Gynecologic Care of HIV-Infected Women

There are nearly 16 million women living with HIV/AIDS worldwide, making up approximately half of all infections.¹ Although the goal of this chapter is to describe the gynecologic care of the HIV-infected woman in industrialized settings, it is critical to understand that the burden of the epidemic lies within resource-limited areas, particularly sub-Saharan Africa.

In 2007, the Centers for Disease Control and Prevention reported 198,544 cumulative cases of AIDS among women in the United States, accounting for almost 20% of overall cases.² AIDS incidence is increasing yearly among American women, particularly among women of color. African American women represent two thirds of current HIV/AIDS cases and have ten times the risk of becoming infected compared to white women.³ The most common risk factor for HIV acquisition in women is heterosexual contact with high risk partners, followed by injection drug use.² In the United States, there is dramatic geographic variation in the prevalence of HIV/AIDS among women, with the highest prevalence of AIDS cases in the northeast and southeast regions of the country. New York City has the highest frequency, with three times the national prevalence of AIDS cases among women.⁴

Gynecologic diseases are commonly encountered when caring for HIV-infected women.⁵ The increasing proportion of reproductive aged women infected with HIV means that the gynecologist should be prepared to care for the unique gynecologic conditions and issues facing this population. Advances in highly active antiretroviral therapy (HAART) mean that more women will enjoy a longer life expectancy, along with the ability to bear children with an extremely low risk of disease transmission. Family planning and preconception counseling, along with screening and management of sexually transmitted infections, menstrual abnormalities, and lower genital tract dysplasia are among the gynecologic issues that may be influenced by HIV disease. As life expectancy increases, and a larger proportion of older women are living with HIV, the impact of infection on menopause and bone mineral density must be considered. The field of research in HIV and gynecology is continuing to evolve, and evidence-based clinical care for women with these conditions may prove to be challenging. In general, these conditions are more common and severe among more immunosuppressed women.⁴

Women and men appear to have similar rates of disease progression and survival, once access to care, use of antiretrovirals, and stage of disease at presentation are accounted for.^{6, 7, 8} Nonetheless, women appear to have less access to care and when in care are less likely than men to receive antiretrovirals and opportunistic prophylaxis.^{9, 10, 11, 12, 13} Most studies addressing gender differences have found that women, on average, have lower plasma HIV RNA levels at any given CD4 count compared to men.^{14, 15} In their systematic review of the published literature, Gandhi and associates found that women's CD4-adjusted HIV RNA levels were approximately two- to six-fold (0.13–0.78 log₁₀) lower than those of men.¹⁵ The meta-analysis by Napravnik and associates confirmed this finding, calculating the CD4-adjusted HIV RNA summary estimate to be 41% lower (0.23 log₁₀) among women compared with men.¹⁴ There are apparent cyclic alterations in viral load among women as well. Greenblatt and associates noted a 0.16-log₁₀ decrease in median HIV RNA levels from the early follicular to midluteal phase among ovulatory women.¹⁶ Nonetheless, menopause does not appear to have an impact on CD4 counts.¹⁷ The full implication of these findings for clinical care remains unclear.

Studies looking at sex differences in response to HAART have shown mixed results, with a recent meta-analysis showing no sex-associated difference in response¹⁸ and guidelines for initiating and changing HAART regimens are the same for men and women.¹⁹ Women have higher plasma levels of antiretroviral medications, due to body weight, renal clearance, and P-glycoprotein activity.²⁰ Studies have shown that women have higher frequencies of lipodystrophy and lipoatrophy, lactic acidosis, and HAART-related rash.^{21, 22, 23, 24, 25} This increased frequency of adverse events may explain the lower levels of adherence observed in several recent studies.^{26, 27} 

Human papillomavirus associated cervical neoplasia

Co-infection with oncogenic human papillomavirus (HPV) and HIV exponentially increases a woman's risk for development of cervical neoplasia.⁶ HIV increases the incidence and persistence of HPV infection.^{28, 29, 30} Increased prevalence and time to clearance of HPV infection has been associated with worsening HIV disease status.³¹

Women with HIV infection have approximately a four- to six-fold increased risk for squamous intraepithelial lesions (SIL) or cervical intraepithelial neoplasia (CIN) compared with uninfected women.^{32, 33, 34} Two large, multicenter cohorts of HIV infected and uninfected women in the United States showed rates of SIL on cytology of 17–18% among HIV-infected women compared with 3–5% among HIV-uninfected women (p <0.001).^{28, 35} HPV infection was 65% among HIV-infected women, versus 30% among seronegative women. A 30-month prospective study of women with no evidence of cervical disease in New York City showed incident cases of SIL in 20% of HIV-infected and in 5% of HIV-uninfected women.³⁶ No women had cervical cancer during the study. A multivariate analysis showed a significant association between HIV infection, transient and persistent HPV infection, and young age with development of SIL. Incidence of cervical disease is correlated both with low CD4+ cell count and increased HIV RNA level.³²

Vulvar, vaginal, and anal neoplasia

Vulvar, vaginal, and anal intraepithelial neoplasia appear to be prevalent among HIV-infected women. In a well-controlled prospective study, vulvar and vaginal condyloma were found in 5.6% of HIV-infected versus 0.8% of HIV-uninfected women. Vulvar intraepithelial neoplasia (VIN) was noted in only two of 396 (0.5%).³⁷ Incident vulvovaginal intraepithelial neoplasia lesions were 16-times more frequent in a prospective study of HIV-infected women compared with uninfected women.³⁸ In this study, after a median follow-up of 3.2 years, 2% of all HIV-infected women had high-grade VIN and one patient had invasive vulvar carcinoma. Cigarette smoking and injection drug use were independently associated with vulvovaginal and perianal lesions in HIV-infected women. A similar prospective study of 280 women showed a 7.5-fold increased incidence in vulvar, vaginal, and perianal intraepithelial neoplasia in HIV-infected women compared to HIV-negative women.³⁹

