Syndromes causing vulvovaginitis and vaginal discharge
Abnormal vaginal discharge is one of the most common reasons for women to seek medical attention. Since the 1960s, the number of women receiving care for vulvovaginal infections increased approximately threefold. In 2000, an estimated three million initial physician office visits in the United States were prompted by vulvovaginal infections.28
The most common causes of an abnormal vaginal discharge are bacterial vaginosis (BV), vulvovaginal candidiasis (VVC), and trichomonal vaginitis (TV) [see Table 2]. Both BV and trichomoniasis have been associated with preterm labor.49,50 However, to date, treatment of these infections has not definitively been shown to decrease preterm delivery.51-54 BV has been identified as a risk factor for PID, and both BV and trichomoniasis may increase the risk of HIV acquisition and transmission.55-59 Consequently, these diagnoses have assumed new importance in HIV prevention.
Evaluation of women with vaginal complaints should include a pelvic examination and a directed laboratory evaluation. Although these infections tend to have different clinical features, a study of patients triaged and selectively treated after a telephone assessment found poor agreement between the diagnosis made by nurses and other providers and the diagnoses obtained after examination and testing.60 Similarly, a study of women purchasing over-the-counter antifungal therapies found that only one third had vaginal candidiasis, and 53% had a diagnosis other than vulvovaginal candidiasis.61 These findings emphasize the need for a complete evaluation in women complaining of vaginal discharge or discomfort. Less frequent causes of vaginitis include atrophic vaginitis with secondary bacterial infection, vaginitis associated with foreign bodies or toxins, Staphylococcus aureus vaginitis associated with toxic-shock syndrome, group A Streptococcus-associated vaginitis, desquamative vaginitis (clin-damycin responsive), erosive lichen planus, allergic vaginitis, vaginitis associated with autoimmune disease, and idiopathic vaginitis.62
Bacterial Vaginosis
BV is the most common cause of vaginal discharge in women of reproductive age. Prevalence studies have found BV in 10% to 40% of women tested, with higher rates of infection in women tested in STD clinics and in African Americans. Douching and use of intrauterine devices (IUDs) have also been associated with BV.63
Table 2 Clinical Features and Management of Vulvovaginitis
|
Feature |
Normal Vaginal Examination |
Vulvovaginal Candidiasis |
Trichomonal Vaginitis |
Bacterial Vaginosis |
|
Etiology |
Uninfected; lactobacilli predominate |
Candida albicans most common; candidiasis caused by species other than C. albicans may be increasing |
Trichomonas vaginalis |
Loss of normal vaginal lacto-bacilli; associated with Gard-nerella vaginalis; increased anaerobic bacteria and mycoplasmas |
|
Symptoms |
None |
Abnormal vaginal discharge, external dysuria, vulvar itching, pain and/or irritation |
Yellow vaginal discharge, external dysuria, vulvar itching |
Increased, abnormal, or malodorous vaginal discharge |
|
Discharge Amount Color Consistency |
Variable Clear or white Nonhomogeneous, patchy (floccular) |
Scant White Clumped; adherent plaques |
Profuse Yellow Homogeneous or frothy |
Moderate White or gray Adherent, homogeneous discharge that uniformly coats vagina |
|
Inflammatory findings |
None |
Vulvar erythema, edema, or fissure; erythema of vaginal epithelium; introitus |
Erythema of vaginal and vulvar epithelium; colpitis macularis |
None |
|
pH of vaginal fluid* |
Usually s 4.5 |
Usually s 4.5 |
Usually > 4.5 |
Usually > 4.5 |
|
Amine (fishy) odor with 10% KOH |
None |
None |
May be present |
May be present |
|
Microscopy |
Normal epithelial cells; lactobacilli predominate |
Leukocytes, epithelial cells; mycelia or pseudomycelia+ (50%-85% of cases) |
Leukocytes; trichomonads seen in 50%-70% of culture-positive cases |
Clue cells (81%-94% of cases); few leukocytes; lactobacilli outnumbered by mixed flora |
|
Recommended treatment |
— |
Intravaginal imidazole (butocona-zole, clotrimazole, miconazole, terconazole, tioconazole) for 3-7 days; fluconazole, 150 mg p.o. (single dose) |
Metronidazole, 2 g p.o. (single dose); metronida-zole, 500 mg p.o., b.i.d., for 7 days |
Metronidazole, 500 mg p.o., b.i.d., for 7 days; metronidazole gel, 0.75%, 5 g intravaginally each night for 5 nights; clindamy-cin cream 2%, 5 g intravagi-nally each night for 7 days |
|
Sexual partner treatment |
— |
None if asymptomatic; topical treatment if candidal dermatitis of the penis or balanitis is detected |
Metronidazole, 2 g orally (single dose) |
None |
*pH determination is not useful if blood is present.
