Fournier's gangrene and sodium–glucose cotransporter 2 inhibitors: Is there a causal association?
Year: 2019
Type: article
Abstract: …cases of a rare but serious infection of the genitals and area around the genitals have been reported with the class of type 2 diabetes medicines called sodium–glucose cotransporter-2 inhibitors (SGLT2i).2 This warning was based on the FDA having identified 12 people with Fournier's gangrene (FG) who had been treated with an SGLT2i between 2013 and 2018. In 2017 alone there were 1.7 million patients treated with an SGLT2i.2 If this is the case, the number of people with FG identified by the FDA appears to be orders of magnitude smaller than the incidence of FG in the general population and the population with diabetes (see below). Our analysis of the literature suggests that the FDA's data and/or analysis is flawed. The epidemiology of FG can be estimated from several published studies. Using the National Inpatient Sample, a US database representing 20% of all hospitalized people between 2004 and 2012, Furr et al.3 identified 9249 cases of FG requiring surgical debridement, leading the authors to estimate that 43 146 people were hospitalized with FG during this period, among more than 350 million hospitalizations,4 suggesting an incidence around 12 per 100 000 hospitalizations. In addition, 52% of those with FG had diabetes.3 In another US analysis, Sorensen et al.5 analyzed the State Inpatient Databases, a hospital administrative dataset of >10 000 000 admissions, using the International Classifications of Diseases (ICD) 9 code identifying FG as a disease of the male genital organs. In that study, the authors found 1641 males with FG, giving a lower incidence of 1.6/100 000 male hospitalizations.5 The respective likelihoods of diabetes and obesity were 3.3- and 3.7-fold greater than those for men not having FG. Furthermore, 66% of hospitals in the database reported no cases and, although fewer than 18% of hospitals in the database were teaching hospitals, 57% of the cases were managed at these hospitals,6 suggesting that such hospitals were either more likely to have a higher FG case incidence or were better at making the diagnosis. Two analyses used the National Surgical Quality Improvement Program database, comprised of fewer hospitalizations.7 Mills et al.8 found a diabetes prevalence of 47% among 688 necrotizing soft tissue infections, compared with 15% of controls, with 64% of the patients treated with insulin prior to development of the infection. In the other study, Kim et al.9 identified 650 people with peroneal or genital necrotizing fasciitis in a smaller 400-hospital dataset, 57% men and 43% women. There are several single-center reports. Sarkut et al.10 found 32 women among 120 people with FG, with 66% of women and 44% of men with FG having diabetes. Wróblewska et al.11 reported that more than 40% of people with FG had diabetes, and noted a relationship with incompletely treated urogenital or anorectal diseases, often with poor hygiene or local trauma. Aridogan et al.12 described 71 patients with FG, 50% with diabetes, and Czymek et al.13 reported on 38 people with FG, 12 women (five with diabetes) and 26 men (eight with diabetes). Cakmak et al.14 reported on 45 men (19 with diabetes) and 20 women (10 with diabetes) treated for FG in their institution over a 22-year period. Similarly, Yanar et al.15 reported on 25 men and 10 women with FG, 46% of whom had diabetes. Given these data, is SGLT2i use actually associated with FG? Although SGLT2i are associated with genital fungal infections in both women and men,16, 17 none of the reviewed studies found an association between such superficial infections and FG, which, rather, was associated with procedures such as hemorrhoidectomy and with