Using fecal immmunochemical cartridges for gut microbiome analysis within a colorectal cancer screening program

Abstract

The colorectal cancer (CRC) screening program B-PREDICT is an invited two-stage screening project using a fecal immunochemical test (FIT) for initial screening followed by a colonoscopy for those with a positive FIT. Since the gut microbiome likely plays a role in the etiology of CRC, microbiome-based biomarkers in combination with FIT could be a promising tool for optimizing CRC screening. Therefore, we evaluated the usability of FIT cartridges for microbiome analysis and compared it to Stool Collection and Preservation Tubes. Corresponding FIT cartridges as well as Stool Collection and Preservation Tubes were collected from participants of the B-PREDICT screening program to perform 16S rRNA gene sequencing. We calculated intraclass correlation coefficients (ICCs) based on center log ratio transformed abundances and used ALDEx2 to test for significantly differential abundant taxa between the two sample types. Additionally, FIT and Stool Collection and Preservation Tube triplicate samples were obtained from volunteers to estimate variance components of microbial abundances. FIT and Preservation Tube samples produce highly similar microbiome profiles which cluster according to subject. Significant differences between the two sample types can be found for abundances of some bacterial taxa (e.g. 33 genera) but are minor compared to the differences between the subjects. Analysis of triplicate samples revealed slightly worse repeatability of results for FIT than for Preservation Tube samples. Our findings indicate that FIT cartridges are appropriate for gut microbiome analysis nested within CRC screening programs.

Keywords: 16S rRNA sequencing; FIT; Microbiome; NORGEN; colorectal cancer.

Figure 1.

Schematic workflow of the presented study.

Figure 2.

Figure based on patient samples. ASVs are colored according to their phylum and sized according to their abundance. A: Scatterplot of ASVs with mean center log ratio (CLR) transformed abundances of FIT and Norgen samples B: Scatterplot showing the relationship between each ASV‘s ICC estimate (with confidence interval) and the sum of the logarithms of its abundances. ICCs above 0.9 (dashed line) indicate excellent reliability. A boxplot of the ICCs is provided additionally. C: ICC estimates and confidence intervals for all calculated metrics computed on FIT and Norgen samples. Table 2 is giving all calculated beta diversities. The ICC of the Shannon, Simpson and Inverse Simpson index are all above 0.75, with the Shannon index providing the highest reliability between FIT and Norgen. The Bray-Curtis dissimilarity and the Jaccard index results are almost in perfect agreement. Unweighted UniFrac also displays an excellent ICC, while the weighted version results in only good reliability. D, E, F: Violin plots comparing Norgen and FIT samples based on the richness of the samples (d), number of reads per sample (e) and the prevalence of ASVs identified in at least 5% of the samples (f).

Figure 3.

A: [gray box] Boxplot of distances between samples of different subjects (regardless of sample type), [green box] between FIT and Norgen samples of the same patient and [red box] between FIT and Norgen samples, [blue box] between FIT samples and [yellow box] between Norgen samples of the same volunteer. B: Result of a hierarchical clustering algorithm based on the distances between all samples. Samples originating from the same subject are connected with colored bars.

Figure 4.

A: Taxonomic tree displaying significant differences between FIT and Norgen samples based on the ALDEX analysis. Taxa are labeled with an ID and the first letters of their name. Full taxa names are given in Table 3. B: Scatterplots of the first four principal components extracted from the volunteer samples. Each of the five volunteers is represented by a number.

Figure 5.

Analyses of triplicate samples of the five volunteers (3x FIT, 3x Norgen). A: Ternary plot displaying the variance components of each ASV as a single point based on a linear model with subject and sample type as explanatory variables. Considering only ASVs detected in at least three volunteer samples. ASVs are colored according to the results of the corresponding significance tests on the patient samples. B: Boxplots of the variance components used in A. C: Violin plots comparing the proportions explained by subject in separate models for FIT and Norgen. Only ASVs detected in at least three samples of the respective sample type were used.