Research paper 109
Influence of sample return time and ambient temperature on the performance of an immunochemical faecal occult blood test with a new buffer for colorectal cancer screening Vincent Dancourta,b, Samia Hamzaa, Sylvain Manfredia, Antoine Drouillarda, Jeanne-Marie Bidanc, Jean Faivrea and Come Lepagea The haemoglobin concentration measured by faecal immunochemical tests (FIT) may be decreased in cases of delayed sample return or high temperature. It is an issue of great importance. The aim of this study was to investigate the effects of sample return time and of season on the performance of an FIT (FOB-Gold) with a new buffer. The study included 20 371 participants involved in the French organized colorectal cancer (CRC) screening programme. The probability of a positive screening test, detection rates and positive predictive values for CRC and advanced adenoma were analysed according to sample return time and season of screening. A sample of positive FIT was stored for 7 days in an incubator at 20°C or 30°C. The positivity rate was 4.1% for a sample return time of up to 3 days, 4.1% for 4–5 days and 4.6% for 6–7 days (P = 0.25). In multivariate analysis, there was no association between positivity rates, detection rates and positive predictive values for CRC and advanced adenoma and the sample return time or the season of screening. At a constant temperature of 20°C, there was a decrease in the
Introduction Colorectal cancer (CRC) meets the requirements for mass screening. It is a major cause of morbidity and mortality in industrialized countries and despite advances in treatment, survival rates remain poor. It can be treated successfully if detected at an early stage and potentially prevented by the removal of a precancerous lesion. In addition, there are effective screening strategies. In this context, population-based studies have shown that guaiac faecal occult blood testing, followed by a colonoscopy in case of positivity, can reduce CRC mortality (Hewitson et al., 2008). However, because the sensitivity of guaiacbased tests is fairly low and as they react to nonhuman haem in food, alternative strategies have been evaluated. A new generation of faecal occult blood tests, faecal immunochemical tests (FIT), is now available. It is clear from available data that FIT outperform the guaiac tests in both detection rates and participation rates (Guittet et al., 2007; Van Rossum et al., 2008; Hol et al., 2010; Faivre et al., 2012). Economic arguments in CRC screening lend further support to the view that FIT is the preferable strategy for average-risk populations (Heitman et al., 2010; Wilschut et al., 2011; Lejeune et al., 2014). FIT should therefore be preferred for CRC screening. Furthermore, automated quantitative FIT are easier to 0959-8278 Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
haemoglobin concentration of 5.1% after 7 days. The decrease reached 20.5% at a temperature of 30°C. It was only 4.5% during the first 4 days of storage in the incubator. With the new buffer, delay in sample return or season did not affect the clinical outcome. When temperatures reach 30°C, the faecal sample must be returned promptly. European Journal of Cancer Prevention 25:109–114 Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved. European Journal of Cancer Prevention 2016, 25:109–114 Keywords: colorectal cancer, faecal immunochemical test, sample stability, screening a Registre Bourguignon des cancers digestifs, INSERM U866; Université de Bourgogne; CHU de Dijon, bADECA 21-58 and cLaboratoire du CES de la CPAM Dijon, France
Correspondence to Jean Faivre, PhD, 7 boulevard Jeanne d’Arc, BP87900, 21000 Dijon, France Tel: + 33 380 393 340; fax: + 33 380 668 251; e-mail: [email protected] Received 23 July 2014 Accepted 18 February 2015
interpret, minimize human error in test processing and enable determination of the concentration of human haemoglobin in faecal samples. One limitation of FIT is that globin (the measured component of haemoglobin in these tests) is not stable in the transport buffer and is prone to degradation (Vilkin et al., 2005). Studies have suggested that high outside temperatures in summer or long delays between faecal sampling and laboratory analysis could decrease the concentration of haemoglobin (Van Rossum et al., 2009; Grazzini et al., 2010; Cha et al., 2012; Van Roon et al., 2012). Taking into account these results, some manufacturers have modified their buffers to increase globin stability. We used the second generation of the FOB-Gold buffer commercialized in December 2010, and the study started in January 2011. This paper is the first to report at the population level on the effect of sample return time and ambient temperature on the performance characteristics of a FIT using a new buffer in a population-based CRC screening programme.
Patients and methods Clinical study
France has an organized screening programme for CRC. Individuals aged 50–74 are invited to undergo a guaiac faecal occult blood test every 2 years. Pilot studies were DOI: 10.1097/CEJ.0000000000000153
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
110 European Journal of Cancer Prevention 2016, Vol 25 No 2
initiated in 2003 and screening was generalized in 2009. Between January 2011 and June 2012, within this programme, individuals living in well-defined areas of Burgundy were invited to undergo both the Hemoccult II (Beckman Coulter, Brea, California, USA) and an FIT: ‘FOB-Gold (Sentinel, Milan, Italy for Beckman Coulter). The FIT test performed in this study used a new buffer introduced in December 2010. For the FIT, each participant was provided with two FOB-Gold tubes and instructed to take a faecal sample from two consecutive stools. The completed tests were sent by mail to a central analysis centre. Samples were assayed by trained staff on the day they were received. After the start of this study, the French authorities decided that a one-sample strategy would be used. For this reason, the analysis took into account only the result of the second sample. An analyser provided by the company was used to quantify faecal haemoglobin in the buffer solution (AU 400; Beckman Coulter). The diagnostic yield of a one-sample FIT at different cut-off values has already been described (Hamza et al., 2013). Briefly, a total of 23 231 individuals aged 50–74 years underwent both a Hemoccult II test and an FIT (FOB-Gold). For 20 381 of these individuals, it was their fourth screening round and for 2850, it was their second screening round. The cut-off value chosen to define positivity was 100 ng/ml of buffer, which corresponds to 17 μg Hb/g of faeces (Fraser et al., 2012, 2013). The sampling date was written on the test tube. The participants were asked to post the FIT samples immediately. They were not asked to keep the samples refrigerated until they were posted. To take into account the manufacturer’s recommendations, the analysis was limited to the 20 371 individuals whose screening test was analysed within the 7 days following deposit of the faeces. Exclusion criteria were recent digestive symptoms, a history of CRC or adenoma, a first-degree relative who had CRC before 65 years of age and those with at least two first-degree relatives with inflammatory bowel disease whatever the age. A screening colonoscopy was recommended in these cases. Individuals with a normal colonoscopy during the previous 5 years or a severe illness contraindicating screening were also excluded. All potential participants received an information letter and an information brochure at the beginning of the screening round. Before this round, general practitioners (GPs) were informed thoroughly about the research project. During the first 6 months of the screening campaign, GPs offered the screening tests free of charge to the eligible individuals seen at their practices. For patients who did not complete the test after a visit to their GP, the coordination centre in charge of organizing the screening campaign subsequently mailed the test. A colonoscopy was indicated if the test proved positive. Preparation for the colonoscopy consisted of drinking 3–4 l of polyethylene glycol solution or a fleet phospho soda solution on the eve of the colonoscopy. Qualified
