Published in the Bulletin of Experimental Treatments for AIDS 1999 Year-End issue, by the San Francisco AIDS Foundation.

Global Perspectives on Women, HIV, and AIDS

When Should Women Take Viral Load Tests?

Women & HIV: The Experts Speak

Year-End 1999 Table of Contents

Main Page

beta@sfaf.org

Women & HIV: Conferences Raise New Questions about HAART Use by Women

Leslie Hanna

This review article discusses selected major medical issues related to HIV positive women's health in the context of the year-and era-drawing to an end. To the extent that the goal is the provision of a broad perspective on trends and developments, an overview of both recent and historically significant data is provided.

Most of the following reports are based on information presented during the course of two extremely important meetings for women with HIV, both of which took place during the autumn. The 1999 National Conference on Women and HIV/AIDS was held in Los Angeles in October, and the Second Conference on Global Strategies for the Prevention of HIV Transmission from Mothers to Infants took place in Montreal in September. Much of the data also have been published in peer-reviewed journals.

[Ed. note: Full coverage of the topics presented at the combined conferences -- let alone full coverage of all currently important topics for women -- is not possible in a single article. Social and political factors that are germane to the women's epidemic are not a focus of this article. Upcoming issues of BETA will feature in-depth reports of selected news items mentioned below, including adherence and gynecologic conditions, among others.]

In the U.S., state and local health departments in concert with the Centers for Disease Control and Prevention (CDC) Division of HIV/AIDS Surveillance and Epidemiology monitor the epidemic using information from anonymous HIV infection prevalence surveys, results of blood testing, and reports of death due to HIV/AIDS. Eighteen years ago, the first five AIDS cases in the U.S. were diagnosed among five gay men in Los Angeles. Now, over half a million cases have been reported in this country, including Puerto Rico and the Virgin Islands, and an increasingly large proportion of cases are reported among minorities and women.

As of December 31, 1998, 16% of total reported U.S. AIDS cases (109,311 of 688,200) were among women 13 years of age or older. In the U.S., women continue to account for an ever-widening margin of total AIDS cases, from 7% in 1985 to 24% in 1998. Between January 1994 and June 1997, 17% of all AIDS diagnoses and 28% of all HIV diagnoses were in women, confirming the predicted trend toward an increasing proportion of female AIDS cases. (Since to date HIV diagnosis reporting has not been required in New York or California, states with significant prevalence, 28% is probably a low figure.)

Also on the increase is the proportion of cases attributable to heterosexual contact: in 1994, for the first time, AIDS cases in women attributable to sex with a man surpassed the number attributed to transmission through injection drug use. Most cases attributed to sexual contact with a man involved a man who injects or has injected drugs. Thus, at the close of the century in the U.S., what was once primarily a condition affecting gay and bisexual men is shifting as women--especially women of color--are increasingly affected.

HIV/AIDS is the third leading cause of death among all women age 25-44 and the leading cause of death in Black women in America. Although Black and Latina women account for less than 25% of U.S. women, they represent 77% of reported AIDS cases in women (according to CDC midyear estimates through June 1999). The leading cause of death in New York City for women age 25-44 is HIV.

In the Southeast, the U.S. region experiencing the highest proportional increases among women of both HIV infection and AIDS cases, many HIV positive women live in small communities or rural areas. These women are likely to be Black or Latina and especially likely to have acquired HIV through heterosexual transmission (as opposed to injection drug use [IDU]).

Wafaa El-Sadr, MD, of Harlem Hospital in New York discussed conflicting data in the scientific literature as to rates of HIV disease progression and survival among women, after referencing recent data indicating that HIV/AIDS is severely impacting U.S. women. In brief, studies that showed worse outcomes for women showed that differences in access and treatment regimens were most likely the causes.

To better understand the implications of such mixed results, Dr. El-Sadr and others looked at access issues for women in the context of antiretroviral therapy. The analysis was a review of data from four Community Programs for Clinical Research on AIDS (CPCRA) studies, in which both men and women had similar access to care and were randomized to various regimens (mono, dual, or triple antiretroviral therapy) and similar follow-up. Across all regimens, no significant differences were noted in rates of disease progression or death between men and women.

Results were consistent over time. Rates of disease progression and death among both men and women were influenced by baseline CD4 cell count and type of treatment regimen (essentially, those on combination therapy fared better than those on monotherapy). However, Dr. El-Sadr said, studies with larger numbers of women are needed to determine accurately whether gender influences the outcome of antiretroviral therapy.

Access problems disproportionately affect women with HIV in the U.S. for many reasons. Compared with men, women are more likely to have less formal education and less knowledge of HIV, as well as language barriers and financial issues. Women are less likely to have a regular source of health care and are frequently unaware that they have any HIV risk factors. Women account for a higher proportion of so-called marginalized people with HIV, such as the homeless. Finally, women must contend with the social barrier of sexism and, often, dual barriers of sexism and racism.