HPV infection of the anus and epithelial abnormalities on anal Pap tests are common in persons with HIV infection.^{40, 41} In a prospective study of 319 women, 26% of the HIV-infected women and 8% of the HIV-negative women had abnormal anal cytology.⁴²  HIV-infected women have an eight-fold risk of anal intraepithelial neoplasia and a nearly seven-fold increased risk for invasive anal cancer compared with HIV-negative women.⁴³

Although vulvar, vaginal, and anal cancers are infrequently reported in HIV-infected women, these conditions may become more common as improved medical care of HIV leads to women living with immunosuppression and intraepithelial neoplasia for longer periods of time. In women with immunosuppression after organ transplantations, a 100-fold increase in vulvar cancer has been reported occurring after an average of 107 months.⁴⁴

Highly active antiretroviral therapy and neoplasia

The impact of HAART on cervical neoplasia has not been clearly elucidated.⁴⁵ Theoretically, HAART-mediated immune reconstitution should decrease the incidence and recurrence of cervical neoplasia. Conversely, increased life expectancy may increase a woman's overall risk of cervical disease. Koshiol et al. showed that HIV did not affect the time to HPV clearance among women receiving HAART.⁴⁶ In a study of 168 HIV-infected women, the 96 who received HAART had twice the likelihood of regression of CIN during a 12-month follow-up period compared with those who did not use HAART.⁴⁷ In a larger study, the Pap smears of women who received HAART were less likely to have progression to a higher grade (OR 0.68; 95% CI 0.52–0.8) and more likely to have regression.⁴⁸ Other studies have not shown a significant effect of HAART on regression or recurrence of SIL.^{49, 50, 51}

Diagnosis

Most controlled studies have shown Pap smear sensitivity and specificity rates similar to those reported for HIV-negative women.^{52, 53} HIV-infected women who have cytologic atypia or SIL on their Pap smears should undergo a colposcopic examination. Some investigators advocate screening colposcopy based on their findings in populations with a higher prevalence of dysplasia.⁵⁴ However, a cost-effectiveness study comparing semi-annual cervical cytology with colposcopic screening of HIV-infected women in a Markov model showed that colposcopy added little and increased costs markedly.⁵⁵ On the basis of the predominant findings, routine colposcopy is not recommended in published guidelines for screening of HIV-infected women.^{56, 57} Because of the high prevalence of HPV in HIV-infected women, HPV typing (by polymerase chain reaction) may not be a useful means of triage of abnormal cervical smears.²⁸ HPV testing by hybrid capture technique may be more useful in HIV-infected women. In a study from Hannover, Germany, HPV testing identified 16 of 17 (94%) of the lesions classified as CIN2 or higher with a specificity of 70%. Cervical cytology identified 11 of 17 (65%) of the lesions classified as CIN2 or higher with a specificity of 98%.⁵⁸ Cost-effectiveness analyses have looked at screening strategies using more frequent Pap smears and high risk HPV testing among HIV-infected women. Goldie et al. showed that cervical cytology screening every 6 months and then annually after two negatives was a cost-effective strategy.⁵⁵ In a follow-up study, the same author modeled HPV testing as a screening strategy, and found semi-annual cytology screening among women with high risk HPV and annual screening among and HPV-uninfected women to be a cost-effective strategy.⁵⁹  n our practice, we do cytology screening every six months until two consecutive negative tests, and then annually thereafter.   

The optimum method of screening for noncervical lesions in HIV-infected women has not been determined. At a minimum, careful inspection of the vulva and perianal region should be performed during pelvic examinations and in any women undergoing colposcopy. There is no consensus of opinion regarding anal Pap smears, although some experts recommend regular screening in settings with equipped for follow-up.^{60, 61} Two recent systematic reviews and a cost-utility analysis have concluded that current evidence does not support regular anal Pap screening among high risk HIV-infected populations.^{62, 63, 64} High-resolution anoscopy (the use of a colposcope to examine the anal canal) is indicated for anal Pap smears that show atypical squamous cells, low-grade squamous intraepithelial lesions (LSIL) or HSIL.^{60, 65} Approximately half of high-grade anal intraepithelial neoplasia (AIN) lesions will have only mild cytological abnormalities on cytology.⁴²

Clinical course and treatment

The clinical course of CIN in HIV-infected women is characterized by more progression and less regression than in HIV-negative women.⁶⁶ A prospective study of women with SIL followed-up for 5 months in the Ivory Coast showed persistence in 76% of HIV-infected compared with only 18% of HIV-uninfected women (RR 4.3; 95% CI 2.4–7.7).⁶⁷ Women with CIN 2 or greater should undergo treatment with an ablative or excisional procedure. Women with CIN 1 can be managed with observation in the absence of severe immunosuppression, as a substantial portion of these lesions will regress.³² Persistence or recurrence of CIN (40–100%) after treatment is common in HIV-infected women, especially those with immunosuppression.^{52, 66, 68, 69, 70, 71} Recurrence rates may by higher among women who undergo cryotherapy, as compared to Loop Electrosurgical Excision Procedure.⁷² A possible explanation for the increased treatment failure is the high likelihood of persistent HPV infection and elevated HPV viral load in HIV-infected women.²⁹

Treatment of VIN in HIV-infected women is difficult. In one study, these women uniformly had recurrence of VIN after excisional or ablative therapy.⁷³ The outcome of surgical treatment of AIN in HIV-infected patients appears to be similar, with approximately 80% having persistent or recurrent disease.⁷⁴ In one retrospective analysis, 65% of women with AIN, VIN or VAIN had recurrences 2 years after treatment, and 10% had invasive cancer.⁷⁵

