Scholarlink

Background <p>To refine a computed tomography (CT)-based algorithm for distinguishing clear cell renal cell carcinoma (ccRCC) from other small renal masses (SRMs ≤ 4 cm) through a simplified heterogeneity score and optimized attenuation criteria.</p> Patients and Methods <p>This multicenter retrospective study included 257 SRMs as an initial cohort and 149 for validation. A simplified heterogeneity score was developed incorporating persistent hypodense areas and tumor-parenchymal enhancement heterogeneity. Attenuation score was re-stratified using classification and regression tree analysis. Two modified systems were created: Modified Score 1 (original attenuation score + simplified heterogeneity score) and Modified Score 2 (new attenuation score + simplified heterogeneity score), compared against the original CT score.</p> Results <p>Both modified scores outperformed the original system. Modified Score 1 achieved higher sensitivity (cutoff ≥ 4: 86.3% versus 69.1%, <i>p</i> &lt; 0.001) and area under the receiver operating characteristic curve (AUC) (0.880 versus 0.793, <i>p</i> &lt; 0.001); Modified Score 2 showed better specificity (85.6% versus 75.4%, <i>p</i> = 0.017) and AUC (0.907 versus 0.793, <i>p</i> &lt; 0.001). The simplified heterogeneity score improved interobserver agreement (kappa: 0.736 versus 0.282) and drove improvements in the corresponding CT scores from 0.466 to 0.663 (score 1) and 0.577 (score 2) and from 0.459 to 0.833 and 0.712 in the five-tiered and dichotomous systems, respectively. The validation cohort confirmed the superior performance of the modified scores, including for SRMs ≤ 3 cm. Moreover, a simplified heterogeneity score of 3 demonstrated high specificity (90.6–94.0%) for diagnosing malignant SRMs.</p> Conclusions <p>The modified CT algorithms significantly improve diagnostic accuracy and reproducibility for ccRCC, supporting clinical decisions between active surveillance and intervention.</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

Improving Clear Cell Renal Cell Carcinoma Diagnosis: Modified CT Algorithms with Simplified Heterogeneity Scoring

  • 期刊论文

  • 发表日期：

    2026年5月25日

查看全文
  • Wenjie Zhang,
  • Qichun Chen,
  • Xianli Zhu,
  • Xinyan Li,
  • Xiaofei Wang,
  • Heng Ma,
  • Xinhong Song,
  • Zhaofeng Zheng,
  • Jianyi Qu

摘要

Background

To refine a computed tomography (CT)-based algorithm for distinguishing clear cell renal cell carcinoma (ccRCC) from other small renal masses (SRMs ≤ 4 cm) through a simplified heterogeneity score and optimized attenuation criteria.

Patients and Methods

This multicenter retrospective study included 257 SRMs as an initial cohort and 149 for validation. A simplified heterogeneity score was developed incorporating persistent hypodense areas and tumor-parenchymal enhancement heterogeneity. Attenuation score was re-stratified using classification and regression tree analysis. Two modified systems were created: Modified Score 1 (original attenuation score + simplified heterogeneity score) and Modified Score 2 (new attenuation score + simplified heterogeneity score), compared against the original CT score.

Results

Both modified scores outperformed the original system. Modified Score 1 achieved higher sensitivity (cutoff ≥ 4: 86.3% versus 69.1%, p < 0.001) and area under the receiver operating characteristic curve (AUC) (0.880 versus 0.793, p < 0.001); Modified Score 2 showed better specificity (85.6% versus 75.4%, p = 0.017) and AUC (0.907 versus 0.793, p < 0.001). The simplified heterogeneity score improved interobserver agreement (kappa: 0.736 versus 0.282) and drove improvements in the corresponding CT scores from 0.466 to 0.663 (score 1) and 0.577 (score 2) and from 0.459 to 0.833 and 0.712 in the five-tiered and dichotomous systems, respectively. The validation cohort confirmed the superior performance of the modified scores, including for SRMs ≤ 3 cm. Moreover, a simplified heterogeneity score of 3 demonstrated high specificity (90.6–94.0%) for diagnosing malignant SRMs.

Conclusions

The modified CT algorithms significantly improve diagnostic accuracy and reproducibility for ccRCC, supporting clinical decisions between active surveillance and intervention.