Kenneth Mayer, MD, of the Brown University AIDS Program, noted that women continue to have relatively higher rates of HIV-related illness and death because they continue to have less access to health care than men. He emphasized that the key to improving outcomes in women is improving access because "when women have access, they do as well [as men] on antiretroviral medications and experience fewer HIV-related illnesses."

While it is clear that antiretroviral therapy and preventive treatment, especially Pneumocystis carinii pneumonia (PCP) prophylaxis, prolong survival in people with HIV disease, studies like one conducted by Michael Stein and others have shown that women, non-Whites, and injection drug users are 33-50% less likely to be offered antiretroviral therapy.

At the 12th World AIDS Conference in July 1998, Valerie Stone of the Memorial Hospital of Rhode Island and others reported that treatment with regimens containing protease inhibitors (PIs) was less likely to be given to women and non-Whites, and with respect to mode of HIV acquisition. That is, 73% of men compared with 50% of women who met the criteria for highly active antiretroviral therapy (HAART) treatment actually received it. Seventy-five percent (75%) of Whites were taking HAART compared with 58% of Blacks. Furthermore, 81% of those who reported male homosexual or bisexual sexual acquisition received HAART, compared with 50% of women who reported heterosexual acquisition and 51% of those who reported IDU.

Access to care and antiretroviral treatment for women is improving. Yet for many women with HIV, due in large part to lower socioeconomic status and poor insurance status, access to HAART remains elusive. In addition, many doctors continue to make false assumptions about women's ability to adhere to HAART, often failing to recommend or prescribe it.

Lisa Hirschhorn, MD, of Demet Community Health Center in Newton, MA, introduced a discussion of the adverse effects of antiretroviral therapy in women by noting that not all drugs have the same toxicities in all people. Risk factors for increased side effects in both women and men include liver or kidney disease, increased age, genetics (some people metabolize some drugs such as isoniazid for tuberculosis at different rates), and use of other medications.

"There are many potential causes of gender-related differences in side effects, which do exist," Dr. Hirschhorn stated. Since many drugs are approved based on studies in adults that do not test different doses specifically in relation to gender, recommended (i.e., approved) doses usually apply to both women and men, a standardization that may cause problems. A gender analysis of ACTG 175, a double-blind, placebo-controlled, randomized clinical trial that looked at a variety of anti-HIV medications, found that women, especially younger women, needed to reduce their doses of ddI (Videx) because of greater adverse effects. Specifically, the study compared AZT (Retrovir) monotherapy to combination AZT/ddI and AZT/ddC (Hivid). This was the first preplanned analysis of gender differences. Of 2,467 study participants, 438 (18%) were women.

A major finding was that time to first dose modification occurred earlier for women than men when the regimen was ddI monotherapy, and especially when those taking it were antiretroviral-naive. Women also were more likely to discontinue ddI and AZT/ddI therapy. There were no overall gender differences found in efficacy.

Biological factors that mediate such differences in women include body composition and weight (i.e., some small or lower-weight women may in effect be receiving double doses); metabolic differences in the CYP450 system, which influences how drugs are broken down and eliminated; hormonal differences, not only between women and men but also among women (for instance, pre- versus post-menopausal women); and the frequency of other, concomitant complicating factors (e.g., a higher rate of anemia in women influences the complication rate of women as a group to AZT, or to hydroxyurea [Hydrea]).

Indirectly, psychosocial factors also influence side effects in a gender-related manner. For example, women tend to have more depression than men. A poster presentation at the National Conference on Women and HIV/AIDS (P113) by Sandra Trusdale, PhD, noted that clinical depression in HIV positive women "drastically affects quality of life and greatly increases the chance of nonadherence to medications. Yet many women living with HIV and their providers do not recognize depression and are unaware of appropriate treatment," which may include Western medications, complementary and alternative therapies, and psychotherapy.

Elsewhere, several women taking efavirenz (Sustiva) have reported disturbing "flashbacks" as well as nightmares, a known central nervous system side effect, that involved traumatic memories of physical abuse and domestic violence. (Researchers are currently evaluating roles domestic violence may play in HIV positive women's health care, from regular access to adherence to mortality.) Men also have reported nightmares, a side effect that generally can be treated through (i.e., it diminishes and resolves), in both men and women.

Recently, elevated lactate levels and lactic acidosis have been reported in people taking nucleoside analog (NRTI) antiretroviral drugs. Women, particularly overweight women, appear to be more likely to develop lactate elevations. At the 1999 Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC; see "Conference Coverage" in this issue), D.E. Boxwell and others presented on 60 cases of lactic acidosis, which included muscle cramping and hepatic steatosis (fat in the liver, in 69% of the 60) in persons taking NRTIs. Regimens in use were combinations of d4T (Zerit), 3TC (Epivir), ddI (Videx), and AZT (Retrovir). Of the 60 who developed lactic acidosis, 83% were women. Among 30 of the 60 who were obese, 15 were women.