+To detect fungal elements, vaginal fluid is digested with 10% KOH before microscopic examination; to examine for other features, fluid is mixed (1:1) with normal saline. Culture may be necessary if microscopy results are negative and the suspicion of Candida is high.
Pathophysiology and transmission The etiology of BV is unknown. The syndrome constitutes a disturbance in normal vaginal bacterial flora characterized by a reduction in the concentration of hydrogen peroxide-producing lactobacilli64 and increased growth of mixed bacterial flora that include Gardnerella vaginalis, anaerobes, and Mycoplasma hominis. There is evidence that BV can be transmitted sexually. This evidence includes the following: studies have demonstrated that the inoculation of women with vaginal fluid from another woman with BV can induce the syndrome65,66; there is a high prevalence of BV in patients being treated at STD clinics; there are high rates of concordant BV among lesbian sexual partners67; longitudinal studies have associated BV with having higher numbers of sexual partners and with having new sexual partners68-70; and most studies have found that BV is absent in virgins.71 Evidence against sexual transmission includes the lack of benefit from treating sexual part-ners72-74 and inconsistent associations with levels of sexual activity.
Diagnosis Physical examination of women with BV typically reveals a homogeneous, white, uniformly adherent vaginal discharge.63 The Amsel criteria for diagnosis of BV include the following: (1) presence of a homogeneous, thin vaginal discharge; (2) vaginal pH greater than 4.5; (3) clue cells (bacteria attached to vaginal epithelial cells on wet mount); and (4) presence of an amine (fishy) odor when vaginal fluid is mixed with 10% potassium hydroxide (KOH).71,75-77 The presence of three of the four criteria establishes the diagnosis [see Table 3].
Treatment BV is treated with metronidazole. A meta-analy-sis found higher cure rates with a dosage of 1 g a day for 7 days than with a single 2 g dose (82% versus 73%).63 Intravaginal metronidazole and intravaginal clindamycin offer efficacy comparable to 7-day courses of metronidazole, with fewer side effects, but are not effective in the treatment of trichomoniasis and are typically more costly. Recurrence of BV is common, occurring in 50% to 70% of cases. Multiple randomized trials have failed to demonstrate any benefit from treating male partners.
Trichomoniasis
T. vaginalis is a sexually transmitted protozoan. In the United States, the number of women seeking care for trichomonal vaginitis declined by over 50% from 1966 to the mid-1980s; in 2000, physicians in the United States saw an estimated 200,000 patients with TV.28 A cross-sectional study of 13,816 pregnant women in the United States found TV in 13%; the vast majority of those infections were subclinical or asymptomatic. Risk factors for TV included African-American ethnicity, cigarette smoking, unmarried status, and lower educational level.77 Untreated infections in women are thought to persist for a median of 3 to 5 years.78
Table 3 Amsel Criteria for the Diagnosis of Bacterial Vaginosis71,75,77
|
Criterion |
Sensitivity (%) |
Specificity (%) |
|
Homogeneous, thin vaginal discharge |
52-65 |
71-97 |
|
Vaginal pH > 4.5 |
92-97 |
53-62 |
|
Clue cells on vaginal wet mount |
81-94 |
94-98 |
|
Amine odor when vaginal fluid is mixed with 10% potassium hydroxide (KOH) |
43-84 |
98–99 |
Note: the presence of three of these four criteria establishes the diagnosis of bacterial vaginosis.
Diagnosis Clinical manifestations of trichomonal infection include yellow vaginal discharge and vulvar itching. Neither is highly sensitive or specific. On physical examination, signs associated with Trichomonas infection include frothy or purulent vaginal discharge, which is sometimes profuse; vulvar or vaginal erythema; and cervical mucopus.79,80 All of these signs have far greater specificity than sensitivity. The finding of colpitis macularis—punctate cervical hemorrhages and ulcers, sometimes referred to as strawberry cervix—has a specificity of 99% for TV but is seen in fewer than 5% of patients on unaided physical examination; colpitis macularis is much more readily visible on colposcopy.79 In expert hands, a finding of motile Trichomonas on wet-mount examination has a sensitivity of 50% to 70%, although in clinical practice, wet-mount examination is usually considerably less sensitive. Culture on Diamond medium is the traditional diagnostic gold standard, but this technique is not available in most practice settings. Recently, InPouch, a relatively simple and inexpensive culture method, became available. The sensitivity of InPouch is comparable to that of Diamond medium and superior to that of wet mount.81 PCR has been successfully used in research settings, but no NAAT is commercially available at present. Antigen detection tests are also under investigation.
