Polyketide synthase positive Escherichia coli one-time measurement in stool is not informative of colorectal cancer risk in a screening setting

Willemijn de Klaver, Meike de Wit, Anne Bolijn, Marianne Tijssen, Pien Delis-van Diemen, Margriet Lemmens, Manon C. W. Spaander, Evelien Dekker, Monique E. van Leerdam, Veerle M. H. Coupé, Ruben van Boxtel, Hans Clevers, Beatriz Carvalho, Gerrit A. Meijer

Research output: Contribution to journalArticleAcademicpeer-review

Abstract

Environmental factors like the pathogenicity island polyketide synthase positive (pks+) Escherichia coli (E. coli) could have potential for risk stratification in colorectal cancer (CRC) screening. The association between pks+ E. coli measured in fecal immunochemical test (FIT) samples and the detection of advanced neoplasia (AN) at colonoscopy was investigated. Biobanked FIT samples were analyzed for both presence of E. coli and pks+ E. coli and correlated with colonoscopy findings; 5020 CRC screening participants were included. Controls were participants in which no relevant lesion was detected because of FIT-negative results (cut-off ≥15 μg Hb/g feces), a negative colonoscopy, or a colonoscopy during which only a nonadvanced polyp was detected. Cases were participants with AN [CRC, advanced adenoma (AA), or advanced serrated polyp (ASP)]. Existing DNA isolation and quantitative polymerase chain reaction (qPCR) procedures were used for the detection of E. coli and pks+ E. coli in stool. A total of 4542 (90.2%) individuals were E. coli positive, and 1322 (26.2%) were pks+ E. coli positive. The prevalence of E. coli in FIT samples from individuals with AN was 92.9% compared to 89.7% in FIT samples of controls (p = 0.010). The prevalence of pks+ E. coli in FIT samples from individuals with AN (28.6%) and controls (25.9%) was not significantly different (p = 0.13). The prevalences of pks+ E. coli in FIT samples from individuals with CRC, AA, or ASP were 29.6%, 28.3%, and 32.1%, respectively. In conclusion, the prevalence of pks+ E. coli in a screening population was 26.2% and did not differ significantly between individuals with AN and controls. These findings disqualify the straightforward option of using a snapshot measurement of pks+ E. coli in FIT samples as a stratification biomarker for CRC risk.

Original languageEnglish
Pages (from-to)217-225
Number of pages9
JournalJournal of pathology
Volume263
Issue number2
Early online date2024
DOIs
Publication statusPublished - Jun 2024

Keywords

  • advanced neoplasia
  • biomarker
  • colorectal cancer
  • pks+ E. coli
  • risk stratification
  • screening

Cite this