This randomized clinical trial compares the effect of a single oral dose of aspirin vs placebo prior to fecal immunochemical testing (FIT) on test sensitivity for detecting advanced colorectal neoplasms.
Key Points
Question
What is the effect of a single oral dose of aspirin prior to fecal immunochemical testing on test sensitivity for detecting advanced colorectal neoplasms?
Findings
In this randomized clinical trial involving 2422 adults aged 40 to 80 years not using aspirin or other antithrombotic medications, there was no statistically significant difference in fecal immunochemical test sensitivity for detecting advanced colorectal neoplasms with administration of a single tablet of 300-mg aspirin, compared with placebo, 2 days before fecal testing (sensitivity at 2 prespecified positivity thresholds: 40.2% vs 30.4%, and 28.6% vs 22.5%, respectively).
Meaning
Administration of a single dose of oral aspirin prior to fecal immunochemical testing, compared with placebo, did not significantly increase test sensitivity for detecting advanced colorectal neoplasms at 2 predefined cutoffs of a quantitative fecal immunochemical test.
Abstract
Importance
Fecal immunochemical tests for hemoglobin are widely used for colorectal cancer (CRC) screening. Observational studies suggested that sensitivity of fecal immunochemical tests for detecting advanced neoplasms could be increased by acetylsalicylic acid (aspirin), especially among men.
Objective
To evaluate the potential to increase sensitivity of fecal immunochemical tests by administering a single 300-mg oral aspirin dose 2 days before stool sampling.
Design, Setting, and Participants
A randomized, placebo-controlled, double-blind trial was conducted in 14 gastroenterology practices and 4 hospitals in Germany, and included 2422 men and women aged 40 to 80 years scheduled for colonoscopy, with no recent use of aspirin or other drugs with antithrombotic effects (enrollment from June 2013 to November 2016, and final follow-up January 27, 2017).
Interventions
Administration of a single tablet containing 300 mg of aspirin (n = 1208) or placebo (n = 1214) 2 days before fecal sampling for fecal immunochemical test.
Main Outcome and Measures
The primary outcome was sensitivity of a quantitative fecal immunochemical test at 2 predefined cutoffs (10.2 and 17-μg Hb/g stool) for detecting advanced neoplasms (colorectal cancer or advanced adenoma).
Results
Among 2422 randomized patients (mean [SD] age, 59.6 [7.9] years; 1219, 50%, men), 2134 were included in the analysis (78% for primary screening colonoscopy, 22% for diagnostic colonoscopy). Advanced neoplasms were identified in 224 participants (10.5%), including 8 participants (0.4%) with CRC and 216 participants (10.1%) with advanced adenoma. Sensitivity was 40.2% in the aspirin group and 30.4% in the placebo group (difference 9.8%, 95% CI, −3.1% to 22.2%, P = .14) at cutoff 10.2-μg Hb/g stool and 28.6% in the aspirin and 22.5% in the placebo group (difference 6.0%, 95% CI, −5.7% to 17.5%, P = .32) at cutoff 17-μg Hb/g stool.
Conclusions and Relevance
Among adults aged 40 to 80 years not using aspirin or other antithrombotic medications, administration of a single dose of oral aspirin prior to fecal immunochemical testing, compared with placebo, did not significantly increase test sensitivity for detecting advanced colorectal neoplasms at 2 predefined cutoffs of a quantitative fecal immunochemical test.
Trial registration
Deutsches Register Klinischer Studien Identifier: DRKS00003252; EudraCT Identifier: 2011-005603-32/DE
Introduction
Colorectal cancer (CRC) accounts for approximately 860 000 deaths each year globally.1 A large proportion of these deaths could be potentially prevented by screening. Randomized clinical trials (RCTs) have demonstrated effectiveness of screening by guaiac-based, chemical fecal occult blood tests (FOBTs) in reducing CRC incidence and mortality.2,3,4 Newer fecal immunochemical tests (FITs) for hemoglobin (Hb) outperform guaiac-based FOBTs in diagnostic accuracy5,6,7,8 and are broadly recommended for CRC screening.9,10,11
FITs detect the majority (70%-80%) of CRCs at high (>90%) specificity but only a minority of advanced adenomas, the precursors of most CRCs.8,10 Detecting advanced adenomas in addition to CRCs could make a major contribution to effectiveness of FIT-based CRC screening.12 Improved sensitivity without simultaneously decreasing specificity could improve the performance of FIT for colorectal cancer screening.
In a previous observational study, enhanced sensitivity of FITs for detecting advanced adenomas was found among participants of screening colonoscopy who used low-dose aspirin for cardiovascular disease prevention.13 A plausible explanation might be that aspirin predisposes to subclinical bleeding and, hence, increased detection of advanced adenomas by FIT. This suggests that administration of aspirin prior to fecal sampling might be a practical intervention to increase FIT sensitivity. The objective of this randomized clinical trial was to investigate the effect of a single 300-mg dose of aspirin before FIT on the sensitivity of FITs to detect advanced colorectal neoplasms.
