Systemic pharmacological treatments for acne: an overview of systematic reviews

Objectives

This is a protocol for a Cochrane Review (overview). The objectives are as follows:

To summarise the findings of systematic reviews that assess the effects (benefits and harms) of systemic pharmacological interventions for acne vulgaris.

Background

A glossary of medical terms is provided in Table 1.

1. Glossary of medical terms.

Term	Meaning
Adrenocorticotropic hormone	Hormone responsible for synthesis and secretion of cortisol (the primary stress hormone) in the adrenal glands
Androgenic alopecia	Hair loss associated with increased levels of androgens (male sex hormones)
Alpha‐melanocyte stimulating hormone	A neuropeptide produced in the pituitary gland that increases the synthesis of melanin and regulates inflammation
Activator protein‐1 (Ap1)	A transcription factor associated with several processes such as cellular apoptosis and proliferation
Cell apoptosis	Programmed cellular death
Cheilitis	Inflammation of the lips that can cause swelling, desquamation (peeling skin), and erosion
Chemotaxis	Oriented migration of cells or organisms in response to a chemical stimulus
Comedones	Result of the obstruction of a hair follicle by keratin or sebum. It can be open (blackhead) or closed (whitehead) by skin.
Corticotropin‐releasing hormone	Hormone that stimulates adrenocorticotropic hormone synthesis and secretion in the pituitary gland
Cytokeratin	Keratin proteins that form intermediate filaments in the intracytoplasmic cytoskeleton of epithelial tissue. It provides mechanical support to a cell.
Cytokines	Small proteins that influence immune responses
Dihydrotestosterone	A type of androgen (male sex) hormone
Dysbiosis	Imbalance in the composition of gut microbiota which causes a wide range of digestive disturbances
Epistaxis	Nose bleeding
Filaggrin	Protein produced by keratinocytes that promotes aggregation of keratin fibres in epithelial cells
Growth hormone, or human growth hormone	Hormone produced by the pituitary gland that in children and adolescents is responsible for the growth of bone and cartilage, and in adults plays a key role in insulin metabolism
Hirsutism	Abnormal increase in the amount of female body hair in male‐like pattern. It can result from an excess of androgens such as testosterone.
Hyperkalaemia	Elevated level of potassium in the blood
Hyperkeratinisation	Abnormal shedding and clumping of dead skin cells
Hyperpigmentation	Abnormal increase in melanin which can be triggered by such factors as sun exposure, hormones, age, and skin inflammation
Inflammatory bowel disease	A disorder of the large intestine characterised by chronic abdominal discomfort associated with alterations of stool consistency and frequency
Monocytes	The largest type of white blood cells, they are responsible for the formation of macrophages and dendritic cells
Metalloproteinases	Enzymes that contribute to extracellular degradation and the tissue remodelling process
Pilosebaceous unit	Structure containing a sebaceous gland with erector pili muscle, hair follicle, and hair shaft
Plurimetabolic syndrome	A group of conditions occurring together that increase a person's risk of developing heart disease, stroke, and type 2 diabetes. It includes increased blood pressure, high triglyceride levels, reduced high‐density lipoprotein cholesterol levels, large waist, and elevated fasting blood sugar.
Polycystic ovarian syndrome	Hormonal disorder common in women characterised by menstrual irregularity, hirsutism, acne, insulin resistance, and increased levels of androgens (male sex hormones)
Prolactin	Hormone produced in the pituitary gland that stimulates milk production in the mammary glands and breast size increase
Pruritus	An irritating skin sensation associated with itching and scratching
Pseudotumor cerebri syndrome	A syndrome that presents with clinical symptoms of high intracranial pressure but without any evidence of intracranial abnormality. It can be associated with systemic diseases and drug exposure.
Ribosomes	Cellular structure responsible for protein synthesis
Sebocytes	Epithelial cells that produce and release sebum
Teratogenicity	Capacity of agents to cause foetal abnormalities if administered to the mother at any stage of pregnancy
Xerosis	Medical disorder characterised by abnormally dry skin

Description of the condition

Acne is a chronic inflammatory and immune‐mediated disease of the pilosebaceous unit. Clinically, it is characterised by non‐inflammatory (open and closed comedones) and inflammatory (papules, pustules, nodules, and cysts) lesions; and it is diagnosed by physical examination. The lesions may be present on face, thorax, and back with variable severity. Acne fulminans and acne conglobata are rare and severe forms (Williams 2012).

According to the Global Burden of Disease study, acne vulgaris is a highly prevalent disease, affecting 231 million people of both sexes in 2019, and causing 4.96 million YLDs (years lived with a disability) (Lancet Global Health Metrics 2019). It exhibits a global distribution with a growing prevalence, mainly within the population between 15 and 19 years of age (Lynn 2016). Acne is one of the three most prevalent cutaneous diseases in different regions of the globe (Bhate 2013; Tan 2015; Wolkenstein 2003). Although some epidemiological studies have demonstrated differences regarding acne prevalence in different ethnicities (Davis 2010; Quarles 2007), there is no definitive evidence that a biological difference in acne pathogenesis can explain higher prevalence or severity of the disease amongst some specific ethnic groups. Recent epidemiological studies from high‐income countries tend to show an increased prevalence of acne amongst dark‐skinned participants, which seems to be associated with the progressive improvement of access to healthcare services in these regions of the globe during the last two decades (Lynn 2016). The prevalence amongst 452 adolescents, aged between 10 and 17 (mean 13.3 years), from elementary and high school in Sao Paulo city, Brazil, was 96.0%, increasing with age: acne was present in 100% of those over 14 years, and was predominately facial and of the comedonal type (61.1%) (Bagatin 2014). Acne is believed to affect 95% of adolescent boys and 85% of adolescent girls. Disease onset for girls is between 11 to 13 years, worsening around 18 years. Disease onset for boys is between 13 to 14 years, worsening around 19 to 21 years (Ghodsi 2009; Wei 2010). Acne may persist after adolescence (after 25 years of age), or occur for the first time in adulthood (late‐onset acne). It may also recur in adults many years after puberty. Prevalence of acne amongst the population over 25 years of age has been increasing in the last 30 years, with women representing the majority of the affected individuals in this age range (Goulden 1999; Khunger 2012).

