Confirmation of a diagnosis may be undertaken using a variety of techniques, including cytological examination of fine needle aspiration samples, tumour specimens from automatic cutting needles and surgical biopsies (Figs. 68-1 and 68-2). Core biopsies yield a higher tumour volume than fine needle aspiration, and may be more suited for tumour biomarker analysis, often required by clinical trials.

Staging Systems

An ideal staging system should be simple, precise, consistent and applicable to all clinical circumstances in oncology, and convey some prognostic information to facilitate best practice. Over the years, many staging systems have borne the name of eminent doctors (e.g. the Robson staging classification of renal tumours or Dukes’ staging classification of colorectal cancer), institutions (e.g. the Royal Marsden Hospital staging classification for testicular germ cell tumours) or organisations (e.g. the Fédération Internationale de Gynécologie et d’Obstétrique (FIGO) classification systems for cervical, uterine and other gynaecological neoplasms). More recently, the tumour–node–metastasis (TNM system), espoused by the American Joint Committee on Cancer (AJCC) and the Union Internationale Contre le Cancer (UICC), has been adopted widely.

The TNM classification of malignant tumours and the AJCC cancer staging handbook are now in their seventh editions.⁴ The system was originally devised by Pierre Denoix in the 1940s and has been modified over the subsequent decades. The ‘T’ category entails evaluation of local tumour extent. The ‘N’ category entails evaluation of nodal involvement. The ‘M’ category entails evaluation of disease at distant sites. The ‘T’ category, which may have the prefix ‘c’ to indicate ‘clinical’ staging, although this is frequently omitted, has several standard forms of notation: Tx indicates that primary tumour cannot be assessed; Tis indicates in situ disease with no evidence of invasion; T0 indicates no visible evidence of primary tumour; T1–T4 indicates increasing degrees of local tumour extent. These divisions may be adapted with the addition of subdivisions indicated by letters (e.g. ‘a’ or ‘b’) for greater flexibility in different tumour types. Although staging of the primary from T1 to T4 follows broad principles and there are some similarities between tumour types, refinements and adaptations for individual tumours are needed usually.

The ‘N’ category has similar notation. Direct spread of the primary tumour into an adjacent lymph node is classified as nodal spread. Nx is where regional lymph nodes cannot be assessed, N0 is where no regional lymph node metastases are present and N1, N2 and N3 indicate increasing involvement of regional lymph nodes. Likewise, these divisions may include subdivisions indicated by letters (e.g. ‘a’ or ‘b’). The‘M’ category assesses distant metastasis where Mx indicates that distant metastasis cannot be assessed, M0 indicates there are no distant metastases and M1 indicates the presence of distant metastasis. The category M1 may be further specified indicating which organs are involved. For example, PUL indicates pulmonary metastases, OSS indicates osseous metastases and HEP indicates hepatic metastases. Again, subdivisions may be indicated by letters (e.g. ‘a’ or ‘b’).

All tumours must be confirmed pathologically. A number of general rules apply when using the clinical TNM staging system. Clinical stage is assigned by physical examination, imaging and other relevant investigations, but may be amended as pathological information becomes available, and given the prefix pTNM stage where microscopic extent of disease is known. If there is doubt what stage should be assigned, the lower category should be used. Therefore, an imaging investigation that is suspicious but not diagnostic of spread to the pelvic sidewall will be disregarded unless supplemented by further imaging or confirmation by histopathology.

Once assigned, the pre-treatment TNM stage is recorded in the patient’s records and remains unchanged through subsequent treatment. For multiple synchronous primary tumours, the tumour with the highest ‘T’ category is used for staging purposes. For synchronous primary tumours arising in paired organs, each tumour should be assigned a separate TNM stage. In modern usage the TNM system has the advantages of clarity of communication, but is complex. This has led to a further system of stage grouping, which is published within the AJCC system. Stage groups of 0–4 are assigned as tumour becomes more extensive and widespread.

National bodies, such as the Royal College of Radiologists in the United Kingdom, have published recommendations or guidelines on the choice of staging investigation (CT, MRI or PET/CT), recognising that local availability of advanced imaging techniques and the preference and experience of individual radiologists are important considerations.⁵ A potential effect of advances in imaging technology over time is stage migration, e.g. via improvements in tumour detection, resulting in upstaging, and artefactual improvement in subgroup prognosis, although overall survival will remain stable unless more effective treatment is given.

In 1932, Dukes highlighted the importance of extramural spread in the prediction of local recurrence and survival. He also observed that lymph node invasion was present in 14% of patients with tumours confined to the bowel wall and 43% of patients with tumours extending beyond the serosa.⁶ A number of other prognostic indicators have subsequently been identified, including the pattern of local spread (a well-circumscribed margin implies a better prognosis than a widely infiltrated tumour with ill-defined borders). Spread beyond the peritoneal membrane results in a high incidence of both local recurrence and transcoelomic dissemination; invasion of extramural veins by tumour and extent and number of tumour-involved lymph nodes are all independent predictors of a poor prognosis.

For many years the preferred operation for rectal cancer was synchronous combined abdominoperineal excision of rectum (AP resection). However, total meso­rectal excision has become the gold standard since its introduction in the 1970s.⁷ In this technique, the surgeon dissects from above and finds the mesorectal plane. As the lymphatic vessels and nodes draining the tumour are located throughout the mesorectum, if this is divided or disrupted during surgical excision, spillage of malignant cells may occur and involved nodes may be left in situ. This increases the likelihood of local recurrence, and excision of the rectum is best achieved by dissection down the mesorectal fascia, allowing the mesorectum and rectum to be removed en bloc without disruption of the presacral fascia and the underlying venous plexus. The surgically excised specimen may then be sectioned transversely and the circumferential resection margin (CRM) examined for tumour involvement. The presence of tumour at the extremity of the resection margin is a major prognostic factor.⁸

In recent years, MRI has been shown to be the investigation of choice in demonstrating spread of tumour within the mesorectum and the mesorectal plane itself (Fig. 68-3). It is possible to demonstrate the relationship of the tumour to the anal canal, indicating the potential for sphincter-preserving surgery (Fig. 68-4). Criteria have been developed in pathologically correlated series for staging of the primary tumour and nodes within the mesorectum. Effectively, MRI can provide a ‘road map’ for the surgeon and studies have shown that the pathological status of the CRM can be predicted.^9,10 Other prognostic factors such as extramural venous invasion and peritoneal penetration can be demonstrated. If locally advanced disease with mural penetration is demonstrated, chemotherapy and radiation therapy may be employed to downstage the primary before an attempt at surgery, allowing a more tailored preoperative strategy.