Methods
Ethics Approval
The trial was approved by the Ethics Committee of the Medical Faculty of Heidelberg University, other responsible ethics committees of state physicians’ boards, and the German Federal Institute for Drugs and Medical Devices (BfArM). Written informed consent was obtained from each participant.14
Trial Design and Trial Population
The study protocol and statistical analysis plan are provided in Supplement 1 and Supplement 2, respectively. Briefly, this randomized, double-blind, placebo-controlled trial was designed to evaluate diagnostic performance of 2 different FITs for detecting advanced neoplasms, after a single 300-mg dose of oral acetylsalicylic acid (aspirin) dose compared with placebo. Advanced neoplasms were defined as the presence of either CRC or advanced adenoma that either was 1 cm or larger in size, had tubulovillous or villous components, or high-grade dysplasia. Men and women aged 40 to 80 years with no recent use of aspirin or other drugs with antithrombotic effects were recruited from June 2013 until November 2016 when visiting 1 of 18 trial centers including 14 gastroenterology private practices and 4 hospitals in Germany for a precolonoscopy appointment. In addition to the participants who had been scheduled for primary screening colonoscopy, participants undergoing diagnostic colonoscopy (including colonoscopy conducted before the recommended screening age, which is reimbursed as diagnostic colonoscopy in the German health care system) could be included. Participants were excluded if they had previous CRC, had overt or FOBT-detected rectal bleeding, or reported recent use of aspirin or other nonsteroidal anti-inflammatory agents (for details see eTable 1 in Supplement 3). In each case, the colonoscopy was planned before recruitment to the trial.
Potentially eligible participants obtained detailed written information on the trial from their physicians. Participants received trial medication (allocated in a blinded manner for both physicians and participants according to randomization list), questionnaire, and patient diary. In addition, participants received 4 stool collection kits for each of 2 FITs: FOB Gold Tube Screen (Sentinel Diagnostics), a widely used quantitative FIT,8,15,16 and the FD Hb/Hp Complex quick (Frost Diagnostika), a qualitative chromatographic FIT,17 referred to as quantitative and qualitative FIT hereafter. At the time this trial was planned, both quantitative and qualitative FITs were widely used in Germany.
Participants were asked to collect a baseline stool sample, then take the trial medication and collect further stool samples 2, 3, and 4 days after trial medication intake (1 stool sample per scheduled day, 4 stool samples overall). If stool collection on these days was not possible, postponement to subsequent day(s) was allowed, but all fecal samples were to be collected prior to initiation of large bowel preparation for colonoscopy, which had to be conducted within 3 months after recruitment. Participants were asked to document the day of trial medication intake and the days of stool sample collection in a diary and to fill out a standardized questionnaire addressing sociodemographic characteristics, comorbidities, lifestyle factors, and other potential CRC risk factors. Stool samples, diaries, and questionnaires were sent by regular mail to the coordinating center at the German Cancer Research Center using prepaid and addressed mailing devices.
Randomization
The Pharmacy of the University Heidelberg compiled randomization lists based on computer-generated random numbers (Randomization in Treatment Arms, RITA, version 1.24) and packed the trial medication (single tablet containing 300-mg aspirin without enteric coating [ASS Ratiopharm 300 mg, Ratiopharm GmbH] or cellulose, lactosemonohydrate, magnesium stearate [placebo]). A randomized block design was used with each block containing 20 tablets (10 assigned to each group). Tablets were taken from their original packaging by trained personnel in a clean room class D environment and repackaged and labeled in a blind manner.
Data Processing
Data collected at the trial centers were entered into electronic case report forms (eCRFs) on site and transferred to a database at the coordinating center. Deidentified colonoscopy and histology reports were sent to the coordinating center, where information was extracted and entered into a colonoscopy database by 2 independent, blinded reviewers. Comprehensive plausibility checks of data entries were performed by medical information specialists.
Laboratory Methods
Participants were asked to collect fecal samples in sampling tubes provided by the manufacturers and send them to the coordinating center, where they were stored at 2°C to 8°C until analysis. Median time between stool sampling and arrival in the coordinating center was 4 days (interquartile range, 3-5 days).
The quantitative FIT was analyzed in an external accredited laboratory (Limbach Laboratory) using Abbott Architect c8000 (analytical range, 0.03-142 μg-Hb/g feces). Analyses of Hb concentration by the qualitative FIT were performed at the coordinating center. All laboratory analyses were performed blinded with respect to treatment assignment and clinical and colonoscopy data.
