Article Text

Download PDFPDF

Histopathological reporting of pT4 tumour stage in colorectal carcinomas: dotting the ‘i’s and crossing the ‘t’s
  1. R Swamy
  1. Correspondence to Dr R Swamy, Department of Histopathology, Queen Elizabeth Hospital II, Welwyn Garden City, Hertfordshire AL7 4HQ, UK; rajiv.swamy{at}


A histological diagnosis of pT4 stage is used as the basis for offering adjuvant therapy to patients with colorectal cancer. Histological features diagnostic of pT4 stage in colorectal cancer include: (1) presence of tumour perforation, (2) invasion of an adjacent organ, (3) direct or discontinuous tumour extending to the non-peritonealised resection margin, and (4) tumour breaching the visceral peritoneum. Histological parameters to identify these features are not always entirely straightforward, making recognition of pT4 stage difficult at times. The presence of a cellular mesothelial reaction and/or inflammation extending to the serosal surface causes considerable confusion and may result in upstaging pT3 tumours. Conversely, involvement of the non-peritonealised margin in colonic tumours other than rectal cancers is at a risk of being understaged. Audits have shown that significant differences remain in the frequency with which serosal involvement is found between individual pathologists and multidisciplinary teams. The recognition of the criteria that define pT4a and pT4b subsets in pT4 stage to highlight different therapeutic connotations is not universally accepted, and this is likely to affect future reporting of the subsets. The review article highlights the contentious issues in histopathological recognition of pT4 tumour stage based on microscopic parameters in colorectal cancer.

  • Mesothelial/serosal surface
  • non-peritonealised resection margin
  • pT4a
  • pT4b
  • pT4 tumour stage pTNM

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


pTNM (pathological tumour, nodes and metastasis) tumour staging is an integral part of cancer reporting in histopathology practice. The pathological staging of tumour extent (pT) is an assessment of locoregional parameters based on tumour size and local invasion mostly.1 2 In colorectal cancer, the pT stage recognises the extent of local invasion across the bowel wall. When colorectal tumour can be assessed, the pT stage includes four levels of invasion from pT1 (tumour invades the submucosa) to pT4 (tumour directly invades other organs and/or involves the visceral peritoneum).1–4 The Royal College of Pathologists3 and the Association of Directors of Anatomic and Surgical Pathology4 recognise two subsets for the highest pT4 category, namely pT4a and pT4b. Both of these schemes define pT4a subset as direct tumour invasion of adjacent organs, and pT4b as tumour penetration of the visceral peritoneum. These two subsets of pT4 have different implications and therapeutic connotations, and the recommendation is to record these two subsets separately.3 4 While the 2000 College of American Pathologists (CAP) Prognostic Factors Conference5 did acknowledge the presence of two pT4 subsets, the 6th edition of the American Joint Committee on Cancer (AJCC) cancer staging manual6 did not recommend the reporting of the two subsets. The understanding was that more data should be collected to determine whether such subsets be adopted in the future.6 7 The 5th (TNM5) and 6th (TNM6) editions of the International Union against Cancer ( UICC) TNM classification of malignant tumours also do not recognise two subsets in the pT4 stage for colorectal cancer reporting.1 2 Interestingly, there has been a significant change in the recently revised CAP protocol for examination of colorectal cancer specimens with respect to the pT4 subsets.8 In the new proposals, the CAP recommends that the definitions for pT4a and pT4b subsets be reversed.8

In the UK, the Royal College of Pathologists' minimum dataset3 is agreed on the basis of TNM52 and does not accept the changes in TNM6,1 as the evidence base for the proposed changes in TNM6 does not warrant adoption.4 9–11 The UK position regarding the TNM6 centres around the issue of pericolonic and mesenteric tumour deposits.3 4 9–11 However, this stance does not affect the reporting guidelines for recognition of pT4 stage or its two subsets. In contrast, the new CAP proposals8 are likely to affect the reporting of the pT4 subsets in the future. The reversal of the pT4 subsets in the new proposals has generated a heated debate, but it is likely to be adopted by the 7th edition of the UICC TNM manual.7 Nevertheless the debate surrounding the two pT4 substages, pT4 tumour stage denoting serosal (mesothelial) surface involvement in colonic cancers, is one of the most important factors in predicting transperitoneal spread and overall prognosis.12–16 Some large studies have evaluated serosal penetration as a separate pathological variable and have demonstrated by multivariate analysis that it has independent adverse prognostic significance.12–15 17–19 Establishing pT4 tumour stage in colorectal cancer at histology may be seemingly straightforward in most instances (figures 1 and 2). However, one recognises that this is not always the case. While there are best practice guidelines to enable easy identification of pT4 stage,3 4 6 8 20 there are some contentious issues in establishing a robust histological diagnosis of the pT4 stage. This is discussed under the following headings:

