The incidence of oesophageal adenocarcinoma (OA) has increased exponentially in the western world over the past few decades. Barrett's oesophagus (BO) is a well known precursor of OA with a risk approximately 20 times more than that of background population. Regular endoscopic surveillance in patients with BO is recommended by most of the national gastroenterological societies. The advantage of Barrett's surveillance is to identify early subtle lesions which could then be managed early to avoid symptomatic and advanced cancers. The detection of such early lesions are challenging as they could be flat and inconspicuous on routine endoscopic examination. In the absence of any lesions, four quadrant biopsies every 1-2 cm of the whole length of Barrett's oesophagus is advised. This technique would map only 5-10% of the surface area of Barrett's segment and hence it is associated with significant sampling error. The improvement in electronics over the past decade has led to the production of endoscopes with better charged coupled devices and image enhancement techniques by altering the spectrum of light. This thesis examines the role of multi modal imaging in Barrett's oesophagus with a focus on detecting dysplasia and early cancer (EC). Firstly, the role of high definition (HD) imaging in routine clinical setting was studied using data from patients who have undergone Barrett's· surveillance. The yield of dysplasia by HD endoscopy was compared to standard definition (SD) endoscopy in this study. The role of narrow band imaging (NBI) with magnification in characterising abnormal lesions detected during BO surveillance was evaluated by performing a meta- analysis of clinical studies. The role of autofluorescence imaging (AFI) in Barrett's oesophagus was examined in detail with a view to understand the biological basis of autofluorescence and to improve the specificity of this technique as it is associated with significant false positive results in clinical studies. A meta-analysis was performed to identify whether AFI has a clinical advantage over white light endoscopy in detecting Barrett's dysplasia and the inter-observer reliability of this technology was studied using AFI expert and AFI non-expert endoscopists. An objective method of measuring the autofluorescence intensity was proposed as a ratio of the red to the green colour tone (AF ratio) of the area of interest. When the AF ratio of the lesion was divided by the AF ratio of the background mucosa, an AF index is obtained. A pilot study was performed to identify a cut-off value of AF index to differentiate high grade dysplasia (HGD) and EC from non-dysplastic BO. Finally, the biological basis of AF intensity was examined using APCmin mouse colonic models. This study looked into the AF ratio of the colonic mucosal lesions and correlated it with the amount of collagen and elastin in the submucosal tissue. Collagen and elastin are known to be the strongest fluorophores of the gastrointestinal tract and the question addressed is whether the low AF intensity associated with dysplastic lesions is due to the thickening of mucosa or to a reduction of collagen and elastin.

Introduction. The anti-idiotypic monoclonal antibody 10SAD7 mimics the tumour associated antigen 791T/ gp72, present on approximately 80% of colorectal cancer cells. A Phase I study using 10SAD7 in 13 patients with advanced colorectal cancer has shown that it is nontoxic, and conferred a survival advantage on patients who received it [Denton GWL 1994]. Aim. There were two aims of this work. The first was to assess whether. the survival advantage seen in the Phase I study was reproducible in a Phase II study. The second was to immunise patients with primary colorectal cancer, in a non-randomised adjuvant study, and explore further the immune responses generated. Materials and Methods. Patients with advanced colorectal cancer were recruited to a randomised, double-blind, placebo controlled survival study. The first patient was recruited to this Phase II study in April 1994, and the last in October 1996. Four trial centres were used- Nottingham, Hull, Leeds, and Newcastle. Eligible patients had a life expectancy of 3 months, and none had received radiotherapy or chemotherapy in the preceding 1 and 3 months, respectively. Patients attended on 3 occasions, 6 weeks apart, receiving 10µg of 10SAD7/alum i.d. followed by 100µg i.m. Venous blood was assayed for blood count and differential, liver function, urea and electrolytes, and CEA. Chest X-rays and CT scans were performed at trial entry and week 12 where possible. Dates of death were recorded following consultation with General Practitioner or referring clinician. In addition, patients with primary colorectal cancer were recruited to a non-randomised adjuvant study, whereby they received 10SAD7 before surgery. Venous blood samples were taken between immunisation and operation, and assayed for lymphocyte subsets. Samples taken from resection specimens were analysed immunohistochemically. Fresh tumours were in addition disaggregated, and separated TIL labelled with a panel of monoclonal antibodies, and analysed by flow cytometry. Control tumours were similarly labelled. All analysis was performed blind. Results. 