This thesis details my contribution to the literature on military surgery, based on both front-line surgical experiences as well as research carried out on causes of death and disability, particularly in relation to limb injuries, the most common site of wounding in conflict. Injury analysis (6 papers). Injury prevention/mitigation (5 papers). Management (8 papers). Outcome (13 papers). Education (9 papers).

Colorectal cancer is the 4th most common cancer in the UK and the second commonest cause of cancer death. Whilst mortality rates from colorectal cancer haven fallen over the last 2 decades, around 40% of those diagnosed with colorectal cancer will die from their disease. Surgery currently remains the only chance of cure. Around 10% of patients present as an emergency with perforation, obstruction or bleeding. Outcomes from these emergency operations are substantially worse than from elective procedures. The presence of a systemic inflammatory response pre-operatively is now widely recognised as a predictor of disease progression and poor outcomes, both long and short term, regardless of tumour stage in those with colorectal cancer. Numerous scoring systems that measure various components of the systemic inflammatory response have been documented, the most commonly used are the modified Glasgow Prognostic Score (mGPS) and the Neutrophil-Lymphocyte Ratio (NLR). The NLR has the advantage of using 2 components of the differential white cell count, which is routinely measured in surgical and oncological practice, whereas CRP is less commonly routinely measured. However, studies utilising the NLR have used a variety of thresholds, making comparison of the results from study to study difficult. Whether one of the components of the NLR is more important than the other remains to be seen and indeed whether there is a more optimal score that utilises the white cell count is not clear. To date no work has examined similar scoring systems in the post-operative period. The present thesis aims to examine the impact of the innate immune response, through such systemic inflammation based scoring systems, on patients undergoing surgery for colorectal cancer. Furthermore, it analyses the nature of the inflammatory response in the post-operative period in order to ascertain whether similar scoring systems may be of clinical utility. Chapter 1 provides an overview of colorectal cancer, its presentation and treatment and its known determinants of outcomes. Furthermore, the immune response to injury and post-operative inflammatory response are discussed. Chapter 2 documents a survey of clinicians who have an interest in systemic inflammation. The survey asks the participants whether they routinely measure systemic inflammation, to what purpose and which scoring system they prefer. Unsurprisingly, the majority of participants use these scoring systems for research purposes only with an even split in terms of which scoring system they prefer to use. Their use in clinical practice remains small but their use in some oncological studies may signify a step towards their incorporation into clinical practice in the future. Chapter 3 presents data from a cohort of patients whom have undergone surgery for colorectal cancer with pre-operative differential white cell counts in order to determine whether any of the white cell count components are important in determining long term outcomes. Only the neutrophil count was independently associated with poor long term survival in patients undergoing surgery for colorectal cancer. These results highlight the importance of both the neutrophil count and the innate immune system in outcomes in patients with colorectal cancer. In chapter 4, a cohort of colorectal cancer patients and a cohort of patients with cancer were utilised in order to determine whether a pre-operative systemic inflammation based score using the neutrophil and platelet count was capable of predicting survival in these patients. This was based on the fact that recent in-vitro work had suggested that a critical checkpoint early in the inflammatory process involved the interaction between neutrophils and activated platelets. The subsequent score – the neutrophil platelet score (NPS)- was shown to be capable of predicting survival, independent of TNM stage, in patients with colorectal cancer and had prognostic value in patients with a variety of other tumours. Chapter 5 describes a systematic review of studies analysing the effect of various surgical procedures on markers of the systemic inflammatory response. Only CRP and IL-6 were found to represent the degree of surgical trauma and