  1. Presence of tumour perforation

  2. Invasion of an adjacent organ

  3. Direct or discontinuous tumour extending to the non-peritonealised resection margin

  4. Mesothelial reaction (cellular or fibrous mesothelial proliferation) in the absence of actual tumour breaching the peritoneal surface.

Figure 1

Tumour cells penetrating peritoneal surface in the crevices (H&E ×2.5).

Figure 2

Tumour perforation across serosa (in both instances, the stage is straightforward pT4) (H&E ×2.5).

Presence of tumour perforation

Perforation is a macroscopically visible defect through the tumour wherein the bowel lumen is in communication with the external surface.3 Tumour perforation into the peritoneal cavity in colonic and rectal cancer should be recorded as pT4 tumour stage. Strictly speaking, localised perforation of the tumour into the mesentery, mesorectum or retroperitoneum, unlike serosal perforation, does not denote pT4 stage, as this does not pose a risk for potential contamination of the peritoneal cavity.3 A caveat to this is rectal cancers where in all forms of perforation, localised or otherwise, are recorded as pT4 stage.3 Even in the absence of tumour perforation into the peritoneal cavity, localised tumour perforation is very likely to have tumour positive non-peritonealised resection margin (NRM) involvement (figure 3A,B). In this situation, the surgeon is more likely to cut into the tumour leaving behind residual tumour. While this scenario is unlikely to pose a risk factor for intraperitoneal metastasis, it is associated with an increased risk of leaving residual tumour behind and hence local recurrence. It is prudent to consider a pT4 stage here to highlight the involvement of the NRM that poses a risk for local recurrence.

Figure 3

(A) Transverse colon cancer with positive, posterior non-peritonealised resection margin (painted diathermy margin) (H&E ×2.5). (B) Shavings from pancreatic bed showing further tumour indicating residual tumour left behind by surgeon. This tumour should be staged as pT4 (H&E ×2.5).

Invasion of an adjacent organ

Invasion of an adjacent organ by colorectal cancer can be either intramural or extramural. Intramural invasion denotes longitudinal extension into an adjacent part of the bowel (eg, extension of a caecal tumour into the terminal ileum or of a rectal cancer into the anal canal) without serosal surface involvement (figure 4).3 In contrast, direct invasion beyond the bowel wall by way of serosa (figure 5) represents extramural invasion (eg, invasion of sigmoid colon by a carcinoma of caecum).3 4

Figure 4

Direct luminal extension of low rectal cancer into anal canal, not to be confused with pT4 stage (H&E ×2.5).

Figure 5

Small bowel involvement across serosa. This should be recognised as pT4 (H&E ×2.5).

Theoretically, in the scenario of extramural spread across the serosa to involve an adjacent organ, there is fulfilment of both of the criteria that define the pT4 stage. In this situation, there is invasion into an adjacent organ, as well as extension beyond the serosal surface. The UK College dataset recognises extramural invasion beyond the bowel wall by way of serosa as pT4 stage only, without qualifying a pT4 subset.3

Direct extension of a rectal cancer beyond the rectal wall into the skeletal muscle of the external sphincter, levator ani and/or puborectalis (without serosal involvement) also denotes extramural invasion.3 The UK College dataset recognises invasion of a low rectal tumour into the levators as pT4a stage separate from serosal involvement due to different therapeutic connotations.3

The 6th edition of the UICC TNM staging recognises colonic tumour that is adherent to other organs or structures macroscopically as pT4 stage.1 If there is no tumour in the adhesion microscopically, the tumour stage should be recognised as pT3. Thus the pT4 stage should be based on the microscopic presence of tumour when colonic tumours are adherent to other organs or structures rather than the mere presence of macroscopic adhesion to adjacent organs (figures 6 A,B and 7).4

Figure 6

A) Rectosigmoid tumour with extensive adhesion to bladder by fibrosis alone (H&E ×2.5). (B) This tumour does not invade adherent bladder (detrusor muscle) to qualify for pT4 stage (H&E ×2.5).

