| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Post mortem studies of the reproductive tract, confirmed by histopathologic analysis of material from surgery, have shown that S. haematobium may be distributed in all the pelvic organs. Ova are most commonly found in the cervix (30–100%), followed by the vagina (12–100%), ovaries (5–57%), fallopian tubes (3–36%), vulva (7–17%), and uterus (0–11%).10–14
Histopathologic studies have shown that S. haematobium ova in genital tissue may be surrounded by eosinophils, lymphocytes, epitheloid cells, macrophages, foreign body giant cells, and multinucleate histiocytes. Schistosoma haematobium ova may be viable, non-viable, intact, or partly sequestered, sometimes with little inflammatory reaction, fibrosis, and calcification.6,12,15–17
Gynecologic S. haematobium has been associated with a range of lesions morphologically similar to the different sexually transmitted infections (STIs) and reproductive tract diseases, such as ulcers, papillomatous tumors, leukoplakia, malignancies, and polyps.2,17–21
The aims of the present study were to further describe the morphologic aspects of S. haematobium-induced pathology and find the predictors for lesions in the lower female genital tract in a population also examined for the common STIs. We sought to establish the prevalence of lower genital tract S. haematobium infection in a rural female Zimbabwean population large enough to allow for multivariate analysis of possible predictors.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) where most laundry and bathing was done. No schistosomiasis control program had been conducted in the adult population. Women were recruited from the area immediately around the clinic of Mupfure from three pick-up points (Chihuri, Kaziro, and Fountains) and from the areas surrounding these pick-up points. Based on the last census, 353 and 706 women between the ages of 20 and 49 lived close to the clinic/pick-up points and surrounding areas, respectively.22 The pick-up points were chosen for their proximity to the river and to the road. The objectives of the study were explained to the women in a series of public meetings, and it was emphasized that gynecologic diseases may be subclinical and that all women should have regular Papanicolaou (Pap) smears. The clinic provided free condoms and treatment of common diseases. Local village health workers listed female household members assumed to be sexually active. Prior to the investigation dates, the village health workers informed pre-listed women who would either walk to the clinic or to a pick-up point. All women in the indicated area were invited.
|
Virgins, pregnant, postmenopausal or menstruating women, those who refused to undergo gynecologic examination or participate, and those who had other serious diseases were excluded. All included and excluded patients found positive for urinary or genital schistosomiasis were treated with prazi-quantel (40 mg/kg) in one dose. Patients with symptoms or signs of STIs or other diseases were treated in accordance with the Zimbabwean standard syndromic approach23 and referred to tertiary institutions when necessary. Partner treatment slips were given to the index STI case; these ensured free treatment in the district or at the research clinic.
Clinical examination. A nurse questioned the women in Shona (the local language) for urogenital symptoms, obstetric history, and current and past STIs, whereupon 10 mL of blood was collected by venipuncture. At the time of the examination, the clinician did not know the result of the S. haematobium examination in urine. The examination started with cervico-vaginal lavage. Saline (5 mL) was sprayed on the vaginal wall and cervix twice, whereupon it was drawn back into a syringe and deposited into four tubes. This was followed by photocolposcopic examination (Leisegang Photocolposcope, Script-O-Flash; Leisegang Feinmechanik-Optik GmbH, Berlin, Germany) at three magnifications (7.5x, 15x, and 30x) using an autoclaved metal speculum, after which Pap smears were done in all consenting women. The mucosal and vulval surfaces were inspected section by section according to a predefined protocol. Upon suspicion and if the patient agreed, specimens for STI or cancer diagnoses were taken. However, if the woman wished, the investigation was interrupted and the husband (or the head of the household) was invited to give his opinion. Acetic acid or iodine application for colposcopic examination was always done last. The investigation was finalized by bimanual examination.
Slides from the photocolposcope in cases of doubt were presented before a diagnostic committee consisting of the clinician for the project, an experienced gynecologist, and a specialist in infectious diseases. Neo-vascularization was defined as pathologic convoluted (corkscrew), reticular, circular, and/ or branched, uneven-caliber blood vessels visible (by 15x magnification) on the mucosal surface. Contact bleeding was defined as fresh blood originating from the mucosal surface. Pre-contact bleeding was defined as darkened blood on the mucosal surface in the absence of recent or present menstruation. A polyp was defined as a single, smooth, pedunculated mass originating from the endocervix or from the mucosal surface. Leukoplakia was defined as a white plaque on the mucosal surface, visible with or without acetic acid. Papilloma was defined as a sessile mass either on the mucosa or vulva that was whitish in color, often with a cauliflower appearance.
