Metformin in Hidradenitis Suppurativa: Is It Worth Pursuing Further?

Abstract

Hidradenitis suppurativa (HS) often coexists with obesity, metabolic syndrome, diabetes mellitus, or impaired glucose tolerance and insulin resistance and polycystic ovarian syndrome. Metformin is a medication used for the treatment of diabetes, acting in multiple ways. There is evidence that it decreases inflammatory cytokines, some of which are implicated in the pathogenesis of HS (TNF-α, IL-17). We performed a systematic review of data regarding the efficacy and safety of metformin for the treatment of HS. Four electronic databases (MEDLINE, ScienceDirect, Cochrane Library, and ClinicalTrials.gov), as well as the abstracts compendia of major dermatologic congresses, were searched. A total of 133 patients received metformin for HS across 6 studies, 117 of whom received it as monotherapy. The great majority of participants were female, in their thirties and overweight or obese, with one study including only children. The efficacy tools employed varied widely. Four studies (106 patients) documented improvement, 1 documented treatment failure, and 1 had mixed results. Only mild and transient side effects were noted. Metformin has been tried in few HS patients with acceptable efficacy in a fair number of them. As it is generally well tolerated and reasonably priced, carefully designed clinical trials comparing it with placebo are worth performing.

Introduction

Hidradenitis suppurativa (HS) is a long-lasting, periodically flaring cutaneous disease characterized by inflammation of the hair follicles of mostly apocrine-gland bearing skin [1, 2]. Skin lesions primarily comprise nodules, abscesses, scars, fistulae, and comedones [1]. HS can take a heavy toll on life quality and often coexists with conditions such as obesity, metabolic syndrome, diabetes mellitus, or impaired glucose tolerance and insulin resistance, as well as polycystic ovarian syndrome (PCOS) [1, 3].

Metformin is a medication (a synthetic biguanide) used for the treatment of diabetes and acts in multiple ways [4]. It impedes hepatic gluconeogenesis, augments glucose utilization in muscles, and limits intestinal glucose absorption, while also lowering insulin resistance and insulin levels as well as enabling weight loss [3, 4, 5]. What is more, it lowers lipids, exhibits anti-aging action, and lessens male hormone levels in cases of PCOS [2, 3, 4, 5]. There is also evidence that metformin decreases inflammatory cytokines, some of which are implicated in the pathogenesis of HS (TNF-α, IL-17) [1].

Metformin is administered orally and is then gradually and only partially absorbed through the (mainly small) intestine; consequently, high pharmacologically active doses are required (500 mg–2,500 mg daily, divided in up to 3 doses) [4, 5]. Its bioavailability is 50–60% and its half-life 1.5–4 h [5]. It is transported within the cells mainly through organic cation transporters and interacts with other organic cation transporter inhibitors, such as proton pump inhibitors [5]. Metformin is not metabolized, and up to 80% of an oral dose is removed through the urine and feces unchanged within 24 h [5].

Metformin can be used from 10 years of age and passes through the placenta [4]. Approximately one in five recipients develops gastrointestinal (GI) disturbance, which is transient in most patients [4]. Lactic acidosis is a rare but serious adverse outcome, which essentially affects diabetic patients with significant renal failure [5]. There has been scant evidence in literature, suggesting that metformin could benefit HS patients [1].

Material and Methods

A systematic review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. Prospective or retrospective, interventional or observational, controlled or uncontrolled studies employing any means of documenting efficacy and safety of metformin administration for the treatment of HS were considered eligible. No restrictions were imposed on language, year of publication, or age of participants. Four electronic databases (MEDLINE, ScienceDirect, Cochrane Library, and ClinicalTrials.gov) as well as the abstracts compendia of the European Hidradenitis Suppurativa Conferences, European Academy of Dermatology and Venereology Congresses, and American Academy of Dermatology Annual Meetings of the last 10 years, were searched (last search date January 10, 2023). The “Reference” section of approved studies was hand-searched for additional eligible records, and a number of experts were contacted via e-mail for any unpublished records of metformin safety and efficacy data that they would like to share with us.

