The number of sections that should be submitted for histological examination varies according to the size of the tumor, as well as the degree of macroscopic tumor response in rectal cancer patients who have undergone preoperative chemoradiation. For large tumors, three or four sections usually suffice, provided that the areas of deepest tumor infiltration are sampled. At times, tumor regression postneoadjuvant rectal cancer therapy is so pronounced that the lesional area is reduced to a minute focus of thickening or ulceration. In these circumstances, it is essential that the entire area be submitted for examination to adequately assess tumor response.

The status of the resection margins is critical for all members of the multidisciplinary team involved in the care of cancer patients as a positive margin may imply additional therapeutic intervention. For the surgeon performing the procedure, it has added value since involvement of the resection margins may be reflective of the quality of the surgery (Nagtegaal and Quirke 2008).

In colon cancer, the proximal and distal margins are the most pertinent but are rarely involved. In rectal cancer, emphasis that once centered on the distal margin has now shifted to the radial or circumferential margins (CRM) . Numerous studies have shown that the incidence of local recurrence in rectal cancer is more frequently related to inadequate circumferential rather than distal margins (Birbeck et al. 2002; Li Destri et al. 2014). The CRM should be measured microscopically and it is generally accepted that the presence of tumor at the CRM or at less than 1 mm represents a positive margin (Fig. 22.3). Due to the impact of the CRM status as a predictor of recurrence, it is essential for pathologists to consistently measure the distance of the tumor to the CRM. It is recommended that, following evaluation of the mesorectal integrity, the perirectal soft tissue surrounding the tumor is inked before the specimen is opened. Scissors should be used to open the specimen longitudinally to avoid cutting through the tumor as this may lead to artifactual retraction of the CRM. It is for this reason that surgeons are discouraged from opening colorectal cancer specimens in the operating room. This is to ensure that for rectal cancer cases, the distal resection margin is tumor-free. In these situations, it is advisable to request an intraoperative pathology consult so that assessment of the quality of the mesorectum and inking of the CRM are accomplished prior to opening the specimen. As previously stated, for cases in which the lesion is <1 cm from the distal margin, the sections should be taken in a perpendicular fashion to include the margin and the tumor in the same section. This allows accurate microscopic calculation of the distance between the margin and the tumor (Fig. 22.4).

The presence of lymph node metastasis not only dictates prognosis but also impacts treatment decisions since the majority of patients with positive nodal disease are offered further treatment. It is important to acknowledge that the pathologist should have the same accountability as the surgeon in adequate lymph node retrieval. It is well known that the number of lymph nodes dissected from colon and rectal specimens has been used to judge the quality of the surgery and that a minimum of 12 nodes is required from each cancer specimen (Dillman et al. 2009; Stocchi et al. 2011). Although the number of retrieved nodes is, to a certain extent, related to the surgeon’s skill, it has been shown that another important variable influencing the number of dissected lymph nodes is the effort and diligence invested by the pathologist (Sarli et al. 2005). Although the adverse impact of positive nodes has been clearly demonstrated in patients with colorectal cancer, more recently it has become evident that the total number of dissected lymph nodes, regardless of the status (positive or negative), influences outcome (Swanson et al. 2003; Le Voyer et al. 2003). Several studies have shown a positive correlation between the total number of lymph nodes dissected and survival (Chang et al. 2007). In a systematic review, Chang et al. showed that in 16/17 reports longer survival was noted with an increased number of lymph nodes evaluated. More recently, Budde et al. analyzed 147,076 colon cancer cases extracted from the SEER database and confirmed a small but significant association between the number of lymph nodes found and survival.

