- 508 对活检—根治标本配对：把病例压缩成「最高分级组（GG）」会丢掉各 GG 的定量负荷信息。
- 在已校正最高活检 GG 与非目标 GG3–5 负荷后：GG4 阳性单位数与根治后 GG4–5 相关（OR 1.72）；GG3 与 pT3+ 相关（OR 1.72）；GG5 与根治后 GG5 相关（OR 3.46）。
- 癌长指标方向一致；318 例严格质控敏感分析估计相近。
- 在已含最高 GG 的模型上再加目标 GG 负荷，乐观校正 AUC 分别提升 0.029 / 0.042 / 0.010——提示日常活检报告中已有的分级组特异性定量字段对根治分级与局部分期有互补价值。
收录范围：PubMed 英文摘要及中文翻译（abstract-only）。原文为订阅制（ScienceDirect，Unpaywall 显示 closed），未采集全文、图表。
摘要
常规前列腺活检报告含有定量信息，但当病例被简化为单一「最高分级组（Grade Group, GG）」时这些信息会丢失。本研究评估常规活检报告中的分级组特异性肿瘤负荷，是否在最高活检 GG 之外，仍能提供关于根治性前列腺切除术（RP）病理的额外信息。本回顾性单中心研究纳入 2014–2026 年可分析的 508 对活检—RP 配对病例。对 GG3、GG4、GG5，负荷重建为「阳性报告单位数」与「癌长总和」。采用 Firth 惩罚 logistic 回归，校正最高活检 GG 与非目标 GG3–GG5 负荷。主要终点：GG4 负荷对应 RP GG4–5；GG3 负荷对应 pT3 及以上；GG5 负荷对应 RP GG5。计数模型的比值比（OR）按「每增加 1 个目标阳性报告单位」表述。GG4 计数与 RP GG4–5 相关（OR 1.72；95% CI 1.29–2.29）；GG3 计数与 pT3 及以上相关（OR 1.72；95% CI 1.41–2.10；n=473）；GG5 计数与 RP GG5 相关（OR 3.46；95% CI 1.40–8.59）。相应癌长指标关联方向一致；严格质控后的 318 例敏感分析得到相近估计。在已包含最高活检 GG 与非目标 GG 负荷的模型中加入目标 GG 负荷后，乐观校正 AUC 分别提高：GG4 计数 0.029、GG3 计数 0.042、GG5 计数 0.010。因此，常规活检报告中已存在的分级组特异性负荷，可在最高活检 GG 之外，为后续 RP 分级与局部分期提供互补信息。
- 508 paired biopsy–RP cases: reducing a case to the single highest Grade Group (GG) discards quantitative GG-specific burden already in routine reports.
- After adjusting for highest biopsy GG and non-target GG3–5 burden: GG4 positive-unit count associated with RP GG4–5 (OR 1.72); GG3 with pT3+ (OR 1.72); GG5 with RP GG5 (OR 3.46).
- Cancer-length measures were concordant; a strict 318-case QC sensitivity subset gave similar estimates.
- Adding target GG burden to models with highest GG yielded optimism-corrected AUC gains of 0.029 / 0.042 / 0.010 — complementary RP grade/stage information already present in routine biopsy reports.
Scope: PubMed abstract only. The article is subscription-only (ScienceDirect; Unpaywall: closed), so full text, figures and tables were not collected.
Abstract
Routine prostate biopsy reports contain quantitative information that is lost when a case is reduced to the single highest Grade Group (GG). We examined whether GG-specific tumor burden in routine biopsy reports provides information on radical prostatectomy (RP) pathology beyond the highest biopsy GG. This retrospective single-institution study included 508 analyzable paired biopsy-RP cases from 2014 to 2026. For GG3, GG4, and GG5, burden was reconstructed as the number of positive reporting units and summed cancer length. Firth penalized logistic regression adjusted for highest biopsy GG and non-target GG3-GG5 burden. The focused endpoints were RP GG4-5 for GG4 burden, pT3 or higher disease for GG3 burden, and RP GG5 for GG5 burden. For count models, odds ratios (ORs) were expressed per additional target reporting unit. GG4 count was associated with RP GG4-5 (OR, 1.72; 95% confidence interval [CI], 1.29-2.29), GG3 count with pT3 or higher disease (OR, 1.72; 95% CI, 1.41-2.10; n=473), and GG5 count with RP GG5 (OR, 3.46; 95% CI, 1.40-8.59). Corresponding cancer-length measures showed concordant associations, and the strict 318-case quality-control sensitivity subset yielded similar estimates. Adding target GG burden to models already containing highest biopsy GG and non-target GG burden yielded optimism-corrected area-under-the-curve gains of 0.029 for GG4 count, 0.042 for GG3 count, and 0.010 for GG5 count. Grade-specific burden already present in routine biopsy reports therefore provides complementary information on subsequent RP grade and local stage beyond the highest biopsy GG alone.