While the high frequency of recurrence of intraepithelial neoplasia after therapy in HIV-infected women is concerning, it is important to keep in mind that invasive cancer occurs rarely in HIV-infected women who are treated for cervical dysplasia. HIV-infected women with dysplasia need careful follow-up after treatment and often need second or third therapeutic procedures. Adjuvant 5-fluorouracil after excisional or ablative therapy for high-grade cervical dysplasia may reduce the need for repeat excisional procedures. In a randomized clinical trial (ACTG 200) of 200 HIV-infected women, women treated with 5% 5-fluorouracil twice weekly for 6 months reduced the recurrence rate of cervical dysplasia to 28%, compared with 47% among women in the control arm.⁷⁶ The 5-fluorouracil not only reduced the frequency of recurrence but also reduced the proportion of recurrences that were high-grade. An additional finding of this study was that women who were receiving antiretroviral therapy had less recurrence, whereas those with lower CD4+ cell counts had more recurrence. Isotretinoin did not significantly prolong time to progression from LSIL to HSIL in a randomized, controlled trial.⁷⁷ A randomized, controlled trial of topical 5% imiquimod cream, an immune-response modifier, in treatment of anogenital warts showed some efficacy in a group of mostly males with HIV infection.⁷⁸ Several small studies have shown a modest response of high-grade VIN to topical imiquimod.⁷⁹

The relationship between cervical cancer and HIV has been recognized since 1987, and the CDC classified cervical cancer as an AIDS-defining malignancy in 1993.⁸⁰ HIV infection has been more clearly associated with invasive cervical cancer in observational studies from Europe and the United States than in most studies from Africa.^{81, 82, 83} This could be related to the lack of screening and diagnostic technologies, differences in access to treatment for both cervical dysplasia and HIV, and increased pressure from competing mortalities in developing countries. Some authors postulate that the association between HIV and cervical cancer is confounded by behavioral factors such as injection drug use or lack of cervical cancer screening among HIV-infected women in the European and American studies.⁸⁴ In the United States Cancer Match Registry for 1978–1996, there was an increased risk for cervical cancer (RR 5.5), vaginal and vulvar cancer (OR 0.8), and anal cancer (OR 6.8) among HIV-infected women.⁸⁵ In the New York State AIDS Registry, the risk for invasive cervical cancer was found to be increased by a relative risk of 6.5 (95% CI 4.1–9.7).⁸⁶ The prevalence of cervical cancer among more than 23,000 HIV-infected women in France, Italy, and Spain was 3.2% among those whose mode of HIV acquisition was injection drug use versus 1.8% among those who acquired HIV through sex.⁸⁷ However, a recent meta-analysis of observational studies looking at cancer risk among people with immunosuppression from either HIV or organ transplantation showed that the increased cervical cancer risk was similar in both groups.⁸⁸ The authors concluded that the increased risk was more likely due to biological than behavioral factors.

Maiman and associates reported that HIV-infected women with cervical cancer were more likely to present at advanced clinical stage than were HIV-negative women.⁸⁹ A review of invasive cervical cancer at two hospitals in New York City showed a HIV seroprevalence of 18%. A greater proportion of HIV-uninfected women presented with early stage disease (I or II) compared with HIV-infected women (69% vs. 42%).⁹⁰ In a multivariate analysis, risk factors associated with advanced cervical cancer were symptom duration more than 3 months and more than 3 years since last Pap smear. HIV infection was not correlated with advanced cervical cancer (OR 0.8; 95% CI 0.2–3.4). Thus, the advanced stage of cancer may reflect behavioral rather than biological factors. In the WIHS cohort, after a median of 10.3 years of follow-up of women undergoing regular screening and treatment of preinvasive lesions, there was no increased risk of cervical cancer among HIV-infected compared with uninfected women (OR 1.32 CI 0.27–3.85).⁹¹

HIV, HPV, cervical dysplasia, and cervical cancer are clearly epidemiologically related. Consideration should be given to counseling for HIV testing of women who have high-grade CIN, VIN, and AIN.  As cervical cancer is an AIDS-defining illness, HIV-testing is clearly indicated among those women who have invasive cervical cancer. Most of the available evidence indicates that lower genital tract neoplasia in women with HIV is extremely common and in many cases may be refractory to standard treatment, especially in the presence of severe immunosuppression. Screening for HPV-mediated lower genital tract disease differs among women with HIV. We recommend semi-annual Pap screening, followed by annual Paps after two normal results. HIV-infected women should undergo colposcopic evaluation if the Pap shows dysplasia or atypia; the role of HPV testing among HIV-infected women remains unclear.

Epidemiologic synergy describes the mutually amplifying epidemics of sexually transmitted infections (STIs) and HIV.^{92, 93} The presence of an STI, particularly an ulcerative STI, increases the risk of acquisition of HIV by two- to five-fold.⁹² The proposed mechanisms include the disruption of genital mucosa and the recruitment of HIV target cells, including CD4 cells, Langerhans cells, and macrophages, to the genital tract in the setting of an STI. In addition, infection, inflammation, and genital ulceration are all correlated with increased presence and magnitude of HIV viral copies in the genital tract, potentially leading to an increased risk of HIV transmission.^{94, 95, 96} Herpes simplex virus type-2 co-infection is associated with the greatest increase in cervicovaginal HIV shedding, even in the absence of genital ulcers.^{97, 98} Increased shedding has also been seen with Chlamydia trachomatis⁹⁹ and Neisseria gonorrhoeae.¹⁰⁰ Three large randomized clinical trials in sub-Saharan Africa looked at STI treatment to reduce HIV transmission; however, only one of the three showed a significant reduction.¹⁰¹ Explanations for the varying results include difference in high risk behaviors and low prevalence of treatable STIs and HIV in the populations in which no effect was seen. Regardless, all providers caring for HIV-infected women should obtain a thorough sexual history (including type and number of partners, specific sexual activities, condom and other barrier use, HIV-status of partner(s), reproductive desires, past STIs, and current STI symptoms), and undergo routine screening for STIs and treatment as indicated.⁵⁷ Appropriate STI screening should be guided by a woman's sexual behavior and local epidemiology in addition to standards of care established for HIV-infected individuals.¹⁰²