Acne is a multifactorial disease. Many aetiopathogenetic factors are involved, but chronic inflammation is considered to be the central one. It has been demonstrated histologically that inflammation is present before the development of clinical inflammatory lesions (Rocha 2014). The activity of sebaceous glands starts at puberty through the linkage of androgens, especially dihydrotestosterone, to nuclear receptors of sebocytes, with subsequent increased lipogenesis and qualitative changes in sebum. The androgens also stimulate follicular hyperkeratinisation. In vitro studies have demonstrated the metabolism of sebaceous glands (Schneider 2018). A specific phylotype (IA1) of Cutibacterium acnes (previously named Propioniobacterium acnes) predominates in the follicles, and its recognition by toll‐like receptor 2, present in sebocytes, keratinocytes and monocytes, activates nuclear pathways like the nuclear factor kappa B (NF‐κB), with cytokines production, and the activator protein 1 (AP1), with metalloproteinases release (Dréno 2018). Histologic and immunohistochemical studies have shown a variety of inflammatory mediators in acne‐prone skin, such as: interleukin (IL)‐1, IL‐6, tumour necrosis factor alpha (TNF‐α), insulin‐like growth factor 1 (IGF‐1) and insulin‐like growth factor 1 receptor (IGF‐1R), peroxisome proliferator‐activated receptor‐gamma (PPAR gamma) and T helper (Th) cells, Th 1 and 17 (Agak 2018). Studies on the influence of occidental insulinogenic diet with high glycaemic load have suggested the specific role of IGF‐1 and IGF‐1R in acne pathogenesis, increasing sebum production. The mechanism is related to a deviation of forkhead box O1 (FoxO1)/mechanistic target of rapamycin complex 1 (mTORC1) signalling, with increased activation of mTORC1 and deficiency of FoxO1 (Melnik 2016). Neurohormones from hypothalamus (corticotropin‐releasing hormone) and hypophysis (adrenocorticotropic hormone, prolactin, growth hormone, alpha‐melanocyte stimulating hormone) also regulate the sebocytes, by direct stimulation or by increasing the sensitivity to testosterone derivatives (Clayton 2020). In adult women, the presence of polycystic ovarian syndrome, with or without hormonal alterations, in addition to hirsutism and menstrual irregularities, are important risk factors for acne (Bansal 2020). There has recently been a relevant discussion about the role of skin and gut microbiome in acne. This involves the decreased diversity of cutaneous and intestinal flora caused by prolonged use of oral and topical antibiotics, which affects host‐microbiome interactions and cutaneous barrier in individuals with acne, with the release of pro‐inflammatory cytokines (Deng 2018; O'Neill 2018; Rocha 2018).

Atrophic and hypertrophic scars, as well as postinflammatory hyperpigmentation, are possible sequelae. Scarring is caused by appearance change and histopathological change of normal skin tissue after skin injury. Skin damage during the healing of active acne may cause scars. Active scars may last 10 years or more, whilst acne scars may last a lifetime (Abdel Hay 2016). Systemic pharmacological treatments for acne cannot resolve scars, but early effective treatment prevents acne scars. Risk factors for scarring include treatment delay, more severe acne, male gender, lesion manipulation habits, family history of acne scars, and relapsing acne (Tan 2017a; Tan 2017b).

It should be highlighted that there may be a negative impact of this disease on quality of life, and that acne is also associated with such psychological sequelae as depression, anxiety, loneliness, and internet addiction (Gieler 2015; Öztekin 2020). Acne may seriously affect an individual's mental health, especially in those with severe acne. Acne tends to occur in adolescents, and may be emotionally devastating at this stage of life when self‐image is emphasised (Lauermann 2016). Anxiety, depression, and even suicidal tendency may appear in those with acne (Lukaviciute 2017), with risk influenced by disease severity, local pigmentation, and scar formation. Acne has a huge impact on the quality of life of those affected (Alanazi 2018). Several studies have found a close two‐way relationship between mental stress and acne (Ramrakha 2016; Wen 2015). In a competitive, modern world where good appearance is often considered advantageous, people with acne can suffer feelings of low self‐confidence, rejection, and are often humiliated by others (Cenk 2020). Positive treatment can help improve the psychological state of those with acne, possibly alleviating feelings of depression and anxiety and improving quality of life (Kaymak 2009). Individuals with acne therefore need not only medication to alleviate their symptoms, but also timely psychological intervention, so as to avoid the occurrence of mental illness.

Description of the interventions

Systemic antibiotics, hormonal antiandrogens, and oral isotretinoin are the main systemic interventions for acne management. According to guidelines, each one of these options may be used to treat moderate to severe papulopustular and nodulocystic acne (Nast 2016), the categorisation of moderate or severe done based on the subjective opinion of the physician (Thiboutot 2018). However, it is important to highlight that only oral isotretinoin is recommended as monotherapy for moderate to severe cases in guidelines for acne treatment, and the use of antiandrogens only applies to female patients who want or agree to avoid pregnancy (Bagatin 2019; Nast 2016; Thiboutot 2018; Zaenglein 2016). Besides these drugs, oral corticosteroids also have a more restricted role in acne vulgaris therapy: they are specifically recommended for acne fulminans, a rare and severe presentation, which may occur with or without systemic symptoms and be de novo or secondary to isotretinoin use when it is initiated in high doses (Greywal 2017; Zaenglein 2016).

Oral antibiotics

Antibiotics are the most prescribed systemic antiacne treatment in daily clinical practice (Barbieri 2019), either in dermatological or non‐dermatological physician practices (Barbieri 2017). However, there are clinical guideline recommendations to limit use and duration of antibiotic therapy (maximum of three to four months) and to avoid prescription of oral antibiotic as monotherapy, in order to minimise the risk of emergent antimicrobial resistance (Nast 2016; Ozolins 2004; Thiboutot 2018; Zaenglein 2016). Oral tetracyclines, especially the second‐generation tetracyclines (minocycline, doxycycline, and lymecycline), are the mainstay when opting for a systemic antibiotic to manage acne (Armstrong 2020). Their advantages in comparison to first‐generation tetracyclines (tetracycline and oxytetracycline) are related to pharmacokinetics. Their dosage is more convenient to patients, due to all three having longer half‐lives that allow a single daily dosage. Furthermore, taking with meals does not impair their absorption (Ochsendorf 2006). However, there is still uncertainty as to the most efficacious and safe therapeutic oral tetracycline regimens in acne management (Armstrong 2020; Garner 2012; Williams 2012). There is no definitive and consistent body of evidence indicating a greater efficacy of minocycline, doxycycline, or lymecycline when compared to tetracycline or to each other (Garner 2012; Zaenglein 2016). Data from one Cochrane Review could not corroborate the contention that extended‐release minocycline formula is safer than other cyclins preparations (Garner 2012). Despite gaps in knowledge and research concerning oral antibiotics in acne management, there is consensus, and an evidence‐based approach, to recommend the concurrent use of topical benzoyl peroxide, a non‐antibiotic antimicrobial agent, when prescribing oral antibiotics to acne patients (Nast 2016; Thiboutot 2018; Tzellos 2011; Zaenglein 2016). This strategy is aimed at preventing treatment failure due to antibiotic‐resistant acne, and at reducing the prevalence of antibiotic‐resistant C acnes strains in communities (Ozolins 2004; Tzellos 2011). In clinical daily practice, the usual dosing of oral tetracyclines for acne is as follows: 50 to 135 mg daily for minocycline; 250 to 500 mg daily or twice daily for tetracycline; 50 to 100 mg daily or twice daily for doxycycline; and 150 to 300 mg daily for lymecycline (Armstrong 2020; Nast 2016; Ochsendorf 2006; Zaenglein 2016).