invasiveness of the procedure. This work provides a framework for analysing the post-operative SIR and how it is affected by surgery and peri-operative programmes such as ERAS that are reported to improve length of stay and sort term outcomes following surgery for colorectal cancer. It was of interest in the previous chapter that white cell count did not reflect the degree of surgical trauma. Whether individual white cell components act differently and represent the degree of surgical trauma was unclear. Chapter 6 sought to clarify this by analysing, in a cohort of patients undergoing surgery for colorectal cancer, the differential white cell count and whether it reflected the magnitude of injury and short term outcomes. Only the neutrophil count reflected the magnitude of trauma and development of infective complications. However, it remains inferior to other well established markers such as CRP. Whilst the pre-operative systemic inflammatory response is a well-recognised determinant of both long term outcomes and short term outcomes such as infective complications, little work has focussed on the post-operative systemic inflammatory response. In chapter 7, the possibility of the post-operative systemic inflammatory response also being capable of predicting both short and long term outcomes was explored in a cohort of patients whom had undergone surgery for colorectal cancer. A score using the combination of post-operative CRP and albumin was created and called the post-operative Glasgow Prognostic Score (poGPS). In this cohort of patients, this score predicted the development of infective complications and also long term survival. Given that these results would indicate that a reduction in the post-operative systemic inflammatory response would improve outcomes, the clinicopathological factors that may alter this post-operative systemic inflammatory response should be investigated as some of these may be modifiable and may therefore improve outcomes following surgery for colorectal cancer. ERAS programmes have changed perioperative management and are reported to be beneficial in reducing length of hospital stay and post-operative complications. It is purposed that this is due to the reduction on the surgical stress response. However it is unclear which of the components of an ERAS programme are responsible for this reduction in the systemic inflammatory response. Chapter 8 describes a systematic review analysing studies of the various ERAS components and whether there is objective evidence of a reduction in the SIR, evidenced by a reduction in either CRP or IL-6. Only laparoscopic surgery was reported to reduce the SIR in these studies, all the remaining components had either little or no evidence of a reduction in the SIR. Further work is required to ascertain whether any of the other components also reduce the SIR. This will hopefully allow streamlining of the ERAS process in order to improve outcomes. Specific clinicopathological factors that may alter the post-operative systemic inflammatory response are examined in chapter 9. Common clinicopathological factors were examined using the poGPS to ascertain which factors resulted in increased poGPS scores. In those patients undergoing elective surgery, year of operation, ASA grade, pre-operative systemic inflammation, and tumour site were associated with increased poGPS scores. These findings may have important clinical consequences as whilst factors such as ASA grade and BMI are not readily modifiable in the short time frame between diagnosis and surgery, pre-operative inflammation could potentially be targeted with anti-inflammatory medication. / However, more work is required to identify the specific agent and the timing of its delivery. In chapter 10, a cohort of patients undergoing surgery for colorectal cancer in whom there was prescription information available. Patients prescribed aspirin or statin were identified and their post-operative inflammatory response and short term outcomes were compared to those not prescribed aspirin or statins. In 446 patients, neither aspirin nor statin prescription was associated with a reduction in the post-operative systemic inflammatory response. Therefore, it would appear that these medications will not be useful in moderating the systemic inflammatory response following surgery. However, further work is required to identify which medications will be of benefit and should take the format of a randomised controlled trial. Chapter 11 provides a summary of the main findings of this thesis, discussed their implications and provides some discussion surrounding future work in this field.