Parasitology. Three types of specimens were used to verify the presence of S. haematobium ova in genital tissue, Pap smears, wet mounts, and biopsy specimens. Pap smears were taken from all consenting women whereas wet mounts24 were taken from consenting women who showed debris or friable, loose, and bleeding tissue. Biopsy specimens were obtained only from consenting women when there was suspicion of malignancy. The specimens were each divided into two sections, one section was used for histologic analysis and the other was retained to make a crushed biopsy specimen that was examined microscopically.5,25 Urine samples were collected on three consecutive days, and 10 mL were processed by filtration and examined by microscopy for ova.26 The presence of a single terminal, spined ovum gave a positive diagnosis. A single stool sample was processed by the Kato-Katz technique.
Other specimens and laboratory tests. Blood specimens were centrifuged and serum was separated. The serum and lavage specimens were stored at –20°C immediately after collection. All laboratories that received samples were blinded to subject characteristics. Examinations for STIs were carried out last; thus, for some women these were done only when there was sample material remaining after other examinations. Antibody to herpes simplex virus type 2 (HSV-2) was detected with the HerpeSelectTM IgG enzyme-linked immunosorbent assay (Focus Technologies, Cypress, CA). Cultures of ulcer exudates and a polymerase chain reaction (PCR) of ulcer lavage samples were conducted for Haemophilus ducreyi.27 Sera were tested for syphilis using the Treponema pallidum hemagglutination assay (TPHA; Randox, Crumlin, United Kingdom) and rapid plasma reagin (RPR; Spinreact SA, Olot, Spain). Sera from women with ulcers, condylomata, a positive TPHA result, or a positive RPR test result were also tested for antibodies to Treponema using the ICE Syphilis (Abbott Murex, Dartford, United Kingdom) and were reinvestigated on follow-up sera for titer changes (after 3 or 12 months). A diagnosis of active syphilis was based on positive serology, evidence of seroconversion, and the presence of an ulcer. Sera were tested for human immunodeficiency virus type 1 (HIV-1) with Genelavia Mixt (Sanofi Diagnostics Pasteur SA, Marnes la Coquette, France) and Recombigen (Trinity Biotech Bray Co., Wicklow, Ireland); the Vironostica assay was used in cases of discrepant results (Organon Technika, Boxtel, The Netherlands). Evidence of trichomonas infections was detected by serology28 and microscopic examination of Pap smears. We searched for Candida in Gram-stained samples and in Pap smears, and gram stains were performed for Neisseria gonorrhoeae in 22 women with a discharge. In addition, a ligase chain reaction was conducted on lavage samples for N. gonorrhoeae and Chlamydia trachomatis (LCX; Abbott Laboratories, Abbott Park, IL). A gram stain of the lavage specimen was read to detect lactobacilli and neutrophils, and bacterial vaginosis was diagnosed using Nugent’s criteria.29 Pap smears (up to two smears per person over a three-month period) were investigated for cell atypia, and cervical intraepithelial neoplasia grades I-III (low to high grade squamous intraepithelial lesion). A random selection of specimens was tested for human papillomavirus by GP5/6 probes and carcinogenic human papillomavirus (HPV) probes (HPV high-risk types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68, 73, and 82 and probably high-risk types 26, 53, and 66) by a PCR.30 The presence of HSV-1 was not analyzed due to the high seroprevalence countrywide (> 90%) and donovaniasis was not tested due to negligible prevalence in Zimbabwe.31 Tests were conducted by the Blair Research Institute (Harare, Zimbabwe), the Biomedical Research and Training Institute (Harare, Zimbabwe), the Department of Medical Laboratory Sciences (Harare, Zimbabwe), and Medical Oncology Section of the Department of Medicine (University of Antwerp, Antwerp. Belgium). The Zimbab-wean laboratories subscribe to the Zimbabwe National Quality Assurance Program.
Statistical methods. Chi-square and Fisher’s exact tests (for numbers < 5) and odds ratios (ORs) with 95% confidence intervals (CIs) were used when studying the association between laboratory results and clinical pathology in the genitals. To study simultaneously the impact of several variables, logistic regression analysis was applied on those variables with a 5% significance level in bivariate analysis. Where there were less than 10 cases in the denominator or numerator, the variable was not included in multivariate analysis. The biopsy and wet smear tests were applied only in the presence of certain clinical findings and due to a high probability of a tautologic argument these tests were not used in analyses against clinical findings. The Pap smear was the only diagnostic test for S. haematobium ova used irrespective of findings and was therefore chosen for analysis against clinical parameters. Statistical analysis was done using the Statistical Package for the Social Sciences version 11 (SPSS, Chicago, IL) and Epi-Info version 2000 (Centers for Disease Control and Prevention, Atlanta, GA).