gastroenterologists practising either in public or in private hospitals performed the colonoscopies, most of the time using general anaesthesia. Pathologists provided copies of their reports. Cancers were classified according to the TNM classification (Sobin et al., 2009). Advanced adenoma was defined as an adenoma of 10 mm or more in size and/or with high-grade dysplasia and/or a villous component. If a patient had more than one adenoma, the most advanced lesion was included in the analysis. Sample return time was defined as the time in days between faecal sampling at home and delivery of the FIT to the laboratory. The mean sample return time with its SD was calculated. Positive tests were classified into three subgroups according to the sample return time: up to 3 days, 4–5 days and 6–7 days. An analysis was carried out to determine the effect of sex, age, urban–rural residence, participation in a previous screening round, season test was done and sample return time on the performance of the FIT. The probability of the FIT being positive and of diagnosing an advanced adenoma was studied using logistic regression. These analyses were not carried out for CRC because the number of cases was too small to carry out a reliable multivariable analysis. We also decided not to pool CRC and advanced adenoma because the result is too strongly influenced by advanced adenoma, which is the most frequent neoplastic lesion. The positivity rate was defined as the percentage of screened participants with a positive test. The colonoscopy rate was the percentage of positive test participants who underwent colonoscopy. The detection rate was defined as the number of participants with CRC or advanced adenoma per 1000 participants. The positive predictive value was calculated as the number of true positives with CRC or advanced adenoma relative to the total number of positives who underwent a colonoscopy. All these values were calculated with their 95% confidence intervals. Laboratory experiment
In total, 247 positive FIT samples were stored at 20°C or 30°C. Samples were received, measured and then used in the stability study. They were retested at intervals of 1 day during the first week. The mean decrease in faecal haemoglobin over time in the sample solution was measured. A linear mixed-effects model was used to estimate the mean percentage haemoglobin decrease per day (McCulloch et al., 2008). The risk of false negatives caused by the degradation of faecal haemoglobin was also estimated.
Results Clinical study
The mean sample return time was 4.4 ± 1.5 days (mean ± SD) and was similar for men and women. The
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Immunochemical faecal occult blood test Dancourt et al. 111
mean sample return time varied between 4.3 and 4.5 days according to the age group. Among the 20 371 screenees, 4.2% (n = 865) had a positive test result at a cut-off value of 100 ng/ml of buffer. Among these, 85.3% underwent a colonoscopy. Table 1 shows the percentage of positive screening tests according to sex, age, sample return time, season, history of screening and place of residence. There was no significant difference between the positivity rate with respect to the sample return time or the season. Sex and age were related to the positivity rate, whereas the place of residence and participation in a previous screening round were not. In a multivariate logistic regression analysis of factors associated with the probability of a positive screening test, no relationship was observed with sample return time or with season. There was a significantly higher positivity rate among men than women and with increasing age. Among positive screenees who underwent a colonoscopy, 47 were diagnosed with CRC and 268 were diagnosed with advanced adenoma. Table 2 shows the detection rate and the positive predictive value for the different sample return times. The detection rate of CRC was 2.1, 2.3 and 2.5% for sample return times of up to 3 days, 4–5 days and 6–7 days, respectively (P = 0.91). For advanced adenoma, the corresponding detection rates were 14.7, 11.3 and 14.3%, respectively (P = 0.14). The positive predictive value for cancer was 6.4% for a sample return time of up to 3 days versus 6.2% for a sample
Percentage of positive screening tests by age, sex, sample return time, season, history of screening and place of residence
Table 1
Univariate analysis n
Positive (%)
95% CI
Sex Men 9285 5.9 5.4–6.3 Women 11 096 2.9 2.6–3.2 Age 50–54 3468 2.8 2.3–3.4 55–59 4586 3.4 2.9–3.9 60–64 5511 4.7 4.1–5.2 65–69 3824 4.6 3.9–5.3 70–74 2982 6 5.1–6.8 Sample return time (days) ≤3 6130 4.1 3.6–4.6 4–5 8643 4.1 3.7–4.5 6–7 5598 4.6 4.1–5.2 Season Winter 4797 4.3 3.7–4.9 Spring 8488 4.1 3.7–4.6 Summer 4818 4.5 4.0–5.1 Autumn 2268 3.9 3.1–4.7 Participation in a previous screening round No 3894 3.8 3.2–4.4 Yes 16 477 4.3 4.0–4.6 Place of residence Urban 8411 4.1 3.6–4.5 Rural 11 946 4.4 4.0–4.7 CI, confidence interval; OR, odds ratio.
Multivariate analysis OR
95% CI
P
1 0.5
0.4–0.6
< 0.001
1 1.2 1.7 1.7 2.2
0.9–1.6 1.3–2.1 1.3–2.2 1.7–2.9
0.16 < 0.001 < 0.001 < 0.001
1 0.9 1.1
0.8–1.2 0.9–1.4
0.90 0.15
1 0.9 1.1 0.9
0.8–1.1 0.9–1.3 0.7–1.2
0.40 0.57 0.55
1 1.0
0.8–1.3
0.84
1 1.1
0.9–1.3
0.30
return time of 6–7 days. For advanced adenoma, it was 44.1 and 35.2%, respectively. No statistically significant correlation was found between the positive predictive values for CRC (P = 0.43) or advanced adenoma (P = 0.93) and season (Table 2). Table 3 shows the probability of detecting advanced adenoma. After adjustment for other studied variables, neither the sample return time nor the season of screening was associated significantly with the detection rate. Laboratory experiment
In total, 124 positive FIT samples were stored at 20°C and 123 at 30°C. The samples retested after 7 days in an incubator corresponded to faecal sampling performed 9–14 days previously. The initial haemoglobin concentration of the selected tests varied between 103 and 1064 ng/ml for samples stored at 20°C and between 104 and 1023 ng/ml for samples stored at 30°C. During storage at 20°C, the mean haemoglobin concentration in the faecal sample decreased on average by 1.1 ng/ml/day, corresponding to a decrease of 0.7%/day (Fig. 1). When samples were stored in an incubator at +30°C, the mean haemoglobin level decreased by 14.0 ng/ml/day. The cumulative decrease in the haemoglobin concentration was 20.5% after 7 days. After 1 week of storage at 20°C, 10.0% of the samples became negative, whereas 23.8% of the samples became negative after storage at 30°C (Fig. 1). After 7 days at a constant temperature of 20°C, 100% of the samples of participants with CRC (4/4) and 97% of samples of participants with advanced adenomas (38/39) remained positive. After correction for sample return time, it was only after 10 days that the haemoglobin concentration for the advanced adenoma decreased below the 100 ng/ml cut-off level at a temperature of 20°C. After 7 days at a constant temperature of 30°C, 78% of the samples corresponding to CRC (7/9) and 78% from those with advanced adenoma (29/37) remained positive. The two samples from participants with CRC became negative after 7 days of storage at a temperature of 30°C. For advanced adenoma, the samples became negative after 1–6 days of storage at 30°C. At 30°C, the samples were relatively stable during the first 4 days with respect to the tests sampled at home 6–11 days earlier. During storage at 30°C, the cumulative decrease in the haemoglobin concentration was 4.5% during the first 4 days after sample analysis and 18.5% during the period of 5–7 days.
