全文

Introduction

Vaginal symptoms suggestive of vaginitis such as vaginal itching, discharge, and dyspareunia are common reasons for women to visit gynecological clinics [1]. In spite of the high prevalence of such complaints, one of three patients will be undiagnosed [2]. Desquamative inflammatory vaginitis (DIV) is an uncommon severe form of chronic purulent vaginitis causing discharge, vestibulo-vaginal irritation, and dyspareunia ∗[3] and ∗[4]. It occurs mainly in Caucasians and although diagnosed in a wide age range, its occurrence peaks in perimenopausal women ∗[4], ∗[5] and ∗[6]. Diagnosis is based on a detailed medical history, physical examination, and wet mount (office-based microscopy). The symptoms and signs are nonspecific and may require the use of vaginal culture, polymerase chain reaction (PCR), rare use of blood tests (hormone levels), and infrequent histology only to exclude other causes of vaginal inflammation. Unfortunately, such thorough evaluation does not often occur in primary gynecological clinics, and the ability to estimate the true prevalence of DIV in the general population is limited. By contrast, specialized studies in vulvovaginal clinics estimate the incidence of DIV as 0.8–4.3% of referred cases. This incidence likely reflects a referral or accrual bias ∗[5], ∗[6] and [7]. The etiology is unknown. There are only 29 published articles in the English literature (as of August 2013) of which seven (24%) are case reports and series and eight (28%) are reviews. The majority of the original studies (10 all together) are mainly retrospective descriptive studies. (Table 1).

 

 

History

The term “desquamative inflammatory vaginitis (DIV)” was first introduced in 1965 when Gray and Barnes described six women from a group of 478 consecutive patients with vaginal complaints who had a “reddened” vagina and “numerous pus cells … with oval and round parabasal cells.” Cultures were sent for all six patients and two were positive for Trichomonas vaginalis. [7] The authors concluded that the other four had an “interesting form of vaginitis … seems to represent a clinical entity … and the true nature is not clear.” [7] The definition was refined only 3 years later when Gardner published his milestone case series of eight patients titled “Desquamative inflammatory vaginitis: a newly defined entity.” [3] All eight patients presented with purulent discharge and demonstrated ecchymotic vaginal spotting and desquamation of vaginal walls. A microscopy evaluation of vaginal discharge showed increased inflammatory cells with excess of parabasal cells and lack of lactobacilli. Vaginal pH was universally elevated. Notably, there was poor response to antimicrobial treatment and local estrogens were ineffective. Gardner concluded that this was a rare condition given that only eight of the 3000 vaginitis patients he evaluated in 15 years matched the description [3]. Once DIV was characterized by Gardner, in retrospect it was possible to recognize case reports describing DIV – although calling the condition by a variety of descriptive terms, as was the case published by Scheffey in 1956 (12 years before Gardner). Scheffey described a 50-year-old women with 14 months of copious vaginal discharge with severe inflammation of vagina and increase of inflammatory and parabasal cells. Scheffey called the condition “exudative vaginitis” – this is probably the first case report describing a patient with DIV [8]. Others attribute the earliest description of DIV to Franken's report in 1956 of a 12-year-old girl with exudative vaginitis who responded well to estrogen, yet as we consider today such a response to treatment excludes the diagnosis of DIV [9] and [10]. During the 30 years following Gardner's publication, there was one letter to the editor addressing Gardner's publication and one review of DIV describing the case reports presented above including cases of erosive lichen planus and vulvo-vaginal–gingival syndrome [9] and [11]. After a gap of four decades from the first case report, a retrospective study of 51 patients diagnosed with DIV based on Gardner's characteristics was published [4]. The study described the short-term follow-up of 51 women with DIV with successful treatment utilizing topical 2% clindamycin. This suggested a bacterial etiology. This assumption changed 17 years later by the same author ∗[4] and ∗[5]. In the last decade, approximately eight new original publications regarding DIV were published, however with only minor contributions to etiology and prognosis ∗[5], ∗[6], ∗[12], ∗[13], [14], ∗[15], ∗[16] and [17].

 

Case definition

Definition

DIV is a clinical syndrome of severe chronic purulent vaginitis. Unfortunately, the case definition is based on symptoms, signs, and laboratory findings that are nonspecific (Table 2). Exclusion of other etiologies causing purulent vaginitis is essential for confirming diagnosis (Fig. 1).

