Keywords

Introduction

Pathology as a diagnostic branch is an important pillar in the multidisciplinary management of most of the diseases including management of cancers. The insight provided by the microscopic features of any disease, pathological evaluation of tissue, is pivotal and an essential component in securing the best outcome in the multidisciplinary/multispecialty management of any cancer including gastrointestinal (GI) cancer.

GI cancer diagnosis involves multiple steps with various specialties including clinical examination for evaluating symptoms and signs, which guides the selection of an appropriate combination of imaging modalities, endoscopy, and various approaches for tissue diagnosis. The ultimate step is tissue diagnosis, the gold standard, with help of various biopsy methods. Sampling artifact due to missing of the actual pathology by random approach may be avoided by applying targeted methods guided by high-resolution endoscopy, such as different types of endomicroscopy in an effort to achieve in vivo histology-like real-time details (optical biopsy) [1,2,3,4,5].

Any of these methodologies has to conclude with appropriate expertise in ruling out various morphological mimickers by weeding out potential pitfalls in marching toward the correct diagnosis. Careful scrutiny of a variety of morphological features in the tissue specimens under examination is the most important step. Generally, the differential diagnosis involves a wide spectrum, spanning from reactive process at one end to various benign and malignant tumors at other end. If the morphological features are not sufficient enough to reach conclusive interpretation, a variety of ancillary tests may have to be applied. These ancillary tests include immunophenotyping by immunohistochemistry (IHC) or flow cytometry, fluorescence in situ hybridization/chromogenic in situ hybridization (FISH/CISH), cytogenetics, various molecular tests, electron microscopy, etc. Because it is easily adaptable to the routine anatomic pathology workflow using light microscopy, IHC is the most frequently used tool for evaluating diagnostic and prognostic immunomarkers. In addition, IHC has many other practical benefits, including feasibility to perform the immunostaining on archivable formalin-fixed paraffin-embedded (FFPE) tissue/cell-blocks. IHC slides can be stored like surgical pathology slides for future record. Ongoing refinement and availability of an ever-widening battery of immunomarkers along with increasing availability of multicolor immunostaining options for improved interpretation are continuously strengthening its ancillary status.

Thus, the interpretation of tissue for the diagnosis of any cancer is based on microscopic evaluation of morphological features with or without ancillary tests including immunophenotyping (immunohistochemistry/flow cytometry), cytogenetics, and variety of molecular pathology tests. Another component of interpretation is proper classification, which by itself, is an ongoing process based on increasing understanding with advances in the field of molecular pathology. Due to this, there are many tumor classifications for various cancers. However, depending on regional/local preferences and standard of practice, one or other classification is favored. In general, some classifications, such as the World Health Organization (WHO) classification [6] are favored over others. The tumors are generally classified based on their morphological features matching with its normal counterpart. This has been termed histogenesis (tissue of tumor origin). However, the preferred approach would be to consider the resemblance of a particular tumor to a particular type of normal tissue as its differentiation into that tissue type rather than as evidence of tissue of origin or histogenesis.

Although not significantly important for all tumors, grading of tumor is an additional component of tissue diagnosis. Most of the approaches involve comparison of the tumor differentiation with the normal counterpart. Tumors with morphological resemblance closer to the normal spectrum would be “well differentiated” and the one lacking significant differentiation as “poorly differentiated,” with “moderately differentiated” falling between the two extremes. This approach may be modified in some specific tumor/organ systems, such as in the application of mitotic figure count (proliferation status) and necrosis for grading neuroendocrine tumors (NET) [7] and gastrointestinal stromal tumors (GIST) [8, 9]. Ancillary tests such as KI-67 index may be applied for improved objectivity in tumor grading based on parameters related to proliferation [9a] are important factors to be considered for making treatment decisions. These features should be included in final pathology report under summary/synoptic report [10].

After tissue diagnosis and its proper classification, staging of that tumor has prognostic significance and is a critical component of any surgical pathology report on the resection specimen for proper clinical management. Currently, TNM (Tumor, Node, Metastasis) staging is the most widely practiced staging system. Based on various experiences, The American Joint Committee on Cancer (AJCC), in cooperation with the TNM Committee of the International Union Against Cancer (UICC), has incorporated these factors and developed a comprehensive TNM staging system, which is revised periodically [10,11,13, 15]. Each of the three components in TNM is given an incremental number as the tumor shows worsening features in that category. T (Tumor topography) is usually based on the size of tumor or the depth of the tumor invasion in tubular GI organs. Larger tumor size and/or deeper tumor invasion equates with a higher stage. N (extent of regional lymph node involvement) and M (evidence of distant metastasis) indicate the status regarding the spread of the tumor beyond the primary site as additional prognostic indicators. Depending on T, N, and M status, the AJCC has compiled various permutations and combinations into progressive groups from Stage 0 to Stage IV. In addition to TNM, other features such as Tumor deposits, Preoperative blood level of CEA, Tumor regresion score, Circumferential resection margin, Lymphovascular invasion, Perineural invasion, Microsatellite instability, KRAS and NRAS mutation, and BRAF mutation. Currently, this staging is one of the most important prognostic determinants and is important information in guiding the treatment plan [13]. Please see Table 2.1 in which “Colon carcinoma” is chosen as the organ system as an example for TNM staging [13]. The prognosis of higher stage cancer is poorer than lower stage cancers with shorter 5-year survival rates, even after curative resection [14].