In 2019, sarecycline, a novel oral tetracycline, was released in the United States specifically to treat moderate to severe acne. Its efficacy in comparison to the other tetracycline‐class antibiotics in clearing acne lesions has not yet been tested in randomised controlled trials. However, sarecycline has a narrower‐spectrum antibacterial activity and lower potential for triggering intestinal dysbiosis, which may result in better tolerability and fewer gastrointestinal side effects than older drugs of its class (Armstrong 2020). A more focused action against Cacnes, associated with a decreased activity against gram‐negative bacteria, may also lessen concerns related to the potential for bacterial resistance following prescription of this fourth‐generation tetracycline (Kircik 2020).

Gram‐negative folliculitis of the face and trunk, which is frequently caused by bacteria with poor response to conventional antimicrobial acne therapies (Lehrhoff 2012), and Candida vaginitis may complicate acne after prolonged treatment with oral tetracycline, as well as after treatment with other oral antibiotics, including all other tetracycline‐class drugs (Lehrhoff 2012; Zaenglein 2016). Besides the more common gastrointestinal side effects, such as nausea, vomiting, abdominal aches, and diarrhoea, oral antibiotics of the tetracycline class when used to treat acne may also be associated with photosensitivity, vulvovaginal fungal infections, pruritus, urticaria, cutaneous rash, vertigo, headache, and pseudotumour cerebri syndrome. The last three central nervous system conditions are more frequently found among minocycline‐treated patients (Armstrong 2020). Two tetracycline‐class antibiotics, doxycycline and minocycline, have been associated with new‐onset inflammatory bowel disease (IBD) (an increase in risk of development of 60% and 20%, respectively) (Margolis 2010). This could be linked to an imbalance in normal intestinal flora due to repeated or long‐term courses of these drugs: individuals with active IBD exhibit modifications in their microbial colonisation regarding functions, abundance, composition, and diversity, which results in dysbiosis, a disruption in intestinal homeostasis and local immune responses (Lee 2021). Greyish‐blue tissues pigmentation (seen in skin, nails, sclera, conjunctiva, thyroid, teeth, and bone during therapy with higher cumulative doses), which may be irreversible (Garner 2012), and systemic lupus erythematosus (SLE) (a rare occurrence, but an over twice times increased risk in comparison to acne treatment with other oral tetracyclines, according to Margolis 2007) are more uncommon adverse effects of minocycline use. In the last 25 years, a few rare serious adverse effects of minocycline have also been documented when used for acne, such as acute organ failure related to early‐onset therapy with higher doses (mainly hepatic); DRESS (drug‐induced reaction with eosinophilia and systemic symptoms); and autoimmune diseases (hepatitis, thyroiditis, polyarteritis nodosa, and SLE‐like syndrome) (Garner 2012; Ochsendorf 2006). Second‐generation tetracyclines show enhanced gastrointestinal absorption, accumulation in body fat deposits, as well as (especially for minocycline) high affinity for calcified tissues and a capacity for crossing the blood‐brain barrier. All of these pharmacokinetics characteristics contribute to the adverse effects cited above (Armstrong 2020; Garner 2012; Ochsendorf 2006).

Azithromycin and erythromycin, antibiotics of the macrolides class, are second‐line systemic therapies for inflammatory acne due to rising antimicrobial resistance to macrolides (Bienenfeld 2017; Sardana 2021). Both drugs are used off‐label to treat acne, and have their best indication when there are restrictions to tetracyclines use, as occurs with pregnant women or children under eight years old (Patel 2020; Schachner 2020; Zaenglein 2016). Consistent with clinical trial dosage regimens, azithromycin is usually recommended to manage acne as pulse therapy: dosages are 500 mg taken once to three times in a week, or four times monthly for 12 weeks (Kim 2018). The usual antiacne dose of erythromycin is 500 mg twice daily (Ochsendorf 2006). The main side effects of macrolides for acne are gastrointestinal, with erythromycin being more associated with gastrointestinal effects than azithromycin. Less often, macrolides can cause hepatic toxicity, cardiac arrhythmias, and skin hypersensitivity reactions (Zaenglein 2016).

Hormonal antiandrogens

Spironolactone

Spironolactone was originally introduced into clinical practice to manage systemic arterial hypertension and heart failure due to its diuretic effect (Garthwaite 2004; Zouboulis 1994). However, besides the antagonistic action on mineralocorticoid receptors in kidneys, which spares potassium and increases water and sodium renal elimination, this synthetic 17‐lactone steroid also has a moderate tendency to bind to progesterone and androgens receptors in the human body (Garthwaite 2004; Layton 2017). This lack of selectivity for the mineralocorticoid receptor, which resulted in endocrine side effects (gynaecomastia, decreased libido, menstrual irregularities, and impotence) in individuals treated for cardiovascular diseases, also made spironolactone a possible treatment for hirsutism, polycystic ovarian syndrome, androgenic alopecia, and, since the 1980s, acne (Garthwaite 2004; Zouboulis 1994). Spironolactone remains an off‐label option for some women, both adult and teenaged, with acne vulgaris (Barbieri 2019; Kim 2012; Zaenglein 2016). It has been recommended as an alternative therapy for female patients who present with moderate to severe inflammatory acne. Guidelines suggest that it be used concurrently with topical agents plus systemic antibiotics or with a combination of topical antiacne drugs. As part of an antiandrogen tailored approach for women, spironolactone alone may comprise the acne therapeutic regimen, or it may be prescribed in combination with oral contraceptives (Nast 2016; Zaenglein 2016). The optimal use of spironolactone is for female patients with prominent seborrhoea, as a clinical manifestation of SAHA (seborrhoea/acne/hirsutism/androgenic alopecia) syndrome, and late‐onset acne vulgaris, even when androgen levels are not raised in laboratory tests (Kim 2012).

In accordance with consensus and expert opinion, spironolactone daily dosage ranges from 100 to 200 mg (Layton 2017; Zaenglein 2016). Adverse events are more common with higher doses (> 100 mg): menstrual irregularities; breast tenderness and enlargement; dizziness, vertigo, and light‐headedness; headache; nausea with and without vomiting; weight gain; abdominal pain; polyuria; fatigue; and lethargy were the most frequent adverse events reported by a systematic review (Layton 2017). The review also found a lower incidence of menstrual disturbances when used in association with a combined oral contraceptive (COC) compared to the prescription of spironolactone without a COC. The risk of hyperkalaemia due to spironolactone is not a concern in the absence of specific risk factors such as: older age; hypertension; hepatic, renal, or adrenal disease; and concomitant intake of drugs which could increase the risk (i.e. angiotensin‐converting enzyme inhibitor, angiotensin receptor blockers, digoxin, potassium‐sparing diuretics, sulfamethoxazole/trimethoprim, non‐steroidal anti‐inflammatories, and potassium supplements). Blood potassium measurements before and during therapy for young, healthy acne patients are therefore unnecessary (Barbieri 2019; Layton 2017; Zaenglein 2016). Avoidance of pregnancy is important throughout treatment, as spironolactone may cause feminisation of the male foetus in animal studies. Combination therapy with oral contraceptives also helps to address this concern (Kim 2012; Zaenglein 2016).

