Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, MD, USA.
Infixion Bioscience, Inc., San Diego, CA, USA.
Acta Neuropathol Commun. 2024 Oct 29;12(1):172. doi: 10.1186/s40478-024-01875-z.
NF1 inactivation is associated with sensitivity to MEK inhibitor targeted therapy in low-grade and some high-grade gliomas. NF1 loss may also be a harbinger of exploitable vulnerabilities in IDH-wildtype glioblastoma (GBM). Accurate and consistent detection of NF1 loss, however, is fraught given the large gene size, challenges with complete coverage and variant calling upon sequencing, and mechanisms of mRNA and protein regulation that result in early degradation in the absence of genomic alterations. Here, we seek to perform a composite analysis for NF1 loss accounting for genomic alterations and protein expression via immunohistochemistry. We also characterize the landscape of NF1 alterations in GBM.
We assembled a single-institution, retrospective cohort of 542 IDH-wildtype GBM with somatic next generation sequencing to investigate the frequency and nature of detected NF1 alterations. We selected 69 GBMs from which to build a tissue microarray (TMA) of 44 NF1-wildtype and 25 NF1-mutant cases. We performed NF1 immunohistochemistry using two different NF1 antibodies (NFC, Sigma-Aldrich; and iNF-07E, iNFixion Bioscience) and correlated results with clinical, genomic, and other immunohistochemical features.
In our retrospective cohort, we identified 88 IDH-wildtype GBM with NF1 alterations (16%). NF1 alterations were mutually exclusive with EGFR and MDM2 alterations (p-adj < 0.001, 0.05, respectively), but co-occurred with PIK3R1 alterations (Log(OR) = - 1.6, p-adj = 0.03). Of the 63 scorable sporadic GBMs in the TMA, 14 harbored NF1 inactivating alterations and of those, 12 (86%) demonstrated minimal NF1 immunoreactivity by NFC antibody, compared to 8 (57%) by iNF-07E antibody. Among the 42 scorable NF1-wildtype GBM in the TMA, NF1 immunostaining was minimal in 18 (43%) by NFC antibody compared to 4 (10%) by iNF-07E antibody, potentially reflecting false positives or differential protein regulation. Minimal immunoreactivity by NFC antibody was associated with decreased median overall survival (8.5 vs. 16.4 months, p = 0.011). Cox proportional hazards model correcting for prognostic variables in this subset revealed HR 3.23 (95% CI 1.29-8.06, p = 0.01) associated with decreased NF1 expression by IHC.
NF1 immunostaining may serve as a sensitive surrogate marker of NF1 genomic inactivation and a valuable extension to next-generation sequencing for defining NF1 status. Minimal NF1 immunoreactivity is a poor prognostic marker, even in IDH-wildtype glioblastoma without apparent NF1 genomic alterations, but the underlying molecular mechanism requires further investigation.
NF1 失活与 MEK 抑制剂靶向治疗在低级别和一些高级别神经胶质瘤中的敏感性有关。NF1 缺失也可能是 IDH 野生型胶质母细胞瘤 (GBM) 中可利用脆弱性的先兆。然而，由于基因较大，测序时完全覆盖和变异调用存在挑战，以及导致在没有基因组改变的情况下早期降解的 mRNA 和蛋白质调节机制，准确且一致地检测 NF1 缺失是困难的。在这里，我们试图通过免疫组织化学对 NF1 缺失进行基因组改变和蛋白质表达的综合分析。我们还描述了 GBM 中 NF1 改变的情况。
我们组建了一个单机构回顾性队列，其中包括 542 例 IDH 野生型 GBM，进行体细胞下一代测序，以研究检测到的 NF1 改变的频率和性质。我们从其中选择了 69 例 GBM 来构建一个包含 44 例 NF1 野生型和 25 例 NF1 突变型病例的组织微阵列 (TMA)。我们使用两种不同的 NF1 抗体 (NFC，Sigma-Aldrich；和 iNF-07E，iNFixion Bioscience) 进行 NF1 免疫组织化学检测，并将结果与临床、基因组和其他免疫组织化学特征相关联。
在我们的回顾性队列中，我们确定了 88 例 IDH 野生型 GBM 存在 NF1 改变 (16%)。NF1 改变与 EGFR 和 MDM2 改变相互排斥 (p-adj < 0.001，0.05)，但与 PIK3R1 改变同时发生 (Log(OR)=-1.6，p-adj=0.03)。在 TMA 中可评分的 63 例散发性 GBM 中，有 14 例存在 NF1 失活改变，其中 12 例 (86%)用 NFC 抗体显示最小的 NF1 免疫反应性，而用 iNF-07E 抗体则为 8 例 (57%)。在 TMA 中可评分的 42 例 NF1 野生型 GBM 中，18 例 (43%)用 NFC 抗体显示最小的 NF1 免疫染色，而用 iNF-07E 抗体则为 4 例 (10%)，这可能反映了假阳性或差异蛋白调节。NFC 抗体的最小免疫反应性与中位总生存期缩短相关 (8.5 与 16.4 个月，p=0.011)。在这个亚组中，校正预后变量的 Cox 比例风险模型显示，与 NF1 免疫组化表达降低相关的 HR 为 3.23 (95%CI 1.29-8.06，p=0.01)。
NF1 免疫染色可能是 NF1 基因组失活的敏感替代标志物，也是对下一代测序进行扩展以定义 NF1 状态的有价值方法。最小的 NF1 免疫反应性是一个不良的预后标志物，即使在没有明显 NF1 基因组改变的 IDH 野生型 GBM 中也是如此，但潜在的分子机制仍需要进一步研究。
