Pharmacovigilance-based drug repurposing: searching for putative drugs with hypohidrosis or anhidrosis adverse events for use against hyperhidrosis

Drug repurposing is the process of identifying and developing novel disease indications for approved drugs [1]. The advantage of drug repurposing over conventional drug development is that it can reduce the time, cost, and risk of failure associated with new drug development [2]. Other than approaches that are primarily focused on omics- or ligand docking, adverse event (AE)-based drug repurposing is a novel strategy [3]. Each drug interacts with multiple pharmacodynamic targets, producing the expected therapeutic activity and undesired effects, which may be either beneficial or harmful AEs. These multitarget profiles explain how one drug might have therapeutic relevance in several diseases, especially when these diseases share common genes, biomarkers, signaling pathways, or environmental factors [4]. Thus, drug screening based on AEs could reveal unknown pharmacodynamic properties.

Pharmacovigilance data consist of spontaneously reported AEs, and several databases are available, such as the US Food and Drug Administration (FDA) Adverse Event Reporting System (FAERS), European Medicines Agency (EMA) EudraVigilance and World Health Organization (WHO) VigiBase [5]. Disproportionate analysis, a drug–event combination that occurs more frequently than expected, helps to estimate whether these AEs occurred randomly or were caused by the drug and can, therefore, be considered a true adverse drug reaction [6]. Pharmacovigilance data have been employed to generate hypotheses for new therapeutic applications of existing drugs [7‐11]. For instance, it was used to detect which drugs were associated with significantly overreported hypotension and assess the opportunity to use them as antihypertensive agents based on the FAERS database [10]. Zamami Y et al. identified prophylactic drugs for oxaliplatin-induced peripheral neuropathy by drug repositioning using data from the FAERS database [11].

Hyperhidrosis is an idiopathic disorder characterized by excessive sweating not related to heat or physical exercise, which may have a significant negative impact on a person’s quality of life, causing both physical problems and psychological distress [12]. The prevalence rate of hyperhidrosis in the general population in the US was reported to be 4.8% [13]. Further estimates of its prevalence have varied widely, ranging from 4.6% in Germany to 5.5% in Sweden; 12.3% in Vancouver, Canada; and 14.5% in Shanghai, China [14‐16]. A variety of nonsurgical and surgical treatments currently exist to manage hyperhidrosis. Nonsurgical treatments include topical therapies, oral anticholinergic medications, iontophoresis, and injectable medications. Aluminum chloride is the most commonly used and effective topical medication, but its satisfaction rate is only 60% [17]. Anticholinergic or astringent medications are topically used as well. However, the common disadvantage of all of these topical therapies is inadequate long-term efficacy [18]. Oral anticholinergic medications, such as glycopyrrolate or oxybutynin, have been used off-label for hyperhidrosis, with a response rate of approximately 70%, but one-third of patients discontinue treatment due to adverse effects [19]. Iontophoresis is another effective treatment for hyperhidrosis, with an effective rate of approximately 80%, but maintenance therapy must be given every 1–4 weeks [20]. Botulinum toxin is an injectable medication used for hyperhidrosis, but adverse effects, including pain and haematoma caused by injection, limit its usage. It is usually highly effective within the first 7–10 days after therapy, but only lasts for 3–9 months or longer and requires repeated injections to maintain its effects [21]. Surgical interventions, such as endoscopic thoracic sympathectomy, are usually reserved for severe cases in which other less invasive and more conservative measures have failed [17]. However, persistent side effects such as compensatory sweating and Horner’s syndrome may result from the surgery. To date, there are no available drugs or methods with stable and durable effects. Compared with other skin disorders of similar prevalence, hyperhidrosis is under researched [22]. Assuming that hypohidrosis or anhidrosis, with decreased sweating or complete absence of sweating, can be regarded as the opposite conditions of hyperhidrosis, potential anti-hyperhidrosis agents may be discovered among drugs that induce hypohidrosis or anhidrosis as side effects.

To the best of our knowledge, this drug discovery approach has not been applied to anti-hyperhidrosis drug repurposing. Therefore, the objectives of the present study were to describe the hypohidrosis or anhidrosis cases submitted to the FAERS database, evaluate the signals between drugs and hypohidrosis or anhidrosis using disproportionality analyses, and examine the onset time of hypohidrosis or anhidrosis events after drug exposure. We hoped this study would provide unique insight into pharmacovigilance-based anti-hyperhidrosis drug repurposing and furnish possible treatment options for hyperhidrosis through indication expansion.

To the best of our knowledge, this is the first study using AE-based drug repurposing strategies to identify potential anti-hyperhidrosis agents. Based on pharmacovigilance analysis of real-world adverse events reported to FAERS, some putative drugs were found to have statistically significant signals with hypohidrosis or anhidrosis, which may be redirected for diminishing sweat production.