Outcomes
The primary outcome was sensitivity of the quantitative FIT for detecting advanced neoplasms at 2 predefined cutoffs: 17-μg Hb/g feces, the cutoff given by the manufacturer, and 10.2-μg Hb/g feces, the cutoff of the qualitative FIT, in fecal samples scheduled 2 days after intake of aspirin or placebo. The main analysis followed an intention-to-screen approach using fecal samples scheduled for day 2 after tablet intake; results of analyses following a prespecified per-protocol approach using fecal samples collected at exactly day 2 are additionally reported. Prespecified secondary outcomes reported in this article were specificity, positive predictive value (PPV) and negative predictive value (NPV) of the quantitative test, potential gain in sensitivity by combining results from multiple fecal samples scheduled on any day between day 2 and day 4, sensitivity, specificity, PPV and NPV of the qualitative test, sex-specific sensitivity, specificity, PPV and NPV of both tests, and serious adverse events. Further secondary outcomes not reported in this article were likelihood ratios and area under the curve.
Sample Size and Power Calculations
Assuming a prevalence of advanced neoplasms of 10% and based on the results of the preceding observational study,13 the trial had been designed to have 90% power (in 2-sided testing at α = .05) to detect a difference in sensitivity between intervention and control group at day 2 of 24 percentage points (60% vs 36%).
Statistical Analyses
Participants were analyzed according to their randomization group. Those who withdrew their consent or did not contribute with any samples and information were excluded as dropouts. Participants who did not provide stool samples prior to colonoscopy were excluded from the primary analysis.
All analyses were performed with R statistical software version 3.4.0. Wald-adjusted confidence intervals were computed for sensitivity and specificity values (Agresti-Coull adjustment)18 and the differences (Agresti-Caffo adjustment)19 with the PropCIs package. Interaction P values were obtained from a generalized linear model fit for binomial data. Statistical significance was defined by a 2-sided P value <.05. No adjustments were made for multiple testing. Analyses of secondary outcomes should therefore be regarded as exploratory.
Sensitivity Analysis
To account for study conduction at multiple sites, a generalized linear mixed model, with a random intercept accounting for the site, was fitted for each outcome of the main intention-to-screen analysis via the lme4 package.20 A likelihood ratio test21 for the null hypothesis of zero variance of the random intercept was performed.
Results
From 2422 recruited participants, 288 dropped out or were excluded for the reasons shown in the Figure, leaving 2134 participants for the analysis. Table 1 displays main characteristics of the trial population, which included similar proportions of women and men. The majority (93.6%) of participants were between 45 and 74 years old, mean age was 59.6 years. Seventy-eight percent of colonoscopies were conducted for primary screening. Advanced neoplasms were identified in 224 participants (10.5%), including 8 participants (0.4%) with CRC and 216 participants (10.1%) with advanced adenoma. Distributions of demographic and clinical characteristics were similar in the intervention and the control group.
Figure. Flow Diagram of Recruitment and Exclusions.
aThe number of participants assessed for eligibility was not collected.
bPatients who did not contribute with any sample or information were considered as dropouts.
cPatients who completed the colonoscopy were considered to have been followed up.
Table 1. Characteristics of the Trial Population.
| Group, No. (%) | ||
|---|---|---|
| Aspirin (n = 1075) |
Placebo (n = 1059) |
|
| Sex | ||
| Men | 547 (50.9) | 528 (49.9) |
| Women | 528 (49.1) | 531 (50.1) |
| Age, y | ||
| <45 | 26 (2.4) | 30 (2.8) |
| 45-49 | 63 (5.9) | 52 (4.9) |
| 50-54 | 165 (15.3) | 164 (15.5) |
| 55-59 | 315 (29.3) | 318 (30.0) |
| 60-64 | 232 (21.6) | 216 (20.4) |
| 65-69 | 144 (13.4) | 153 (14.4) |
| 70-74 | 92 (8.6) | 83 (7.8) |
| ≥75 | 38 (3.5) | 43 (4.1) |
| Primary indication for colonoscopya | ||
| Screening | 805 (77.9) | 794 (78.2) |
| Diagnostic | 228 (22.1) | 221 (21.8) |
| Most advanced finding at colonoscopy | ||
| Colorectal cancer | 3 (0.3) | 5 (0.5) |
| Advanced adenoma | 112 (10.4) | 104 (9.8) |
| Nonadvanced adenoma | 286 (26.6) | 280 (26.4) |
| None of above | 674 (62.7) | 670 (63.3) |
Information on primary indication for colonoscopy was missing for 42 participants in the aspirin group and 44 participants in the control group
eTable 2 in Supplement 3 shows the numbers of returned and valid FITs by both scheduled and actual day of stool sampling. A total of 1030 (95.8%) and 1012 (95.6%) valid samples for the quantitative FIT scheduled for day 2 were obtained from participants in the intervention and control group, respectively. Numbers for the qualitative FIT were 974 (90.6%) and 969 (91.5%), respectively.