Clinical management of STIs among HIV-infected women differs from that of HIV-uninfected women in certain cases. For instance, women co-infected with HIV and herpes simplex virus (HSV) may have more frequent, painful, and prolonged episodes of genital lesions as compared with HIV-negative women.⁵⁷ Immune reconstitution with HAART has been shown to reduce the frequency of symptomatic outbreaks, although women may still have subclinical HSV-viral shedding.¹⁰³ Suppressive therapy with oral antiviral agents such as acyclovir and valacyclovir is effective in decreasing the clinical manifestations of HSV among HIV-infected women.^{104, 105} While acyclovir-resistant HSV is uncommon, it should be considered in HIV-infected women with persistent lesions despite adequate antiviral therapy. In such cases, a viral isolate should be obtained for sensitivity testing. Because valacyclovir has the same active metabolite as acyclovir, patients with verified acyclovir-resistant strains should be treated with foscarnet under the supervision of an expert in the field as recommended by the Centers for Disease Control.⁵⁷

For most co-infected individuals, HIV does not appear to alter the clinical course of syphilis. Nonetheless, HIV-infected individuals appear to be at slightly increased risk for both neurosyphilis and persistent disease despite treatment, although overall the absolute risk of these events is low.⁵⁷ The CDC recommends treating HIV-infected patients with primary, secondary, and early latent syphilis with the standard single dose of 2.4 million units of intramuscular benzathine penicillin. In the absence of neurological signs or symptoms, pretreatment cerebral spinal fluid (CSF) assessment is not necessary. Because of the increased risk of treatment failure, it is important to monitor these patients clinically and serologically using a nontreponemal antibody titer at 3, 6, 9, 12, and 24 months after initial treatment. Cases in which nontreponemal titers do not decrease four-fold by 6 months after treatment should be considered treatment failures, undergo a CSF evaluation, and, if normal, get treated with weekly benzathine penicillin injections for 3 weeks. Women with late latent or latent syphilis of unknown duration should undergo a pretreatment CSF examination and, if normal, receive weekly benzathine penicillin injections for 3 weeks. These women should be monitored clinically and serologically at 6, 12, 18, and 24 months. Neurological symptoms or a four-fold increase in nontreponemal titers necessitate a repeat CSF evaluation and treatment as indicated. Those women in whom the nontreponemal titer does not decrease by four-fold within 12 months should also undergo a repeat CSF evaluation and be treated accordingly. Neurosyphilis may occur at any stage of syphilis. Treatment includes intravenous aqueous crystalline penicillin G 3–4 million units every 4 hours for 10–14 days. The CSF should be monitored at 6-month intervals until the cell count normalizes.

Treatment recommendations for gonorrhea, Chlamydia, trichomonas, lymphogranuloma venereum, granuloma inguinale, pediculosis pubis, and scabies are the same for HIV-infected and uninfected women.⁵⁷

Hepatitis B (HBV) is a vaccine-preventable STI. Among those with HIV, HBV is more likely to proceed to chronic infection. It is, therefore, particularly important to vaccinate HIV-infected individuals against HBV with the three-part vaccination series over the course of 6 months (0, 1, and 6 months). Because immunization may be impaired in immunocompromised hosts, clinicians should obtain a HBV surface antibody (HBsAb) approximately 1–2 months after the third vaccine injection to confirm immunity.¹⁰⁶ Modification of dosing, including doubling the antigen load or giving additional doses, may improve the immune response in HIV-infected individuals.¹⁰⁷

As with all STIs, sexual partners of infected patients should be evaluated and treated appropriately. For those clinicians who do not regularly treat STIs, the CDC maintains updated treatment guidelines that are easily accessible.⁵⁷

HIV and the organisms associated with pelvic inflammatory disease are sexually transmitted, therefore it is not surprising that these conditions are frequently coincident. Because HIV seroprevalence among women with PID ranges from 6.7% to 22%, serologic testing for HIV infection should be offered to women with a diagnosis of PID.¹⁰⁸

Earlier studies suggested that HIV altered the clinical presentation of PID, with more frequency of severe disease, prolonged hospitalization, and increased need for surgical intervention among HIV-infected women.^{109, 110} However, more recent investigations have shown little difference in the presentation and course of disease between HIV-infected and uninfected women, with HIV-infected women experiencing more tubo-ovarian abscesses.^{111, 112, 113, 114, 115} Prospective studies show little difference in response to antibiotic treatment (combined with drainage of abscesses greater than 4 cm in some of these studies) when comparing HIV-infected and uninfected women. Outpatient treatment of PID has been used with success in HIV-infected women.^{112, 113, 114, 115} Current CDC recommendations for treatment of PID are the same for HIV-infected and uninfected women, with the caveat that indications for hospitalizations and more aggressive management are yet to be determined.⁵⁷ We recommend vigilance for treatment failure in patients with severe immunosuppression and suggest use of antibiotic regimens with optimal anaerobic coverage in this group.

Background

HIV-related immunosuppression can impact the frequency and/or severity of vaginal infections. In turn, these infections may influence HIV acquisition/transmission by several mechanisms, including an increased presence and magnitude of HIV shedding, increased presence of inflammatory cells in the cervico/vaginal epithelia, or by adverse effects on protective vaginal microflora.

