Knowledge of indication of procedure

While it goes without saying that the prosector should be aware of the indication for which the excision/resection was performed, it is not uncommon to have incomplete or even misleading information on requisitions. Requisitions are often completed by ancillary staff who may not be completely aware of the clinical scenario and/or may be merely reproducing partial information from the patient’s history. Clinical information is equally important for benign specimens whose pathology may not be obvious. One such example is resection for fistulae. After resection, the fistula or perforation may collapse and may not be instantly apparent on gross examination, especially after fixation. Hence, being cognizant of the indication of the procedure allows the prosector to actively evaluate the disease process and ink/mark the area and submit sections most pertinent to the clinical indication.

Importance of reviewing operative note

Similarly reviewing the operative note before sectioning the specimen helps the prosector to identify and account for additional structures that are typically not present in that surgical procedure but may be adherent or attached to the specimen. For example, a small portion of the bladder wall or seminal vesicle may adhere to a rectosigmoid colon resection for cancer. Unless indicated by the surgeon or aware of that by reviewing the operative note, this piece of tissue can escape the prosector’s attention. Demonstration of direct invasion of another organ by colorectal tumour would potentially upstage the tumour to a pT4b tumour (figure 1A).1

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

(A) Low anterior resection from rectal cancer showing that seminal vesicles (highlighted by arrow) are densely adhered to the resection specimen, raising a gross suspicion of direct extension/invasion of colorectal cancer to seminal vesicles (potentially upstaging the patient to pT4b). These areas should be carefully noted and submitted for histological evaluation; (B): Gross evaluation of endoscopic mucosal resection from a case of Barrett’s neoplasia highlighting that the local excision specimen resection specimen should ideally be pinned out right after the procedure. After adequate fixation, serial sections are to be obtained at 2–3 mm intervals perpendicular to the long axis (highlighted by red lines); (C) Histological sections highlighting an intramucosal adenocarcinoma in this properly oriented section. The black ink shows the cauterised base. The lateral mucosal edges are negative for dysplasia or carcinoma. Adequate fixation and proper sectioning make the assessment of depth of invasion and margins easier and more accurate (H&E ×40); (D): Partial colectomy specimen highlighting an endoscopically unresectable polyp with a focus grossly suspicious for submucosal invasion (highlighted by arrow). Inset highlights a perpendicular section taken through this grossly suspicious focus depicting a submucosally invasive adenocarcinoma (H&E ×100).

Documentation of number of pieces

Documentation of the number of pieces with dimensions is of importance as well. This information comes in handy while comparing the notes with the interventional teams when there is discordance or in the unusual scenario of specimen mix-up. Additionally, the specimen container should always be evaluated after removal of the main specimen for any small pieces of tissue which may have been detached or removed separately but could be of diagnostic importance.

Adequate fixation and optimal sectioning

Resection specimens should be fixed adequately (typically 12–24 hours, depending on the specimen size) to enable proper tissue slicing and gross evaluation. A general recommendation is that when using 10% neutral buffered formalin, specimens should be fixed for at least 6 hours and no more than 72 hours.2 An optimal fixation duration is crucial for adequate tissue processing and staining.3 Inadequately fixed tissue can be difficult to slice, and autolytic changes can obscure the pathological findings. Additionally, sections taken from improperly fixed specimens can be suboptimal, leading to diagnostic challenges when looking at the morphology of tumour on slides or for margin assessment and when assessing the relationship between the lesion of interest and surrounding tissue. Ancillary stains may also not work properly if the tissue is not adequately fixed. To ensure proper fixation of the tissue/tumour, the tissue should be sectioned into thin slices. The permeation speed of 10% neutral buffered formalin is 1 mm/hour.3 4 For tubular GI specimens, the lumen should be opened at the time of fixation and the specimen pinned on a paraffin board. For bulky tumours, the mass may be sliced to enable penetration of formalin. If, for some reason, it is undesirable to cut through the tumour and open the lumen, the gut adjacent to the tumour may be opened, and a formalin-soaked gauze can be inserted through the lumen to enable fixation. The sections taken should not be more than 3–5 mm in thickness and should easily be able to fit between the top and bottom of the processing cassette.2 3

Small therapeutic resections (endoscopic mucosal resection and endoscopic submucosal dissection)