Table 2.1 TNM staging based on AJCC eighth edition using colon as example (comparable approach with organ-specific details is applied for other tubular GIT) (see Fig. 2.5)
Full size table

Various organ systems have comparable methods to TNM staging, which may be modified in some cases based on the type of neoplasm. For example, TNM staging of the appendix for adenocarcinoma including goblet cell carcinoid (crypt cell carcinoma) is different than for neuroendocrine tumor (carcinoid) for the same organ (Table 2.2) [15].

Table 2.2 Appendix: Comparative TNM staging according to AJCC applied to carcinoma versus neuroendocrine tumor [15]. (Note that for Appendix, in addition to TNM, grade of the tumor is also a consideration for staging of carcinoma, especially subcategorization of stage IV)
Full size table

The role of molecular pathology is evolving due to the ongoing introduction of a variety of targeted therapy for various GI cancers. The classical example is the role of KIT (CD117) in establishing the diagnosis of gastrointestinal stromal tumor (GIST) with evaluation for various KIT mutations related to the response to tyrosine kinase inhibitors such as Gleevec [16]. Other molecular tests are evolving continuously with an increasing role not only in treating GI cancer patients but also in monitoring/evaluating their relatives. An example includes evaluation for mismatch repair (MMR) genes for microsatellite instability (MSI), which is linked with the hereditary form of colorectal cancer in Lynch syndrome [17].

Most of this information is currently included as part of the final report on most of the definitive resections and some of the biopsies as per the College of American Pathologists (CAP) checklist for a particular tumor/organ [10, 18].

Pathological Evaluation

The standard of practice requires tissue diagnosis prior to initiation of treatment. Many lesions – both benign (including benign ulceration [usually due to ischemia or inflammatory processes, such as Helicobacter pylori infection in the stomach or cytomegalovirus infection in the colon], inflammatory conditions such as inflammatory bowel disease [Crohn’s disease or ulcerative colitis], solitary rectal ulcer syndrome, and diverticular disease with mural stricturing, hamartomas, endometriosis, and adenomas) and malignant (including neuroendocrine tumors, lymphomas, mesenchymal tumors [e.g., GIST]), metastatic tumors with tendency for gastrointestinal tract metastases (e.g., melanomas), and malignancies growing into GI tract (GIT) from adjacent organs (e.g., cancers of the ovary, endometrium, urinary bladder, or prostate]) – may clinically resemble GI carcinomas. Due to this, it is critical to confirm the tissue diagnosis prior to definitive therapy as a standard of practice for the best outcome.

Tissue diagnosis and pathological evaluation may be achieved by various biopsy methods including fine-needle aspiration (FNA) biopsy (with its variants such as endoscopic ultrasound [EUS]-guided FNA, which is very important for evaluation of lesions of deeper organs such as the pancreas and other sites accessible through the tubular GI system) and other cytopathology methods including brushings, washings/lavages, and cyst aspirations. Surgical pathology approaches include endoscopic forceps biopsies/resection of small lesions such as polyps, core biopsy (including image-guided core biopsy), wedge biopsy (including laparoscopic biopsies), and ultimately resection specimens. Each of these approaches has benefits and limitations discussed briefly as follows.

Cytopathological Evaluation

Cytology has multiple advantages with the ability to evaluate excellent cytomorphological details (Fig. 2.1) over surgical pathology biopsy (Fig. 2.2). The principal mechanism by which the diagnostic material is retrieved by FNA facilitates selective suction of poorly cohesive neoplastic cells (Fig. 2.3) over supporting stroma, as compared to coring out of both stroma and tumor cells by core biopsy along the tract for that core (Fig. 2.3). FNA procedure samples a relatively wider area of the lesion because of the nature of the procedure in which the sampling FNA needle has to be moved back and forth in different directions in the tumor. Most of the sampled material is seen directly on the slides under scrutiny (in contrast to just a tiny fraction of the sampled surgical biopsy tissue as just a 4-micron thick tissue section) [19]. In addition to rapid turnaround time and lower cost, these specimens provide the opportunity to evaluate the cytomorphological features of tumor/lesion cells at a higher level of clarity with excellent nuclear details allowing precise diagnosis even with limited material (Fig. 2.1). In addition to the initial tissue diagnosis (Fig. 2.1), cytopathology contributes to the staging of many GI cancers such as TNM staging of colon cancer. Positivity of tumor cells in peritoneal fluid cytology is equivalent to the distant metastasis properly assigning a status of AJCC stage IV to these cases.