Vaginal isolates of Candida albicans from HIV-infected women produce higher levels of secretory aspartyl proteinase, a virulence enzyme, than do those of HIV-uninfected women.¹¹⁶ A recent meta-analysis of inflammatory correlates of HIV genital shedding showed Candida albicans to be significantly associated with the increased frequency of HIV detected in the genital tract.¹¹⁷ A study performed in Mombasa, Kenya, showed a decrease in cell-associated and cell-free HIV after treatment of candidal vaginitis, a decrease only in cell-free HIV after treatment of Trichomonas vaginalis, and no effect on cell-free or cell-associated HIV after treatment of bacterial vaginosis (BV).¹¹⁸ In this report, the most significant decrease in genital-tract HIV shedding occurred in women with Candida who had vulvar erythema and in women with Trichomonas who had more than 10 neutrophils per high-power field on vaginal Gram stain.

BV is associated with HIV seropositivity.¹¹⁴ A recent meta-analysis of 23 studies, including a total of 30,739 women found that all except one study showed an increased risk of HIV acquisition among women with BV.¹¹⁹ This is biologically plausible, as women with BV have lower concentrations of vaginal lactobacillus. This leads to a higher vaginal pH, which is more favorable to HIV than the normal acidic pH. It is unclear whether BV may increase female susceptibility to heterosexual transmission of HIV, whether immunosuppression may lead to increased susceptibility to bacterial vaginosis, or both. Prospective trials of BV and HIV acquisition among low- and high-risk women, and possibly a randomized trial looking at the impact of BV control measures on acquisition are needed to determine the clinical significance of this association.

Lower genital tract mucosal inflammation caused by Trichomonas vaginalis infection could lead to increased shedding of HIV in seropositive women and increased susceptibility to infection with HIV in seronegative women. A recent prospective cohort study of 150 HIV-infected women showed that trichomonas treatment reduced the detection of HIV-1 in genital secretions (RR 0.34, CI 0.12–0.92).¹²⁰ To date, however, studies show an inconsistent association between trichomonal infection and HIV acquisition and transmission.¹¹⁴   

Prevalence and incidence

The association between HIV infection and vaginal candidiasis is debated.^{121, 122} A prospective cohort of HIV-infected and uninfected women showed no difference in prevalence of vaginal Candida albicans colonization in the absence of immunosuppression. Among women with lower CD4+ counts, the rates of colonization and symptomatic infection tripled.¹²³ The proportion of nonalbicans isolates did not differ among groups and Candida colonization was not associated with antibiotic or oral contraceptive use. In contrast, in a cross-sectional study of patients referred to a vaginitis clinic, HIV was associated with nonalbicans Candida infection.¹²⁴ In a 2-year follow-up study, there was a very slow incremental increase in nonalbicans yeast (mostly Candida glabrata) isolated from the vaginas of HIV-infected women.¹²⁵ This increase may be related to systemic azole exposure among HIV-infected women. In this study, the HIV-infected women reported recent vulvovaginal candidiasis (VVC) approximately 50% more often than did the HIV-uninfected women. A large prospective study (HERS) with 4.5 years of follow-up showed a significantly higher incidence of VVC among HIV-infected than among uninfected women.¹²⁶ The 4.5-year cumulative frequency of vaginal candidiasis in the HIV-infected cohort who did not have VVC at the baseline visit was as high as 64.5% depending on how VVC was defined. The risk of acquiring VVC was elevated among women who had severe immunosuppression with CD4+ cells less than 200 cells/μL and for those with elevated HIV viral load. HAART use, systemic antifungal use, and older age were associated with reduced risk of VVC. Interestingly, antibacterial drug use did not increase the risk of VVC. In this study, VVC was more likely to persist among HIV-infected women but was no more severe than in HIV-uninfected women. There is no evidence from controlled studies that the severity or clinical presentation of vaginal candidiasis is different in HIV-infected compared with HIV-uninfected women.¹²⁷

In a large cross-sectional study, there was a similar prevalence of Trichomonas vaginalis and BV among the HIV-infected and uninfected participants and no association with CD4+ cell count.¹²⁸ A longitudinal study showed a higher prevalence and more persistence of BV among HIV-infected women than HIV-uninfected women.¹²⁹ In this study, prevalence and severity, but not incidence, of BV were correlated with degree of immunosuppression.

Diagnosis and treatment

Diagnosis of vaginitis among women who are HIV-infected is similar to amongst those who are uninfected. A retrospective analysis of clinical and laboratory diagnosis of vaginitis from the WIHS study showed no difference in diagnostic performance between HIV-infected and uninfected individuals.¹³⁰ When vaginal candidiasis is encountered in HIV-infected women, a variety of treatment options are available. These include topical antifungal agents such as clotrimazole, miconazole, terconazole, boric acid vaginal suppositories (600 mg twice daily for 2 weeks), and single-dose, oral fluconazole. For immunosuppressed women who have recurrent vaginal candidiasis, 7–14 days of topical therapy may be effective.¹³¹ Fluconazole (200 mg/week) is also an effective means of prophylaxis in this group.¹³² Unfortunately, there is evidence that this regimen leads to an increase in vaginal colonization with nonalbicans yeast¹³³ and an increase in the proportion of fluconazole-resistant nonalbicans yeast over time.¹²⁵ The decision to use long-term fluconazole, especially for primary prophylaxis, must be carefully considered because of the risk of development of resistance, the effectiveness of therapy for acute disease, the negligible incidence of serious invasive disease, and the possibility of drug–drug interactions.¹³⁴ Clotrimazole powder (100 mg capsules) used intravaginally once weekly reduced the rate of vaginal candidiasis in a cohort of HIV-infected women.¹³⁵

A prospective study showed no difference in the treatment outcome of Trichomonas but a significant decrease in cure rates of BV among HIV-infected compared with HIV-uninfected women.¹³⁶ The treatment recommendations for both infections are the same in HIV-infected and uninfected women.