Other side effects included nausea and vomiting, abdominal pain, weight loss, and shortness of breath. Also at ICAAC, L. Maulin and a team of French researchers presented on 11 cases of lactate elevation in women who had been taking d4T and concomitant PI therapy for about 11 months. Symptoms included asthenia (physical weakness), weight loss, abdominal pain, and shortness of breath. Eight developed peripheral neuropathy (numbness or tingling in the feet and sometimes hands) and four developed steatosis. Nine discontinued antiretroviral therapy, six took riboflavin, and four took carnitine. (Some evidence suggests that these complications from NRTI anti-HIV drugs are due in part to a vitamin deficiency.) Symptoms resolved in 10 of the 11.

Data from other clinical trials suggest that women are more likely to develop rash from taking nevirapine (Viramune). In the Abbott 247 study, which saw a survival benefit from ritonavir (Norvir) versus placebo in people with advanced HIV, investigators also noted that women were more likely to discontinue both ritonavir and placebo due to side effects of nausea and vomiting. After controlling for differences in the placebo group, researchers concluded that there were few significant gender effects related to ritonavir. Higher rates of hair loss, white blood cells in urine, and kidney stones, most of which resolve within several weeks of stopping the drug, have been reported in women taking indinavir (Crixivan), compared with men. Higher rates of side effects from ritonavir and indinavir appear possibly related to higher blood levels of each drug.

In terms of the still emerging side effects associated with body fat abnormalities ("lipodystrophy") and metabolic disturbances, Dr. Hirschhorn said, "There is no evidence yet to suggest that women are developing them more than men, but we are getting clues that women are experiencing different changes than men." For instance, increased breast size, rarely seen in men, is fairly common in women taking antiretroviral therapy (specifically protease inhibitors [PIs]) and represents lipodystrophy. Women also may be less likely than men to develop hyperlipidemia (elevated plasma blood lipid levels) and lipoatrophy (fat wasting or "loss"), and slightly more likely to develop truncal obesity.

To date, the etiology (causes) of this emerging syndrome appears to be multifactorial. No single factor can explain the condition and there are likely several that contribute, including antiretroviral history, age, and gender. Baseline metabolic abnormalities seen in HIV positive women may be unrelated to PI therapy but exacerbated by PI-related immune recovery and associated weight gain. However, this does not rule out a direct association with PIs.

Dr. Hirschhorn explained that side effects are the most common reason people report for decreased adherence and for discontinuing PIs. She described one urban clinic at which 37% of women versus 25% of men reported discontinuing PI therapy because of side effects. In large part, adherence in women is influenced by unique psychosocial factors; effective interventions must address the multidimensional needs of women. (See "The Challenge of Adherence" in the January 1999 issue of BETA.)

At the National Conference on Women and HIV/AIDS, investigators from the Women's Interagency HIV Study (WIHS) presented data on the prevalence of anemia in this cohort of women. WIHS, a prospective, longitudinal cohort study of HIV positive and at-risk HIV negative women, has enrolled over 2,600 women since 1994; the six sites are the Bronx/Manhattan, Brooklyn, Chicago, Los Angeles, Washington, DC, and San Francisco. WIHS's objective is to evaluate and better understand the spectrum of HIV disease in women.

Anemia is the most common hematologic (related to blood) abnormality in people with HIV/AIDS, estimated to occur in about 20% of new cases and in 65% of people with advanced HIV disease. Anemia may occur as a medication side effect, as a result of HIV-related illnesses, or as a consequence of HIV itself. A unique issue for women is menstruation, which may contribute to the development of anemia. Prior studies of anemia in HIV have enrolled mostly or exclusively men. In this study, 2,625 (2,056 HIV positive and 569 HIV negative) women were evaluated at the first visit. Anemia was defined as a hemoglobin level less than 11 g/dL (grams per deciliter).

HIV positive women not surprisingly were found more likely than HIV negative controls to have anemia. On multivariate analysis significant risk factors included a self-reported history of clinical AIDS; CD4 cell counts less than 200 cells/mm3; high viral load (risk is increased by 1.8 at 49,000-99,999 copies/mL and increased 2.2-fold when greater than 100,000 copies/mL); current AZT use; and mean corpuscular volume (MCV) of red blood cells lower than 80, suggesting iron deficiency. Further analysis of the WIHS cohort will consider the impact of anemia in HIV positive women. In general the data showed roles for low MCV and hemoglobin in anemia, but did not control for conditions such as thallasemia (a type or types of anemia that are hereditary, occurring most commonly in people of Mediterranean, African, or Southeast Asian ancestry) or sickle cell disease.

