Article Text

Download PDFPDF

A pathologist's survey on the reporting of sessile serrated adenomas/polyps
  1. Runjan Chetty1,
  2. Adrian C Bateman2,
  3. Emina Torlakovic1,
  4. Lai Mun Wang3,
  5. Pelvender Gill3,
  6. Adnan Al-Badri4,
  7. Mark Arends5,
  8. Leigh Biddlestone6,
  9. Susan Burroughs7,
  10. Frank Carey8,
  11. David Cowlishaw9,
  12. Stephen Crowther10,
  13. Philip Da Costa11,
  14. Mahomed A Dada12,
  15. Charles d'Adhemar13,
  16. Kaushik Dasgupta14,
  17. Chandima de Cates15,
  18. Vikram Deshpande16,
  19. Roger M Feakins17,
  20. Bineeta Foria18,
  21. Vipul Foria2,
  22. Clare Fuller7,
  23. Bryan Green2,
  24. Joel K Greenson19,
  25. Paul Griffiths20,
  26. Sara Hafezi-Bakhtiari1,
  27. James Henry21,
  28. Eleanor Jaynes2,
  29. Michael D Jeffers10,
  30. Philip Kaye22,
  31. Robert Landers23,
  32. Gregory Y Lauwers16,
  33. Maurice Loughrey24,
  34. Nicholas Mapstone25,
  35. Marco Novelli26,
  36. Robert Odze27,
  37. David Poller9,
  38. Corwyn Rowsell28,
  39. Scott Sanders29,
  40. Patrick Sarsfield30,
  41. John B Schofield31,
  42. Kieran Sheahan32,
  43. Neil Shepherd15,
  44. Ali Sherif33,
  45. James Sington34,
  46. Shaun Walsh8,
  47. Namor Williams35,
  48. Newton Wong36
  1. 1University Health Network, Toronto, Canada
  2. 2University Hospital Southampton NHS Foundation Trust, Southampton, UK
  3. 3Oxford University Hospitals Trust, Oxford, UK
  4. 4Royal Hampshire County Hospital, Winchester, UK
  5. 5Western General Infirmary, Edinburgh, UK
  6. 6Royal United Hospital, Bath, UK
  7. 7Salisbury NHS Foundation Trust, Salisbury, UK
  8. 8Ninewells Hospital, Dundee, UK
  9. 9Queen Alexandra Hospital, Portsmouth, UK
  10. 10Tallaght Hospital, Dublin, Ireland
  11. 11Airdale Hospital, Keighley, UK
  12. 12West Suffolk Hospital, Bury St Edmunds, UK
  13. 13Midland Regional Hospital, Tullamore, Ireland
  14. 14University Hospital of North Tees, UK
  15. 15Cheltenham General Hospital, Cheltenham, UK
  16. 16Massachusetts General Hospital, Boston, USA
  17. 17Barts and London NHS Trust, UK
  18. 18Royal Bournemouth Hospital, Bournemouth, UK
  19. 19University of Michigan Hospitals, Ann Arbor, USA
  20. 20Swansea NHS Trust, Swansea, UK
  21. 21Queen Elizabeth Hospital, Gateshead, UK
  22. 22Nottingham University Hopsitals, Nottingham, UK
  23. 23Waterford Regional Hospital, Waterford, Ireland
  24. 24Royal Victoria Hospital, Belfast Trust, UK
  25. 25Royal Lancaster Infirmary, Lancaster, UK
  26. 26University College London Hospitals, London, UK
  27. 27Brigham and Womens Hospital, Boston, USA
  28. 28Sunnybrook Health Sciences Centre, Toronto, Canada
  29. 29South Warwickshire Foundation Trust, Warwick, UK
  30. 30Royal Devon and Exeter Hospital, Exeter, UK
  31. 31Kent Oncology Centre, Maidstone, UK
  32. 32St Vincent's University Hospital, Dublin, Ireland
  33. 33Russell Hall Hospital, Dudley, UK
  34. 34Norfolk and Norwich University Hospital, Norwich, UK
  35. 35Singleton Hospital, Swansea, UK
  36. 36Bristol Royal Infirmary, Bristol, UK
  1. Correspondence to Professor Runjan Chetty, Department of Pathology, University Health Network, 11th floor Eaton wing, Toronto General Hospital, 200 Elizabeth Street, Toronto, Ontario, Canada M5G 2C4; runjan.chetty{at}gmail.com

Abstract

Aim The purpose of this survey was to ascertain reporting habits of pathologists towards sessile serrated adenomas/polyps (SSA/P).