Discussion This study describes the relationship between the positivity rate and the diagnostic yield of a second generation of FIT, which uses an improved buffer, with particular attention paid to sample return time, season of screening
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
112 European Journal of Cancer Prevention 2016, Vol 25 No 2
Detection rate and positive predictive value for CRC and advanced adenoma for the different sample return times and season of screening
Table 2
Detection rate (%) Number of colonoscopies
Positive predictive value (%)
Cancer
95% CI
Advanced adenoma
95% CI
Cancer
95% CI
Advanced adenoma
95% CI
2.1 2.3 2.5
0.9–3.2 1.3–3.2 1.2–3.8
14.7 11.3 14.3
11.7–17.7 9.1–13.6 11.2–17.4
6.4 6.5 6.2
3.0–9.8 3.7–9.3 3.1–9.3
44.1 31.9 35.2
37.3–50.9 26.7–37.1 29.0–41.4
1.7 2.9 1.9 2.2
0.5–2.8 1.8–4.1 0.6–3.1 0.3–4.1
12.9 12.7 13.7 14.1
9.7–16.1 10.3–15.1 10.4–17.0 9.2–19.0
5 7.9 4.8 6.9
1.6–8.4 4.9–10.9 1.8–7.8 1.1–12.7
38.7 34.2 34.9 43.8
31.2–46.3 29.0–39.4 28.1–41.7 32.4–55.2
Sample return time (days) ≤3 204 4–5 307 6–7 227 Season Winter 160 Spring 316 Summer 189 Autumn 73
CI, confidence interval; CRC, colorectal cancer.
Logistic regression of the risk of having an advanced adenoma detected
Table 3
n
Odds ratio
Sex Men 9285 1 Women 11 096 0.4 Age 50–54 3468 1 55–59 4586 1.7 60–64 5511 2.5 65–69 3824 2.4 70–74 2982 2.7 Sample return time (days) ≤3 6130 1 4–5 8643 0.8 6–7 5598 1 Season Winter 4797 1 Spring 8488 0.9 Summer 4818 1.0 Autumn 2268 1.1 Participation in a previous screening round Yes 3894 1 No 16 477 1.3 Place of residence Urban 8411 1 Rural 11 946 1.0
95% CI
P
0.3–0.5
< 0.001
1.0–2.8 1.5–4.1 1.4–4.1 1.6–4.5
0.05 < 0.001 0.001 < 0.001
0.6–1.1 0.7–1.3
0.06 0.93
0.7–1.5 0.7–1.5 0.7–1.8
0.55 0.83 0.53
0.9–1.8
0.24
0.8–1.3
0.85
CI, confidence interval.
and outside temperatures. The main strength of this study is the inclusion of a large number of screenees in an asymptomatic average-risk population involved in an organized screening programme for CRC in France. This is the first study to be carried out using an FIT (FOBGold) with a new buffer developed by the company after the suggestion that tests performed in summer or samples with long return times might yield false negatives because of haemoglobin degradation (Van Rossum et al., 2009; Grazzini et al., 2010; Cha et al., 2012; Van Roon et al., 2012). Our results differ from those reported with the first-generation buffers. A study from Italy showed that the probability of detecting CRC and advanced adenoma was 13% lower in the summer than in the winter (Grazzini et al., 2010). This result was related to differences in the ambient temperature. The data reported in this study suggest that this is no longer a problem. A study from Korea indicated that an ambient
temperature over 25°C was not a significant risk factor for positive FIT results or the detection of colorectal neoplasms (Cha et al., 2012). A previous study carried out in Nijmegen and Amsterdam found that a delay over 5 days between faecal sampling and laboratory delivery at room temperature decreased the performance of the FIT (Van Rossum et al., 2009). These results were not confirmed by another Dutch study (Van Roon et al., 2012) and an Israeli study (Levi et al., 2007). In the Israeli study, no significant haemoglobin degradation was observed when samples were stored at 20°C for 21 days or at room temperature and for 10 days as shown in the Dutch study (Van Roon et al., 2012). Our results show that the positivity rate, the detection rate and the positive predictive value of CRC and advanced adenoma did not decrease when samples had been stored for up to 7 days at 20°C. It can be concluded that haemoglobin degradation at a temperature of 20°C is limited, in particular with the new buffer used in our study. However, at temperatures of 30° C and above, haemoglobin deterioration is more marked. A decrease in the haemoglobin concentration of 3.7%/day was observed in Israel with the first generation of buffer when the test was kept at a temperature of 28°C on average (Levi et al., 2007). In-vitro results have been confirmed by in-vivo studies. In a previous laboratory study (Guittet et al., 2011), stool samples from 10 healthy volunteers were supplemented with human blood to compare the stability of measurement at varying storage temperatures for three FITs. At 20°C, the decrease in the haemoglobin concentration was 1.7% daily with the OC-Sensor and 7.8% with FOBGold. After modification of the buffer, our results indicate that the haemoglobin concentration at 20°C with FOB-Gold deceased on average by 0.7%/day. In another study based on a pool of 15 positive faecal samples, the percentage of cumulative faecal haemoglobin decrease was significantly lower with the new FOB-Gold buffer compared with the previous one (Gnatta et al., 2014). It can be concluded that more effective ways have been found to improve sample stability. However, a decrease in the haemoglobin concentration was observed after 4 days in an oven when the test was kept at a permanent
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Immunochemical faecal occult blood test Dancourt et al. 113
Fig. 1
Haemoglobin concentration
Incubator = 20°C
Incubator = 30°C
100 90 80 70 60 50 40 30 20 10 0
100 90 80 70 60 50 40 30 20 10 0
1
2
3
4 Days
5
6
7
1
2
3
95% CI
4 Days
5
6
7
95% CI
Cumulative decrease in haemoglobin concentration according to storage temperature. CI, confidence interval.
