| |
Herpes in pregnancy may raise HIV risk for baby
|
| |
| |
Wed Dec 28, 2005 10:16 PM ET
By Anne Harding
NEW YORK (Reuters Health) - HIV-positive women diagnosed with genital herpes during pregnancy appear to be quite likely to pass HIV to their infants, according to a study conducted in New York City.
Other studies are needed to confirm the relationship, lead investigator Dr. Katherine T. Chen of Columbia University told Reuters Health.
At present, she noted, HIV-positive women who contract herpes during pregnancy are treated for herpes infection and given medications beginning at approximately 36 weeks to prevent the transmission of herpes to the infant -- the same way such infections are treated in pregnant women without HIV.
The big question that remains unanswered, Chen explained, is whether HIV-positive women with any prior or current history of herpes infection should receive anti-herpes drugs throughout pregnancy in order to prevent transmission not only of herpes but also HIV.
As reported in the journal Obstetrics and Gynecology, Chen and her team evaluated 402 HIV-positive pregnant women, 5 percent of whom had a diagnosis of herpes during pregnancy. Six of the 21 women with herpes delivered HIV-infected infants.
After accounting for other risk factors for perinatal HIV transmission, the researchers found that herpes infection raised the risk of the baby becoming infected with HIV by nearly fivefold.
One of the study's strengths is that herpes diagnosis was clinically confirmed; its weaknesses include the lack of information on HIV levels for the pregnant women, Chen told Reuters Health. She and her colleagues are currently conducting a trial in which they are analyzing viral levels for both HIV and herpes in the genital fluid of HIV-positive pregnant women.
SOURCE: Obstetrics and Gynecology, December 2005.
Genital Herpes Simplex Virus Infection and Perinatal Transmission of Human Immunodeficiency Virus
Obstetrics & Gynecology Dec 2005;106:1341-1348
Katherine T. Chen, MD, MPH1,2, Marta Segu, MD, MPH5, L. H. Lumey, MD, PhD2, Louise Kuhn, PhD3, Rosalind J. Carter, PhD6, Marc Bulterys, MD, PhD7, Elaine J. Abrams, MD4 for the New York City Perinatal AIDS Collaborative Transmission Study (PACTS) Group*
From the Department of 1Obstetrics and Gynecology, Departments of 2Epidemiology, Columbia University, 3Sergievsky Center and 4Harlem Hospital Center, Columbia University, New York, New York; 5Fundacio Barcelona SIDA 2002, Barcelona, Spain; 6Medical and Health Research Association, New York, New York; and 7Division of HIV/AIDS Prevention, National Center for HIV/STD/TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia.
ABSTRACT
OBJECTIVE: To assess the risk of perinatal human immunodeficiency virus (HIV) transmission in HIV-infected women clinically diagnosed with genital herpes simplex virus (HSV) infection during pregnancy.
METHODS: This retrospective analysis included 402 HIV-infected pregnant women who enrolled from 1994-1999 in a multicenter prospective cohort study in New York City, who delivered a liveborn singleton infant with known HIV infection status, and who had information on diagnosis of genital HSV infection during pregnancy. Study participants were determined to have genital HSV infection during pregnancy by documentation of clinical diagnosis.
RESULTS: Forty-six (11.4%) of the study participants delivered HIV-infected infants. Twenty-one (5.2%) had clinical diagnosis of genital HSV infection in pregnancy. Six (28.6%) of the 21 HIV-infected women with a clinical diagnosis of genital HSV infection delivered an HIV-infected infant. In univariate analyses, HIV-infected pregnant women with clinical diagnosis of genital HSV infection during pregnancy had a significantly increased risk of perinatal HIV transmission (odds ratio 3.4, 95% confidence interval 1.3-9.3; P = .02). When other factors associated with perinatal HIV transmission were included in a logistic regression model (lack of zidovudine therapy during pregnancy or delivery, prolonged rupture of membranes, and preterm delivery), clinical diagnosis of genital HSV infection during pregnancy remained a significant independent predictor of perinatal HIV transmission (adjusted odds ratio 4.8, 95% confidence interval 1.3-17.0; P = .02).