Primary End Point
The sensitivities of the quantitative FIT for detecting advanced neoplasms using fecal samples scheduled 2 days after tablet intake (intention-to-screen analysis) are shown in Table 2. At the 10.2-μg Hb/g stool cutoff, sensitivity was 40.2% in the intervention group and was 30.4% in the placebo group (difference, 9.8%; 95% CI, −3.1% to 22.2%; P = .14); at the 17-μg Hb/g stool cutoff, the sensitivities were 28.6% and 22.5%, respectively (difference, 6.0%; 95% CI, −5.7% to 17.5%, P = .32).
Table 2. Sensitivity and Specificity for Detecting Advanced Neoplasms at Day 2 After a Single 300-mg Aspirin Dose or Placeboa.
| Test | Cutoff, μg, Hb/g |
No. of Detections | Difference, % (95% CI) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Aspirin | Placebo | ||||||||||||||
| TP | FN | TN | FP | Sensitivity, % | Specificity, % | TP | FN | TN | FP | Sensitivity, % | Specificity, % | Sensitivity | Specificity | ||
| All participants | |||||||||||||||
| Quantitative | 10.2 | 45 | 67 | 755 | 163 | 40.2 | 82.2 | 31 | 71 | 807 | 103 | 30.4 | 88.7 | 9.8 (−3.1 to 22.2) |
−6.4 (−9.6 to −3.2) |
| Quantitative | 17.0 | 32 | 80 | 842 | 76 | 28.6 | 91.7 | 23 | 79 | 863 | 47 | 22.5 | 94.8 | 6.0 (−5.7 to 17.5) |
−3.1 (−5.4 to −0.8) |
| Qualitative | 10.2 | 35 | 66 | 732 | 141 | 34.7 | 83.8 | 22 | 78 | 798 | 71 | 22.0 | 91.8 | 12.7 (0.1 to 24.7) |
−8.0 (−11.0 to −4.9) |
| Men | |||||||||||||||
| Quantitative | 10.2 | 33 | 44 | 360 | 90 | 42.9 | 80.0 | 15 | 39 | 387 | 68 | 27.8 | 85.1 | 15.1 (−1.6 to 30.6) |
−5.1 (−10.0 to −0.1) |
| Quantitative | 17.0 | 25 | 52 | 408 | 42 | 32.5 | 90.7 | 13 | 41 | 424 | 31 | 24.1 | 93.2 | 8.4 (−7.5 to 23.3) |
−2.5 (−6.1 to 1.1) |
| Qualitative | 10.2 | 28 | 43 | 354 | 74 | 39.4 | 82.7 | 12 | 43 | 400 | 48 | 21.8 | 89.3 | 17.6 (1.3 to 32.6) |
−6.6 (−11.2 to −2.0) |
| Women | |||||||||||||||
| Quantitative | 10.2 | 12 | 23 | 395 | 73 | 34.3 | 84.4 | 16 | 32 | 420 | 35 | 33.3 | 92.3 | 1.0 (−19.1 to 21.4) |
−7.9 (−12.0 to −3.8) |
| Quantitative | 17.0 | 7 | 28 | 434 | 34 | 20.0 | 92.7 | 10 | 38 | 439 | 16 | 20.8 | 96.5 | −0.8 (−17.9 to 17.2) |
−3.7 (−6.7 to −0.8) |
| Qualitative | 10.2 | 7 | 23 | 378 | 67 | 23.3 | 84.9 | 10 | 35 | 398 | 23 | 22.2 | 94.5 | 1.1 (−17.7 to 20.9) |
−9.6 (−13.5 to −5.5) |
Abbreviations: FN, false-negative; FP, false-positive; Hb, hemoglobin; TN, true-negative; TP, true-positive.
Results of the intention-to-screen analysis.
Secondary End Points
At the 10.2-μg Hb/g stool cutoff, specificity of the quantitative FIT was 82.2% in the intervention group and 88.7% in the control group (difference, −6.4%, 95% CI, −9.6% to −3.2%, P < .001) and at 17-μg Hb/g stool cutoff was 91.7% and 94.8%, respectively (difference, −3.1%; 95% CI, −5.4% to −0.8%; P = .008) (Table 2).
For the qualitative test, sensitivity was higher by 12.7 percentage points in the intervention group (34.7% vs 22.0%) than it was in the control group (95% CI, 0.1 to 24.7, P = .048) but specificity was lower by 8.0 percentage points in the intervention group (83.8% vs 91.8%) than it was in the control group (95% CI, −11.0 to −4.9; P < .001).
