目的：评价R2-ISS分期（Revision 2 of the International Staging System）在初治多发性骨髓瘤（NDMM）患者预后分层价值。

方法：对 2013年1月至2019年12月就诊于中国医学科学院血液病医院淋巴瘤中心的505例NDMM患者回顾性分析，并分别应用R-ISS分期、R2-ISS分期对患者进行危险度分层，比较不同分期系统预后分层价值。

结果：（1）本研究共纳入505例NDMM患者，男294例，女211例，中位年龄58（29-83）岁，480例（95.0%）患者接受新药为主的诱导治疗方案，163例（32.3%）患者接受自体移植，所有患者中位无进展生存时间（PFS）41.9月，中位总生存时间（OS）71.5月。1q21+与NDMM患者不良预后密切相关，而1q21不同拷贝数异常患者预后无明显差异。多因素分析显示，del(17p)、LDH升高、1q21+、ISS II、ISS III期是PFS独立不良预后因素，而del(17p)、LDH升高、t(4;14)、1q21+、年龄＞65岁、ISS III期是OS独立不良预后因素。（2）R-ISS和R2-ISS分期均可对NDMM患者进行有效的预后分层，不同分期患者PFS和OS差异均存在统计学意义（P＜0.001）。而应用R2-ISS分期可将325例R-ISS II期患者进一步分为低中危组（82例，25.2%），中高危组（236例，72.6%），高危组（7例，2.2%），且这3组患者PFS、OS均存在显著差异（P=0.01，P=0.003）。R-ISS分期、R2-ISS分期C-index得分分别为0.699（95%CI：0.654-0.744），0.711（95%CI：0.664-0.758），证实R2-ISS分期预测效能优于R-ISS分期。（3）亚组分析显示R2-ISS分期可实现非高龄组、非自体移植组、PIs组和IMiDs组患者危险分层（P＜0.05），而在高龄组、PIs+IMiDs组中不同分期患者仅OS差异有统计学意义（P＜0.05）。（4）纳入年龄因素（＞65岁）可实现R2-ISS分期再分层，可将患者重新分为低危组（57例，11.3%）、低中危组（105例，20.8%）、中高危组*（198例，39.2%）、高危组*（145例，28.7%），各组OS差异具有统计学意义（P＜0.001）。

结论：1q21+是NDMM患者PFS和OS独立不良预后因素，包含1q21+这一危险因素的R2-ISS分期预后分层价值明显优于R-ISS分期，可较好克服R-ISS II期MM人群的预后异质性，且在不同年龄、非自体移植和不同治疗方案患者中均具有极佳的预测价值。此外，年龄因素对MM患者具有重要预后意义，可实现对R2-ISS中高危组再分层。

Objective：To evaluate the prognostic stratification value of Revision 2 of the International Staging System (R2-ISS) in newly treated multiple myeloma (NDMM) patients.

Method：The clinical data of 505 NDMM patients who were treated in our hospital from January 2013 to December 2019 were collected and analyzed retrospectively. According to the R-ISS and R2-ISS staging, 505 NDMM patients were respectively stratified. Compare the prognostic stratification value of different staging systems, and further explore the significance of age factor in the current prognostic model.

Results：(1) Our study included 505 NDMM patients, of 294 males and 211 females, with a median age of 58 (29-83) years. And 480 patients (95.0%) received novel-drug based induction treatment, then 163 patients underwent auto stem cell transplant. The median progression-free survival (PFS) for this cohort was 41.9 months, and the median overall survival time (OS) was 71.5 months. The presence of 1q21+ was related to the poor prognosis of NDMM patients, while the aberrations of 1q21 with different copy numbers didn’t show significant difference on poor outcome. The multivariate analysis showed that del(17p), elevated LDH, 1q21+, ISS II, and ISS III could exert independent poor prognostic impact on PFS. Meanwhile, del(17p), elevated LDH, t(4;14), 1q21+, age>65 years, and ISS III are independent poor predictors of OS. (2) The R-ISS staging and R2-ISS staging could effectively predict different prognoses for NDMM patients in the era of new drugs. And the survival outcomes of patients classified in different stages had statistically significant differences (P＜0.001). The application of R2-ISS staging could furtherly divide the heterogeneous R-ISS II patients into the low-intermediate risk group 82 cases (25.2%), 236 cases (72.6%) in the intermediate-high risk group, 7 cases (2.2%) in the high-risk group. And the differences in PFS and OS of the three groups of patients were statistically significant (P=0.01, P=0.003). The C-index scores of R-ISS and R2-ISS were 0.699 (95 %CI: 0.654-0.744) and 0.711 (95%CI: 0.664-0.758), respectively. Our results confirmed that the prognostic predictive power of R2-ISS is better than R-ISS. (3) R2-ISS staging has statistically significant differences in PFS and OS among patients with different risk stratifications in the non-elderly age group, no-ASCT group, PIs group, and IMiDs group (P<0.05), while only the OS difference was statistically significant in the elderly group and the PIs +IMiDs group (P<0.05). (4) With including the factor of elder age (age> 65 years), these patients could be re-classified into low-risk group (11.3%), low-intermediate risk group 105 cases (20.8%), intermediate-risk group*198 cases (39.2%), high-risk group*145 cases (28.7%), the median OS in different groups got statistical significance (P<0.001).

Conclusion：1q21+ is an independent poor prognostic factor for PFS and OS in NDMM patients. The R2-ISS staging prognostic prediction value is better than R-ISS staging, and it could achieve further stratification for R-ISS II patients. The new model could discriminate diverse prognosis for patients of different ages, non-ASCT, and different treatment options groups. In addition, the age factor has important prognostic significance for MM patients and could attain the re-stratification of R2-ISS intermediate-high risk groups.