Search

Background: Despite advances in treatment strategies and improvements in overall survival rates, oesophageal cancer remains an important cause of cancer-related deaths worldwide. Oesophageal cancer is an aggressive disease and most patients present with locally advanced or metastatic disease. As a result, only a minority of patients are suitable for curative resection. Over the last two decades, UK practice has been influenced by numerous large clinical trials demonstrating the benefit of neoadjuvant treatment compared to surgery alone. The specific gains afforded by adjuvant treatment remain largely unknown and only approximately 50% of patients scheduled for post-operative treatment complete it, due to the accumulation of acquired therapeutic toxicities. This thesis examines the current role of neoadjuvant and adjuvant therapies in patients with oesophageal adenocarcinoma. Aims and Methods: The main aim of this thesis was to assess patients with oesophageal adenocarcinoma as responders or non-responders in the context of the peri-operative treatment pathway. As the benefit of neoadjuvant treatment is well established, this thesis addresses the possibility of identifying clinical and radiological factors in relation to pathological response to neoadjuvant chemotherapy and if a combination of these factors could be used to predict chemoresponse. There is currently still a lack of evidence to support a clear benefit of adjuvant treatment in these patients. As such, this thesis addresses the potential survival benefit of adjuvant therapies in this group of patients, sub-classified and assessed based on surgical resection margin. This thesis comprises three main studies; one assessing neoadjuvant chemotherapy and two assessing adjuvant treatments in patients with oesophageal adenocarcinoma treated on a pre-determined, peri-operative pathway. A prospectively maintained, hospital-based operative database was used for all data collection. Adjuvant therapies were assessed using an ethically approved, collaborative database from two high-volume tertiary referral centres in the UK. Main outcomes for the studies included overall survival and time to recurrence in addition to recurrence patterns. A variety of statistical methods were employed across the studies including Kaplan-Meier survival analysis, Cox proportional hazards regression analysis, logistic regression analysis and receiver-operating characteristics (ROC) curve analysis. Results: Analysis of patients in the neoadjuvant treatment study demonstrated that a decrease in radiological tumour volume, a 'partial response' RECIST score, a clinically improved dysphagia score and lymphovascular invasion all influenced pathological response to neoadjuvant chemotherapy. Additional ROC curve analysis supported the association between these factors and the pathological response to neoadjuvant chemotherapy. The main finding demonstrated in the adjuvant study assessing patients with positive resection margins (R1), was an overall survival and recurrence-free survival benefit of adjuvant chemoradiotherapy when compared with adjuvant chemotherapy. In those with clear resection margins (R0), no independent survival benefit was observed, however adjuvant treatment was favoured in those with chemoresponsive tumours. Conclusion: In conclusion, this thesis has addressed the current role of neoadjuvant and adjuvant therapies in patients with oesophageal adenocarcinoma, classified as responders or non-responders, and in the context of overall and disease-free survival. This work has identified factors that can be used to pre-determine chemoresponse to neoadjuvant therapy and has highlighted the potential benefit of adjuvant chemoradiotherapy in margin-positive patients. These results can be utilised clinically to encourage a more individualised and tailored approach to the peri-operative pathway for oesophageal cancer patients.

Background: Oesophageal cancer survival has improved with the introduction of neoadjuvant treatment. Compared to surgery alone, neoadjuvant chemotherapy and chemoradiotherapy reduce rates of positive resection margins and overall tumour recurrence. Despite this, positive resection margin rates and recurrence remain disappointingly high following surgery. There is ongoing debate as to the optimum neoadjuvant treatment strategy and more emphasis is now being placed on tailoring neoadjuvant treatments to individual patients. Understanding risk factors associated with margin positivity and recurrence is essential in creating individualised neoadjuvant treatment pathways. Methods: A large prospectively collected database of consecutive resections was used. Included in the study cohort were patients who underwent oesophagectomy between 2000 and 2014 for oesophageal adenocarcinoma (AC) or squamous cell carcinoma (SCC). This data was rigorously updated and cross checked and was then analysed by a team of affiliated bio-statisticians. The aim of this thesis was to analyse risk factors associated with recurrence patterns in adenocarcinoma (study 1), assess the validity of current definitions of a positive circumferential resection margin and analyse the association between incremental oesophageal resection margin distance with recurrence and survival (study 2) and determine if clinicopathological and radiological variables can be used to predict circumferential resection margin positivity (study 3). Results: Pathological nodal disease in both pT1-2 (pN1 OR 2.72; 95%CI 1.35-5.48, pN2,3 OR 5.00; 95%CI 2.35-10.66) and pT3-4 patients (pN1 OR 3.03, 95%CI 1.51-6.07; pN2-3 OR 5.75, 95%CI 3.15-10.49) predisposed to systemic recurrence. Lymphovascular invasion (OR 2.09, 95%CI 1.18-3.71) and preoperative stenting (OR 3.70, 95%CI 1.34-10.23) were independent risk factors for isolated locoregional recurrence in oesophageal adenocarcinoma. Poor or no response to chemotherapy was also an independent risk factor for isolated systemic recurrence (OR 1.85, 95%CI 1.05-3.26). An involved circumferential resection margin (CRM) was not associated with significantly increased risk of isolated locoregional recurrence (OR 1.37, 95%CI 0.81-2.33). Kaplan-Meier and unadjusted analyses found a significant incremental improvement in overall survival (p<0.0001) and recurrence rates (p trend < 0.0001) with increasing distance from the CRM. Tumour distance of >2mm remained a significant predictor of survival on multivariable analysis (risk of death HR 0.66; 95%CI 0.44-1.00). Multivariable analysis of overall survival demonstrated a significant difference between a positive and negative CRM with the Royal College of Pathologists definition (HR 1.37 95%CI 1.01-1.85) and not with the College of American Pathologists' definition (HR 1.22 95%CI 0.90-1.65). Poor tumour differentiation (OR 2.84, 95%CI 1.39-6.01) and advanced clinical T-stage (T3-4) (OR 2.93 95%CI 1.03-9.48) were independently associated with an increased risk of a positive CRM. CT non-response (stable or progressive disease) following chemotherapy independently corresponded with an increased risk of CRM positivity (OR 3.38 95% CI 1.43-8.50). Additional CT evidence of local invasion/pleural thickening and higher CT tumour volume (14cm3) improved the performance of a prediction model, including all above parameters; with AUC (c-index) of 0.76 (0.68-0.83). Conclusion: Analysis of clinicopathological risk factors associated with positive circumferential resection margins and recurrence patterns demonstrate the complexity of adenocarcinoma of the oesophagus. Adenocarcinoma of the oesophagus has a propensity to recur systemically and neoadjuvant treatment strategies should take account of this. Understanding the mechanisms behind the aggressive nature of some cell lines to disseminate early will be vital in the next step in developing a patient centred neoadjuvant and adjuvant treatment strategy.