CONCLUSION: Clinical diagnosis of genital HSV infection during pregnancy in HIV-infected women may be a risk factor for perinatal HIV transmission. If future studies confirm this association, therapy to suppress genital HSV reactivation during pregnancy may be a strategy to reduce perinatal HIV transmission.
LEVEL OF EVIDENCE: II-3
BACKGROUND
Accumulating evidence suggests that human immunodeficiency virus (HIV)-infected persons who are coinfected with herpes simplex virus (HSV)-2 may have an increased risk of sexual transmission of HIV.1,2 Both clinical and subclinical genital HSV-2 reactivations increase the magnitude of HIV viral load in the genital tract. The recruitment of HIV-infected+ cells into mucosal ulcerations due to genital HSV-2 reactivations, along with the potential in vivo interaction of the 2 viruses, may account for the high HIV viral load in the genital tract.6 When genital HSV-2 reactivation occurs, HIV viral load increases not only in the genital tract but in the plasma as well.7 Given the potential role of HSV-2 in sexual transmission of HIV, we sought to investigate the role of genital HSV infection in perinatal HIV transmission.
Previous studies conducted in African countries have examined the role of genital ulcers and the risk of perinatal HIV transmission, with 2 studies showing no association 8,9 and 1 study revealing a trend.10 Although HSV infection is a major cause of genital ulcers, other illnesses can also produce them, including syphilis, chancroid, and HIV itself.11 Several studies have examined the role of sexually transmitted diseases as a group (including genital HSV infection) and risk of perinatal HIV transmission with inconclusive results. One Rwandan study and 1 U.S. study showed no association;12,13 another U.S. study revealed a trend;14 and 2 French studies showed a significant association.15,16
To our knowledge, only the study by Van Dyke and colleagues17 examined the role of genital HSV infection during pregnancy. They found no association of self-reported genital HSV infection with perinatal transmission in 200 HIV-infected pregnant women in a U.S. multicenter prospective cohort. Given the paucity of information on the association of genital HSV infection during pregnancy and perinatal HIV transmission, we examined this question in a multicenter prospective cohort in New York City that documented clinical diagnosis of genital HSV infection during pregnancy.
DISCUSSION
In this analysis of 402 HIV-infected women who enrolled in a multicenter prospective study of perinatal HIV transmission in New York City during 1994 to 1999, we found clinical diagnosis of genital HSV infection during pregnancy to be 1 of several significant risk factors for perinatal HIV transmission. This association is independent of the other factors found to be significant in this analysis, including lack of zidovudine prophylaxis during pregnancy or delivery, prolonged rupture of membranes, and preterm delivery.
Previous research on pregnant women coinfected with HIV and HSV is limited. One study demonstrated that pregnant women coinfected with both HIV and HSV-2 were more likely to shed HSV-2 in the genital tract than nonpregnant women.29 Hitti and colleagues30 demonstrated that pregnant women coinfected with HIV and HSV-2 reactivate HSV in labor more often than pregnant women infected with HSV-2 only. In a study by Van Dyke and colleagues,17 the risk of perinatal transmission was not associated with self-reported genital HSV infection during pregnancy. In contrast, the present study with an increased number of study participants and HIV-infected infants did detect an association between perinatal HIV transmission and clinical diagnosis of genital HSV infection in pregnancy assessed by medical record documentation. Clinical diagnosis of genital HSV infection (supported by positive genital herpes culture in one half of the patients) may be a better measure of genital HSV infection than self report.
The mechanism by which genital HSV infection in pregnancy may increase perinatal HIV transmission may occur through the effect of HSV to increase maternal plasma and genital tract HIV viral load. Herpes simplex virus-1 can activate latent HIV or enhance its replication.31 When HSV-2 reactivation occurs, plasma HIV viral load increases.7 Maternal plasma HIV viral load is 1 of the most important risk factors for perinatal HIV transmission.23 Both clinical and subclinical HSV-2 reactivation increase the frequency and quantity of mucosal HIV shedding. Previous studies have demonstrated that the presence of genital HIV RNA32 and DNA33 in HIV-infected pregnant women receiving no or minimal antiretroviral therapy increases the risk of perinatal HIV transmission. Tuomala and colleagues34 recently demonstrated that detection of genital HIV DNA may also be an independent risk factor for perinatal HIV transmission in a cohort of HIV-infected women receiving antiretroviral therapy. Because reactivation with HSV-2 is thought to increase both plasma7 and genital HIV viral load, genital HSV infection may also be important in perinatal HIV transmission.
