Introduction: Diagnostic and reporting errors are one of the major challenges encountered in provision of quality health care services. Incorrect diagnosis interferes with clinical decision making and ends up with serious consequences in terms of patient management. Practicing regular departmental audits is a milestone towards professional development and excellence. The aim is to highlight and address errors and discrepancies in routine practices to maintain high quality standards and take corrective actions for future. Shaukat Khanum Memorial Cancer hospital being one of the largest cancer centers in Pakistan, colorectal polyps are not an uncommon biopsy specimen received in surgical pathology lab. We conducted an intra-departmental audit of colorectal polyps. The aim of our study was to investigate diagnostic accuracy and quality of reporting of colorectal polyps according to standardized CAP guidelines.

Material and Methods: We performed a retrospective study of 229 cases of colorectal polyps diagnosed in year 2019 after approval by institutional review board. The data was available in hospital archives. Slide review was done by two consultant pathologists to assess diagnostic and reporting errors. The conflicting cases were referred to a third pathologist for final opinion.

Results: The diagnostic discrepancies were observed in 20 (6.1%) cases and reporting errors were present in 61 (18.6%) out of 229 cases (328 biopsies).

Conclusion: Overall the results were satisfactory in terms of patient management. There was only one case with discrepant grading of dysplasia and there were four cases with overcall of adenocarcinoma which turned out be high grade dysplasia on review.

Keywords: Colorectal Cancer (CRC); College of American Pathologists (CAP); Mucin Rich Traditional Serrated Adenomas (MR-TSAs); Serrated Tubulovillous Adenomas (S-TVAs)

Colorectal cancer is the second most common cancer in women and third most common malignancy in males worldwide [1]. The incidence is higher in well developed countries compared to low-income countries. The established risk factors are dietary habits like consumption of alcohol and processed meat, sedentary lifestyle and genetic predisposition. Colorectal polyps are precursor lesions that can harbor genetic rearrangements, develop high grade dysplasia and can undergo malignant transformation. Colorectal screening programs are used worldwide for early detection of precursor lesions and identification of patients at high risk of developing colorectal cancer.

Unfortunately, there is no such screening strategy operating at present in Pakistan and majority of polyps are diagnosed as incidental endoscopic finding. Few cases of larger polyps present with abdominal discomfort, bleeding and diarrhea. The primary objective of our study was to assess diagnostic accuracy of colorectal polyps. Furthermore, we also aimed to investigate quality of reporting of colorectal polyps according to standardized CAP guidelines.

After approval by institutional review board, we retrospectively reviewed 241 consecutive cases of colorectal polyps diagnosed in the year 2019. The data was retrieved through the hospital archives. The small mucosal biopsies diagnosed as colorectal polyp, history of polyposis syndrome or with endoscopic impression of polyp (229 cases) were included in the study. Larger resections including total and hemi-colectomies (12 cases) were excluded from the study. After applying inclusion-exclusion criteria we were left with a total of 229 cases. A double-blind slide review was done by two independent consultant pathologists with significant experience in gastrointestinal pathology to avoid personal bias. Diagnostic accuracy was assessed on the basis of correct categorization of polyps, grading of dysplasia and correct assessment of malignancy arising in a polyp. The cases marked as discrepant by two pathologists were re-discussed on a multi-head session. The cases with difficult assessment and high subjectivity were referred to a third pathologist for final opinion. 

Dysplasia in conventional adenomas was graded according to two-tiered stratification as low and high grade dysplasia based on cytological and architectural changes. However, grading the severity of dysplasia into low or high grade in serrated lesions is not recommended by WHO [1]. The quality of reporting was assessed according to CAP protocol for examination of excisional biopsy (polypectomy) specimen April 2021 [2-4]. The core data items recommended in CAP guidelines were tumor site, tumor size (including polyp size, configuration and size of invasive carcinoma if present), histological type (use of standardized terminologies), grading of dysplasia, extent of local invasion (for malignant polyps), comment on mucosal margin (for sessile polyp), deep margin (for pedunculated polyp with stalk) and lymphovascular invasion.

Results:
Table 1 shows the baseline histomorphological features of 229 cases (328 biopsies) of colorectal polyps. Majority of cases had multiple biopsy specimens from different sites making the total biopsy count as 328. There was male pre-dominance (73.8%). Location wise, left colon harbored maximum number of polyps 184 (56.1%), pre-dominantly in rectum and anal canal 74 (22.6%). There were 107 (31.7%) single polyps ranging in size from 1 - 25 mm. Rest of the polyps were received as multiple mucosal fragments. We broadly classified the data into conventional colorectal adenomas, serrated lesions and polyps, inflammatory polyps, adenocarcinoma (with and without precursor lesion), genetic syndrome associated polyps, mucosal prolapse polyps, no significant pathology and cases signed out as descriptive without histological features of polyp Table 2. Tubular adenomas with low grade dysplasia 72 (18.3%) constituted the largest group.

