Enhancement of the drug efficacy and elimination of the side effects resulting from drug overdoses are an essential aspect in drug therapy. To achieve these demands two general guidelines have been used; producing new drugs with higher selectivity and therefore less side effects and improving controlled/sustained drug delivery agents based on polymers. Thus, the relationship between the active pharmaceutical ingredient and the polymeric system is important in the development of a drug delivery system and several considerations need to be taken in to account, for example the polymer should be biocompatible, biodegradable, and non-toxic and physiochemical properties. Because mucus is the first barrier with which food and drugs can interact with and diffuse through to be absorbed and enter the circulatory system, characterisation of mucin is an essential step towards establishing suitable pharmaceutical excipients. Therefore, the aim of the present study was to investigate the potential to construct and study drug delivery systems based on polysaccharides. The physicochemical characterisation of extensively degraded pig gastric mucin was studied and revealed that this type of mucin contains: protein, carbohydrate (Fuc, Gal, GalN, GlcN) and sialic acid, which provides the negative charges that becomes progressively stronger with increasing pH. The measurements of viscosity vs. shear rate showed that mucin has a shear thinning behaviour and a relatively low viscosity which is consistent with a high critical overlap concentration (c*), small hydrodynamic size and hence compact structure. The insight in to the compositional, hydrodynamic and viscoelastic properties support the understanding of mucin interactions with polysaccharide based drug delivery systems. Several polysaccharides including chitosan (Cs), two grade of alginates; high guluronate alginate (HGA) and low guluronate alginate (LGA) (which differ in structural conformation) and two kinds of pectin; high methoxyl pectin (HMP) and low methoxyl pectin (LMP) (with different degrees of esterification) have been characterised. The structure of these polysaccharides as powder have been studied; Fourier transform infrared spectroscopy ) findings indicate the structure and the function group for each polysaccharide whereas powder X-ray diffraction measurements displays that all the polysaccharide which were analysed are amorphous in nature except LMP which has a number of sharp crystalline peaks. In addition, solution properties of these polysaccharides such as zeta potential and intrinsic viscosity were investigated at several ionic strengths and pH. Furthermore the molecular weights were evaluated based on intrinsic viscosity and the Smidsrød-Haug stiffness parameter (B) and intrinsic persistence length (Lp) were estimated using the novel ionic strength dependency of zeta potential method and intrinsic viscosity (traditional method). The interaction between polysaccharides and pig gastric mucin were evaluated based on relative viscosity. It has been suggested that polysaccharide–mucin interactions are not only driven by electrostatic forces, but also by the molecular weight, conformation and flexibility of the polymer also played significant roles. As the mucin-HGA system displayed exceptionally high viscosity, the viscoelastic properties of this system were extensively studied. The mechanical spectra of the mucin-HGA blends indicate that with the exception of the system involving only HGA (0 % mucin) and 60 % mucin, all mixtures including mucin itself displayed typical ‘weak gel’ rheological behaviour and the gel became stronger with decreasing HGA content in the system. Moreover 80 % of mucin was successfully encapsulated within phospholipids bilayer using liposomal encapsulation technology. The liposomal vesicles with encapsulated mucin display larger sizes than the control vesicles (prepared in DI water) this may be due to the electrostatic interaction between mucin molecules and phospholipid which is the main component the vesicles. In the final part of the thesis the hydrogel containing chitosan and naturally occurring polyanions and its potential for drug release were studied. Chitosan - polyanion (HGA, LGA, HMP and LMP) hydrogels complexes were successfully prepared (in acetate buffer 0.05M, 4.3 pH) at various ratios (10 %, 30 %, 50 %, 70 % 90 % of Cs) using the ionotropic gelation method. The freeze dried hydrogels were characterized by FT-IR and XRD and the results confirmed the electrostatic interactions between chitosan and polyanions at all ratios and percentage yield of hydrogel ζ and ηsp results of the supernatant was determined and it was found that the optimum ratios 3:7 and 1:1 of chitosan-pectins and chitosan-alginates respectively. The hydrogels of ideal ratios were studied by determining zeta potential, particles size, water uptake, morphology by scanning electron microscopy for freeze dried hydrogels and optical microscopy analysis for homogenous suspension. In addition, dynamic small deformation oscillatory measurements and adhesion property were studied. Finally, ibuprofen was successfully encapsulated by the chitosan-polyanion hydrogel complexes and the encapsulation efficiency of the formulations was calculated. Finally the drug release behaviour of the formulations was in vitro assessed over the time. The findings demonstrated that HMP and LGA hydrogels displayed the highest percentage of retained ibuprofen followed by HGA and LMP. This could be attributed to the fibrous appearance small size of pores which may impedes movements of entrapped molecules.