A major limitation of this study is the inability to assess maternal plasma HIV viral load at the time of delivery in a large number of the study participants. Consequently, we cannot reliably determine the role of plasma HIV viral load as a risk factor for perinatal HIV transmission. Plasma HIV viral load at the time of delivery has been shown to be a significant risk factor for perinatal HIV transmission in a number of other studies.13,23,35 We did find an increased odds ratio for women with an HIV viral load of 1,000 copies or more per milliliter and CD4 count less than 500 cells per cubic millimeter close to delivery, but the association was not statistically significant. Therefore, it would be important that other studies further assess the relationship between HIV viral load, CD4 count, and HSV infection to the risk of perinatal transmission.
Another limitation is the inability to verify the clinical diagnosis of genital HSV infection during pregnancy in 9 patients who did not have a positive genital herpes culture. We do not have information on whether a genital herpes culture was not performed or the genital herpes culture was negative. We assume that the clinical diagnosis of genital HSV infection during pregnancy by examination only can be made accurately when characteristic multiple vesicular lesions on an erythematous base are present.36 In addition, a negative genital herpes culture may be a false negative, because the sensitivity of HSV viral isolation in culture depends on the stage of lesions and on whether the patient has a first or a recurrent episode of the disease.36 In 1 study, 10% of pregnant women who had a negative genital herpes culture done on genital specimens collected during labor were HSV positive by polymerase chain reaction.37
Only 1 woman who had a culture-confirmed diagnosis of genital HSV infection delivered an HIV-infected infant. In addition, no woman who delivered within 30 days of diagnosis of genital HSV infection delivered an HIV-infected infant. These findings may result from an increased likelihood of receiving treatment for genital HSV infection, a longer duration of therapy, or an increased use of therapy to suppress genital HSV reactivation during pregnancy. These practices may in turn decrease risk of genital HSV reactivation at time of delivery.
This study must be interpreted in the context of the study design. We do not know whether the women who had culture-confirmed diagnosis of genital infection were infected with HSV-1 or HSV-2. Although HSV-2 is primarily responsible for genital HSV infection, HSV-1 can also cause genital disease.38 The association between HSV-1 and transmission of HIV is less clear. We are unable to determine whether study participants had asymptomatic genital HSV infections during pregnancy. Viral detection by HSV DNA polymerase chain reaction would have led to more accurate and higher HSV detection rates.39 This method of detection was not available to the researchers of this study. In addition, we are only able to assess genital HSV infection during pregnancy and not during labor and delivery because data were not collected. Genital HSV infection during the latter time period is likely to be a more important predictor of perinatal HIV transmission.
Although we have shown that clinical diagnosis of genital HSV infection during pregnancy in HIV-infected women seems to be a significant risk factor for perinatal HIV transmission, we acknowledge that our finding could have occurred by chance. In addition, these data were collected before the introduction of highly active antiretroviral therapy during pregnancy. The use of potent combination therapy to reduce HIV viral burden to undetectable values is the most effective strategy to decrease perinatal HIV transmission. However, in low-resource settings where highly active antiretroviral therapy is not routinely available for prevention of perinatal HIV transmission, short-course treatment of HSV coinfection could hypothetically reduce infant HIV infection. Furthermore, our work suggests a need for a more complete evaluation of the role of HSV as a risk factor for perinatal HIV transmission to gain insight into the mechanisms of disease and perinatal transmission.