Methods A questionnaire designed to highlight diagnostic criteria, approach and clinical implications of SSA/P was circulated electronically to 45 pathologists in the UK and North America.

Results Forty-three of 45 pathologists agreed to participate. The vast majority (88%) had a special interest in gastrointestinal (GI) pathology, had great exposure to GI polyps in general with 40% diagnosing SSA/P at least once a week if not more, abnormal architecture was thought by all participants to be histologically diagnostic, and 11% would make the diagnosis if a single diagnostic histological feature was present in one crypt only, while a further 19% would diagnose SSA/P in one crypt if more than one diagnostic feature was present. The vast majority agreed that deeper sections were useful and 88% did not feel proliferation markers were useful. More than one-third did not know whether, or did not feel that, their clinicians were aware of the implications of SSA/P.

Conclusions 98% of pathologists surveyed are aware that SSA/P is a precursor lesion to colorectal cancer, the majority agree on diagnostic criteria, and a significant number feel that there needs to be greater communication and awareness among pathologists and gastroenterologists about SSA/P.

  • Sessile Serrated Polyp
  • Sessile Serrated Adenoma
  • Serrated Polyp
  • Gastrointestinal Polyps

Statistics from Altmetric.com

Sessile serrated adenomas/polyps (SSA/P) often referred to as sessile serrated lesions (SSL) in the UK, have been the subject of much clinical and molecular research in the last 10 years. Despite this, the degree of awareness among many pathologists regarding the diagnostic criteria, terminology and the implications of making this diagnosis is unknown. There have been publications in which the diagnostic criteria for SSA/P have been discussed and reiterated, and these were consolidated in the 2010 WHO classification.1–3 The following types of serrated polyp were recognised: hyperplastic polyp, sessile serrated adenoma/polyp with or without dysplasia, and traditional serrated adenoma.1 Recently, the American College of Gastroenterology (ACG) produced a paper that delineated and highlighted the already well-established pathological features for a diagnosis of SSA/P.4

Two recent studies have shown that SSA/P is still being underdiagnosed in academic and district general hospital settings in the UK.5 ,6 The purpose of this survey was, therefore, to further assess the practice habits and reporting of SSA/P among a range of pathologists in academic and district general hospital departments, especially in the UK and Ireland, with regards to SSA/P in their routine practice.

Materials and methods

A questionnaire (see box 1) was designed (by RC, ACB, LMW, ET and PG) to determine/ascertain participant practice habits specifically related to SSA/P. These included questions on gastrointestinal (GI) pathology expertise, exposure to SSA/P, diagnostic criteria for SSA/P and, opinions regarding the clinical implications/consequences of this diagnosis. A total of 37 pathologists from the UK and Ireland were invited to participate in the questionnaire which was circulated via SurveyMonkey. Additionally, 8 GI monospecialist pathologists from North America were also invited to participate. The invitees were diagnostically involved in GI pathology including GI monospecialists, as well as pathologists with a proclaimed special interest in GI pathology despite reporting other subspeciality areas. Furthermore, several of the participants were members of the Pathology section of the British Society of Gastroenterology and/or of the UK Bowel Cancer Screening Programme pathology committee. Free comments were also allowed pertaining to each question, and the results were then collated by RC and ACB.

Box 1

Questions asked in the survey and responses from the 43 participants

Question 1: Type practice

 I am a specialist GI pathologist (monospecialist): 15

 I am a specialist GI pathologist (not monospecialist): 9

 I am a general pathologist with a special interest in GI pathology: 14

 I am a general pathologist who reports GI biopsies but without a defined special interest in this area: 5

Question 2: How many GI polyps do you see?

 I see GI polyps rarely (eg, 0–10 per month): 0

 I see GI polyps occasionally (eg, 11–30 per month): 8

 I see GI polyps frequently (eg, >30 per month): 35

Question 3: How often do you make a diagnosis of SSA/P?