This work was performed between 1979 and 1985 when there was great interest in the role of the recently identified prostacyclin in vascular function and disorders. The discovery of this substance with its powerful vasodilatory and platelet anti-aggregatory powers raised the hypothesis that prostacyclin might be involved in the pathogenesis of portal hypertension, its associated hyperdynamic circulation and typically catastrophic haemorrhage from bleeding oesophageal varices. Partial portal vein ligation in a rat model of portal hypertension was used because of its simplicity and absence of hepato-cellular dysfunction. This model was found to result in short lived hypertension with return to normal pressure by two weeks. Anatomical studies (using venography and corrosion casting) of the changes to the portal venous circulation after partial portal vein ligation, revealed the development of a dominant portosystemic collateral draining into the left renal vein via the left anterior lumbar vein. Ligation of this collateral at the same time as partial portal vein ligation gave a reliable model of permanent portal hypertension. Accurate measurement of prostacyclin proved to be difficult. Initially a bioassay of prostacyclin-like activity was used with success. The rat was found to produce high levels of prostacyclin well within the range of accurate measurement of this assay. Prostacyclin production was shown to increase directly with pressure increase in the portal vein. This direct relationship was confirmed in the acute model where prostacyclin production fell as the portal pressure returned to normal; in the model of chronic portal hypertension, prostacyclin production remained permanently elevated. A radioimmunoassay for 6 ketoPGFIα, the stable breakdown product of prostacyclin, was developed to allow measurement of prostacyclin in human tissue and serum samples (at this time there were no commercially available RIA kits). Initially the Wellcome antiserum was used with accurate measurement in incubated human tissue samples. These studies confirmed greater intrinsic prostacyclin activity in normal mesenteric and portal vein compared to peripheral venous tissues but failed to show any difference in tissue or plasma levels of portal hypertensive compared to normal patients. A second antiserum, the Cardeza antiserum, was then used in the radioimmunoassay. Unlike the Wellcome assay, this antiserum did not require a prostanoid extraction process. Comparison of the two assays in identical samples revealed major differences with large quantities of 6 ketoPGFIα being measured using the Wellcome antiserum with its extraction step compared to virtually none detected using the Cardeza antiserum. The extraction step was resulting in production of cross-reacting prostanoid substances giving falsely high readings. Both radioimmunoassays were abandoned at this stage because of the inaccuracy of the first, and the inability of the second to detect the low levels of 6 ketoPGFIα in human plasma. The human studies were continued using a highly specific and sensitive assay of 6 ketoPGFIα, namely gas chromatography/negative ion chemical ionisation mass spectroscopy (GC/NICIMS). At the time of this work, this methodology was complex, cumbersome, with limited access and the numbers studied were small. Very low levels of 6 ketoPGFIα were found in peripheral blood of normal and portal hypertensive patients who were not bleeding from oesophageal varices, in patients without portal hypertension, portal blood levels of prostacyclin were higher compared to peripheral levels, confirming the finding in the rat and pig that prostacyclin activity is higher in the normal portal circulation compared to peripheral vein. Significantly elevated prostacyclin production was found in both the peripheral and portal blood of portal hypertensive patients who were actively bleeding from oesophageal varices. Portal prostacyclin production was found to be significantly higher in patients with portal hypertension who were actively bleeding compared to normal patients undergoing laparotomy for other conditions. These findings in bleeding patients support the hypothesis that prostacyclin activity is increased in portal hypertension and may play a role in the severity of haemorrhage. In both the animal and human studies a clear effect of surgical intervention on increased prostacyclin production was found. These studies demonstrated for the first time increased prostacyclin production in both developing, and established portal hypertension in both the experimental animal situation and in man. High levels of prostacyclin production were found in the portal circulation of portal hypertensive patients undergoing surgery for uncontrolled variceal bleeding.