Fig. 2.1
figure 1

Diagnostically crisp cytomorphological details in cytopathology samples (e.g., pancreatic ductal adenocarcinoma) (Pap stain – direct smear). (a) (inset): Cohesive group of neoplastic cells with sudden nucleomegaly. Variation in size of tumor nuclei: The difference in size between smallest (red arrowhead) and largest (blue arrow) nucleus in the group is at least 1:4. (b): 1. Large cell with high nuclear:cytoplasmic ratio; 2. irregular nuclear margin; 3. coarsely clumped irregularly distributed hyperchromatic chromatin; 4. parachromatin clearing; 5. nucleoli with irregular outlines; 6. cytoplasmic vacuoles with secretion (all these features collectively are consistent with adenocarcinoma)

Full size image
Fig. 2.2
figure 2

Pancreatic adenocarcinoma (H&E) (a). Surgical pathology biopsy samples all tissue in the trajectory of the biopsy needle. Four-micron section of pancreatic ductal adenocarcinoma shows only fraction of the neoplastic epithelial component with predominance of stroma in section from tumors with predominance of desmoplastic stroma (compare with Fig. 2.3) (b). The morphology of individual tumor cells is relatively suboptimal as compared to cytology specimen (compare with Fig. 2.1). Similarly, the evaluation of sudden nucleomegaly is also relatively less dependable in surgical pathology sections

Full size image
Fig. 2.3
figure 3

Pancreatic adenocarcinoma (H&E). Cell-block section of FNA biopsy specimen (H&E stain). Note tumor with predominantly neoplastic epithelial component without significant proportion of stroma

Full size image

However, depending on a particular situation, invasion cannot be evaluated directly in the cytology specimens, although some indirect evidence such as tumor diathesis in the background with relatively higher cellularity may suggest that. Similarly, although some architectural details may be observed, it may not be comparable to that seen in surgical pathology (histopathology) tissue sections . Both these limitations could be overcome by using improved techniques for achieving best cellularity in cell-block sections from an adequately cellular cell-block [20, 21]. Recent advances for improving cellularity of cell-blocks allows maximum retrieval of diagnostic material in cell-block sections [19] [20a]. Cell-block also allows application of ancillary tests including IHC for differential diagnosis, for evaluation of prognostic markers, and for evaluating primary versus metastatic nature of the tumor. With the ever-increasing role of molecular tests, the cell-block is an excellent resource for many of these tests to be performed as indicated synchronously or at a later time on the archived FFPE cell-blocks. During on-site adequacy evaluation, it should be recommended to submit dedicated passes/material for cell-block preparation for future elective tests as clinically indicated. All these advantages of cell-blocks with recent advances are discussed in detail in the recent review article on CellBlockistry [21a].

Cytological approaches facilitate preoperative tissue diagnosis of lesions, especially those that otherwise may not be accessible with conventional biopsy due to complex locations (most of the pancreatic lesions) or due to potential risks of biopsy-associated complications such as needle tracking. However, due to the relative complexity in interpreting cytopathology material, the availability of expertise may be limited to some special centers.

Onsite adequacy evaluation is an important component to navigate the exact area to be sampled and provide real-time input for retrieval of adequate diagnostic material with triage feedback for appropriate supporting tests such as flow cytometry, microbiology cultures, and cytogenetics. The final goal of onsite adequacy evaluation is to achieve diagnostic material for unequivocal cytopathological interpretation , which for adenocarcinoma and other nonhematopoietic lesions is heavily dependent on evaluation of Papanicolaou (Pap) stained smears. Due to this, it is important to ensure retrieval of diagnostic material on the Pap-stained smears (instead of Diff-Quick [DQ]-stained smears), especially when only a suboptimal scant specimen could be available for final interpretation. In such cases, use of DQ stain initially for onsite adequacy may compromise the final interpretation especially if the lesion turns out to be a well-differentiated adenocarcinoma with scant material, leading to atypical/suspicious type suboptimal final report. Conventionally, wet-fixed smears are needed for Pap staining and air-dried counterpart for DQ staining . However, air-drying of all smears allows application of either Pap stain (after rehydration with post-fixation) or DQ stain electively [22]. Based on published study and long personal experience, using air-dried smears with routine availability of rapid Pap staining protocol during onsite adequacy evaluation is recommended for increasing the chances of final unequivocal cytopathological interpretation of most GI nonhematopoietic lesions [22].

Commonly used approaches for cytological sampling of lesions in various organ systems are summarized in Table 2.3 [23, 24].

Table 2.3 Cytopathological evaluation of GI lesions
Full size table

Surgical Pathological (Histopathological) Evaluation

Surgical pathological (histopathological) evaluation of lesions suspicious for cancer identified after clinical examination in concert with different imaging modalities and/or various endoscopic studies with suspicion for malignancy is another modality available in addition to cytopathological methods. Similar to cytopathological evaluation, the role of the biopsy is to distinguish benign lesions from clinically neoplastic mimickers and to rule out or rule in malignancy along with histological typing of the tumor.

Similar to cytopathological evaluation , a variety of approaches may be used to retrieve tissue for surgical pathological (histopathological) evaluation of suspicious lesions. The methodology may range from minimal representative sampling to total resection in various forms and may be categorized mainly into:

  1. 1.

    Diagnostic sampling

    1. A.

      Diagnostic biopsies (may be supported by guidance from onsite adequacy evaluation for precise sampling of the lesion by intra-procedural cytology smears)

      1. (a)

        Needle core biopsies

      2. (b)

        Endoscopic forceps biopsies

      3. (c)

        EUS-guided core biopsies

    2. B.

      Wedge biopsies

    3. C.

      Excisional biopsies

  2. 2.

