Scholarlink

Rationale <p>It is challenging to identify which patients with Stage II colorectal cancer (T3N0M0 and T4N0M0) are at high risk of recurrence and might benefit from additional therapies. This preliminary study examines whether multiparametric [<sup>18</sup>F]FDG PET/CT is superior to T-stage in predicting overall survival in this patient group.</p> Materials and methods <p>This multicentre, prospective observational study included 66 patients (41 male, 25 female, mean age 69.1 ± 10.3yrs) with biopsy-proven Stage II colorectal cancer who underwent [<sup>18</sup>F]FDG PET/CT prior to resection. Kaplan-Meier analysis was performed on PET and image texture parameters and clinico-histopathological markers to identify associations with survival. P-values were adjusted using the Benjamini-Hochberg procedure, and the most statistically significant radiomic parameters underwent 3-fold cross-validation. Multivariate Cox regression analysis was used to determine independence of prognostic markers.</p> Results <p>49 patients survived the follow up period, with a mean overall survival of 88.4 (± 42.3 months). Univariate analysis showed no significant survival association with clinical variables such as patient sex (<i>p</i> = 0.295), age (<i>p</i> = 0.085), tumour side (<i>p</i> = 0.662), location (<i>p</i> = 0.848) or volume (<i>p</i> = 0.782), or high-risk histopathological metrics including low number lymph node sampling (<i>p</i> = 0.363), perineural, lymphovascular or extramural tumour invasion (<i>p</i> = 0.196), poor tumour differentiation (<i>p</i> = 0.372), and tumour perforation (<i>p</i> = 0.475). No significant survival difference was observed between T3 and T4-stages (<i>p</i> = 0.748). An increased mortality risk was observed for patients with a pixel mean intensity &lt; 20.390 on CT texture (coarse texture scale), HR: 3.40 (1.36–8.49); <i>p</i> = 0.005, and where SUVmax ≥ 25.560&#xa0;g/mL on [<sup>18</sup>F]FDG PET/CT, HR: 3.45 (1.31–9.12); <i>p</i> = 0.008. These parameters remained significant after 3-fold cross validation and were independent predictors of survival after Multivariate Cox regression analysis. A higher mortality risk was indicated when the parameters were combined (5 of 66 patients), hazard ratio: 5.44 (1.75–16.91); <i>p</i> = 0.001.</p> Conclusion <p>Multiparametric [<sup>18</sup>F]FDG PET/CT can potentially provide prognostic markers for patients with stage II colorectal cancer with superior risk stratification compared to T-stage.</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

Texture analysis of [18F]FDG PET/CT may stratify risk in stage II colorectal cancer – discovery findings

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年10月6日

查看全文
  • Louise Rice,
  • Balaji Ganeshan,
  • Simon Wan,
  • Shih-hsin Chen,
  • David M.L. Lilburn,
  • Manuel Rodriguez-Justo,
  • Stuart Taylor,
  • Robert I. Shortman,
  • Raymond Endozo,
  • Saif Khan,
  • Luke R. Hoy,
  • Daren Francis,
  • Tan Arulampalam,
  • Nicholas Reay-Jones,
  • Kenneth A. Miles,
  • Ashley M. Groves

摘要

Rationale

It is challenging to identify which patients with Stage II colorectal cancer (T3N0M0 and T4N0M0) are at high risk of recurrence and might benefit from additional therapies. This preliminary study examines whether multiparametric [18F]FDG PET/CT is superior to T-stage in predicting overall survival in this patient group.

Materials and methods

This multicentre, prospective observational study included 66 patients (41 male, 25 female, mean age 69.1 ± 10.3yrs) with biopsy-proven Stage II colorectal cancer who underwent [18F]FDG PET/CT prior to resection. Kaplan-Meier analysis was performed on PET and image texture parameters and clinico-histopathological markers to identify associations with survival. P-values were adjusted using the Benjamini-Hochberg procedure, and the most statistically significant radiomic parameters underwent 3-fold cross-validation. Multivariate Cox regression analysis was used to determine independence of prognostic markers.

Results

49 patients survived the follow up period, with a mean overall survival of 88.4 (± 42.3 months). Univariate analysis showed no significant survival association with clinical variables such as patient sex (p = 0.295), age (p = 0.085), tumour side (p = 0.662), location (p = 0.848) or volume (p = 0.782), or high-risk histopathological metrics including low number lymph node sampling (p = 0.363), perineural, lymphovascular or extramural tumour invasion (p = 0.196), poor tumour differentiation (p = 0.372), and tumour perforation (p = 0.475). No significant survival difference was observed between T3 and T4-stages (p = 0.748). An increased mortality risk was observed for patients with a pixel mean intensity < 20.390 on CT texture (coarse texture scale), HR: 3.40 (1.36–8.49); p = 0.005, and where SUVmax ≥ 25.560 g/mL on [18F]FDG PET/CT, HR: 3.45 (1.31–9.12); p = 0.008. These parameters remained significant after 3-fold cross validation and were independent predictors of survival after Multivariate Cox regression analysis. A higher mortality risk was indicated when the parameters were combined (5 of 66 patients), hazard ratio: 5.44 (1.75–16.91); p = 0.001.

Conclusion

Multiparametric [18F]FDG PET/CT can potentially provide prognostic markers for patients with stage II colorectal cancer with superior risk stratification compared to T-stage.