Major vascular surgery is associated with a substantial risk of cardiovascular events and death. This risk is of increased importance in prophylactic elective open Abdominal Aortic Aneurysm (AAA) repair, where a balance of risk of rupture and postoperative outcome is assessed prior to management decisions. Further, the UK Small Aneurysm Trial has shown that prophylactic repair of an AAA has no survival benefit for the first three years due to the major adverse cardiac event (MACE) rate of 5-15%. There is however no ideal method of predicting this risk. Cardiac Troponin I (cTnI) is a contractile protein that is a highly sensitive and specific marker of myocardial necrosis. A few case reports have commented on the finding of preoperative asymptomatic elevated cTnI levels and poor outcome in a small number of patients undergoing major vascular surgery. There are however no studies looking at its incidence in the vascular surgical population or its utility as a preoperative marker. Several studies have noted that B-type natriuretic peptide (BNP), a diagnostic and prognostic marker of heart failure, may have a role in predicting MACE in settings including major vascular surgery. There are no studies that have investigated this role in AAA repair alone. The aim of this thesis is to investigate the incidence of, and to determine a possible role for, preoperative elevated cTnI in major vascular surgery. The further aim is to determine if a single preoperative BNP level correlated with MACE and all-cause mortality in elective open AAA repair in both the short and long-term. Comparisons to current accepted risk indices in AAA, and a possible role for BNP in EndoVascular Aneurysm Repair (EVAR) will also be investigated. Patients were recruited in two cohorts: Firstly, a prospective, 2 year observational single centre cohort study of all patients undergoing a vascular procedure, with an expected cardiac event rate >5%, recruited patients who had no clinical or ECG evidence of myocardial ischaemia. Preoperative cTnI was performed in all and postoperative screening (clinical assessment, ECG and cTnI) for cardiac events was performed at days 2, 5 and 30. 213 patient were recruited, of whom 11 (5.2%) had an asymptomatic elevated preoperative cTnI (>0.02 ng/ml). Eight of these patients proceeded directly to theatre, and 2 were delayed but later underwent surgery with a persistently elevated cTnI. Of these 10 patients, 5 (50%) died and 4 (40%) suffered MACE. The remaining patient was delayed due to the poor outcome of the preceding patients, and later underwent an uncomplicated aortic bifurcation graft with a normal cTnI level which had been preceded by coronary intervention. Secondly, a prospective, 2 year observational multi-centre cohort study in the 3 largest vascular units in Glasgow (Gartnavel General Hospital, Glasgow Royal Infirmary and Southern General Hospital) was performed between August 2005 and August 2007, recruiting all patients who were admitted for both elective open AAA repair and EVAR. Preoperative BNP levels were performed and batch analysed at the end of the study. Postoperative screening for cardiac events was performed as described above. Data was collected to allow calculation of risk indices associated with outcome in AAA repair (Glasgow Aneurysm Score [GAS], Vascular physiology only Physiological and Operative Severity Score for enUmeration of Mortality [V{p}-POSSUM], Vascular Biochemical and Haematological Outcome Model [VBHOM], Revised Cardiac Risk Index [RCRI] and Preoperative Risk Score of the Estimation of Physiological Ability and Surgical Stress Score [PRS of E-PASS]). Follow-up was continued to a minimum of 3 years, where possible, with cause of death recorded. 106 of 111 patients were recruited. The median [interquartile range] BNP concentrations in the 16 patients (15%) who suffered immediate postoperative MACE was 206 [118-454] vs 35 [17-61] pg/ml in the remainder (p=0.001). ROC analysis indicated a BNP concentration of 99.5 pg/ml best predicted MACE (area under the curve 0.927), with sensitivity of 88% and specificity of 89%. The BNP in patients who suffered cardiac death was significantly higher than in those that did not (median BNP 496 [280-881] vs 38 [18-84] pg/ml, p=0.043). ROC analysis revealed a cut-off of 448 pg/ml (AUC 0.963), with sensitivity 80%, specificity 100%, positive predictive value 100% and negative predictive value 99%. Not only did higher values of BNP predict MACE, but it was also found to predict all-cause mortality in the immediate (median BNP 100 [84-521] vs 35 [17-81], p=0.028), intermediate (median BNP 201 [97-496] vs 35 [17-73], p<0.001) and long-term (median BNP 98.5 [58-285] vs 32 [17-71.5], p<0.001) postoperative periods. ROC analysis revealed decreasing BNP levels to predict outcome over time, with a BNP of >60.5 pg/ml (AUC 0.761) found to best predict death at 3 years. Whilst BNP was found to predict outcome, most risk indices performed poorly. The GAS, VBHOM and RCRI performed poorly and did not predict any outcome measure. V(p)-POSSUM was, however, found to predict all outcome measures (p=0.028, p=0.030, p=0.038 for MACE, cardiac death and all-cause mortality respectively). The PRS component of E-PASS was found to predict MACE (p=0.019) and cardiac death (p=0.017). BNP performed better than any risk index. During the study period only 40 of 42 patients admitted for elective EVAR were recruited. Of these 40, only 3 suffered a non-fatal MI and 1 died of respiratory failure. BNP was not found to predict MACE or death in this cohort, and due to the small number of patients, and events, no strong conclusions could be drawn. Whilst preoperative elevated cTnI was found to identify patients that were at an increased risk of both postoperative MACE and death following their major vascular surgical procedure, its use in elective open AAA repair is limited due to infrequent occurrence. Preoperative serum BNP concentration, however, predicted postoperative MACE, cardiac death and all-cause mortality in patients undergoing elective open AAA repair on immediate, intermediate and long term follow-up. Further, BNP performed better than any current risk index for elective open AAA surgery. This simple blood test, therefore, offers valuable information regarding risk stratification of prospective surgical patients and should be considered a part of routine preoperative assessment in this prophylactic procedure.

