A Review of hormone-based therapies to treat adult acne vulgaris in women.

Hormone-based therapies can be separated into two broad categories: androgen synthesis inhibitors and androgen receptor antagonists. Estrogen and progesterone derivatives, which make up the components of COC medications, are generally considered androgen synthesis inhibitors. Commonly known androgen receptor antagonists include agents such as spironolactone, flutamide, cyproterone acetate, and progestins alone ( Trifu et al., 2011 ). We will consider COC medications, spironolactone, and flutamide separately with regard to their use to treat patients with adult acne. We will briefly discuss the use of metformin although its use to treat adult acne is not formally reviewed in this paper. Oral contraceptive medications are thought to primarily exert anti-androgen effects through the actions of estrogen. Estrogen is known to stimulate the hepatic synthesis of sex hormone and bind globulin, which binds androgens and decreases levels of free testosterone, with dihydroepiandrosterone sulfate. Estrogen also inhibits 5-alpha reductase, which prevents the conversion of testosterone to the more potent dihydrotestosterone. Ovarian and adrenal androgen synthesis is also reduced due to the effects of estrogen and its negative feedback on gonadotropin, which releases hormone and leutinizing hormone ( Haider and Shaw, 2004 ). Progestins have a lesser role in the production of anti-androgenic activity because early generation progestins have androgenic properties. When choosing oral contraceptive pills (OCPs) for the treatment of acne, third generation progestins (e.g., norgestimate or desogestrel) or later generations (e.g., fourth or fifth generation progestin-containing OCP formulations) are preferred because they have a lower androgenic activity overall. Progestins may contribute to the anti-androgenic properties of combined oral contraceptive medications through the reduction of gonadotropin-releasing hormone pulsatility and therefore luteinizing hormone production ( Arowojolu et al., 2012 ). Although numerous formulations of oral contraceptive medications exist, the U.S. Food and Drug Administration (FDA) has only approved three medications to date for the treatment of acne vulgaris: Ortho Tri-Cyclen (combination of norgestimate and ethinyl estradiol [EE]), Estrostep (combination of norethindrone acetate and EE), and Yaz (combination of drospirenone and EE; Ebede et al., 2009 ). A number of clinical trials on the safety and efficacy of different COC formulations to treat acne vulgaris have been published to date. Table 1 consolidates this evidence by COC formulation, duration of study, type of study, dosages used, and published results. To date, the superiority of particular OCP formulations for the treatment of acne have not been clearly demonstrated. Multiple COC formulations have been studied in the context of treating adult women with acne vulgaris. All combinations that are reported in Table 1 exhibit some degree of efficacy in the reduction of acne lesion count. In terms of adverse events in patients with acne who were treated with COC medications, the most commonly reported events include headache, menstrual cycle irregularities, and emotional lability. Three of the studies reported severe adverse events related to COC use, with two cases of clinical depression and one case of ovarian cyst formation. However, the most common side effects that were reported in these studies are well known and expected for low-dose combined oral contraceptive medications. None of the studies reported above demonstrated any unexpected adverse events. In terms of comparing efficacies, desogestrel may be slightly more effective than levonorgestrel COC formulations. However, one of the three comparison studies reported that the two have equal efficacy. The chlormadinone acetate (CMA) formulation may be more efficacious than levenorgestrel. However, CMA has not been marketed in the United States since 1972 after reports of mammary gland nodule formation in a study using canines. Several European countries, however, still use CMA COC formulations ( Lingeman, 2012 ). All other comparison studies revealed results of non-inferiority. Studies that compared COC formulations head-to-head and had relevant results are listed in Table 2 . A meta-analysis of 24 randomized trials that was conducted by Arowojolu et al., 2012 demonstrated no consistent differences in acne reduction between different combined OCPs ( Arowojolu et al., 2012 ).

Hormonal therapy is often considered to treat acne vulgaris in the settings of hyperandrogenism, late-onset acne (> 25 years of age), jawline acne distribution, acne with menstrual flare, comedonal acne with seborrhea, and acne that is resistant to conventional therapies. Prior to the initiation of hormone therapy, expectations about long-term systemic treatment should be set in terms of family planning. Additionally, an endocrinologic evaluation should be prompted by certain clinical signs and symptoms including hirsutism, androgenetic alopecia, virilization, acanthosis nigricans, central obesity, oligo- or amenorrhea, infertility, sudden and/or severe onset acne, and acne that is significantly resistant to conventional therapy ( Zaenglein et al., 2016 ). It should also be noted that antiandrogen therapy should always be used with an effective contraceptive method if the patient can get pregnant ( Fauser et al., 2012 ). Hormonal therapy to treat acne may take longer to show results compared with other acne treatments and often requires 3 months to 6 months of treatment for efficacy ( Koo et al., 2014 ). Therefore, it is important to set appropriate expectations for patients prior to treatment initiation.

A systematic search of the PubMed Database was conducted with the following search terms: contraceptives, hormone therapy, spironolactone, flutamide, and acne treatment. The terms polycystic ovarian syndrome, endometriosis, and hidradenitis supparativa were excluded from the search. Filters included human studies that were published after January 1, 2000. A total of 236 results were then manually filtered with a systematic approach to exclude review papers, unrelated topics, commentaries, and articles that were published in a language other than English, which yielded 36 relevant studies.

Flutamide is a selective androgen receptor antagonist, which is a well-known hormone-based chemotherapeutic agent to treat patients with prostate cancer. This agent has been purported as a potential therapy for adult onset acne vulgaris due to its antiandrogenic properties. However, very few studies on the efficacy of flutamide in acne treatment have been published thus far. A 2002 study conducted by Carmina and Lobo demonstrated the equal and significant efficacy of flutamide and a COC formulation (EE + cyproterone acetate) in the reduction of acne Cook scores in 48 hyperandrogenic women ( Carmina and Lobo, 2002 ). However, the study did not directly assess the effects of flutamide in nonhyperandrogenic women who make up a significant portion of patients with adult acne. Paradisi et al. (2011) published a retrospective study on the effects of low dose flutamide in patients with acne who were over 15 years of age. More than 97% of patients reported satisfaction with the control of their acne on low-dose flutamide regimens over a 6-year period ( Paradisi et al., 2011 ). However, flutamide has been used cautiously by clinicians because of potential hepatotoxic effects. In the previous study, 5% of patients experienced significant increases in transaminase levels. The androgen excess group of the American Society for Reproductive Medicine noted that flutamide had very limited value in the treatment of hyperandrogenism because of its hepatotoxic effects ( Fauser et al., 2012 ).

Interestingly, this well known drug that has been used to treat patients with diabetes for years, has recently gained attention as a potential treatment for hormonal acne in patients with polycystic ovarian syndrome (PCOS). Levels of facial sebum production are positively correlated with levels of the hormone insulin-like growth factor 1 (IGF-1), which is already known to be correlated with acne severity in women. IGF-1 is known to stimulate androgenic hormone production, which may explain this phenomenon ( Vora et al., 2008 ). This finding is independent of a diagnosis of PCOS and suggests that targeting hyperinsulinemia and IGF-1–mediated metabolic changes may be potentially beneficial to patients outside of the PCOS population. Metformin has been shown to improve acne in several studies in patients with PCOS and hyperinsulinemia. It is thought to act by increasing insulin sensitivity and thereby reducing levels of IGF-1 and its downstream effects ( Bubna, 2016 ).

Hormone therapy is a safe and effective option for women who suffer from adult acne. Most COC formulations demonstrate a certain level of efficacy in the treatment of patients with acne. Large meta-analyses that compared the different COC formulations do not demonstrate significant differences in efficacy between the formulations. Limitations of this review of current studies include difficulty to meaningfully compare the trials that are included because outcome measures varied across the studies. Lesion counts, sebum production, and investigator and patient-assessments were a few of the efficacy measures that were reported. Also, none of the trials that were reported were against active comparators. Most studies did not use follicle-stimulating and/or luteinizing hormone ratios or measurements of insulin resistance as exclusion criteria for screening. The most common side effects that were reported throughout the literature with the use of COC formulations included headache, menstrual irregularities, nausea, and breast tenderness, which are all common side effects that are expected with the use of low-dose COC formulations. Despite the relative safety of COC formulations, discussions on risks and side effects with individual patients is necessary to provide the safest and most beneficial care. Independent cardiovascular risk factors such as smoking, hypertension, and diabetes should be addressed. When discussing clotting risk, the increased risk of thrombophilia should be conveyed but contextualized because pregnancy confers a significantly higher risk of clotting. Although the FDA has issued warning labels for increased clotting risk with drosperinone-containing OCPs, a recent large prospective trial has shown that there is no increased clotting risk compared with other OCP formulations ( Dinger et al., 2016 ). It is important to know the general risks, side effects, and contraindications of COC formulations prior to prescribing them as an acne treatment. It is also important to inform patients that COC therapy improves acne over time. Most studies showed some improvement at 3 months of treatment and the greatest levels of improvement at 6 months. Oral spironolactone is a highly effective and safe option to treat patients with acne and contrary to prior practice potassium monitoring is not required in patients without a history of prior renal disease. Spironolactone may provide additional benefit in the treatment of patients with acne when used together with OCP. It is also a lower cost option for patients who cannot afford expensive oral isotretinoin therapy. Overall, anti-androgen therapy represents a safe and effective option for the treatment of patients with adult acne vulgaris and therapy should be prescribed on the basis of patient preference, cost, and individual patient characteristics because differences in effectiveness do not drastically vary from one formulation to another.

COC formulations can provide significant relief to patients who suffer from severe acne vulgaris, especially when conventional therapy has failed. However, to achieve maximum benefits and safety for each individual patient, counseling and discussion must be pursued during the initial visit. Contrary to previous practice, pelvic examinations and Pap smears are not required prior to prescribing COC medications according to recommendations by the World Health Organization, the American Congress of Obstetricians and Gynecologists (ACOG), and Planned Parenthood. This makes it more feasible for dermatologists to prescribe COC medications during routine clinic visits ( Frangos et al., 2008 ). Patients commonly ask about cardiovascular, clotting, and cancer risks that are associated with oral contraceptive use. These issues are addressed separately below. When counseling patients about cardiovascular risk, it is important to address modifiable risk factors such as smoking, hypertension, and diabetes and how control of these factors are the most important in reducing long-term risks of myocardial infarction and stroke. The risk of clotting should be discussed in context of baseline risks for women in the general population. The risk of deep vein thrombosis in women at baseline is 1/10,000 woman-years; at 1 year of COC use 3.4/10,000 woman-years; and during pregnancy 5 to 12/10,000. Peragallo Urrutia et al. (2013) conducted a meta-analysis of 14 different studies in which OCP users had three times increased odds for venous thromboembolism (VTE) compared with non-users. Interestingly, older progestins seem to confer a lower VTE risk than newer progestin-based COC formulations. In 2012, drospirenone-based COC formulations were FDA-labeled with a warning for an increased risk of clotting ( Lidegaard et al., 2012 , Manzoli et al., 2012 ). However, this was contradicted when the results of a prospective controlled trial entitled “Long-term active surveillance study for oral contraceptives” were published in 2016 and demonstrated that drosperinone-based OCPs did not pose a higher risk for VTE compared with other OCPs ( Dinger et al., 2016 ). In summary, there is a slightly increased risk of DVT in COC users but much less than that of the thrombophilic state of pregnancy. In 1996, the Collaborative Group on Hormonal Factors in Breast Cancer conducted a meta-analysis of 54 different epidemiological studies in which data from 53,297 women with breast cancer and 100,239 women without breast cancer was analyzed. The findings suggested a slightly increased risk of breast cancer in women who take a COC formulation (relative risk = 1.24, 95% confidence interval [CI]). This risk was increased up to 10 years after discontinuing birth control but did not extend past this time frame. Notably, women who were diagnosed with breast cancer and had taken COC medications had clinically less advanced tumors compared with women who were diagnosed with breast cancer and had not taken COC formulations ( Collaborative Group on Hormonal Factors in Breast Cancer, 1996 ). The more recent 2002 Women’s CARE study in which 4575 women were interviewed demonstrated contrary to the previously mentioned pooled analysis that there was no significant increase in the risk of breast cancer in COC users between the ages of 35 and 64 years (relative risk = 1.0). Similar to the pooled analysis study, the risk of breast cancer was not significantly related to the dose of estrogen or duration of COC use ( Marchbanks et al., 2002 ). However, a 36-year follow-up in the Nurse’s Health Study, which aimed to assess the risk of breast cancer with COC use, reported a higher risk of premature deaths due to breast cancer in the population that used COC formulations ( p < .0001). The authors hypothesized that this higher risk may be attributable to higher hormone doses that were used in earlier COC formulations ( Charlton et al., 2014 ). The risk of breast cancer with COC use is still controversial because these major studies have inconsistent outcomes. When discussing the risk with patients, it is important to highlight that the long-term risks remain unclear at this time. Newer formulations of OCPs use 5- to 10-fold less hormone than the original formulations so more data is needed to determine the true risk based on today’s available formulations. Providers should prescribe COC medications with an understanding of each patient’s unique history and other independent risk factors such as gene mutations, family history, age, and obesity. It is also important to discuss the reduced overall risk of endometrial and ovarian cancer ( Collaborative Group on Hormonal Factors in Breast Cancer, 1996 , Frangos et al., 2008 , George et al., 2008 , Haider and Shaw, 2004 ). The follow-up on the Nurse’s Health Study mentioned earlier also reported a decrease in mortality rates from ovarian cancer in patients who used COC formulations ( p = .002; Charlton et al., 2014 ). A 2008 meta-analysis of 45 different studies reported a significant correlation between duration of COC use and reduction in risk of ovarian cancer ( p < .0001; Collaborative Group on Epidemiological Studies of Ovarian Cancer, 2008 ). The risk of endometrial cancer is also reduced, theoretically due to the suppression of endometrial growth through contraceptive cycle control. A 1999 Swedish case-control study noted that COC use decreased the risk of endometrial cancer by 30% and that this reduction was noticeable after 3 or more years of use ( Weiderpass et al., 1999 ). Other important questions include the risk of weight gain, common side effects, rare side effects, and use of OCPs with antibiotic medications. In 30% of patients, 1 kg to 2 kg of weight gain occur mostly due to fluid retention. Other common side effects include unscheduled bleeding, nausea, and breast tenderness. These symptoms except for unscheduled bleeding are lessened with decreased estrogen doses. Less common side effects include decreased libido, melasma, and mood changes ( Junkins-Hopkins, 2010 ). It is important to discuss the interactions of antibiotic medications with OCPs especially in women who utilize OCPs as their primary means of birth control. Since many antibiotic medications are inducers of 3A4, practitioners are concerned that OCP efficacy can be reduced with concurrent antibiotic medication use. However, according to the 2006 ACOG Practice Bulletin, pharmacokinetic evidence of reduced steroid levels with antibiotic medication use only exists for rifampin. Studies have failed to demonstrate similar findings for doxycycline, ampicillin, metronidazole, tetracycline, or quinolones ( Kaunitz, 2006 ). Important contraindications to consider prior to the initiation of COC therapy include cardiovascular risk factors, severe hypertension, history of stroke, VTE or myocardial infarction, smoking combined with age > 35 years, history of migraine with focal aura, history of migraine combined with age > 35 years, current or past history of breast cancer or endometrial cancer, cholestatic jaundice of pregnancy (or jaundice with previous OCP use), diabetes with complications, hepatic neoplasias, abnormal liver function, hypersensitivity to OCPs, pregnancy, and major surgery with prolonged immobilization ( Bonnema et al., 2010 , Frangos et al., 2008 ).

Acne in women over the age of 25 years, also known as adult-onset or persistent acne, affects approximately 12 to 22% of women in the United States. Historically, acne vulgaris has been primarily associated with the adolescent population. However, the burden of disease in the adult population has been highlighted in recent literature and women are noted to have greater incidence of disease compared with their male counterparts ( Tanghetti et al., 2014 ). It should be noted that a number of adult women with acne may have underlying polycystic ovarian syndrome (PCOS) because approximately 70 to 80% of patients with PCOS suffer from cutaneous manifestations of the disease ( Schmidt et al., 2015 ). Adult acne in women is often associated with anxiety, depression, and a reduced quality of life ( Tanghetti et al., 2014 ). Furthermore, this condition is characterized by high rates of treatment failure. In one prospective study of 200 women over the age of 25 years, approximately 80% of women failed multiple courses of systemic antibiotic medications and approximately 30% of patients relapsed after several therapeutic cycles of isotretinoin ( Goulden et al., 1997 ). The high rates of treatment failure with traditional therapies along with more consciousness about antibiotic stewardship in dermatology patients, many of whom are on systemic antibiotic therapy for acne treatment, have motivated clinicians to reconsider the therapeutic targets of treatment in this population ( Dreno et al., 2014 ). There is considerable evidence on the role of sex hormones and specifically relative to androgen excess in the pathophysiology of acne vulgaris. For example, the development of acne in pre-pubertal girls has been correlated with high levels of dihydroepiandrosterone sulfate. Polycystic ovarian syndrome in which women experience hyperandrogenism is also associated with an increased prevalence of acne vulgaris. Individuals who are androgen insensitive do not experience acne. Most importantly and perhaps the strongest evidence is the effective use of combined oral contraceptive medications and anti-androgen therapies to treat women for acne. Hormonal therapy has been shown to be effective in postmenarchal adult women (age > 14 years) even in those women with normal androgen levels ( Harper, 2008 , Lolis et al., 2009 , Lucky et al., 1991 , Lucky et al., 1994 , Lucky et al., 1997 ). A myriad of combined oral contraceptive (COC) formulations and anti-androgen therapies exist and have been used in the clinical setting for the treatment of acne vulgaris with various efficacies and side effect profiles. The main objective of this systematic review is to provide a broad overview of currently available therapies and their efficacies, risks, side effects, and contraindications as reported by controlled clinical trials to date. Important counseling points for patients prior to the initiation of hormonal therapy and other types of contraception and their implications for acne are discussed.

Spironolactone is an aldosterone antagonist that is often used as a diuretic and provides mortality benefit for patients with severe congestive heart failure. However, in higher doses, this agent also exhibits anti-androgen effects through inhibition of the cytochrome p450 system, inhibition of 5 alpha-reductase activity, and increase of hepatic synthesis of sex hormone-binding globulin ( Armanini et al., 2016 ). Although not an FDA-approved medication for the treatment of acne, clinicians have often used this agent when there is inadequate acne control in patients who are already treated with a COC, have a contraindication to COC formulations, have coexisting hypertension, or their economic barriers prevent the purchase of other acne-controlling medications. The usual dosing of this medication is 50 to 200 mg/day. Contraindications to the use of spironolactone include renal insufficiency, hyperkalemia, pregnancy, or unevaluated abnormal uterine bleeding ( Zaenglein et al., 2016 ). There have been no randomized prospective placebo controlled trials on the efficacy of spironolactone in the treatment of adult-onset acne vulgaris. However, a few observational and retrospective analyses have been conducted in the last decade. These studies demonstrate excellent rates of efficacy for oral spironolactone to treat female adult patients with acne. The treatment doses ranged from 100 to 200 mg/day. Sato et al. (2006) conducted a study in which 139 Japanese patients were treated with a 20-week tapered regimen of oral spironolactone. A total of 64 female patients completed the study and there was a 100% response rate with approximately 50% of patients who had an excellent response. A total of 23 male patients were also initially included in the study but treatment was stopped after several men developed gynecomastia ( Sato et al., 2006 ). Yemisci et al. (2005) performed a similar study and demonstrated an 85% response rate with significant improvements in acne for responders. In another study on the effects of treatment with topical spironolactone, 78 patients were enrolled in a randomized, double-blind, placebo controlled trial. The effects of 5% topical spironolactone on acne were mixed with a significant decrease in total acne lesion count but no difference in acne severity index between the two groups ( Afzali et al., 2012 ). In conclusion, on the basis of initial studies, oral spironolactone demonstrates adequate levels of efficacy especially for patients with recalcitrant and/or severe forms of acne and those who have failed with COC therapy. Spironolactone has also been studied in combination with other acne therapies including topical retinoids and COC formulations that contain drosperinone ( Krunic et al., 2008 , Lessner et al., 2014 ). Both studies demonstrated that combination treatment with these agents is both efficacious and well-tolerated. Although spironolactone has been considered generally safe, there are certain common side effects including menstrual irregularities, breast tenderness, fatigue, and headache ( Shaw and White, 2002 ). Hyperkalemia and the necessity of potassium level monitoring has been a subject of controversy in previous years. However, in 2015, Plovanich et al. published a landmark paper with the results of a retrospective analysis of hyperkalemia rates in women who were administered spironolactone from 2000 to 2014. The researchers concluded that routine potassium monitoring in healthy women who take this medication is unnecessary. Healthy women were defined as women ages 18 to 45 years who were diagnosed with acne, with the exclusion of women with heart failure, on medication that affects the renin-angiotensin-aldosterone pathway, or who were diagnosed with renal disease ( Plovanich et al., 2015 ). However, physicians should screen patients for bizarre diets that are high in potassium-containing substances such as coconut water prior to the initiation of spironolactone therapy. Other safety considerations have included the effect of spironolactone on the body’s thrombotic state. Unlike COC formulations, spironolactone seems to improve the thrombotic profile because aldosterone is known to be prothrombotic. Two individual case studies report the successful use of spironolactone-based treatment for Budd-Chiari syndrome in one patient with protein C deficiency and portal vein thrombosis in another patient with hepatitis B-related chronic liver disease ( Hiroe et al., 2008 , Kumar et al., 2011 , Struthers and MacDonald, 2004 ). At very high doses, spironolactone has also been associated with benign tumor formation in studies on animals. In a study on rats that were administered spironolactone at doses of 50, 150, and 500 mg/kg/day, increases in benign adenomas of the thyroid and testes were reported after 18 months. Similar doses in another study of rates led to the formation of hepatocellular adenomas in the subjects after 24 months, which caused the FDA to issue a black box warning (FDA, Aldactone Black Box Warning, revised 2008). However, long-term safety data does not support tumorigenic potential in humans ( Danielson et al., 1982 , Friedman and Ury, 1980 , Shaw and White, 2002 ).