temperature of 30°C. It is recommended that faecal samples should be posted to the interpretation laboratory immediately after sample collection and tested immediately on receipt, in particular when the outside temperature is above 20°C. The current organization in France allows this rapid testing. In cases of delay before transportation, the samples should be stored in a refrigerator (Grazzini et al., 2010; Guittet et al., 2011; Gnatta et al., 2014). The haemoglobin level in samples from individuals with advanced adenoma is lower than that in samples from individuals with CRC. The delayed return of samples from the former could therefore more frequently lead to a false-negative test than is the case for CRC. However, it does not seem to be a frequent problem at a temperature of 20°C as the test becomes negative for 3% of advanced adenomas. After 7 days of storage at 30°C, 24% of the positive tests became negative. The data from this study have important implications for the organization of FIT-based screening programmes. Delay in sample return and season did not affect the diagnostic yield of the FIT. An extra 7 days at a temperature of 20°C after sample interpretation, which corresponds to 9–14 days after sample collection, is acceptable. In this condition, globin degradation is not an obstacle to the implementation of FIT screening. However, at a temperature of 30°C and above, the decrease in the haemoglobin concentration was significant after 4 days. Permanent exposure to such a high temperature is rare. However, individuals invited to take part in CRC screening should be informed that it is essential to return the faecal sample promptly and to delay the screening test if the outside temperature is high.
Acknowledgements The authors thank the GPs, gastroenterologists and pathologists for their active participation. This study was part of the research project of team 5 ‘epidemiology and clinical research on digestive cancers’ of the Dijon centre of the National Institute for Medical Research (INSERM U866). Conflicts of interest
There are no conflicts of interest.
References Cha JM, Lee JI, Joo KR, Shin HP, Park JJ, Jeun JW, et al. (2012). Performance of the fecal immunochemical test is not decreased by high ambient temperature in the rapid return system. Dig Dis Sci 57:2178–2183. Faivre J, Dancourt V, Denis B, Dorval E, Piette C, Perrin P, et al. (2012). Comparison between a guaiac and three immunochemical faecal occult blood tests in screening for colorectal cancer. Eur J Cancer 48:2969–2976. Fraser CG, Allison JE, Halloran SP, Young GP (2012). A proposal to standardize reporting units for fecal immunochemical tests for hemoglobin. J Natl Cancer Inst 104:810–814. Fraser CG, Halloran SP, Allison JE, Young GP (2013). Making colorectal cancer screening FITTER for purpose with quantitative faecal immunochemical tests for haemoglobin (FIT). Clin Chem Lab Med 51:2065–2067. Gnatta E, Zaninotto M, Epifani MG, Padoan A, Gjini R, Plebani M (2014). A new sampling device for faecal immunochemical testing: haemoglobin stability is still an open issue. Clin Chem Lab Med 52:1203–1209. Grazzini G, Ventura L, Zappa M, Ciatto S, Confortini M, Rapi S, et al. (2010). Influence of seasonal variations in ambient temperatures on performance of immunochemical faecal occult blood test for colorectal cancer screening: observational study from the Florence district. Gut 59:1511–1515. Guittet L, Bouvier V, Mariotte N, Vallee JP, Arsene D, Boutreux S, et al. (2007). Comparison of a guaiac based and an immunochemical faecal occult blood test in screening for colorectal cancer in a general average risk population. Gut 56:210–214. Guittet L, Guillaume E, Levillain R, Beley P, Tichet J, Lantieri O, Launoy G (2011). Analytical comparison of three quantitative immunochemical fecal occult blood tests for colorectal cancer screening. Cancer Epidemiol Biomarkers Prev 20:1492–1501. Hamza S, Dancourt V, Lejeune C, Bidan JM, Lepage C, Faivre J (2013). Diagnostic yield of a one sample immunochemical test at different cut-off
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
114 European Journal of Cancer Prevention 2016, Vol 25 No 2
values in an organised screening programme for colorectal cancer. Eur J Cancer 49:2727–2733. Heitman SJ, Hilsden RJ, Au F, Dowden S, Manns BJ (2010). Colorectal cancer screening for average-risk North Americans: an economic evaluation. PLoS Med 7:e1000370. Hewitson P, Glasziou P, Watson E, Towler B, Irwig L (2008). Cochrane systematic review of colorectal cancer screening using the fecal occult blood test (Hemoccult): an update. Am J Gastroenterol 103:1541–1549. Hol L, van Leerdam ME, van Ballegooijen M, van Vuuren AJ, van Dekken H, Reijerink JC, et al. (2010). Screening for colorectal cancer: randomised trial comparing guaiac-based and immunochemical faecal occult blood testing and flexible sigmoidoscopy. Gut 59:62–68. Lejeune C, Le Gleut K, Cottet V, Galimard C, Durand G, Dancourt V, et al. (2014). The cost-effectiveness of immunochemical tests for colorectal cancer screening. Dig Liver Dis 46:76–81. Levi Z, Rozen P, Hazazi R, Vilkin A, Waked A, Maoz E, et al. (2007). A quantitative immunochemical fecal occult blood test for colorectal neoplasia. Ann Intern Med 146:244–255. McCulloch C, Searle S, Neuhaus J (2008). Generalized, linear, and mixed models, 2nd ed. Chichester: Wiley.
Sobin L, Gospodarowicz M, Wittekind C (2009). TNM classification of malignant tumours, 7th ed. Oxford: Wiley-Blackwell. pp. 19–20. Van Roon AH, Hol L, van Vuuren AJ, Francke J, Ouwendijk M, Heijens A, et al. (2012). Are fecal immunochemical test characteristics influenced by sample return time? A population-based colorectal cancer screening trial. Am J Gastroenterol 107:99–107. Van Rossum LG, van Rijn AF, Laheij RJ, van Oijen MG, Fockens P, van Krieken HH, et al. (2008). Random comparison of guaiac and immunochemical fecal occult blood tests for colorectal cancer in a screening population. Gastroenterology 135:82–90. Van Rossum LG, van Rijn AF, van Oijen MG, Fockens P, Laheij RJ, Verbeek AL, et al. (2009). False negative fecal occult blood tests due to delayed sample return in colorectal cancer screening. Int J Cancer 125:746–750. Vilkin A, Rozen P, Levi Z, Waked A, Maoz E, Birkenfeld S, Niv Y (2005). Performance characteristics and evaluation of an automated-developed and quantitative, immunochemical, fecal occult blood screening test. Am J Gastroenterol 100:2519–2525. Wilschut JA, Hol L, Dekker E, Jansen JB, Van Leerdam ME, Lansdorp-Vogelaar I, et al. (2011). Cost-effectiveness analysis of a quantitative immunochemical test for colorectal cancer screening. Gastroenterology 141:1648.e1–1655.e1.