The reason for the influence of lymph node harvest on outcome remains unclear. It would be logical to assume that the higher the number of lymph nodes examined, the higher the possibility of identifying positive nodes (upstaging). However, other factors likely play a role. It is known that deep infiltrating (Swanson et al. 2003) and poorly differentiated lesions (Chen and Bilchik 2006) and patients’ age (Soreide et al. 2006) are all associated with a higher number of lymph nodes retrieved. It is also possible that interactions between the tumor and the host occur at the molecular level, where malignant cells somehow stimulate the host’s immune response. In this regard, it has been shown that microsatellite unstable tumors tend to be correlated with high lymph node yield (Soreide et al. 2006). In order to increase lymph node harvest, several enhancing techniques have been tested over the years and have been shown to boost the number of dissected lymph nodes. Universal acceptance of these enhancing methods among pathologists has been disparate due to higher cost and lengthier procedures. Furthermore, certain techniques contain toxic chemicals such as xylene and acetone (Sanchez et al. 1997). More recently, simpler methods that entail soaking the mesocolon in pure alcohol for 24 h have been applied with great success (Wang et al. 2009). It is now well established that patients with advanced rectal cancer benefit from preoperative chemoradiation (Bosset et al. 2006). It is also well known that lymph node assessment in these situations is challenging as radiation leads to shrinkage of the lymphoid tissue, rendering smaller lymph nodes more difficult to identify using the traditional palpation and visualization methods (Wang et al. 2009). The utilization of techniques such as xylene, acetone, and alcohol soaking to enhance lymph node retrieval has been encouraged in this situation (Wang et al. 2009; Chapman et al. 2013).

As previously mentioned, patients with advanced rectal tumors are often offered neoadjuvant chemoradiation (Bosset et al. 2006). The pathologist’s role in these cases is to determine the amount of residual tumor burden or tumor regression. In the absence of obvious tumor after a known history of chemotherapy and/or radiation, areas of scarring should be searched and correlated with the patient’s prior diagnostic procedure for the location of the original tumor. Once the area of interest is identified, it should be sampled in its entirety to evaluate for residual malignancy (MacGregor et al. 2012).

Since Dukes’ initial observations regarding the prognostic impact of depth of tumor invasion into the colorectal wall, pathologists have consistently assessed and reported these variables. The tumor stage, both in Dukes and the TNM classification, is based on the degree of the anatomical involvement of the colorectal wall (Table 22.1). The influence on prognosis of the depth of tumor invasion into the colorectal wall is of great importance (Edge et al. 2010). This is best exemplified in the substratification of stage II colorectal cancer into Stages IIA, IIB, and IIC. In Stage IIA, the tumor infiltrates through the muscularis propria into the pericolorectal tissues (pT3). In stage IIB, tumor penetrates into the surface of the visceral peritoneum (pT4a), and in stage IIC, tumor directly invades or is adherent to other organs or structures (pT4b). Deeper tumor invasion through the wall is associated with a worse prognosis. According to the 7th edition of the AJCC staging manual, the 5-year observed survival rate for stage IIA is 67 %; for stage IIB, 59 %; and for stage IIC, 37 % (Edge et al. 2010).

Although the task of determining the depth of tumor invasion may appear simple and relatively straightforward, pathological staging may occasionally pose difficulties. The most common example is differentiation of extensive involvement of the pericolonic soft tissues (pT3) from the infiltration of the serosa by neoplastic cells (pT4a) (Fig. 22.5). Both situations are generally classified as stage II; however, the former falls into substage IIB while the latter would be classified as substage IIC, which has been related to poorer outcomes justifying the use of postoperative chemotherapy. Not uncommonly, histological sections are incomplete due to poor fixation or suboptimal technique and poor serosal visualization. This phenomenon may lead to substaging of a pT4 lesion to a pT3 lesion. It is therefore of outmost importance that pathologists carefully evaluate the quality of the histological slides being reviewed.

Generally, histological evaluation of the lymph nodes is a relatively simple task. However, small foci of metastatic cells or individual cells may be difficult to recognize, leading to understaging. Although the significance of isolated metastatic cells within the lymph nodes has been controversial over the years, more recently several authors have shown that micrometastasis in colorectal cancer is associated with poorer outcomes than actual pN0 (Nissan et al. 2012). It is for this reason that some pathologists have advocated the use of immunohistochemical stains such CK20 to highlight tumor cells (Nissan et al. 2012). Lymph nodes that have been entirely replaced by tumor and the so-called tumor deposits that result from vascular invasion in the soft tissue surrounding the colon and the rectum may be extremely difficult to locate. Moreover, this distinction is based on rather subjective and ever-changing criteria (Fig. 22.6) (Rock et al. 2014). In general, a stellate or irregular appearance of the tumoral focus would favor a metastatic tumor deposit due to venous invasion, whereas a round configuration would support a lymph node metastasis. This issue has generated much controversy as reflected by the evolving terminology applied in the various editions of the AJCC TNM classification in which the definition of tumor deposits and metastatic lymph nodes overlap according to each specific edition (Nagtegaal et al. 2011).

