude, our selected organogels (5% w/w 12-HSA/MCT and 14% w/w 12-HSA/PG) met the goal of our work firstly, by showing the strength and the elasticity to be injected. Secondly, they were able to slow down the release of N4-myristoyl gemcitabine.

Para investigar o efeito de diferentes sistemas de manejo do solo no estoque de carbono e na composição química da matéria orgânica ao longo do perfil do solo e sobre sua capacidade de retenção de carbono, foram desenvolvidos dois estudos em Latossolos subtropicais. No primeiro estudo, foi investigada a influência do sistema de plantio direto (PD) e do sistema de preparo convencional (PC) do solo sobre o estoque de carbono e qualidade da matéria no perfil do solo e nas frações leve livre (FLL), leve oclusa (FLO) e pesada (FP) até 20 cm de profundidade de um Latossolo Bruno (LB) e de um Latossolo Vermelho (LV). No segundo, foi estudado o efeito de sistema de manejo do solo (PD e PC) sobre a capacidade de retenção de carbono nas frações silte e argila dos mesmos Latossolos do estudo I. Para tais finalidades, foram coletadas amostras deformadas de solo nas camadas de 0-2,5; 2,5-5; 5-10; 10-20; 20-30; 30-40; 40-60; 60-80 e 80-100 cm e indeformadas nas profundidades 0-5; 5-10 e 10-20 cm. O teor de carbono do solo e das frações foi determinado por combustão seca e a composição química da matéria orgânica foi analisada por técnicas espectroscópicas. Sob PD, o estoque de C foi superior (9,7 e 12,2 Mg ha-1 nas camadas de 0-2,5 cm e de 2,5-5 cm respectivamente) ao observado sob PC até 5 cm de profundidade (6,4 e 8,5 Mg ha-1 nas camadas de 0-2,5 cm e de 2,5-5 respectivamente) no LB e até 2,5 cm no LV (5,6 Mg ha-1 sob PD e 4,7 Mg ha-1 sob PC). A subsuperfície do solo armazenou 69% (no LB) e 75% (no LV) do carbono aportado anualmente ao perfil do solo em virtude do manejo do solo sob PD. Tanto no solo inteiro quanto nas frações (FLL, FLO e FP), o PD favoreceu a preservação de estruturas orgânicas lábeis e tal efeito foi mais pronunciado no LB em comparação ao LV. O LB apresentou maior retenção de C nas frações silte (52 g kg-1 sob PD e 41 g kg-1 sob PC) e argila (61 g kg-1 sob PD e 54 g kg-1 sob PC) do que o LV (Csilte = 32 g kg-1 sob PD e 31 g kg-1 sob PC e Cargila = 25 g kg-1 sob PD e sob PC), independente do manejo de solo. No LB, o sistema PD promoveu maior retenção de C nas frações comparativamente ao PC, enquanto no LV, a capacidade de retenção de C das frações não diferiu em função do manejo. Em ambos os Latossolos, a retenção de C pode ser explicada por dois processos principais de sorção; interação organo-mineral e auto-associação da matéria orgânica em estruturas tipo micelares. No LV, este último processo parece ser predominante, ao passo que no LB, ambos os mecanismos contribuem igualmente para o sequestro de carbono. / Two studies were developed in Oxisols for investigating the effect of different management systems on soil C stocks and on chemical composition of organic matter along the profile, and on the carbon retention capacity. In the first study, we investigated the the effect of no-tillage (NT) and conventional tillage (PC) on C stocks and composition of organic matter of the soil profile and of the free light fraction (FLF), occluded light fraction (OLF) and heavy fraction (HF) to 20 cm depth of a Brown Oxisol (BO) and a Red Oxisol (RO). In the second study, we investigated the C retention capacity in the silt and clay fractions of the Oxisols used in the first study. For these purposes, soil samples were collected in soil layers of 0-2.5, 2.5 to 5, 5-10, 10-20, 20-30, 30-40, 40-60, 60-80 and 80 -100 cm and undisturbed soil samples in the depth of 0-5, 5-10, 10-20 cm. The C content in soil and fractions was determined by dry combustion and chemical composition of organic matter was analyzed by spectroscopic techniques. Under NT, C stock was greater (9.7 e 12.2 Mg ha-1 in surface layers 0-2.5 cm and 2.5-5 cm respectively) than that found under CT (6.4 e 8.5 Mg ha-1 in surface layers 0-2.5 cm and 2.5-5 cm respectively) in the BO and up to 2.5 cm in the RO (5.6 Mg ha-1 under NT and 4.7 Mg ha-1 under CT). The subsoil stored 69% (in BO) and 75% (in RO) of carbon added annually to the soil profile due to soil management under NT system. In bulk soil and in the fractions (FLF, OLF and HF) as well, the NT promoted the preservation of labile organic structures. Such effect was more pronounced in BO in comparison to RO. The BO showed a greater C retention in silt (52 g kg-1 under NT and 41 g kg-1 under CT) and clay fractions (61 g kg-1 under NT and 54 g kg-1 under CT) than RO (Csilt = 32 g kg-1 under NT and 31 g kg-1 under CT and Cclay = 25 g kg-1 under NT and CT), regardless the soil management. Furthermore, in BO, the NT system promoted a greater C retention in the two fractions compared to CT, while in RO, the C retention capacity of the silt and clay fraction was not affected by the management system. In both Oxisols, C sequestration can be explained by two main sorption processes: organo-mineral interaction and self-assemblage of organic matter in micelle-like structures. In the RO, the latter process seems to be predominant, whereas in BO both mechanisms contribute equally to the C sequestration.