162 patients were randomised to the Phase II study, between April 1994 and October 1996. 85 received 105AD7 and 77 placebo. The mean ages and sex-ratios of the two groups were comparable, as was the time from diagnosis of advanced disease to trial entry (172v179 days). Median survival from date on study was 124 and 184 days, in 105AD7 and placebo arms, respectively (p=O.38). Survival from date of diagnosis of advanced disease was 456 and 486 days (p=O.82). Chemotherapy and radiotherapy all prolonged survival in a multivariate analysis. Only one serious adverse event was seen in the 105AD7 arm, and this was felt unlikely to be attributable to the vaccine. Twenty-four patients were recruited to the adjuvant study. Immunohistochemical analysis of tumour sections from 16 patients showed increased infiltration of CD4 and CD8 expressing lymphocytes, relative to a well matched control group (p<O.05). Infiltration of CD4, CD8 and CD56 expressing lymphocytes combined was significantly higher, as was that of the mitochondrial antigen 7A6, expressed on cells undergoing apoptosis (p<O.005). The activation marker CD25 was also significantly increased (p<O.05). Flow cytometric analysis of disaggregated tumours from 16 trial and 22 control patients, confirmed the increased expression of CD25 on TIL in the 105AD7 group (p<O.01). Peripheral blood phenotyping failed to show any significant increase in any lymphocyte subset, following immunisation. A separate analysis was performed comparing 2 year survival and recurrence in 23 patients immunised by the previous CRC Fellow, with 97 matched controls from the Trent Audit. No significant difference was seen between the two groups. Discussion. No survival difference was seen between patients receiving l05AD7 and placebo, in the Phase II study. This suggests that any immune responses generated by l05AD7 are insufficient to have a significant effect on tumour growth, in patients with advanced disease. Work has therefore focused on immunising patients with primary colorectal cancer. Patients receiving l05AD7 prior to resection of their primary tumours, showed an increased number of activated lymphocytes, and apoptosis, at the tumour site, relative to a well-matched control group. The numbers in the survival analysis based on patients recruited by the previous CRC fellow, are insufficient to show whether any of these immunological changes confer a survival advantage. This question can only be answered in a large, prospective, placebo-controlled study in patients with primary colorectal cancer.

The incidence of the idiopathic inflammatory bowel diseases ulcerative colitis and Crohn’s disease appears to have risen markedly during the 20th century. These diseases now account for a considerable proportion of the workload of gastroenterologists in the developed world, and may affect as much as 1% of the population at some point in their lives. The aetiology of these diseases has been subject of much research over a number of decades and it is clear that both genetic and environmental factors are involved. The certain knowledge of environmental risk factors however remains scant. Similarly although inflammatory bowel diseases cause considerable morbidity and a small amount of mortality for their sufferers directly there is little agreement as to their overall impact once indirect effects are accounted for. This thesis contains studies contributing to the knowledge of both these areas using routinely or semi-routinely collected data. It examines two hypotheses relating to the aetiology of IBD (that risk is related to the season of birth, and that it is related to antibiotic use), and two areas of the impact of the diseases (overall mortality and fracture risk). With regard to aetiology the studies described show no variation in the risk of IBD with season of birth. They do show an increase in risk associated with the use of antibiotics, but since this is not specific (it is seen to occur with other groups of drugs also) it is far from clear that the association is causal. With regard to the indirect impact of the diseases a significant excess in overall mortality is demonstrated which is greater in Crohn’s disease than in ulcerative colitis, and is greatest in relative terms in the young but in absolute terms in the elderly. An excess is also shown for hip fractures in those with inflammatory bowel diseases, which is only partially explained by the use of corticosteroids.