Among the subgroup of men, point estimates of sensitivity were between 8.4 (95% CI, −7.5 to 23.3) and 17.6 (95% CI, 1.3 to 32.6) percentage points higher, and specificities were between 2.5 (95% CI, −6.1 to 1.1) and 6.6 (95% CI, −11.2 to −2.0) percentage points lower in the intervention group than in the control group (Table 2). In contrast, among women sensitivity was similar but specificity was lower in the intervention group than in the control group (Table 2). P values for interaction effects between sex and intervention were all statistically nonsignificant and ranged from .08 to .49.
eTable 3 in Supplement 3 shows sensitivities and specificities obtained when test positivity was defined by at least 1 positive result in up to 3 fecal samples scheduled for days 2, 3, and 4. Compared with results for day-2 samples only, higher sensitivities and lower specificities were observed in both the intervention group and the control group, with little change in the differences in sensitivity between the intervention and control groups.
eTable 4 in Supplement 3 shows sensitivities and specificities using fecal samples collected exactly 2 days after tablet intake (per-protocol analysis). Compared with intention-to-screen analysis, differences in sensitivity between the intervention and control group were of greater magnitude among men, with sensitivities that were 23.2 (95% CI, 5.0 to 39.5) and 22.5 (95% CI, 4.5 to 38.6) percentage points higher in the intervention group at the lower cutoff of the quantitative test and for the qualitative test, respectively. Differences in specificity between the intervention and control groups were similar for the per-protocol and intention-to-screen analyses.
Table 3 and eTable 5 in Supplement 3 provide PPVs and NPVs at day 2 after participants took a single 300-mg aspirin dose or placebo. Overall, no significant differences were seen, neither in intention-to-screen nor in per-protocol analysis.
Table 3. Positive and Negative Predictive Values for Detecting Advanced Neoplasms at Day 2 After a Single 300-mg Aspirin Dose or Placeboa.
| Test | Cutoff μg Hb/g | Aspirin | Placebo | Difference (95% CI) | |||
|---|---|---|---|---|---|---|---|
| PPV | NPV | PPV | NPV | PPV | NPV | ||
| All participants | |||||||
| Quantitative | 10.2 | 21.6 | 91.8 | 23.1 | 91.9 | −1.5 (−10.7 to 7.4) | −0.1 (−2.7 to 2.5) |
| Quantitative | 17.0 | 29.6 | 91.3 | 32.9 | 91.6 | −3.2 (−17.2 to 10.5) | −0.3 (−2.8 to 2.3) |
| Qualitative | 10.2 | 19.9 | 91.7 | 23.7 | 91.1 | −3.8 (−14.4 to 6.5) | 0.6 (−2.1 to 3.3) |
| Men | |||||||
| Quantitative | 10.2 | 26.8 | 89.1 | 18.1 | 90.8 | 8.8 (−3.0 to 19.8) | −1.7 (−5.9 to 2.4) |
| Quantitative | 17.0 | 37.3 | 88.7 | 29.5 | 91.2 | 7.8 (−10.3 to 24.8) | −2.5 (−6.4 to 1.4) |
| 10.2 | 27.5 | 89.2 | 20 | 90.3 | 7.5 (−6.4 to 20.2) | −1.1 (−5.3 to 3.0) | |
| Women | |||||||
| Quantitative | 10.2 | 14.1 | 94.5 | 31.4 | 92.9 | −17.3 (−31.8 to −2.5) | 1.6 (−1.7 to 4.8) |
| Quantitative | 17.0 | 17.1 | 93.9 | 38.5 | 92 | −21.4 (−42.2 to 0.8) | 1.9 (−1.4 to 5.2) |
| Qualitative | 10.2 | 9.5 | 94.3 | 30.3 | 91.9 | −20.8 (−37.8 to −4.0) | 2.3 (−1.1 to 5.8) |
Abbreviations: Hb, hemoglobin; NPV, negative predictive value; PPV, positive predictive value.
Results of the intention-to-screen analysis.
Adverse Events
Overall, the trial medication was well tolerated. In the aspirin group, 11 adverse events in 7 participants were reported. Two serious adverse events in 2 participants (acute appendicitis and acute suppurative cholangitis accompanied by acute kidney injury), which required hospitalization, were considered as not related to the trial drug. In the placebo group, 5 participants reported 6 adverse events. Serious adverse events were not reported. All participants recovered completely.
Sensitivity Analysis
Results of the sensitivity analysis for the effect of study site are shown in eTable 6 in Supplement 3. The likelihood ratio test for the null hypothesis of zero variance of the random intercept never achieved statistical significance (P values 0.16 to 0.50), therefore the model without random intercept was retained.
Discussion
In this RCT involving adults aged 40 to 80 years not using aspirin or other antithrombotic medications, administration of a single dose of oral aspirin prior to fecal immunochemical testing, compared with placebo, did not result in a statistically significant increase in test sensitivity for detecting advanced colorectal neoplasms at 2 predefined cutoffs of a quantitative fecal immunochemical test.
The lack of statistically significant findings in this study contrast with findings of a previous observational study involving similar populations that motivated the conduct of this trial.13 The observational study found that the sensitivity of 2 other FITs (RIDASCREEN Haemo-/Haptoglobin Complex) for detecting advanced neoplasms was markedly increased (at a modest decrease in specificity) in users of low-dose aspirin among men but not among women. However, this study compared regular daily use of low-dose aspirin with no use in a nonrandomized design. An observational study from Israel also found substantially higher sensitivity of yet another FIT (OC-Micro) among users of aspirin or nonsteroidal anti-inflammatory drugs.22 Taken together, results from these observational studies and this trial suggest that further research is needed to assess whether low-dose aspirin may have an effect on FIT test performance, such as in a larger well-powered trial. This trial was designed to detect a 24% absolute increase in sensitivity and was not adequately powered to detect small differences that may nevertheless be clinically meaningful given the low morbidity observed, the low cost of a single dose of aspirin, and the ease of implementation of this intervention across health systems.
