To the Editor, I read with interest the recent article published by Bajor J et al., assessing the effect of HLA‐DQB1*02 allele on several clinical (e.g. classical/atypical phenotype, type of extra‐gastrointestinal symp‐ toms and association with other autoimmune diseases) and labo‐ ratory (e.g. anti‐tTG IgA levels, histology scoring) features of their cohort of patients affected with coeliac disease (CD). (Bajor et al., 2019) Actually, in this letter I would like to highlight and comment some aspects emerging from their HLA‐DQ genotype analyses, in the light of some previous evidences and publications regarding the specific (and relevant) role played by the HLA‐DQB1*02 allele to determine the necessary immunogenetic background of predisposition to CD. (Megiorni et al., 2009; Megiorni & Pizzuti, 2012) However, to start with, it is important to remind that HLA‐ DQ2 and HLA‐DQB1*02 are not exactly synonyms, as they refer to a serotype and a genotype, respectively, although actually the HLA‐DQ2 serotype is basically determined by the presence of the β‐chain of this class II MHC molecule, which is encoded by HLA‐ DQB1*02 allelic variant, exactly. Indeed, class II MHC molecules are heterodimers and, of course, the HLA‐DQ2 molecule includes also the α‐chain, in its structure and definition. Different α chains are associated with different HLA‐DQ2 isoforms, like DQ2.5 and DQ2.2, whose α chains are encoded by the HLA‐DQA1*05 and HLA‐ DQA1*02 alleles, respectively. The same reasoning could be applied to the HLA‐DQ8 heterodimer, which is less frequent in the general population and, importantly, much less represented in CD patients. (McCluskey, Kanaan, & Diviney, 2017) Recently, we published a meta‐analysis assessing the HLA gen‐ otype in children with CD, and we found that 90%–95% of paedi‐ atric patients carry at least one copy of HLA‐DQB1*02 allele [5], which is consistent with the present study, where only 7.6% of CD patients (mostly adults) are deprived of it, according to table 1 in the article by Balor J et al. In this table, the authors reported a “B1*02 allele dose” as equivalent to “zero” even in some patients that actually carry at least one HLA‐DQB1*02 copy (see first 4 lines of “Low risk” group). (Bajor et al., 2019) Anyway, my point is that the immunogenetic predisposition in this cohort of CD patients might be established even by assessing the presence of HLA‐DQB1*02 only, in almost all cases. In the perspective of a potential widened two‐step (genetics and, eventually, serology in predisposed sub‐ jects) screening strategy for CD, this approach (searching for HLA‐ DQB1*02 only) would lead to the loss of very few cases. (Poddighe, 2018; De Silvestri et al., 2018) Moreover, if we consider the co‐ hort described by Bajor et al., six out of eight patients negative for HLA‐DQB1*02 presented with nonclassical phenotype. It would be interesting to know the Marsh score and the CD serological pa‐ rameters in these specific patients; moreover, according to table 2 (in the article by Bajor J et al.), two of these patients presented an autoimmune comorbidity (thyroiditis and alopecia areata), that may have helped the diagnosis. (Bajor et al., 2019) In conclusion, the relevant role of HLA‐DQB1*02 allele might be exploited to implement potential cost‐effective two‐step screening strategies to extend the disease research to the general population, especially in children, who might develop worse long‐term conse‐ quences from a missed diagnosis, potentially.

中文翻译：

---

HLA-DQB1*02等位基因与乳糜泻易感性的相关性

对于编辑，我很感兴趣地阅读了 Bajor J 等人最近发表的文章，评估 HLA-DQB1*02 等位基因对几种临床（例如经典/非典型表型、胃肠外症状的类型以及与其他自身免疫性疾病）和实验室（例如，抗 tTG IgA 水平、组织学评分）特征的乳糜泻 (CD) 患者队列。(Bajor et al., 2019) 实际上，在这封信中，我想强调和评论他们的 HLA-DQ 基因型分析中出现的一些方面，鉴于之前关于特定（和相关）作用的一些证据和出版物HLA-DQB1*02 等位基因以确定 CD 易感性的必要免疫遗传背景。(Megiorni et al., 2009; Megiorni & Pizzuti, 2012) 然而，首先，重要的是要提醒的是 HLA-DQ2 和 HLA-DQB1*02 并不是完全同义词，因为它们分别指的是血清型和基因型，尽管实际上 HLA-DQ2 血清型基本上是由 β 链的存在决定的这种由 HLA-DQB1*02 等位基因变体编码的 II 类 MHC 分子。事实上，II 类 MHC 分子是异源二聚体，当然，HLA-DQ2 分子在其结构和定义中也包括 α 链。不同的 α 链与不同的 HLA-DQ2 亚型相关，如 DQ2.5 和 DQ2.2，其 α 链分别由 HLA-DQA1*05 和 HLA-DQA1*02 等位基因编码。同样的推理也适用于 HLA-DQ8 异源二聚体，它在一般人群中较少见，重要的是，在 CD 患者中很少出现。（麦克拉斯基、卡南和迪维尼，2017) 最近，我们发表了一项评估 CD 儿童 HLA 基因型的荟萃分析，我们发现 90%–95% 的儿科患者携带至少一个 HLA-DQB1*02 等位基因 [5] ，这与本研究一致，根据 Balor J 等人文章中的表 1，只有 7.6% 的 CD 患者（主要是成年人）被剥夺了它。在该表中，作者报告了“B1*02 等位基因剂量”相当于“零”，即使在一些实际上携带至少一个 HLA-DQB1*02 拷贝的患者中（参见“低风险”组的前 4 行）。（Bajor 等人，2019 年）无论如何，我的观点是，即使在几乎所有情况下，仅通过评估 HLA-DQB1*02 的存在，也可以确定这组 CD 患者的免疫遗传易感性。从潜在扩大的两步（遗传学和最终，易感受试者的血清学）CD 筛查策略，这种方法（仅搜索 HLA-DQB1*02）将导致极少数病例的丢失。（Poddighe，2018 年；De Silvestri 等人，2018 年）此外，如果我们考虑 Bajor 等人描述的队列，八分之六的 HLA-DQB1*02 阴性患者呈现非经典表型。了解这些特定患者的 Marsh 评分和 CD 血清学参数会很有趣；此外，根据表 2（在 Bajor J 等人的文章中），其中两名患者出现自身免疫性合并症（甲状腺炎和斑秃），这可能有助于诊断。(Bajor et al., 2019) 总之，