Scholarlink

Purpose <p>To evaluate the predictive value of three-dimensional (3D) CT-based volumetric shrinkage for post-chemotherapy retroperitoneal lymph node dissection (RPLND) pathology in patients with non-seminomatous germ cell tumors (NSGCTs).</p> Methods <p>We retrospectively reviewed patients who underwent post-chemotherapy RPLND at our tertiary care center between 2016 and 2025, of whom 52 were included in the final analysis. Retroperitoneal lymph node volumes were segmented using 3D Slicer software on pre- and post-chemotherapy contrast-enhanced CT scans. Volumetric shrinkage (Δ%Volume) and serum tumor marker changes (ΔAFP, Δβ-hCG, ΔLDH) were evaluated. Residual pathology was classified as necrosis/fibrosis, teratoma, or viable germ cell tumor (GCT). Internal validation of the predictive model was performed using 1000 bootstrap resamples.</p> Results <p>Final pathology revealed necrosis/fibrosis in 19 (36.5%) patients, teratoma in 23 (44.2%), and viable GCT in 10 (19.2%). Median volumetric shrinkage was the greatest in the necrosis/fibrosis group (68.0%), intermediate in the teratoma group (54.4%), and the lowest in the viable GCT group (28.0%) (<i>p</i> = 0.001). In multivariable multinomial logistic regression, Δ%Volume was the only independent predictor of viable GCT with each 10% increment in volumetric shrinkage associated with a 51.1% reduction in the odds of viable residual disease (OR = 0.489, 95% CI 0.263–0.914, <i>p </i>= 0.025). ROC analysis demonstrated good discriminatory performance for predicting viable GCT (AUC = 0.904, 95% CI 0.819–0.988), with 90.0% sensitivity and 76.2% specificity at the optimal cut-off of 49.5% volumetric shrinkage. Bootstrap internal validation demonstrated limited optimism, yielding an optimism-corrected AUC of 0.866.</p> Conclusion <p>3D CT-based volumetric shrinkage was independently associated with viable residual disease and demonstrated good discriminatory performance for identifying viable GCT after chemotherapy. As a simple quantitative imaging biomarker, volumetric assessment may complement preoperative risk stratification in patients undergoing post-chemotherapy RPLND; however, external validation is required before routine clinical implementation.</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

3D CT-based volumetric shrinkage predicts viable residual disease after chemotherapy in non-seminomatous germ cell tumors

  • 期刊论文

  • 发表日期：

    2026年7月21日

查看全文
  • Kadir Can Sahin,
  • Feyyaz Irmak,
  • Berk Yilmaz,
  • Muhammed Fatih Simsekoglu,
  • Cetin Demirdag,
  • Bulent Onal

摘要

Purpose

To evaluate the predictive value of three-dimensional (3D) CT-based volumetric shrinkage for post-chemotherapy retroperitoneal lymph node dissection (RPLND) pathology in patients with non-seminomatous germ cell tumors (NSGCTs).

Methods

We retrospectively reviewed patients who underwent post-chemotherapy RPLND at our tertiary care center between 2016 and 2025, of whom 52 were included in the final analysis. Retroperitoneal lymph node volumes were segmented using 3D Slicer software on pre- and post-chemotherapy contrast-enhanced CT scans. Volumetric shrinkage (Δ%Volume) and serum tumor marker changes (ΔAFP, Δβ-hCG, ΔLDH) were evaluated. Residual pathology was classified as necrosis/fibrosis, teratoma, or viable germ cell tumor (GCT). Internal validation of the predictive model was performed using 1000 bootstrap resamples.

Results

Final pathology revealed necrosis/fibrosis in 19 (36.5%) patients, teratoma in 23 (44.2%), and viable GCT in 10 (19.2%). Median volumetric shrinkage was the greatest in the necrosis/fibrosis group (68.0%), intermediate in the teratoma group (54.4%), and the lowest in the viable GCT group (28.0%) (p = 0.001). In multivariable multinomial logistic regression, Δ%Volume was the only independent predictor of viable GCT with each 10% increment in volumetric shrinkage associated with a 51.1% reduction in the odds of viable residual disease (OR = 0.489, 95% CI 0.263–0.914, p = 0.025). ROC analysis demonstrated good discriminatory performance for predicting viable GCT (AUC = 0.904, 95% CI 0.819–0.988), with 90.0% sensitivity and 76.2% specificity at the optimal cut-off of 49.5% volumetric shrinkage. Bootstrap internal validation demonstrated limited optimism, yielding an optimism-corrected AUC of 0.866.

Conclusion

3D CT-based volumetric shrinkage was independently associated with viable residual disease and demonstrated good discriminatory performance for identifying viable GCT after chemotherapy. As a simple quantitative imaging biomarker, volumetric assessment may complement preoperative risk stratification in patients undergoing post-chemotherapy RPLND; however, external validation is required before routine clinical implementation.