The following search strategy was used for MEDLINE database and was modified accordingly for the other three databases: “hidradenitis” [Title] OR “acne inversa” [Title] AND “metformin” [MeSH Terms] OR “metformin” [All Fields] OR “metformine” [All Fields] OR “metformin s” [All Fields] OR “metformins” [All Fields]. The search, duplicate check, title screening, full text reading, study inclusion, and data extraction were performed independently by two authors (AT and ES). Any disagreements were resolved in consultation with a third author (EV). As this is a review of existing studies, the Research Ethics Committee of the Hospital of Venereal and Cutaneous Diseases of Thessaloniki, Greece, confirmed that no ethics approval is required.

Results

Six studies were identified, the basic characteristics of which are presented in Table 1. One study exclusively included pediatric patients [6]. A total of 133 patients received metformin for HS, 117 of whom received it as monotherapy. The great majority of participants were female, in their thirties, and overweight or obese. Some included patients had PCOS or diabetes, but results were not reported separately for these groups. The only adverse effects mentioned were mild gastrointestinal disturbance and mood change (the latter only in pediatric patients), which tended to be transient, but resulted in treatment cessation in 8 patients.

Table 1.

Basic characteristics of studies included in the review

Study, year, country, type	Ptsa on MTFb Sex, age	Weight, diabetes, PCOSc	Hurley stage	Monotherapy, dose	Control group	Txd duration	Follow-up duration	Tx outcome	Adverse events
Arun and Loffeld (2009) [3], UK, case report	1 (fe) 50	146 kg Type II diabetes No PCOS	NRf	Yes 500 mg odg for 3 months, then 1 g od	No	NR (>4 months)	4 months	Less frequent and shorter flares at 3 months, no leaking of previously persistent left-axilla fistula and reduced pain at 4 months	None
Verdolini et al. (2013) [2], UK, prospective	25 (22 f) 31.5 (17–49)	22 f pts were “moderately overweight, had impaired glucose tolerance and PCOS” (no diabetics)	NR	Yes 500 mg od for a week, 500 mg bdh after a week, and then 500 mg tdsi after a week	No	“Most patients continued on the treatment well past the trial”	24 weeks	SSj mean±SDk dropped from 30.1 ±9.2 to 18.4±7.2 (p < 0.0001). DLQI'sl mean±SD dropped from 13.6±4.2 to 6.3±5.5 (p < 0.000001) Lost work-days dropped from 1.5 per month to 0.4	Mild Glm disturbance at the beginning of treatment
Sanz Bueno et al. (2017) [8], NFn	11 (5 f) NF	NF	NF	Yes Doses ranging from 450 to 2,550 mg daily	No	NF	24 weeks	SS ≥30% reduction in 2 pts Glucose tolerance normalization in 5 pts DLQI improvement in 4 pts (median reduction 5.5 points) DLQI increase in 4 pts (median increase 4.2 points) 2 pts discontinued tx because of more severe flares	Mild Gl discomfort (number of pts NR)
Jennings et al. (2019) [1], Ireland, retrospective	53 (45 f) 37 (19–62)	mean weight 102 kg (67–160 kg) 4 diabetes 5 PCOS	I 4% II 72% III 24%	Yes 500 mg od and then bd after 2 weeks (further increases applicable) – mean daily dose 1.5 g/day	No	11.3 months (mean) (1–36 months)	At least 3 months	19% complete response (no active HS) – only HIIp 58% of HII and 55% of HIIIq significant partial response 68% subjective clinical improvement Non-significant CRPr reduction 25% stopped due to non-response 21% needed 2nd agent	Gl complaints (6 patients) (3 stopped tx)
Moussa et al. (2020) [6], USA, retrospective	16 peds pts NF NF	NF	I 69% II 31%	No	No	NF	NF	5 pts improved (decreased flare frequency) 5 pts did not improve 6 pts were lost to f/ut 1 pt stopped tx due to lack of improvement	Gl distress and mood changes (2, stopped tx)
Segura Palacios et al. (2021) [7], Spain, retrospective	27 (18 f) 32.1	70.4% overweight or obese NR NR	I 51.8% II 48.2%	Yes between 1.7 and 2.55 g per day	No	At least 12 weeks	24 weeks	DLQI reduction (median: 13 to 9), p = 0.001 No significant PGAu reduction	Gl complaints (3 pts stopped tx because of them)

^aPts: patients.

^bMTF: metformin.

^cPCOS: polycystic ovarian syndrome.