Combined oral contraceptives (COCs)

The efficacy of COCs to clear inflammatory and non‐inflammatory acne lesions and to improve patient‐reported outcomes, when compared to placebo, has been demonstrated in randomised controlled trials (RCTs) (Arowojolu 2012). COCs with ethinyl oestradiol and an antiandrogenic progestin (especially cyproterone acetate, drospirenone, chlormadinone acetate, or dienogest) in combination are most frequently recommended by physicians for teenage and adult women (Bagatin 2019; Dréno 2013). There is evidence that COCs containing chlormadinone acetate and cyproterone acetate as the progestins improve acne more than COCs containing levonorgestrel, a progestin with androgenic activity. RCT evidence suggests that COCs with drospirenone as the progestin combined with ethinyl oestradiol work less well for acne than COCs containing cyproterone acetate. However, they may be more effective than COCs containing norgestimate plus ethinyl oestradiol, or those containing nomegestrol acetate plus 17β‐oestradiol (Arowojolu 2012). COC use increases the risk of thromboembolic events as well as causing a small increase in the relative risk of breast cancer (de Bastos 2014; Raymond 2012). A detailed medical history and blood pressure measurement are therefore required before starting therapy. The most common side effects of COCs are nausea, breast enlargement, headaches, and weight gain. Concomitant use with antibiotics may reduce the contraceptive effect, as in the case of rifampicin and griseofulvin. Combined treatment with tetracyclines is safe, unless patients experience vomiting or diarrhoea, when an additional contraceptive method should be adopted (Dréno 2004; Raymond 2012).

Isotretinoin

Amongst systemic treatments, isotretinoin is the only one that is recommended as monotherapy in current acne treatment guidelines. It is the first‐line choice for severe nodulocystic acne and is also recommended therapy for severe papulopustular and moderate cystic nodular acne which relapses rapidly after initial therapeutic approach, has concomitant scarring, or compromises quality of life (Bagatin 2019; Nast 2016; Zaenglein 2016). After analysing the best available evidence, a task force study from the European Academy of Dermatology and Venereology found that isotretinoin improves quality of life in acne patients more than any other intervention (Chernyshov 2018).

Conventional dosage is 0.5 to 1.0 mg/kg daily, which must be taken with meals (an exception is the novel formula isotretinoin‐lidose, which is not affected by fasting), and, in accordance with the drug label, until a total cumulative dose of 120 to 150 mg/kg (Bagatin 2020a). However, a recent Cochrane Review failed to find RCT data that supported a specific cumulative dose to yield acne clearance in accordance to the severity of the disease (Costa 2018). There is a current tendency, with a basis in expert experience and opinion, to consider individual clinical aspects, when deciding upon duration of therapy with isotretinoin, and to maintain the dose for two to four months after total resolution of the lesions (which usually occurs after four to six months of therapy) (Fallah 2020; Rademaker 2013; Tan 2016; Thiboutot 2018). Low‐dose daily isotretinoin (< 0.5 mg/kg) improves moderate to severe acne less than conventional‐dose therapy but is associated with fewer adverse effects (Costa 2018). The last update of the European guideline for the treatment of acne recommended a dose of 0.3 to 0.5 mg/kg/day to treat severe papulopustular and moderate nodular acne (Nast 2016). Intermittent schemes (i.e. pulsed therapy) have a lower efficacy than low and conventional‐daily dosage, and international expert guidelines do not recommend their use (Costa 2018; Zaenglein 2016).

Mucocutaneous adverse events, especially those linked to secondary xerosis, may affect up to 90% of individuals treated with isotretinoin. Lips desquamation, cheilitis, dry skin, pruritis, dermatitis, dry nose, rhinitis sicca, epistaxis, dry eyes, keratoconjunctivitis, and dry mouth are frequently reported in clinical trials (Costa 2018; Vallerand 2018). As observed with laboratory changes associated with the use of oral isotretinoin (the most frequent, with an incidence of 2%, are increases in triglycerides, cholesterol and transaminases serum levels, but reduction in blood cell counts may also occur), mucocutaneous side effects are dose‐dependent and easily manageable with dosage reduction (Bagatin 2020a; Lee 2016a; Nast 2016). Laboratory monitoring, with lipid and hepatic panel at baseline and at one to two months of treatment, is required. Additional tests are recommended only in the presence of comorbidity that increases the risk of laboratory abnormalities (plurimetabolic syndrome and use of hepatotoxic drugs constitute the main indications), or if some change is detected in one of the two blood assessments (Altman 2002; Bagatin 2020b; Barbieri 2020a; Barth 1993; Lee 2016a). Teratogenicity is dose‐independent and represents the most serious adverse event. Pregnancy blood tests are required for women in reproductive age just before starting therapy and monthly until the end of the treatment. Also, patients must adhere to a very rigorous contraception scheme, preferably with intrauterine device or COC combined with condoms, during the entire treatment and up to one‐month post‐therapy (Bagatin 2019; Barbieri 2020b; Lee 2016a). IBD and psychiatric events (depression, suicide, psychosis) have been associated with oral isotretinoin for acne, but a causal relationship between these serious adverse events and isotretinoin has thus far not been established (Costa 2018; Etminan 2013; Huang 2017; Lee 2016b). Regarding IBD, the higher risk amongst isotretinoin users in a few retrospective studies is probably related to the frequent exposure to systemic antibiotics for acne (chiefly from tetracycline‐class antibiotics) prior to commencing isotretinoin in patients with moderate to severe acne (Margolis 2010; Wright 2021). Prospective controlled trials have shown a greater improvement of depression symptoms and quality of life with oral isotretinoin when compared to other antiacne therapies (Chernyshov 2018; Costa 2018). However, due to the possibility of serious psychiatric events as a rare, idiosyncratic reaction to isotretinoin, monitoring for behaviour changes in physician visits during isotretinoin treatment is recommended (Bagatin 2020b).