RESULTS
Between April 4, 1994, and April 21, 1999, 463 HIV-infected pregnant women delivered a liveborn singleton infant with known HIV-infection status. Four hundred two (86.8%) of the study participants had information on diagnosis of genital HSV infection during pregnancy, given adequate medical records for review. The rest of the analysis is therefore limited to these 402 women. One hundred eighty-five (46.0%) of the 402 women were enrolled postpartum. Forty-six (11.4%) of the 402 women delivered HIV-infected infants.
The 402 women were recruited from the following hospitals in New York City: Bronx-Lebanon (29.9%), Harlem (29.4%), Montefiore (20.4%), Metropolitan (16.9%), and Bellevue (3.5%). The mean age of the study participants at delivery was 29 years (range 16-46) for 399 women. Most women were of non-Hispanic black (54.7%) and of Hispanic (33.3%) origin. With respect to education, 11.9% of 370 women reported 8 years or less and 9.4% reported 13 years or more. Table 1 shows maternal demographics according to perinatal transmission status.
Twenty-one (5.2%) of the 402 women were determined to have a clinical diagnosis of genital HSV infection during pregnancy. Six (28.6%) of the 21 HIV-infected women with a clinical diagnosis of genital HSV infection delivered an HIV-infected infant. Women with clinical diagnosis of genital HSV infection during pregnancy had an increased risk of delivering an HIV-infected infant in a univariate analysis (odds ratio 3.4, 95% confidence interval 1.3-9.3, P = .02; Table 2).
Twelve (57.1%) of the 21 women who had a clinical diagnosis of genital HSV infection during pregnancy had the diagnosis confirmed by culture. Data were not available as to whether the culture was positive for HSV-1 or HSV-2. One woman of the 12 who had a culture-confirmed diagnosis of genital HSV infection during pregnancy delivered an HIV-infected infant.
Twenty (95.2%) of the 21 women who had a genital HSV infection during pregnancy had a date of clinical diagnosis recorded. Seven (35.0%) of the 20 women who had a date of clinical diagnosis recorded delivered within 30 days of diagnosis of genital HSV infection. None of these 7 women who delivered within 30 days of diagnosis of genital HSV infection delivered an HIV-infected infant. The mean number of days from clinical diagnosis to delivery was 63 days; median, 35 days; and range, 0-240 days.
Women who lacked zidovudine prophylaxis during pregnancy or delivery, women who experienced rupture of the membranes 4 hours and women who delivered infants at < 37 weeks of gestation all had a significantly increased risk of perinatal HIV transmission (Table 2). Women with maternal plasma HIV viral load of 1,000 copies or more per milliliter had an increased odds of perinatal HIV transmission that was not statistically significant. However, 45% of the study participants had unknown viral load. Analyses with other cut-points of maternal HIV viral load23 did not yield significant associations with perinatal HIV transmission (data not shown). Other indicators of maternal health status such as CD4 count and diagnosis of AIDS were also not significantly associated with perinatal HIV transmission. Illicit drug use during pregnancy and mode of delivery showed no association.
The association between clinical diagnosis of genital HSV infection during pregnancy and the risk of perinatal HIV transmission was assessed for confounding by the other maternal variables previously mentioned (HIV viral load, CD4 count, diagnosis of AIDS, lack of zidovudine prophylaxis during pregnancy or delivery, illicit drug use during pregnancy, duration of membrane rupture, mode of delivery, and gestational age at delivery). Clinical diagnosis of genital HSV infection during pregnancy was not significantly associated with any of the other variables (data not shown).
Women with test-positive gonorrhea or Chlamydia infection during pregnancy did not have an increased risk of perinatal HIV transmission (Table 3). Similarly, women with a clinical diagnosis of syphilis during pregnancy did not have an increased risk of perinatal HIV transmission. Clinical diagnosis of genital HSV infection during pregnancy was not significantly associated with diagnosis of any other sexually transmitted disease during pregnancy (data not shown).