 Less than once a month: 15

 Once a month: 11

 Once a week: 11

 More than once a week: 6

Question 4: Do you diagnose SSA/P in only the right or left colon, or both?

 I make the diagnosis only in the right colon: 8

 I make the diagnosis in only the left colon: 0

 I make the diagnosis in the right and left colons: 35

Question 5: What are the key histological criteria that you use to make the diagnosis of SSA/P? You can tick as many options as required

 Bland cytology: 22

 Elongation of crypts: 18

 Prominent serration: 24

 Serration at crypt base: 33

 Overall distortion of the architecture: 20

 Goblet cell presence at the base of the crypts: 14

 Dilated crypts: 32

 Abnormally shaped crypts (L-shaped or T-shaped): 43

Question 6: What are the minimum histological criteria that you accept for a diagnosis of SSA/P?

 A single crypt showing one of the characteristic features is sufficient: 5

 A single crypt is sufficient if it shows >1 of the characteristic features: 8

 Several crypts showing at least one of the characteristic features are required: 30

Question 7: How often do you get levels prepared on lesions that could be SSA/P?

 Levels done automatically and additional levels are not ordered very often: 29

 Levels done automatically but additional levels still ordered: 8

 Levels not done automatically and additional levels are not ordered very often: 3

 Levels not done automatically but additional levels are often ordered: 3

Question 8: Are proliferation markers used for the reporting of SSA/P?

 Yes-always: 0

 Yes-occasionally: 5

 No: 38

Question 9: When you make a diagnosis of SSA/P, are any of the following stated in the pathology report?

 No evidence of dysplasia: 25

 No evidence of high-grade dysplasia: 4

 No evidence of ‘conventional’ dysplasia: 9

 None of the above: 5

Question 10: Are you aware of the implications of a diagnosis of SSA/P?

 Yes: 42

 No: 1

Question 11: Are your clinicians aware of the implications of a diagnosis of SSA/P?

 Yes: 27

 No: 3

 Don't know: 13

Question 12: If your clinicians are unaware of SSA/P is the diagnosis flagged with a comment in your report?

 n/a—clinicians are aware of the implications of this diagnosis: 16

 Yes: 13

 No: 14

Question 13: Is there a surveillance programme for SSA/P in your institution?

 Yes: 14

 No: 17

 Don't know: 20

GI, gastrointestinal; SSA/P, sessile serrated adenomas/polyps.

Results

These are presented in the box below, in conjunction with the questionnaire.

Of the total of 45 pathologists who were invited, 43 (96%) agreed to participate and respond to the questionnaire. All 37 pathologists from the UK, and 6 of 8 from North America agreed to be participants.

The vast majority of the respondents (88%) had a special interest in GI pathology, and 56% were GI specialist pathologists (including GI monospecialists). Only 12% of participants stated that they did not have a special interest in GI pathology.

Eighty-one per cent of participants diagnosed more than 30 polyps (of all kinds) per month, while 26% made a diagnosis of SSA/P at least once a week, and 14% more than once a week. 81% diagnosed SSA/P in the left and right colons, while 19% made the diagnosis in the right colon only.

In terms of diagnostic criteria for SSA/P, abnormally shaped crypts (L-or T-shaped) were the only histologic feature used by all (100%) participants. Serrations at the base and dilated crypts were the second and third most commonly nominated features used to make a diagnosis of SSA/P (77% and 74%, respectively). Almost 70% of pathologists required several crypts to display at least one characteristic histological feature before rendering a diagnosis of SSA/P; 11% of participants said they would make a diagnosis of SSA/P if a single crypt showed a single typical histological feature, while 19% would make the diagnosis in a single crypt as long as it showed more than one diagnostic criterion.

Routine levels were the standard laboratory handling procedure according to 86% of participants. Only 7% of pathologists indicated that their laboratory did not routinely prepare levels for this type of specimen, and that they did not feel the need to request deeper sections to make or facilitate making the diagnosis of SSA/P. The remaining respondents used levels when considering the diagnosis, either those prepared routinely by the laboratory and/or additional levels subsequently requested by the pathologist. The vast majority (88%) of participants did not resort to the use of proliferation marker immunohistochemistry to make a diagnosis of SSA/P.