Small, early-stage lesions are now amenable to conservative yet complete excision using advanced endoscopic techniques. The type of procedure undertaken depends on the type of lesion, location and expertise of the gastroenterologist. Hot snare polypectomy is the procedure of choice for pedunculated colonic polyps more than 10 mm.5 Endoscopic mucosal resection (EMR) may be considered for non-pedunculated polyps more than 10 mm and is the method of choice for polyps/lesions more than 20 mm. For larger polyps with a higher likelihood of submucosal invasion, an endoscopic submucosal dissection (ESD) is preferable.5 Similarly, for more advanced lesions associated with Barrett’s oesophagus (eg, size>1.0–1.5 cm, large bulky lesions with nodularity, suspected possible minimal submucosal invasion, prior biopsy with high-grade histology, prior EMR with positive margins, recurrent disease) and all gastric lesions, ESD is the preferred modality.6 7 EMRs are limited to the mucosa and muscularis mucosae, while ESDs also have a portion of superficial submucosa. The advantage of the ESD procedure includes the possibility of en bloc resection of larger lesions, higher likelihood of achieving complete (R0) resection and a lower recurrence rate. Regarding macroscopic evaluation and handling, EMR and ESD specimens are almost handled similarly.7 Of note, measurement of these specimens is best performed before fixation. The ideal recommended fixation time for EMR is 12–72 hours but no less than 2 hours.6 7 Similar protocol can also be followed for specimens from transanal excision, transanal endoscopic microsurgery or transanal minimally invasive microsurgery of early rectal lesions.8 9

The local excision specimens in the fresh state should ideally be pinned out on a small board with a hard surface (cork, paraffin or polystyrene) with the mucosal surface up, to prevent rolling of the specimen and its shrinkage. This can be performed right after the procedure in the endoscopy suite itself or the specimen can be sent fresh to the gross room immediately after the procedure and be handled and pinned in the same manner in the gross room.7 A specimen sent in saline can also be pinned in the gross room before fixation. Endoscopic mucosal rection specimens are often received piecemeal, and in these cases, an attempt should be made to pin the larger fragments before fixation and grossing. Overstretching can result in tears, and hence polypoid fragments should be loosely stretched and pinned. Proper pinning allows for appropriate orientation and sectioning, and hence optimal evaluation of margins.7 The deep and peripheral margins of the specimen should be inked. If the mucosal aspect has been oriented (proximal vs distal), mucosal margins can be differentially inked. In some centres, piecemeal EMR specimens or even ESD specimens maybe sent fixed in formalin. This can make measurements, inking and subsequent pinning of larger fragments very challenging. Hence, it is worth having a discussion with gastroenterologists regarding sending the samples fresh or in saline immediately or pinning them in the endoscopy suite properly before placement in formalin.

Often a definitive lesion may not be visible, but any mucosal abnormality should be noted, and the distance from the margins should be documented. Once the specimen is fixed adequately, serial sections should be obtained at 2–3 mm intervals perpendicular to the long axis (figure 1B,C). The tips or ends can be submitted en face or perpendicularly submitted depending on the distance of margin from the lesion, the preference of the pathologist and institutional practice.7 10 Tissue embedding is a critical step for optimal sections, and hence, all sections should be embedded ‘en face’ or on edge.7

Alternative techniques, such as use of Captivator device, may be used to ensure proper fixation of endoscopic resection specimens. The captivator device includes a cassette into which the mucosal resection specimen can be directly placed and flattened at the time of fixation due to pressure from the lid.7 Similarly, in some centres the specimen may be directly fixed without pinning out. A study comparing the three methods showed that specimens handled by pinning method were most effective for final histological interpretation, followed by the non-pinning method and the Captivator device, respectively.6

Another therapeutic small resection is a polypectomy specimen for pedunculated polyps. The stalk margin should be inked, and a perpendicular section should be taken through the polyp head and stalk.9 This is important to show the relationship of any potential invasive carcinoma arising in the polyp to the polypectomy margin and assess submucosal invasion by the carcinoma. Any additional friable pieces of the polyp should be submitted separately in a different cassette. A similar approach (perpendicular and well-oriented full-thickness sections of the polyp) should be taken for a colectomy specimen obtained for an endoscopically unresectable polyp (figure 1D inset). If there is gross evidence of submucosal or deeper invasion, representative full-thickness sections from the deepest extent of invasion may be submitted. However, if there is no gross evidence of submucosal invasion, the entire polyp should be submitted for histological evaluation to look for microscopic invasion.