Background: Irritable bowel syndrome is a medically unexplained phenomena relating to the lower gastrointestinal tract with symptoms such as altered bowel habit and abdominal pain. Patients experience poor quality of life and consume significant healthcare resources. Mechanisms for the delivery of evidence based psychological interventions for irritable bowel syndrome within the National Health Service are lacking and the feasibility of these interventions is poorly understood. Methods: A novel, low-intensity, nurse-led psychological intervention has been developed and trialled within a mixed methods feasibility study. Twenty participants were randomly allocated across four treatment conditions consisting of; a treatment as usual control (n = 5), self-help (n = 5), low-intensity (n = 5) and high-intensity (n = 5) cognitive behavioural therapy interventions. A total of ten participants took part in post-intervention interviews analysed using a group thematic analysis. Results: Recruitment to this feasibility study was a significant challenge with 22 participants recruited of which, 20 were randomised to the feasibility interventions. Of the 104 patients approached within secondary care gastrointestinal clinics, 27.7% of patients volunteered to enrol into the study. Reasons provided relate to difficulties with committing to taking part and personal circumstances. Themes derived from post-intervention interviews suggest participant’s valued face-to-face therapist interaction and described their perceived treatment utility along with a variety of barriers and facilitators to engagement in CBT interventions. Conclusion: Low-intensity and self-help cognitive behavioural therapy may be feasible mechanisms for the delivery of evidence based psychological interventions for patients with IBS, although significant concerns regarding recruitment of participants to future trials will need to be addressed. Further development of these lower-intensity interventions in collaboration with service users is required in order to improve the acceptability and relevance of the interventions.

Faecal incontinence is an embarrassing and socially debilitating condition which is primarily acquired and increases in prevalence with age. Current conservative measures are aimed at dietary modification and changing the consistency of stool with no targeted treatments available to address the underlying cause of incontinence. Surgical treatments are either unsatisfactory or carry significant morbidity. There is currently increasing interest in the use of α adrenoceptor agonists to increase the tone of the anus and thereby improve continence. One of the potential drugs, L-erythro-methoxamine, has been shown to increase mean anal resting tone in healthy volunteers and is well tolerated as suppositories. Extensive data exists on the neuromyogenic properties of the human internal anal sphincter (IAS) and its response to various drugs, particularly α adrenoceptor agonists. Little data exists on the response of the rectum to α agonists. The ideal drug treatment for incontinence would cause a contraction in the IAS and a relaxation in the rectum – increasing the reservoir of stool while augmenting the sphincter to aid continence. We performed in vitro experiments on sheep internal anal sphincter (IAS) using an organ bath method and subjected the tissues to electrical field stimulation to mimic nerve stimulation. Our results were comparable to results of previous authors who also examined the sheep IAS. Using this validated protocol, we investigated the sheep rectum to identify the neuronal mediators of the EFS response and to investigate the effect of α1 adrenoceptor agonists. Our results showed that sheep rectum relaxes in response to nerve stimulation and this relaxation is the result of the release of nitric oxide. Contraction in response to nerve stimulation is primarily mediated by acetylcholine acting on muscarinic receptors. Methoxamine caused a contraction in the sheep rectum. We also examined the pig IAS and rectum in vitro. An identical organ bath technique was used with Electrical Field Stimulation to mimic nerve stimulation. For both IAS and rectum in the pig, nerve stimulation caused a relaxation via nitric oxide and a contraction mediated primarily by noradrenaline acting on α1 adrenoceptors with a small component mediated by acetylcholine acting via muscarinic receptors. Methoxamine caused a contraction in both IAS and rectal tissue with similar potency in each. We were part of an industry-sponsored multi-centre randomised placebo-controlled clinical trial investigating the safety and efficacy of L-erythro-methoxamine on patients with faecal incontinence. The nine patients recruited from our centre showed no significant improvement in the number of episodes of incontinence or in the questionnaire scores measuring the impact of incontinence on quality of life after eight weeks of daily suppositories. The drug was well tolerated with few adverse events. There were no significant safety concerns although there was a prolongation of the PR interval in post-treatment ECGs and a positive correlation between the QT intervals on ECGs with serum concentrations of the drug. Our results were typical of those obtained in the other centres. The overall result of the trial was that there was no improvement in episodes of faecal incontinence following treatment with L-erythro-methoxamine at the chosen doses. The results from our in vitro experiments suggest that α adrenoceptor agonists may not be the best methods of treating faecal incontinence. The results from the clinical study support this finding. We believe that more in vitro studies need to be performed on human rectal tissue to confirm our findings that α adrenoceptor agonists cause a contraction in the rectum.