A estabilidade de agregados é influenciada pelos sistemas de manejo do solo, com reflexo na estabilização da matéria orgânica e estoques de C e N no solo. O efeito de longo duração (30 anos) de dois sistemas de preparo [plantio direto (PD) e preparo convencional (PC)] e de três sistemas de culturas: [aveia/milho (A/M), ervilhaca/milho (V/M) e aveia+ ervilhaca/milho+ caupi (A+V/M+C)] sobre os índices de agregação, conteúdo e frações de C e N de um Argissolo Vermelho distrófico tipico, foram avaliados. Amostras de solo foram coletadas nas camadas de 0-5, 5-10 e 10-20 cm e separadas em seis classes de agregados (>4,76; 4,76-2,00; 2,00-0,50; 0,50-0,250, 0,250-0,053 e <0,053 mm) por via úmida e seca, e avaliadas em relação ao teor de C e N e submetidas ao fracionamento granulométrico da matéria orgânica do solo (MOS). As concentrações de C e N no solo foram maiores no solo em PD e na camada de 0-5 cm. Os estoques de C e N foram 23 e 32% maiores respectivamente no sistema de culturas (0-20 cm) com elevado aporte de resíduos (A+V/M+C), em relação ao sistema com gramíneas (A/M). O diâmetro médio ponderado (DMP) e o índice de estabilidade dos agregados do solo [ (DMPu/DMPs) x100], foram superiores no solo em PD em relação ao solo em PC, devido ao predomínio de macroagregados (≥ 4,76-2,00 mm). Nos estoques das frações da MOS, o carbono orgânico particulado (COP) e o nitrogênio particulado (NP) foram maiores nos sistemas de culturas com leguminosas no solo em PD, em relação ao PC, enquanto que o C e o N associado aos minerais foram influenciados somente pelos sistemas de culturas. As taxas anuais de acúmulo de C na camada de 0-20 cm nos sistemas de culturas variaram de 35,3 a 67,6 kg C ha-1 ano-1, o acúmulo de N variou de 4,7 a 16,7 kg N ha-1 ano-1 sendo maiores no sistema A+V/M+C. Os estoques de C e N no solo foram influenciados em maior grau pelo sistema de culturas do que pelo sistema de preparo de solo. Esses resultados demonstram que em sistemas conservacionistas o aumento nos estoques de C e N do solo e a estabilidade dos agregados estão associados tanto com a não mobilização do solo quanto com o aporte de resíduos vegetais pelos sistemas de culturas. / The aggregate stability is influenced by soil management systems, with reflex on soil organic matter stabilization and C and N stocks. We evaluated the long term (30-yr) effect of two tillage systems [no-tillage (NT) and conventional tillage (CT) ] and three cropping systems [oat/maize (O/M), vetch/maize (V/M), and oat + vetch/maize + cowpea (O+V/M+C)] on stability of soil aggregates, C and N content in total soil organic matter (OM) and particulate and mineral-associated fractions of a sandy clay loam Typic Paleudult. Soil samples were collected at 0-5, 5-10 and 10-20 cm. Disturbed soil samples were collected for C and N analysis and OM fractionation; undisturbed soil samples were collected for soil aggregates analysis. Aggregates were separated in six classes (> 4.76; 4.76-2.00; 2.00-0.50; 0.50-0.250, 0.250- 0.053 and <0.053 mm) by dry (d) and wet (w) methods, the mean weighted diameter (MWD) of aggregates and the stability index [(MWDw/MWDd)x100] were calculated. ). The concentrations of C and N in soil were higher in NT in 0-5 cm layer. The C and N stocks were 23 and 32% higher in cropping system (0-20 cm) with a high crop residue input (O+V/M+C) than of grasses only based cropping system (O/M). The average diameter and stability index of soil aggregates were higher in NT in relation to CT, mainly due to predominance of macroaggregates (≥ 4.76 to 2.00 mm). In SOM fractions, particulate organic carbon (POC) and particulate nitrogen (NP) were higher in legume-based cropping systems in NT, related to CT, while the C and N in mineral-associated soil organic matter was affected by only cropping systems. The annual rates of C accumulation at 0-20 cm in cropping systems ranged from 35.3 to 67.6 kg ha-1 yr-1, N accumulation ranged from 4.7 to 16.7 kg ha-1 yr-1 and it was higher in the system O+V/M+C. Soil C and N stocks were influenced mainly by cropping systems than the tillage systems. Our results highlight that increase of soil C and N stocks are related both by no soil disturbance plus higher C input under conservationist soil management systems.