Large resections for masses

One of the most common specimens received on GI service is oncological resections for masses. Often, there is a prior biopsy diagnosis available, and the specimen can be adequately grossed per the staging requirements for the given site. The staging parameters across all GI sites (except the anus and squamous cell carcinoma of the oesophagus) are similar, making it easier for prosecutors to follow standard procedures.

Margins

It is a standard practice to evaluate the mucosal margins (proximal and distal) and radial, circumferential or mesenteric resection margins when analysing an oncological GI resection. While it has been shown in multiple studies that evaluation of mucosal margins and anastomotic donuts is of limited value, especially if the tumour is >2 cm from the margin,11–13 in rare instances, tumour may extend submucosally or tumour deposits may be present at the margin. Additionally, in tumours that have been treated with neoadjuvant therapy or in specimens harbouring diffuse-type gastric carcinoma, the gross inspection can miss subtle involvement of margin by tumour. Second, the presence of dysplasia at the mucosal margins may also be significant for subsequent management, depending on the grade and underlying disease process. Typically, mucosal margins are taken en face, if the tumour is grossly at a reasonable distance from the margin. However, if one of the mucosal margins is close to the tumour(≤2 cm), then a perpendicular section is preferable so that the relation between the tumour to the margin and the exact distance between the two can be documented. This is important in low anterior resection specimens, where clearance of >2 cm is ideal but even 1 cm may be sufficient, especially in T1 and T2 tumours.1 14 15 Additionally, in most instances, distal intratumoural spread is limited to 2 cm, in both treated and treatment naive cases.14 16 17 A perpendicular section in such cases will be helpful to document the distance between the most distal tumour focus and the distal margin, even if the mucosal disease is further away from the margin. Tumours involving non-peritonealised or partially peritonealised parts of the GI tract such as oesophagus, ascending colon, descending colon and distal rectum/anus will have a radial resection margin or circumferential soft tissue resection margin. For GI segments that are completely encased by peritoneum (ie, stomach, small intestine, cecum, transverse colon, sigmoid colon), the mesenteric resection margin is the only pertinent ‘radial’ margin.1 Gross measurement of the distance from the point of deepest penetration the tumour to the closest radial/circumferential or mesenteric resection margin should be documented before taking a section of the same. If the tumour is grossly >1 cm for this margin, we usually take an ‘en face’ section; however, we recommend taking a perpendicular section with inked margin for closer tumours to document the exact distance.

Sampling of the tumour

In addition to margins, multiple full-thickness tumour sections to document the extent of deepest invasion, and relationship to the serosa are important for staging purposes. In cases where additional organs such as the uterus, cervix, prostate and bladder are present along with the colorectal resection specimen (pelvic exenteration), documentation of the relationship of the tumour to the adjacent organs and sections to determine the direct involvement of the adjacent organ by the tumour are necessary for staging. Special attention should be paid to the serosa close to the tumour. Serosal involvement may be evidenced by puckered, irregular or sclerotic areas on the serosa, or by creeping fat, fibrinous adhesions, and plaques on the serosal surface overlying the tumour (figure 2A).18 Sometimes these findings can be very subtle and easily missed, if not inspected carefully. Involvement of the serosa upstages many luminal GI tract tumours to pT4a stage, and the patient may qualify for adjuvant therapy (figure 2B).18 Thorough and astute inspection of the serosa both in fresh and fixed state is therefore a key step in staging of GI tumours. Such areas can be inked before sampling for easy correlation during microscopy and should be carefully sampled.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

(A, B) Gross examination from a colectomy specimen highlighting a focus of serosal puckering (highlighted by arrow). The histological section from this focus shows a black-inked serosal cleft with tumour involving the serosal surface (staged as pT4a). (H&E ×100); (C): Gross examination of a low anterior resection specimen for rectal cancer highlighting a complete total mesorectal excision (characterised by smooth, regular circumferential resection margin, with no coning, and intact and smooth mesorectum); (D): Partial small intestine resection specimen with neuroendocrine tumour and associated large mesenteric mass (long, thin narrow), measuring more than 2 cm (upstaging the patient to pN2). Additionally, an area of serosal nodule (short, thick arrow), representing either a small satellite nodule or possible serosal invasion by overlying neuroendocrine tumour (pT4a) is seen. Adequate sectioning would help make this distinction.

For colonic tumours, it is usually recommended to take at least two blocks from deepest extent of tumour to closest peritoneal surface, and at least additional one block from deepest extent of tumour to adjacent pericolonic fat.18 In our clinical practice, we recommend taking multiple full-thickness sections of the luminal GI tract tumour where possible, with adequate representation of the tumour relationship with adjacent tissue. We would take multiple additional sections and deeper levels if we find a deeply invasive tumour close to the serosal surface on initial histological sections or if we see gross serosal abnormality but no histological correlate of serosal invasion in initial sections submitted or their deeper levels.

