SEXUALLY TRANSMITTED INFECTIONS DURING PREGNANCY are, as a group, associated with increased risks of preterm birth and other adverse pregnancy outcomes. Gonococcal and chlamydial infections have been associated with preterm birth and premature rupture of membranes.1,2 In 1997, the Vaginal Infections and Prematurity Study Group reported a significant increase in low birth weight and preterm delivery with a 40% attributable risk in preterm low birth weight resulting from Trichomonas vaginalis (TV).3 However, metronidazole treatment of women with asymptomatic trichomoniasis was associated with an increased risk of preterm birth in a large multicenter trial.4 The biologic mechanism for the increase in preterm birth among trichomonas-infected women has not been elucidated. One hypothesis is that the organisms cause a release of inflammatory molecules, which results in a host response favoring early labor.
There is evidence of a local host inflammatory response to TV,5 Neisseria gonorrhoeae (GC),6 and Chlamydia trachomatis (CT).7 These same inflammatory molecules, when found in the cervix or vagina, have been linked to preterm birth.8,9 However, little work done characterizing the systemic immune response to sexually transmitted infections has been published. Hedges et al examined the systemic immune response in nonpregnant women with gonococcal infection. They did not find a significant systemic immune response to infection with GC alone, but found elevated concentrations of inflammatory cytokines when women were concomitantly infected with other sexually transmitted pathogens. Interestingly, they found that in women coinfected with GC and TV, the systemic immune response was intensified after treatment.10 There is a need for a more robust body of literature regarding the presence or absence of systemic inflammation both as it relates to pregnancy and to sexually transmitted infection.
C-reactive protein (CRP) is a marker of systemic inflammation that is frequently elevated in response to infection. It has not been shown to reliably predict preterm birth risk in unselected populations but is associated with overall increased risks of preterm birth, infection, and inflammation.11,12 A recent study was the first to examine the relationship among a sexually transmitted infection, CRP, and preterm birth risk. They found that CRP levels using a highly sensitive assay were higher in the women with high levels of antibodies to CT and preterm birth.13 Interestingly, the study did not confirm active CT infection but used only the presence of antibodies to assess infection status, and the CRP levels alone were not predictive. The objective of the current study was to assess whether there is a systemic immune response in pregnant women infected with TV as compared with pregnant women who are not infected. A secondary objective was to determine if women with multiple concomitant sexually transmitted infections have a greater systemic response compared with those infected with TV alone.
Materials and Methods
A nested case control study was performed using stored serum from the Vaginal Infections and Prematurity Study (VIP) conducted by the National Institute of Child Health and Human Development of the National Institutes of Health (NIH). The VIP study was a multicenter prospective study of genital infections and pregnancy outcomes. Between 1984 and 1989, pregnant women were enrolled in the study between 23 and 26 weeks gestation. Demographic, medical, and sexual information was collected at the original enrollment. Cervical cultures were obtained for GC and CT, and vaginal cultures were obtained for TV. For the TV culture, specimens on sterile cotton swabs were inoculated directly into a tube of modified Diamond's medium and incubated at 37°C. Aliquots of media were examined microscopically for the presence of motile, flagellated protozoa. The specimens were examined at 2- to 3-day intervals and were considered negative after incubation without organism detection after 5 days. The reproducibility of detection, as previously described, was excellent.3 Gram stain was performed to evaluate for the presence of bacterial vaginosis. Exclusion criteria for the VIP study were age less than 16 years, Rh isoimmunization, use of antibiotics in the preceding 2 weeks, multiple gestation, cervical cerclage, prior use of tocolytic agents during the current pregnancy, hypertension requiring treatment with medication, insulin-dependent diabetes, current use of corticosteroids, chronic renal or symptomatic organic heart disease, and the intention to deliver at a nonstudy hospital. The current study was approved by the Institutional Review Board at the University of Pittsburgh.