Chemotherapy-induced alopecia (CIA) is widely regarded as the most traumatic side effect associated with cancer treatment and the associated stress can be detrimental to overall outcomes. Yet there has been little research into its pathobiology and no pharmaceutical intervention is available. CIA is caused by chemotherapy-mediated damage of the rapidly dividing cells of the hair follicle and although it is normally reversible, on regrowth the hair is often different in colour and/or texture and only grows gradually. The only effective treatment for CIA currently available is scalp cooling. It has been hypothesised that scalp cooling works by a combination of vasoconstriction, a reduction in the metabolic rate and/or reduced drug uptake by cells in the hair bulb. The ability of cooling to protect from CIA has been clinically demonstrated for years yet, to date, no cell biology is available to support its cytoprotective effects. The overall aim of this work was to for the first time provide a systematic investigation of the effects of cooling on chemotherapy-induced toxicity in human cells. The work established cellular models to determine the efficacy of cooling in rescuing from toxicity, investigate the temperature conditions providing maximal rescue and understand not only the mechanisms responsible for drug-mediated cytotoxicity, but also the way in which cooling regulates such mechanisms. Various human keratinocyte models were established, including normal (epidermal, NHEK, and follicular, HHFK) cells and adapted HaCaT (HaCaTa) cells. Viability, cell cycle and apoptosis assays were used, alongside Reactive Oxygen Species (ROS) detection, mitochondrial integrity assays and Western blotting, as well as functional pharmacological inhibition experiments. A panel of chemotherapy drugs commonly used in the clinic were employed, including doxorubicin, docetaxel and active metabolite of cyclophosphamide, 4-hydroxy-cyclophosphamide (4-OH-CP) and 5-FU, whilst a series of temperature conditions were tested, including 22°C as well as more severe cooling, particularly 18°C and 14°C (and even extreme cooling at 10°C). This study showed that cooling dramatically reduces or completely prevents the cytotoxic effects of docetaxel (T), doxorubicin (A), 5-FU (F) and particularly 4-OH-CP (C); however, optimal rescue was observed in conjunction with mono-therapy treatments (and substantial rescue with dual therapies, e.g. AC), whereas combinatorial treatment (TAC) showed relatively poor response to cooling, in agreement with clinical observations. Importantly, the work demonstrated that lowering the temperature below the widely accepted 22C threshold, even by a small number of degrees (e.g. 18C), resulted in significantly improved or even complete cytoprotection, a striking observation strongly suggesting that the scalp temperature achieved clinically is of critical importance in dictating the success of head cooling in CIA prevention. The panel of chemotherapy drugs tested caused differential effects on keratinocyte cell cycle progression and drug-mediated cell cycle arrest was significantly attenuated by cooling. Notably, cooling alone appeared to decelerate cell cycle progression, providing evidence for metabolic effects. More importantly, protective pre-conditioning (PPC) achieved either by growth factor removal or pharmacological inhibition of EGFR activation enhanced the cytoprotective effects of cooling and significantly reduced the effects of the chemotherapy drugs. As the ability of PPC to enhance protection from drug cytotoxicity could be attributed to its propensity to regulate the cell cycle progression, the work provided evidence that one mechanism via which cooling cytoprotects might be due to its ability to decelerate cell cycle progression. Disruption of mitochondrial membrane potential and elevation of ROS indicated the activation of an apoptotic pathway, which was confirmed by cell death-specific assays that confirmed a mitochondrial apoptotic pathway, as evident by plasma membrane disruption, caspase activation and DNA fragmentation. Importantly, cooling at a variety of temperatures (but mainly at or below 18C) attenuated drug-mediated apoptosis. To further investigate the precise mechanisms of growth arrest and/or cytotoxicity, activation/regulation of critical intracellular signalling mediators was investigated at the protein level. The majority of the drugs used induced activation of p53 and subsequent induction of p53-inducible mediators such as p21, as well as pro-apoptotic mediators associated with the mitochondrial pathway, such as Bak, PUMA and Noxa, whilst induction of pro-apoptotic FasL and Bid cleavage was detected, suggesting possible cross-talk with the extrinsic apoptotic pathway. Strikingly, cooling attenuated or blocked in a time- and, more importantly, temperature-dependent fashion induction of these pro-apoptotic mediators (an effect that became more marked as the temperature was reduced from 37C, to 22C, 18C and 14C); these results have provided for the first time a more detailed mechanistic explanation for the cytoprotective effects of cooling. As ROS appeared to be important in cytotoxicity, the hypothesis raised was that the cytoprotective effect of cooling might be enhanced via co-treatment with an antioxidant (e.g. NAC), aimed at enhancing the cytoprotective capacity of cooling at sub-optimal temperatures (such as 26°C). The findings presented here suggested that cooling plus topical treatment with antioxidants might represent a promising approach to improve the cytoprotective effects without compromising the anticancer effects of chemotherapy. Overall, despite their reductive nature, these in vitro models have provided experimental evidence for the ability of cooling to rescue from chemotherapy drug-mediated toxicity and shown that the choice of temperature may be critical in determining the efficacy of cooling in the clinic. This, whilst generating a novel combinatorial approach that has the potential to significantly enhance the ability of scalp cooling to protect against CIA in the clinic.

