Soutenance mémoire
Participation in screening campaigns
METHODS
Screening
People aged from 50 to 74 years, with no personal or family history of CRC and adenoma, with no colonoscopy for the past five years, no inflammatory bowel disease and no terminal disease or bedridden patient, received a letter from CAP SANTE 49, the departmental analysis center that organizes CRC screening in Maine-et-Loire, and were invited to contact their general practitioner to undergo the screening test.
General practitioners were giving them information about the test, verifying the absence of contraindications (blood in the stool, anorexia, transit disorder…) and delivered the screening test. Individuals did the test at home and sample were sent to IRSA (Institut interRegional pour la Santé, Tours) for g-FOBT and CERBA Laboratory (Cergy-Pontoise) for FIT, to be analyzed.
If result was positive, patients were invited to contact their general practitioner and a gastroenterologist to perform a colonoscopy under general anesthesia. If a polyp or a cancer was found, it was sent to a pathologist.
Inclusion criteria
Patients with polyps or cancer at the colonoscopy after a positive screening test between 01/01/2013 and 31/12/2016 in Maine-et-Loire were included.
We also get data from CAP SANTE 49 database about all the individuals invited to participate in CRC- screening between 01/01/2013 and 31/12/2016.
Exclusion criteria
Individuals with a positive test from CAP SANTE 49 database, who have refused or have a contraindication to colonoscopy were excluded from the analysis.
Fecal occult blood tests
Hemoccult II® was used from 01/01/2013 to 31/12/2014 in the first campaign in our study and OC Sensor® test was used in the second campaign from 01/06/2015 to 31/12/2016. No invitation was sent from 01/11/2014 to 01/06/2015.
Data collection
Gastroenterologists and pathologists have sent a copy of results of colonoscopy and histopathological results of the samples to CAP SANTE 49.
Thanks to their registry, individuals with polyps or cancers were identified. Age, gender and date of positive test were collected for all individuals.
Concerning the colonoscopy: date of colonoscopy, quality of preparation (insufficient, average, good), if the exam was complete or not, localization (rectum, left, transversal and right colon) and size of lesions were collected.
Histological data was collected thanks to CAP SANTE 49 and thanks to histopathological departments of Maine-et-Loire in Centre de Pathologie de l’Ouest or Angers University Hospital. Only information about the 3 most pejoratives lesions per colonoscopy were analyzed. Histological analysis concerning the type of polyp (tubular adenoma, tubulovillous or villous adenoma, serrated adenoma, hyperplastic polyp and others), dysplasia (low or high-grade) and the presence of carcinoma with or without presence of a colloid component were collected.
If a cancer was diagnosed, other datas were collected: date of resection, differentiation (low – moderately – well-differentiated), staging tumor according to the AJCC classification (25), anterior participation to screening CRC campaign and date of the last screening test result.
The criteria for diagnosing cancer, in accordance with the international classification, was an invasion of malignant cells beyond the muscularis mucosa. Intramucosal carcinoma and carcinoma in situ were classified as adenoma with high grade dysplasia (26).
Advanced adenoma (AA) was defined by size ⩾10 mm, tubulovillous or villous adenoma, high grade dysplasia (26).
Advanced neoplasia (AN) was an advanced adenoma or a colorectal cancer.
Collection data was retrospectively performed to 1/05/2017.
Data analysis
The participation rate was calculated as the ratio of number of individuals returning the screening test to the number of individuals invited during the corresponding screening campaign. Detection rate was calculated as the number of persons with a lesion at colonoscopy relative to the number of participants in CRC screening. The positive predictive value (PPV) was the proportion of true positives (persons with a lesion at colonoscopy) relative to the total number of patients who were screened positive and underwent colonoscopy or Computed Tomographic Colonography (CTC). Detection rate and PPV are expressed as percentages with 95% confidence intervals (95% CI).
Characteristics of subjects were expressed as mean +/- standard deviation or frequency (%). Continuous variables were compared between groups (FIT versus g-FOBT) using the Student’s t-test. Categorical variables were compared between groups using the Chi-square test or the Fisher’s exact test where it was appropriate.
All statistical analyses were performed using IBM SPSS version 15 for Windows.
RESULTS
Participation in CRC screening campaigns
In overall 391 932 individuals were invited to participate in CRC screening campaigns between 01/01/2013 and 31/12/2016; 188 815 received an invitation for g-FOBT-based CRC screening (2013-2014) and 203 117 for a FIT-based CRC screening (2015-2016). 143 408 individuals performed a screening test, 69 326 did an Hemoccult® test (g-FOBT) and 74 082 an OC Sensor® test (FIT). Main characteristics are presented in the Flow chart (Figure 1).
4565 participants had a positive screening test: 1346 g-FOBT (1.9%) and 3219 FIT (4.3%) (p<0.0001).
Among participants who were screened positive, 93.5% (4266 of 4565) patients underwent a colonoscopy. Colonoscopy rate was 95.5% for g-FOBT-based campaign and 92.6% for FIT-based campaign (p=0.0003). 10 patients in g-FOBT group and 23 in FIT group refused or had contraindication for colonoscopy. Furthermore, 51 patients in g-FOBT group and 215 in FIT group did not undergo further exploration after positive test; these patients were excluded of the study (61 and 238 respectively).
Colonoscopy was performed within 3 months after the test reading for 50% of the patients and within 8 months for 97% of patients. Among them 55 had an incomplete colonoscopy, 7 (1.1%) in the g-FOBT group and 48 (2.4%) in the FIT group.