Introduction The response to neoadjuvant chemotherapy in oesophago-gastric (OG) cancer is only 40%, so over half of the patient’s disease will progress, whilst they also suffer the toxic chemotherapy side-effects. A model to predict chemotherapy response would provide a marked clinical benefit, by enabling personalised treatment of OG cancer. Methods Live chemo-naïve tumour biopsies were obtained following informed consent at staging endoscopy, before patients underwent their routine neoadjuvant chemotherapy. Tumour cells from the endoscopic biopsies were expanded, using an in vitro feeder layer system and supplemented medium. With ethics committee approval and under Home Office guidance, these individual patient cancer cells were engrafted into immuno-compromised mice, where they formed representative tumour xenografts. Primary patient tissue, the corresponding individual patient cancer cells and their matching xenografts were analysed using immunohistochemistry, demonstrating that the in vitro and in vivo cells had retained the characteristics of the original patient’s oesophageal adenocarcinoma. To model the human tumour micro-environment (TME), a three dimensional tumour growth assay (3D-TGA) was developed, whereby the individual patient’s primary tumour cells were grown as 3D cancer cell clusters. This was performed by seeding individual patient’s primary tumour cells within a biological basement membrane extract, rich in extracellular matrix (ECM) components, with and without human mesenchymal stem cells to provide stromal support. The individual patient cancer clusters in the 3D-TGA were subjected to detailed chemotherapeutic assessment, to quantify their chemo-sensitivity to the standard chemotherapy which was administered to the patient in the clinic. This 3D-TGA predicted chemo-sensitivity was then compared with the patient’s actual clinical chemotherapy response, as measured by the histological tumour regression grade, which directly relates to prognosis. In combination with standard platinum-based chemotherapy, the 3D-TGA was assessed as a platform for evaluating new chemotherapeutics: the novel emerging HDAC inhibitor Panobinostat, and the phosphodiesterase type 5 inhibitor Vardenafil, which has recently been shown to be active against cancer stroma, were evaluated. Results Individual patient tumours were grown from primary endoscopic biopsy tissue in over half of samples obtained within a clinically applicable timescale of 2-4 weeks. Incorporating human mesenchymal cells into the 3D-TGA significantly changed the growth and drug resistance profiles (p < 0.005). This 3D-TGA chemo-response in the presence of stroma reflected the clinical chemo-sensitivity, with an accurate correlation between the 3D-TGA predicted chemo-resistance and actual clinical response for the patients evaluated. As well as predicting potential chemo-sensitivity for individual patients, the method allows individual drugs and combinations to be evaluated, trends in chemo-sensitivity between patients to be appraised, and analysis of the effect of the TME on tumour growth and chemotherapy resistance. Combination with Panobinostat enhanced response and proved efficacious in otherwise chemo-resistant tumours. Addition of PDE5i demonstrated an overall significantly enhanced chemotherapeutic response (p=0.003), and consequently provided efficacy in 60% of the otherwise chemo-resistant tumours. Discussion The novel method of growing individual patient OG cancers, using a 3D model with specific components of the tumour micro-environment in particular ECM and mesenchymal cells, provides a clinically-relevant oesophageal cancer model with application for chemo-sensitivity testing. Mesenchymal cells have a significant effect enhancing chemotherapy drug resistance in OG cancer, and this 3D model allowed identification of patients in which stromal targeting using PDE5i provided a significant reduction in chemotherapy drug resistance. In these patients, addition of PDE5i to routine chemotherapy could result in a marked change in the clinical efficacy of their chemotherapy regimen. The 3D model’s chemo-response accurately reflects individual patients’ clinical chemo-sensitivity and so this research has direct clinical application: if this assay proves to be predictive across a wider patient population, then following clinical trials, it could potentially be used to routinely guide individual patient therapy in the clinic, with administration of tailored chemotherapy for individual patient benefit.