Regulatory toxicology in the 21st century is faced with the challenge of having to replace its use of experimental animals in chemical risk assessment with alternative methods. This is due to the introduction of the REACH legislation and the seventh amendment to the cosmetics directive. Such alternative methods include the use of in vitro (cell culture/tissue etc.), in chemico (chemical experiments e.g. determination of reactivity) and in silico (computational) approaches. Importantly, it is envisaged that data from all these alternative sources will be required for the prediction of the animal-based endpoints used in regulatory toxicology. One of the key computational approaches used for data gap filling is category formation and read-across. When using this approach to assess the potential toxicity of a chemical, a chemical category is best defined based on a common molecular initiating event e.g. the formation of a covalent bond with biological nucleophile via the same chemical mechanism. The structural features that define a chemical's membership of such a category can be encoded computationally as structural alerts, which in turn, can be grouped together to form an in silico profiler. The work discussed in this thesis addresses the key shortcoming of traditional in silico profilers, this being that current in silico profilers provided no information about the rate of covalent bond formation for chemicals containing the same structural alert but with different substituents. The research within this thesis addresses this problem through the introduction of a fragment-based approach to in silico profiler development. This fragment-based approach introduces the use of calculated activation energies determined through the use of quantum mechanics calculations which enable chemical reactivity to be predicted. Chapter 3 outlines the development of the approach for α,β-unsaturated aldehydes, ketones and esters which form covalent bonds through Michael addition. Chapter 4 extends the work outlined in Chapter 3 demonstrating how the fragment-based profiler can be used to predict both chemical reactivity and skin sensitisation and toxicity to Tetrahymena pyriformis. Finally, Chapter 5 extends the approach to chemicals capable of reacting with proteins via an SN2 mechanism demonstrating the approach can be applied to any mechanistic domain for which data exist. Overall, this thesis outlines an approach for the development of novel fragment-based in silico profilers capable of quantitatively predicting chemical reactivity and by extension toxicity. It is envisaged that the work outlined in this thesis will be of use primarily in regulatory toxicology, within such tools as the OECD QSAR toolbox.