Lymph nodes

During gross examination of oncological resection for tumours, all grossly negative lymph nodes must be submitted for microscopic evaluation. Larger lymph nodes should be either bisected or even serially sectioned before submission. Lymph nodes should be differentially inked prior to sectioning if multiple nodes are being submitted in the same cassette. Unequivocal grossly positive lymph nodes may be partially submitted. For colorectal tumours, a minimum requirement of 12 regional lymph nodes in the surgical specimen is a quality control measure for the surgical resection and is associated with survival benefit.19 This number has also shown some prognostic benefit in patients with stages II and III colon cancer.20 However, it is important for prosecutors to note that 12 is the minimum requirement and there is no maximum. It is prudent to submit all the lymph nodes in an attempt to capture any nodal metastases present. In postneoadjuvant therapy specimens, lymph nodes may be difficult to identify and reevaluation of the specimen to identify additional smaller lymph nodes may be necessary.21 22 Various fat clearing techniques (such as acetic acid, xylene and acetone/alcohol) and special fixatives (such as GEWF solution) have been tried to increase the yield of lymph nodes.23 24 While the dissolution of fat might help increase the yield, the additional time, effort and usage of laboratory resources may not be able to justify the routine use of these special techniques.25 We recommend carefully re-examining colorectal specimens in cases where lymph node yield is <12 on initial gross evaluation. Any fibrous-looking areas may be submitted for microscopic evaluation during this second look. In these cases, we often also submit multiple sections of pericolic/perirectal fat to retrieve minute lymph nodes that may have been missed by gross examination to reach this minimum recommendation. Studies have shown that ‘second look’ at the specimen to reach a benchmark of 12 lymph nodes may result in pathological upstaging (pN) in a subset of cases.26 27 In one study, a second attempt at lymph node dissection increased the mean lymph node count from 8.3±7.5 on initial attempt to 14.6±8.0 following submission of additional sections. The number of cases meeting the target of 12 lymph nodes increased from 14 to 69, and examination of the additional lymph nodes resulted in pathological upstaging (pN) of 5% (5 of 99) cases in this study.26 In an another study, a second search increased retrieval of lymph node count by an average of 10 additional nodes. Additionally, in 4.4% (5 of 114) cases, pathological node negative (N0) disease after the first search converted to node positive (N+) after a second search that yielded 1– 4 positive nodes.27 Some experts even suggest not reporting any colorectal cancer resection specimens with <12 lymph nodes unless all the pericolorectal soft tissue has been submitted for histological evaluation.18

Rectum: total mesorectal excision and treated rectal tumours

Total mesorectal excision (TME) refers to complete removal of the soft tissue envelope around the rectum by dissection along the visceral plane. Total or complete mesorectal excision ensures complete removal of the perirectal lymph nodes and is an indicator of the quality of the surgery as well as enables adequate examination of the circumferential resection margin. TME has been associated with reduced local recurrence and has become a standard of care for rectal cancer surgery.28 The quality and completeness of mesorectal resection can be graded as complete, nearly complete and incomplete based on the criteria outlined in table 1.28 Gross pathological assessment of the completeness of the mesorectum is shown to accurately predict both local recurrence and distant metastasis29 and has been included as a reporting parameter in the College of American Pathology (CAP) guidelines (figure 2C).1 Not surprisingly, pathologists and especially prosecutors play an important role in evaluating TME.30 However, receiving an incomplete mesorectum (some causes include difficult anatomy secondary to prior therapy or adhesions or due to the nature of the disease or obese patient) and documenting that in the pathology report can initiate a challenging conservation with the surgeon and at tumour boards. It may be the best for the pathologist to discuss this with the surgeon before releasing the pathology report and starting the conversation by acknowledging the inherent difficulty of resection in some cases and understanding how this may been such a case. Taking gross photographs in all these cases for documentation and discussion with the surgeon before signing out the report are recommended.