How the intervention might work

Oral antibiotics

The role of tetracyclines in acne control is not only related to the bacteriostatic action, which inhibit C acnes growth and proliferation: this class of drugs, along with macrolides and sulfamethoxazole‐trimethoprim, shows anti‐inflammatory properties (Armstrong 2020; Eady 2003; Zaenglein 2016). First‐ and second‐generation cyclins, and the newest tetracycline sarecycline, all have antiacne effects by interrupting bacterial protein synthesis after the linkage to 30S subunit in ribosomes, but also by promoting inhibition of neutrophil chemotaxis, cytokines, and matrix metalloproteinases release (Armstrong 2020; Monk 2011). The use of sub‐antimicrobial and lower doses may be sufficient to acne treatment due to the non‐antimicrobial, anti‐inflammatory actions of tetracyclines (Simonart 2008; Thiboutot 2018; Tzellos 2011). However, concern has been raised that the prescription of sub‐antimicrobial lower daily doses of second‐generation tetracyclines might promote C acnes antibiotic resistance, chiefly in patients with intense seborrhoea (Eady 2003; Nord 2006). Tetracyclines provide immunomodulation in acne management by both specific cell‐mediated and innate immune response (Eady 2003).

Similar to tetracyclines, macrolides have anti‐inflammatory properties (though the mechanism is unclear) as well as antiacne activity due to bacteriostatic effect on C acnes (this class of antibiotics act by binding to the 50S subunit in ribosomes) (Ochsendorf 2006; Zaenglein 2016). Sulfamethoxazole inhibits C acnes growth by halting synthesis of folic acid, which is indispensable to replication of bacterial cells. Trimethoprim, a folic acid analogue, interferes with the same metabolic by blocking the enzyme dihydrofolate reductase. These last two drugs work synergically when combined to treat acne, restricting bacterial processing of amino acids and nucleotides synthesis (Zaenglein 2016).

Hormonal antiandrogens

Spironolactone exerts its antisebogenic properties by the competitive blockage of the interaction between dihydrotestosterone and androgen receptors inside sebaceous cells of the skin, especially on the face. There is a local decrease in sebogenesis secondary to androgenic stimulus (Zouboulis 1994). COCs work on acne lesions due to the same mechanism as spironolactone (antiandrogenic effects on sebaceous glands cells), but the oestrogen component also contributes to sebosuppression by decreasing free testosterone serum levels by half. This effect is due to two mechanisms: inhibition of luteinising hormone secretion, with consequent reduction of androgens production; and increase in serum levels of sex hormone‐binding globulin, which binds to free testosterone (Thorneycroft 1999). Despite differences in the antiandrogenic activity of progesterones (only cyproterone acetate, drospirenone, chlormadinone acetate, and dienogest may not exhibit androgenic effect when used alone) (Regidor 2017), the presence of oestrogen results in antiacne action in all COCs (Arrington 2012).

Isotretinoin

Isotretinoin is the single systemic antiacne therapy to address all recognised acne aetiopathogenic mechanisms (Bagatin 2020b). It is a 13‐cis isomer of the all‐trans retinoic acid (ATRA), same tretinoin, and a prodrug that has effects on nuclear retinoic acid receptors (RARs) only after the intracellular conversion to metabolites, such as ATRA, 4‐oxo‐tretinoin, 4‐oxo‐isotretinoin, 9‐cis acid retinoic, and 9‐cis‐4‐oxo retinoic acid (Khalil 2017). These retinoids compounds bind to RARs and down‐ or upregulate the transcription of different genes, modifying cellular processes within keratinocytes and sebocytes (Nelson 2009). Isotretinoin normalises follicular hyperkeratinisation by increasing the production of cytokeratin 7, 13, and 19, laminin B1, and IL‐1, whilst promoting a decrease in the release of cytokeratin 1, 10, and 14, filaggrin, and metalloproteinases. The changes to the corneal layer of the skin favour cellular proliferation and shedding, with consequent renewal within the follicular units and reduced comedogenesis (Törmä 2011). Also, isotretinoin enables local immune regulation, with a decrease in the acne inflammatory process, chiefly by downregulation of gene expression related to cell membrane receptors Toll‐like 2 e 4 (TLR‐2 and 4) and T helper cells (Chen 2019; Dispenza 2012). Isotretinoin inactivates the androgen nuclear receptor in sebaceous glands and increases cell apoptosis, which explains the great reduction in sebum excretion (Melnik 2017). Finally, isotretinoin changes the microenvironment within the pilosebaceous follicle, especially regarding sebum quantity and composition, making it unfavourable to hypercolonisation by C acnes (Ganceviciene 2010; McCoy 2019).

Why it is important to do this overview

Despite important advances in the more than 50 years since physicians first prescribed oral antibiotics as a first‐line therapy to manage inflammatory acne (Tzellos 2011), there is still a lack of consistent evidence regarding the superiority of an oral antibiotic combined with a topical agent compared to oral antibiotic alone, even though monotherapy with systemic antibiotic is widely and strongly contraindicated by expert acne management guidelines around the globe (Nast 2016; Thiboutot 2018; Zaenglein 2016). There are also key questions that remain unelucidated with high‐certainty evidence regarding tetracyclines, the main class of systemic antibiotics in acne management. The following questions remain unresolved: optimal duration of therapy; most efficacious dosage regimens (i.e. once daily versus twice daily, specifically with respect to second‐generation tetracyclines; higher, full therapeutic dose versus lower dose and tapering dose for all drugs of the class); differences in efficacy and safety between the tetracyclines (tetracycline, oxytetracycline, doxycycline, minocycline, lymecycline, and sarecycline); and superiority of extended‐release formulas of minocycline and doxycycline compared to regular preparations (Garner 2012; Moore 2015; Williams 2012; Zaenglein 2016).

Regarding the use of spironolactone to manage acne in women, evidence to support its use is still lacking, despite it being increasingly prescribed over the last decade as an important antibiotic‐sparing therapeutic option (Barbieri 2017), and its use being supported by acne experts and treatment guidelines (Bagatin 2019; Layton 2017). The best dosing regimen to achieve acne clearance without important side effects is uncertain. Relevant questions regarding spironolactone are: effectiveness of spironolactone for acne, compared to oral antibiotics, COCs, and isotretinoin; the benefit of prescribing spironolactone in association with other antiacne therapies (both topical and systemic); and the specific acne patients who are more responsive to spironolactone (Layton 2017).

Regarding COCs, there is no high‐certainty evidence of the superiority of pills containing progestins with antiandrogenic activity, such as cyproterone acetate, drospirenone, chlormadinone acetate, and dienogest. Furthermore, the efficacy and safety of COCs for acne in comparison to other systemic treatments is still a knowledge‐gap (Arowojolu 2012; Bagatin 2019; Zaenglein 2016).

A recent Cochrane Review analysed the use of oral isotretinoin for acne and reported some gaps in scientific knowledge. These were related to:

1. the best therapeutic regimen, cumulative dose, and duration of therapy to achieve better efficacy outcomes and fewer adverse effects;

2. the lack of RCTs evaluating the long‐term effectiveness of the drug in comparison to oral antibiotics plus topical agents for moderate to severe acne;

3. the need of standardised outcome measurements for future well‐designed RCTs, which would enable the production of high‐certainty evidence about isotretinoin for acne after pooling results from multiple studies;

4. scarcity of prospective, observational studies with long duration and large sample size to clarify uncertainties with regard to very rare adverse events, including the possible causal relationship between the drug and serious psychiatric disorders and IBD; and

5. lack of data regarding the true benefits and harms of isotretinoin for acne of varying severities (specially for milder forms) and for specific subgroups of patients, such as adult women, pre‐adolescents, and people with truncal acne (Costa 2018).