    Therapeutic excisions

    1. A.

      Wide excisions, including endoscopic mucosal resections (EMR) [25]

    2. B.

      Radical resections

The major benefit with most of the surgical pathology specimens is the ability to evaluate the tissue architecture and invasion (Fig. 2.2). Even though FNA with good cell-block has numerous benefits as stated previously, there may be a tendency to prefer needle core biopsy over the relatively skill-dependent FNA procedure due to perceived ease in performing core biopsies. Generally, the cytopathological approach has a higher chance of diagnostic outcome as compared to core biopsies with small/tiny tissue for surgical pathology, especially for tumors with a tendency for sclerotic/desmoplastic stroma (e.g., pancreatic ductal carcinoma) [26, 27].

However, the final result with surgical pathology depends on a variety of factors including how the tissue is collected, from where it is collected, how it is fixed and processed, and the final quality of tissue sections with elective application of ancillary tests for final interpretation.

For diagnostic biopsies, it is important to sample the proper area of any lesion. For sampling ulcerated lesions and retrieving representative diagnostic material, the specimens should be taken from all 4 quadrants of the ulcer edge (e.g., for ulcerated carcinomas) and its base (e.g., for ulcerated lymphoma and sarcoma). The surface of the polypoid lesions would be the representative tissue. However, superficial biopsies such as from tubular gut or the ampulla of Vater, extrahepatic bile ducts, and pancreatic ducts may be limited by the difficulty in evaluating the invasion and its depth. For sampling obstructive lesions, an endoscope may not be negotiable and so may be difficult to biopsy. In such cases, brush cytology is an appropriate alternative. Some deeper lesions such as lymphomas, neuroendocrine tumors, GIST, and sarcomas usually have deeper submucosal mural growth pattern. Such lesions may be missed in superficial luminal biopsies and so, in this clinical situation , the same specific biopsy site should be sampled repeatedly to retrieve the representative deeper tissue. Tumors with extensive necrosis may not provide a sample with viable diagnostic component. In such cases, sampling multiple biopsies, especially from the periphery of the lesion, would enhance the possibility of sampling viable diagnostic tissue. In addition, core biopsies may not sample diagnostic material or if it samples diagnostic tissue, it may not be sufficient for precise grading of some lesions such as NETs and GIST. Calculation of Ki67 (MIB1) index (need at least 500 to 1000 tumor cell nuclei) and mitotic figure counting (need up to 50 high power fields) may not be precise on specimens with scant viable tumor components [7, 8].

Intraoperative Consult (Including Frozen Sectioning and Imprint/Scrape Cytology Smears)

The final management, especially resection, may need intra-procedural input to guide the surgical treatment. The most common indication is evaluation of the resection margins for the tumor. Other benefits of the intraoperative consult include triaging of the fresh specimen for ancillary studies such as cytogenetics, flow cytometry, microbiology culture, and ultrastructural (electron microscopic) studies as indicated based on preliminary morphological evaluation. Some of these ancillary tests may not be possible at a later stage once the tissue is fixed. It is important not to use frozen sections (FS) routinely just for the diagnosis, especially on tiny biopsies and tissues with predominance of fat. Performance of FS without considering this limitation may compromise the morphology required for optimal final interpretation, including interference with some studies such as elective IHC. In case the tissue diagnosis input is a must on such specimens, intra-procedural imprint/scrape cytology smears is a better option [27a].

Specimen Handling

For the best interpretation outcome, both cytopathology and surgical pathology specimens have to be collected, handled, and processed properly. All personnel associated in this process should be aware of limitations and precautions with emphasis on coordination and communication between different entities involved in it for the best outcome. Compromisation may affect the integrity of the specimen needed for the best outcome. Improper fixative, inappropriate fixation time, or prolonged ischemic time (time from excision to putting the specimen in the fixative) may compromise the results of ancillary tests, especially the immunostaining pattern/immunophenotype.

Although cytopathology specimens have many benefits as mentioned previously, they also have many challenges due to the complexity in choosing an appropriate collection protocol [20]. Close collaboration with the cytopathology laboratory is needed to achieve the best outcome. The simplest approach would be to submit a fresh specimen to the cytopathology laboratory for immediate processing. Similarly, air-dried direct cytology smears allow more flexibility and may be processed for both Pap and Diff-Quik staining with multiple benefits [22]. If this is not possible, it should follow the protocol standardized for their particular laboratory/institution (Table 2.4) [20, 22].

Table 2.4 Cytopathology specimen submission protocols
Full size table

Small surgical pathology specimens such as core/forceps biopsies in general can be submitted in 10% formalin. Large specimens may be submitted in 10% formalin or as fresh, but fresh specimens must be processed immediately for appropriate final outcome. Fresh unfixed specimen provides the benefit and flexibility of applying different protocols, but not without the risk of compromising tissue integrity if immediate processing cannot be guaranteed. Some specimens may need special attention with preliminary orientation and processing to avoid a sub-optimal outcome. A good example in this category is endoscopic mucosal resection (EMR) specimens. These specimens should be oriented and mounted by pinning onto a paraffin wax block or cork board before submitting in fixative prior to transportation to the laboratory [25].

Application of Various Ancillary Tests

Routine morphological evaluation may not be sufficient for reaching a definitive interpretation, especially with limited biopsy specimen, scantly cellular cytology specimen, or some lesions such as poorly differentiated tumors. Ancillary methods including immunohistochemistry, in situ hybridization (FISH and CISH) , other molecular tests, ultrastructural studies (electron microscopy), or histochemistry may be indicated.