Serum was collected at enrollment into the VIP study and stored at McKesson BioServices Corporation, Rockville, Maryland, the biologic repository center used by the NIH. The samples were stored at −80°C and never thawed before their use in the current study. A pilot analysis was performed on a subset of the samples to determine the cytokine and chemokine profile that would be detectable in the sera of pregnant women in the absence of well-established normal values. The 11 cytokines and chemokines that were included in this pilot were granulocyte–macrophage colony-stimulating factor (GM-CSF), interleukin (IL)-1β, IL-6, IL-8, macrophage inflammatory protein (MIP)-1α, regulated on activation, normally T-cell expressed and secreted (RANTES), IL-2, IL-4, IL-10, interferon-γ, and IL-12p40. The only cytokines and chemokines that were routinely detected in maternal serum and thus subsequently included in the study were GM-CSF, IL-1β, IL-6, IL-8, MIP-1α, and RANTES. Although the serum samples had prolonged storage, they were held at −80°C, which is the temperature at which the least degradation of cytokines is expected.14
A nested case control study was performed on 195 serum samples selected from the NIH-maintained VIP database matched for race and concurrent gonococcal or chlamydial infection comparing concentrations of inflammatory markers between 96 women infected and 99 women uninfected with TV. The sample was a convenience sample in which cases were defined as women infected with TV and controls defined as TV-uninfected. The cytokine and chemokine concentrations were assessed using the Luminex multiplex assay in duplicate (Luminex Corp., Austin, TX). Serum samples were first diluted 1:1 with human serum diluent. The dilute serum was mixed with Beadlyte multicytokine beads (Upstate USA Inc., Charlottesville, VA). The sample was then mixed with antihuman multicytokine reporter and fluorescence labeled. The Luminex laser then detected and quantified the analyte present on the beads. The Luminex multicytokine bead method has been previously validated and shown to have excellent correlation with standard enzyme-linked immunosorbent assay (ELISA) methods.15 The lower limit of detection for all cytokines except IL-12p40 was 2.3 pg/mL. The lower limit of detection for IL-12p40 was 6.9 pg/mL.
The CRP assay was a simple sandwich ELISA adapted from that of Erhardt et al.16 The plates were coated in diluted rabbit polyclonal anti-CRP antibody (Dako, Glostrup, Denmark) and incubated. Serum samples were diluted 1:1,000 to 1:10,000, to fit a standard curve, and incubated in the antibody-coated wells. The samples were then incubated with the secondary antibody, diluted polyclonal HRP-conjugated rabbit anti-CRP antibody (Dako). After washing, TMB peroxidase substrate (Bio-Rad, Hercules, CA) was added for color reagent.
Demographic information was compared between the groups using the χ2 test. The cytokine and chemokine concentrations were evaluated for normal distribution. Transformation of the data to normality was not possible. Therefore, nonparametric analyses were performed. The Kruskal-Wallis test was performed to compare median concentrations of cytokines and chemokines. Nonparametric test for trend was performed to compare cytokine and chemokine concentrations across groups. For these analyses, P < 0.05 was considered significant. Univariate logistic regression was used to determine the association of particular confounding variables with cytokine and chemokine concentrations in the highest quartiles with a P <0.25 considered significant. Significant confounders as well as appropriate demographics were included in a forward-stepping multiple logistic regression model to determine odds ratios for association.
The log transformation of CRP approximated normal. A 2-sided Student t test was used to compare log-transformed values of CRP between groups. A logistic regression model was used to evaluate the association of having a CRP value in the above the 90th percentile with various confounders. Univariate analysis was used to identify the associations with potential confounders. Variables that were found to be significant with a P value <0.25 were included in a forward-stepping multivariable model. Statistical analysis was performed using Stata 8.0 (Stata Corp., College Station, TX).
A comparison of women infected with TV with those who were uninfected is presented in Table 1. There were no significant differences between the groups in terms of age, parity, or median gestational age at delivery. Infected women were more likely to smoke and less likely to have bacterial vaginosis detected at enrollment. By nature of the matched design, there was no difference in infection by GC or CT between the groups. Twenty percent of women in each group were coinfected with CT and an additional 20% were in each group were infected with GC. Also, by nature of the matched design, there were no significant differences in racial composition between the groups. The groups were approximately 50% black, 25% white, and 25% Hispanic.
The median serum concentration of GM-CSF was significantly higher in the group of women infected with TV, 8.9 pg/mL compared with 5.7 pg/mL (P <0.001). There were no significant differences in median concentrations of any of the other cytokines measured. The relationship between TV infection and GM-CSF persisted after controlling for confounding variables. The adjusted odds ratio for having GM-CSF concentration in the highest quartile was 6.0 (95% confidence interval [CI], 2.9–12.5) in the group of women having trichomoniasis. The increased risk of having elevated GM-CSF was found after controlling for age, tobacco use, marital status, and bacterial vaginosis using multiple logisticregression.