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Influence of sample return time and ambient temperature on the performance of an immunochemical faecal occult blood test with a new buffer for colorectal cancer screening Vincent Dancourta,b, Samia Hamzaa, Sylvain Manfredia, Antoine Drouillarda, Jeanne-Marie Bidanc, Jean Faivrea and Come Lepagea The haemoglobin concentration measured by faecal immunochemical tests (FIT) may be decreased in cases of delayed sample return or high temperature. It is an issue of great importance. The aim of this study was to investigate the effects of sample return time and of season on the performance of an FIT (FOB-Gold) with a new buffer. The study included 20 371 participants involved in the French organized colorectal cancer (CRC) screening programme. The probability of a positive screening test, detection rates and positive predictive values for CRC and advanced adenoma were analysed according to sample return time and season of screening. A sample of positive FIT was stored for 7 days in an incubator at 20°C or 30°C. The positivity rate was 4.1% for a sample return time of up to 3 days, 4.1% for 4–5 days and 4.6% for 6–7 days (P = 0.25). In multivariate analysis, there was no association between positivity rates, detection rates and positive predictive values for CRC and advanced adenoma and the sample return time or the season of screening. At a constant temperature of 20°C, there was a decrease in the
Introduction Colorectal cancer (CRC) meets the requirements for mass screening. It is a major cause of morbidity and mortality in industrialized countries and despite advances in treatment, survival rates remain poor. It can be treated successfully if detected at an early stage and potentially prevented by the removal of a precancerous lesion. In addition, there are effective screening strategies. In this context, population-based studies have shown that guaiac faecal occult blood testing, followed by a colonoscopy in case of positivity, can reduce CRC mortality (Hewitson et al., 2008). However, because the sensitivity of guaiacbased tests is fairly low and as they react to nonhuman haem in food, alternative strategies have been evaluated. A new generation of faecal occult blood tests, faecal immunochemical tests (FIT), is now available. It is clear from available data that FIT outperform the guaiac tests in both detection rates and participation rates (Guittet et al., 2007; Van Rossum et al., 2008; Hol et al., 2010; Faivre et al., 2012). Economic arguments in CRC screening lend further support to the view that FIT is the preferable strategy for average-risk populations (Heitman et al., 2010; Wilschut et al., 2011; Lejeune et al., 2014). FIT should therefore be preferred for CRC screening. Furthermore, automated quantitative FIT are easier to 0959-8278 Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved.
haemoglobin concentration of 5.1% after 7 days. The decrease reached 20.5% at a temperature of 30°C. It was only 4.5% during the first 4 days of storage in the incubator. With the new buffer, delay in sample return or season did not affect the clinical outcome. When temperatures reach 30°C, the faecal sample must be returned promptly. European Journal of Cancer Prevention 25:109–114 Copyright © 2016 Wolters Kluwer Health, Inc. All rights reserved. European Journal of Cancer Prevention 2016, 25:109–114 Keywords: colorectal cancer, faecal immunochemical test, sample stability, screening a Registre Bourguignon des cancers digestifs, INSERM U866; Université de Bourgogne; CHU de Dijon, bADECA 21-58 and cLaboratoire du CES de la CPAM Dijon, France
Correspondence to Jean Faivre, PhD, 7 boulevard Jeanne d’Arc, BP87900, 21000 Dijon, France Tel: + 33 380 393 340; fax: + 33 380 668 251; e-mail: [email protected] Received 23 July 2014 Accepted 18 February 2015
interpret, minimize human error in test processing and enable determination of the concentration of human haemoglobin in faecal samples. One limitation of FIT is that globin (the measured component of haemoglobin in these tests) is not stable in the transport buffer and is prone to degradation (Vilkin et al., 2005). Studies have suggested that high outside temperatures in summer or long delays between faecal sampling and laboratory analysis could decrease the concentration of haemoglobin (Van Rossum et al., 2009; Grazzini et al., 2010; Cha et al., 2012; Van Roon et al., 2012). Taking into account these results, some manufacturers have modified their buffers to increase globin stability. We used the second generation of the FOB-Gold buffer commercialized in December 2010, and the study started in January 2011. This paper is the first to report at the population level on the effect of sample return time and ambient temperature on the performance characteristics of a FIT using a new buffer in a population-based CRC screening programme.
Patients and methods Clinical study
France has an organized screening programme for CRC. Individuals aged 50–74 are invited to undergo a guaiac faecal occult blood test every 2 years. Pilot studies were DOI: 10.1097/CEJ.0000000000000153
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
110 European Journal of Cancer Prevention 2016, Vol 25 No 2
initiated in 2003 and screening was generalized in 2009. Between January 2011 and June 2012, within this programme, individuals living in well-defined areas of Burgundy were invited to undergo both the Hemoccult II (Beckman Coulter, Brea, California, USA) and an FIT: ‘FOB-Gold (Sentinel, Milan, Italy for Beckman Coulter). The FIT test performed in this study used a new buffer introduced in December 2010. For the FIT, each participant was provided with two FOB-Gold tubes and instructed to take a faecal sample from two consecutive stools. The completed tests were sent by mail to a central analysis centre. Samples were assayed by trained staff on the day they were received. After the start of this study, the French authorities decided that a one-sample strategy would be used. For this reason, the analysis took into account only the result of the second sample. An analyser provided by the company was used to quantify faecal haemoglobin in the buffer solution (AU 400; Beckman Coulter). The diagnostic yield of a one-sample FIT at different cut-off values has already been described (Hamza et al., 2013). Briefly, a total of 23 231 individuals aged 50–74 years underwent both a Hemoccult II test and an FIT (FOB-Gold). For 20 381 of these individuals, it was their fourth screening round and for 2850, it was their second screening round. The cut-off value chosen to define positivity was 100 ng/ml of buffer, which corresponds to 17 μg Hb/g of faeces (Fraser et al., 2012, 2013). The sampling date was written on the test tube. The participants were asked to post the FIT samples immediately. They were not asked to keep the samples refrigerated until they were posted. To take into account the manufacturer’s recommendations, the analysis was limited to the 20 371 individuals whose screening test was analysed within the 7 days following deposit of the faeces. Exclusion criteria were recent digestive symptoms, a history of CRC or adenoma, a first-degree relative who had CRC before 65 years of age and those with at least two first-degree relatives with inflammatory bowel disease whatever the age. A screening colonoscopy was recommended in these cases. Individuals with a normal colonoscopy during the previous 5 years or a severe illness contraindicating screening were also excluded. All potential participants received an information letter and an information brochure at the beginning of the screening round. Before this round, general practitioners (GPs) were informed thoroughly about the research project. During the first 6 months of the screening campaign, GPs offered the screening tests free of charge to the eligible individuals seen at their practices. For patients who did not complete the test after a visit to their GP, the coordination centre in charge of organizing the screening campaign subsequently mailed the test. A colonoscopy was indicated if the test proved positive. Preparation for the colonoscopy consisted of drinking 3–4 l of polyethylene glycol solution or a fleet phospho soda solution on the eve of the colonoscopy. Qualified