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). In the surrounding area, enrollment was 233 (33%) of the 706 women living there. Women in the two groups did not differ with respect to prevalence of the investigated diseases, age, or personal characteristics and the results have therefore been combined.32 Almost all (486, 92%) of the women had grown up in a rural area, and 225 (46%) of them came from the study area itself. Most (444, 84%) women were married, 33 (6%) were widowed, 23 (4%) were divorced, 19 (4%) cohabited, and 8 (2%) were single. Three-fourths (397, 76%) of the women were housewives. The rest were vegetable vendors, small-scale cash crop farmers, or teachers. Most (463, 88%) women reported regular water body contact (bathing, washing clothes, etc.) and 91% had water body contact during childhood. We excluded 59 women: 1 had rectal cancer, 3 had cervical cancer, and 1 did not agree to have a gynecologic examination, as well as 13 virgins and 41 post-menopausal women. Ninety-three of the non-attendees listed by village health workers were approached at home, and 47 of these had an address at a pick-up point. Among the non-attendees listed by the village health workers, 14% had menstruated at the appointment time, 9% were pregnant, 11% had left for one or several visits to other areas, 21% of the non-attendees declined to come, approximately 33% did not participate in the study because their husbands refused, 37% were not found on three visits or were said to be busy in the fields, and 8% of the non-attendees proved not to be within the target group. We therefore have reason to believe that non-attendance was not systematic and that the examined women were representative for the area.
Table 1 shows that sandy patches was the most common pathologic finding in the genitals, found in almost half of the female population between 20 and 49 years of age, 44% of the women had neo-vascularization, and nearly 25% had contact bleeding. Each woman was tested for a median of seven STIs (range = 1–8), and 382 (73%) had a laboratory confirmed history or presence of at least one STI, of which HSV-2 se-ropositivity (64%) was the commonest, 29% had HIV-1,32 23% had carcinogenic human papillomavirus, and 10% had low-risk HPV.30 Urinary S. haematobium ova were found in 39% of the women. No women were found to have S. mansoni in genital specimens, but 16% were found to have S. mansoni in feces.
|
). The grains (approximately 0.05 x 0.2 mm long and shaped like minuscule rice grains in clusters of up to 300) were deeply or superficially situated in the mucosa. The mucosal surface over the deeply grained patches was smooth and grains were not moveable. Superficial grains were situated within the mucosa. The movable, distinct, minuscule, crust-like, superficial protrusions could be felt with a metal spatula and sometimes even heard. The other presentation showed yellow homogenous foci, found in 154 (29%) women (Figure 3). The foci were seen at a magnification of 15x as sandy areas with no distinct grains. Fifteen women had both grainy and homogenous yellow sandy patches that were found up to 6.0 cm from each other on the genital mucosal surface. The sandy patch diagnosis was dependent on the use of a strong light, and in 39% of the cases the magnification of the colposcope was needed.
|
|
. Table 3 shows that S. haematobium ova in genital specimens were the only laboratory results significantly associated with genital grainy sandy patches. However, the second type of sandy patch, homogenous yellow sandy patches, was also significantly associated with other STIs (Table 4) and thus cannot be said to be pathognomonic for presence of ova. The homogenous yellow sandy patches were tested against all clinical entities listed in Table 3, but only significant results are presented. Less than 50% (222 of 527) of the cases were analyzed for high-risk human papillomavirus and there were less than 10 cases of C. trachomatis; thus, these two variables were excluded from the multivariate model. Multivariate analysis was done on the chosen diagnostic test (Pap smear) and HSV-2. Both S. haematobium ova on a Pap smear and HSV-2 are associated with homogenous yellow sandy patches (adjusted OR = 3.03, 95% CI = 1.29–7.1, P = 0.011 and adjusted OR =1.62, 95% CI =1.02–2.58, P = 0.042, respectively). Age, S. mansoni infection, use of hormonal contraceptives, or use of intravaginal substances did not influence the association between S. haematobium ova and sandy patch types significantly.
|
|
|
and 4). Genital S. haematobium ova in Pap smears were analyzed along with laboratory test results for other diseases to find predictors for mucosal abnormalities (Table 5). Neo-vascularization was associated with both S. haematobium ova in Pap smears and cervical intraepithelial neoplasia grades I–III in bivariate analysis, as well as in multivariate analysis. Similarly, the presence of genital S. haematobium ova in Pap smears, along with other clinical entities, was significantly associated with pre-contact bleeding, contact bleeding, mucosal edema, and erosion and remained significantly associated by multivariate analysis (Table 5).
|
|
Thirty-nine women had either a papillomatous lesion and/ or leukoplakia, 15 of whom had schistosome ova in genital tissue. Leukoplakia and papillomata (together or separately) were not significantly associated with the finding of S. haematobium ova in genital tissue (
2 = 0.39–0.74). The same was true for ulcers and polyps. None were significantly associated with the presence of S. haematobium ova in the mucosal membrane or with sandy patches (2 = 0.21–0.62).