Twelve per cent of respondents did not qualify their diagnosis of SSA/P with a comment on the presence or absence of ‘conventional’ cytological dysplasia, and 98% of pathologists believed that they were aware of the clinical implications of making this diagnosis. However, over one-third of respondents did not know whether, or did not feel that, their clinicians were aware of the implications of a diagnosis of SSA/P. Despite this, about half the pathologists who believed that their clinicians were unaware of the implications of a diagnosis of SSA/P did not provide a comment in their reports alerting the clinicians to this.

Lastly, over two-thirds of participants either did not know if their institution had a surveillance programme for SSA/P, and/or claimed that a surveillance programme did not exist in their institution.

Discussion

The architectural feature of serration was attributed to adenomatous polyps by Longacre and Fenoglio-Preiser7 in 1990, who introduced the term ‘serrated adenoma’. However, this terminology was applied for the most part to lesions that we recognise today as ‘traditional serrated adenoma’.6 SSA/P were distinguished from hyperplastic polyps in 1996 when a detailed histopathological examination of the polyps present in ‘hyperplastic polyposis’ revealed subtle but important differences from sporadic hyperplastic polyps.8 The hallmark of uncomplicated SSA/P is ‘architectural dysplasia’ and evidence of ‘dysmaturation’ believed to be secondary to displacement of proliferation zone from the muscularis mucosae.8 This results in an abnormal shape of the crypts as well as the proliferation zone and maturation sequence being altered and detected in unexpected areas away from muscularis mucosae.8 ,9 It is now believed that these changes are intrinsic to SSA/P. Furthermore, conventional, adenomatous-type dysplasia (low-grade and high-grade) and colorectal cancer (CRC) can develop within SSA/P via the serrated pathway.6 ,7 SSA/P are believed to be more common in the right colon, often attain a size of >10 mm and contain genetic alterations distinct to those of ‘classical’ adenomas. It is postulated that progression to CRC along this serrated pathway may occur at a faster rate than via the chromosomal instability adenoma-carcinoma sequence. In particular, BRAF mutations and DNA methylation across the genome, leading to the CpG island methylator phenotype, are commonly present.10 The latter event results in microsatellite instability and loss of DNA mismatch repair protein (especially hMLH-1 and PMS-2) expression within areas of conventional dysplasia and CRC, when they arise in association with SSA/P.10

Recognition of the neoplastic potential of SSA/P has led to recommendation that these polyps should be managed differently to hyperplastic polyps.2 There is growing acceptance of the need for complete local excision of SSA/P and for endoscopic follow-up that is broadly similar to that recommended for classical adenomas of similar size.2 This contrasts markedly with patients in whom hyperplastic polyps are identified and for the majority of whom no further surveillance is needed.

The implications for patient management following the diagnosis of a lesion as a SSA/P rather than a hyperplastic polyp mean that accurate recognition of SSA/P by histopathologists is essential.

Despite several publications in the last 10 years, SSA/P remains an enigmatic and underdiagnosed entity.11–14 Its absence as a diagnostic entity in some polyp surveillance programmes (such as the bowel cancer screening programme in the UK, which is in the process of being revised), has compounded the lack of awareness. Given its importance, it is incumbent on the pathologist to identify these precancerous lesions. The intent of this survey was to ascertain the degree of consensus among a diverse range of pathologists with regards to SSA/P.

The overwhelming majority of participating pathologists were aware of the implications and importance of SSA/P. However, over one-third of pathologists did not know, or believed that their clinical colleagues were aware, of the implications of SSA/P. This may tie in with the absence of surveillance programmes in several institutions and was pointed out in 2009 by Pettus and colleagues.15 In a survey of 52 university and community hospital-based gastroenterologists, 40% were unaware that SSA and SSP were the same lesion; 50% claimed that they received a diagnosis of SSA/P with an accompanying note in the pathology report, and 73% thought SSA/P showed conventional adenomatous change and were variants of tubular/tubulovillous adenomas.15 Almost all participants (98%), however, were aware that SSA/P were indeed precursor lesions to CRC.15 This discordance of awareness, and/or knowledge about SSA/P between pathologists and clinicians, is of concern, and should be a major focus of education sessions and even routine multidisciplinary team meetings. Despite knowing that clinicians might be unaware of SSA/P, over one-third of participants still did not enter a comment flagging the importance of SSA/P in the pathology report. We believe that the pathologist should take the lead in this role as they make the diagnosis, and appear to be well informed about the entity.