Women coinfected with TV and either GC or CT had significantly greater GM-CSF concentrations than women infected with TV alone (P <0.001). Additionally, there was a significant nonparametric test for trend for increasing concentrations of GM-CSF in the setting of coinfections of TV with GC or CT as compared with women infected with TV alone (P <0.001). There were no significant differences in the medians of any of the other cytokines or chemokines measured in these groups.
The mean log-transformed CRP values were higher in the TV-infected group compared to the TV-uninfected group (1.66 vs. 1.27; P = 0.03). The variables that were significantly associated with having a CRP in the 90th percentile on univariate analysis included presence of bacterial vaginosis and trichomoniasis. After controlling for bacterial vaginosis, the odds ratio for having a CRP above the 90th percentile in women with trichomoniasis was 3.3 (95% CI, 1.14–9.56; P = 0.03). There did not appear to be a trend toward higher CRP values in the presence of coinfections with other sexually transmitted infections. Nor were there any differences in mean log-transformed CRP values in women infected with GC or CT as compared with women without any sexually transmitted infections. The median cytokine, chemokine, and CRP concentrations for the 2 groups are shown in Table 2.
The results of this study suggest that vaginal infection with TV during pregnancy is associated with a systemic inflammatory response in some women as exhibited by an increase in the serum concentrations of both GM-CSF and CRP. There have been relatively few studies to examine the systemic response to sexually transmitted infections, and none have been done in pregnant women.
Hedges et al found that women coinfected with GC and either CT or TV had elevated serum concentrations of the proinflammatory cytokines IL-1β and IL-6.10 A slight systemic immune response to the inoculation of healthy male volunteers with gonococcal urethritis has been reported as represented by increased plasma concentrations of IL-8, TNF-α, IL-1β, and IL-6 in some of the participants.6 The current study did not confirm these earlier reports linking IL-8, IL-1β, or IL-6 to gonococcal or chlamydial infection. Our study differs from the Hedges study in the larger number of women included having gonococcal or chlamydial infection with (n = 39) or without (n = 40) concomitant trichomoniasis, whereas the Hedges study reported on a total of 41 infected women. A possible limitation of our study is the lower sensitivity of the tests used for detecting sexually transmitted infections in the 1980s. It is possible that there was an ascertainment bias in that only women with a large organism burden might have been included. Future studies will use newer detection methods that will allow for more accurate diagnosis.
Another limitation of our study is the prolonged duration of storage of serum specimens before analysis. However, because the specimens were stored at −80°C in a central repository and the only published data suggests stability for 1 year under these storage conditions, we believe that these findings are valid for comparison within this subset.14 It is certainly possible that there was degradation of certain cytokines with prolonged storage. If there was degradation, however, it would be expected to be seen across all groups. It is may be, however, that with a more recently collected sample, we would have detected more differences than are presented here. The ranges of cytokine concentrations that were recorded in our study are in accordance with those expected in serum at basal levels according to the product manufacturer (Upstate USA Inc.). Established normal serum values for these particular cytokines and chemokines in pregnancy are lacking. The concentrations of GM-CSF detected in this study were similar, although slightly lower, than observed among patients with infectious and noninfectious systemic inflammatory response syndrome, suggesting that the values are in the range one would expect for a mild inflammatory condition.17
Infection with TV during pregnancy is a significant concern for the clinician both because of adverse pregnancy outcomes as well as public health interests related to infection. The association of preterm birth with trichomoniasis was the focus for the Maternal Fetal Medicine Units (MFM-U) trial investigating screening and treatment of asymptomatic women. Unfortunately, treatment of asymptomatic trichomoniasis did not decrease the risk of preterm birth in that study, and in fact, the study was stopped because of an increased risk in the treated group.4 The MFM-U study, however, was not designed to address either the mechanism of the underlying cause for the increased prematurity in infected women or the mechanism of the increased risk of treatment. A secondary analysis of a study designed to evaluate the impact of treatment for sexually transmitted infections for prevention of HIV did not detect an increase in preterm birth among women treated for trichomoniasis in any trimester, although there was an increased risk of low birth weight among women receiving cefixime, azithromycin, and metronidazole.18 It is possible that the inflammation caused by the parasite is such that its damage cannot be remediated by short-course therapy. Alternatively, it is possible that cysteine proteases or protozoal fragments released after antibiotic treatment lead to inflammatory changes that enhance the risk of preterm delivery many weeks later. It is important to note that the MFMU Network treatment trial used an unusual dosing regimen, and there was significant crossover in the placebo group with almost one fourth of the women receiving metronidazole therapy off protocol.
