Department of Pathology, Peking University First Hospital, Beijing, China.
Department of Pathology, Peking University Health Science Center, Beijing, China.
Cancer Med. 2019 Jul;8(8):3831-3845. doi: 10.1002/cam4.2316. Epub 2019 May 31.
The expression of programmed cell death ligand 1 (PD-L1) is a biomarker for immunotherapy, but approved detection method is absent in diffuse large B-cell lymphoma (DLBCL). Here, we performed three methods including immunohistochemistry (IHC) (clone SP263 and SP142), RNAscope, and fluorescence in situ hybridization (FISH) to evaluate PD-L1 status on a cohort of DLBCL including 94 of DLBCL-NOS, 25 of primary mediastinal large B-cell lymphoma (PMBCL) and 7 of double-hit lymphoma (DHL). SP263 with 25% for immune cell (IC) or combined cell and SP142 with 10% for tumor cell (TC), 20% for both of IC and combined cell were proved to have corresponding survival prognostic. Combined showed comparable prognostic value with TC and IC . SP263 and SP142 showed strong concordance (k = 0.788) with combined rates of 33.3% (42/126) and 34.9% (44/126), respectively. In DLBCL-NOS, TC by SP263 preferred to non-GCB and immunoblastic variant DLBCL-NOS (P = 0.029 and P = 0.004). Combined (SP263 and SP142) were associated with more than one extranodal site involved (P = 0.006, P = 0.042), higher ECOG PS scores (P = 0.001, P < 0.001), high IPI risk (P = 0.012, P = 0.005), and poor treatment response (P = 0.095, P = 0.002). IC by SP263 and SP142 were both independent risk factors (P = 0.027, P = 0.037). 9p24.1 locus amplification and gain were identified in 4.3% and 7.6% DLBCL-NOS and indicated shorter overall survival (P = 0.004). Positive rate of PD-L1 by RNAscope was 36.5%, while no clinical significance shown. PD-L1 positive rates were all higher in PMBCL and DHL than in DLBCL-NOS by SP263, SP142, RNAscope, and FISH (P = 0.001, P < 0.001, P = 0.005 and P < 0.001, respectively). In conclusion, combined PD-L1 expression by IHC was potentially reliable and convenient as a predicting biomarker. SP263 staining was easier to evaluate and recognized more PD-L1-stained cells, but SP142 presented a better prognostic indicator. FISH and RNAscope could be used as supplementary assays. PMBCL itself was a sensitive cohort for immunotherapy.
程序性细胞死亡配体 1（PD-L1）的表达是免疫治疗的生物标志物，但在弥漫性大 B 细胞淋巴瘤（DLBCL）中缺乏批准的检测方法。在这里，我们使用三种方法，包括免疫组织化学（IHC）（克隆 SP263 和 SP142）、RNAscope 和荧光原位杂交（FISH），对包括 94 例弥漫性大 B 细胞淋巴瘤非特指型（DLBCL-NOS）、25 例原发性纵隔大 B 细胞淋巴瘤（PMBCL）和 7 例双打击淋巴瘤（DHL）在内的 DLBCL 队列进行 PD-L1 状态评估。SP263 的免疫细胞（IC）或联合细胞的 25%，SP142 的肿瘤细胞（TC）的 10%，IC 和联合细胞的 20%被证明具有相应的生存预后。联合显示与 TC 和 IC 具有可比的预后价值。SP263 和 SP142 与联合率为 33.3%（42/126）和 34.9%（44/126）的联合显示出较强的一致性（k=0.788）。在 DLBCL-NOS 中，TC 由 SP263 染色更倾向于非生发中心 B 细胞和免疫母细胞型 DLBCL-NOS（P=0.029 和 P=0.004）。联合（SP263 和 SP142）与超过一个结外部位受累相关（P=0.006，P=0.042），ECOG PS 评分较高（P=0.001，P<0.001），国际预后指数（IPI）风险较高（P=0.012，P=0.005），治疗反应较差（P=0.095，P=0.002）。SP263 和 SP142 的 IC 都是独立的危险因素（P=0.027，P=0.037）。9p24.1 基因座扩增和获得在 4.3%和 7.6%的 DLBCL-NOS 中被识别，并且提示总生存期更短（P=0.004）。RNAscope 检测到的 PD-L1 阳性率为 36.5%，但无临床意义。PMBCL 和 DHL 中 SP263、SP142、RNAscope 和 FISH 检测到的 PD-L1 阳性率均高于 DLBCL-NOS（P=0.001，P<0.001，P=0.005 和 P<0.001）。总之，IHC 联合 PD-L1 表达可能是一种可靠且方便的预测生物标志物。SP263 染色更容易评估，并且识别出更多的 PD-L1 染色细胞，但 SP142 呈现出更好的预后指标。FISH 和 RNAscope 可作为补充检测方法。PMBCL 本身是一个对免疫治疗敏感的队列。