An overview of systematic reviews of the systemic pharmacological therapies for acne will provide a summary of the best available evidence and answer the following question: what is the evidence for the comparative efficacy and safety of systemic interventions for treating acne vulgaris? There are not any other Cochrane or non‐Cochrane overviews on this topic. The title of this overview of systematic reviews has been prioritised by Cochrane Skin in their 2020 prioritisation exercise, which aimed to identify the most important systematic review titles within the group's scope (Cochrane Skin 2020). Another overview of systematic reviews focusing on non‐systemic interventions for acne is also being produced (Yuan 2021).

Objectives

To summarise the findings of systematic reviews that assess the effects (benefits and harms) of systemic pharmacological interventions for acne vulgaris.

Methods

Criteria for considering reviews for inclusion

Types of reviews

We will include Cochrane Reviews and non‐Cochrane Reviews of systemic pharmacological treatments for acne vulgaris. We will exclude reviews that are rated as high risk of bias using the ROBIS tool (see Assessment of methodological quality of included reviews).

We will only include reviews that include data from RCTs.

We will include all non‐overlapping systematic reviews. However, for groups of overlapping reviews, we will primarily include the latest Cochrane Review unless there is a more recent high‐quality non‐Cochrane Review. Where there is no Cochrane Review, we will include the highest‐quality, most relevant, or most comprehensive systematic review.

We will use a template table (Table 2) to show any overlap in included RCTs across reviews.

2. Template for a table mapping the primary randomised controlled trials (RCTs) contained within each included systematic review.

	Review 1	Review 2	Review 3	[...]	Review 'k'
Primary RCT 1	Yes/No/Related Note#	Yes/No/Related Note	Yes/No/Related Note	[...]	Yes/No/Related Note
Primary RCT 2	Yes/No/Related Note	Yes/No/Related Note	Yes/No/Related Note	[...]	Yes/No/Related Note
[...]	[...]	[...]	[...]	[...]	[...]
Primary RCT 'i'	Yes/No/Related Note	Yes/No/Related Note	Yes/No/Related Note	[...]	Yes/No/Related Note

^#Mark 'Yes', 'No', or a related note in each cell to display which original RCTs are included in each included review, so that it is possible to determine whether different reviews included the same RCTs.

Types of participants

People with a diagnosis of mild, moderate, or severe acne vulgaris defined by any classification system or healthcare professional. We will include individuals regardless of ethnicity, sex, age, or healthcare setting.

If a systematic review considered other types of acne as well as acne vulgaris, we will include the review and analyse only the data from RCTs addressing acne vulgaris.

Types of interventions

We will consider oral isotretinoin, oral antibiotics, spironolactone or combined contraceptive pills for acne in any dose, regimen, or duration of treatment.

We will also include systematic reviews in which the intervention groups included concomitant use of any other systemic or topical antiacne pharmacological intervention.

We will not include systematic reviews that mainly focus on complementary therapies.

Types of comparisons

• Any other topical or systemic pharmacological acne treatment, when used alone or in therapeutic schemes based on the association of antiacne drugs.

• Placebo or no intervention.

• Itself in a different dose and therapeutic regimen (e.g. pulsed or intermittent therapy, when the drug is taken only during one prespecified period per week or month; and use of the intervention only in alternate days).

Types of outcomes

The included reviews should have addressed at least one of the outcomes listed below; however, we will include reviews that aimed to report at least one of the outcomes listed but found no data.

Primary outcomes

1. Improvement in acne severity assessed by a decrease in total lesion counts and inflammatory and non‐inflammatory lesion counts (if counted separately).

2. The frequency of participants experiencing at least one serious adverse event (defined as those associated with death, hospitalisation, foetal losses, miscarriage, malformation, or permanent disability, in accordance with FDA 2021).

Secondary outcomes

1. Improvement in acne severity assessed by participant's global self‐assessment using a Likert‐type scale.

2. Improvement in acne severity assessed by investigator’s global assessment using a Likert‐type scale or a previously published subjective grading system.

3. Quality of life measured by any specific or generic validated tool.

4. The frequency of participants experiencing at least one less serious adverse event. We intend to analyse frequency of serious and less serious adverse events only as reported in the included reviews.

These efficacy outcomes are at present the most frequently adopted in acne clinical trials according to the work of the Acne Core Outcomes Research Network (ACORN) (Thiboutot 2019).

Timing of outcome assessment

We will assess all of the aforementioned outcomes at any time point of measurement, during treatment and follow‐up after the end of therapy, in accordance with the information reported in the included systematic reviews. We will consider time points of assessment as short term (up to 12 weeks, inclusive) or long term (more than 12 weeks). If more than one time point is reported in the short‐ or long‐term category, we will use the final time point in these categories.

Search methods for identification of reviews

We aim to identify all relevant systematic reviews regardless of language, date, or publication status (fully published, unpublished, in press, or in progress).

Electronic searches

The Cochrane Skin Information Specialist will search the following databases for relevant systematic reviews with no date restriction:

• the Cochrane Database of Systematic Reviews (CDSR) in the Cochrane Library;

• MEDLINE Ovid (from 1946 onwards);

• Embase Ovid (from 1974 onwards); and

• Epistemonikos (www.epistemonikos.org/en/).

We will use terms for acne to search CDSR (see Appendix 1) and Epistemonikos (see Appendix 2). We will use filters developed for identifying systematic reviews by the Scottish Intercollegiate Guidelines Network (SIGN) with acne terms to search MEDLINE (see Appendix 3) and Embase (see Appendix 4).

Searching other resources

Searching reference lists

We will check the bibliographies of included systematic reviews for further references to relevant reviews.

Correspondence with original authors

We will contact the original authors of systematic reviews for clarification and further data if necessary.

Errata and retractions

The Cochrane Skin Information Specialist will run a specific search to identify errata or retractions related to our included reviews, and we will examine any relevant retraction statements and errata that are retrieved.

Data collection and analysis

The methodology for data collection and analysis is based on Chapter V of the Cochrane Handbook of Systematic Reviews of Interventions (Pollock 2021).

Selection of reviews

Two overview authors (CSC and RR; or RLP and ZY for the screening of any systematic reviews authored by CSC or RR) will independently select systematic reviews by screening each title and abstract against the inclusion criteria. Two overview authors (CSC and RR; or RLP and ZY) will independently assess the full texts and then select reviews for inclusion by consensus. Any disagreements will be resolved by discussion for by involvement of a third overview author (RLP or TP) if necessary. During the title and abstract screening stage, all systematic reviews will be included regardless of quality. During the full‐text screening stage, systematic reviews judged to be at high risk of bias according to the ROBIS tool will be excluded (see Assessment of methodological quality of included reviews).