^dTx: treatment.

^ef:female.

^fNR: not reported.

^god: once daily.

^hbd: twice daily.

ⁱtds: three times daily.

^jSS: Sartorius score.

^kSD: Standard deviation.

^lDLQI: Dermatology life quality index.

^mGI: Gastrointestinal.

ⁿNF: Not found.

^oEADV: European Academy of Dermatology and Venereology.

^pHII: Hurley stage II.

^qHIII: Hurley stage III.

^rCRP: C-reactive protein.

^sped: pediatric.

^t:f/u: follow-up.

^uPGA: Physician Global Assessment.

The tools employed to assess efficacy were widely varied (Table 1). Three studies (79 patients in total) documented improvement in lesion-severity and quality-of-life endpoints (reaching up to complete response in some cases) in the majority of treated patients [1, 2, 3], while an additional one reported significant reduction in DLQI scores only (27 patients) [7]. One study showed no treatment success on the whole (11 patients) [8], while the children study reported improvement and non-improvement in equal number of participants (5 and 5 − a significant number of patients were lost to follow-up) [6].

Only three studies used a specific disease-severity tool (sartorius score − 2 studies − and Physician Global Assessment − 1 study) to assess metformin efficacy; statistically significant mean-score reduction, though, was achieved only in one [2]. Descriptive terms of disease activity were used in the rest of the studies, and treatment was considered successful, when inflammation was reduced (less frequent flares, less suppuration, less pain); such a response was documented in two studies [1, 3]. Patient-reported improvement, in terms of subjective clinical evaluation and DLQI questioning, was documented in the majority of patients in three studies [1, 2, 7]. One study examined the impact of metformin on work absenteeism, reporting a reduction in lost work-days, which is a recognized indirect measure of disease activity [2]. Laboratory markers were investigated in two studies: glucose tolerance was normalized in 5 out of 11 patients in one study [8], and CRP levels remained unchanged in one study [2]. Drug survival was documented in three studies [1, 6, 8]. In one study, 2 out of 11 patients discontinued treatment due to a worsening of flares; in another study, 25% of patients had to stop treatment because of no-response and 21% of them needed an additional agent; in the pediatric study, one child discontinued metformin due to lack of efficacy.

Discussion

According to this review, metformin has been tried in only few HS patients, but has shown acceptable efficacy and safety in a fair number of them. This observation needs to be appraised keeping in mind that no existing treatment modality (either drug or surgical intervention) can guarantee complete remission of HS; even if significant improvement is achieved, relapse is not rare [2, 6, 8].

Metformin seems to mostly be efficacious in overweight or obese HS patients with insulin resistance, in whom it may be indicated as adjuvant treatment or in case previous systemic treatments have failed [3, 7]. According to Jennings et al. [1], though, response to metformin is not associated with insulin resistance or body weight. This contradictory evidence is likely the result of limited existing data, which is derived from small studies of mostly poor methodological quality. Patients with PCOS seem to also be good candidates for metformin, which could be prescribed as a first-line drug, alone or in combination with other agents [2]. Higher doses than those routinely used for the treatment of diabetes seem to be warranted in HS patients [7].

As metformin is generally well tolerated and reasonably priced, carefully designed clinical trials comparing it with placebo are worth performing, in both adult and pediatric populations [1, 6, 8]. A double-blind randomized clinical trial (clinicaltrials.gov registration number: NCT04649502) examining the difference in International Hidradenitis Suppurativa Severity Score System (IHS4) in HS patients receiving metformin and doxycycline versus HS patients receiving placebo and doxycycline at 24 weeks is currently underway. Studies trying to elucidate metformin's way of action in HS patients would also be of significant help [2].

Conflict of Interest Statement

None to disclose.

Funding Sources

No funds, grants, or other support was received for conducting this study.

Author Contributions

A.T. conceived the presented idea, performed the search, extracted data, and wrote the manuscript. E.S. performed the search, extracted data, co-wrote the manuscript, and supervised the whole project. I.P. assisted with the search and edited the final manuscript. E.V. conceived the presented idea, helped reach consensus when needed, and edited the final manuscript. D.I. assisted with data curation, edited the final manuscript, and co-supervised the project.

Funding Statement

No funds, grants, or other support was received for conducting this study.