| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Coinfection with schistosomiasis and HIV does occur in areas where both infections are endemic. There are only a few epidemiologic studies on S. mansoni coinfection with HIV,1–4 and only one on coinfection with S. haematobium.5 Consequently, a lot remains unknown about the interactions between S. haematobium and HIV with respect to treatment efficacy, pattern of acquired immunity, clinical manifestations, and diagnosis.
It seems very probable that the effects of HIV on the immune system will also influence the efficacy of treatment, since host immune responses have been shown to enhance the efficacy of several schistosomicidal drugs, including praziquantel.6,7 Possible effects of HIV on praziquantel treatment are important because it is the most widely used drug due to its ease of administration, tolerable side effects, and low cost.
The evidence of a relationship between the immune response and praziquantel derives mainly from experiments on S. mansoni infections in mice with their immune status variously modulated.8 This synergy between chemotherapy and host immune response is antibody mediated.6 The exact mechanism of action of praziquantel on schistosomes is not well understood, but there is evidence that the drug functions by rendering the parasite more susceptible to elimination by the immune response of the host.7 Therefore, in humans coinfected with HIV the efficacy of chemotherapy against schistosomiasis might be impaired, depending on the extent to which the humoral immunity is dysfunctional.
Infection with HIV results in the depletion and dysfunction of CD4+ T cells. This may affect the humoral immunity, though it is the cell-mediated immunity that is impaired most. So far, only one study has evaluated the efficacy of praziquantel in the treatment of schistosomiasis in people coinfected with HIV. Karanja and others2 studied 15 subjects in Kenya coinfected with S. mansoni and HIV compared with 32 subjects who had only S. mansoni infection. They treated both groups with standard regimen of praziquantel and monitored them by fecal examination and plasma circulating cathodic antigen. They could not demonstrate any impairment in the efficacy of praziquantel.
We present here the first study on the interactions between S. haematobium and HIV-1. It was conducted in Zambia, in an area endemic for S. haematobium and HIV infections. The goal was to compare the efficacy of praziquantel treatment, in terms of egg excretion and presentation of symptoms, in schistosomiasis haematobium cases with and without concomitant HIV infection. In addition, the diagnostic performance of reported symptoms was compared in schistosomiasis cases with and without concomitant HIV infection prior to treatment.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Ethical considerations. Research and ethical clearance was granted by the Tropical Diseases Research Center, Ministry of Health (Ndola, Zambia). Written consent, including willingness to give blood and have it analyzed for HIV infection, was obtained from each subject or guardian before enrollment into the study. Confidentiality of immunologic results was guaranteed. Subjects who wanted to know their HIV serostatus were counseled before and after their status was revealed to them.
Recruitment and baseline survey. All permanent residents of the two study villages were eligible for enrollment into the study as long as they had all the inclusion criteria and none of the exclusion criteria. The inclusion criteria were an age of 10–55 years, S. haematobium infection, and written consent. The exclusion criteria were pregnancy, breastfeeding, chronic illness, or other acute conditions. These criteria excluded all AIDS patients.
Subjects were initially screened for S. haematobium infection. In Kawama village, where houses are close to each other, the Kawama primary school was used as the screening site. In Milopa village, where houses are spaced many kilometers apart, an outreach health center and two cereal-grinding mill stations where villagers take their maze grain for grinding were used as screening sites. Urine filtrations were performed on three consecutive days between 9:00 AM and 3:00 PM for eggs/10 mL using Nuclepore (Whatman International Limited, Maidstone, England) membranes (12 µm). The membranes were reused after each urine examination after a rigorous cleaning process to rid them of any S. haematobium ova using a technique described by Mshinda and others.9 Ten percent of the urine samples were re-examined by a second technician.
On the fourth day, subjects who had S. haematobium eggs in at least one of the three urine samples and fulfilled the enrollment criteria underwent HIV testing, which was done using the Capillus (Cambridge Diagnostics Ireland Limited, Galway, Ireland) HIV-1/HIV-2 test. The second confirmatory test, which was used on all samples found to be positive with the screening test and on 10% of HIV-negative samples, was the Bionor HIV-1 & 2 test (Bionor AS, Gulset, Norway). If there was any discrepancy between the results of these two tests, a Western blot test was performed as the final confirmatory test.
The enrolled subjects were treated at baseline with a single oral dose of praziquantel (40 mg/kg of body weight). Demographic data and data pertaining to activities likely to expose the subject to S. haematobium-infected water were collected, together with full clinical history data to establish the presence of symptoms of S. haematobium infection. Subjects were followed-up at 3, 6, and 12 months after the initial treatment.
Sample size. We calculated first the expected number of subjects in the two villages with dual infection who could be eligible for enrollment into the study. Based on unpublished results (Dr. M. Shehata, Personal Communication, 1999), we considered a 10% prevalence of HIV infection and two scenarios for the prevalence of S. haematobium infection; 30% or 40%. Assuming that the two infections are randomly distributed and that only 60% of the total population of 7,000 in the two villages were possibly available for inclusion, we estimated that there would be 126 co-infected subjects if schistosomiasis prevalence was 30% and 168 if it was 40%.