gastroenterologists practising either in public or in private hospitals performed the colonoscopies, most of the time using general anaesthesia. Pathologists provided copies of their reports. Cancers were classified according to the TNM classification (Sobin et al., 2009). Advanced adenoma was defined as an adenoma of 10 mm or more in size and/or with high-grade dysplasia and/or a villous component. If a patient had more than one adenoma, the most advanced lesion was included in the analysis. Sample return time was defined as the time in days between faecal sampling at home and delivery of the FIT to the laboratory. The mean sample return time with its SD was calculated. Positive tests were classified into three subgroups according to the sample return time: up to 3 days, 4–5 days and 6–7 days. An analysis was carried out to determine the effect of sex, age, urban–rural residence, participation in a previous screening round, season test was done and sample return time on the performance of the FIT. The probability of the FIT being positive and of diagnosing an advanced adenoma was studied using logistic regression. These analyses were not carried out for CRC because the number of cases was too small to carry out a reliable multivariable analysis. We also decided not to pool CRC and advanced adenoma because the result is too strongly influenced by advanced adenoma, which is the most frequent neoplastic lesion. The positivity rate was defined as the percentage of screened participants with a positive test. The colonoscopy rate was the percentage of positive test participants who underwent colonoscopy. The detection rate was defined as the number of participants with CRC or advanced adenoma per 1000 participants. The positive predictive value was calculated as the number of true positives with CRC or advanced adenoma relative to the total number of positives who underwent a colonoscopy. All these values were calculated with their 95% confidence intervals. Laboratory experiment
In total, 247 positive FIT samples were stored at 20°C or 30°C. Samples were received, measured and then used in the stability study. They were retested at intervals of 1 day during the first week. The mean decrease in faecal haemoglobin over time in the sample solution was measured. A linear mixed-effects model was used to estimate the mean percentage haemoglobin decrease per day (McCulloch et al., 2008). The risk of false negatives caused by the degradation of faecal haemoglobin was also estimated.
Results Clinical study
The mean sample return time was 4.4 ± 1.5 days (mean ± SD) and was similar for men and women. The
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Immunochemical faecal occult blood test Dancourt et al. 111
mean sample return time varied between 4.3 and 4.5 days according to the age group. Among the 20 371 screenees, 4.2% (n = 865) had a positive test result at a cut-off value of 100 ng/ml of buffer. Among these, 85.3% underwent a colonoscopy. Table 1 shows the percentage of positive screening tests according to sex, age, sample return time, season, history of screening and place of residence. There was no significant difference between the positivity rate with respect to the sample return time or the season. Sex and age were related to the positivity rate, whereas the place of residence and participation in a previous screening round were not. In a multivariate logistic regression analysis of factors associated with the probability of a positive screening test, no relationship was observed with sample return time or with season. There was a significantly higher positivity rate among men than women and with increasing age. Among positive screenees who underwent a colonoscopy, 47 were diagnosed with CRC and 268 were diagnosed with advanced adenoma. Table 2 shows the detection rate and the positive predictive value for the different sample return times. The detection rate of CRC was 2.1, 2.3 and 2.5% for sample return times of up to 3 days, 4–5 days and 6–7 days, respectively (P = 0.91). For advanced adenoma, the corresponding detection rates were 14.7, 11.3 and 14.3%, respectively (P = 0.14). The positive predictive value for cancer was 6.4% for a sample return time of up to 3 days versus 6.2% for a sample
Percentage of positive screening tests by age, sex, sample return time, season, history of screening and place of residence
Table 1
Univariate analysis n
Positive (%)
95% CI
Sex Men 9285 5.9 5.4–6.3 Women 11 096 2.9 2.6–3.2 Age 50–54 3468 2.8 2.3–3.4 55–59 4586 3.4 2.9–3.9 60–64 5511 4.7 4.1–5.2 65–69 3824 4.6 3.9–5.3 70–74 2982 6 5.1–6.8 Sample return time (days) ≤3 6130 4.1 3.6–4.6 4–5 8643 4.1 3.7–4.5 6–7 5598 4.6 4.1–5.2 Season Winter 4797 4.3 3.7–4.9 Spring 8488 4.1 3.7–4.6 Summer 4818 4.5 4.0–5.1 Autumn 2268 3.9 3.1–4.7 Participation in a previous screening round No 3894 3.8 3.2–4.4 Yes 16 477 4.3 4.0–4.6 Place of residence Urban 8411 4.1 3.6–4.5 Rural 11 946 4.4 4.0–4.7 CI, confidence interval; OR, odds ratio.
Multivariate analysis OR
95% CI
P
1 0.5
0.4–0.6
< 0.001
1 1.2 1.7 1.7 2.2
0.9–1.6 1.3–2.1 1.3–2.2 1.7–2.9
0.16 < 0.001 < 0.001 < 0.001
1 0.9 1.1
0.8–1.2 0.9–1.4
0.90 0.15
1 0.9 1.1 0.9
0.8–1.1 0.9–1.3 0.7–1.2
0.40 0.57 0.55
1 1.0
0.8–1.3
0.84
1 1.1
0.9–1.3
0.30
return time of 6–7 days. For advanced adenoma, it was 44.1 and 35.2%, respectively. No statistically significant correlation was found between the positive predictive values for CRC (P = 0.43) or advanced adenoma (P = 0.93) and season (Table 2). Table 3 shows the probability of detecting advanced adenoma. After adjustment for other studied variables, neither the sample return time nor the season of screening was associated significantly with the detection rate. Laboratory experiment
In total, 124 positive FIT samples were stored at 20°C and 123 at 30°C. The samples retested after 7 days in an incubator corresponded to faecal sampling performed 9–14 days previously. The initial haemoglobin concentration of the selected tests varied between 103 and 1064 ng/ml for samples stored at 20°C and between 104 and 1023 ng/ml for samples stored at 30°C. During storage at 20°C, the mean haemoglobin concentration in the faecal sample decreased on average by 1.1 ng/ml/day, corresponding to a decrease of 0.7%/day (Fig. 1). When samples were stored in an incubator at +30°C, the mean haemoglobin level decreased by 14.0 ng/ml/day. The cumulative decrease in the haemoglobin concentration was 20.5% after 7 days. After 1 week of storage at 20°C, 10.0% of the samples became negative, whereas 23.8% of the samples became negative after storage at 30°C (Fig. 1). After 7 days at a constant temperature of 20°C, 100% of the samples of participants with CRC (4/4) and 97% of samples of participants with advanced adenomas (38/39) remained positive. After correction for sample return time, it was only after 10 days that the haemoglobin concentration for the advanced adenoma decreased below the 100 ng/ml cut-off level at a temperature of 20°C. After 7 days at a constant temperature of 30°C, 78% of the samples corresponding to CRC (7/9) and 78% from those with advanced adenoma (29/37) remained positive. The two samples from participants with CRC became negative after 7 days of storage at a temperature of 30°C. For advanced adenoma, the samples became negative after 1–6 days of storage at 30°C. At 30°C, the samples were relatively stable during the first 4 days with respect to the tests sampled at home 6–11 days earlier. During storage at 30°C, the cumulative decrease in the haemoglobin concentration was 4.5% during the first 4 days after sample analysis and 18.5% during the period of 5–7 days.
