Scholarlink

Background <p>Research on drugs that contribute to colon adenocarcinoma (COAD) remains insufficient. This study seeks to find drugs associated with this condition and to elucidate common key genes implicated in both drugs and COAD.</p> Methods <p>Through the integration of real-world pharmacovigilance and network toxicology, we identified the associations between drugs and COAD occurrence. Subsequently, key genes in COAD were identified using bioinformatics analyses, including univariate Cox, expression, and Kaplan-Meier survival analyses. Finally, a single gene study was performed to investigate the functions of this gene in COAD.</p> Results <p>Direct oral anticoagulants (DOACs)-related drugs—dabigatran (reporting odds ratio (ROR) = 5.13, 95% confidence interval (CI) = 4.43–5.93), rivaroxaban (ROR = 1.92, 95% CI  1.60–2.31), and apixaban (ROR = 3.59, 95% CI  3.14–4.11)—were identified to be associated with COAD, exhibiting moderate toxicity classified as 3 and 4. A total of 3 key genes linked to drugs—CA2, LGALS2, and PAK6—were discovered in COAD. Elevated levels of CA2 and LGALS2 were associated with improved survival in COAD, whereas PAK6 exhibited the opposite correlation. Molecular docking and MD simulations indicated that the CA2-apixaban exhibited a stable structure with a binding energy of – 8.033 Kcal/mol. Ultimately, the CA2 diverse expression groups exhibited significant disparities in immune infiltration, drug sensitivity, and mutation profiles.</p> Conclusion <p>Dabigatran, rivaroxaban, and apixaban showed statistical associations with COAD reports, with CA2, LGALS2, and PAK6 identified as potential prognostic targets. However, detection bias from gastrointestinal bleeding may partly explain these signals by uncovering occult tumors. Our findings are hypothesis-generating and warrant further studies to rule out surveillance bias.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

An in-depth exploration of the association between dabigatran, rivaroxaban, apixaban and colon adenocarcinoma based on real-world data and network toxicology analysis

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月5日

查看全文
  • Jiawen Liu,
  • Manduo Ouyang,
  • Xinchang Shangguan,
  • Chaorong Xue,
  • Junrong Zhang,
  • Xianqiang Chen

摘要

Background

Research on drugs that contribute to colon adenocarcinoma (COAD) remains insufficient. This study seeks to find drugs associated with this condition and to elucidate common key genes implicated in both drugs and COAD.

Methods

Through the integration of real-world pharmacovigilance and network toxicology, we identified the associations between drugs and COAD occurrence. Subsequently, key genes in COAD were identified using bioinformatics analyses, including univariate Cox, expression, and Kaplan-Meier survival analyses. Finally, a single gene study was performed to investigate the functions of this gene in COAD.

Results

Direct oral anticoagulants (DOACs)-related drugs—dabigatran (reporting odds ratio (ROR) = 5.13, 95% confidence interval (CI) = 4.43–5.93), rivaroxaban (ROR = 1.92, 95% CI  1.60–2.31), and apixaban (ROR = 3.59, 95% CI  3.14–4.11)—were identified to be associated with COAD, exhibiting moderate toxicity classified as 3 and 4. A total of 3 key genes linked to drugs—CA2, LGALS2, and PAK6—were discovered in COAD. Elevated levels of CA2 and LGALS2 were associated with improved survival in COAD, whereas PAK6 exhibited the opposite correlation. Molecular docking and MD simulations indicated that the CA2-apixaban exhibited a stable structure with a binding energy of – 8.033 Kcal/mol. Ultimately, the CA2 diverse expression groups exhibited significant disparities in immune infiltration, drug sensitivity, and mutation profiles.

Conclusion

Dabigatran, rivaroxaban, and apixaban showed statistical associations with COAD reports, with CA2, LGALS2, and PAK6 identified as potential prognostic targets. However, detection bias from gastrointestinal bleeding may partly explain these signals by uncovering occult tumors. Our findings are hypothesis-generating and warrant further studies to rule out surveillance bias.