Background: Colorectal cancer (CRC) is the third most prevalent and second most fatal type of cancer in the Western world. 20q amplification has been identified in approximately 50% of CRC cases. 20q amplification and may be associated with liver metastasis. In our lab, a previous array comparative genomic hybridization (aCGH) study of a series of primary CRCs and corresponding liver metastases, found that the 20q13 was frequently amplified in both groups. Several known oncogenes are located at Chromosome 20q (for example SRC, TPX2 and CSE1L) suggesting that 20q amplification may be an important event in the development and metastasis of CRC. Hypothesis and Aim: Although there are many genes located in 20q identified to have role in developing CRC, other genes with potential role in tumour progression have not been studied in CRC. We hypothesized that 20q harbouring genes have a potential role in CRC progression. Therefore, we aimed to (i) study the frequency of 20q amplification in CRC by validating our aCGH data by alternative method and screen larger sample set for 20q amplification. (ii) Protein expression of candidate genes was investigated. Then, candidate genes were evaluated in several cancer related processes including proliferation and cell motility. Materials and Methods: The evaluation of 20q amplification was performed in DNA from FFPE of micro-dissected 20 CRCs and their matched liver metastasis, using comparative quantitative PCR (qPCR) based on 6 genes (PTPN1, CSE1L, ADNP, PREX1, ELMO2 and PTGIS), which are located in the same region. It was found that quantification of two genes was sufficient to evaluate amplification and thus separate series of 103 cases were screened for 20q amplification based on qPCR of two genes (ELMO2 and PTPN1). For studying the functional activity, PTPN1, CD40, PREX1, ELMO2 and PTGIS were studied in CRC cell lines. To evaluate gene function, small interference RNAs (siRNAs) were used to knockdown genes of interest in CRC cell lines. Knockdown was validated by western blot and qPCR and the effect of knockdown was evaluated on cellular functions such as proliferation, cell migration, cell invasion and wound healing. Immunohistochemistry was used to investigate protein expression in primary CRCs. Results: The amplification of 20q in CRC sample was studied by qPCR and aCGH. High concordance was found between the two methods aCGH, and high correlation coefficient between the qPCR results of the genes demonstrated the utility of this method for measuring amplification. In the combined data set, 20q amplification was seen in both primary tumours (51%) and metastases (60%). TP53 mutation was associated with 20q amplification (p =0.010). Then, we investigated the functional activity of 5 genes located on 20q, which might be has a role in CRC progression. (1) PTPN1 has been reported as tumour suppresser and oncogene. Therefore, we studied protein expression and its functional role in CRC cell lines. PTPN1 was overexpressed in 59% of primary tumours. Knockdown of PTPN1 reduced cell motility but did not affect cell proliferation. (2) CD40, previously, has been identified as a tumour suppresser in CRC. However, high CD40 expression was detected 25% of tumours. Functional assays showed CD40 promotes proliferation by altering subG1 phase of the cell cycle. Therefore, CD40 might promote tumour growth by inhibiting apoptosis. (3) PREX1 was expressed in 61% of primary tumours. Functional analysis showed that it promotes both cell proliferation and motility. (4) ELMO2 was expressed in 38% of primary tumours and functional analysis showed it promotes proliferation and cell migrations. (5) Finally, PTGIS is highly expressed in CRC and its expression associated with liver metastasis. However, PTGIS effect on tumour progression has not been studied. PTGIS knockdown had no effect on CRC cell lines growth and motility. Conclusion: Using alternative methods to evaluate the 20q13 amplification, we have confirmed our previous aCGH and found a frequency similar to that reported in the literature. Amplification of 20q appears frequently in primary tumours but it not positively associated with metastasis. Many genes located in 20q have oncogenic effects which would support the hallmarks of cancer. Understanding their role in CRC development could reveal new avenue in understanding cancer biology and treatment.

Colorectal cancer (CRCa) is the 4th most common cancer in the United Kingdom with 41,265 new cases diagnosed in 2014 (Cancer Research UK, 2017). Advancing age is an established risk factor for the development of CRCa (Figure 1.1); between 2012 and 2014 44% of new cancers were diagnosed in patients aged 75 years and over (Cancer Research UK, 2017). Due to our ageing population, national screening programmes and improved diagnostic techniques, a greater number of older people are now being diagnosed with CRCa. However older people are not surviving the disease as well as their younger counterparts. The 5 year survival rate from diagnosis of bowel cancer between 2009 and 2013 was 67.5% for 60-69 year old men and 45.5% for 80-99 year old men. This illustrates an increased burden of this disease in an ageing population (Cancer Research UK, 2017). In general terms, due to improvements in health screening and perioperative care, mortality from CRCa in all populations is decreasing over time. Indeed, the England and Wales age standardised 5 year survival rates for men with CRCa have increased from 24% in the early 1970s to 59% in 2010-11 (Cancer Research UK, 2017). Overall CRCa incidence is increasing over time but this is mirrored by improved survival rates. Age is a significant factor in reduced survival from colorectal cancer and efforts to improve outcomes in elderly CRCa patients should be investigated.