 

 

 

Symptoms and signs

DIV is a chronic condition. Most patients will have complaints for more than a year before being diagnosed with a mean duration of symptoms of 15–31 months ∗[4] and ∗[5]. Patients are typically symptomatic although asymptomatic DIV occasionally occurs, the frequency of which is unknown. Approximately 90% of patients will complain of purulent discharge, severe dyspareunia, and vaginal discomfort ∗[4] and ∗[5].

 

A purulent discharge is present which ranges from mild to profuse. Based on case definition, signs of vaginal inflammation need to be present, at least focal as petechiae, or diffuse vaginal erythema. A typical spotted vaginal rash (petechial or ecchymotic) is noticed in 30–70% of cases. Occasionally, annular lesions described as erythematous papules with a pale center resembling donuts are seen ∗[4], ∗[5] and [14]. This could involve the cervix with the appearance of colpitis macularis in up to 27% of patients [4]. Van der Meijden and Ewing described this entity as papular colpitis, the biopsy of which consistently showed dense lymphocytic infiltrates [18]. The vestibule frequently shows signs of focal or diffuse erythema [5]. Dermatologic changes may be evident in the most severe cases with a symmetrical erythematous macular vulvar rash. In contrast to erosive lichen planus, vaginal adhesions, synechiae, and stenosis are extremely rare and might suggest incorrect diagnosis.

 

Wet mount and pH

There is a marked increase in inflammatory cells (a ratio of inflammatory cells to epithelial cells >1:1), predominantly polymorphonuclear leukocytes (PMNs), together with a mandatory increase in parabasal epithelial cells (immature squamous cells). Vaginal flora is abnormal with the loss of dominant lactobacillus morphotype and pH is always elevated above 4.5 ∗[3], ∗[4], ∗[5] and [10].

 

Diagnosis

The case definition of DIV is unfortunately not specific and the exclusion of other conditions that cause purulent vaginitis is essential; see Table 3∗[3], ∗[4] and ∗[6]. Clinicians need to be adequately trained to use the microscope for evaluating vaginal discharge. The response to treatment with local clindamycin or topical corticosteroids is useful and supports the diagnosis of DIV. By contrast, a good response to estrogen or antibiotics other than clindamycin likely excludes the diagnosis of DIV.

 

 

Etiology

During the past 60 years, different etiologies were suggested to play a key role in the pathogenesis of DIV.

 

Estrogen deficiency

The symptoms and signs of women with DIV and advanced atrophic vaginitis (but not uncomplicated vaginal atrophy) are often similar including erythema, discharge, purulence, elevated pH, inflammatory, and parabasal cell presence and loss of Lactobacillus spp. dominance. However, the typical annular rash is absent in atrophic vaginitis and the dominant picture is that of atrophy (pale, dry vaginal mucosa with loss of rugae). In both conditions, erythematous linear markings may be present. Accordingly, differentiating these two entities is more difficult in middle-aged and older women compared to younger women. Fifty percent of patients with DIV are diagnosed in postmenopausal women [5]. Given the age prevalence, it has been suggested that estrogen deficiency may play a role in the evolution of DIV. Unfortunately, consistently, there is no clinical response to local or systemic estrogen treatment in women with DIV. This phenomenon serves as a critical diagnostic step in differentiating the two entities. In addition, a large percentage of patients diagnosed with typical DIV are not estrogen deficient. Nevertheless, it is accepted today that a trial of treatment with topical estrogen is a useful method to differentiate between DIV and atrophic vaginitis. Menopausal women should continue to apply vaginal estrogen in addition to the anti-inflammatory treatment they receive for the DIV.

 

It is noteworthy that many women with DIV have a history of being incorrectly diagnosed with vaginal trichomoniasis. This occurs because the inflammatory discharge and the rash resembling colpitis macularis are similar to the physical examination findings in women with trichomoniasis. However, DIV does not respond to metronidazole therapy. This should tip off the practitioner to reconsider another etiology. It is not unreasonable to exclude trichomoniasis by obtaining a PCR test for this common cause of vaginitis, especially as the sensitivity of microscopy for the diagnosis of trichomoniasis is poor.