Squamous intraepithelial lesions. The clinical suspicion of malignancy was significantly associated with the finding of genital S. haematobium ova in Pap smears (OR = 3.21, 95% CI = 1.03–10.07, P < 0.045). However, in multivariate analysis the histopathologic finding of atypical cells in this population was significantly associated only with high-risk human papillomavirus30 and not with the presence of genital S. haematobium ova.
Urinary schistosomiasis and female genital schistosomiasis.
Urinary schistosomiasis was not included in the multivariate analysis as a predictor for the sandy patch types because it has the same causal pathway. However, in univariate analysis, the grainy and the homogenous yellow genital sandy patch types were significantly associated with urinary S. haematobium infection (OR = 1.94, 95% CI = 1.34–2.79, P < 0.001). Tested individually, the presence of ova in the urine was significantly associated with homogenous yellow sandy patches (OR = 1.82, 95% CI = 1.23–2.76, P = 0.003) and grainy sandy patches (OR = 2.22, 95% CI = 1.49–3.29, P < 0.001). The overall prevalence of genital schistosomiasis (sandy patches and/or S. haematobium ova in genital tissue) was 49% (Table 2). Genital schistosomiasis was found in 58% of the women with S. haematobium ova in urine, but also in 41% of the women without detectable S. haematobium ova in urine.
Urine ova excretion decreased significantly with age (Table 6). There was no significant difference between the age groups with respect to genital sandy patches, although there was a tendency that the young have less sandy patches than the older women. Schistosoma haematobium ova in genital tissue was also tested against age and followed the same trend as that of sandy patches, but the cases were too few to reach significance.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Other reproductive tract diseases such as cell atypia, infection with HSV-2, and syphilis were also associated with homogenous yellow sandy patches, bleeding, or pathologic vessel morphology.30,32 However, the presence of S. haematobium ova was consistently a stronger predictor for these clinical entities.
The sandy patch grains were measured in retrospect upon evaluation of the photocolposcopic slides. The length of the S. haematobium ovum may be up to 0.186 mm, which is very close to our roughly estimated length of the superficial and deep sandy grains. It has also been suggested by other investigators that each grain represents one S. haematobium ovum, but this fact still needs to be ascertained.10
Previous studies in women have reported that the observed proportion of genital schistosomiasis is an underestimate of the true prevalence.2 In the present study population, 49% of the women had genital schistosomiasis, and 26% of the sandy patches were confirmed by the presence of genital S. haematobium ova. Biopsy specimens were obtained under suspicion of malignancy only and, although deemed to be the golden standard for diagnosis, S. haematobium ova are located in highly focal clusters and may be missed in a biopsy specimen.6,14,25 In addition, the two other test methods in this study, the Pap test and wet smears, reportedly have low sensitivity for the diagnosis of genital schistosomiasis.2,5,25 We therefore assume that our prevalence figures underestimate the true prevalence of genital schistosomiasis in this rural female population.
To our knowledge, this is the first study on female genital schistosomiasis in which women have been tested comprehensively for STIs in a sample large enough for multivariate analysis. Unexpectedly, and contrary to previous reports,7,19 after controlling for other causative agents, S. haematobium in the genitals was not associated with leukoplakia, papillomata, polyps, or ulcers. Schistosoma haematobium ova may be found in almost every organ and our data indicate that the coexistence of eggs and lesions may not be causal.2,11–15 The clinical appearance of genital schistosomiasis may be mistaken for malignancy, but as reported by others, we found no association between genital schistosomiasis and cell atypia.12,19,35,36 In countries where schistosomiasis is endemic, laboratory diagnostic tests for preclusion of STIs or malignancy are not readily available. These data seem to indicate that women could be carriers of ova while a particular lesion may be caused by another undiagnosed phenomenon. Biopsy alone is therefore an imprecise tool in diagnosing a particular lesion.