Cell-based therapeutic interventions are being developed for a variety of clinical indications, including irreversible retinal pathologies and stroke. Numerous cell therapy procedures use injection-based administration to deliver high density cell preparations, either systemically or directly. The mode of delivery of fragile cells can compromise treatment efficacy, which is dependent on cell viability and functionality post-injection. Reviewing current literature, there is a lack of comprehensive testing of the effects of injection-based cell delivery on the various parameters of cell function. This study investigated the effects of the administration process on a range of cell characteristics, and aimed to answer critical questions regarding possible reasons for failure to deliver a sufficient numbers of viable cells. Biomaterial-assisted cell delivery was also investigated to determine improvement of cell recovery and possible influence on cell fate. Primary human mesenchymal stem cells (hMSCs) and Swiss mouse embryonic fibroblast cell line (NIH 3T3) suspensions were drawn up into 100 μL Hamilton syringes with 30- and 34G needles. They were then ejected at rates ranging from 10-300 μL/min. A comprehensive toolset was employed to assess the effects of various injection parameters, including ejection rate and needle size. Cell dose recovery, viability, apoptosis, senescence and other parameters of cellular health were evaluated using various standard and multiplex assays. Trilineage differentiation potential of ejected hMSCs was also assessed. Moreover, various injectable cell carriers were explored in terms of improvement of cell recovery and potential influence on differentiation capacity. Ejections at slower flow rates resulted in a significantly lower percentage of dose delivered as viable cells, with ejections at 300 μL/min showing the maximum percentage of hMSCs dose delivered at 77.6 ± 11.7%. Lower cell numbers delivered at slower ejection rates were mainly attributed to cell retention within the delivery device. Normalised caspase-3/7 activity measurements ejected at 10 μL/min were also significantly higher than control. Quantification of differentiation of ejected hMSCs revealed that both ejection rate and cell carrier employed may exert an effect on differentiation capacity. The use of biomaterials as cell carriers significantly improved cell recovery. Ejection of hMSCs in gelatin solution resulted in 87.5 ± 14% of the cell dose being delivered as viable cells, in comparison with 32.2 ± 19% of the dose ejected in phosphate buffered saline (PBS) at 10 μL/min. This study shows that ejection rate, needle size and cell carrier have a significant impact on the percentage of cell dose delivered as viable cells, cellular health and differentiation potential post-ejection. Optimal delivery strategies for injectable cell-based therapeutics are required to enhance their efficacy and reproducibility. This study emphasises the importance of careful consideration of administration protocols, according to the nature of the administered cells and cellular responses post-ejection. The combination of the investigated factors, among others, may also influence the fate of stem cells injected, thereby affecting the success of cell-based therapies.

The aim of this study was to study the electrophysiological and mechanical effects of the gap junction uncoupler 1-heptanol in the left ventricle (epicardial surface) of the artificially perfused Langendorff rabbit heart. Specifically, electrical restitution and the dispersion of repolarisation were studied. Methods. Using a single monophasic action potential (MAP) electrode, in the healthy and failing (coronary ligated) heart, the effect of 1-heptanol was studied on rate dependent changes in action potential duration. Dispersion of repolarisation was measured sequentially. A 32 MAP electrode array was developed to simultaneously record dispersion of repolarisation from the epicardial surface of the left ventricle of healthy hearts. Restitution was studied using an extrastimulus protocol that involved electrically stimulating the heart with 16 S1 stimuli (350ms intervals), and an extrastimulus S2. S1-S2 interval was increased progressively from 70 to 600ms. S1-S2 changes of 5ms were made between 70 and 150ms, 10ms between 150 and 350ms, and 50ms between 350 and 600ms. Protocols were run at 37°C, initially in Tyrode's solution, then after addition of 0.3mM 1-heptanol. Results. The single catheter study showed that failure significantly (P<0.05) prolongs MAP duration between cycle lengths of 250ms and 650ms. No base to apex changes, changes in dispersion of repolarisation or ventricular fibrillation thresholds were observed. 1-Heptanol, at cycle lengths above 350ms, significantly (P<0.05) decreased MAP duration in failing and healthy hearts. 