At the National Conference on Women and HIV/AIDS, Alexandra Levine, MD, and others in a poster presentation (P31) described positive results in correcting hemoglobin levels and in self-reported quality of life among HIV positive women with anemia who took once weekly epoetin alfa (Procrit) at a dose of 40,000 units subcutaneously for 16 weeks. (The poster discusses preliminary, eight-week data.) Once weekly epoetin alfa effectively resolved anemia in women of all disease stages regardless of whether or not they were taking AZT. The drug also was well tolerated and safe at the dose tested.

In her introduction to a panel session on routine testing to monitor the efficacy of antiretroviral therapy, Vivian Torres reminded those at the National Conference on Women and HIV/AIDS of the importance of routine monitoring of CD4 cell count and viral load. She also recommended regular testing of hemoglobin levels to make sure that values are normal (12-16 g/dL). On behalf of the Black Women's Health Project, which seeks to support women to receive regular health care, she recommended that women take a proactive stance toward their own health care, quipping, "We are the ones we've been waiting for."

Kathryn Anastos, MD, of the Catholic Medical Center in Jamaica, NY, described a "two-fold miracle in late 1995-early 1996 that changed things for individuals with HIV: protease inhibitors and viral load tests." Today, viral load monitoring is part of normal clinical testing, the purpose of which is to optimize an individual's care. Dr. Anastos used a popular analogy to describe the roles of HIV monitoring tests: if HIV disease is likened to "a train approaching a cliff's edge, [then] T-cells tell you how close to the edge of the cliff it is, and viral load tells you how fast it's approaching the edge."

Much of what is currently known about viral load testing is based on data gathered from men, particularly the Multicenter AIDS Cohort (MACS) study and influential data published by John Mellors, MD, in the Annals of Internal Medicine in 1997. The basic concept gleaned from that study was the higher the viral load, the sicker the person and the faster the progression to AIDS and death. In Dr. Mellors' study, 100% of people with a viral load over 10,000 copies/mL and fewer than 200 CD4 cells/mm3 became sick or died (based on ten-year data). Therefore, particularly in conjunction with CD4 cell information, viral load information allows estimation of how quickly or slowly a person will develop an opportunistic condition (infection, cancer, or wasting) or die.

A study published by Dr. Anastos and others earlier this year (see "Recent News about Women and HIV" in the April 1999 issue of BETA) raised important questions about the use of viral load information in antiretroviral treatment decision-making. At the 6th Conference on Retroviruses and Opportunistic Infections (CROI) in February 1999, WIHS data was presented in what may be viewed as the first of a second wave of studies on viral load and disease progression (whereas the MACS data may be viewed as early or first-wave data).

In this analysis, investigators compared viral loads in women and men at all disease stages. A slight difference was found at every CD4 cell level, with men having slightly higher viral loads than women did at every phase (i.e., fewer than 200, 200-349, 350-500, and greater than 500 CD4 cells/mm3). Investigators also looked at the potential roles of other factors such as race and history of IDU, and found that people with a history of IDU tended to have lower viral loads at respective stages of illness.

The main question raised by the study's findings is whether women need to adjust viral load test results relative to men, especially when determining when to start antiretroviral therapy, because of apparent differences in viral load thresholds and disease progression and staging. Issues and differences related to that analysis relative to the MACS data remain to be resolved; for instance, women in the WIHS cohort were slightly older than men in MACS, and age is an independent factor for increased disease progression. Also, the WIHS women were more likely than MACS men to have a history of IDU and were clinically sicker. Dr. Anastos therefore emphasized that "all treatment guidelines [as of late 1999] are based on data derived from studies of men but there are not yet cogent data suggesting that the guidelines should be different for women."

Many studies evaluating gender and viral load have been hampered by unknown dates of seroconversion. Recent data on gender and viral load from the AIDS Linked to the Intravenous Experience (ALIVE) cohort study published in the Journal of Infectious Diseases by Timothy Sterling, MD, and others suggest that differences between men and women are present from the beginning of HIV infection.

This study followed viral load in injection drug users who seroconverted to evaluate gender differences both at seroconversion and over time. Basically, at baseline men had a median viral load of 148,354 copies/mL compared with women, who had a median viral load of 14,918 copies/mL. (Differences remained after controlling for age at seroconversion and CD4 cell count.) Yet while women began with lower post-seroconversion viral loads, they also displayed steady and significantly greater viral load increases over time than did men. Within five years, at follow up, median viral loads had become equivalent between men and women.

What happens in the longer run--beyond five years--is still unclear; more data are needed. In terms of CD4 cell counts, women's were higher than men's. (In HIV negative adults too, women's CD4 cell counts are on average about 100 cells/mm3 higher than men's, regardless of race). These observations highlight the problem of grouping people with HIV according to CD4 cell levels, which assumes duration of disease: stratifying women by CD4 cells along with men may obscure significant gender differences.