The most powerful and practical tool widely used currently is IHC. Other tools have relative limitations and are used sparingly. Electron microscopy needs planning from the beginning of the biopsy procedure when the tissue is still fresh, so that it is appropriately processed with special fixative (glutaraldehyde). In addition, it takes several days to obtain results and is labor intensive. Due to this, the role of electron microscopy has been decreasing steadily with ongoing refinement in IHC. Histochemistry may be performed for neutral and acidic mucins (adenocarcinoma), glycoproteins (adenocarcinoma or hepatocellular carcinoma), neurosecretory granules (neuroendocrine tumors), melanin (primary or metastatic melanoma), and other tumor cell products or associated proteins. But most of these are detected by IHC with better specificity and sensitivity even for detecting some organisms such as Helicobacter pylori in gastric biopsies, thus limiting the role of histochemistry in today’s practice environment. However, histochemistry is still used for some indications such as for detection of various organisms such as fungi (Periodic acid–Schiff for fungus [PAS-F] and Gomori’s methenamine silver [GMS] stain) or acid-fast organisms (various acid-fast bacillus [AFB] stains).

Immunohistochemical Assessment

An increasing number of antibodies that may be applied to FFPE tissue are continuously being added to the ever-expanding spectrum of diagnostic and prognostic immunomarkers. This has facilitated widespread application of immunohistochemistry [28, 29] in routine diagnostic pathology. However, for some lesions, such as lymphomas, there is preference for fresh tissue in isotonic medium for immunolabeling and evaluation by flow cytometry. Although immunophenotyping (either IHC or flow cytometry) is a very powerful tool, it is absolutely essential to understand that it is an ancillary tool and has to be used in the context of a carefully structured differential diagnosis with reference to the clinical details and morphological findings. There are many pitfalls with potential false positivity if this caveat is not taken into consideration. It may be applied for a variety of indications including differential diagnosis of primary site, grading, and increasingly expanding prognostic/therapeutic reasons.

For example, recently, IHC has been made available for evaluation of programmed death ligand 1 (PD-L1) in the tumor cells [30]. Programmed death (PD)-1 (CD279) is a co-inhibitory receptor present on the cell surface of monocytes, T lymphocytes, B lymphocytes, and natural killer cells [31]. It has 2 ligands: PD-L1 (B7-H1) and PD-L2 (B7-DC). Interaction between PD-1 and its ligands down-regulates the T-cell response by inhibiting T-cell receptor signaling. PD-L1 on tumor cells is upregulated. Studies revealed that barricading this interaction with antibodies to PD-1 or PD-L1 reverses this inhibition to regain anti-tumor T-cell activity with therapeutic benefits [31].

Discussing application of IHC in detail is beyond the scope of this chapter [28]. A few immunomarkers applicable to GI cancers are shown in Table 2.5 [16, 17, 28,29,32].

Table 2.5 Application of immunomarkers in gastrointestinal cancers: a few examples
Full size table

Molecular Pathology

The role of molecular tests in GI cancer is continuously increasing. Please refer to the chapter on this topic in this book for more details in addition to other publications on this topic [17, 31,32,35]. Here, it is important to understand some basic details related to these. The molecular tests may be DNA-based or RNA-based. Recently, the role of microRNA (miRNA) is evolving. DNA is very robust and miRNA is relatively stable. In contrast, RNA is quite unstable and requires special precautions and protocols due to ubiquity of RNAase (RNA-destroying enzyme) present in tissue samples and in the devices/steps at different stages of processing. However, currently, many refinements have been achieved in the application of RNA-based molecular tests performed on FFPE [36]. Thus, like IHC, most of the molecular tests could be performed on FFPE, which in general is the most easily available clinical material for performing elective molecular pathology test at any stage on the archived FFPE tissue. Also, it is important to know the proportion of viable tumor component in the FFPE section in comparison with background nontumor nucleated component. Many tests require a minimum fraction of tumor component for optimum results. One should check with the laboratory performing a particular molecular pathology test regarding the minimum tumor proportion required for a specific test in their laboratory. This may be overcome by selectively dissecting out the tumor by various microdissection methodologies. For other molecular pathology tests , there may be specific protocols requiring fresh or frozen tissue or tissue collected in special medium/preservative such as RNAlater® [37]. All of these limitations should be taken into consideration prior to proceeding with any molecular tests on any specimen. The overview for approaching molecular pathology tests on GI cancer specimens is summarized in Fig. 2.4 [17, 28,29,32, 36,37,38,39,40,41,42,45].

Fig. 2.4
figure 4

Approach to evaluate commonly used molecular pathology tests and methodologies applicable to GI cancers (∗See references [17, 28,29,32, 38, 61,62,63,64,65,66,67,70])

Full size image

Classification of Gastrointestinal Tumors

GI cancers have been classified traditionally at two levels: macroscopic and microscopic.

Macroscopic Classification

Ultimately, similar to other cancers, microscopic findings in GI cancers decide the final interpretation and classification. But, the macroscopic gross evaluation including tumor configuration, size, and anatomic site is an important step with extended practical application, especially during endoscopic examination. The tumors of tubular GIT may be classified based on the approach used for gastric tumors, which are generally divided into four types: type I (polypoid), type II (fungating), type III (ulcerated), and type IV (infiltrative, also called linitis plastica) [46]. Some macroscopic features of ulcerated lesions may help to distinguish a benign ulcer from an ulcerated carcinoma (type III). A small, punched-out, well-circumscribed ulcer with a smooth base and edematous regular margin favors a benign gastric ulcer. In comparison, an irregular ulcer with raised, firm borders with necrotic and hemorrhagic base, typically favors a malignant ulcer [47].