Figure 7

Rectosigmoid tumour with direct invasion into adherent bladder (detrusor muscle) which qualifies for pT4 stage (H&E ×2.5).

Direct or discontinuous tumour extending to the NRM

The NRM, previously known as the radial or circumferential resection margin, represents the ‘bare’ area in the connective tissue at the surgical plane of excision that is not covered by a serosal surface.3 Its extent varies greatly according to the site of the tumour. Its involvement is predictive of local recurrence and poor survival in rectal tumours,3–6 20–23 and in those that have not received neoadjuvant therapy, it may be an indication for postoperative adjuvant therapy.20 23–31

The presence of a non-peritonealised margin (NPM) in colonic tumours other than the rectum has also been recently recognised.3 4 8 13 32 33 Tumours of the ascending and descending colon usually have a NPM margin posterolaterally.3 The NPM around caecal cancers is typically small and narrow. Transverse and sigmoid colon tumours do not have a NPM, as they are entirely intraperitoneal and have an associated mesentery (mesocolon). It is important to recognise the NPM margin separate from the serosal surface in tumours of the ascending and descending colon. The pathologist should ink this margin in the fresh specimen with a suitable marker to enable the subsequent identification of margin involvement.

Involvement of this margin may be through direct continuity with the main tumour, by tumour in veins, lymphatics or lymph nodes, or by tumour deposits discontinuous from the main growth (figures 3A and 8). The Gloucester group12 13 recommends that the NPM should be regarded as being involved if the tumour is continuous with this surface through an area of inflammation. Tumour can apparently seed across an inflammatory focus, present at this margin, to allow metastasis, even if tumour is not demonstrated at the margin.12 The assessment of the NRM involves recording of the minimum distance (in millimeters) between the tumour and the NRM from histological slides.3 20 If this is less than or equal to 1 mm, then the NRM is regarded as involved in the assessment of completeness of the resection.3 Involvement of this margin is a risk factor for local recurrence and is recognised separate from peritoneal surface involvement.20 22 32–35 The concept of the NPM in colonic tumours other than the rectum suggests that colonic tumours other than those arising in the rectum may invade an adjacent organ or the abdominal wall, and may qualify for a pT4 stage (figure 8).

Figure 8

Caecal cancer at non-peritonealised resection margin (abdominal wall). Note diathermy margin indicating the surgeon cut through the tumour (H&E ×2.5).

Mesothelial reaction (cellular or fibrous mesothelial proliferation) in the absence of actual tumour breaching the peritoneal surface

Involvement of the serosal (peritoneal) surface is defined as ‘tumour breaching of the serosa with tumour cells visible either on the peritoneal surface or free in the peritoneal cavity’ (figures 1 and 9).3 12 When tumour extends close to the peritoneal surface without direct involvement of the surface, it is recommended that serial levels be cut through the block, to confidently exclude peritoneal involvement.3 20 The implication of the definition and the subsequent statement is that tumour close to but not at the peritoneal surface does not qualify for pT4 stage (figure 10). Disruption of the mesothelial cells on the bowel surface may be elicited by actual tumour cells or by an inflammatory reaction (figure 11). Thus, disruption of the mesothelial surface may include the following:

  1. mesothelial inflammatory and/or hyperplastic reactions with tumour close to but not at the serosal surface

  2. tumour present at the serosal surface with inflammatory reaction or mesothelial hyperplasia

  3. tumour cells actual breaching the serosal surface.

Figure 9

Tumour kisses the mesothelial surface, but no actual breach is seen. This would qualify for pT4 stage (H&E ×5).

