Bowel Preparation for Colonoscopy Changes Serum Composition as Detected by Thermal Liquid Biopsy and Fluorescence Spectroscopy

Abstract

(1) Background: About 50% of prescribed colonoscopies report no pathological findings. A secondary screening test after fecal immunochemical test positivity (FIT+) would be required. Considering thermal liquid biopsy (TLB) as a potential secondary test, the aim of this work was to study possible interferences of colonoscopy bowel preparation on TLB outcome on a retrospective study; (2) Methods: Three groups were studied: 1/514 FIT(+) patients enrolled in a colorectal screening program (CN and CP with normal and pathological colonoscopy, respectively), with blood samples obtained just before colonoscopy and after bowel preparation; 2/55 patients from the CN group with blood sample redrawn after only standard 8-10 h fasting and no bowel preparation (CNR); and 3/55 blood donors from the biobank considered as a healthy control group; (3) Results: The results showed that from the 514 patients undergoing colonoscopy, 247 had CN and 267 had CP. TLB parameters in these two groups were similar but different from those of the blood donors. The resampled patients (with normal colonoscopy and no bowel preparation) had similar TLB parameters to those of the blood donors. TLB parameters together with fluorescence spectra and other serum indicators (albumin and C-reactive protein) confirmed the statistically significant differences between normal colonoscopy patients with and without bowel preparation; (4) Conclusions: Bowel preparation seemed to alter serum protein levels and altered TLB parameters (different from a healthy subject). The diagnostic capability of other liquid-biopsy-based methods might also be compromised. Blood extraction after bowel preparation for colonoscopy should be avoided.

Keywords: colonoscopy; colorectal cancer; fluorescence spectroscopy; thermal liquid biopsy.

Figure 1

Sample study population flow chart.

Figure 2

Thermograms from patients with (A) normal (CN, n = 247) and (B) pathological (CP, n = 267) colonoscopy diagnosis.

Figure 3

Average and standard deviation of the thermograms from healthy blood donors (BD, green, n = 55), and from patients with a normal colonoscopy diagnosis (CN, black, n = 247).

Figure 4

Box-plot diagrams indicating Q1, Q2, and Q3, together with the average value (square) for each TLB parameter, illustrating the distribution of the p-values (asterisks indicate the order of magnitude of p-values) between patients with a normal colonoscopy diagnosis (CN, black) and healthy blood donors (BD, green). It was calculated according to Student’s t-test or a Wilcoxon test for independent samples, depending on the normality character of the parameter distribution. Note: n.s.: p-value is not significant; **: p-value = 0.050–0.010; ***: p-value = 0.009–0.001.

Figure 5

(A) Average and standard deviation of the thermograms from healthy BD (green, n = 55) and patients (blue, n = 55); (B) average and standard deviation of the thermograms from 55 selected patients with a normal colonoscopy diagnosis in two blood draw conditions: with colonoscopy preparation (CN, black) and with standard analytical conditions (CNR, blue).

Figure 6

Box-plot diagrams indicating Q1, Q2, and Q3, together with the average value (square) for each TLB parameter, illustrating the difference between 55 samples without preparation and 55 samples with preparation (asterisks indicate the order of magnitude of the p-value). Note: n.s.: p-value is not significant; **: p-value = 0.050–0.010; ***: p-value = 0.009–0.001. p-values were calculated according to Student’s t-test or a Wilcoxon test, both for dependent samples, depending on the normality character of the parameter distribution.