There is potential for haemorrhaging injuries that become contaminated with toxic chemicals e.g. sulphur mustard (SM). There are no specific medical countermeasures for such injuries at present. It is proposed that haemostats could simultaneously stop bleeding and decontaminate wounds. Products must be able to clot SM-contaminated blood and reduce SM percutaneous absorption to be considered suitable. Sulphur mustard did not significantly affect coagulation in vitro or in vivo. Overall SM did not affect the pro-coagulatory function of haemostats. However, clot strength in WoundStat™ treated blood was adversely affected by SM in vitro. Superficial damage to the skin significantly increased the percutaneous absorption of SM with good agreement between in vitro and in vivo studies. Moreover, the altered toxicodynamics imply that the latent period associated with SM pathology is due, in part, to rate of absorption and not solely caused by biochemical pathways. Application of WoundStat™ significantly reduced the amount and rate of SM penetration through the skin in vitro. Furthermore, WoundStat™ reduced SM percutaneous absorption and pathology in vivo. Microarray analysis has identified that SM exposure via damaged skin causes changes in gene expression which may warrant further investigation as potential therapeutic targets.

Title: A study of some of the psychological issues affecting women undergoing treatment for urinary incontinence. Background: There is evidence suggesting that effective clinical treatment of urinary incontinence (UI) is not reflected in patients’ quality of life reports or in psychological well-being. There may be other psychological issues that are not routinely captured but may be influential in recovery. Aims and objectives: To explore women’s experiences of UI and identify what aspects are not altered by stress incontinence (SUI) surgery, to identify themes with which to develop a treatment-planning tool and to identify measures to improve patient satisfaction. Methods: Qualitative interviews and repertory grids were performed on women with UI. Themes generated from these women, together with themes from a literature review and a focus group, form the basis for future questionnaire development. Data were scrutinised for additional potential to improve satisfaction. Results: Wide-ranging themes were identified. Psychosocial aspects are not fully addressed by traditional UI treatment. Many suggestions for improvements in management were identified. Conclusions: There is much that can be done to try to improve satisfaction for women with UI. Many themes have been identified as a basis for the further development of a treatment-planning tool.

Background: Osteoarthritis affects approximately 40% of older adults but molecular mediation of OA pain in the dorsal horn is unexplored clinically. This study explored amino acids and cytokines related to pain signalling and sensitisation to determine whether significant differences existed in their concentration in comparison with pain-free controls after adjustment for age, gender and psychological distress. Method: After ethical approval people having primary hip or knee arthroplasty (OA group) or urological surgery (pain-free controls) were recruited. Pain at rest, (PAR), pain on movement (POM) (0-10 numerical rating scale), and HADS data was collected before aspiration of 2ml sample of CSF. HPLC and multiplex bead array assay was conducted and data explored using ANCOVA and logistic regression. Results: Data from 21 control (75% male) and 59 OA (46% male) participants revealed that HADS, serine, leucine, valine, and TNFα were significantly higher and IL-12 was significantly lower in the OA group. IFNγ was significantly lower in the PAR group. Discussion: This study suggests central sensitisation is involved in OA. Psychological distress is an integral part of the OA experience. Amino acid and cytokine involvement in pain transmission is complex; further work exploring human CSF in painful conditions with clinical follow up is recommended.