SSA/P is characterised by the presence of several morphological features that are not seen in other serrated polyps (hyperplastic polyps or traditional serrated adenomas). However, there is not yet complete agreement as to which of these are the most important or useful and crucially, how widespread they need to be to make a diagnosis of SSA/P. All participating pathologists did agree that architectural abnormalities of the gland/crypt were diagnostic of SSA/P. The survey has shown that the major differences of opinion and difficulties lie in defining the minimum diagnostic criteria and in terminology. It is apparent that the ACG guideline of a single crypt showing unequivocally typical SSA/P features, is not universally adhered to, and the majority of participants (70%) required diagnostic features to be present in more than one crypt. Additionally, the single histologic feature that all pathologists required was the presence of dilated crypts displaying the characteristic architecture. When these changes are well developed, especially in larger SSA/P, then the diagnosis can be made with relative ease. The difficulty lies in small and/or poorly oriented biopsies. The group of participants in this survey almost unanimously used levels to facilitate the diagnosis. Furthermore, our future understanding of the biological importance of these lesions will need to include a clearer elucidation of the risk of disease progression in relation to the minimum diagnostic criteria required, and to other features, such as the lesion size and anatomical location.

The vast majority of participants (81%) were diagnosed with SSA/P in the right and left colons. Although there is a distinct predilection for the right colon, left-sided SSA/P do exist, and should be diagnosed once histological mimics (especially mucosal prolapse with hyperplastic tubules) have been excluded.3 There is no doubt that bona fide examples of SSA/P do exist in the left colon, and the same histological criteria should be applied in this location as when assessing right-sided lesions.14

A brief discussion regarding terminology raised some divisions around the use of the term, ‘SSL’. Although not an integral part of the survey, the inevitable trans-Atlantic divide on nomenclature did surface. ‘SSL’ is a Eurocentric term and endorsed mainly by UK/European pathologists.16 Some North American participants appeared sanguine about its use, while others were more opposed to the term. There is a contention that using the term ‘adenoma’, by definition, traditionally implies the presence of ‘conventional’ cytological dysplasia, and that ‘lesion’, on the other hand, is perhaps too vague a descriptor. Although the vast majority of SSA/P do not show ‘conventional’ cytological dysplasia, it is ‘architectural dysplasia’ that defines SSA/P. In most non-epithelial tissues, the term ‘dysplasia’ is not limited to nuclear atypia, as it is in preinvasive neoplastic epithelial lesions, but it is used in its alternative and wider meaning/definition, that is, ‘abnormal development or growth of tissues, organs, or cells’. SSA/P is defined by evidence of the abnormal growth of cells in crypts resulting in their disordered development.9 The current terminology that is endorsed by WHO is ‘sessile serrated adenoma/polyp’, but there is a contention, mainly from the UK, that ‘sessile serrated lesion’ be adopted. Suffice it to say, that recognising this characteristic pathology and communicating its importance, no matter what terminology one uses, is the emphasis of this study.

Awareness among clinicians and pathologists needs to increase, and standardised handling and reporting guidelines are needed to ensure greater uniformity and consistency with regards to SSA/P.

Take home messages

  • Sessile serrated adenoma/polyp (SSA/P) must be distinguished from hyperplastic polyps.

  • SSA/P are precursor lesions to colorectal cancer.

  • There needs to be a greater awareness among pathologists and gastroenterologists about SSA/P.

  • An internationally agreed approach to minimum diagnostic criteria and terminology is needed.

References

Footnotes

  • Contributors RC, ACB, ET, LMW and PG contributed to the planning, conduct and reporting of the work described in the article. All others participated in the questionnaire and are aware of the contents of the paper. RC is responsible for the overall content as guarantor.

  • Competing interests None.

  • Provenance and peer review Not commissioned; internally peer reviewed.

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.

Linked Articles