Article Text
PDF
Abstract
Background Faecal occult blood testing (FOBT) in population screening has proved to be effective in reducing mortality from colorectal cancer. In Italy a latex agglutination FOBT has been adopted for a single-sample screening programme. The aim of this study was to examine the performance of FOBTs in the Florence screening programme over several seasons to evaluate the impact of variations in ambient temperature on the performance of the screening test.
Methods Measured haemoglobin (Hb) concentrations were aggregated into seasons with their average ambient temperature (AAT). Using logistic regression, the AAT over the period preceding the test measurement was analysed. This period included the time between faecal sampling and return of the test sample (mean 7 days) and the time in the laboratory refrigerator before analysis (mean 4 days). The AAT from days 5–11 before analysis of the test sample was considered a determinant of test positivity. The Kruskal–Wallis rank test was used to evaluate the significance of seasonal and/or AAT-related differences in Hb concentration. A logistic regression model adjusted for sex, age, season and screening episode (first or repeated examination) was constructed.
Results 199 654 FOBT results were examined. Mean FOBT seasonal Hb concentrations (ng/ml) were: spring 27.6 (95% CI 26.2 to 29.1); summer 25.2 (95% CI 23.1 to 27.3); autumn 29.2 (95% CI 27.7 to 30.6); winter 29.5 (95% CI 27.9 to 31.1). Logistic regression showed that there was a 17% lower probability of the FOBT being positive in summer than in winter. The results of the logistic regression showed that an increase in temperature of 1°C produced a 0.7% reduction in probability of a FOBT being positive. In the summer the probability of detecting a cancer or an advanced adenoma was about 13% lower than in the winter.
Conclusions This study showed that there is a significant fall in Hb concentration at higher ambient temperatures. These results will have important implications for the organisation of immunochemical FOBT-based screening programmes, particularly in countries with high ambient temperatures.
- colorectal cancer screening
- faecal occult blood test
- immunochemical test
Statistics from Altmetric.com
Significance of this study
What is already known about this subject?
Haemoglobin (Hb) is not stable in faeces. High moisture levels can reduce stability and impair the performance of dry collection guaiac-based faecal occult blood tests (gFOBTs) such as Hemoccult. In the wet collection systems used by most immunochemical FOBTs (iFOBTs), globin (the measured component of Hb in these tests) is even more susceptible to denaturation than haem.
Although iFOBTs are gradually replacing gFOBTs as the test of choice for screening, little is known about their stability under the range of storage conditions which occurs in population-based screening programmes where environmental conditions cannot be easily controlled.
A recent study has described the occurrence of false negative iFOBTs with delayed return of the test sample.
What are the new findings?
This study is the first to describe a relationship between the seasons and the Hb concentrations measured in faecal samples in a screening programme for colorectal cancer.
The results show a significant difference in the proportion of iFOBT positive results in the summer than in the winter, a difference which is likely to be related to ambient temperature differences.
In the summer the probability of detecting a neoplastic lesion is about 13% lower than in the winter.
How might seasonal variations have an impact on clinical practice?
During the summer, significant neoplasia will be missed which will increase the number of interval cancers.
These observations have significant implications for the organisation of national screening programmes. New mechanisms will need to be considered which will minimise the effect of seasonal temperature variations on positive rates.
Background
Screening by faecal occult blood testing (FOBT) has proved to be effective in decreasing mortality from colorectal cancer.1 Most studies have used the traditional guaiac-based faecal occult blood test (gFOBT) which detects the haem component of blood using its inherent peroxidase activity. However, gFOBT has low analytical and clinical specificity and sensitivity.2 Immunochemical faecal occult blood tests (iFOBTs) detect the human globin moiety of haemoglobin and are therefore more analytically specific and have a much lower detection limit. iFOBTs eliminate interference from dietary meat and vegetables and are more effective in detecting cancers, particularly adenomas.3–7
All recent screening programmes implemented in Italy use iFOBTs.8 In the Florence district the latex agglutination test OC-Sensor (Eiken, Tokyo, Japan) has been adopted, using a single-day test with a positivity threshold of 100 ng/ml Hb.9 The Eiken system enables a quantitative measurement of faecal Hb with a choice of positivity cut-off concentrations. The device enables automated and objective analysis, enhances measurement precision and facilitates improvements in quality control. The possibility of using a positivity cut-off optimised for both the clinical objectives and available resources has been described in several studies7 9–14 which have reached different conclusions, with the choice of cut-off being influenced greatly by financial and logistical considerations.