Antithrombotic effects of aspirin are well established.23 They seem to be stronger among men than women,24,25,26 and sex differences in platelet number, function and responsiveness to aspirin have been reported.26,27,28 Also, longer colonic transit time of feces and higher prevalence of constipation among women29,30 might favor intracolonic Hb degradation, especially in the case of bleeding from proximal neoplasms that are relatively more common among women.31 By contrast, prevalence of conditions potentially associated with false-positive FITs, especially under aspirin treatment, such as hemorrhoids, seems not to show major sex differences.32 Additional adequately powered studies are needed to evaluate potential sex differences and reasons for differential aspirin effects on FIT sensitivity.
Other important questions include necessary aspirin dose and preparation and optimal timing of aspirin intake and fecal sampling. This trial tested a 300-mg dose to achieve a relevant hemorrhagic effect even with a single tablet. Possibly, an equivalent or better effect might be achieved with single or multiple tablets (taken over several days) containing lower doses of aspirin, such as 75 mg or 100 mg, which should be explored in further research. FIT performance 2 days following the dose of aspirin was chosen as the primary analysis to allow for adequate colonic transition time. Sensitivity analyses do not suggest performing fecal sampling over multiple days would improve diagnostic performance of FIT.
Limitations
This trial has several limitations. First, despite the overall large number of participants undergoing FIT screening with or without aspirin, estimates of sensitivity were based on relatively few participants with advanced neoplasms. The trial was designed to detect a large effect size of 24% absolute increase in sensitivity of FIT and was not designed to detect more moderate but nevertheless clinically meaningful between-group differences. Second, multiple secondary analyses were conducted without adjustments for multiple testing. Third, although the sample size for analyses of sensitivity was mainly limited by the low prevalence of advanced neoplasms, sample sizes for the primary analyses were reduced by approximately 4% due to exclusions. However, such exclusions were unrelated to treatment group and should not have biased between-group comparisons. Fourth, the limited size of the study sample precluded more detailed analyses of sensitivity for specific subtypes of advanced neoplasms, eg, by advanced adenoma size. Fifth, the number of potential participants excluded due to preexisting use of aspirin or nonsteroidal agents was not quantified. Sixth, the trial included participants scheduled for either screening or diagnostic colonoscopies. However, the 22% of the trial participants who had FIT prior to diagnostic colonoscopies were included only if the indication was unrelated to rectal bleeding and therefore was unlikely to have an effect on FIT results. Seventh, only effects on diagnostic performance of 1-time screening were assessed. FIT-based screening programs typically include multiple screening rounds, and potential effects on detection of advanced neoplasms and reduction of CRC incidence and mortality in the long run are yet to be determined.
Conclusions
Among adults aged 40 to 80 years not using aspirin or other antithrombotic medications, administration of a single dose of oral aspirin prior to fecal immunochemical testing, compared with placebo, did not significantly increase test sensitivity for detecting advanced colorectal neoplasms at 2 predefined cutoffs of a quantitative fecal immunochemical test.
Trial Protocol
Statistical Analysis Plan
eTable 1. Full list of inclusion and exclusion criteria
eTable 2. Numbers of valid FIT results according to day of fecal sampling in relation to day of tablet intake
eTable 3. Sensitivity and specificity for detecting advanced neoplasms (colorectal cancer or advanced adenoma) after single 300 mg aspirin dose or placebo, based on at least one positive test result from up to three stool samples scheduled on day 2, 3 and 4 after intake of aspirin or placebo
eTable 4. Sensitivity (Se) and specificity (Sp) for detecting advanced neoplasms (colorectal cancer or advanced adenoma) at day 2 after single 300 mg aspirin dose or placebo.