In addition to the pTN status, there are several other histopathological factors that correlate with outcome and should always be present in the pathology report. These include tumor differentiation, the presence of lymphovascular and perineural invasion, and the pattern of growth at the edge of the tumor (Compton et al. 2000; Betge et al. 2012; Liebig et al. 2009; Nakamura et al. 2008). Identification of these elements is of particular importance in stage II cases (negative lymph nodes), as further treatment in the form of chemotherapy may be appropriate for these patients.

Several studies have shown the impact of tumor differentiation on prognosis (Compton et al. 2000; Compton 2006). Tumor differentiation ranges from well-differentiated in which there are well-formed glands, to poorly differentiated in which glands are no longer distinguishable and instead consist of solid sheets of markedly atypical malignant cells with numerous mitoses (Fig. 22.7). As a results of a notable degree of interobserver variability in histological grading, the CAP and AJCC UICC recommends a two-tiered grading system with <50 % gland formation designated as high-grade and >50 % as well-to-moderately differentiated disease (Compton et al. 2000; Compton 2006).

Both lymphovascular (LVI) and perineural invasion (PI) have long been recognized as prognostic indicators in colorectal cancer (Betge et al. 2012) (Fig. 22.8). However, the power to predict outcome and influence treatment is somewhat overshadowed not only by the lack of consistency among pathologists in reporting these variables, but also by the inherent inter- and intraobserver variability in the assessment of these parameters (Compton 2006). Harris et al. (2008) studied this issue in a group of 7 expert gastrointestinal pathologists. These authors demonstrated that evaluation of H&E stained slides led to an overall fair level of agreement (k = 0.24). As the result of this phenomenon, some authors have encouraged the use of special stains to enhance lymphatic and blood vessel identification (Kingston et al. 2007). However, the application of these techniques remains controversial as they not only add time but also expense and are therefore not routinely performed.

Recognition of and reporting perineural invasion presents similar difficulties. This parameter is often underreported in published detection rates that range from 9 % to 42 % (Ceyhan et al. 2010). Similar to lymphovascular invasion, some authors have proposed the application of nerve-specific immunostains (S100) with the purpose of increasing the recognition of this marker (White et al. 2013). However, as previously stated, the added expense related to this technique renders this method unrealistic for use in routine practice.

Although the impact on prognosis of an infiltrative pattern of growth at the tumor border has been recognized for decades (Amato et al. 1994), in the past few years numerous publications have confirmed the finding that the phenomenon of “tumor budding” is a strong marker of prognosis (Harris et al. 2008). Tumor budding has been defined as the presence of individual cells or small clusters of cells at the invasive edge of the tumor. It has been shown that tumors that exhibit this feature are more frequently associated with lymph node and distant metastasis. It has been shown that, even in the presence of positive nodes (stage III), the presence of severe budding remains an independent poor prognostic factor (Choi et al. 2007).

Many of the postoperative treatment decisions in cancer patients depend on the pathological findings present in the surgical resection specimen. The anatomical extent of the disease represented by pathological tumor (pT) and nodal stage (pN) most commonly conveyed through the TNM staging system is of outmost importance to plan additional therapy and to predict outcome. Although pTNM is the most powerful marker of prognosis in cancer patients, it does not provide any information concerning the adequacy of the resection (margin status), the quality of the surgery, and the presence or absence of other histopathological or molecular factors that influence outcome. It is therefore imperative that all the relevant data are communicated to the clinician through the pathology report. In this regard it has been shown that the use of structured synoptic reports significantly decreases the possibility of omitting critical information (Zarbo 1992).