Immunochemical FOBTs detect the intact globin moiety of human Hb or its early degradation products, but little is known about its stability under different storage conditions. Hb is not stable in faeces; the globin component is degraded quite rapidly while the haem component is degraded more slowly by processes which include the removal of iron by colonic bacteria.15 The stability of globin in the wet collection systems used by most iFOBTs is less than that of the haem used in gFOBTs because globin is more susceptible to denaturation than haem. Moreover, the concentration of Hb in iFOBT samples is very low, which increases its susceptibility to degradation. In 1996 Young et al15 demonstrated an increase in Hb deterioration when the application of wet faecal samples onto gFOBT cards was delayed. A study by Faure et al16 which evaluated the influence of temperature and moisture on gFOBT sensitivity showed significant differences in the positivity rate of Hemoccult II throughout the seasons.
A simulation study performed by Vilkin et al3 using faecal samples collected into the same iFOBT device as that used in the present study (OC-Hemodia) showed no significant deterioration over 21 days at 4°C or 20°C but a fall in the daily Hb concentration of 3.7±1.8% at 28°C. The OC-Hemodia iFOBT device contains a stabilising buffer designed to reduce Hb degradation.3 Nevertheless, as the storage temperature increases, the rate of degradation of Hb increases. This fact presents a challenge for screening programmes which have very limited control over environmental conditions. In a more recent study, Van Rossum et al17 described the occurrence of false negative iFOBTs with delayed return of test samples.
A population-based FOBT screening programme active in the Florence district since 1982 has shown a significant reduction in mortality from colorectal cancer.18 Since the latex agglutination test has been in use in the Florence screening programme, refrigerated storage of tests has been strongly recommended, but faecal samples are likely to be exposured to higher temperatures during transport to the laboratory. The aim of this study was to investigate potential seasonal variations in programme performance according to ambient temperatures in the population-based screening programme of the Florence district.
Methods
The Florence district of Italy has a population-based screening programme for colorectal cancer coordinated by the Cancer Prevention and Research Institute (ISPO). Colorectal cancer screening is available to all subjects in the Florence district who are aged 50–70 years. Subjects are invited by mail every 2 years to perform a single iFOBT with no dietary restriction requirements.
Negative FOBT results are mailed to participants with a recommendation to repeat the screening in 2 years. Non-respondents to the first invitation receive a reminder within 6 months.19 Subjects who test positive are offered a full colonoscopy during dedicated sessions at an accredited assessment clinic. If the colonoscopy is incomplete, a double-contrast x-ray is then performed. Subjects with screen-detected neoplasms are referred for surgical or endoscopic therapy and thereafter enrolled in a surveillance programme. Quality indicators are used to monitor screening performance19 and all screening and assessment data are recorded and available in the ISPO computer archives.
The Florence screening programme started in the 1980s and several different screening tests have been used since then. A latex agglutination test,3 OC-Hemodia, was developed with the OC-Sensor instrument and launched in 2000. In 2004 a new generation of instruments (OC-Sensor Micro) and a sample tube was adopted. The OC-Sensor assay is based on the flocculation reaction between human Hb and multiple monoclonal anti-Hb latex-adsorbed antibodies. The Hb concentration is determined by measuring the optical absorbance of the turbid solution at 660 nm and relating it to a calibration curve; the positivity cut-off concentration used to trigger a diagnostic investigation is 100 ng/ml Hb.
Participants in the screening programme receive detailed written and oral information about preparing faecal samples and are asked to return the test samples as soon as possible in order to avoid degradation of Hb. If the test samples cannot be returned immediately, participants are advised to store samples in the zip-lock bag provided by the manufacturer and to place them in a domestic refrigerator. Tests returned to local healthcare units are stored at 4°C (or at room temperature for a few hours) before being transported every 2 days to the centralised ISPO laboratory. Refrigeration of samples is recommended during transport. Test kits are received by the local healthcare units throughout the year except for a week in the middle of August when both transportation and laboratory activity is suspended.
Once received by the laboratory, samples are stored at 4°C until analysis which is usually performed within 1–2 weeks of sample collection. Analysis of the samples is performed by a technician under the supervision of a laboratory scientist. Irrespective of the period between collection and analysis, all samples are analysed and their Hb concentration recorded in the ISPO archives. The date of kit delivery and test analysis is routinely recorded in the ISPO archives, but the dates of faecal sampling or test returns are not recorded. The average time between test delivery date and analysis is 12.1 days. Periodical reports from the ISPO laboratory show an average storage period in the laboratory refrigerator before analysis of 3–5 days. For this study, ambient temperatures throughout the period of sample collection and analysis were provided by the Laboratory for Environmental Monitoring and Modelling of Florence (LAMMA).
Advanced adenoma was defined as an adenoma with a diameter >0.9 cm, or with at least 20% of villous component or with high-grade dysplasia.