Introduction : Oesophageal cancer represents a formidable challenge to both patients and clinicians. Due to its propensity for early systemic dissemination, the majority of patients are not eligible for curative treatment. In the minority of patients suitable for surgical resection, there remain many controversies in management. Good prospective data from high volume centres is vital in attempting to improve staging algorithms and management strategies. Methods : A large prospectively collected database was utilised, the result of a research collaboration between two high volume oesophageal cancer centres (St Thomas’ hospital and Royal Marsden hospital) in London, UK. The database consisted of consecutive patients undergoing surgery for oesophageal or oesophago-gastric junction (type 1 and 2) tumours between 2000 and 2010. The database was rigorously cross-referenced to ensure accuracy. Data were analysed independently at an aligned academic unit by an experienced bio-statistician. The aim was to assess factors predicting early recurrence and death after oesophagectomy (study 1), the influence of surgical radicality on outcomes (study 2), and the down-staging effects of neo-adjuvant chemotherapy (study 3). Each of these was to be presented as a scientific article, accepted for peer-review publication. Results : Of the 680 patients included, the median age was 64 years with a male preponderance (81%). The majority of patients had adenocarcinoma (82%), although patients with squamous cell carcinoma (14%) and high grade dysplasia (4%) were also included in the database. Multivariable analysis showed T and N stage (T3-4 N2-3 OR 10.6; 95% CI 2.8-40.0), poor differentiation (OR 2.8; 95% CI 1.4-5.5), involved resection margins (OR 2.7; 95% CI 1.2-6.0) and poor response to pre-operative chemotherapy (OR 3.2; 95% CI 1.1-8.8) to independently predict early recurrence and death after surgery. The predominant mode of recurrence was with distant metastases. Surgical approach, comparing transhiatal with transthoracic oesophagectomy, had no impact on overall survival (HR 1.07, 95% CI 0.84 – 1.36) or tumour recurrence (HR 0.99, 95% CI 0.76 – 1.29) when adjusted for potential confounding factors. In patients undergoing neo-adjuvant chemotherapy, outcome was determined by tumour stage after chemotherapy rather than that of initial presentation. This may have a significant impact on staging algorithms with a shift in focus to assessment of tumour stage after chemotherapy in order to improve clinical decision making and predict outcome. The three studies were published in the Journal of Surgical Oncology, British Journal of Surgery and Journal of Clinical Oncology respectively. Conclusions : Predictive models that could guide individualised patient management are achievable. These could include selecting patients for specific neo-adjuvant treatment strategies, guiding surgical approach in patients deemed suitable for resection and improved selection of patients who are most likely to benefit from surgery.

This thesis aimed to study the occurrence and treatment of oesophageal and gastric cancer patients in England between 1998 and 2009. Variation in incidence over time, by socioeconomic deprivation, and between ethnic groups for six subgroups of these cancers (upper and middle oesophagus, lower oesophagus, oesophagus not otherwise specified (NOS), cardia, non-cardia, gastric NOS) was investigated. Factors that affected who received surgery for these cancers were examined and survival in relation to hospital volume (measured as the annual number of operations per hospital) was assessed. Key findings include an increase in the incidence of lower oesophageal cancer in men, a higher incidence of this and cardia cancers in men, a higher incidence of these subgroups in White men compared with the other ethnic groups studied, a higher incidence of upper and middle oesophageal cancer in Bangladeshi women compared with White women, and in each subgroup a higher incidence in more deprived areas and a poor prognosis. The different patterns of incidence are likely to be explained by differences in exposures to risk factors. These findings confirm patterns reported in other developed countries, and highlight the importance of preventative initiatives. Older patients and patients resident in more deprived areas were less likely to undergo surgery. This could be explained by several valid reasons, including differences in performance status, disease stage, or willingness to undergo surgery. Data on these factors were not available for analysis but should be further investigated to ensure that surgery is available to all who will benefit clinically from it. This work found lower mortality with increasing hospital volume, both in the short- and long-term. For the first time in England, this work assessed the effect of hospital volume on survival and found that the centralisation of surgical services for these patients has been beneficial. The work supports the continued centralisation of these services.