eTable 5. Positive predictive value (PPV) and negative predictive value (NPV) in % for detecting advanced neoplasms
eTable 6. Sensitivity analysis for the effect of study site
Data Sharing Statement
References
- 1.Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394-424. doi: 10.3322/caac.21492 [DOI] [PubMed] [Google Scholar]
- 2.Hewitson P, Glasziou P, Watson E, Towler B, Irwig L. Cochrane systematic review of colorectal cancer screening using the fecal occult blood test (hemoccult): an update. Am J Gastroenterol. 2008;103(6):1541-1549. doi: 10.1111/j.1572-0241.2008.01875.x [DOI] [PubMed] [Google Scholar]
- 3.Scholefield JH, Moss SM, Mangham CM, Whynes DK, Hardcastle JD. Nottingham trial of faecal occult blood testing for colorectal cancer: a 20-year follow-up. Gut. 2012;61(7):1036-1040. doi: 10.1136/gutjnl-2011-300774 [DOI] [PubMed] [Google Scholar]
- 4.Shaukat A, Mongin SJ, Geisser MS, et al. Long-term mortality after screening for colorectal cancer. N Engl J Med. 2013;369(12):1106-1114. doi: 10.1056/NEJMoa1300720 [DOI] [PubMed] [Google Scholar]
- 5.Park DI, Ryu S, Kim YH, et al. Comparison of guaiac-based and quantitative immunochemical fecal occult blood testing in a population at average risk undergoing colorectal cancer screening. Am J Gastroenterol. 2010;105(9):2017-2025. doi: 10.1038/ajg.2010.179 [DOI] [PubMed] [Google Scholar]
- 6.Faivre J, Dancourt V, Denis B, et al. Comparison between a guaiac and three immunochemical faecal occult blood tests in screening for colorectal cancer. Eur J Cancer. 2012;48(16):2969-2976. doi: 10.1016/j.ejca.2012.04.007 [DOI] [PubMed] [Google Scholar]
- 7.Brenner H, Tao S. Superior diagnostic performance of faecal immunochemical tests for haemoglobin in a head-to-head comparison with guaiac based faecal occult blood test among 2235 participants of screening colonoscopy. Eur J Cancer. 2013;49(14):3049-3054. doi: 10.1016/j.ejca.2013.04.023 [DOI] [PubMed] [Google Scholar]
- 8.Gies A, Bhardwaj M, Stock C, Schrotz-King P, Brenner H. Quantitative fecal immunochemical tests for colorectal cancer screening. Int J Cancer. 2018;143(2):234-244. doi: 10.1002/ijc.31233 [DOI] [PubMed] [Google Scholar]
- 9.Armaroli P, Villain P, Suonio E, et al. European code against cancer, 4th edition: cancer screening. Cancer Epidemiol. 2015;39(suppl 1):S139-S152. doi: 10.1016/j.canep.2015.10.021 [DOI] [PubMed] [Google Scholar]
- 10.Robertson DJ, Lee JK, Boland CR, et al. Recommendations on fecal immunochemical testing to screen for colorectal neoplasia: a consensus statement by the US Multi-Society Task Force on colorectal cancer. Am J Gastroenterol. 2017;112(1):37-53. doi: 10.1038/ajg.2016.492 [DOI] [PubMed] [Google Scholar]
- 11.Allison JE, Fraser CG, Halloran SP, Young GP. Population screening for colorectal cancer means getting FIT: the past, present, and future of colorectal cancer screening using the fecal immunochemical test for hemoglobin (FIT). Gut Liver. 2014;8(2):117-130. doi: 10.5009/gnl.2014.8.2.117 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Haug U, Knudsen AB, Lansdorp-Vogelaar I, Kuntz KM. Development of new non-invasive tests for colorectal cancer screening: the relevance of information on adenoma detection. Int J Cancer. 2015;136(12):2864-2874. doi: 10.1002/ijc.29343 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Brenner H, Tao S, Haug U. Low-dose aspirin use and performance of immunochemical fecal occult blood tests. JAMA. 2010;304(22):2513-2520. doi: 10.1001/jama.2010.1773 [DOI] [PubMed] [Google Scholar]
- 14.Tikk K, Czock D, Haefeli WE, Kopp-Schneider A, Brenner H. Clinical trial protocol of the ASTER trial: a double-blind, randomized, placebo-controlled phase III trial evaluating the use of acetylsalicylic acid (ASA) for enhanced early detection of colorectal neoplasms. BMC Cancer. 2018;18(1):914. doi: 10.1186/s12885-018-4826-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Grobbee EJ, van der Vlugt M, van Vuuren AJ, et al. A randomised comparison of two faecal immunochemical tests in population-based colorectal cancer screening. Gut. 2017;66(11):1975-1982. doi: 10.1136/gutjnl-2016-311819 [DOI] [PubMed] [Google Scholar]
- 16.Brenner H, Chen H. Fecal occult blood versus DNA testing: indirect comparison in a colorectal cancer screening population. Clin Epidemiol. 2017;9:377-384. doi: 10.2147/CLEP.S136565 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Tao S, Brenner H. Well adjusted qualitative immunochemical faecal occult blood tests could be a promising alternative for inexpensive, high-quality colorectal cancer screening. Eur J Cancer Prev. 2013;22(4):305-310. doi: 10.1097/CEJ.0b013e32835b6991 [DOI] [PubMed] [Google Scholar]
- 18.Agresti A, Coull B. Approximate is better than “exact” for interval estimation of binomial proportions. Am Stat. 1998;52:119-126. [Google Scholar]
- 19.Agresti A, Caffo B. Simple and effective confidence intervals for proportions and difference of proportions result from adding two successes and two failures. Am Stat. 2000;54:280-288. [Google Scholar]
- 20.Package Lme4: Linear Mixed-Effects Models Using “Eigen” and S4. Version 1.1-21. https://cran.r-project.org/web/packages/lme4/lme4.pdf. Accessed April 15, 2019.