Data extraction and management

Two overview authors (CSC and EB; or RLP and ZY) will independently extract the following information using an electronic form, which we will pilot on each included systematic review.

Data from included systematic reviews

• Basic data from systematic reviews (title; authors; date of last assessment as up‐to‐date; number of included studies and participants; and year of publication).

• Search strategies of the included systematic reviews (searched databases; date ranges used in each searched database; and date of the last update of the search).

• Characteristics of participants of the included systematic reviews (age; gender; ethnicity; care setting; and severity of acne).

• Interventions analysed by the systematic reviews (dose and therapeutic scheme; duration of therapy; and use alone or in combination with other antiacne treatments).

• Outcomes assessed, as prespecified in the systematic reviews.

• Data of effects of interventions (effect estimates); 95% confidence intervals (CIs); measures of heterogeneity when results of primary studies have been pooled in the reviews; and assessments of the certainty of the effects.

• Any other characteristics required to assess and report on review quality (see Assessment of methodological quality of included reviews).

• Conflicts of interest and funding sources for the authors of included systematic reviews.

• Data of any limitations in the systematic review.

Data from primary studies analysed in the included systematic reviews

• Basic data about RCTs analysed in the included systematic reviews (authors; country and year of publication).

• Risk of bias assessments of the primary studies of the included systematic reviews.

• Conflicts of interest and funding sources for the authors of the primary studies.

A third‐party author (RR or TP) will resolve any disagreements during data collection and management. Where data are missing from the systematic reviews, we will email the review authors to request further information. However, we will not extract missing data directly from the underlying primary studies, but will simply report the gap in coverage in our overview (Pollock 2021). We will enter data into Cochrane Review Manager 5 software (Review Manager 2020).

Assessment of methodological quality of included reviews

Methodological quality of included Cochrane and non‐Cochrane Reviews

We will analyse each included review to determine to what degree it satisfies the criteria specified in the ROBIS tool, which assesses risk of bias in a systematic review (Whiting 2016). The tool is completed in three phases: 1) assess relevance (optional), 2) identify concerns with the review process, 3) judge risk of bias. Phase 2 includes four domains: study eligibility criteria; identification and selection of studies; data collection and study appraisal; and synthesis and findings. Each domain has five or six signalling questions which can be answered with 'yes', 'probably yes', 'probably no', 'no', or 'no information', with 'yes' indicating low concerns. The level of concern for each domain can then be judged 'low', 'high', or 'unclear'. If the answers to all signalling questions are 'yes' or 'probably yes', then the level of concern can be judged as low. However, if any of the signalling questions are answered with 'no' or 'probably no', then there is potential concern regarding the existence of bias. Phase 3 of the tool considers whether the systematic review is at risk of bias overall. In this phase, the overview author needs to summarise any concerns from the four domains in phase 2, and then answer three signalling questions. As in phase 2, if these signalling questions are answered 'yes' or 'probably yes', then the overall risk of bias can be judged as low. If one or more signalling questions are answered 'no' or 'probably no', then the overall risk of bias is likely to be high. Two overview authors (CSC and EB; or ZY and RLP) will independently apply all items of the ROBIS tool to each of the selected reviews. A third party author will resolve any disagreements during collection and management of data (RR or TP). We will present the results of the assessments in both tabular and graphic format.

Quality of the trial evidence included in reviews

Two overview authors (CSC and EB; or RLP and ZY) will independently extract data from the risk of bias assessments conducted by the review authors for each individual included RCT. A third overview author (RR or TP) will resolve any discrepancies throughout the process. We will not reassess risk of bias in RCTs analysed by our included systematic reviews. When there are available results with quantitative data related to any one of our outcomes, we will extract quality assessments conducted using the GRADE approach (Guyatt 2008), which assesses the confidence in the effect estimates, and present these in the summary of findings tables of our included systematic reviews. If GRADE assessments are not reported in the included reviews, we will attempt to perform GRADE evaluations using GRADEpro GDT software for summary measures of effect (GRADEpro GDT), but only if sufficient information is available in the systematic review (Pollock 2021).

Data synthesis

Summary of findings and assessment of the certainty of the evidence

We will present summarised data related to our primary and secondary outcomes using a narrative approach in the 'Effects of interventions' section of the overview. We will tabulate data within our summary of findings tables when there is available data related to any of our primary and secondary outcomes in summary of findings tables of our included reviews. We will follow the template for an overview summary of findings table described in Pollock 2021, where the results will be organised by outcome and then by comparison. In the case of sufficient data, we will address the following main comparisons in our summary of findings table:

• oral isotretinoin versus no treatment or placebo;

• oral antibiotic versus no treatment or placebo;

• oral antiandrogens versus no treatment or placebo;

• oral isotretinoin versus oral antibiotics;

• oral antiandrogens versus oral isotretinoin;

• oral antiandrogens versus oral antibiotics.

Regarding quantitative results measurements of the systematic reviews, we will narratively summarise: effects estimates; 95% CIs; measures of heterogeneity of meta‐analysis (I² value and the respective thresholds as extracted by us from included systematic reviews that pooled data from the primary studies); and GRADE assessments to provide the certainty of the evidence as reported in the included systematic reviews, or as undertaken by us if not reported (see Assessment of methodological quality of included reviews) (Pollock 2021). We will not re‐analyse outcome data from the included systematic reviews other than to conduct subgroup analyses. We do not plan to conduct any additional indirect comparisons or network meta‐analyses, as we expect that the assumptions of similarity and consistency for these analyses are probably invalid in the evidence network of current therapeutic field of acne.

For each of our outcomes, we will create a bubble plot, where on the x axis is the treatment effect (categorised as large, moderate, small, or trivial or small unimportant effect or no effect) (Schünemann 2021), and on the y axis is the GRADE assessment (certainty of the evidence categorised as high, moderate, low, or very low). The bubble size will correspond to the number of participants involved in each pooled analysis of a systematic review, and the colour of the bubbles will represent different interventions.

Subgroup analysis

If data are available in the included systematic reviews, we will perform subgroup analyses according to:

• severity of acne (mild, moderate or severe), as we expect better outcomes for less severe cases;

• age of the participants, as we expect worst outcomes for older participants, with longer duration of the disease;

• ethnicity, as worse outcomes are possible for white adolescents, who may have greater tendency to more severe disease (Cheng 2010);

• gender, as we expect differences between women and men with regard to responsiveness to each systemic pharmacological intervention for acne (e.g. women, especially those ones with hormonal changes, may present worse outcomes when treated with oral isotretinoin and oral antibiotics) (Bagatin 2019; Bagatin 2020a).

