Colorectal cancer

Introduction

Colorectal cancer is a significant healthcare issue in terms of prevention, timely diagnosis, treatment and survivorship. Improving outcomes remains a key healthcare challenge in the UK, where bowel cancer is the second most common cause of cancer death, accounting for over 16 000 deaths annually. Over 42 000 new cases are diagnosed each year, the majority arising in the elderly, although increasing incidence in younger patients has been recorded worldwide. Overall numbers are higher in men, especially in the rectum, but right colon cancer incidence is higher in women. The overall 5-year survival rate is around 58%, with ‘early’ disease yielding 5-year survival over 90%, compared to only 10% in those with metastases at diagnosis.

New National Institute for Health and Care Excellence (NICE) guidance focuses on information for patients and shared decision making, reflecting changes in UK societal expectations around valid consent; quality of life considerations also feature, as do minimum numbers for rectal cancer major resections. Accurate outcome data are key to quality improvement and the National Bowel Cancer Audit Project underpins this process in England and Wales.

Distinguishing the colon and rectum is important – treatment regimens differ and comparisons between treatments, or minimum numbers for site-specific treatments, cannot be measured without uniform definitions. The colon comprises the large bowel proximal to the rectum but, historically, the definition of the rectum has been variable. Anatomical texts describe the top of the rectum as the point where the sigmoid mesocolon ends or that part of the large bowel level with the third sacral vertebra, whilst some surgeons think of the rectum as the large bowel segment lying within the true pelvis. Another intra-operative definition is the fusion of the two anti-mesenteric taenia into an amorphous area where the true rectum begins. In the UK, a rectal cancer is a tumour within 15 cm of the anal verge, whereas US authorities have preferred 11 or 12 cm. Standardised use of cross-sectional radiology for staging and expanding non-operative options for rectal cancer mean a radiological definition may be most appropriate. The ‘sigmoid take-off’ is the point where the fixed mesorectum ends and no longer tethers the rectum to the sacrum, whilst the mesocolon elongates, and consensus around this reproducible radiological definition may grow over time.

Natural history

Approximately 50% of cancers arise in the rectum and left colon, 25% in the right, with synchronous lesions in around 5% of cases ( Fig. 2.1 ). Most colorectal cancers arise from pre-existing polyps ( Box 2.1 ). The majority of adenomas diagnosed in the West are polypoid or exophytic, but the flat adenoma, where the depth of the dysplastic tissue is no more than twice that of the mucosa, may account for up to 40% of all adenomas. In addition, the serrated lesion, which is histologically distinct and related to the hyperplastic polyp, is now a recognised pre-malignant condition. These characteristically right-sided lesions, are easily missed and may represent pre-malignant lesions in an ‘accelerated’ cancer pathway, perhaps explaining some interval cancers that arise between endoscopies. Reliable diagnosis requires a skilful, experienced endoscopist, using dye spray on the colonic mucosa and narrow band imaging (NBI) to aid detection. Newer techniques, including mechanical adjuncts that improve luminal views and digital software that highlights irregular mucosa for closer inspection, are discussed in Chapter 1 .

When invasion has taken place, colorectal cancer can spread directly and by the lymphatic, blood and transcoelomic routes.

Aetiology

Colorectal cancers arise as a result of a combination of environmental and inherited risk factors. In most, environmental factors are the main cause, with only 5% of cancer patients having a high-risk inherited genetic predisposition, although lower risk genetic variants probably play a part in many more. Cancers arise due to a complex interplay between the intra-luminal environment, host-tumour immune responses and the colonic mucosa, wherein genetic alterations drive neoplasia. Knowledge of the molecular genetics of colorectal cancer has increased rapidly and it is now evident that numerous pathways exist, layers of redundancy allowing cancers to escape host responses and progress. The gut microbiome may act as a mediator of environmental factors. The microbiomes of colorectal cancer patients display distinct properties, although it is not clear whether this is a ‘cause or effect’ of disease. Certain bacterial species demonstrate altered profiles in bowel cancer patients and perhaps some ability to modulate disease progression and response to treatment ( Fig. 2.2 ). Further studies in this area might drive improvements in prevention, diagnosis, individualised treatment and quality of life – the use of pre-operative antibiotics with mechanical bowel preparation demonstrates these concepts influencing current practice.

Presentation

Around 20% of colorectal cancers present as emergencies, despite established pathways that encourage urgent referral for recognised chronic symptoms. This highlights that patients with colorectal cancer at all stages may be asymptomatic, fail to consult for symptoms, or only report vague symptoms. The commonest emergency presentations are bowel obstruction and rectal bleeding. Occasionally, emergency presentation may relate to a palpable mass, more rarely a sigmoid cancer causes pneumaturia or urinary infection by fistulation into the bladder, or a gastrocolic fistula may cause faecal vomiting or severe diarrhoea.

Most diagnoses are made after primary care referral and since 2000, pathways to investigate ‘high risk’ patients urgently, based on age and symptoms, have been implemented in the UK ( Table 2.1 ). However, notwithstanding an initial decrease in emergency presentations, these pathways and sustained public awareness campaigns have increased the number of patients being investigated without positive effect on the stage at diagnosis, and similar numbers of diagnoses are made on routine pathways. Updated guidelines broadened referral criteria in 2015, reducing the threshold for urgent referral criteria to a positive predictive value of 3%, aiming to increase diagnosis at early stage and improve clinical outcomes. This further exacerbated demand for diagnostics without any demonstrable benefit, but also re-introduced the concept of testing for occult blood in faeces. There is now significant interest in the use of quantitative faecal immunochemical testing (FIT) for detection of faecal haemoglobin (fHb) in symptomatic patients, although initial guidance recommended use specifically in ‘low-risk’ patients that do not qualify for urgent referral.

FIT is a risk stratification tool and effectively segments the symptomatic population into a majority with very low risk (around 0.5%), and groups with higher risk.

These data broadly concur with other research studies and three large service evaluations where FIT was used to stratify symptomatic patients in Primary Care. In Northern Spain and Tayside, all symptoms were included, and in Nottingham rectal mass and bleeding were excluded; the cancer diagnosis rate after reassurance without investigation based on low fHb results in primary care was consistently 0.3% or less at follow-up. FIT misses a small number of colorectal cancers, but far fewer than referral criteria based on symptoms and age and identifies a population with low risk in whom investigation might be deferred or undertaken routinely. Anaemia, iron deficiency and thrombocytosis may all have added value in defining optimal thresholds, or may yet combine with FIT and demographics to improve on existing scoring systems. Patients with very high fHb readings harbour increased risk of bowel cancer, or significant bowel pathology, and should be prioritised for urgent investigation and ‘one-stop’ services. Point-of-care testing platforms are emerging but clear guidelines around current uncertainties are awaited ( Box 2.2 ).