deep space infections, such as perianal14 or vulvar abscess,13 so a mechanism whereby SGLT2i may cause FG is not apparent. Certainly diabetes is present in many people with FG. With increasing use of SGLT2i for cardiovascular and renal indications, as well as for glucose lowering, the drugs will certainly be given to people who will develop FG. However, the epidemiologic analysis suggests that the FDA warning that SGLT2i use increases the incidence of FG may be inaccurate. The FDA determined that "only six cases of FG [all in men] were identified in review of other antidiabetic drug classes over a period of more than 30 years".2 The FDA warning also stated that "although most cases of FG have previously been reported in men, our 12 cases included 7 men and 5 women".2 In our analysis, some 40–50% of people with FG have diabetes, and many thousands are referenced in the above studies, so that the FDA assertion of just six people with diabetes being treated with agents other than SGLT2i represents a peculiar underestimation. Clearly, FG occurs in both women and men, further suggesting the inaccuracy of the FDA analysis. The existence of a causal relationship between the use of SGLT2i and FG seems, at present, to be unproven. In the report of the Dapagliflozin Effect on Cardiovascular Events (DECLARE) study, the authors state, "Six cases of Fournier's gangrene were reported, one [of 8574] in the dapagliflozin group and five [of 8569] in the placebo group".18 在本期杂志中收录了一篇关于1名使用达格列净治疗的糖尿病患者在阴囊与腹股沟区域出现坏死性筋膜炎的病例报告1。2018年8月29日,美国食品与药品监督管理局(FDA)发布了一项警告,指出: …有研究报道一种2型糖尿病治疗药物:钠-葡萄糖共转运体-2抑制剂(SGLT2i)可导致罕见却很严重的生殖器以及生殖器周围区域感染2。 FDA确认在2013年至2018年期间有12名使用某种SGLT2i治疗的患者出现了Fournier's坏疽(Fournier's gangrene,FG),因而提出了这一警告。仅仅在2017年就有170万名患者在使用一种SGLT2i治疗2。如果是这样的情况,FDA确认患FG的人数与普通人群以及糖尿病人群中的FG发病率相比似乎要小几个数量级(参见下文)。我们对文献进行分析后发现,FDA的数据和/或分析存在缺陷。 我们可以从几项已经发表的研究中评估FG的流行病学。Furr等3使用全国住院患者样本(National Inpatient Sample),其样本量代表了2004-2012年间美国所有住院患者的20%,从中确认有9249例需要手术清创的FG患者,作者因此估计在这段时间内有43146名FG住院患者,这意味着在超过3.5亿人次的住院患者中4每10万名住院患者的发病率大约为12名。另外,在FG患者中有52%为糖尿病患者3。在另一项美国的分析中,Sorensen等5对州住院患者数据库(State Inpatient Databases),一个住院人数> 1千万人次的医院管理数据集,进行了分析,使用国际疾病分类编码 (ICD)-9将FG分类为男性生殖器官疾病。在该项研究中,作者发现有1641名男性FG患者,由此得出的FG发病率更低,每10万名男性住院患者中只有1.6名5。与没有FG的男性患者相比,男性FG患者合并糖尿病与肥胖的可能性分别要高出3.3倍与3.7倍。此外,在数据库中有66%的医院没有报告此病例,虽然在数据库中只有不到18%的医院是教学医院,但是57%的病例都是在这些医院中治疗的6,这意味着在这些医院中FG病例的发生率可能更高,或者是这些医院可能更擅长诊断FG。 有两项分析使用了全国外科质量改善计划(National Surgical Quality Improvement Program)数据库,纳入的住院人数更少7。Mills等8发现在688例坏死性软组织感染患者中糖尿病的患病率为47%,而在对照组中为15%,并且64%的患者在发生感染之前正在使用胰岛素治疗。在另一个研究中,Kim等9在一个规模较小的400家医院的数据集中确认有650名患者合并腓侧或者生殖器坏死性筋膜炎,57%为男性,43%为女性。 还有几个单中心的报告。Sarkut等10发现在120名FG患者中有32名为女性,并且66%的女性FG患者合并糖尿病,而男性FG患者合并糖尿病的比例为44%。据Wróblewska等11报告,FG患者合并糖尿病的比例超过了40%,并且指出了与未完全治疗的泌尿生殖系统疾病或者肛肠疾病之间的关系,通常患者的卫生条件差或有局部创伤。Aridogan等12描述了71名FG患者,发现50%合并糖尿病,Czymek等13报告了38名FG患者,其中女性12名(5名合并糖尿病),男性26名(8名合并糖尿病)。Cakmak等14报告了在过去超过22年的时间里在他们的机构中接受治疗的FG患者,共有45名男性(19名合并糖尿病)与20名女性(10名合并糖尿病)。与此相似,Yanar等15报告了25名男性与10名女性FG患者,他们中有46%合并有糖尿病。 从这些数据来看,使用SGLT2i治疗真的与FG有关?虽然在男性与女性使用SGLT2i治疗与生殖器真菌感染都有相关性16,17,但是没有一项回顾性研究发现这类浅表性感染与FG之间具有相关性,相反FG却与手术(例如痔疮切除术)后深部空间感染有关,例如肛周脓肿14或外阴脓肿13,因此SGLT2i可能引起FG的机制并不明确。当然,很多FG患者都有糖尿病。因为SGLT2i具有心血管与肾脏适应症,除此之外还有降糖作用,所以目前使用的人越来越多,这类药物当然也适用于有可能会发生FG的患者。然而,流行病学分析表明FDA所提出的警告, 认为使用SGLT2i会增加FG的发生率,可能并不准确。FDA在"回顾分析了三十多年以来其他类别的降糖药物,只发现了6例FG患者(全部是男性)"2。FDA的警告还指出了"虽然既往报告的FG病例大部分都是男性,但是我们的12名病例包括7名男性与5名女性"2。而在我们的分析中,大约40%-50%的FG患者都合并糖尿病,并且在上述的研究中纳入了数千名的患者,因此,FDA声称在接受除了SGLT2i以外药物治疗的糖尿病患者中只有6名出现了FG,这意味着FG发病率被异常地低估了。显然,男性与女性都会发生FG,这进一步表明FDA的分析并不正确。使用SGLT2i与FG之间是否存在因果关系目前似乎还尚未明确。在达格列净对心血管事件影响(Dapagliflozin Effect on Cardiovascular Events,DECLARE)的研究报告中,作者提到,"在报告的6例Fournier's坏疽患者中,有1名出现在达格列净组(总共8574名受试者),有5名出现在安慰剂组(总共8569名受试者)"18。
Source: Journal of Diabetes
Language: English
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W2907326134