It is not known whether the sandy patch types represent coincidental manifestations, genetic virulence variations, differences in the infected humans, or duration of the presence of ova in the tissues. The different urinary tract lesions have been suggested to represent depositions of ova commencing at different times but evolving at similar rates.15 The lack of lesion differences in the age groups of this study might indicate that genital lesions do not undergo a change with age, or that women by the age of 20 have already passed into a chronic state of infection. In the latter case, emphasis should be put on treatment of girls because we would expect little or no effect of treatment on lesions in woman.37 However, the effect of treatment on genital S. haematobium infection has not yet been explored satisfactorily.
The concomitant blood vessel friability and inflammation may give weight to the circumstantial evidence for bidirectional transmission of HIV.38 There is growing evidence of a rural HIV epidemic in Zimbabwe.39 Thus, the possible attributable risk of genital S. haematobium infection on HIV transmission and the effect of treatment on the genital sandy patches, concomitant inflammation, and blood vessel pathology need to be established to prioritize intervention strategies in countries hit severely by the two epidemics.
Received May 20, 2004. Accepted for publication June 21, 2004.
Acknowledgments: Technical, medical, or cultural assistance was provided by Dr. S. Charimari (Provincial Medical Director and Supervisor); the Mupfure community; the staff at Madziwa, Harare Central, and Mount Darwin Hospitals; the personnel from the Blair Research Laboratory; Drs. T. Magwali, T. Mhlanga, M. Chirara, B. Vennerwald, and I. Lyngstad-Vik; and Professors F. Jerve and L. Sandvik. We are indebted to the Medical Research Council of Zimbabwe; the staff at Mupfure Secondary School; headmistress V. Mugabe; and the following indispensable people: Sister J. Chikoore, the late Sister P. Dungare, Councillor C. Chadzimura, the village health workers, the environmental health technicians, and in particular N. Taremeredzwa, C. Mukahiwa, R. Manyaira, and T. Mushipe for prolonged hard work under very difficult circumstances.
Financial support: This study was supported by the Director’s Initiative Grant, United Nations Development Program/World Bank/ World Health Organization Special Program for Research and Training in Tropical Diseases, The Norwegian Research Council (NORAD), The Department for Infectious Diseases, Competence Centre for Imported and Tropical Diseases and Research Forum, Ullevaal University Hospital (Oslo, Norway), and the Danish Bilharziosis Laboratory.
Authors’ addresses: Eyrun Floerecke Kjetland, Centre for Imported and Tropical Diseases, Department of Infectious Diseases, Ullevaal University Hospital, 0407 Oslo, Norway, Telephone: 47-97-00-85-79, Fax: 47-22-11-91-81, E-mail: e.f.kjetland{at}medisin.uio.no. Patricia D. Ndhlovu and Takafira Mduluza, College of Health Sciences, University of Zimbabwe, Harare, Zimbabwe, E-mails: pndhlovu{at}rfc.ucl.ac.uk and mduluza{at}medic.uz.ac.zw. Exnevia Gomo, College of Health Sciences, University of Zimbabwe, Harare, Zimbabwe and Blair Research Institute, Harare, Zimbabwe, E-mail: exgomo{at}mweb.co.zw. Lovemore Gwanzura, College of Health Sciences, University of Zimbabwe, Harare, Zimbabwe, E-mail: lgwanzura{at}healthnet.zw. Peter R. Mason, College of Health Sciences, University of Zimbabwe, Harare, Zimbabwe and Biomedical Research and Training Institute, Harare, Zimbabwe, E-mail: pmason{at}vet.uz.ac.zw. Edith Nyaradzai Kurewa and Nicholas Midzi, Biomedical Research and Training Institute, Harare, Zimbabwe, E-mails: enkurewa{at}hotmail.com and midzi{at}blair.co.zw. Henrik Friis, Department of Epidemiology, Institute of Public Health, University of Copenhagen, Copenhagen, Denmark, E-mail: h.friis{at}pubhealth.ku.dk. Svein Gunnar Gundersen Research Unit, Sorlandet Hospital/Department of Health and Sports, Agder University College, Kristiansand, Norway, E-mails: s.g.gundersen{at}medisin.uio.no and s.g.gundersen{at}vas.no.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. F. Kjetland, P. D. Ndhlovu, E. N. Kurewa, N. Midzi, E. Gomo, T. Mduluza, H. Friis, and S. G. Gundersen Prevention of Gynecologic Contact Bleeding and Genital Sandy Patches by Childhood Anti-schistosomal Treatment Am J Trop Med Hyg, July 1, 2008; 79(1): 79 - 83. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. H. King and M. Dangerfield-Cha The unacknowledged impact of chronic schistosomiasis Chronic Illness, March 1, 2008; 4(1): 65 - 79. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Fenwick Waterborne Infectious Diseases--Could They Be Consigned to History? Science, August 25, 2006; 313(5790): 1077 - 1081. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