Statistical analysis
Using the time described above between kit delivery and faecal test analysis, we tested the most likelihood model by a logistic regression also considering gender, age and rank of screening test (first vs repeated examinations). The most reliable model resulted in 4 days of storage in the laboratory refrigerator and 4 days between faecal sampling and receipt at the ISPO laboratory (this period of 7 days is that which best explains the variance of our data). In our statistical analysis we considered the average ambient temperatures from days 5–11 before the actual test analysis. Concentrations of Hb in the iFOBT samples did not show a normal distribution and constancy of the variances so a logarithmic transformation was performed without normalisation of data distribution. A non-parametric test (Kruskal–Wallis rank test for equality) was therefore used to evaluate the difference across the seasons. The probability of the iFOBT being positive (Hb concentration ≥100 ng/ml) and of diagnosing a neoplastic lesion (cancer or advanced adenoma) was studied using a logistic model adjusted for sex, age, season and episode of screening (first or repeated test).
Results
A total of 199 654 iFOBTs were performed in the colorectal cancer screening programme in the Florence district between January 2001 and December 2008. The average ambient temperature was 16.0°C (25th and 75th percentiles 13.1, 19.3) in spring; 25.1°C (23.5, 26.9) in summer; 14.0°C (10.8, 17.7) in autumn; and 7.9°C (5.8, 9.9) in winter.
Figure 1 shows the mean Hb concentration (with 95% CI) at different ambient temperatures. There was a decrease in Hb concentration with increasing temperature, shown in intervals of 5°C. The mean Hb concentration was 31.1 ng/ml (95% CI 27.5 to 34.8) at temperatures below 5°C and 23.9 ng/ml (95% CI 21.2 to 26.6) at temperatures over 25°C.
Mean (95% CI) haemoglobin (Hb) concentration observed at ambient temperature intervals of 5°C.
Table 1 shows the mean Hb concentration (with 95% CI) according to the seasons. The mean Hb concentration was 27.6 ng/ml (95% CI 26.2 to 29.1) in spring, 25.2 ng/ml (95% CI 23.1 to 27.3) in summer, 29.2 ng/ml (95% CI 27.7 to 30.6) in autumn and 29.5 ng/ml (95% CI 27.9 to 31.1) in winter (Kruskal–Wallis test for equality of rank p<0.001). The lowest concentration was recorded in August (mean 23.4 ng/ml (95% CI 16.7 to 30.1)) and the highest in January (mean 30.4 ng/ml (95% CI 27.1 to 33.8)).
Mean (95% CI) seasonal variation in haemoglobin (Hb) concentration
Logistic regression adjusted for sex, age and history of screening (first or repeated examination) of the screened subjects was used to evaluate the effect on positivity of the reduction in mean FOBT concentration. The results are reported in table 2 and show that there was a 17% lower probability of the iFOBT being positive in summer than in winter.
Logistic regression of the probability of positive screening tests by season and adjusted for age, sex and history of screening (95% CI)
When the same analysis was made using the average ambient temperature (as a continuous variable) from the days 5–11 before test analysis, the regression shows that an increase in temperature of 1°C resulted in a 0.7% reduced probability of iFOBT being positive (OR 0.993, 95% CI 0.989 to 0.996).
Table 3 shows the probability of diagnosing cancer or advanced adenoma. In summer the probability of detecting a neoplastic lesion is about 13% lower than in winter, although the result is only of borderline significance (OR 0.87, 95% CI 0.74 to 1.03). Replacing season as the parameter with average ambient temperature gave the same result (OR 0.993, 95% CI 0.986 to 1.001).
Logistic regression of the probability to have a neoplastic lesion diagnosed (cancer or advanced adenoma) (95% CI)
Compliance with colonoscopy assessment was a little lower in summer than in other seasons but, after adjusting for this factor, the detection rate for cancer or advanced adenomas remained lower in the summer than in the other seasons (see table 4).
Number of screened subjects, iFOBT positive subjects, colonoscopies, screen-detected lesions, crude and adjusted detection rate for cancer or advanced adenomas
Discussion
This study describes the relationship between the seasons and the Hb concentration measured in faecal samples received as part of a screening programme for colorectal cancer. Our results span a period of 8 years and are extracted from a large screening database. They show that Hb measurements made using the OC-Sensor immunochemical device during the summer are significantly lower than those during the winter. The study also shows that, in the summer period, the detection rate of neoplastic lesions (advanced adenomas and cancers) is similarly decreased. The observed seasonal difference in Hb concentration was related to average ambient temperatures. The susceptibility to denaturation of Hb in faecal samples has been previously described.3 The manufacturers of the iFOBT device recommend minimising the period between collection and measurement, and some screening programmes including that in Florence recommend refrigerated storage of specimens prior to analysis. The possibility that seasonal variations in temperature might influence the performance of screening programmes has not been described previously. Throughout the period studied, no significant change in programme organisation or laboratory analysis influenced the data. The programme does recommend that participants store samples hygienically in a dedicated plastic bag, but it has no evidence that this recommendation is adhered to. It is possible that the return time of samples is extended during the summer period, but the programme has no evidence of this. The English NHS Bowel Cancer Screening Programme20 has detailed records of test collection used to monitor elapsed time and has only observed delays during late December which are thought to be due to disrupted mail delivery during the Christmas period.