Similar to gastric cancer, colorectal cancer (CRC) can also be classified macroscopically [48]:

  1. 1.

    Exophytic tumors : usually large, polypoid lesions (typically in the cecum) are rarely obstructive.

  2. 2.

    Infiltrative ulcerating tumors : ulcer with irregular raised edges.

  3. 3.

    Constricting annular tumors : functionally obstructive lesion with firm consistency due to desmoplasia resulting in proximal dilatation with typical double-contrast “apple-core” sign.

  4. 4.

    Diffuse tumors : similar to linitis plastica of the stomach with infiltrative growth along the bowel wall.

Although macroscopic classification does not have a prognostic significance independent of the histological subtype [49], anatomic site does. Right-sided tumors – located in the cecum, ascending colon, hepatic flexure, or transverse colon – have a better prognosis as compared to left-sided tumors – located in the splenic flexure, descending colon, or sigmoid colon [50]. This may be related to tendency for microsatellite instability (MSI) in the right colon.

With the increasing role of endoscopy, macroscopic classification has evolved to categorize early neoplasia (type 0) of the digestive tract [49,50,53]. This classification distinguishes polypoid/protruded (type 0–I); nonpolypoid/nonprotruded, nonexcavated (type 0–II); and nonpolypoid, and excavated (type 0–III) lesions. Type 0–II lesions are subdivided by the absence (type 0–IIa-elevated and type 0–IIb-flat) or presence (type 0–IIc) of a depression. This morphological macroscopic terminology applies to esophagus, stomach, and colon with increasing clinical relevance in the era of endoscopy [51]. But macroscopic features of GI cancers have limited diagnostic, predictive, and prognostic significance. Absolute dependence of staging on imaging findings without meticulous grossing of resection specimen is discouraged. Generally, malignant tumors are nonencapsulated with irregular infiltrative borders. They are usually large and solid with foci of necrosis/hemorrhages. As standard of practice, microscopic surgical pathology examination with tissue diagnosis is critical for appropriate management.

Microscopic Classification

CAP and other professional bodies have recommended internationally accepted terminology and diagnostic criteria established by the WHO for consistency and uniformity in pathological reporting (Table 2.6) [6, 54].

Table 2.6 Pathological classifications of various GI tumors (WHO 2000) [6]
Full size table

Traditionally, tumor classification is based on type of tissue differentiation and is termed histogenetic classification , which categorizes different tumors with reference to various morphological features including: (1) site of primary tumor, (2) differentiation/histogenesis, (3) architectural phenotype, and (4) degree of differentiation (grade).

  1. 1.

    Site of Primary Tumor : Neoplasms of epithelium may be benign (papillomas/adenomas) or malignant (carcinomas). Similarly, those of connective tissue may be benign (various –omas) or malignant (sarcomas). Although generally there is good concordance between the type of normal tissue and type of neoplasm, some tumors with discordant differentiation may be seen in odd tissues. For example, carcinomas with total (as squamous cell carcinoma) or partial (adenosquamous carcinoma) squamous differentiation may be seen in organs such as the colon, rectum, and pancreas, which normally do not have squamous epithelium.

  2. 2.

    Differentiation/Histogenesis : Carcinomas demonstrating glandular growth pattern are adenocarcinomas versus squamous cell carcinomas with squamous differentiation. Other than in the esophagus and anus (which in a significant proportion in these sites are squamous cell carcinoma), most of the GI carcinomas are adenocarcinomas.

Adenocarcinomas may be subdivided morphologically into various subtypes such as usual type (with glands of variable size, shapes, and maturity in the background of variable proportion of desmoplastic stroma); mucinous type (adenocarcinomas comprising of more than 50% component producing abundant secretory mucin (the term “adenocarcinoma with mucinous differentiation” may be used for tumors with marginal proportion of mucinous component >10% but <50%); signet-ring cell type (adenocarcinomas showing at least 50% signet-ring cells with cytoplasmic mucin vacuole pushing the nucleus).

Benign/malignant neoplasms of connective tissue, adipose tissue, smooth muscle, skeletal muscle, vessels, cartilage, and bone are broadly labeled respectively as fibroma/fibrosarcoma, lipoma/liposarcoma; leiomyoma/leiomyosarcoma; rhabdomyoma/rhabdomyosarcoma, angioma/angiosarcoma; chondroma/chondrosarcoma; and osteoma/osteosarcoma.

The tumors of hematopoietic and lymphoid tissues are leukemias and lymphomas. In the adult population, the majority of malignant neoplasms of tubular GIT are carcinomas followed by lymphomas and sarcomas, which are relatively the predominant tumor in the pediatric population.

  1. 3.