In a poster presentation at the National Conference on Women and HIV/AIDS, H.L. Cheng and R.M. Gulick remarked that "it is possible or even likely that women will respond 'better' to treatment, because of their lower [initial] viral loads." Preliminary data from AIDS Clinical Trials Group (ACTG) study 359 suggest that this may be so; Cheng and Gulick reported that indinavir-experienced, NNRTI-naive men and women who were "well matched except for age" (median age was 40 for men and 36 for women) and took one of six saquinavir soft-gel capsule (Fortovase) salvage regimens experienced similar time to first viral suppression (defined as fewer than 500 copies/mL, compared with a median entry viral load of 32,000 copies/mL) in pooled ritonavir or nelfinavir (Viracept), delavirdine (Rescriptor), or delavirdine/adefovir (Preveon) arms. However, a significant gender difference was found in the pooled adefovir arm, with women having a much better virologic response (P91). [Ed. note: On November 1, an advisory committee of the U.S. Food and Drug Administration (FDA) failed to recommend approval for adefovir. See "News Briefs" in this issue.]

Obviously, more research is needed to tease out gender differences. Dr. Anastos and others are continuing to study viral load in women to elucidate possible gender differences and treatment implications. "Right now, however," Dr. Anastos concluded, "treatment recommendations should be made without regard to race or gender, but it is imperative that the research community continue to evaluate these issues."

Today, there are three primary factors that guide recommendations to begin treatment: viral load, CD4 cell count, and clinical disease. Most people with viral loads greater than 10,000 copies/mL also have fewer than 500 CD4 cells/mm³. Some people have discordant monitoring test results (i.e., high viral load and high CD4 cell count or low viral load and low CD4 cell count), and more research is needed to understand the relationships between the two and with clinical health.

New tools available for people with HIV to use in making treatment decisions are genotypic and phenoptypic resistance tests, both of which are still being studied to learn how best to use them. Victoria Johnson, MD, outlined possible clinical situations for HIV-1 drug resistance testing in women.

In addition to those uses relevant to all HIV positive adults (including testing for primary HIV infection in areas with a high prevalence of drug-resistant virus and for selecting salvage therapy in antiretroviral-experienced persons), she emphasized the possible utility of resistance testing in pregnant women, both antiretroviral-naive and antiretroviral-experienced, to help protect infants from vertical transmission of resistant virus. She described a related, growing wave of interest in measuring and comparing viral load in cervicovaginal fluids with plasma viral load, tools that may be particularly useful in pregnancy. (See "Genotypic and Phenotypic Resistance Testing" in the Summer 1999 issue of BETA.)

Dr. Johnson raised as a public health issue "the reconsideration of the blanket use of AZT in pregnant women" with respect to concern with vertical transmission of multidrug-resistant HIV-1. (Variants can occur in the setting of incomplete suppression of maternal plasma viremia.) Potential contributing factors include advanced maternal HIV disease with high viral load, drug interruptions and adherence problems, and prior use of combined AZT and d4T. Furthermore, she emphasized the importance of monitoring for primary (maternal) drug resistance.

Dr. Johnson recommended continued study to monitor and evaluate drug resistance in perinatal settings. Yet she added, "I'm ambivalent about the great deal of commercialization going on right now in this field [resistance testing]. Anyone who stands up and tells you we know exactly how to use these tests is potentially commercially biased, in my opinion."

According to the Joint United Nations Programme on HIV/AIDS (UNAIDS), women age 15-49 currently account for 43% of all new HIV infections (up from 40% a few years ago). Cumulative evidence indicates that women are about four times as susceptible to HIV infection as heterosexual men, based on biological/anatomical factors. (Men or women who have receptive anal sex are probably most vulnerable to HIV.)

In many societies, women are at increased risk because lower social or economic status renders them dependent upon a husband or male partner or places them in a diminished position to request that a husband or male partner be faithful or use a condom, the most widely available method to prevent HIV transmission during intercourse worldwide.

Many studies have documented that a majority of women are infected by their sole sex partners, usually their husbands. Other women with more than one partner include those who are "driven by economic necessity into sex work" and who may fear loss of income or violence, and are at risk for acquisition of HIV through multiple, often unprotected, sexual exposures. These observations highlight the need for the development of an effective, female-controlled protection method so that women may protect themselves.

Ultimately, the most important long-range goal of HIV education and prevention involves the prevention of infection. To that end, several strategies are being pursued, including most notably vaccines and microbicides. Because vaccines may not be equally effective at all mucosal sites, for women in particular the ability to combine a vaginal microbicide with a vaccine may afford a better level of protection. (See the "Status Report on HIV Vaccine Development" in this issue.)

Although female-controlled prevention methods are urgently needed, only one option exists today; but while the so-called female condom has been marketed for several years, it is prohibitively expensive for many women. Vaginal microbicides, still being researched and developed, are extremely promising because of the many characteristics a strong candidate will likely have: nondetectibility (including lack of smell and taste), safety and efficacy, antimicrobicidal properties (against HIV and other causes of STDs), and the option of including a non-inflammatory spermicide or not.