We then calculated the total sample size adequate to measure treatment failure at three months after chemotherapy, assuming a 10% prevalence of S. haematobium at three months in the HIV-uninfected group and a risk ratio of 2.00. For an HIV-uninfected to HIV-infected ratio equal to 4:1, the required number of HIV-infected individuals was 127, while for an HIV-uninfected to HIV-infected ratio equal to 2:1, the required number of HIV-infected individuals was 157. A similar approach was followed for calculating the sample size adequate to measure re-infection rates at 6 and 12 months. It thus appeared possible to enroll the necessary sample size using the earlier estimated number of subjects who would have dual infections to be between 126 and 168. All sample size calculations were based on a 95% confidence interval (CI) and on a power of 80%.
Statistical methods.
Data were double-entered in Epi Info version 6.04 (Centers for Disease Control and Prevention, Atlanta, GA) and analysis was performed using Epi Info and Stata version 7.0 (Stata Corp., College Station, TX) statistical software. The primary outcome measures used to evaluate the efficacy of a single dose of praziquantel treatment were 1) the parasitologic cure rate and the rate of reduction in infection intensity (at 3 months), 2) the rate and intensity of S. haematobium reinfection (at 6 and 12 months), and 3) the time to S. haematobium reinfection. We considered as secondary outcome measures 1) the reduction in prevalence of visible hematuria, heavy infection (
50 egg/10 mL), and reported history of S. haematobium-associated symptoms (at 3 months) and 2) the reported history of S. haematobium-associated symptoms and the prevalence of visible hematuria and heavy reinfection (at 6 and 12 months).
The parasitologic cure rate was defined as the proportion of individuals whose urine was clear of S. haematobium ova three months after an initial single dose of praziquantel. The rate of reduction in intensity of S. haematobium infection at three months was calculated for each individual as the difference between the baseline egg count and the egg count at three months after treatment divided by the egg count at baseline. The measure used for the egg count for each individual and each screening occasion was the maximum number of eggs observed in any of the three urine samples submitted by the individual.
Two cohorts of subjects were extracted to study S. haematobium reinfection. A cohort of those seen at three and six months after treatment who were cleared at three months and a cohort of those with complete follow-up until 12 months who were cleared at three and six months. Reinfection with S. haematobium at six (or 12) months was defined as the presence of S. haematobium ova in the urine at six (or 12) months in subjects who had been cleared of infection at three (and six) months. A third cohort of those who had parasitologic cure at three months and remained clear at 6 and 12 months was extracted from the data to study the evolution of reported history of S. haematobium between the two groups.
Logistic regression models were fitted to compare the parasitologic cure rates or infection prevalences between the two exposure groups at baseline and at subsequent follow-up examinations. Negative binomial models10,11 were applied to compare the average intensities of infection between the HIV-positive and HIV-negative groups, and to assess whether the reduction of egg counts after treatment was significantly related to exposure group. Random effects were introduced in the models whenever paired measurements per individual were analyzed to allow for between-individual variation. Adjustment for confounders such as age, sex, and village was also made whenever necessary. The reductions in prevalence of the different outcome measures between the baseline and the three-month follow-up surveys were compared among the two study groups by testing the significance of the interaction between the study groups and the survey using the likelihood ratio test (LRT) obtained by the relevant logistic regression models. The Cox proportional hazards model was used to compare times to reinfection between the HIV-infected and HIV-uninfected groups after adjusting for confounders. Sensitivity, specificity, positive predictive value, and negative predictive value were compared between the study groups to assess whether the performance of reported S. haematobium symptoms to diagnose heavy infection was altered in the HIV-coinfected individuals.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Three hundred fifty-six enrolled subjects completed all follow-ups, and 47 of them were coinfected with HIV. Losses to follow-up indicated in Figure 1 were the result of migration, refusals, death, absence from home, or the people were simply not traceable. There was no statistically significant difference in the number of losses to follow-up between the HIV-infected and HIV-uninfected groups in all cross-sectional surveys.
|
compares baseline data for HIV-infected and uninfected groups for age, sex, exposure to S. haematobium assessed by reported contact with infected waters, reported history of S. haematobium-associated symptoms, and parasitologic examination results. There was no statistically significant difference in the sex ratio or in the different reported contacts with S. haematobium-infected waters between the study groups. However, individuals coinfected with HIV were older than those without HIV infection. Subsequent analyses controlled for age.
|
2 = 5.64, P = 0.018).
The intensity of S. haematobium infection, measured by geometric mean egg count, was lower in subjects coinfected with HIV than in subjects without HIV coinfection (LRT = 8.35, P = 0.004) (Table 1
). There was no statistically significant difference in the prevalence of heavy infection (50 eggs/10mL of urine), visible hematuria, or reported history of dysuria and abdominal pains between the two groups. However, a significantly lower proportion of subjects coinfected with HIV reported a history of hematuria than subjects without HIV infection (LRT = 6.28, P = 0.012).