Results of multivariate analyses are shown in Table 4. In logistic regression models containing variables significantly associated with perinatal HIV transmission (n = 348, because not all study participants had data available for all variables), clinical diagnosis of genital HSV infection during pregnancy, lack of zidovudine prophylaxis during pregnancy or delivery, rupture of membranes 4 hours or more, and preterm delivery at less than 37 weeks remained significant independent predictors of perinatal HIV transmission. Addition of 2 covariates not significantly associated with perinatal HIV transmission to the model, CD4 count less than 500 cells or 500 cells or more per cubic millimeter and diagnosis of AIDS, yielded similar results (data not shown). Multivariate analyses based on the total number of study participants (n = 402) but with a "missing" category yielded similar results (data not shown).
MATERIALS & METHODS
The study group included 463 HIV-infected pregnant women, who enrolled during pregnancy or the immediate postpartum period (6 weeks) between April 1994 and April 1999 in a multicenter prospective study of perinatal HIV transmission in New York City and who delivered a liveborn singleton infant with known HIV infection status. The ethical review boards of all participating institutions approved the prospective study. Each study participant provided written informed consent.
Study participants had study visits at scheduled intervals throughout pregnancy and the immediate postpartum period. Trained research staff interviewed participants and collected maternal demographic information and medical, drug, and sexual histories at enrollment and at regular follow-up visits when appropriate. Information on physical examinations performed by clinicians providing obstetric care and laboratory data were collected during pregnancy and the postpartum period. Intrapartum information, not collected by a member of the study team, was abstracted through medical record review.
The following information was obtained by medical record review: 1) Did the patient have a clinical diagnosis of genital HSV infection during pregnancy documented in the medical record? 2) If the patient did have a clinical diagnosis of genital HSV infection during pregnancy, what was the most recent date of diagnosis? and 3) If the patient had a genital herpes culture performed that was positive, what was the most recent date of positive culture?
Other maternal variables that have been linked to perinatal HIV transmission (high maternal HIV viral load, low maternal CD4 count, increased maternal illness severity, lack of antiretroviral therapy, illicit drug use, prolonged rupture of membranes, vaginal delivery, and prematurity)21,22 were also studied in the cohort. The plasma HIV viral load and CD4 count closest to delivery (sample obtained within 90 days of delivery either antepartum or postpartum), was used to evaluate maternal virologic and immunologic status. Women were classified by viral load measurement of 1,000 copies or more per milliliter or less than 1,000 copies per milliliter, because 1,000 copies or more per milliliter is associated with risk of perinatal HIV transmission.23 Women were also categorized by CD4 count less than 500 cells per cubic millimeter or 500 cells or more per cubic millimeter, because 500 cells or more per cubic millimeter was considered a relatively high CD4 count during the time of the multicenter cohort study and women who had a CD4 count of 500 cells or more per cubic millimeter may have been managed differently from those who had a CD4 count less than 500 cells per cubic millimeter.24 Study participants were classified as having AIDS according to the 1993 Centers for Disease Control and Prevention surveillance criteria.25 Use of zidovudine prophylaxis during pregnancy or delivery was reported by study participants or site clinics. Details regarding adherence to zidovudine prophylaxis were not available for this analysis. Illicit drug use during pregnancy included any use of heroin, cocaine, crack, phencyclidine, or marijuana. Obstetric factors such as length of ruptured membranes (in hours), mode of delivery, and gestational age at delivery were collected by medical record review. Prolonged rupture of membranes was defined as rupture of membranes 4 hours or more.26,27
Other sexually transmitted diseases were also studied by collecting the following information by medical record review: 1) What was the gonorrhea test result during pregnancy? 2) What was the Chlamydia test result during pregnancy? and 3) Did the patient have a clinical diagnosis of syphilis infection during pregnancy? Infants were considered HIV-infected according to criteria established by the Center for Disease Control and Prevention in 1994.28
Differences in the distribution of genital HSV infection during pregnancy and other maternal variables possibly related to perinatal HIV transmission were cross-tabulated and compared with likelihood ratio 2 tests and Fisher exact test when appropriate. A P value of less than .05 was considered to represent a statistically significant difference. All statistically significant risk factors for HIV transmission were then entered into a logistic regression model to estimate the independent association of each with perinatal HIV transmission, adjusted for the other covariates used in the model. We used SPSS 12.0 software (SPSS Inc., Chicago, IL) for all analyses.
|
|
| |
| |
|
|
|