Extrusion based three dimensional (3D) printing is defined as a process used to make a 3D object layer by layer directly from a computer aided device (CAD). The application of extrusion based 3D printing process to manufacture functional oral solid tablets with relatively complex geometries is demonstrated in this thesis. In Chapter 3 the viability of using a basic desktop 3D printer (Fab@Home) to print functional guaifenesin bilayer tablets (GBTs) is demonstrated. Guaifenesin is an over the counter (OTC) water soluble medicine used as expectorant for reduction of chest congestion caused by common cold and infections in respiratory system. The bilayer tablets were printed using the standard pharmaceutical excipients; hydroxypropyl methyl cellulose (HPMC) 2208, 2910, sodium starch glycolate (SSG), microcrystalline cellulose (MCC) and polyacrylic acid (PAA) in order mimic the commercial model formulation (Mucinex®) guaifenesin extended-release bilayer tablets. The 3D printed guaifenesin bilayer tablets (GBTs) were evaluated for mechanical properties as a comparison to the commercial GBTs and were found to be within acceptable range as defined by the international standards stated in the USP. Drug releases from the 3D printed GBTs were decreased as the amount of HPMC 2208 increased due to the increased wettability, swelling properties and gel barrier formation of the HPMC. The 3D printed GBTs also showed, as required, two release profiles: immediate release (IR) from the top layer containing disintegrants; SSG and MCC and sustained release (SR) profile from the lower layer containing HPMC 2208. The kinetic drug release data from the 3D printed and commercial GBTs were best modelled using the Korsmeyer–Peppas model with n values between 0.27 and 0.44. This suggests Fickian diffusion drug release through a hydrated HPMC gel layer. Other physical characterisations: X-Ray Powder Diffraction (XRPD), Attenuated Total Reflectance Fourier Transform Infrared Spectroscopy (ATR-FTIR), and Differential Scanning Calorimetry (DSC) showed that there was no detectable interaction between guaifenesin and the used excipients in both 3D printed and commercial GBTs. A more complex printer (RegenHu 3D bioprinter) was subsequently used to print complex multi-active tablets containing captopril, nifedipine, and glipizide as a model therapeutic combination. These drugs are frequently used to treat hypertension and diabetes mellitus. The 3D printed tablets were evaluated for drug release and showed that captopril was released by osmosis through permeable cellulose acetate (CA) film and both glipizide and nifedipine were released by diffusion through the hydrophilic HPMC 2208 matrix. According to XRPD and ATR-FTIR results, there was no detectable interaction between the actives and the used excipients. In the final experimental chapter, a combined treatment regimen: atenolol, ramipril, hydrochlorothiazide (anti-hypertensive medications), pravastatin (cholesterol lowering agent), and aspirin (anti-platelets) were printed into more complex geometry (polypill) using the RegenHu 3D bioprinter. This combined drug regimen is manufactured by Cadila Pharmaceuticals Limited as a capsule formulation under the trade name of Polycap™ and is currently the only polypill formulation commercially available and is used to treat and prevent cardiovascular diseases. The printed polypills were characterized for drug release using USP dissolution testing and showed the intended immediate and sustained release profiles based upon the active/excipient ratio used. Aspirin and hydrochlorothiazide were immediately released after the polypill contacted the dissolution medium, and atenolol, ramipril, and pravastatin were released over a period of 12 hrs. XRPD and ATR-FTIR showed that there was no detectable interaction between the actives and the used excipients. In this work, extrusion based 3D printing technique was used to print oral solid dosage forms with complex and well-defined geometries and function. The technology of 3D printing could offer the opportunity to print oral tablets with high and precise drug dosing and controlled drug release profiles tailored for sub-populations or individuals. If the manufacturing and regulatory issues associated with 3DP can be resolved such personalised medicine delivered by 3D printing could improve patient compliance and provide more effective treatment regimes.