Our study was designed to assess the relationship between infection with TV and maternal systemic immune response, not preterm birth. One limitation of this study is that the number of women delivering preterm was not adequate to allow for adequate power to assess these risks. Therefore, no conclusions can be drawn between systemic immune response to TV and preterm birth. Also, the serum was drawn only one time between 23 and 26 weeks gestation. We do not know the clinical implications of immune activation at this particular point in gestation, nor do we know whether treatment alters the immune response profile.
Some experts have recommended against treating asymptomatic trichomoniasis during pregnancy in addition to recommending against routine screening.19 The option of not treating a patient with a sexually transmitted infection remains unattractive for several reasons. First, TV infection is associated with an increased risk of HIV transmission and acquisition. Trichomoniasis is associated with an increased number of CD4 lymphocytes, the cellular target for HIV, in the female genital tract.20 It also has a cytopathogenic effect on human vaginal cells in vitro and causes punctate microhemorrhages that might facilitate HIV acquisition.21,22 In populations with high prevalence of both HIV and TV, it is estimated that up to 20% of HIV infections are related to trichomoniasis.23 Metronidazole treatment has been shown to decrease HIV shedding in TV-infected men.24
Trichomoniasis is especially prevalent among black women and adolescents, 2 groups with an increased risk of adverse pregnancy outcomes.22,25 It may be unethical to withhold information regarding a positive test result for a sexually transmitted infection. In turn, it would then be unethical not to offer therapy to a patient who desires it.
The elevation of GM-CSF and CRP in the serum of pregnant women with trichomoniasis represents a biologically plausible link between a local vaginal infection and a systemic response. The major function of GM-CSF is to increase production of granulocytes, macrophages, and dendritic cells. Therefore, the finding of elevated serum GM-CSF in women with trichomoniasis is not particularly surprising given the known involvement of leukocytes in host defense against TV. GM-CSF has also been shown to be elevated in the vaginas of macaques infected with trichomoniasis.26 In the current study, women having either gonococcal or chlamydial infection did not have elevated GM-CSF (data not shown), suggesting that this finding was specifically related to trichomoniasis and not other sexually transmitted infections.
CRP is an acute-phase protein produced in response to cytokine stimulation and plays a role in a variety of responses to infectious agents. CRP is a known marker of systemic inflammation. Because labor is a systemic process, it may be that there is systemic inflammation in the pathway leading to birth both at term and preterm. The intent of this study is to identify potential mechanistic pathways that need to be further elucidated in the association between TV infection and preterm birth. Future research should be focused on defining the mechanisms by which trichomoniasis increases the risk of preterm birth and the impact that treatment of TV infection has on pregnancy. Only once mechanisms are delineated can we then investigate the optimum therapy for trichomoniasis in pregnancy.
1. Alger LS, Lovchik JC, Hebel JR, et al. The association of Chlamydia trachomatis, Neisseria gonorrhoeae
, and group B streptococci with preterm rupture of the membranes and pregnancy outcome. Am J Obstet Gynecol 1988; 159:397–404.
2. Andrews WW, Goldenberg RL, Mercer B, et al. The Preterm Prediction Study: Association of second-trimester genitourinary chlamydia infection with subsequent spontaneous preterm birth. Am J Obstet Gynecol 2000; 183:662–668.
3. Cotch MF, Pastorek JG 2nd, Nugent RP, et al. Trichomonas vaginalis associated with low birth weight and preterm delivery. The Vaginal Infections and Prematurity Study Group. Sex Transm Dis 1997; 24:353–360.
4. Klebanoff MA, Carey JC, Hauth JC, et al. Failure of metronidazole to prevent preterm delivery among pregnant women with asymptomatic Trichomonas vaginalis
infection. N Engl J Med 2001; 345:487–493.
5. Simhan HN, Anderson BL, Krohn MA, et al. Host immune consequences of asymptomatic Trichomonas vaginalis
infection in pregnancy. Am J Obstet Gynecol. In press.