2621 subjects had screen-detected colorectal lesion: 648 in g-FOBT group (50.4% of patients who had been screened positive) and 1973 in FIT group (66.2% of patients who had been screened positive) (p<0.0001).
Characteristics of patients with screen-detected lesion according to screening campaign
Mean +/ SD age of patients was 63.1 years (63.1+/-6.93), no difference in age was observed according to screening campaigns (Table I): 62.7 (62.7+/-7.1) for g-FOBT group and 63.4 (63.4+/-6.9) for FIT group, p=0.062. For each screening campaign, there were a higher proportion of men than women among patients with screen-detected lesion: 60.8% versus 39.20% in the g-FOBT group and 63.2% versus 36.8% in the FIT group (p=0.273).
3. Performance comparison between g-FOBT and FIT-based screening campaigns
A total of 292 colorectal cancers were diagnosed, 63 cancers in g-FOBT and 229 cancers in FIT groups (Table II). 2329 patients were diagnosed with at least one polyp and no cancer was registered, with 5019 polyps.
The detection rate for any type of colorectal lesions was significantly higher in FIT group than in g-FOBT (2.66, 95%CI [2.55-2.78] in FIT versus 0.93, 95%CI [0.86-1.01] in g-FOBT, p<0.0001). Significant higher detection rates in FIT group were also observed for advanced lesions (AA, AN and CRC) (Table II).
Predicative positive value (PPV) was significantly higher in FIT (65.92, 95%CI [64.20-67.60]) versus in g-FOBT (50.23, 95%CI [47.51-52.96], p<0.0001), significant differences in favor of FIT were also found for advanced lesions (AA, CRC and AN).
Characteristics of screen-detected lesions
2329 patients with at least one polyp and no cancer were registered, with 5019 polyps. Among them, there were 884 hyperplasic polyps, 343 with g-FOBT and 541 with FIT. There were 122 lesions non-collected after colonoscopy, 28 for the g-FOBT and 94 for the FIT.
Polyps
Sex-repartition wasn’t different between screening campaigns. There were 293 men and 176 women with polyps diagnosed by g-FOBT whereas 963 men and 542 women in FIT group (p=0.552).
There was a significant difference in types of screen-detected polyps between the two groups (p=0.008), with more adenomas and less serrated polyps in FIT group as compared to g-FOBT group (Table III). But, there was no significant difference in polyps localization (p=0.225) nor in their dysplasia degree (p=0.142). Polyps size was available only in 67.5% in g-FOBT group and in 80.9% in FIT group. Polyps size was significantly more important in FIT group (p<0.0001).
Tubular and/or villous Adenomas
FIT detected more tubulovillous and villous adenomas than g-FOBT (p<0.0001). Sizes were significantly different for adenomas between groups, with bigger sizes in FIT group (p<0.0001) as shown in Table IV.
Advanced Adenomas (AA) were defined by size ≥ 10mm, villous component or high-grade dysplasia (including In Situ Carcinomas). There were significantly more AA diagnosed by FIT (48.0%) versus g-FOBT (35.1%) (p<0.0001), without difference in AA localizations. Nevertheless, there was no significantly difference of ISC between the groups (p=0.591) (Table V).
In sex-specific analyses, types and size of polyps weren’t significantly different between the groups (Table VI). There wasn’t more AA in male (140 in g-FOBT and 612 in FIT group) than in female group (71 in g-FOBT group and 346 in FIT group) (p=0.498).
Serrated polyps
There were statistically less serrated polyps in FIT group than in g-FOBT (p=0.003), but no difference concerning adenomas size and localization (p=0.412) (Table VII). One serrated polyp was in high dysplasia grade in FIT group.
Invasive cancers
The detection rate for CRC was significantly higher in FIT group (0.31, 95%CI [0.27-0.35] versus g-FOBT (0.09, 95%CI [0.07-0.12]) (p<0.0001).
There were 63 invasive cancer diagnosed after a g-FOBT: 61 adenocarcinomas, 1 lymphoma and 1 neuro endocrine carcinoma, and 229 after a FIT: 227 adenocarcinomas and 2 neuroendocrine carcinomas. Sex repartition wasn’t different between the groups (40 and 149 cancers in men; 23 and 80 cancers in women respectively for g-FOBT and FIT) (p=0.817).
CRC localization, type of cancer, differentiation and stage of disease were not statistically different between the two screening campaigns (Tables VIII), even in sex-specific analyses (Tables IX).
Among the 229 CRC diagnosed after a positive FIT, 64.6% individuals had made a g-FOBT test 2 years before. As shown in Table X, there were statistically more right colon cancer in subjects who made a g-FOBT 2 years before versus subjects who had screening test for the first time (p=0.049), with no statistical difference between men (64.3%) and women (35.7%) compared to other localizations (p=0.840). However, there was no difference in the stages of CRC at screening between these groups (p=0.396).
|
Table des matières
Plan
LIST OF ABBREVIATIONS
ABSTRACT
INTRODUCTION
METHODS
RESULTS
1. Participation in screening campaigns
2. Characteristics of patients with screen-detected lesion according to screening campaign
3. Performance comparison between g-FOBT and FIT-based screening campaigns
4. Characteristics of screen-detected lesions
4.1. Polyps
4.1.1. Tubular and/or villous Adenomas
4.1.2. Advanced adenomas and In Situ Carcinomas
4.1.3. Serrated polyps
4.2. Invasive cancers
DISCUSSION AND CONCLUSION
BIBLIOGRAPHY
LIST OF FIGURES
LIST OF TABLES
ANNEX
Télécharger le rapport complet