Successful tissue repair requires orchestrating a range of biochemical and biophysical factors to direct cell differentiation towards tissue specific lineages. Biodegradable poly DL lactic acid-co-glycolic acid (PDLLGA) microspheres have been reported as a promising injectable cell delivery system with controllable growth factor release potential for different tissue engineering applications. Injectable PDLLGA microspheres have been shown to form highly porous scaffolds at body temperature with a mechanical strength comparable to bone tissues. However, as the elastic properties of the injury microenvironment were shown to have a pivotal role on directing stem cell lineage specification, this work has proposed photo-crosslinkable gelatine methacrylate (gel-MA) hydrogels as promising surface coatings with tunable elastic properties. Moreover, as PDLLGA microspheres have limited functional groups on their surface, this work has evaluated different surface modification and grafting approaches to enable proper grafting of thick gel-MA hydrogel layer to the surface. Surface adsorption, surface entrapment, and oxygen plasma treatment approaches have been proposed and evaluated to modify the surface of PDLLGA microspheres with high density of gel-MA molecules. Surface analytical techniques such as ToF SIMs and XPS have been used to evaluate and quantify the density of gel-MA molecules on the surface, while fluorescent and scanning electron microscopies have been used to visualise the fluorescent deposition of fit-C gel-MA to the surface. Later, grafting-to and encapsulation approaches have been investigated to graft a thick layer (10-20 μm thick) of gel-MA hydrogel to the surface of PDLLGA microspheres following modification with gel-MA. Fluorescein isothiocyanate labelled human serum albumin Fit-C HSA has been loaded into PDLLGA microspheres as a model protein to study its release behaviour from the proposed system. Release data have shown a comparable release profile between PDLLGA microspheres before and after coating with the hydrogel layer suggesting no adverse effect of the proposed coating approach on the release behaviour. Gel-MA hydrogels with tunable elastic properties have been prepared and analysed using texture analyser and atomic force microscopy (AFM). Hydrogels have been later imaged with focused ion beam scanning electron microscope (FIB-SEM) using a novel approach to capture the hydrated structure of the hydrogel. Data obtained from the texture analyser using the compression and indentation mode tests have shown that the elastic modulus values were significantly higher than the values obtained from tension mode tests. In comparison, the values obtained from the texture analyser with the tension mode test were comparable with the values obtained using the AFM nano-indentation tests. This has been explained with the poroelastic behaviour of the hydrated hydrogel structure where a micron size pores have been observed. To verify findings, human mesenchymal stem cells have been cultured on the surface of gel-MA hydrogels to study their phenotypic behaviour and stained with anti-osteogenic or anti-neurogenic immunofluorescent markers to define their fate accordingly. Images have shown that cells cultured on hydrogels with AFM analysed elastic values of (~26, ~9.3, and ~0.1 KPa) have committed to a phenotypic behaviour related to the elastic modulus values of bone, muscle, and neuronal tissues respectively. In comparison, the elastic modulus values obtained from gel-MA hydrogel microbeads with AFM have been notably higher and appeared to be dependent on the cross-linking temperatures. Finally, the proposed cell delivery system can be used to control the chemical and the mechanical properties of the stem cell microenvironment which may pave the way towards directing stem cell differentiation into tissue specific cell lineages for different tissue repair applications.