- 21.Zhang D, Lin X. Variance component testing in generalized linear mixed models for longitudinal/clustered data and other related topics In: Dunson DB, ed. Random Effect and Latent Variable Model Selection. Lecture Notes in Statistics. Vol 192 New York, NY: Springer; 2008. doi: 10.1007/978-0-387-76721-5_2 [DOI] [Google Scholar]
- 22.Levi Z, Rozen P, Hazazi R, et al. Sensitivity, but not specificity, of a quantitative immunochemical fecal occult blood test for neoplasia is slightly increased by the use of low-dose aspirin, NSAIDs, and anticoagulants. Am J Gastroenterol. 2009;104(4):933-938. doi: 10.1038/ajg.2009.14 [DOI] [PubMed] [Google Scholar]
- 23.Lee TK, Chen YC, Kuo TL. Comparison of the effect of acetylsalicylic acid on platelet function in male and female patients with ischemic stroke. Thromb Res. 1987;47(3):295-304. doi: 10.1016/0049-3848(87)90143-5 [DOI] [PubMed] [Google Scholar]
- 24.Canadian Cooperative Study Group A randomized trial of aspirin and sulfinpyrazone in threatened stroke. N Engl J Med. 1978;299(2):53-59. doi: 10.1056/NEJM197807132990201 [DOI] [PubMed] [Google Scholar]
- 25.Pace S, Sautebin L, Werz O. Sex-biased eicosanoid biology: impact for sex differences in inflammation and consequences for pharmacotherapy. Biochem Pharmacol. 2017;145:1-11. doi: 10.1016/j.bcp.2017.06.128 [DOI] [PubMed] [Google Scholar]
- 26.Patti G, De Caterina R, Abbate R, et al. ; Working Group on Thrombosis of the Italian Society of Cardiology . Platelet function and long-term antiplatelet therapy in women: is there a gender-specificity? a ‘state-of-the-art’ paper. Eur Heart J. 2014;35(33):2213-23b. doi: 10.1093/eurheartj/ehu279 [DOI] [PubMed] [Google Scholar]
- 27.Otahbachi M, Simoni J, Simoni G, et al. Gender differences in platelet aggregation in healthy individuals. J Thromb Thrombolysis. 2010;30(2):184-191. doi: 10.1007/s11239-009-0436-x [DOI] [PubMed] [Google Scholar]
- 28.Berlin G, Hammar M, Tapper L, Tynngård N. Effects of age, gender and menstrual cycle on platelet function assessed by impedance aggregometry. Platelets. 2018;30(4):473-479. doi: 10.1080/09537104.2018.1466387 [DOI] [PubMed] [Google Scholar]
- 29.Sadik R, Abrahamsson H, Stotzer PO. Gender differences in gut transit shown with a newly developed radiological procedure. Scand J Gastroenterol. 2003;38(1):36-42. doi: 10.1080/00365520310000410 [DOI] [PubMed] [Google Scholar]
- 30.McCrea GL, Miaskowski C, Stotts NA, Macera L, Varma MG. A review of the literature on gender and age differences in the prevalence and characteristics of constipation in North America. J Pain Symptom Manage. 2009;37(4):737-745. doi: 10.1016/j.jpainsymman.2008.04.016 [DOI] [PubMed] [Google Scholar]
- 31.Schoenfeld P, Cash B, Flood A, et al. ; CONCeRN Study Investigators . Colonoscopic screening of average-risk women for colorectal neoplasia. N Engl J Med. 2005;352(20):2061-2068. doi: 10.1056/NEJMoa042990 [DOI] [PubMed] [Google Scholar]
- 32.Peery AF, Sandler RS, Galanko JA, et al. Risk factors for hemorrhoids on screening colonoscopy. PLoS One. 2015;10(9):e0139100. doi: 10.1371/journal.pone.0139100 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Trial Protocol
Statistical Analysis Plan
eTable 1. Full list of inclusion and exclusion criteria
eTable 2. Numbers of valid FIT results according to day of fecal sampling in relation to day of tablet intake
eTable 3. Sensitivity and specificity for detecting advanced neoplasms (colorectal cancer or advanced adenoma) after single 300 mg aspirin dose or placebo, based on at least one positive test result from up to three stool samples scheduled on day 2, 3 and 4 after intake of aspirin or placebo
eTable 4. Sensitivity (Se) and specificity (Sp) for detecting advanced neoplasms (colorectal cancer or advanced adenoma) at day 2 after single 300 mg aspirin dose or placebo.
eTable 5. Positive predictive value (PPV) and negative predictive value (NPV) in % for detecting advanced neoplasms
eTable 6. Sensitivity analysis for the effect of study site
Data Sharing Statement