Discussion This study describes the relationship between the positivity rate and the diagnostic yield of a second generation of FIT, which uses an improved buffer, with particular attention paid to sample return time, season of screening
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
112 European Journal of Cancer Prevention 2016, Vol 25 No 2
Detection rate and positive predictive value for CRC and advanced adenoma for the different sample return times and season of screening
Table 2
Detection rate (%) Number of colonoscopies
Positive predictive value (%)
Cancer
95% CI
Advanced adenoma
95% CI
Cancer
95% CI
Advanced adenoma
95% CI
2.1 2.3 2.5
0.9–3.2 1.3–3.2 1.2–3.8
14.7 11.3 14.3
11.7–17.7 9.1–13.6 11.2–17.4
6.4 6.5 6.2
3.0–9.8 3.7–9.3 3.1–9.3
44.1 31.9 35.2
37.3–50.9 26.7–37.1 29.0–41.4
1.7 2.9 1.9 2.2
0.5–2.8 1.8–4.1 0.6–3.1 0.3–4.1
12.9 12.7 13.7 14.1
9.7–16.1 10.3–15.1 10.4–17.0 9.2–19.0
5 7.9 4.8 6.9
1.6–8.4 4.9–10.9 1.8–7.8 1.1–12.7
38.7 34.2 34.9 43.8
31.2–46.3 29.0–39.4 28.1–41.7 32.4–55.2
Sample return time (days) ≤3 204 4–5 307 6–7 227 Season Winter 160 Spring 316 Summer 189 Autumn 73
CI, confidence interval; CRC, colorectal cancer.
Logistic regression of the risk of having an advanced adenoma detected
Table 3
n
Odds ratio
Sex Men 9285 1 Women 11 096 0.4 Age 50–54 3468 1 55–59 4586 1.7 60–64 5511 2.5 65–69 3824 2.4 70–74 2982 2.7 Sample return time (days) ≤3 6130 1 4–5 8643 0.8 6–7 5598 1 Season Winter 4797 1 Spring 8488 0.9 Summer 4818 1.0 Autumn 2268 1.1 Participation in a previous screening round Yes 3894 1 No 16 477 1.3 Place of residence Urban 8411 1 Rural 11 946 1.0
95% CI
P
0.3–0.5
< 0.001
1.0–2.8 1.5–4.1 1.4–4.1 1.6–4.5
0.05 < 0.001 0.001 < 0.001
0.6–1.1 0.7–1.3
0.06 0.93
0.7–1.5 0.7–1.5 0.7–1.8
0.55 0.83 0.53
0.9–1.8
0.24
0.8–1.3
0.85
CI, confidence interval.
and outside temperatures. The main strength of this study is the inclusion of a large number of screenees in an asymptomatic average-risk population involved in an organized screening programme for CRC in France. This is the first study to be carried out using an FIT (FOBGold) with a new buffer developed by the company after the suggestion that tests performed in summer or samples with long return times might yield false negatives because of haemoglobin degradation (Van Rossum et al., 2009; Grazzini et al., 2010; Cha et al., 2012; Van Roon et al., 2012). Our results differ from those reported with the first-generation buffers. A study from Italy showed that the probability of detecting CRC and advanced adenoma was 13% lower in the summer than in the winter (Grazzini et al., 2010). This result was related to differences in the ambient temperature. The data reported in this study suggest that this is no longer a problem. A study from Korea indicated that an ambient
temperature over 25°C was not a significant risk factor for positive FIT results or the detection of colorectal neoplasms (Cha et al., 2012). A previous study carried out in Nijmegen and Amsterdam found that a delay over 5 days between faecal sampling and laboratory delivery at room temperature decreased the performance of the FIT (Van Rossum et al., 2009). These results were not confirmed by another Dutch study (Van Roon et al., 2012) and an Israeli study (Levi et al., 2007). In the Israeli study, no significant haemoglobin degradation was observed when samples were stored at 20°C for 21 days or at room temperature and for 10 days as shown in the Dutch study (Van Roon et al., 2012). Our results show that the positivity rate, the detection rate and the positive predictive value of CRC and advanced adenoma did not decrease when samples had been stored for up to 7 days at 20°C. It can be concluded that haemoglobin degradation at a temperature of 20°C is limited, in particular with the new buffer used in our study. However, at temperatures of 30° C and above, haemoglobin deterioration is more marked. A decrease in the haemoglobin concentration of 3.7%/day was observed in Israel with the first generation of buffer when the test was kept at a temperature of 28°C on average (Levi et al., 2007). In-vitro results have been confirmed by in-vivo studies. In a previous laboratory study (Guittet et al., 2011), stool samples from 10 healthy volunteers were supplemented with human blood to compare the stability of measurement at varying storage temperatures for three FITs. At 20°C, the decrease in the haemoglobin concentration was 1.7% daily with the OC-Sensor and 7.8% with FOBGold. After modification of the buffer, our results indicate that the haemoglobin concentration at 20°C with FOB-Gold deceased on average by 0.7%/day. In another study based on a pool of 15 positive faecal samples, the percentage of cumulative faecal haemoglobin decrease was significantly lower with the new FOB-Gold buffer compared with the previous one (Gnatta et al., 2014). It can be concluded that more effective ways have been found to improve sample stability. However, a decrease in the haemoglobin concentration was observed after 4 days in an oven when the test was kept at a permanent
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Immunochemical faecal occult blood test Dancourt et al. 113
Fig. 1
Haemoglobin concentration
Incubator = 20°C
Incubator = 30°C
100 90 80 70 60 50 40 30 20 10 0
100 90 80 70 60 50 40 30 20 10 0
1
2
3
4 Days
5
6
7
1
2
3
95% CI
4 Days
5
6
7
95% CI
Cumulative decrease in haemoglobin concentration according to storage temperature. CI, confidence interval.