 

Bacterial infection

The absence of normal vaginal lactobacillus morphotypes in patients with DIV and the presence on culture of an abnormal vaginal flora with frequent growth of group B streptococci and Escherichia coli led to the original assumption and more recent suggestions that the etiology is infectious in origin ∗[4] and ∗[12]. The favorable response to 2% clindamycin (cream) strengthened the hypothesis that the etiology is a bacterial infection [4]. Despite these findings, failure of other antibiotics especially metronidazole, with a spectrum of antibacterial activity similar to clindamycin, to treat the condition argues against an infectious etiology. Moreover, studies demonstrating that corticosteroids are as effective as clindamycin undermined the theory of a bacterial cause for DIV [5]. It is of interest that when Gardner described his case series of eight patients, he did not think that the etiology was bacterial. As he wrote, there was “no relationship between the isolated bacteria and development of lesion …. The essential absence of lactobacilli indicates that the predominant bacteria may have become established only as opportunistic organisms, a not uncommon sequence of vaginal microbial change …,” that is, suggesting that the vaginal microbiome found in DIV represents a secondary non-contributing phenomenon only. Extensive culture studies performed by the authors failed to identify any consistent microbial pattern other than the absence of Lactobacillus spp. and overgrowth of coliform and other organisms normally present in the vagina in the postmenopausal women [unpublished]. Although DIV is not an infectious disease, it is reasonable to send a vaginal culture in search for Group A Streptococcus, as part of excluding other forms of vaginitis. Despite being rare, it is considered a pathogen capable of causing purulent vaginitis [19]. Recently, Peteira et al. postulated that DIV resembled a vaginal toxic shock reaction possibly triggered through TSST-1 toxins produced by Staphylococcus aureus or like microorganisms [20].

 

Vitamin D deficiency

There are two case reports by a single author that postulate an association between vitamin D deficiency and DIV [21] and [22]. However, based on Edwards' documentation, there was no improvement after adequate treatment with vitamin D and normalization of levels [23]. Little credence is given to this theory.

 

Potential role for Kallikrein-related peptidase in vaginitis

The Kallikrein-related peptidase (KLK) family consists of 15 genes located on chromosome 19q13.4. These genes encode secreted serine proteases that play a crucial role in skin desquamation [15]. The levels of KLKs are elevated in peeling skin syndrome, psoriasis, and atopic dermatitis, resulting in increased serine proteases activity and subsequent over-desquamation of corneocytes. Relatively large levels of many KLKs are present in human cervicovaginal fluid (CVF) and in the supernatant of cultured human vaginal epithelial cells. KLKs are also hormonally regulated in vaginal epithelial cells, particularly by glucocorticoids and estrogens. The physiological role of KLK in the vagina is currently unknown; however, analysis of the CVF proteome has revealed clues for potential KLK functions in this environment [15].

 

In his original hypothesis, Shaw suggests that KLKs play a role in the normal desquamation of vaginal epithelial cells similar to their role in the desquamation of skin corneocytes [15]. It is suggested that in clinical conditions such as DIV or atrophic vaginitis the levels of KLK and/or activity of KLK are elevated and contribute to over-desquamation [15]. Furthermore, Shaw's group has shown that treatment of vaginal epithelial cells with corticosteroids and estrogen reduces KLK expression. His hypothesis is that treatment with corticosteroids and/or estrogen helps to reduce KLK levels associated with vaginitis, therefore reducing proteolytic activity and desquamation [15]. Why KLKs are potentially increased in women with DIV is entirely unknown.

 

Autoimmune – or immune induced

Although etiology and pathogenesis remain unknown, symptoms and signs overlap with other immune-mediated vaginal disease including erosive lichen planus and benign membrane pemphigus. This together with the favorable response to anti-inflammatory agents suggest that the etiology is immune mediated ∗[4], ∗[16], [24] and ∗[25]. Unfortunately, serologic studies to confirm an immune basis have not been forthcoming. Pro-inflammatory cytokines were identified and reported by Donders, even though the etiologic role of these inflammatory markers has not been hypothesized [12]. In a retrospective longitudinal study of DIV patients, 23% had a history of thyroid disease, unpublished data [5].

 

Despite the lack of specific etiologic findings, it is feasible, if not likely, that DIV represents a common end stage of chronic vaginal inflammation brought about by multiple pathogenic mechanisms.