Maintenance of anaesthesia using the intravenous agent propofol has increased following development of pharmacokinetic models. An analyser capable of determining propofol concentrations at the point of care may lead to an improved accuracy of drug delivery. Validation work on a novel analyser measuring propofol concentration in near real time demonstrate a high level of precision for samples in the clinical range. Further work in the clinical setting was carried out using the novel propofol analyser to further research its potential use in a diverse patient cohort. Studies were performed in intensive care correlating blood propofol concentrations with depth of sedation, demonstrating a correlation with organ failure. The Marsh model of Target Controlled Anaesthesia was poorer at predicting propofol concentration in patients with significant organ dysfunction than in those without organ failure (correlation coefficient 0.36 vs. 0.73 respectively). Studies in the operating room were performed in which measured propofol concentrations were compared with those predicted using the Marsh model. Results demonstrated significant inaccuracies of the model (bias 32%, precision -8.7 to 72.6%). A method of Marsh model bias correction using a single blood propofol measurement was tested. Results demonstrated insufficient predictability to allow a single point calibration.

Bacterial biofilm infections cause significant morbidity in orthopaedic joint replacement. One of the most common bacteria in orthopaedic prosthetic infections is Staphylococcus aureus. Infection causes implant failure due to bacterial adherence and subsequent biofilm production. Nanotopography refers to the topography of a surface at the nanometre level and has major effects on cell behaviour. Studies suggest that surface nanotopography impacts the differential ability of staphylococci species to adhere, and may reduce orthopaedic implant infection rate. This research thesis focuses on bacterial adhesion on nanofabricated materials, and investigates the related metabolic changes and possible interventions. Staphylococcus aureus growth and quantification methods were optimised, with regard to growth media, incubation time and lysozyme incubation time. Both polystyrene and titanium (Ti) nanosurfaces were studied. Adhesion analysis was performed using fluorescence imaging, quantitative PCR, and bacterial percentage coverage. Metabolomic analysis was conducted by substitution with ‘heavy’ labelled glucose into growth medium, thus allowing for bacterial metabolomic analysis and identification of up-regulated, labelled metabolites and pathways. Bacterial growth was optimal using DMEM + supplement media, with adhesion occurring after 1hr bacterial incubation. Optimal lysozyme incubation for bacterial quantification using qPCR was 2hr. These parameters were used for all subsequent experimentation. Surface topography affects cell behaviour, bacterial adhesion and long term implant survival can be affected. This study found reduced bacterial adhesion on the SQ and HEX polystyrene patterns. While not found to be significant, this trend was supported by a lower average percentage bacterial coverage on both the SQ and HEX patterns (P=0.05 and P=0.01, respectively). It may be that the SQ and HEX nanopatterns are the optimal nanopit orientation required to prevent bacteria microcolony formation, keeping the bacteria in small, isolated clusters. In addition, this series of investigations showed an increase in bacterial concentrations on both the 2.5Hr and 3Hr treated Ti nanowire discs when compared to the polished Ti control disc, suggesting nanoroughness increases are associated with elevated bacterial adhesion. This theory was further supported by average percentage coverage, being significantly higher on the 2.5Hr and 3Hr treated discs. If, however, a disordered NC Ti nanopattern, hexagonal in nature, is used bacterial adhesion is significantly reduced when compared to a polished, control surface. The bacterial percentage coverage was also noted to be significantly lower on the NC surfaces, with over a 10-fold reduction when compared to the control surface. It is postulated that this reduction is through similar mechanisms to those described by Ivanova et al, and primarily related to altered surface interactions. Metabolomic analysis demonstrated increased intensity counts for key metabolites (pyruvate, aspartate, alanine and carbamoyl aspartate) involved in bacterial aggregation, proteoglycan and DNA synthesis. These pathways are also known to be important in bacterial biofilm production. Therapeutic targeting of these pathways was found to result in significantly reduced bacterial adhesion. This study shows that by altering nanotopography bacterial adhesion, and therefore, biofilm formation can be affected. Specific nanopatterned surfaces may reduce implant infection associated morbidity and mortality. The identification of metabolic pathways involved in adhesion allows for a targeted approach to biofilm eradication in S. aureus. This is of significant benefit to the patient, the surgeon and the NHS, and may well extend far beyond the realms of orthopaedics.

No description available.