Although the important observations made in this 8-year retrospective analysis of data from the Italian programme could be explained by seasonal differences in diet, changes in bowel habit or seasonal pathological changes, the susceptibility of iFOBTs to Hb degradation at higher temperatures is the most convincing explanation. An unpublished retrospective analysis of data from the South of England NHS Bowel Cancer Screening Programme has found no seasonal variation in gFOBT positivity to date, which is consistent with both the higher limit of detection and greater stability of gFOBT. Seasonal variations in diet, bowel habit or disease that might alter the concentration of blood in faeces would be likely to be observed in both gFOBT and iFOBT screening programmes. A causal relationship between the ambient temperature during sampling, transport and storage prior to analysis and the measured Hb concentration is therefore a likely explanation.
Our results predict a higher rate of false negative iFOBT results during the summer period, and this was observed with a significant difference in positive iFOBT results between summer and winter which relates to predictable differences in ambient temperature. Van Rossum17 observed a larger proportion of false negative tests in patients with minor neoplastic lesions where the Hb concentration in faeces is usually lower21 than in patients with larger adenomas or invasive cancers. It is likely that the increased Hb degradation observed during the summer will primarily affect smaller or preneoplastic lesions which have samples with Hb concentrations close to the analytical cut-off value used by the screening programme. In our study, 17 (4.3%) of 398 screen-detected cancers had Hb levels <120 ng/ml.
These observations have significant implications for national screening programmes and might prove of relevance to the recently reported fall in test positivity observed by the Australian Screening Programme22 which used an alternative iFOBT (Hem-Tube/Magstream; Fuiirebio Inc, Japan).
If the instability of iFOBTs results in seasonal variation in the screening positivity rate with increased false negative test rates during periods of hot weather, more effective ways need to be found to improve sample stability.3 The stability of Hb decreases in the presence of microbial contamination but can be enhanced by changes in buffer pH and protein concentration. Other specific moieties such as haptoglobin can increase Hb stability in sample collection buffers. The inherent instability of transporting wet samples might be overcome by collecting samples on filter paper, in a manner similar to that used by gFOBT systems, and extracting the Hb from the dried sample into a test buffer on receipt by the testing laboratory.23 The need for high volume automated analysis in national screening programmes currently precludes this manual alternative.
Until improved collection systems have been developed, screening programmes will need to consider methods which minimise the effect of seasonal variations in temperature on positive rates. Some of the mechanisms which might be adopted could include:
requiring that samples be posted to the laboratory immediately after sample collection;
requiring that samples be stored prior to transportation to the laboratory in a domestic refrigerator;
insisting that participants record the sample collection date and then excluding those that fall outside a designated ‘safe’ period (this approach is used for gFOBT in the English screening programme with the application of a 21-day ‘safe’ period);
improving the rapidity of transport between the participant and the testing laboratory;
refrigerating samples (perhaps using dedicated ice packs) during transport to the laboratory, a method used for DNA faecal tests (dFOBTs)24;
testing immediately on receipt by the laboratory or refrigerating stored samples prior to analysis (freezing is known to interfere with some iFOBT systems);
adopting a more sensitive ‘seasonal’ strategy such as reducing the cut-off points from 100 ng/ml to 80 ng/ml or using a positive from either of two samples instead of a single sample during the summer period. An Italian study12 found higher sensitivity using dual samples compared with the standard strategy (cut-off at 100 ng/ml and a single sample);
reducing the positivity cut-off of FOBT for the whole programme to avoid the increase in false negatives during the summer, accepting an increase in false positives during the winter;
modifying the period of invitation perhaps to 2.5 years so that a subject invited in summer for the first test would be invited during winter for the next test;
suspending screening activity during the hot period of the year;
reducing the interval between screening episodes for those tested during the summer.
All of the mechanisms described above would have significant implications for a screening programme—some would decrease programme participation, others would increase programme complexity and cost and decrease efficiency.
In conclusion, our observations have important implications for the organisation of iFOBT-based screening programmes. The implications are greatest for those countries with wide seasonally-related temperature variations. Programmes need to be aware of the seasonal temperature profiles for samples collected into iFOBT devices as they move between collection and analysis. The use of temperature tracking devices to gather these temperature profiles during the summer period should be considered by all programmes. The observations described in this study need to be confirmed by other programmes working under a variety of different climatic conditions.
References
Footnotes
Linked articles: 216291
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.