Pathology in Slovenian CRC screening programme: Organisation and quality assurance Snježana Frković Grazio and Matej Bračko
June 2009 to December 2013 (first three rounds) 33 969 colonoscopies were performed in 33 207 persons specimens were submitted in 65% of all colonoscopies (this proportion has remained stable over time) 21 976 pathology requests were submitted ( 4800 per year) the number of submitted lesions per pathology request ranged from 1 to 26 (the average number has steadily increased from 2.1 in 2009 to 2.5 in 2013) a total of 51 268 specimens were examined
Finding N % Carcinoma 1163 2.27% Suspicious for carcinoma 162 0.32% Adenoma 35275 68.81% SSL or HP 8476 16.53% SSL 1101 HP 7375 Other 6192 12.08% Inflammatory polyp 669 Leiomyoma 146 Lipoma 71 NET 26 Lymphoma 3 Metastatic ovarian carcinoma 3 Other (inflammation, normal... 5274 Total 51268 100.00%
Pathology in CRC screening the pathology service plays an important role in CRC screening since the management of participants depends on quality and accuracy of the diagnosis pathologic findings affect the decision to undergo further local or major resection as well as surveillance after screening
Factors that affect pathology endoscopists expertise and experience of pathologist organised collection and analysis of data quality control
Factors that affect pathology endoscopists expertise and experience of pathologist oganised collection and analysis of data quality control
endoscopists E1 E2 pathologists P1 E3 E4 E5 P2 P3 E7 E8 E9 E10 E11 P4 P5 P6
Factors that affect pathology endoscopists expertise and experience of pathologist organised collection and analysis of data quality control
Concentration of cases 4 histopathology units (17 pathologist) with specific experience in gastrointestinal pathology, colorectal cancer diagnosis & treatment and participation in MDT meetings Each participating pathologists reports at least 200 screening biopsies per year
Education Introductory course Training course / workshop led by the leading British gastrointestinal pathologists in October 2011 Refresher course planned in spring 2015
Collection of data In addition to written reports, diagnoses and all necessary data are entered into structured online computer database system For each lesion, pathology data are linked with corresponding endoscopic data Data can be easily retrieved and analyses performed (e.g., comparisons between pathologists, pathology units, etc.)
Quality control analysis and comparison of data internal quality control participation in an external quality assurance (EQA) programme
Quality control analysis and comparison of data internal quality control Turnaround times (TAT) Proportion of various types of lesions Proportion of lesions with HG dysplasia Proportion of adenomas with HG dysplasia Proportion of adenomas with villous component. participation in an external quality assurance (EQA) programme
Proportion of pathology reports signed out in 5 working days by histopathology unit 100% 90% 80% 70% 60% 50% 40% 30% MF LJ OI LJ A B UKC MB C GK LJ D skupaj all 20% 10% 0% 12/1 12/2 13/1 13/2
100% Proportion of pathology reports signed out in 5 working days by pathologists 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 12/1 12/2 13/1 13/2
25% Proportion of adenomas with high-grade dysplasia 20% 15% 10% 5% 0% 2009 2010 2011 2012 2013
60% Proportion of adenomas >= 10 mm 50% 40% 30% 20% 10% 0% 2009 2010 2011 2012 2013
40% 35% Proportion of adenomas with high-grade dysplasia training course / workshop 30% 25% 20% 15% all >=10 mm <10 mm 10% 5% 0% 2009 2010 2011 2012 2013
35% Adenomas with HG dysplasia by pathologist 30% 25% 20% 15% 2009-11 2012-13 10% 5% 0% A B C D E F G H I J K L M N O P R S T total
50% Proportion of adenomas with villous component (>10%) 45% 40% 35% 30% 25% 20% all <10 mm >10 mm 15% 10% 5% 0% 2009 2010 2011 2012 2013
40% Adenomas with villous component by pathologist 35% 30% 25% 20% 2009-11 2012-13 15% 10% 5% 0% A B C D E F G H I J K L M N O P R S T total
Quality contol analysis and comparison of data internal quality control Turnaround times (TAT) Proportion of various types of lesions Proportion of lesions with HG dysplasia Proportion of adenomas with HG dysplasia Proportion of adenomas with villous component. participation in an external quality assurance (EQA) programme
UK BCSP EQA uses virtual slides (10 cases) slides accessed online http://www.gieqa.org.uk/ 4 possible answers for each slide Other Low grade dysplasia High grade dysplasia Adenocarcinoma
UK BCSP EQA A case is valid only if the diagnosis is agreed by 80% of the regional lead pathologists Points per case: 2 points for same diagnosis as consensus 1 point for one category removed (e.g. high grade dysplasia/carcinoma) 0 points otherwise Participant score is sum of points for the valid cases (score for 10 cases can be from 0 to 20)
Results of the b13b circulation K of the NBCS EQA sheme score (0-16) SHA Leads 8 1 (N=9) 88,9% 11,1% 16 15 14 13 12 11 10 9 UK participants 221 156 52 16 3 2 (N=450) 49,1% 34,7% 11,6% 3,6% 0,7% 0,4% Irish participants 3 4 2 1 (N=10) 30,0% 40,0% 20,0% 10,0% Slovenian participants 5 4 4 3 1 (N=17) 29,4% 23,5% 23,5% 17,6% 5,9% score 12 = poor performer
Pathology in CRC screening European guidelines for quality assurance in colorectal cancer screening and diagnosis Pathology: Chapter 7 & Annex 7a 23 recommendations
EG recommendations participating pathologists should have specific training in colorectal pathology pathologist should develop a network in order to share experience double reading in cases of T1 cancer participation in MDT meetings
EG recommendations mucosal neoplasia should be used instead of dysplasia only two grades of neoplasia should be used (low grade and high grade) adenomas should be classified as tubular, tubulovillous or villous, using 20% rule * the terms intra-mucosal carcinoma or in situ carcinoma should not be used (= HG mucosal neoplasia) the WHO definition of carcinoma should be used: an invasion of neoplastic cells through the muscularis mucosae into submucosa
what should be reported - type of lesion - in case of adenoma: - type (tubular, tubulovillous, villous, traditional serrated) - grade of neoplasia / dysplasia (LG, HG) - size of adenoma - involvement of resection margins - in case of polyp cancer (pt1 cancer) - tumor grade (low 1, 2 or high 3) - lymphovascular invasion (present, absent, suspicious) - margin involvement ( 1 mm is generally regarded as an indication for further therapy - endoscopic or surgical) - substaging - Kikuchi / Haggitt levels or measurement of depth and width
EG recommendations all lesions should be reported by proforma or structured reporting and the data returned to the screening programme (in a minimum 90% of all cases) *departments and individual pathologists should audit their own reporting practices for key features - distribution of the type and size of lesions - frequency of grades of neoplasia and villousnes (not more than 10% of HG) - the number of LN retrieved (median 12), the frequency of extramural vascular invasion ( 25%), peritoneal invasion (colon 20%, rectum 10%)... in surgical resection specimens participation in an external quality assurance (EQA) programme