    Architectural Phenotype : Like other tumors, GI tumors may be classified based on growth pattern and microscopic architecture, which also provides important histogenetic clues while interpreting the tumor biopsies or resection specimens. Architectural pattern of epithelial tumors may be tubular (branching tubules of variable sizes); papillary (finger-like projections with fibrovascular central cores); solid or trabecular (seen in medullary carcinoma of the colon, neuroendocrine tumors, and hepatocellular carcinoma). Some tumors may show a cystic pattern (seen in the pancreas but relatively uncommon in tubular GI tumors, as mucinous carcinomas and endothelial tumors such as lymphangioma or hemangioma). Even solid tumors including stromal tumors/sarcomas, lymphomas, and carcinomas with central necrosis may present as cystic lesions, especially at imaging level. However, in general, growth pattern of GI tumors has little prognostic significance [55, 56]. Recently, a polyp with serrated glandular architecture has been linked as a precursor lesion for colorectal carcinomas [57].

A few examples suggesting applications of growth patterns for tumor classifications include Lauren classification , categorizing gastric cancers into different types: Intestinal, diffuse, mixed, and indeterminate/unclassified [58], in which diffuse growth pattern with highly unfavorable prognosis has macroscopic linitis plastica appearance with signet-ring cells at the microscopic level [59]. Colon cancers with tumor budding in the form of single cells or groups of less than 4 tumor cells at the invasive margin have worse prognosis and are associated with a diffuse growth pattern [58,59,60,61,62,65].

  1. 4.

    Degree of Differentiation (Grade): Tumor grade reflects the biological properties of the tumor. In general high-grade tumors are associated with aggressive biological behavior. The clinical significance of grading may be different for each tumor category. As an example, carcinomas or sarcomas with lower grade may be biologically less aggressive and amenable to surgical excision as compared to higher grade counterparts. On the other hand, low-grade lymphomas, although more indolent and slow growing than high-grade lymphomas, are difficult to be cured by medical therapy.

Although there are various approaches in grading tumors, the most commonly applied is the degree of resemblance of the tumor morphology to its non-neoplastic counterpart. Several microscopic features are taken into consideration for grading a tumor, including the anatomic site of origin of the tumor, the class of the tumor (i.e., carcinoma, sarcoma, or lymphoma), and the histological subtype within the class. The simplest approach applied for grading includes degree of gland formation in adenocarcinomas versus degree of keratinization in squamous cell carcinomas [56]. Most grading systems assign the grade based on the most poorly differentiated area. Some consider average of grades in different areas of the tumor. Arbitrarily most pathologists grade GI cancers into 4 grades: Well differentiated (grade 1), moderately differentiated (grade 2), poorly differentiated (grade 3), and undifferentiated (grade 4). Due to this subjective judgment left to the individual observer, there may not be reproducible outcome with significant degree of interobserver variability [66]. Despite these limitations, grading has some prognostic significance in most gastrointestinal malignancies [55, 56]. In addition, if the grade of the primary tumor is known, it may help while evaluating the interpretation of metastases at later stage during comparative review.

The CAP-suggested grading system is based on a semiquantitative approach for improved reproducibility and considers the proportion of neoplastic glands in the tumor: grade X (grade cannot be assessed); grade 1 (well differentiated) – more than 95% glands; grade 2 (moderately differentiated) – 50–95% glands; grade 3 (poorly differentiated) – 5–49% glands; and grade 4 (undifferentiated) – fewer than 5% glands [56]. Further simplification of this grading system has suggested a 2-tiered system for improved reproducibility [49]. Higher grade tumors demonstrate adverse prognosis independent of the stage. However, some poorly differentiated colorectal adenocarcinomas, such as those with MSI, may have better prognoses [67]. This simple approach has to be modified for some subtypes of carcinoma (e.g., medullary carcinoma of the colon is left ungraded; signet-ring carcinoma is defined as poorly differentiated or high-grade).

Other tumors including neuroendocrine tumors, sarcomas, and lymphomas have a special grading system based on different parameters such as proliferation index (mitotic figures or Ki-67 index estimation), necrosis, and other features.

Staging of Malignant Gastrointestinal Tumors

Staging is one of the best but simplest time-tested approaches for stratifying malignant neoplasms for prognostic grouping and is very important for planning the therapeutic management of the case. A staging system based on TNM classification standardized by the AJCC and UICC is recommended by CAP [12, 13, 15]. It has been used all over North America by national, regional, and local tumor registries and is also accepted internationally.

General Principles of the TNM Staging

TNM staging is based on classification and grouping of: “T” for the primary tumor status, “N” for regional lymph node status, and “M” for distant metastatic disease status (Table 2.7) [15, 68]. Final AJCC stage is assigned progressively from stage I through stage IV based on various combinations of staging in each category standardized for the individual organ system (see TNM staging of colon cancer as example in Table 2.1, Fig. 2.5) [15]. Lymphoma has a special staging system without applying the TNM approach for most lymphomas, except some types such as primary cutaneous lymphoma [15]. Although in general AJCC staging criteria are practiced, some ongoing approaches continue to evolve and claim better prognostic correlation [69].

Table 2.7 TNM staging: general guidelines [15]
Full size table
Fig. 2.5
figure 5

T staging of colon carcinoma as example (see Table 2.1) [13]

Full size image

More features are added to include other details: prefix “p” refers to the pathological classification; prefix “c” for the clinical classification. Prefix “r” is used for recurrent tumors following curative therapy (subject to the documentation of disease free interval) (Table 2.8) [12, 13, 15].