The product may be formulated as a cream, gel, film suppository, vaginal ring, or diaphragm. Despite research setbacks such as disappointing 1997 results of a large nonoxynol-9 film trial that showed no protective efficacy compared to placebo, today an estimated 23 candidates are in early clinical testing, and more than 50 compounds are being evaluated for their clinical potential.

At the National Conference on Women and HIV/AIDS, Kenneth Mayer, MD, presented results of a phase I trial of BufferGel, which (as its name indicates) is a gel and a leading microbicide candidate. BufferGel is designed to reduce vaginal pH below 5 in the presence of semen, i.e., to create a more acidic environment and thereby inhibit HIV replication. In the Rhode Island study, 27 "low-risk abstinent" (16) or sexually active monogamous (11) women used the product once or twice a day for 14 days. Male partners used condoms. Eighteen women reported predominantly mild adverse effects, mostly vaginal itching. Three women developed vaginal candidiasis and one developed a minor lesion; these women withdrew from the study. Adverse effects did not appear dose-related.

After seven days, median baseline pH decreased from 4.6 to 4.01 and was about 4.3 after 14 days of use. In conclusion, the product appeared effective in reducing vaginal pH, nontoxic, and generally well tolerated in low-risk U.S. women. Investigators reported that the women in the study said they would use the product if it were available (on the market). Additional phase I trials are underway in Malawi, Zimbabwe, Thailand, and India.

Other candidates include Savvy, a clear gel that acts as a spermicide and microbicide, and PRO 2000 gel, which blocks gonorrhea and chlamydia in the test tube as well as HIV. Human tests are underway.

Dramatic progress in understanding and preventing perinatal HIV transmission (PHT) has been made over the past five years, largely through laboratory, epidemiological, and transmission trials. In the U.S. and Europe, the ACTG 076 trial (see pages 68-74 in the September 1994 issue of BETA) alone led to dramatic declines in PHT rates, especially since 1994. Currently 300 infants in the U.S. and 100 in Europe are infected each year. Progress in reducing PHT rates in developed countries has been so great that today, at the end of the millennium, experts have raised as a real possibility the idea of eliminating mother-to-child transmission in the U.S.

For women in developed nations, issues today include AZT pretreatment and resistance and the unknown teratogenic effects of HAART. Most women already taking combination therapy when they become pregnant continue through pregnancy, which may make side effects of antiretrovirals harder to distinguish and manage (i.e., during pregnancy). Other issues relate to the fact that the gold standard for preventing PHT, the 076 regimen, is AZT monotherapy, which is now considered suboptimal HIV treatment and no longer in use outside of perinatal settings.

In contrast, in resource-poor settings including most African and Asian nations where access to antiretroviral therapy is severely restricted at best and breast-feeding continues to be the norm, the PHT rate is 20-35%.

Mary Glenn Fowler, MD, MPH, of the CDC enumerated risk factors for PHT pre- and post-076.

• advanced illness as gauged by CD4 cell count and viral load
• maternal (illicit) drug use
• low vitamin A levels and anemia, particularly in developing countries
• preterm delivery
• breast-feeding

• maternal viral load at delivery (ACTG 185 and the Thai-Bangkok study)
• mode of delivery, particularly vaginal vs. elective cesarean
• breast-feeding
• duration of rupture of membranes (ROM) (combined data from the Perinatal AIDS Collaborative Transmission Studies [PACTS] and the Ariel Project for the Prevention of HIV Transmission from Mother to Infant, or ARIEL studies)

Also under continued scrutiny are the relative risks of CD4 cell count, chorioamnionitis (infection of the placenta), and preterm delivery, among others.

Future trials with special relevance for developing nations will evaluate the ability to decrease PHT through prenatal treatment of STDs and chorioamnionitis, nutritional supplementation, and breast-feeding strategies.

When does PHT occur?

According to J. Bertolli and others in a key article published in the October 1996 issue of the Journal of Infectious Diseases, 23% of PHT occur before birth, or prenatally. About 12% of HIV positive children are infected late, through breast-feeding. The majority of children (65%) are infected intrapartum (during labor and delivery), an observation which provides an obvious target for researchers.

Viral load and risk

Studies of PHT suggest an important correlation between increased viral load and risk. At present, there is no threshold below which risk is low, although recent studies involving viral load assays suggest a low risk of transmission among mothers whose viral load is undetectable. There also is no known threshold above which risk is 100%. Furthermore, the 076 regimen has been observed to protect against PHT across all viral levels, suggesting a role for antiretroviral prophylaxis.