1-Heptanol however did not alter the dispersion of repolarisation or ventricular fibrillation threshold in healthy and failing hearts. The last SI MAP in the 16 beat train and the S2 MAP obtained using the 32 electrode array were analysed at 90% repolarisation (MAPD90). S2 MAPD90 increased with S I-S2 interval up to -180ms but decreased at longer intervals. 0.3mM l-heptanol exacerbated this negative slope in the restitution curve from (mean±SEM) -0.031±O.004 in Tyrode's compared to -O.063±O.005 in O.3mM 1-heptanol (P<O.OOI). Dispersion of repolarisation was increased (p>O.05) in the presence of O.3mM l-heptanol. Conduction delay was increased from (mean±SEM) 44.2±0.82ms to 49.2±O.87ms (P<O.OOI). The possibility of an effect on the single cell being the mechanism behind the exacerbation of the negative slope by l-heptanol was investigated in a single cell study. The effect of l-heptanol on single cell fractional shortening and Ca2+ handling was examined. At 1 and 3Hz l-heptanol decreases fractional cell shortening from (mean±SEM): at 1Hz 9.3±0.8% in Krebs to 5.7±O.7% (O.03mM 1-heptanol), 5.6±O.9% (O.lmM l-heptanol) and 3.2±0.8% (O.3mM lheptanol) (P<O.Ol); at 3Hz lO.6±O.8% in Krebs to 6.S±1.4% (O.03mM l-heptanol), 7.7±1.4% (O.lmM l-heptanol) and 4.7±l.S% (O.3mM l-heptanol) (P<O.Ol). The negative inotropic (Le. the reduction in contractility) effect of I-heptanol indicated that this agent is not a specific gap junction uncoupler. SR ci+ release was reduced at 3Hz only by (mean±SEM) 7.8±O.04% (O.03mM l-heptanol), 5.0±0.03% (O.lmM 1-heptanol) and 7.9±O.02% (P<O.OS),indicating an effect on the L-type Ca2+ channel or ryanodine receptor. The effect on the L-type Ca2+ channel was investigated by the use of nifedipine firstly in single cells then in the whole heart. Nifedipine decreased single cell fractional shortening at 1Hz from (mean±SEM) 8.9±O.7% in Krebs to S.8±O.7% (O.lJ,tM nifedipine), 3.4±O.9% (O.l5J,tM nifedipine, P<O.OOI)and O.8±O.2% (O.2JlM nifedipine, P<O.OOl). SR Ca2+ release was also reduced from (mean±SEM) O.75±O.017% in Krebs to O.72±O.024% (O.IJlM nifedipine), O.73±0.024% (O.l5JlM nifedipine), O.71±O.016% (O.2J,tM nifedipine) (P<O.OS). In the whole heart nifedipine did not induce a negative slope in the restitution curve indicating no role of the L-type Ca2+ channel in this phenomenon. Carbenoxolone, a novel specific gap junction uncoupler failed to induce a negative slope in the electrical restitution curve of the whole heart but did increase dispersion of repolarisation (P>0.05) and caused a significant conduction slowing from (mean±SEM) 45.50±2.I2ms in Tyrode to 55.11±2.82ms in carbenoxolone (P<0.05). Carbenoxolone has an inconsistent effect on single cell fractional shortening and Ca2+ handling. Conclusions. The biphasic relationship and the increased dispersion of repolarisation in the presence of 0.3mM I-heptanol may have implications for the development of alternans and/or arrhythmias (Gilmour and Chialvo, 2000). The cause of the negative slope is as yet unknown. but it is likely that it is an effect on the single cell rather than gap junction uncoupling.

Herein, two translational issues related to the development of cell therapy discoveries into therapeutic products, are addressed. Firstly, analysis of whether the manipulations required to transfer cells from cell culture conditions to a target tissue affect cellular characteristics was performed. It was shown that processing primary murine mesenchymal stem cells (MSCs) into a concentrated cell suspension, drawing them up into a syringe and immediately ejecting them, caused a significant viability decrease. Leaving the cells within the syringe chamber at room temperature for prolonged time periods caused to a further decrease in viability. However, cells that were viable post-ejection were found to be functional with regard to their ability to attach and proliferate. Reducing the ejection rate or using the antioxidant n-acetyl cysteine did not significantly improve viability, although using a wider bore 22g needle did improve viability. Secondly, the feasibility of a cell transportation device that could store viable cells under room conditions was assessed. The development of such a transportation device would remove the need to cyropreserve cells during transit, which has many flaws. Human MSCs were found to enter a reversible proliferative arrest phase whilst under room conditions, allowing them to be stored for up to 11 days, before rapidly decreasing in viability. The accumulation of ammonia was identified within the cultures, and the introduction of a zeolite material was found to partially remove this ammonia and improve viability over 7 days. A means of developing a concentrated cell culture media, using a freeze-drying technique, was found not to comprise cell viability, whilst allowing the nutrient volume to be reduced, thus potentially making the transportation device more compact. However, the introduction of a modified release nutrient gel, consisting of alginate, did not provide any significant effect on viability.

