To the Editor,

In a recent issue of the Journal of Diabetes Investigation, Wang et al.¹ report that individuals with an HNF1B gene deletion are more likely to have hypomagnesemia than people with an intragenic mutation. This was based on an analysis of published data from 539 cumulative cases and the odds ratio was 3.1 (95% confidence interval 1.8–5.4). In 2016, we published data on a small series of patients with HNF1B-associated disease (n = 38) and found that the median serum magnesium level was lower in those with a 17q12 microdeletion (0.58 mmol/L, interquartile range [IQR] 0.53–0.69) compared with those with a mutation (0.7 mmol/L, IQR 0.67–0.75; P = 0.01). However, we also found that ‘renal function was worse in the mutation group with a median estimated glomerular filtration rate (GFR) of 42.6 mL/min/1.73 m² (IQR 31–60) compared with 81.4 (IQR 56–91) in the deletion group, P = 0.002’². This finding was corroborated in a much larger series of individuals with HNF1B-associated disease (n = 201) described by Dubois-Laforgue and colleagues³. They found that patients with an HNF1B mutation had a worse renal prognosis than those with a 17q12 deletion, as they had a higher frequency of chronic kidney disease stages 3–4/end-stage renal disease and a lower median eGFR at follow-up (42.5 vs 75 mL/min/1.73 m², respectively; P = 0.008). Wang et al. do not comment on renal function in the 539 analyzed cases in their paper; we suggest that the higher prevalence of hypomagnesemia in those with an HNF1B deletion is likely to be explained by a higher eGFR in this group compared with those with an intragenic mutation.

They also present a case of an individual with a spontaneous heterozygous HNF1B whole-gene deletion with dorsal pancreatic agenesis, multiple renal cysts and hypomagnesemia. They state in the abstract that ‘magnesium supplementation effectively alleviated the symptoms of diarrhea’ although this was started at the same time as digestive enzymes and other medications. His symptoms of severe diarrhea and significant weight loss are typical of pancreatic insufficiency, which we would usually confirm by fecal elastase measurement. It seems more likely that the improvement in his symptoms was due to starting treatment with digestive enzymes, which usually need to continue long-term. Oral magnesium supplementation itself can be associated with diarrhea as a side-effect and is usually ineffective in HNF1B-associated disease due to the renal magnesium wasting that is seen⁴.

致编辑，

在最近一期的《糖尿病调查杂志》中，Wang等人。¹报告称，HNF1B基因缺失的个体比基因内突变的个体更容易出现低镁血症。这是基于对 539 个累积病例的已发表数据的分析，优势比为 3.1（95% 置信区间 1.8-5.4）。2016 年，我们发表了一小批 HNF1B 相关疾病患者 ( n  = 38) 的数据，发现 17q12 微缺失患者的中位血清镁水平较低（0.58 mmol/L，四分位距 [IQR] 0.53） –0.69) 与突变者 (0.7 mmol/L, IQR 0.67–0.75; P  = 0.01) 相比。然而，我们还发现“突变组的肾功能更差，中位估计肾小球滤过率 (GFR) 为 42.6 mL/min/1.73 m ² (IQR 31–60)，而突变组的肾功能为 81.4 (IQR 56–91)”。删除组，P  = 0.002' ²。这一发现在 Dubois-Laforgue 及其同事描述的更大规模的 HNF1B 相关疾病个体 ( n  = 201)中得到了证实³。他们发现，具有HNF1B突变的患者的肾脏预后比具有 17q12 缺失的患者更差，因为他们患有慢性肾病 3-4 期/终末期肾病的频率较高，且随访时中位 eGFR 较低。分别为 42.5 与 75 mL/min/1.73 m ²；P  = 0.008)。王等人。不对他们论文中分析的 539 例病例中的肾功能发表评论；我们认为，与基因内突变的患者相比， HNF1B缺失患者中低镁血症的发生率较高可能是因为该组患者的 eGFR 较高。

他们还提出了一个病例，该病例患有自发性杂合HNF1B全基因缺失，伴有背侧胰腺发育不全、多发性肾囊肿和低镁血症。他们在摘要中指出，“补充镁有效缓解了腹泻症状”，尽管这是与消化酶和其他药物同时开始的。他的严重腹泻和体重显着减轻的症状是典型的胰腺功能不全的症状，我们通常通过粪便弹性蛋白酶测量来确认。他的症状的改善似乎更有可能是由于开始使用消化酶进行治疗，而这种治疗通常需要长期持续。口服镁补充剂本身可能会导致腹泻这种副作用，并且由于可见肾脏镁消耗，因此通常对 HNF1B 相关疾病无效⁴。