Summary

Vaginal infections play an important role in transmission of HIV. Treatment of vaginal Candida infections can be challenging in HIV-infected women. Antifungal prophylaxis may be necessary but should be administered judiciously.

A key element of providing care to a HIV-infected woman includes reproductive counseling with provision of contraception or preconception care. Most studies have found that individuals continue to be sexually active after receiving a HIV diagnosis.^{137, 138, 139, 140, 141, 142} The reasons individuals and couples decide to use contraception, conceive, and continue or terminate a pregnancy are complex, particularly in the setting of HIV.^{139, 143, 131, 144, 145, 146, 147, 148}  The current consensus of opinion is that HIV-infected and at risk women can use all available contraceptive methods, however, condoms represent the only method that will significantly reduce risk of HIV transmission or acquisition during heterosexual intercourse.^{149, 150}

Hormonal contraceptive methods are safe and effective in the prevention of unwanted pregnancies. However, there are several issues relevant to the HIV-infected woman. Hormonal contraception appears to be associated with increased genital shedding of HIV-1, a potential surrogate of increased transmissibility.^{151, 152, 153} Mostad and colleagues found an association between hormonal contraceptives and cervical HIV-1 shedding.¹⁵³ In particular, this study detected a significant association between cervical HIV-1 shedding and high-dose oral contraceptives (50 μg estradiol) (adjusted odds ratio [aOR]: 12.3; 95% CI: 1.5 to 101), low-dose oral contraceptives (aOR 3.8; 95% CI 1.4–9.9), and depot medroxyprogesterone acetate (DMPA) (aOR 2.9; 95% CI 1.5–5.7). While no studies have evaluated Norplant and HIV shedding in humans, Marx and associates noted an increase in simian immunodeficiency virus (SIV) DNA-positive cells in the vaginal lamina propria of infected macaques with subcutaneous progesterone implants.¹⁵⁴

Several large, prospective studies have looked at hormonal contraception (both DMPA and oral contraceptives) and HIV-1 disease progression. Baeten reported an association between women using DMPA at the time of HIV acquisition and both a higher baseline viral load and increased viral diversity,¹⁵⁵ which has been shown to be predictive of HIV disease progression and mortality.^{156, 157} As a result, HIV-infected women using DMPA may be at increased risk for faster progression of HIV disease. The US Women's Interagency HIV Study (WIHS) looked at 177 women receiving hormonal medication and 544 women not using contraception, and showed no difference in viral load and slightly increased CD4+ counts in women on hormonal contraception.¹⁵⁸ Stringer et al. looked at disease progression and death among 595 HIV-infected women in Zambia, randomized to either hormonal contraception (DMPA or OCPs) or a copper IUD. The study was designed to evaluate IUD safely among women in the postpartum period, all of whom were not eligible for HAART. Although a large number of women were of lost to follow-up (36%) or switched contraceptive methods (31%), the analysis showed an increase in disease progression to CD4+ count <200 cells/μL (HR 1.6; 95% CI 1.04–2.3) among women on hormonal contraception compared to IUD users.¹⁵⁹ The results remained significant in a secondary analysis of the same study in which the participants were stratified according to type of hormonal contraceptive (DMPA or OCPs).¹⁶⁰ These results, while not definitive, underline the urgent need for further studies specifically designed to determine the safety and efficacy of long term contraceptive methods among HIV-infected women.

In addition to biological implications, there may be behavioral consequences of hormonal contraceptives among HIV-infected women. Women using hormonal contraception, regardless of their HIV status, are often less likely to concurrently use condoms. Not surprisingly, one study found that HIV-infected women using oral contraceptives were significantly less likely to use condoms, putting them at risk for acquiring STIs and transmitting HIV to uninfected partners.¹⁶¹

Lastly, there are interactions between the ethinyl estradiol (EE)/norethindrone acetate (NE) in OCPs and some of the antiretrovirals. In particular, amprenavir, atazanavir, efavirenz, and indinavir all appear to increase levels of EE/NE.¹⁶² However, lopinavir/ritonavir, nelfinavir, nevirapine, tipranavir, and ritonavir decrease levels of EE/NE, likely by inducing CYP450 metabolism.^{162, 163, 164} There is also limited evidence that EE/NE may alter levels of antiretrovirals. Concurrent administration of 0.035 mg EE and 1 mg NE for instance, is associated with decreased levels of amprenavir and tipranavir.^{162, 165} No significant dose alterations have been reported with concomitant administration of medroxyprogesterone and antiretrovirals.¹⁶² Although the pharmacokinetic studies show decreased hormonal levels, there is a paucity of clinical data measuring actual contraceptive efficacy to provide an evidence-based recommendation for or against specific hormonal contraception and HAART combinations. The US Department of Health and Human Services and the World Health Organization recommend use of a second, nonhormonal contraceptive method if combined oral contraceptives are given to women taking nonnucleoside reverse transcriptase inhibitors or protease inhibitors.^{1, 166}