Treatment A single 2 g dose of metronidazole is the treatment of choice for TV. Reported cure rates are 82% to 88%.80 Sexual partners should be treated concurrently, and couples should be advised to abstain from sex for 1 week after treatment. Topical metronidazole is not effective.82 Resistance to metronidazole occurs infrequently, and most cases respond to prolonged courses of metronidazole therapy. Some authors have reported successful treatment of metronidazole-resistant cases using either tinidazole or paromomycin cream.83
Vulvovaginal Candidiasis
Because VVC is not a reportable infection, only limited epi-demiologic data are available. In the United States, a study of female university students found that over half experienced at least one episode of VVC by 25 years of age,84 and 6.5% of women who participated in a national random-digit-dialing survey reported that a health care provider had told them they had candidal vaginitis at least once in the preceding 2 months.85 Higher rates of VVC have been observed in African Americans and in users of oral contraceptives, vaginal sponges, or IUDs.86 Although VVC is not clearly identified as an STD, it has been associated with the onset of sexual activity in young women and with cunnilingus.84,86 Other predisposing factors include recent use of antibiotics; diabetes mellitus; pregnancy; and immunodeficiency, including that from HIV infection.
Diagnosis Vulvovaginal pruritus is generally the most common symptom of VVC.87 Other findings sometimes associated with VVC include a cottage-cheese-like discharge; external dysuria; external genital burning or pain; perineal edema or erythema; and vulvar erythema, edema, and fissures.87,88 However, several studies have reported the absence of any signs or symptoms significantly associated with VVC.4,89 As a result, the diagnosis requires microscopic and, at times, microbiologic assessment. A 10% KOH preparation of fluid taken from the vagina has a sensitivity of 50% to 85% in the diagnosis of VVC9091; if this test is negative but the clinical picture is consistent with VVC and there is no alternative diagnosis, culture for yeast should be performed.
Treatment Topical azoles (e.g., butoconazole, clotrimazole, miconazole, econazole, tioconazole, and terconazole) are 80% to 90% effective in treating VCC [see Table 2]. Most of these agents are available over the counter. No clear advantage favors one azole over another. Oral azoles (fluconazole or itraconazole) are comparably or slightly more effective and may be more convenient, but these agents also pose a small risk of systemic reactions. Because there are no compelling data favoring any one agent or route of administration, patient preference should guide the choice of treatment. Immunosuppressed patients and those with candidal infections caused by a species other than Candida albicans may require more prolonged therapy (e.g., 14 days).
Long-term therapy is indicated for patients with recurrent VVC, which is defined as four or more episodes of VCC in a year. Approximately 5% of women with VVC experience recurrences. Treatment may require 14 days of induction therapy followed by once-weekly maintenance therapy. Patients with C. glabrata VVC who do not respond to prolonged courses of azole therapy may benefit from topical boric acid (600 mg once a day for 2 weeks) or topical flucytosine.92
Mucopurulent Cervicitis
Mucopurulent cervicitis (MPC) is an inflammatory process affecting the columnar epithelium and subepithelium of the endo-cervix and adjacent exocervix. As with NGU in men, MPC is common and has most frequently been associated with N. gonorrhoeae or C. trachomatis and, less frequently, with HSV or T. vagi-nalis. Unlike NGU, MPC typically produces no symptoms; or it may produce nonspecific symptoms, such as a yellow vaginal discharge, that often do not prompt women to seek treatment. In recent years, as the prevalence of gonorrhea and chlamydial infections have decreased in some settings, MPC with no defined microbiologic etiology has come to constitute the majority of cas-es.4 MPC is important because of its association with known infections and because patients with MPC have an elevated risk of PID and adverse pregnancy outcome.
Diagnosis Different diagnostic criteria have been used for MPC. According to current CDC guidelines, the diagnosis of MPC is made on the basis of a finding of a visible purulent or mucopurulent exudate on cervical examination or on endocervi-cal swab. The finding of cervical mucopus is 28% to 52% sensitive and 82% to 94% specific for the presence of either C. tra-chomatis or N. gonorrhoeae.88,93,94 Some investigators use additional criteria for MPC, including a finding of from 20 to 30 PMNs per high-power field (hpf) on cervical Gram stain or easily induced cervical bleeding.4,94 These factors have been associated with the likelihood of C. trachomatis or N. gonorrhoeae infection, but they have not consistently been included as diagnostic criteria of MPC; with regard to the use of cervical Gram stain, these findings have not consistently been useful in defining a population in need of empirical therapy.