Table 2.8 Staging classifications/designator rules [15]
Full size table

“R” classification is for residual tumor after primary therapy (e.g., curative surgical resection):

  • R0 – negative for residual disease after definitive therapy (after curative surgical resection or total remission without detectable residual tumor)

  • R1– residual tumor with microscopically positive resection margin

  • R2 – residual tumor with macroscopically positive resection margin

R classification is not usually followed by most institutions; instead, the report includes information on resection margins .

T Category

T staging (Table 2.9) for tubular GI cancer is assigned based on the depth of invasion of the primary tumor into various layers with incremental status as it invades from superficial to deeper layers (Tables 2.1 and 2.2, Fig. 2.5) [15]. For some tumors, for example liver tumors, it is based on other features such as size, vascular invasion, and multifocality.

Table 2.9 Summary of TNM classification rules based on the AJCC Cancer Staging Manual, Eighth Edition (2017) [15]
Full size table

Carcinoma in situ (pTis) includes intraepithelial carcinoma (when malignant cells are still restricted superficial to the basement membrane and have not invaded beyond it) and intramucosal carcinoma (in which tumor cells invade lamina propria without invading muscularis mucosa into submucosa). However, use of these terminologies may be confusing if applied randomly. In the colon, both intraepithelial carcinoma and intramucosal carcinoma are equivalent and have been used interchangeably.

Tumor invading an adjacent organ in contiguity (e.g., colonic carcinoma invading liver or even other segment of tubular GIT) is part of T staging and is not distant metastasis [15]. Similarly, sideways horizontal spread of tumor to the adjacent segment of tubular GIT (e.g., cecal carcinoma spreading along the lumen to adjacent ascending colon and/or adjacent terminal ileum) is also part of pT staging and not distant metastasis [15]. On the other hand, penetration of tumor through a lymph node capsule into a regional lymph node is considered nodal metastasis for N staging.

For multiple primary tumors of tubular GIT, T stage is assigned as per the highest category. However, multiplicity of tumor assigns it a specific T stage in the liver [15].

T staging for some special tumors such as GIST and NET have a special approach. It is based on the size of the tumor in GIST [15] and on the extent of invasion with tumor size in NET (Table 2.2) [15].

N Category

N staging is assigned based on status of regional lymph nodes evaluated conventionally by examining HE-stained sections (Table 2.9) [15]. If lymph nodes are grossly positive, only a representative section is submitted for confirmation. However, grossly negative or equivocal lymph nodes are submitted entirely [49]. The number of lymph nodes that could be evaluated from any resection specimen depends on a variety of factors including anatomic nature of specimen, the length of the resected segment, type of surgical procedure, chemo/radiation therapy status prior to resection, and/or technical skill/diligence on the part of the dissector grossing the specimen. The number of lymph nodes sampled from a node-negative colorectal cancer specimen has been suggested to be at least 12 lymph nodes [70, 71]. At least 1 positive or negative lymph node is needed for assigning pathological N (pN) staging.

Discontinuous spread or tumor deposits (TD) in subserosa, mesentery, and nonperitonealized pericolic or perirectal tissues, although not nodal metastases, are considered under N category. These should be distinguished from totally replaced lymph nodes (which are counted as lymph nodes) or venous invasion with extravascular spread (considered as V1/V2).

Positivity of nonregional lymph nodes for tumor is considered distant metastasis and is not part of pN staging, but belongs to pM staging [12, 13, 15].

M Category

Metastasis to any distant organ or tissue including any nonregional lymph node is considered for M staging (Table 2.9) [15]. Presence of isolated tumor cells in the bone marrow, peritoneal seeding, and positive serous fluid cytology are also considered metastases [15].

Satellite lesions (skip lesions) present as multiple tumor foci in adjacent bowel along the mucosa or submucosa are not distant metastases [15]. These must be distinguished from synchronous primary tumors.

Additional Features

There are a few additional features (Table 2.8 [15]) that should be communicated in the final surgical pathology report of excised GI cancer specimens (Table 2.10) [10, 12, 13, 15, 49, 55, 56, 58,59,60,61,62,65, 70,71,72,]. Although these features are not reported specifically as an individual category, currently they are a routine part of the CAP cancer protocol in the final pathology report (see colon cancer CAP protocol as example in Table 2.11) [10].

Table 2.10 Additional features to be communicated in final surgical pathology report of excision specimens [10]
Full size table
Table 2.11 Recommended reporting protocol of various resections of cancersa using colon/rectal cancer as an example standardized by the College of American Pathologists (CAP) [10]
Full size table

Conclusion

Morphological evaluation with ancillary tests such as immunophenotyping and histochemistry is, and will continue to be, the most critical pivotal component in the management of GI cancers. The current advances in molecular pathology have increased its role and have become an integral part of management in addition to conventional AJCC staging [10].

In future, increasing insight into the molecular biology of all GI cancers including overexpression and/or repression of various genes as well as epigenetic changes would establish a better understanding with ongoing advances in achieving improved tumor classification, diagnosis, prognosis, and targeted personalized therapies [17, 31,32,35]. Generally, both conventional pathological examination and new molecular tests are required for proper evaluation of any GI cancer for diagnostic and therapeutic decisions. Application of any new biomarkers cannot be justified until the findings demonstrate a convincing positive impact on clinical management. The ongoing advances would improve the understanding in molecular biology of various GI cancers and develop treatment algorithms with targeted therapies tailored for individual patient care as personalized medicine evolves [75].