Background. Antipsychotic medications are widely prescribed for schizophrenia and other related psychiatric conditions, but can have serious financial costs and adverse effects. The evidence base concerning the efficacy; effectiveness; and adverse effects of antipsychotic medications is extensive but variable in quality and applicability, and controversy continues to exist as to whether the newer medications are superior to the older antipsychotics. Locally derived data on the use and effectiveness of these medications can inform their future use, and complement the national and international studies. Aims & objectives To review the pertinent literature regarding antipsychotic medication, and to examine the use and clinical effectiveness of antipsychotic medication in a local context. Also to develop a valid but pragmatic scale to monitor the adverse side effects of antipsychotic medications. Methods A pre-existing case register was analysed to describe the contemporary patterns of antipsychotic usage. For the original data containing studies, one prospective and two retrospective attributions of clinical global impression (CGI) scores, as well as continuation and hospitalization rates were examined. The side effect scale (GASS) was devised after literature review and patient consultation, and tested on consenting patients in comparison to a well established existing scale (LUNSERS) and on healthy individuals. Results Data from the Glasgow city case register shows antipsychotics are widely and appropriately prescribed there but polypharmacy is common. In the prospective 6 month study, olanzapine and risperidone produced significant improvements in CGI but lack of power precluded similar conclusions with amisulpride, clozapine, and quetiapine. In the retrospective studies, clozapine was clinically superior to other oral antipsychotics but there was no significant clinical difference between the main 3 depot or long acting antipsychotics studied. The new side-effect scale – the GASS - was found to be easy to use and as discriminating as the LUNSERS. Discussion There can be difficulties generalizing data from short term RCTs to routine clinical practice. However this thesis demonstrates that simple but robust measures such as the CGI or GASS can be used to structure and inform everyday clinical practice. Consistent with the evolving debate on the relative merits of individual medications, this thesis showed there was little difference in clinical effectiveness between various oral antipsychotics, with the exception of clozapine. The lack of a significant difference between the old and newer long acting injectable antipsychotics is a new finding, and this area merits further study. Conclusions Structured routine monitoring of outcomes is possible in the NHS with regard to antipsychotic medication. Oral and LAI (or depot) antipsychotic medications continue to differentiated more by their adverse side effect profile rather than their relative effectiveness, with the exception of clozapine A new short, inclusive, and valid side-effect monitoring scale – the GASS - is introduced.

Back pain is very common; almost half the adult population of the UK (49%) report low back pain. Typically, the pain is in one area of the lower back, but sometimes it spreads to one or both buttocks or thighs. Most cases of lower back pain are known as 'non-specific' because they are not caused by serious damage or disease, but by sprains, muscle strains, minor injuries or a pinched or irritated nerve. The Bowen Technique is a soft tissue remedial therapy, which involves the therapist using fingers or thumbs to move over muscle, ligament tendon and fascia in various parts of the body. The aim of this pilot study is to evaluate the feasibility of conducting a trial into the effectiveness of the Bowen Technique as a treatment for the management of chronic, non-specific low back pain (CNSLBP), using a Randomised Control Trial approach. Thirty-seven participants, 21 female, with a mean age of 44.5 years took part in the study. Nineteen were allocated to the experimental, Bowen, group, and 18 to a control 'Sham Bowen' group. Participants were blinded to their group allocation. Each participant received three weekly treatments, and were asked to complete a questionnaire comprising six different measures before treatment, one week after their final treatment, and four weeks later. Measuring pain and functioning levels, psychosocial/somatic changes and general health, 24 'categories' were created from the measures. The Bowen group recorded a positive change by the second follow up in 20 of these categories. By contrast the Sham group showed an improvement in twelve at the same time point. This will only be the second ethically approved study in the Bowen Technique to date, and shows that with modifications to the study, it is feasible to conduct a larger-scale trial into the effectiveness of the Bowen Technique as a treatment for CNSLBP.

Background: Fractures are common. The annual incidence of fresh fractures is estimated to be 3.6 per 100 people in England and Wales. Traditional management, employing nonoperative immobilisation strategies, has been replaced by a much more operative approach. The aim of this thesis is to understand and add to the evidence that supports available biological adjuncts to fracture management. Exploration of the evidence through synthesis: Several systematic reviews have been performed to review the currently available evidence for the clinical effectiveness of biological adjuncts. Adjuncts were chosen that are currently available for use in routine clinical practice. These reviews particularly focused on identifying where the evidence was of poor quality or did not exist to support such clinical applications. These reviews found that overall the quality and breadth of the evidence was limited. This was particularly true for the evidence to support the use of demineralised bone matrix and platelet-rich therapies. Ultrasound and electromagnetic induction were more established and few opportunities were identified to further test their effectiveness. Development and reporting of a test of clinical effectiveness: The development of a protocol followed, which involved determining an appropriate experimental model in which to test one of the biological adjuncts. This protocol reflected a determination to test the clinical effectiveness, rather than efficacy, of platelet-rich plasma. The trial did not demonstrate a significant effect of platelet-rich plasma in the treatment of patients with an internally fixed intracapsular fracture of the proximal femur. The effect estimate included potentially important benefits or harms from the treatment. Discussion: This thesis summarises the diverse evidence available concerning biological adjuncts, proposes a protocol to test the clinical effectiveness of one, and tests this in a pragmatic controlled trial. The challenges, exemplified in the trial reported here, to the conduct of randomised trials of complex interventions in surgery are highlighted and approaches to overcome these obstacles discussed.