Treatment The decision to treat MPC is based largely on the local prevalence of C. trachomatis or N. gonorrhoeae and on the patient’s risk. In areas where both gonorrhea and chlamydial infection are common, empirical therapy should be directed at both pathogens. In areas where gonorrhea rates are low, treating for Chlamydia infection alone is reasonable. Recent evidence suggests that in areas where the prevalence of both infections is low, older patients (i.e., those older than 30 years) suspected of having MPC need not be treated until microbiologic test results are available, provided follow-up care is ensured.
Pelvic Inflammatory Disease
PID is an inflammatory process involving a variable combination of endometritis, salpingitis, tubo-ovarian abscess, and pelvic peritonitis. PID can be blood-borne (e.g., tuberculosis) or result from extension of an intra-abdominal process. At present, however, PID most often develops when bacteria ascend from the vagina or cervix into the endometrium, fallopian tubes, and pelvic peritoneum. Although the number of women seeking care for PID has declined by over 25% since the 1980s, 8% of participants in the 1995 National Survey of Family Growth, a national representative sample of United States women, reported a history of PID.95 Identified risk factors for PID include a previous history of PID, higher numbers of lifetime sex partners, douching, and a history of bacterial STD. In the past, IUD use was identified as a risk factor, but its importance beyond the first 30 days after insertion is now controversial; a recent case-control study found no association between the use of currently available copper IUDs and the occurrence of PID.96 Gynecologic procedures that disrupt the protective cervical barrier (e.g., pregnancy termination, IUD insertion, dilatation and curettage, and hys-terosalpingography) elevate the risk of PID and may lead to PID in the absence of classic sexually transmitted pathogens.
Microbiology
Studies of PID conducted in the United States and Europe in the 1980s typically implicated C. trachomatis, N. gonorrhoeae, or both as a cause of PID in approximately half of cases.97 Frequently, these bacteria were part of a polymicrobial infection involving diverse normal vaginal flora, including anaerobic bacteria, facultative anaerobes, and genital mycoplasmas. M. genitalium has been associated with endometritis and PID.98 Actinomyces is-raelii is a cause of PID in women with IUDs.
Diagnosis
The diagnosis of PID is difficult. To date, studies have been unable to identify any single clinical finding or constellation of findings that allow accurate identification of women with PID.99,100 Moreover, PID studies have typically enrolled only women with overt disease and, consequently, have not provided an accurate picture of the full spectrum of the clinical entity. Indeed, most cases of PID probably go undiagnosed. Approximately two thirds of women with postinfectious fallopian tube occlusion report no history of PID, although many have sought care for abdominal pain.101 When the diagnosis is made clinically, it may not be supported by surgical findings. Only 60% to 70% of women with clinically diagnosed PID typically have laparo-scopic evidence of PID.
In clinically detected cases, the cardinal symptom of PID is pelvic or abdominal pain. The pain is typically dull or aching. Onset can be acute or subacute and frequently occurs at the beginning of menses. Typically, patients present after having symptoms for less than 2 weeks. In a Swedish study of 623 patients with PID, all had pelvic or abdominal pain, cervical motion tenderness, and increased inflammatory cells in vaginal or cervical secretions. Other symptoms and laboratory findings included an erythrocyte sedimentation rate (ESR) of 15 mm/hr or higher (75%), leukocytosis greater than 10,000/ml (60%), abnormal vaginal discharge (55%), fever higher than 38° C (100.4° F) (41%), abnormal vaginal bleeding (36%), dysuria (19%), vomiting (10%), and anorectal symptoms (anorectal pain, tenesmus, or rectal bleeding or discharge) (7%).1Q3 A large study showed that only temperatures higher than 38° C (100.4° F) had a specificity of more than 90% for the diagnosis of PID, although its sensitivity was only 11%.
Although laparoscopy has been the traditional gold standard for diagnosing PID, many women with abnormal fimbrial biopsies have normal results on laparoscopy. Moreover, some women have histologic evidence of endometritis without salp-ingitis,104 which suggests that laparoscopy may be insensitive for the detection of milder cases or of PID that is restricted to the uterus.
Transvaginal ultrasound (TVUS) should be performed when symptoms are severe, when the physical examination reveals a pelvic mass, or when the diagnosis of PID is uncertain. Studies assessing the performance of different imaging modalities in the diagnosis of PID have been small, with no single study enrolling more than 50 patients with the diagnosis.102 Small studies of TVUS have reported sensitivities of 81% to 93%, but specificities have been highly variable, ranging from 5% to 100%, with the test performing best in patients with more severe infection.102 A case-control study of power Doppler TVUS reported a sensitivity of 100% and a specificity of 80%,1Q5 suggesting it may offer advantages over conventional TVUS. In women with tubo-ovarian abscess, repeat TVUS is often indicated to assess response to therapy. Small studies of CT and pelvic MRI have also reported high sensitivity and specificity.