temperature of 30°C. It is recommended that faecal samples should be posted to the interpretation laboratory immediately after sample collection and tested immediately on receipt, in particular when the outside temperature is above 20°C. The current organization in France allows this rapid testing. In cases of delay before transportation, the samples should be stored in a refrigerator (Grazzini et al., 2010; Guittet et al., 2011; Gnatta et al., 2014). The haemoglobin level in samples from individuals with advanced adenoma is lower than that in samples from individuals with CRC. The delayed return of samples from the former could therefore more frequently lead to a false-negative test than is the case for CRC. However, it does not seem to be a frequent problem at a temperature of 20°C as the test becomes negative for 3% of advanced adenomas. After 7 days of storage at 30°C, 24% of the positive tests became negative. The data from this study have important implications for the organization of FIT-based screening programmes. Delay in sample return and season did not affect the diagnostic yield of the FIT. An extra 7 days at a temperature of 20°C after sample interpretation, which corresponds to 9–14 days after sample collection, is acceptable. In this condition, globin degradation is not an obstacle to the implementation of FIT screening. However, at a temperature of 30°C and above, the decrease in the haemoglobin concentration was significant after 4 days. Permanent exposure to such a high temperature is rare. However, individuals invited to take part in CRC screening should be informed that it is essential to return the faecal sample promptly and to delay the screening test if the outside temperature is high.
Acknowledgements The authors thank the GPs, gastroenterologists and pathologists for their active participation. This study was part of the research project of team 5 ‘epidemiology and clinical research on digestive cancers’ of the Dijon centre of the National Institute for Medical Research (INSERM U866). Conflicts of interest
There are no conflicts of interest.
References Cha JM, Lee JI, Joo KR, Shin HP, Park JJ, Jeun JW, et al. (2012). Performance of the fecal immunochemical test is not decreased by high ambient temperature in the rapid return system. Dig Dis Sci 57:2178–2183. Faivre J, Dancourt V, Denis B, Dorval E, Piette C, Perrin P, et al. (2012). Comparison between a guaiac and three immunochemical faecal occult blood tests in screening for colorectal cancer. Eur J Cancer 48:2969–2976. Fraser CG, Allison JE, Halloran SP, Young GP (2012). A proposal to standardize reporting units for fecal immunochemical tests for hemoglobin. J Natl Cancer Inst 104:810–814. Fraser CG, Halloran SP, Allison JE, Young GP (2013). Making colorectal cancer screening FITTER for purpose with quantitative faecal immunochemical tests for haemoglobin (FIT). Clin Chem Lab Med 51:2065–2067. Gnatta E, Zaninotto M, Epifani MG, Padoan A, Gjini R, Plebani M (2014). A new sampling device for faecal immunochemical testing: haemoglobin stability is still an open issue. Clin Chem Lab Med 52:1203–1209. Grazzini G, Ventura L, Zappa M, Ciatto S, Confortini M, Rapi S, et al. (2010). Influence of seasonal variations in ambient temperatures on performance of immunochemical faecal occult blood test for colorectal cancer screening: observational study from the Florence district. Gut 59:1511–1515. Guittet L, Bouvier V, Mariotte N, Vallee JP, Arsene D, Boutreux S, et al. (2007). Comparison of a guaiac based and an immunochemical faecal occult blood test in screening for colorectal cancer in a general average risk population. Gut 56:210–214. Guittet L, Guillaume E, Levillain R, Beley P, Tichet J, Lantieri O, Launoy G (2011). Analytical comparison of three quantitative immunochemical fecal occult blood tests for colorectal cancer screening. Cancer Epidemiol Biomarkers Prev 20:1492–1501. Hamza S, Dancourt V, Lejeune C, Bidan JM, Lepage C, Faivre J (2013). Diagnostic yield of a one sample immunochemical test at different cut-off
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
114 European Journal of Cancer Prevention 2016, Vol 25 No 2
values in an organised screening programme for colorectal cancer. Eur J Cancer 49:2727–2733. Heitman SJ, Hilsden RJ, Au F, Dowden S, Manns BJ (2010). Colorectal cancer screening for average-risk North Americans: an economic evaluation. PLoS Med 7:e1000370. Hewitson P, Glasziou P, Watson E, Towler B, Irwig L (2008). Cochrane systematic review of colorectal cancer screening using the fecal occult blood test (Hemoccult): an update. Am J Gastroenterol 103:1541–1549. Hol L, van Leerdam ME, van Ballegooijen M, van Vuuren AJ, van Dekken H, Reijerink JC, et al. (2010). Screening for colorectal cancer: randomised trial comparing guaiac-based and immunochemical faecal occult blood testing and flexible sigmoidoscopy. Gut 59:62–68. Lejeune C, Le Gleut K, Cottet V, Galimard C, Durand G, Dancourt V, et al. (2014). The cost-effectiveness of immunochemical tests for colorectal cancer screening. Dig Liver Dis 46:76–81. Levi Z, Rozen P, Hazazi R, Vilkin A, Waked A, Maoz E, et al. (2007). A quantitative immunochemical fecal occult blood test for colorectal neoplasia. Ann Intern Med 146:244–255. McCulloch C, Searle S, Neuhaus J (2008). Generalized, linear, and mixed models, 2nd ed. Chichester: Wiley.
Sobin L, Gospodarowicz M, Wittekind C (2009). TNM classification of malignant tumours, 7th ed. Oxford: Wiley-Blackwell. pp. 19–20. Van Roon AH, Hol L, van Vuuren AJ, Francke J, Ouwendijk M, Heijens A, et al. (2012). Are fecal immunochemical test characteristics influenced by sample return time? A population-based colorectal cancer screening trial. Am J Gastroenterol 107:99–107. Van Rossum LG, van Rijn AF, Laheij RJ, van Oijen MG, Fockens P, van Krieken HH, et al. (2008). Random comparison of guaiac and immunochemical fecal occult blood tests for colorectal cancer in a screening population. Gastroenterology 135:82–90. Van Rossum LG, van Rijn AF, van Oijen MG, Fockens P, Laheij RJ, Verbeek AL, et al. (2009). False negative fecal occult blood tests due to delayed sample return in colorectal cancer screening. Int J Cancer 125:746–750. Vilkin A, Rozen P, Levi Z, Waked A, Maoz E, Birkenfeld S, Niv Y (2005). Performance characteristics and evaluation of an automated-developed and quantitative, immunochemical, fecal occult blood screening test. Am J Gastroenterol 100:2519–2525. Wilschut JA, Hol L, Dekker E, Jansen JB, Van Leerdam ME, Lansdorp-Vogelaar I, et al. (2011). Cost-effectiveness analysis of a quantitative immunochemical test for colorectal cancer screening. Gastroenterology 141:1648.e1–1655.e1.
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
