The growth pattern (confluent/expansile versus infiltrative) in primary ovarian mucinous carcinoma (OMC) is prognostically important, and the International Collaboration on Cancer Reporting (ICCR) currently recommends recording the percentage of infiltrative growth in this tumor type. Histologic grading of OMC is controversial with no single approach widely accepted or currently recognized by the World Health Organization Classification of Tumours. Since ovarian carcinoma grade is often considered in clinical decision-making, previous literature has recommended incorporating clinically relevant tumor parameters such as growth pattern into the OMC grade. We herein validate this approach, termed Growth-Based Grade (GBG), in an independent, well-annotated cohort from 2 institutions. OMCs with available histologic material underwent review and grading by Silverberg, International Federation of Obstetrics and Gynecology (FIGO), and GBG schema. GBG categorizes OMCs as low-grade (GBG-LG, confluent/expansile growth, or ≤10% infiltrative invasion) or high-grade (GBG-HG, infiltrative growth in >10% of tumor). The cohort consisted of 74 OMCs, 53 designated as GBG-LG, and 21 as GBG-HG. Using Silverberg grading, the cohort had 42 (57%) grade 1, 28 (38%) grade 2, and 4 (5%) grade 3 OMCs. Using FIGO grading, 50 (68%) OMCs were grade 1, 23 (31%) grade 2, and 1 (1%) grade 3. Follow-up data was available in 68 patients, of which 15 (22%) had cancer recurrence. GBG-HG tumors were far more likely to recur compared with GBG-LG tumors (57% vs. 6%; χ²P<0.0001). Silverberg and FIGO grading systems also correlated with progression-free survival in univariate analysis, but multivariate analysis showed only GBG to be significant (hazard ratio: 10.9; Cox proportional regression P=0.0004). Seven patients (10%) died of disease, all of whom had GBG-HG (log-rank P<0.0001). Multivariate analysis showed that the percentage of infiltrative growth was the only factor predictive of disease-specific survival (hazard ratio: 25.5, Cox P=0.02). Adding nuclear atypia to GBG categories did not improve prognostication. Our study validates the prognostic value of the GBG system for both disease-free survival and disease-specific survival in OMC, which outperformed Silverberg and FIGO grades in multivariate analysis. Thus, GBG should be the preferred method for tumor grading.

原发性卵巢粘液性癌 (OMC) 的生长模式（融合/扩张与浸润）对预后很重要，国际癌症报告协作组织 (ICCR) 目前建议记录浸润性生长的百分比在这种肿瘤类型中。OMC 的组织学分级存在争议，目前没有一种方法被世界卫生组织肿瘤分类广泛接受或认可。由于在临床决策中经常考虑卵巢癌分级，以前的文献建议将临床相关的肿瘤参数（如生长模式）纳入 OMC 分级。我们在此验证了这种方法，称为基于增长的等级 (GBG)，在来自 2 个机构的独立、注释良好的队列中。具有可用组织学材料的 OMC 接受了 Silverberg、国际妇产科联合会 (FIGO) 和 GBG 模式的审查和分级。GBG 将 OMC 分类为低等级（GBG-LG，融合/膨胀生长，或≤10% 浸润性侵袭）或高等级（GBG-HG，浸润性生长）在 >10% 的肿瘤中）。该队列由 74 个 OMC 组成，53 个指定为 GBG-LG，21 个指定为 GBG-HG。使用 Silverberg 分级，该队列有 42 个 (57%) 1 级、28 个 (38%) 2 级和 4 个 (5%) 3 级 OMC。使用FIGO分级，50个（68%）OMC为1级，23个（31%）2级和1个（1%）3级。68名患者的随访数据可用，其中15名（22%）患有癌症复发。与 GBG-LG 肿瘤相比，GBG-HG 肿瘤更容易复发（57% 对 6%；χ ² P <0.0001）。Silverberg 和 FIGO 分级系统在单变量分析中也与无进展生存期相关，但多变量分析显示只有 GBG 显着（风险比：10.9；Cox 比例回归P = 0.0004）。7 名患者（10%）死于疾病，所有患者均患有 GBG-HG（对数秩P<0.0001)。多变量分析显示浸润性生长百分比是预测疾病特异性生存的唯一因素（风险比：25.5，Cox P = 0.02）。将核异型添加到 GBG 类别中并没有改善预后。我们的研究验证了 GBG 系统对 OMC 中无病生存和疾病特异性生存的预后价值，在多变量分析中优于 Silverberg 和FIGO 等级。因此，GBG 应该是肿瘤分级的首选方法。