6. Ramsey KH, Schneider H, Cross AS, et al. Inflammatory cytokines produced in response to experimental human gonorrhea. J Infect Dis 1995; 172:186–191.
7. Morrison SG, Morrison RP. In situ analysis of the evolution of the primary immune response in murine Chlamydia trachomatis
genital tract infection. Infect Immun 2000; 68:2870–2879.
8. Arntzen KJ, Kjollesdal AM, Halgunset J, et al. TNF, IL-1, IL-6, IL-8 and soluble TNF receptors in relation to chorioamnionitis and premature labor. J Perinat Med 1998; 26:17–26.
9. Hitti J, Hillier SL, Agnew KJ, et al. Vaginal indicators of amniotic fluid infection in preterm labor. Obstet Gynecol 2001; 97:211–219.
10. Hedges SR, Sibley DA, Mayo MS, et al. Cytokine and antibody responses in women infected with Neisseria gonorrhoeae
: Effects of concomitant infections. J Infect Dis 1998; 178:742–751.
11. Watts DH, Krohn MA, Hillier SL, et al. Characteristics of women in preterm labor associated with elevated C-reactive protein levels. Obstet Gynecol 1993; 82:509–514.
12. Pitiphat W, Gillman MW, Joshipura KJ, et al. Plasma C-reactive protein in early pregnancy and preterm delivery. Am J Epidemiol 2005; 162:1108–1113.
13. Karinen L, Pouta A, Bloigu A, et al. Serum C-reactive protein and Chlamydia trachomatis
antibodies in preterm delivery. Obstet Gynecol 2005; 106:73–80.
14. Porter AE, Auth J, Prince M, et al. Optimization of cytokine stability in stored amniotic fluid. Am J Obstet Gynecol 2001; 185:459–462.
15. de Jager W, te Velthuis H, Prakken BJ, et al. Simultaneous detection of 15 human cytokines in a single sample of stimulated peripheral blood mononuclear cells. Clin Diagn Lab Immunol 2003; 10:133–139.
16. Erhardt JG, Estes JE, Pfeiffer CM, et al. Combined measurement of ferritin, soluble transferrin receptor, retinol binding protein, and C-reactive protein by an inexpensive, sensitive, and simple sandwich enzyme-linked immunosorbent assay technique. J Nutr 2004; 134:3127–3132.
17. Torre D, Tambini R, Manfredi M, et al. Circulating levels of granulocyte macrophage colony-stimulating factor in patients with the systemic inflammatory response syndrome. J Infect 2003; 47:296–299.
18. Gray RH, Wabwire-Mangen F, Kigozi G, et al. Randomized trial of presumptive sexually transmitted disease therapy during pregnancy in Rakai, Uganda. Am J Obstet Gynecol 2001; 185:1209–1217.
19. Klein LL, Gibbs RS. Use of microbial cultures and antibiotics in the prevention of infection-associated preterm birth. Am J Obstet Gynecol 2004; 190:1493–1502.
20. Levine WC, Pope V, Bhoomkar A, et al. Increase in endocervical CD4 lymphocytes among women with nonulcerative sexually transmitted diseases. J Infect Dis 1998; 177:167–174.
21. Gilbert RO, Elia G, Beach DH, et al. Cytopathogenic effect of Trichomonas vaginalis
on human vaginal epithelial cells cultured in vitro. Infect Immun 2000; 68:4200–4206.
22. Miller W, Swygard H, Hobbs M, et al. The prevalence of trichomoniasis in young adults in the United States. Sex Transm Dis 2005; 32:593–598.
23. Sorvillo F, Smith L, Kerndt P, et al. Trichomonas vaginalis
, HIV, and African-Americans. Emerg Infect Dis 2001; 7:927–932.
24. Price MA, Zimba Da, Hoffman IF, et al. Addition of treatment for trichomoniasis to syndromic management of urethritis in Malawi: A randomized clinical trial. Sex Transm Dis 2003; 30:516–522.
25. Van Der Pol B, Williams JA, Orr DP, et al. Prevalence, incidence, natural history, and response to treatment of Trichomonas vaginalis
infection among adolescent women. J Infect Dis 2005; 192:2039–2044.
26. Patton D, Agnew K, Hillier SL. Abstracts of the Annual Scientific Meeting and Symposium/macaque model for Trichomonas vaginalis
infection. Infect Dis Obstet Gynecol 2005; 13:103.