While there are theoretical concerns of the use of an intrauterine device (IUD) among HIV-infected women, no study has found an increase in viral shedding or incident infections associated with its use.¹⁶⁷ Among 98 HIV-infected women in Kenya, there were no differences in HIV-1 DNA genital shedding before and after insertion of a copper-containing IUD.¹⁶⁸ Sinei and associates found no significant differences in overall complications, incident pelvic inflammatory disease, and overall infection-related morbidity between 156 HIV-infected and 493 HIV-uninfected women who underwent IUD insertion.¹⁶⁹ Stringer and associates also found no increased risk of PID among women with an IUD in place, compared to hormonal contraceptive users.¹⁵⁹ The levonorgestrol-containing IUD has also been used with success in HIV-infected women. A study with a 2-year follow-up of 12 HIV-infected women with levonorgestrol-containing IUDs showed reduced menstrual bleeding, biochemical evidence of suppressed ovulation, and no change in HIV-1 genital shedding after placement.¹⁷⁰ While it is often difficult to find the ideal method of contraception for any woman, it is crucial to assist the HIV-infected woman in weighing the risks and benefits of all options and choosing a realistic method for her circumstances.¹⁷¹

Fertility desires among HIV-infected women have become the subject of a growing body of literature, most of which suggests that HIV alone does not decrease women's or couples' desire for childbearing.^{131, 144, 172} However, HIV-infected women's desire to conceive is often overlooked, and many women feel that their fertility desires are not adequately addressed by their physicians.^{173, 174} In the 1980s, both the CDC and the American of Obsetrics and Gynecology recommended that physicians counsel their HIV-infected patients to avoid pregnancy and to inform them of termination options if they became pregnant.^{175, 176} Reasons included the life-threatening nature of HIV and the unacceptably high risk of HIV transmission to the infant. Dramatic advances in antiretroviral development and efficacy have transformed HIV into a chronic condition and reduced the risk of perinatal HIV transmission to less than 1% among women with maximally suppressed virus at the time of delivery.¹⁷⁷ In 2001, the CDC changed its recommendations to state that HIV-infected women should receive information about all reproductive options and counseling should be nondirective and supportive of the patient's decision.¹⁷⁸ 

While optimizing the HIV-specific needs of infected women is clearly outside the realm of most gynecologists, there are several issues that obstetricians/gynecologists can and should address with their HIV-infected patients considering pregnancy. Goals of preconception care include identifying risk factors for adverse maternal and neonatal outcomes, treating and stabilizing medical conditions and psychosocial problems before conception, and providing patient-specific health education.

Clearly, the provision of high-quality preconception care requires eliciting a thorough sexual and reproductive history, including goals and plans for conception. One important aspect of this care is to ensure disclosure of HIV status and encourage partner testing to help reduce the potential risk of HIV-transmission during conception. For serodiscordant couples in which the male partner is HIV-negative, low-cost techniques can be used to successfully achieve conception without risk of HIV transmission. Prior to attempting conception, men should undergo a semen analysis to exclude male factors that would adversely affect fertility. Self-insemination with a needleless syringe or diaphragm during ovulation can be used successfully to achieve pregnancy without HIV transmission. Clinicians should counsel couples that timed conception with unprotected intercourse during ovulation may result in HIV transmission and is not recommended.¹⁷⁹ If couples choose to pursue this strategy, clinicians should reinforce the use of condoms during other sexual encounters and discuss the benefit of achieving maximal viral suppression on antiretroviral therapy to reduce the risk of transmission to negative partners.¹⁸⁰ 

Serodiscordant couples who are unable to conceive using low-cost techniques as well as women with sub- or infertility should be counseled about the availability of assisted reproducive technology for the prevention of horizontal transmission, including artificial insemination and in vitro fertilization. The increased awareness of the fertility desires among HIV-infected couples has resulted in a growing body of literature addressing the ethics and science of providing assisted reproduction techniques to HIV-infected women and HIV-serodiscordant couples.^{181, 182, 183, 184}  Recently, the Ethics Committee of the American Society for Reproductive Medicine stated, “Unless health care workers can show that they lack the skill and facilities to treat HIV-positive patients safely or that the patient refused reasonable testing and treatment, they may be legally as well as ethically obligated to provide requested reproductive assistance.”¹⁸²

Medications should be reviewed and teratogenic agents discontinued or changed. Overall, the use of HAART during conception and pregnancy has not been associated with an increase in congenital birth defects in registry data from United States and Europe.^{185, 186, 187}  However, medications with teratogenic potential in animal studies or case reports include the antiretrovirals efavirenz, amprenavir, and lipid-lowering agents such as HMG-CoA reductase inhibitors.^{188, 189, 190} Exposure to zidovudine (AZT) in the first trimester has been associated with an increased risk of hypospadias in one long-term cohort, however, that relationship has not been reproduced in larger studies.^{191, 192} Women using stavudine and/or didanosine should switch to another antiretroviral regimen, if possible, given the association between these medications and fatal maternal lactic acidosis.^{193, 194, 195, 196}

Observational studies have shown an increase in preterm delivery among women receiving HAART,^{197, 198, 199} and a meta-analysis of the literature suggests that the risk is greater among women receiving HAART prior to conception and during the first trimester.²⁰⁰ As with all observational studies, it is difficult to control for confounding factors among these different groups of women, so we would not recommend withholding HAART prior to conception or in the first trimester when it is indicated for maternal health. 

In addition to optimizing the medication profile, it is important to confirm that all vaccination series are up-to-date. Both hepatitis A and hepatitis B vaccines are safe among HIV-infected individuals. The measles-mumps-rubella (MMR) and varicella vaccine are both live-attenuated viruses, however, the Advisory Committee on Immunization Practices (ACIP) recommends vaccination for all susceptible, asymptomatic HIV-infected individuals in the absence of severe immunosuppression.^{201, 202} 

Due to shared transmissibility through blood and blood products, approximately 30% of HIV-infected women in the United States and Europe are co-infected with hepatitis C.²⁰³ Several studies have shown that antiretroviral use increases the risk of cirrhosis and overall mortality compared to women not on medication.²⁰⁴ Coinfection with HIV and hepatitis C increases the risk of perinatal transmission of both viruses, so coinfected women should be seen by an experienced hepatologist prior to pregnancy to optimize medication and control of viral load. ^{205, 206, 207}

As with all women with chronic medical conditions, ideally pregnancies are planned while in the care of a medical provider. Preconception counseling should be an ongoing process that should be integrated into the gynecologic and primary care of the patient, rather than taking place at a single visit. Providers with questions about preconception counseling and antenatal care of HIV-infected women can obtain free clinical advice and referral resources from the National Clinicians' Consultation Center at www.nccc.ucsf.edu or the NCCC Perinatal HIV Consultation & Referral Service at 1-888-448-8765.