Figure 10

Tumour is close but does not kiss or breach the peritoneal surface to qualify for pT4 stage (H&E ×2.5).

Figure 11

There is abscess formation with a zone of inflammation extending to the serosa; this does not denote pT4 stage (H&E ×2.5).

The first situation can often be frustrating to the reporting pathologist. There is also the question of ‘how close is close?’ in the situation of tumour close to the peritoneal surface. The Gloucester group specifically studied local peritoneal involvement (LPI) in colonic cancers by dividing them into four subgroups as follows12:

  1. tumour well clear of closest peritoneal surface (LPI 1)

  2. inflammatory and/or hyperplastic reaction with tumour close to, but not at, the peritoneal surface (LPI 2)

  3. tumour present at peritoneal surface with inflammatory reaction and/or ‘ulceration’ (LPI 3)

  4. tumour cells shown free in peritoneum and evidence of adjacent ‘ulceration’ (LPI 4).

The results of these studies showed that the last two groups (LPI 3 and 4) had a similar prognostic outlook and should therefore be recognised as actual peritoneal involvement in colonic resections.12 15 The second group (LPI 2) did not show a similar outcome and hence does not warrant a pT4 designation. While the question of how close is close remains unanswered, the recommendation of the Gloucester group to examine multiple levels to assess the closeness of the tumour to the peritoneal surface in histological terms is good practice.20 This would avoid an overcall of the Gloucester LPI group 2 as pT4 stage. Given the similar prognostic outcome for LPI groups 3 and 4 in the Gloucester study, both of these groups should be designated as pT4 in histopathology reporting (figures 1 and 9).

The Association of Directors of Anatomic and Surgical Pathology recommend that pT4 stage should apply when tumour cells are observed on the free peritoneal surface even if admixed with fibrin.4 The pT4 stage should not be used even when tumour is within 1 mm of the peritoneal surface and staged as pT3 on the basis that the peritoneum serves as an effective barrier to tumour spread.4 Both the UK College dataset3 and the Association of Directors of Anatomic and Surgical Pathology4 recommend staging of tumour penetration of the peritoneum as pT4b subset.

The question of whether acellular mucin lakes in a mucinous adenocarcinoma at the serosal surface constitute pT4 stage is an interesting issue (figure 12). The Gloucester study group found that mucinous tumour phenotype is more strongly associated with peritoneal involvement and intraperitoneal dissemination.12 36 This suggests that the situation of acellular mucin at the serosal surface warrants a pT4 tumour stage.

Figure 12

Mucinous cancer with tumour cells and acellular mucin at peritoneal surface which qualifies for pT4 stage (H&E ×2.5).

Tumour continuity with the peritoneal surface through an area of inflammation (the inflamed peritoneal surface) cannot be technically covered by the pT4 tumour stage definition, as there is no actual tumour breach (figure 11). One must also remember that serosal involvement through direct continuity with the primary tumour is recorded as pT4 stage, while peritoneal tumour deposits that are separate from the primary tumour are recognised as distant metastases (pM1) without further qualification.3 The pM1 stage is evidently aimed at recognition of peritoneal seedling akin to the Krukenberg tumour of the ovary.

Take-home messages

  1. pT4 represents the highest stage of local tumour extension in colorectal cancer and some histological features may obscure its recognition resulting in upstaging or understaging of the tumour.

  2. At present, recognition of the two subsets for pT4 tumour stage in colorectal cancer and their definitions are not uniformly practised.

  3. Only tumour present at the peritoneal surface and/or tumour cells actually breaching the peritoneal surface qualify for pT4 stage. Tumour close to the peritoneal surface or continuity with the surface through an area of inflammation does not denote pT4 stage.

  4. The non-peritonealised surgical resection margin (NRM) in ascending and descending colonic cancers other than the rectum should be recognised and inked in the fresh specimen. Involvement of NRM may be through direct continuity with the main tumour, discontinuous tumour deposits or tumour continuous through an area of inflammation.


I am grateful to Ms Cara Lydon for her immense help in preparing the manuscript. I am thankful to Mr K Harbottle and Dr L Semokin for their support in producing the images.



  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed