In a small group of women, the menstrual cycle has been associated with a spectrum of dermatologic diseases including eczema, erythema multiforme, stomatitis, papulopustular lesions, folliculitis, angioedema, urticaria, and others (Table 1) [1–8]. As progesterone sensitivity has been the most commonly identified cause, dermatologic diseases associated with the menstrual cycle have been labeled autoimmune progesterone dermatitis (APD) . The first documented case of APD was in 1921, in which a patient's premenstrual serum caused acute urticarial lesions. In addition, it was shown that the patient's premenstrual serum could be used to desensitize and improve her symptoms . Since 1921, approximately 50 cases of APD have been published in the medical literature.
The clinical symptoms of APD (eczema, urticaria, angioedema, etc.) usually begin 3–10 days prior to the onset of menstrual flow, and end 1–2 days into menses. Severity of symptoms can vary from nearly undetectable to anaphylactic in nature, and symptoms can be progressive [10, 11]. There are no specific histological features on biopsy in APD . The age of onset is variable, with the earliest age reported at menarche . Some studies have noted that a majority of patients had taken an oral contraceptive (OCP) prior to the onset of APD , but multiple cases exist in which women have never been exposed to exogenous progesterone [15–17].
The symptoms of APD correlate with progesterone levels during the luteal phase of the menstrual cycle. Progesterone begins to rise 14 days prior to the onset of menses, peaks 7 days prior to menses, and returns to a low baseline level 1–2 days after menses begins. In studies where an etiologic agent has been sought, progesterone has been found most frequently. However, estrogen, prostacyclin, and gonadotropin levels have correlated with symptoms in some cases [18–21].
Symptoms may first appear, improve, or worsen during pregnancy and the peripartum period [2, 22–24]. In addition, APD during pregnancy has been associated with spontaneous abortions [2, 25]. Pregnancy is associated with an increase of maternal progesterone levels, which may explain the initiation or worsening of symptoms. In regards to an improvement of symptoms during pregnancy, a number of theories have emerged. Explanations include a slow rise of progesterone during pregnancy that acts as a method of desensitization, a decrease in maternal immune response during pregnancy, or an increased production of anti-inflammatory glucocorticoids [13, 25, 26].
The exact pathogenesis of APD is unknown. If exogenous progesterones (i.e. OCPs) are initially used, it is conceivable that uptake by antigen presenting cells and presentation to TH2 cells could result in subsequent IgE synthesis; however this mechanism would not explain the pathogenesis in patients such as ours who have the onset of APD prior to exogenous progesterone exposure. Some authors have suggested that hydrocortisone or 17-α-hydroxyprogesterone have cross-sensitivity with progesterone and may cause initial sensitization, but this has not been observed in all studies [27, 28].
To further delineate the pathogenesis, antibodies against progesterone have been investigated. Using immunofluorescent techniques and basophil degranulation tests, studies have found that such antibodies do exist in certain patients with APD [1, 13, 29]. However, negative results looking for antibodies have also been reported . In addition, skin test results with progesterone have shown immediate reactions (within 30 minutes), delayed reactions (24–48 hours later), and reactions with features of both immediate and delayed features [13, 14, 30, 31]. This presumably indicates both type I and type IV hypersensitivity reactions. Progesterone has also been reported to have a direct histamine releasing effect on mast cells, yet very little research has been done to support this hypothesis . Additionally, one study found an in vitro increase of an interferon-γ release assay, possibly implying a role for TH1-type cytokines in APD .
Eosinophils may also be involved in the pathogenesis of APD. Eosinophilia has been correlated with cutaneous symptoms in some cases, and studies have found a decrease in total eosinophil count after therapy [13, 29, 34]. Whether increased eosinophils are a response to cytokines from lymphocytes or play a primary mechanistic role in APD remains to be determined.
The diagnosis of APD requires an appropriate clinical history accompanied by an intradermal injection test with progesterone. An aqueous suspension or aqueous alcohol solution of progesterone is the preferable vehicle of testing as progesterone in oil can cause an irritant reaction , though many published case reports have used progesterone in oil for testing. Various authors have advocated different amounts of progesterone or medroxyprogesterone to be used for testing [12, 33, 36]. As had been done in some prior studies, the patient presented here was tested with progesterone in aqueous solution at a concentration of 50 mg/mL.
As mentioned above, APD may be due to an immediate or delayed hypersensitivity reaction. Therefore, intradermal testing may not become positive until 24–48 hours later [14, 24]. In addition, some authors have advocated patch testing with progesterone to further evaluate for a hypersensitivity reaction . Of note, intradermal testing has been negative in some patients with typical clinical symptoms of APD and who improved after APD treatment [2, 3, 24].
Some authors have recommended further tests to evaluate the immunologic evidence in APD. These include circulating antibodies to progesterone, basophil granulation tests, direct and indirect immunofluorescence to luteinizing cells of the corpus luteum, in vitro interferon-γ release, and circulating antibodies to 17-α-hydroxyprogesterone [1, 7, 13, 29, 33, 36]. However, most case reports in the medical literature do not routinely check for serologic evidence of APD, and when checked these markers have not always been found to be reliable. This is most likely due to the fact that, as mentioned above, the pathogenesis of APD is incompletely understood.
Autoimmune progesterone dermatitis is usually resistant to conventional therapy such as antihistamines. The use of systemic glucocorticoids, usually in high doses, has been reported to control the cutaneous lesions of APD is some studies, but not in others [3, 10, 37]. Early reports of APD describe attempts of progesterone desensitization, and some authors even attempted injections derived from the corpus luteum [18, 24, 38]. However, results were usually temporary, and such methods of treatment have now fallen out of favor.
Current therapeutic modalities often attempt to inhibit the secretion of endogenous progesterone by the suppression of ovulation. Table 2 lists some of the pharmacologic strategies used in APD. Oral contraceptives (OCPs) are often tried as initial therapy, but have had limited success, possibly due to the fact that virtually all OCPs have a progesterone component. Conjugated estrogens have also been used in the treatment of APD. These did show improvement in many of the patients, but often required high doses [2, 16, 22]. However, due to the increased risk of endometrial carcinoma with unopposed conjugated estrogens, this treatment is not commonly used today .
Various other therapy modalities are currently used in APD, and there is no clear treatment of choice. GnRH agonists, such as buserelin and triptorelin, have been used to induce remission of symptoms by causing ovarian suppression [7, 11, 15]. However, side effects include symptoms of estrogen deficiency (hot flashes, vaginal dryness, decreased bone mineral density), and estrogen supplementation may be needed . Alkaylated steroids such as stanozol have been used to successfully suppress ovulation, sometimes in combination with chronic low doses of corticosteroids . Side effects of alkaylated steroids include abnormal facial or body hair growth, hepatic dysfunction, and mood disorders, any of which may limit their use. To decrease the risk of side effects, some authors have recommended using the alkaylated steroid only in the perimenstrual period . Another therapeutic option used in APD has been the antiestrogen tamoxifen, which also can suppress ovulation [3, 5]. As with GnRH agonists, patients on tamoxifen may experience symptoms of estrogen deficiency. In addition, tamoxifen has been associated with an increased risk of venous thrombosis and cataract formation. In some patients with unremitting symptoms of APD, bilateral oopherectomy has been required [10, 15, 24]. While this definitive treatment has been successful in controlling symptoms, today it is rarely used before all medical options have been exhausted.