An understanding of the criteria governing lymphatic absorption of drugs from the gastrointestinal tract may lead to selective uptake of drugs via this route. Lymphatic absorption may have potential advantages over the portal absorption of drugs; first pass elimination may be avoided as the drug reaches the systemic circulation before it reaches the liver, the absorption of some poorly absorbed drugs may be improved and it may provide a means of targeting anticancer agents directly to the lymphatics. In the present study, the lymphatic uptake of a series of retinoids was investigated after oral administration. The most important factors which were found to affect lymphatic absorption were, the chemical nature of the compound and the nature of the oily vehicle in which the retinoid was administered. It was found that the greater the lipophilicity of the retinoid, (as defined by the logarithm of its octanol : water partition coefficient, log. P), the greater was its lymphatic absorption. Temarotene, Ro 15-0778, (log. P = 8.7) exhibited a maximum dose adjusted lymphatic absorption rate of 4100 ng/h compared with the less lipophilic Ro 04-3780 (log. P = 6.8) which showed a maximum rate of only 150 ng/h. The oily vehicle in which the retinoid was orally administered strongly influenced the rate of absorption via the lymphatic route. Ro 04-3780 demonstrated a 150 fold difference between the selected oils giving the maximum and the minimum lymphatic absorption rate. The best oils in this role appeared to be those in which the retinoid showed the least solubility. Mesenteric lymph flow rate was also shown to vary depending upon the oil. Basal lymph flow rate in the fasted rat, after dosing with saline, was 1.6 m1/h/kg. Cottonseed oil and soyabean oil promoted an increase in this flow rate to greater than 3.0 ml/h/kg (p < 0.01) when orally administered with Ro 04-3780. Conversely, linoleic acid suppressed the mean lymph flow rate to 0.8 ml/h/kg (p < 0.01) after oral administration with Ro 04-3780. Lymph turbidity was evaluated as an indication of chylomicron formation and transport in the lymph. Since the chylomicron is the particle in which dietary fats enter the lymphatic system, it was thought that lipophilic drugs, which are soluble in dietary lipid, may be carried into the lymphatic system via this pathway. The mixed long chain fatty tri-acyl glycerol oils, cotton seed oil, soyabean oil and peanut oil, when orally dosed to rats, produced the most turbid lymph (25 - 48 times greater than the turbidity produced after an oral dose of saline). These type of oil also promoted the highest lymphatic uptake rate for the retinoids. Other oils including, oleic acid, linoleic acid and MTS (a Miglyol S12 based self-emulsifying oil system), demonstrated much wider ranging extents of lymphatic absorption, but produced lymph with similar but lower turbidity (10 - 12 times greater than the turbidity produced after an oral dose of saline). A self-emulsifying oil system was developed for use in the oral administration of a retinoid. This system (MTS), produced a stable emulsion with a particle size of 500 nm after gentle mixing with an aqueous solvent and contained 80% Miglyol 812 and 20 % surfactants. MTS increased both the lymphatic and portal absorption rates for Ro 15-0778 by three fold compared with Miglyol S12 alone, improving the overall bioavailability but without selective promotion of lymphatic uptake. The effect of feeding, prior to orally dosing with an oil (linoleic acid) containing Ro 10-9359, was to suppress greatly the portal absorption rate of the retinoid from 310 ng/h to less than 25 ng/h. A number of factors, which were believed to be important in the lymphatic absorption of the retinoids, have been investigated here, using the rat as an animal model. The data obtained in this work suggest that lymphatic absorption is a very complex process and the factors which govern this absorptive pathway vary depending upon the nature of the drug being studied and the nature of the orally dosed vehicle.