In neoadjuvant treated rectal cancers, the tumour may not be grossly visible and may be replaced by a scar, representing the tumour bed. In such cases, it is advisable to document the size and depth of the tumour bed and submit full-thickness sections of the tumour bed. In cases of only a small (≤3 cm) residual abnormality, it is recommended to submit the entire lesional area as full thickness sections.31 For larger tumours, it is recommended to have at least one section submitted for every centimetre of tumour diameter. However, more sections (and if needed, the entire lesional tissue) should be submitted, if the initial sections show no residual tumour on histological evaluation. Unlike treated pancreatic adenocarcinoma cases where the recommendation is to perform consecutive mapping sections along the largest tumour dimension or whole mount sections of the pancreas in an attempt to validate the gross measurement of the tumour size by histology,32 similar recommendation has not been advocated in treated rectal cancers yet. One of the reasons may be that unlike pancreatic cancer where pathological T stage is based on tumour size, the pathological T staging in colorectal cancer is based on the depth of invasion (and not the tumour size). Of note, tumour may be seen up to 1 cm from the edges of gross mucosal irregularity, and hence, it is also recommended to sample the edge of the residual lesion. Of note, although lateral tumour spread under normal mucosa that is adjacent to a residual mucosal abnormality is not uncommon in treated rectal cancers, the epicentre for maximum depth of invasion has usually been shown to be under the gross focus of residual mucosal abnormality in almost all cases33; hence judicious sampling the edges of the mucosal abnormality should suffice in most cases, and there is presently no recommendation or justification to sample the entire segment. Identifying any residual tumour is important for accurate staging and prognostication. Pathologial complete response of the tumour to neoadjuvant therapy is strongly predictive of improved outcomes and has shown to be associated with improved local recurrence and distant metastasis-free, disease-free and overall survival.34 For a tumour to be considered to have a complete pathological response (ypT0), the entire residual mucosal abnormality must be submitted and evaluated.

Small intestine

Small intestine is a frequent primary suspect site for metastatic neuroendocrine tumours (NET) of unknown origin.35 Some patients may present with mesenteric masses without a diagnosed primary in the small bowel. In such cases, the small bowel mucosa should be carefully inspected for any visible nodules or lesions, and the specimen should be carefully sectioned to identify small submucosal nodules. The size and extent of invasion of the intestinal wall should be carefully recorded for every bowel lesion, as this is important for staging. NETs≤1 cm and limited to the submucosa are staged as pT1 tumours.36 Also, the size of mesenteric masses should be carefully documented as masses >2 cm upgrades the N-stage to pN2 in this setting (figure 2D).37 Mesenteric masses are often associated with dense fibrosis, which can lead to the encasement of mesenteric vessels. Existing data support that the presence of mesenteric masses has also been associated with frequent liver metastasis and a poor prognosis.38 Recent data also suggest that mesenteric tumour deposit multifocality is associated with shorter disease-specific survival and should be recorded and incorporated into future staging criteria.37

Appendix

The appendix is a unique GI tract organ and home to low-grade mucinous neoplasms. For accurate diagnosis and staging of these neoplasms, careful gross observation of the external surface of the appendix is necessary (figure 3A). The presence of any mucin or disruption/perforation must be documented in the gross description and preferably inked for easy identification. It is common to see a grossly dilated appendix with a distended lumen filled with mucinous material and a thin, often calcified appendiceal wall. The term mucocele is discouraged in pathology reports as it can encompass a range of benign and neoplastic conditions and may be misleading to the treating physicians and patients. In such cases, if a grossly invasive neoplasm may not be apparent, it is prudent to submit the entire appendix for microscopic evaluation. It is important to submit well-oriented sections as the staging depends on the extent of involvement of the appendiceal wall by tumour or mucin. It is also important to clean the blade frequently. While true serosal mucin can, for the most part, be easily distinguished from knife carryover of mucin on the serosa, in some cases, it may be confusing and critical. The presence of true mucin, although acellular, would upstage the neoplasm to a pT4a. Documentation of the relationship of the appendix with other structures is also critical for sampling and staging. Any foci of adhesion between the appendix and adjacent small bowel or colon must be documented and submitted for histological evaluation. Direct involvement of an adjacent organ by tumour or mucin would upstage that neoplasm to pT4b. Separate implants (both acellular and cellular) on the peritoneum, surface of the small bowel or colon should also be documented and sampled, as they would upstage the tumour to M1 disease (pM1a: Intraperitoneal acellular mucin, without identifiable tumour cells in the peritoneal mucinous deposits; pM1b: intraperitoneal metastasis only, including cellular peritoneal mucinous deposits; pM1c: microscopic confirmation of metastasis to sites other than peritoneum).39