We will use the formal Chi² test for subgroup differences to test for subgroup interactions, available in Review Manager 5 (Review Manager 2020)

We will analyse statistical diversity by checking the estimates of treatment effect within each subgroup. We will adopt the random‐effects model to pool data and produce forest plots using Review Manager 5 (Review Manager 2020). We will use the I² statistic and forest plot inspection to identify the percentage of total variation across subgroup results due to heterogeneity rather than chance (Higgins 2021). We will use the following thresholds to interpret the I² statistic:

• 0% to 40%: might not be important;

• 30% to 60%: may represent moderate heterogeneity;

• 50% to 90%: may represent substantial heterogeneity;

• 75% to 100%: considerable heterogeneity.

However, we acknowledge that the I² statistic can be misleading, as it depends on the magnitude and direction of effects and the strength of evidence for heterogeneity (e.g. P value from the Chi² test, or a confidence interval for I², as uncertainty in the I² value is substantial when the number of studies is small), therefore we will interpret it with caution (Higgins 2021).

Appendices

Appendix 1. Search strategy for the Cochrane Database of Systematic Reviews (the Cochrane Library)

#1 MeSH descriptor: [Acne Vulgaris] explode all trees
#2 acne:ti,ab
#3 #1 or #2

Appendix 2. Search strategy for Epistemonikos

We will use the Advanced Search facility. We will search for the term 'acne' in the title or abstract of a record, and limit results to Publication type 'systematic review' and Systematic Review Question 'interventions'.

Appendix 3. Search strategy for MEDLINE (Ovid)

1. exp Acne Vulgaris/
2. acne.ti,ab.
3. 1 or 2
4. meta analy$.tw.
5. metaanaly$.tw.
6. (systematic adj (review$1 or overview$1)).tw.
7. exp "Review Literature as Topic"/
8. Meta‐Analysis as Topic/
9. Meta‐Analysis/
10. systematic review.pt.
11. or/4‐10
12. cochrane.ab.
13. embase.ab.
14. (psychlit or psyclit).ab.
15. (psychinfo or psycinfo).ab.
16. (cinhal or cinahl).ab.
17. science citation index.ab.
18. bids.ab.
19. cancerlit.ab.
20. or/12‐19
21. reference list$.ab.
22. bibliograph$.ab.
23. hand‐search$.ab.
24. relevant journals.ab.
25. manual search$.ab.
26. or/21‐25
27. selection criteria.ab.
28. data extraction.ab.
29. 27 or 28
30. "Review"/
31. 29 and 30
32. Comment/
33. Letter/
34. Editorial/
35. Animals/
36. Humans/
37. 35 not (35 and 36)
38. or/32‐34,37
39. 11 or 20 or 26 or 31
40. 39 not 38
41. 3 and 40

Lines 4‐40: search filter designed by the Scottish Intercollegiate Guidelines Network (SIGN) to retrieve systematic reviews. The filter is available at https://www.sign.ac.uk/what-we-do/methodology/search-filters/. Line 10 added by the Cochrane Skin Information Specialist as this term has become available since the SIGN filter was developed.

Appendix 4. Search strategy for Embase (Ovid)

1. exp acne vulgaris/
2. acne.ti,ab.
3. 1 or 2
4. exp meta analysis/
5. ((meta adj analy$) or metaanalys$).tw.
6. (systematic adj (review$1 or overview$1)).tw.
7. or/4‐6
8. cancerlit.ab.
9. cochrane.ab.
10. embase.ab.
11. (psychlit or psyclit).ab.
12. (psychinfo or psycinfo).ab.
13. (cinahl or cinhal).ab.
14. science citation index.ab.
15. bids.ab.
16. or/8‐15
17. reference lists.ab.
18. bibliograph$.ab.
19. hand‐search$.ab.
20. manual search$.ab.
21. relevant journals.ab.
22. or/17‐21
23. data extraction.ab.
24. selection criteria.ab.
25. 23 or 24
26. review.pt.
27. 25 and 26
28. letter.pt.
29. editorial.pt.
30. animal/
31. human/
32. 30 not (30 and 31)
33. or/28‐29,32
34. 7 or 16 or 22 or 27
35. 34 not 33
36. 3 and 35

Lines 4‐35: search filter designed by the Scottish Intercollegiate Guidelines Network (SIGN) to retrieve systematic reviews. The filter is available at www.sign.ac.uk/what-we-do/methodology/search-filters/.

Contributions of authors

CSC was the contact person with the editorial base.
CSC co‐ordinated the contributions from the co‐authors and wrote the final draft of the protocol.
CSC, RR, RLP, ZY, EB, PM, and TP worked on the Methods section.
CSC, EB, PM, RR, RLP and ZY drafted the clinical sections of the Background and responded to clinical comments of the referees.
CSC responded to methodology and statistics comments of the referees.
CSC, EB, RR, RLP, PM and ZY contributed to writing of the protocol.
LSUS was the consumer co‐author and checked the protocol for readability and clarity. She also ensured that the outcomes are relevant to consumers.

Disclaimer

This project was supported by the National Institute for Health Research (NIHR), via Cochrane Infrastructure funding to Cochrane Skin. The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, NHS, or the Department of Health.

Sources of support

Internal sources

• No sources of support provided

External sources

• The National Institute for Health Research (NIHR), UK

  The NIHR, UK, is the largest single funder of Cochrane Skin

Declarations of interest

Caroline S Costa: has published review articles in two medical journals (Expert Review of Clinical Pharmacology and Anais Brasileiros de Dermatologia) on the use of oral isotretinoin for acne (with Ediléia Bagatin). CSC works as a Dermatologist at a private clinical practice in Unimed Teresina, Piaui, Brazil, which includes seeing acne patients daily and prescribing systemic pharmacological treatments when necessary.

Ediléia Bagatin: declares financial activities with LeoPharma Inc. Denmark (Consultancy, Advisory Board, Speaker and Investigator for studies and research on adult female acne, personal payments); Greencare, Brazil (Consultant, Advisory Board, Speaker and Investigator for cannabis in dermatology, personal payment); Douglas Pharmaceuticals, New Zealand (Consultancy about oral isotretinoin, personal payment); USK (Under Skin), Brazil (Advisory Board and Speaker on acne vulgaris, oral isotretinoin, and antibiotics for acne, personal payment); and Theraskin, Brazil (Speaker on acne vulgaris, personal payment).

Zhirong Yang: declares that they have no conflict of interest.

Rafael L Pacheco: declares that they have no conflict of interest.

Parker Magin: declares that they have no conflict of interest.

Luiza de Sá Urtiga Santos: works as a family medicine physician, employed by the City Hall of São Miguel do Tapuio, Brazil, but declares that they do not have any direct activity related to acne treatments other than referring patients to specialist care.

Tiago Pereira: declares that they have no conflict of interest.

Rachel Riera: declares that they have no conflict of interest.

New