In the Women and Infants Transmission Study (WITS) data published in the August 1999 issue of the New England Journal of Medicine, researchers concluded that maternal viral load predicts the likelihood but not the timing of PHT. WITS is a natural history study of vertical HIV transmission that enrolled over 2,000 mothers and infants. In this analysis, 552 women who delivered singletons (i.e., not twins) were studied. Women not treated with AZT and with viral loads greater than 100,000 copies/mL had the highest risk of transmitting HIV. No women with viral loads below 1,000 copies/mL transmitted the virus to their infants. Treatment with AZT was associated with a lower rate of PHT but not with a lower maternal viral load. In other studies, two-thirds of risk of PHT has been attributed to high maternal viral load.

Taken altogether, studies thus far suggest that therapy that reduces maternal viral load is highly effective, yet not 100% effective in preventing mother-to-child transmission. Other factors should continue to be regarded as significant, including length of rupture of membranes (greater than four hours before delivery has been previously shown to increase the risk of PHT) and chorioamnionitis.

Mode of delivery

Both with and without antiretroviral therapy, PHT is reduced by 30% when mothers elect to deliver by cesarean section, compared with vaginal delivery. Cumulative data suggest that combining cesarean delivery with AZT treatment reduces the risk of PHT to 2%.

The American College of Obstetricians and Gynecologists (ACOG) has recommended that all HIV positive women be offered cesarean delivery at the 38th week of pregnancy. The ACOG recommendations urge health-care providers to respect a woman's right to elect whether or not to have a cesarean delivery, which entails risks of infection and illness.

Breast-feeding

Breast-feeding is known to contribute significantly to PHT rates in developing countries. P.G. Miotti and others reported in the August 25, 1999 issue of the Journal of the American Medical Association that the greatest risk of breast-feeding transmission in their Malawi study of 672 infants born to HIV positive mothers (no antiretroviral therapy was used) occurred immediately after birth. For infants who continued on breast milk after one month, the cumulative risk was 3.5% after five months, 7.0% after 11 months, 8.9% after 17 months, and 10.3% after 23 months.

The declining rate of transmission is thought to be due more to the fact that susceptible infants are infected early in life, rather than that risk is truly lower as time goes on. Richard Semba and others published in the Journal of Infectious Diseases in 1999 findings that detectable viral load in breast milk translated to a high risk of PHT. They diagnosed subclinical mastitis (inflammation of the breast, usually caused by infection) on the basis of sodium in breast milk, which they found was also predictive of PHT.

In the August 7, 1999 issue of the Lancet, Anna Coutsoudis and others published landmark results of a trial that looked at risk of PHT by type of breast-feeding: exclusive breast-feeding vs. exclusive bottle-feeding vs. mixed feeding, or both breast- and bottle-feeding. At three months of age, 18.8% of babies who were never breast-fed were HIV positive, compared with 21.3% of breast-fed babies. When researchers looked at the children who were breast-fed, they divided them into two groups: babies who had been both breast- and bottle-fed (mixed feeding) and babies who were exclusively breast-fed.

The astonishing results were that 14.6% of babies who were exclusively breast-fed were HIV positive, compared with 24.1% of the mixed feeding babies. Unexpectedly, it appears that exclusive breast-feeding may be preferable to mixed breast and formula feeding, which could have enormous public health implications in developing countries, where many mothers currently provide mixed feeding to their infants (largely based on the understanding that breast milk may harm their babies). Researchers hypothesize that while breast milk is known to contain HIV, it may also contain antiviral factor as well as other protective antibodies that may help protect against HIV and other pathogens. Formula, on the other hand, may introduce new allergens or contaminants to the babies, which then can increase absorption in babies' intestinal tracts and thus increase their risk of being HIV-infected through breast-feeding by their HIV positive mothers.

In the developing world, several trials have already been completed. Most notable among recent trials are the HIVNET 012 trial in Uganda and the multicenter African Perinatal Transmission (PETRA) trial.

Short-course, placebo-controlled AZT trials also have been conducted and completed in Bangkok, Thailand, and in Abidjan, Ivory Coast. Both involved giving 300 mg of AZT twice daily or placebo beginning at 36 weeks to the pregnant women, and oral dosing through labor. A significant difference between the two trials was that a single dose was given at the beginning of labor and then AZT was also taken for one week postpartum by mothers in Abidjan (whereas AZT dosing in the Thai trials continued throughout labor). Also, there was no breast-feeding in the Thai trail. Since the Thai trial reported 50% efficacy and the African trial reported 37% efficacy, efficacy for AZT was clearly better in non-breast-feeding cases (50% vs. 37%). No additional benefit was apparently conferred by giving the mothers AZT for one week.