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

(A) Gross examination of appendiceal mucinous neoplasm with serosal disruption and extra appendiceal mucin (highlighted by arrow). This would make this tumour at least pT4a in pathological staging; (B) A case of hereditary diffuse gastric cancer with no abnormalities seen on gross examination of the stomach. An exhaustive sectioning (with some experts recommending complete submission of stomach) shows a focus of intramucosal signet ring cell carcinoma (arrow) in this case (B inset; H&E ×100). (C) Gross examination from a case of treated oesophageal/gastro-oesophageal junction carcinoma. No discrete mass is seen; however, an area of ulcerated/puckered mucosa (tumour bed, black arrow) is seen. The tumour bed was entirely submitted; (D) shows a focus of residual submucosally invasive adenocarcinoma (H&E ×100); (E): Gross examination of the total colectomy specimen for idiopathic inflammatory bowel disease. The left upper field shows an area of polypoid lesion, histological sections showed polypoid low-grade dysplasia (not shown here). However, adjacent to the polypoid lesion is a subtle area of flattened mucosa with some mucin and possible mucosal irregularity (highlighted by arrow); (F): Histological sections from this focus showed variant (hypermucinous) dysplasia (H&E ×100).

While grossing routine appendectomy specimens, the tip of the appendix should always be bisected and submitted due to high incidence of incidental well-differentiated NET.40 These tumours are usually small and early stage but could possibly need further workup or treatment depending on additional prognostic factors such as depth of invasion, grade and lymphovascular invasion.41 Similarly, any foci of appendiceal wall thickening should also be submitted to rule out incidental neoplasms such as goblet cell adenocarcinoma which often do not form grossly visible masses. If a neoplastic process is identified on initial sections of the appendix, it is prudent to submit the entire appendectomy specimen for accurate staging of the tumour. For the same reason, it is important to submit the appendectomy proximal resection margin in every case and an additional mesoappendiceal margin where a neoplasm is suspected.

Stomach

Poorly cohesive (diffuse type) gastric cancer may not manifest as a grossly visible mass lesion in a gastrectomy specimen. In contrast, they might have diffuse wall thickening or thickened gastric folds on inspection. In such cases, it may not be possible to give a definitive size of the tumour but an attempt should be made to demarcate and document the area of thickening and its relationship to the margins. If the area of thickening is small, it can be entirely sampled for microscopic evaluation. However, large areas can be representatively submitted with at least one section per centimetre. Any foci of serosal adhesions should also be sampled. Submitting the entire peripheral margins (proximal and distal) in these cases is also recommended, as microscopic submucosal involvement is very common.

Another relatively rare but unique setting is a prophylactic total gastrectomy specimen for hereditary diffuse-type gastric cancer with underlying germline mutations such as CDH1 gene, as these cases often don’t show any grossly identifiable lesional tissue (figure 3B). In such cases, it is important to be aware of the institutional/departmental policy for gross evaluation of such prophylactic resections. Many institutions favour submitting the entire specimen for evaluation with gastric mapping in the absence of a grossly visible lesion. However, some believe that complete submission and pathological examination of the entire stomach provides little additional value for routine clinical management.42 Depending on the availability of resources, three levels of pathological examination are suggested. Level 1 is the minimum sampling of a grossly normal gastrectomy in this setting to obtain sufficient results, and in addition to a photograph, includes a sampling of proximal and distal margins (to confirm all the gastric mucosa has been resected and there is no tumour at the margins), entire lymph node sampling, and mapped sampling from all different zones of the stomach. Level 2 represents a compromise between clinical reporting and preserving tissue for future research (ie, embed all mucosa, process to paraffin blocks, cut a subset of blocks, sample all three gastric zones and examine sampled slides.). Level 3 is total gastric embedding and mapping. It is also recommended that if the entire stomach is not submitted for histological evaluation and no foci of cancer are found in the sections submitted, the gastrectomy specimen be reported as ‘no carcinoma found in xx% of the mucosa examined’.43

Oesophagus/gastro-oesophageal junction

Careful inspection of treated oesophagectomy specimen needs to be done for evaluation of any residual mass lesion or abnormal area (tumour bed). Existing data suggest that patients reported to have complete pathological response in which the entire tumour bed was submitted for histological evaluation have significantly improved survival compared with patients reported to have complete tumour regression but with an incomplete sampling of the tumour bed.44 Hence, it is recommended to submit the entire area of the tumour bed (including any abnormal appearing mucosa) for histological evaluation in such cases to evaluate for any possible residual tumour (figure 3C,D).31