Table 1: Baseline histopathological features of 229 cases (328 biopsies).

IRB Approval number: EX-04-06-21-03.

Table 2: Original diagnosis and discrepancies in 229 cases (328 biopsies) of colorectal polyps.

The diagnostic discrepancies were observed in 20 cases (6.1%). Errors in correct categorization were present in 15 cases. 6 cases of conventional adenomas had incorrect diagnosis. 2 of these 6 cases were under-diagnosed as non-neoplastic, one case as focal hyperplastic change, juvenile rectal polyp (1 case) and two cases were signed out as focal active colitis. 3 cases of sessile serrated lesions were incorrectly diagnosed as conventional adenomas. Hyperplastic polyp (1 case) was misdiagnosed as tubular adenoma with low grade dysplasia. Prolapse associated polyp (1 case) and 1 case of juvenile rectal polyp were incorrectly categorized as hamartomatous polyp and inflammatory polyp respectively. 3 cases had no adenomatous change and were over-diagnosed as conventional adenomas. Error in grading of dysplasia was observed in 1 case of tubular adenoma with high grade dysplasia which was under-graded as low grade dysplasia. There was overcall of adenocarcinoma in 4 cases which turned out to be conventional adenomas with high grade dysplasia on review.

The reporting discrepancies were observed in 61 cases (18.6%). Major reporting discrepancies encountered during review were use of non-standardized terminologies, inappropriate sub-typing of conventional adenomas, missing comment on size of invasive tumor in a malignant polyp , margin status and lymphovascular invasion. 21 cases were reported as adenomatous polyps which is not a standardized term. Conventional adenomas should ideally be reported as tubular, villous or tubulovillous adenomas. Furthermore 10 cases with colonoscopic impression of polyp were signed out as polypoidal colonic fragments with hyperplastic change which is again not a correct terminology. On review, all 10 cases showed no significant pathology highlighting the bias of the signing out pathologist towards colonoscopic impression of polyp. Problem in correct sub-typing was observed in two cases of conventional adenomas, one having <25% villous architecture and wrongly reported as tubulovillous adenoma and the other having >25% villous architecture and misdiagnosed as tubular adenoma.

Out of 107 intact polyps, there was no comment on margin status (free or involved by dysplasia) in 28 cases of conventional adenomas with low grade dysplasia and 1 with high grade dysplasia. 2 cases had adenocarcinoma arising in a background of tubulovillous adenoma. There was no comment on size of invasive tumor arising in polyp, margin assessment and lymphovascular invasion. Rest of the polyps were received fragmented.

Over the past decade, there has been significant advancement in endoscopic procedures of superficial colorectal lesions. These advanced endoscopic procedures have opened a new era of interdisciplinary collaborative efforts of a pathologist and gastrointestinal endoscopist to understand each other's perspective. These advanced procedures are more targeted and demand more precise, detailed and meticulous pathology reporting to optimize patient management. The quality assurance audit we performed brought into light many diagnostic and reporting errors. Diagnostic difficulty in correct categorization of sessile serrated lesions and conventional adenomas was observed in 3 cases (3.7%). According to WHO 2019 the diagnostic criteria for sessile serrated lesion is “The Presence of Single Unequivocal Architecturally Distorted Crypt”. The architecturally distorted crypt was explained by presence of at least one of these features i.e. Serrations extending to crypt base, horizontal growth along muscularis propria (T or L-shaped crypts) and dilatation of basal third of crypts Figure 1. Three types of dysplasia have been described in serrated lesions with dysplasia by Crockett and Nagtegaal (2019) [5] i.e. Intestinal dysplasia, serrated dysplasia and minimal deviation dysplasia [5]. Intestinal dysplasia is relatively rare and is similar to dysplasia in conventional adenoma. MLH1 satin shows retained expression and very rarely show malignant transformation. Serrated dysplasia is characterized by eosinophilia cytoplasm, closely packed glands, nuclear atypia and mitotic activity. Loss of MLH1 staining is not very common. Minimal deviation dysplasia is similar to sessile serrated lesion with characteristic loss of MLH1 expression.