The landmark results of the HIVNET 012 trial were published in the September 3, 1999 issue of the Lancet. In this phase I/IIb trial, one group of mother and infants received nevirapine treatment and one group received AZT treatment. In the nevirapine arm, one 200 mg dose was taken by the mother at the onset of labor and one 2mg/kg dose was given to the newborn within the first 72 hours of life. The results were a 1.3 log drop in maternal viral load and effective transplacental and breast milk penetration by the drug. Mothers in the AZT arm took 600 mg at the onset of labor and 300 mg every three hours throughout labor; their infants received 4 mg/kg twice daily during the first week of life. In the AZT arm, there was a 10% rate of PHT at birth that increased to 25% at 14 weeks, compared with an 8% transmission rate at birth and 13.1% rate at 14 weeks in the nevirapine arm.

This trial proved that nevirapine (Viramune) can effectively prophylax infants through the first three months in a breast-feeding setting and that it is more effective in preventing PHT than (an equivalent blood duration of) AZT. The ideal length of effective neonatal prophylaxis in a breast-feeding setting has not been determined but will be a major focus of future studies. The HIVNET 012 study represents a major landmark in the HIV/AIDS epidemic, since it proves that a treatment regimen consisting of two doses of nevirapine can decrease PHT by one-half. With a wholesale cost of $4, this highly effective treatment should be much more accessible to women in the developing world. Nevertheless, implementation will likely pose many problems.

In the PETRA study of AZT/3TC, a factorial design permitted a comparison of different lengths of courses of AZT/3TC or placebo. Each of the four arms had 480 mother/child pairs. In the longest regimen, mothers received 300 mg of AZT and 150 mg of 3TC beginning in the 36th week of pregnancy and during labor and delivery, and infants received therapy for one week postpartum. In the intrapartum group, mothers received 300-600 mg of AZT/3TC during labor and delivery, and infants received one week of therapy. In the third arm only mothers received intrapartum therapy. The fourth group received placebo. The two longer regimens in which infants also received therapy showed efficacy; the intrapartum portion showed no efficacy compared to placebo. The placebo arm was stopped in February 1998.

At the time of the Global Strategies conference, early efficacy results (to six weeks) were available, and researchers stated that they hoped to have 18-month or late efficacy results available sometime in the first half of the year 2000. Of 1,796 total participants, data were reported for 1,447. In this study, breast-feeding and cesarean delivery were also considered. The transmission rate in women who did not have a cesarean section was 16.9%, in women who had an emergency cesarean section, 9.3%, and in women who had an elective cesarean section, 10.4%.

Strangely, the transmission rate was not significantly different between those who breast-fed and those who did not-this observation at six weeks may be too preliminary to be conclusive. Thus far, the two main factors seen to influence PHT were CD4 cell count and elective cesarean section; both were independent risk factors. Researchers concluded that giving AZT and 3TC at the onset of labor and for one week postpartum to both mother and child reduced the PHT rate by 38% even in areas where a majority of women breast-feed.

Some data presented by N.A. Wade and others from New York City show that six weeks of neonatal AZT may be effective in a non-breast-feeding setting. Among infants treated with AZT begun within 48 hours of life for six weeks, the PHT was 9.3% compared with 26.6% in those not treated. How helpful this approach would be in a breast-feeding setting is unknown.

Both the PETRA and HIVNET 012 trials have shown significant efficacy for either combination AZT/3TC or nevirapine, compared with both placebo and AZT. It is now also clear from perinatal studies that there is an important role for infant prophylaxis, and that intrapartum prophylaxis alone is insufficient. In addition to ongoing study of the role of neonatal prophylaxis, other topics for future study include the role of virus in breast milk.

A variety of nonantiretroviral strategies have been and continue to be studied, particularly for use in low-resource settings. These include vaginal cleansing and nutritional interventions. Overall results have been mixed-some appear mildly to moderately effective in reducing adverse maternal and infant health outcomes not related to HIV, while others seem ineffective. To date, none has been shown to be effective in reducing PHT.

Leslie Hanna is Editor of BETA.

Selected Sources

Bertolli, J. and others. Estimating the timing of mother-to-child transmission of human immunodeficiency virus in a breast-feeding population in Kinshasa, Zaire. Journal of Infectious Diseases 174(4): 722-726. October 1996.

Centers for Disease Control and Prevention. HIV/AIDS Surveillance Report 11(1). 1999.

Coutsoudis, A. and others. Influence of infant-feeding patterns on early mother-to-child transmission of HIV-1 in Durban, South Africa: a prospective cohort study. The Lancet 354 (9177): 442-443. August 7, 1999.

Guay, L.A. and others. Intrapartum and neonatal single-dose nevirapine compared with zidovudine for prevention of mother-to-child transmission of HIV-1 in Kampala, Uganda: HIVNET 012 randomised trial. The Lancet 354(9181): 795-802. September 4, 1999.

Lindegren, M.L. and others. Trends in perinatal transmission of HIV/AIDS in the United States. Journal of the American Medical Association 282(6): 531-538. August 11, 1999.

Miotti, P.G. and others. HIV transmission through breast-feeding: a study in Malawi. Journal of the American Medical Association 282(8): 744-749. August 25, 1999.

Page last updated 8 January 2000