 

Treatment

Both topical vaginal clindamycin and local vaginal corticosteroids are useful, and sometimes curative, in the treatment of DIV. Both have anti-inflammatory affect. Clindamycin is a lincosamide, which is a class of antibiotics capable of binding to the 23s segment of the 50s subunit of bacterial ribosomal RNA and inhibits protein synthesis. The anti-inflammatory properties of clindamycin emerge from its ability to prevent synthesis of pro-inflammatory cytokines including interleukin 1 (IL-1), IL-6, IL-8, and tumor necrosis factor (TNF). It has been shown that macrolides have overlapping binding sites with clindamycin to the 23s segment, suggesting a similar immunomodulatory mechanism for clindamycin, macrolides, and cortisone [26], [27], [28], [29], [30] and [31].

 

The initial treatment, with either clindamycin or steroid, is administered daily for 2–4 weeks followed by maintenance therapy as needed (Table 4).

There is no single randomized control study comparing the two treatment regimens. However, a large retrospective chart review study, evaluating 98 women with DIV who used one of three treatment protocols, did not find a significant difference in the response to treatment. The treatment protocols included topical vaginal 2% clindamycin 5 g daily, 53 women (57%), vaginal hydrocortisone 10% cream 3–5 g daily, 35 women (37%), and vaginal cortisone acetate suppository 25 mg twice daily six women (7%). All patients adhered to their assigned treatment protocol until the first follow-up visit that occurred at a median of 3 weeks following the initial visit. The initial therapeutic response was extremely encouraging, with 84 patients (86%) dramatically improved and a marked decline in symptoms and signs. During the following weeks, treatment was tapered to twice a week and after a median of 8 weeks, treatment was discontinued in 53 asymptomatic patients. Unfortunately, 17 patients (33%) relapsed in 6 weeks [5]. In a different study, a good response to treatment with clindamycin only was observed with a similar relapse rate (29%) 4 weeks after cessation of treatment [4]. It is important to recognize that DIV is a chronic condition in the majority of cases, and to continue maintenance treatment as needed [5].

 

After 1 year of initial diagnosis, the majority of DIV patients, 87/98 (84%), were asymptomatic; however, 57 women required maintenance therapy and only 25 women were off therapy. It is worth noting that 15 women (16%) were only partially controlled with frequent attempts to intensify local anti-inflammatory vaginal and vestibular therapy (0.05% clobetasol, 0.01% tacrolimus, or 0.5 mg/g estriol cream) [5].

 

Women with DIV who are on maintenance therapy require ongoing evaluation in order to assess improvement and remission. The evaluation includes assessing the improvement of symptoms and signs together with the measurement of inflammatory and parabasal cell number. Vaginal pH as a marker for the presence of normal healthy lactobacillus morphotype will be a part of the assessment only after discontinuing treatment with clindamycin. Microscopy plays an important role in determining the need for additional therapy.

 

The major adverse outcome from the treatment with topical clindamycin or corticosteroids is developing a symptomatic yeast infection. This occurs in 25% of patients. It is recommended to add oral fluconazole to the treatment regimen in patients prone to develop such an infection [5].

 

Summary

DIV is an uncommon severe form of purulent vaginitis causing discharge and dyspareunia. It is a devastating condition with a profound effect on sexual relations. It occurs mainly in Caucasians with a peak occurrence in the perimenopause, yet 50% of women are diagnosed in their reproductive age. The symptoms and signs are nonspecific. DIV is in part a diagnosis of exclusion, and other causes of purulent vaginitis should first be excluded including Trichomonas vaginalis, Group A Streptococcus, atrophic vaginitis, and erosive lichen planus. Definition includes the presence of markers that necessitates the evaluation of vaginal secretion with microscopy (wet mount). Per definition, it is mandatory to show an increase in inflammatory cells and parabasal epithelial cells (immature squamous cells). Vaginal flora is abnormal and pH is always elevated above 4.5. From the author's point of view, one cannot make a diagnosis of DIV without including a description of wet mount microscopy [14], [20] and [21]. Although etiology and pathogenesis remain unknown, symptoms and signs overlap with other immune-mediated vaginal disease including erosive lichen planus and benign membrane pemphigus. This together with the favorable response to anti-inflammatory agents suggest that the etiology is immune mediated. Both topical vaginal clindamycin and local vaginal corticosteroids are useful for treating DIV; the choice of treatment should consider the availability and cost of these medications. The initial therapeutic response is extremely encouraging, with 86% of patients experiencing dramatic improvement; however, relapse and chronic manifestations frequently require long-term topical therapy. It is important to recognize that DIV is a chronic condition in the majority of cases, and to continue maintenance treatment including topical estrogen as needed.