Our study first systematically assessed hypohidrosis or anhidrosis reports based on the FAERS database and is the largest summary of drug-associated hypohidrosis or anhidrosis cases to date. In our study, 540 reports of 192 drugs with suspected drug-associated hypohidrosis or anhidrosis from the FAERS database were assessed. Among them, 213 reports were submitted by healthcare professionals, including physicians, pharmacists and other health professionals. Detailed information on the clinical characteristics was summarized. Furthermore, 39 drugs were identified with hypohidrosis or anhidrosis signals according to disproportionality analysis. Nervous system drugs were most frequently reported, including antiepileptics, antidepressants and antipsychotics. The results were consistent with other previous studies [26]. A review demonstrated that the drugs with a higher correlation with hypohidrosis or anhidrosis were anticholinergics, tricyclic antidepressants, and antiepileptics [26], although evidence mainly came from case reports or a small sample population in a specific area [27‐33].

To avoid false signals, 18 drugs were further chosen given  ≥ 5 reported cases [34]. In addition, the pharmacological characteristics of the drugs were also considered [35]. Carboplatin, which might be clinically infeasible, was removed from the candidate drug list. In addition to detecting hypohidrosis or anhidrosis signals, the time to onset of drug-induced hypohidrosis or anhidrosis was also evaluated, which would possibly be the time to onset of anti-hyperhidrosis therapy. Our results indicated that the median onset time of botulinum toxin type A, glycopyrronium, oxybutynin and solifenacin-associated hypohidrosis or anhidrosis was 0–6 days, whereas that of lamotrigine, agalsidase beta, topiramate, finasteride, ciprofloxacin, zonisamide, duloxetine, citalopram, risperidone and aripiprazole was 34–1210 days. If the interval between adverse reactions of hypohidrosis or anhidrosis after drug exposure is too long, it may indicate that the drug is inappropriate for the treatment of hyperhidrosis.

Considering disproportionality signals, pharmacological characteristics of drugs and appropriate onset time, our study suggests that the main putative drugs for anti-hyperhidrosis are glycopyrronium, solifenacin, oxybutynin, and botulinum toxin type A. Other drugs, such as topiramate, zonisamide, agalsidase beta, finasteride, metformin, lamotrigine, citalopram, ciprofloxacin, bupropion, duloxetine, aripiprazole, prednisolone, and risperidone need further investigation. This is in line with studies that used VigiBase to investigate drugs associated with hypohidrosis or anhidrosis and revealed that the most at-risk drugs are oxybutynin, topiramate, and zonisamide [36].

The spectrum of drugs highlights the various mechanisms of drug-related hypohidrosis or anhidrosis. Undoubtedly, understanding the components of the human thermoregulatory system can help classify the effects of drugs on sweating [37]. The sweating pathway begins in the medial preoptic area of the hypothalamus and extends to the intermediolateral column, from which preganglionic sympathetic fibers arise that leave the spinal cord and synapse within the sympathetic chain ganglia. Postganglionic sympathetic nerves emerge from these ganglia and travel alongside arteries to reach their subdermal targets, the eccrine sweat glands. Along this pathway, the most clinically important neuroendocrine mediator is acetylcholine, which binds to cholinergic muscarinic receptors in the basolateral membrane of the clear cell. Released acetylcholine leads to myoepithelial cell contraction, and sweat production is designated sudomotor activity [37‐39]. In theory, drugs that interact with each level of this pathway can induce sweating disorder. These putative drugs for decreasing sweat production were categorized and discussed in the following pharmacological categories:

There are several limitations in our study. First, FAERS is a spontaneous reporting system with inherent disadvantages, including false reporting, under-reporting, and incomplete reporting, all of which may lead to reporting bias. Although the identity of the reporter is visible, inaccuracy and omissions of the report content cannot be avoided. Second, although the basic information of the patients is provided, the underlying disease status is unknown, which will involve many confounding factors and introduce uncertainty into the analysis. Third, as with other pharmacovigilance studies, the signals obtained through disproportionate analysis can only provide a correlation between drugs and hypohidrosis or anhidrosis but cannot prove the causal relationship between them, which means that the approach could be used for hypothesis generation only, not for proof. Fourth, different drugs are approved at different times, which may result in fewer adverse event reports for drugs that have been on the market for a short period.

Although these drawbacks do exist, the FAERS database can identify signals of drugs associated with hypohidrosis or anhidrosis. More importantly, this study demonstrated that a pharmacovigilance system could alternatively be used to identify anti-hyperhidrosis treatment agents and sustainably create opportunities for drug repurposing.

Using pharmacovigilance for drug repurposing is an approach that is fundamentally different from all other currently employed methodologies. Based on pharmacovigilance analysis of FAERS, we have identified a plethora of candidate drugs for anti-hyperhidrosis treatment. However, more preclinical, and clinical trials are necessary to evaluate these hypotheses. Candidate drugs that were not explainable by our research should stimulate further study. The raw data and the results presented here might guide and thus accelerate these trials.