Introduction: Irritable bowel syndrome (IBS) has substantial impact on Quality of Life (QoL) and patients have high healthcare utilization. Guidelines recommend diagnosis and management within primary care, yet around 25% of patients are referred to gastroenterology. These studies aimed to assess the incidence of organic gastrointestinal disease in patients diagnosed with IBS, the cost of healthcare utilization and the QoL in patients with IBS before and after seeing a gastroenterologist and to estimate the cost-effectiveness of a gastroenterology appointment. Methods: Patients with IBS were identified within the UK Clinical Practice Research Dataset. Incidence rates of coeliac disease, colorectal cancer (CRC) and inflammatory bowel disease (IBD) were calculated. Individual-level healthcare utilization data were extracted for IBS patients who first visited a gastroenterologist in 2008 or 2009. Mean costs of total healthcare utilization were calculated before and after gastroenterology attendance. A questionnaire study of patients with IBS attending a gastroenterology outpatient clinic for the first time measured QoL and utility before and after the appointment. Quality Adjusted Life Years (QALYs) were modeled from these utility values. Cost-effectiveness of a referral to gastroenterology in IBS was assessed using mean cost per QALY. Results: Fifteen years after IBS diagnosis, the combined cumulative excess incidence of coeliac disease, IBD and CRC in IBS is 3.7%. Over one year following gastroenterology appointment, the expected QALY gain compared to no appointment was 0.03 and the expected extra total healthcare costs were £657. The incremental cost-effectiveness ratio was £27865.64/QALY. Referral for patients younger than 30, men, and increasing the time horizon, reduces the expected cost effectiveness. Conclusions: My findings provide reassurance that non-specialists are unlikely to be missing an organic condition in the majority of IBS patients. Referral to a gastroenterologist for IBS might be cost-effective for the NHS but more data, especially on potential QALY gains, are needed.

Introduction: Perioperative allogeneic red blood transfusions (ARBT) are associated with impaired short and long term outcomes. Consequently, perioperative ARBT should be avoided, yet preoperative anaemia increases this need. The study aimed to compare the efficacy of preoperative intravenous (IVI) and oral iron (OI) in reducing ARBT use in anaemic patients undergoing colorectal cancer (CRC) surgery. Methods: 116 anaemic patients with non-metastatic CRC adenocarcinoma were recruited preoperatively and randomised to receive either OI (ferrous sulphate) or IVI (ferric carboxymaltose). Perioperative changes in Haemoglobin (HB) and ARBT were recorded across groups. Parametric data was compared with 2 tailed T-test and non-parametric paired data with Wilcoxon Rank test, and Mann-Whitney U test. Nominal data was compared with 2-tailed Chi squared test. Results: There was no difference in demographic data between groups. HB levels at recruitment were comparable (OI 10.4g/dL 95%CI 10.1-10.7; IVI 10.2g/dL 95%CI 9.8-10.5, P=0.24), as was median treatment duration (OI 21 days IQR 15-33; IVI 21 days IQR 15-34, P=0.75). However, HB levels were higher on the day of Surgery in IVI (11.9g/dL 95%CI 11.5-12.3 vs OI 11g/dL 95%CI 10.6-11.4, P<0.01). Median preoperative HB change in patients not transfused preoperatively was higher in IVI (1.5g/dL IQR 0.9-2.6 vs OI 0.5g/dL IQR-0.1-1.3, P<0.01). There were fewer anaemic patients at surgery in the IVI group after treatment (75% vs 90%, P<0.05). OI patients received a mean 0.63u (95%CI 0.26-1) from recruitment to day 28 postoperatively vs mean 0.47u (95%CI 0.1-0.84) for IVI. Neither number of patients transfused (P=0.33) nor mean units transfused (P=0.54) differed over this period. When patients with heavy intraoperative losses (>1.5L) were excluded in subgroup analysis, a significant difference in mean units of blood transfused was seen up to 7 days post operatively (n= 108; OI 0.6u 95%CI 0.23-0.96; IVI 0.16u 95%CI 0.01-0.3, P< 0.05) and significantly less IVI patients were transfused (10% vs 25%, P<0.05) Conclusions: In patients undergoing CRC surgery, IVI appears more efficacious than OI at treating preoperative anaemia. It does not appear to minimise overall ARBT requirement, but may reduce ARBT use in the immediate perioperative period when the implications of ARBT are probably at their greatest.