HIV-infected women frequently report changes in their menstrual cycles.^{208, 209, 210, 211, 212} It is not clear, however, whether HIV, per se, is a risk factor for such disorders, and several studies have found no difference in menstrual patterns between HIV-infected and uninfected women.^{210, 213} Menstrual abnormalities can be related to conditions associated with HIV, including weight loss, comorbid psychiatric illnesses and psychotropic medications, illicit substance use, thrombocytopenia, and renal dysfunction. Chirgwin and colleagues demonstrated a higher prevalence of amenorrhea among 248 HIV-infected versus 82 high risk HIV-uninfected women, after adjusting for age, ethnicity, and drug use (aOR 7.1; 95% CI 1.1–1000).²⁰⁸ Another study found a higher frequency of amenorrhea among HIV-infected women with late wasting (38%), defined as less than 90% ideal body weight, as compared with infected women without wasting (17%) and healthy controls (5%).²¹⁴ In the large combined HERS and WIHS prospective cohorts, polymenorrhea, defined as fewer than 19 days between cycles, was seen more often among HIV-infected women (aOR 1.45; 95% CI 1.00–2.11) as compared with HIV-uninfected women, adjusting for age, ethnicity, body mass index, psychotropic medication, and drug use.²¹³ Polymenorrhea was particularly evident among women with HIV viral load values above 168,000 copies/mL. A separate analysis of the same cohort found an independent association between psychotropic medication and both amenorrhea and polymenorrhea.²¹⁵ Not surprisingly, they also found an association between both methadone maintenance and injection drug use and amenorrhea. One study has shown a greater prevalence of amenorrhea (19%) among women with severe immunosuppression (CD4+ count <100 cells/uL) when compared to women with CD4+ counts between 200 and 500 cells/uL (9%).²¹⁶  Regardless of whether HIV is independently associated with menstrual disorders, prevalence of such menstrual irregularities is high; therefore, it is crucial to enquire about menstrual function among HIV-infected women seeking care.

The widespread availability and advances in HAART have resulted in markedly increased life expectancy for HIV-infected women. Women are now living through menopause and beyond, so healthcare providers should have an understanding of the effects of HIV on menopause. Although studies are limited, existing research suggests that the process of menopause may be affected by HIV infection.^{217, 218} Results of the WIHS Study show that HIV-infected women are more likely than HIV-uninfected women to have prolonged periods of amenorrhea without ovarian failure,²¹⁹ making the investigation of age at menopause difficult. HIV-infected women may experience menopause at an earlier age, making them more susceptable to menopause related health risks.^{220, 221} Although some studies show no difference in the experience of perimenopausal symptoms related to HIV status, several studies suggest that HIV-infected women experience and report greater symptomatology than HIV-uninfected women.^{222, 223} HIV infection also appears to affect bonemineral density, cardiovascular disease, and cognition.²²⁴

There is a growing body of literature addressing metabolic alterations among HIV-infected individuals. While most of the attention has focused on changes in central fat accumulation and peripheral subcutaneous fat loss, researchers are now turning their attention to altered bone metabolism.^{225, 226, 227, 228} A recent meta-analysis of studies showed that HIV-infected women had a 6-fold greater odds of decreased bone mineral density and an almost 4-fold greater odds of osteoporosis than HIV-uninfected women.²²⁹ The etiology of bone loss among HIV-infected women is multifactorial, and may include increased bone resporption by proinflammatory cytokines, as well as weight loss, hypogonadism, and vitamin D deficiency associated with HIV disease. HAART, specifically protease inhibitors, can cause a decrease in bone mineral density, which is thought to be one of the main mechanisms for increased osteopenia and osteoporosis among infected women.^{230, 231, 232} However, most studies are unable to control for confounding factors, such as comorbidities, CD4+ count, and overall immune status.²²⁹ Fracture prevention should be a priority in caring for HIV-infected women during menopause. Lifestyle modifications, such as reduction in alcohol comsumption and increased weight bearing exercises should be encouraged. This population may benefit from early bone mineral density (BMD) testing and treatment as indicated.²³³ Bisphosphonate alendronate and calcium have been shown to be effective in treating HIV-associated bone loss, and may be a good option for perimenopausal women with low BMD who are at increased fracture risk.^{234, 235} As HIV-infected women live longer lives and more of these women reach menopause, the clinical management of osteopenia and osteoporosis in this population will become only more relevant.

Recent well-designed prospective studies have shed some light on the interactions between HIV and gynecologic diseases, but our understanding is far from complete. The most significant challenges ahead seem to be the gaining of better understanding of sex differences in HIV disease progress and the effects of HIV on aging along with the development of better treatment methods for lower genital tract neoplasia. Gynecologic diseases are frequently encountered when caring for HIV-infected women, especially those with greater degrees of immunosuppression. Most gynecologic conditions in HIV-infected women will respond to standard therapy but require increased vigilance for treatment failure. Some will require multiple courses of treatment or use of innovative treatment methods. All HIV-infected women deserve careful evaluation for and treatment of gynecologic diseases.