To the Editor:

We read with great interest the manuscript by Halloran and the INTERCOMEX investigators recently published in AJT on the molecular phenotype of antibody-mediated kidney transplant rejection (ABMR).¹

We described earlier that the intrarenal molecular signatures of ABMR histology did not differ between patients with and without HLA-DSA.² The INTERCOMEX data now confirm that also molecularly defined ABMR (mABMR) did not differ between patients with and without HLA-DSA. It is reassuring that the molecular approach by Halloran et al.¹ agrees with our previous results based on light microscopic classification of microvascular inflammation.

However, we do not share the conclusion of Halloran et al. that the lack of major differences in gene expression profiles between HLA-DSA-positive and -negative cases suggests that all cases with microvascular inflammation and an increased expression of above transcripts can be diagnosed as ABMR, and hence the management should not differ between HLA-DSA-positive and -negative cases.

Neither our study² nor Halloran's¹ can make claims on underlying causality. With these analyses, top genes and pathways are essentially uncovered by statistical comparison of samples with versus without inflammation, thus assessing primarily the cellular composition of the kidneys.

The mechanism behind this immune cell infiltration and activation, and related tissue/endothelial injury, is not necessarily reflected by their gene expression changes. For instance, the observation that Fc receptors are upregulated in HLA-DSA-negative mABMR is not proof that these receptors contributed to the infiltration and activation of the cells. It is possible that this merely reflects the presence of (Fc receptor-expressing) cells, potentially activated by totally different mechanisms. Hence the transcripts of a molecular phenotype are as specific/non-specific as the lesions of a histologic phenotype.

In the absence of data on potential causes or specific risk factors for the HLA-DSA-negative mABMR, it is premature to conclude that this phenotype is explained by antibodies, either missed HLA-DSA (below the levels of current detection methods) or non-HLA antibodies. Recent studies demonstrated the activation of NK cells by antibody- and Fc-receptor independent mechanisms. For instance, the lack of inhibitory signals dependent upon KIR/HLA-I interactions (“missing self”) is sufficient to trigger microvascular inflammation, with a cellular composition like that observed in ABMR.³ It is also conceivable that other innate allorecognition mechanisms, including myeloid- and monocyte-driven allorecognition,⁴ or even primary T-cell activation by mismatched HLA molecules⁵ could lead to similar histological/molecular pictures.

We agree with Halloran et al. that it is time to consider including well-defined HLA-DSA-negative microvascular rejection in clinical trials but urge that this phenotype be kept delineated from its HLA-DSA-positive counterpart. Lumping all mABMR cases together bears the risk of underestimating the heterogeneity of the phenotype and reminds us of earlier discussions on the abandoned chronic allograft nephropathy (CAN) concept, just at the molecular level instead of at the histological level. In the era of precision diagnostics, disease- and pathogenesis-specific approaches are needed. Therefore, an international consensus definition of HLA-DSA-negative microvascular rejection seems necessary to study the mechanisms operating more systematically, and to find better, targeted therapies.

致编辑：

我们怀着极大的兴趣阅读了 Halloran 和 INTERCOMEX 研究人员最近在AJT上发表的关于抗体介导的肾移植排斥反应 (ABMR) 分子表型的手稿。^1个

我们之前描述过，ABMR 组织学的肾内分子特征在有和没有 HLA-DSA 的患者之间没有差异。² INTERCOMEX 数据现在证实，分子定义的 ABMR (mABMR) 在有和没有 HLA-DSA 的患者之间也没有差异。令人欣慰的是，Halloran 等人的分子方法。^图1与我们之前基于微血管炎症的光学显微镜分类的结果一致。

然而，我们不同意 Halloran 等人的结论。HLA-DSA 阳性和阴性病例之间的基因表达谱没有显着差异，这表明所有具有微血管炎症和上述转录物表达增加的病例都可以诊断为 ABMR，因此 HLA-DSA 之间的管理不应有差异DSA 阳性和阴性病例。

我们的研究²和 Halloran 的研究¹都不能断言潜在的因果关系。通过这些分析，主要基因和通路基本上可以通过对有炎症和没有炎症的样本进行统计比较来揭示，从而主要评估肾脏的细胞组成。

这种免疫细胞浸润和激活以及相关组织/内皮损伤背后的机制不一定反映在它们的基因表达变化上。例如，观察到 Fc 受体在 HLA-DSA 阴性 mABMR 中上调并不能证明这些受体有助于细胞的浸润和活化。这可能仅反映了（Fc 受体表达）细胞的存在，可能由完全不同的机制激活。因此，分子表型的转录物与组织学表型的病变一样具有特异性/非特异性。

由于缺乏关于 HLA-DSA 阴性 mABMR 的潜在原因或特定风险因素的数据，现在断定这种表型是由抗体解释的还为时过早，无论是漏检 HLA-DSA（低于当前检测方法的水平）还是非-HLA 抗体。最近的研究证明了通过抗体和 Fc 受体独立机制激活 NK 细胞。例如，依赖于 KIR/HLA-I 相互作用的抑制信号的缺失（“缺失自身”）足以引发微血管炎症，其细胞组成与 ABMR 中观察到的相似。³还可以想象，其他先天性同种异体识别机制，包括骨髓和单核细胞驱动的同种异体识别，⁴或什至由不匹配的 HLA 分子激活的初级 T 细胞⁵可能导致类似的组织学/分子图片。

我们同意 Halloran 等人的观点。现在是时候考虑在临床试验中包括定义明确的 HLA-DSA 阴性微血管排斥反应，但敦促将这种表型与其 HLA-DSA 阳性对应物区分开来。将所有 mABMR 病例集中在一起存在低估表型异质性的风险，并提醒我们早期关于废弃的慢性同种异体移植肾病 (CAN) 概念的讨论，只是在分子水平而不是组织学水平。在精确诊断的时代，需要特定于疾病和发病机制的方法。因此，HLA-DSA 阴性微血管排斥的国际共识定义似乎有必要研究更系统地运作的机制，并找到更好的靶向治疗。