The internal validity of a reported history of S. haematobium associated symptoms in the diagnosis of heavy infection of S. haematobium was compared between the two groups (Table 2), taking the maximum egg count observed in any one of the three urine samples examined as the "gold standard." The results indicate significantly lower sensitivity and positive predictive value of reported history of hematuria in the HIV-infected group. Also, among individuals with heavy infection, the proportion who reported a history of hematuria was statistically significantly lower in the group coinfected with HIV (Table 3).
|
|
). Results obtained using a negative binomial model fitted on egg counts of the cohort present at three months show that after adjusting for age the average intensity was reduced by 99.81% (LRT = 1,136.14, P < 0.0001). This reduction did not differ significantly between the two study groups (LRT = 0.78, P = 0.37). The average egg intensity of the HIV-positive group at three months was lower by 17% than the egg intensity of the HIV negative group (Incidence Rate Ratio = 0.83, 95% CI = 0.58, 1.18), but this difference was not significant.
|
) was not significantly different between the two study groups. Dysuria was the only reported symptom where the prevalence reduction was statistically significant lower in subjects coinfected with HIV than in subjects without HIV (Table 4). Moreover, the prevalence of hematuria, heavy infection, and reported symptoms at three months was not statistically different between the two groups (Table 5).
|
. The two groups also did not differ significantly in the proportion of visible hematuria and heavy infection (Table 5). The Cox proportional hazards regression adjusted for age showed no significant difference between the two groups in the time to reinfection (LRT = 0.651, P = 0.52) although reinfection commenced earlier in subjects without HIV infection (Figure 2). Younger subjects tended to be at a higher risk of reinfection than older ones (LRT = 7.73, P = 0.005).
|
), although the prevalence of reported history of hematuria was lower (but not significantly) in the HIV-infected group at both follow-up times.
When we studied the evolution of reported history of S. haematobium in a cohort of 44 HIV-infected and 286 uninfected subjects who had parasitologic cure at three months, and had remained free of S. haematobium infection at 6 and 12 months (Figure 3), we observed that of the three symptoms, reported hematuria is the most closely associated with S. haematobium morbidity. The graph shows an initial sharp decrease at three months in the proportion of subjects reporting hematuria, but between 3 and 6 and 6 and 12 months the decrease was gradual. The decrease was the same in both HIV-infected and HIV-uninfected groups.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
None of the study subjects in either the HIV-infected or HIV-uninfected group had signs and symptoms of a chronic or terminal illness. There may have been people in the villages with AIDS, but they were not enrolled in the study because the budget and the logistics of work in rural areas did not allow us to carry out any estimation of HIV viral load or CD4+ T cell counts. However, in view of the absence of any obvious symptoms and signs of opportunistic infections (OIs) in all study subjects in the HIV-infected group, we could assume that the CD4+ T cell count in these subjects was above the 200 cells/µL cut-off point, since the probability of OIs is very high when the CD4+ T cell count is less than 200 cells/µL. Therefore, there was no possible confounding effect of severe immunodepression in our study.
The mean age of subjects in the HIV-infected group was significantly higher than in the uninfected group. This was expected since HIV is mainly transmitted heterosexually in this region and older subjects are more likely to be sexually active. The peak age for prevalence and intensity of S. haematobium in areas with an endemicity like that in our study region is 10–20 years. The intensity of S. haematobium infection was lower in subjects coinfected with HIV. The study design did not include another measure such as circulating antigen levels to evaluate the level of infection in the two groups. However, this was a randomized study and it is unlikely that the HIV-positive individuals were a selected group with lower intensity of infection. Lower egg output in the HIV-infected group was also found in another study with S. haematobium5 and for S mansoni.1. The most probable explanation is immunologic interaction between Schistosoma and HIV infections.13,14 In S. haematobium infection, granuloma formation around the schistosome egg is critical for the passage of the egg through the urinary bladder wall. It is generally accepted that granuloma formation is initiated by CD4+ T helper lymphocytes that promote cellular recruitment around the schistosome eggs,15 and that antibodies have no evident effect on granulomas.16–18 Dysfunction and depletion of CD4+ T cells are the hallmarks of HIV infection and the predominant causes of immune deficiency.19,20 Depletion of CD4+ T cells is a progressive process going on even in the absence of HIV disease/AIDS. Thus, in our study, although the group did not include anyone with AIDS, dysfunctional and depleted CD4+ T cells might still have led to impaired granuloma formation around the schistosome eggs and thus to a reduction in egg excretion in the urine.
Another possible effect of dysfunctional and depleted CD4+ T cells in coinfected subjects is the adaptation of the schistosome parasite to the definitive host. It has been reported that a female adult worm requires intact host cell-mediated immunity for normal egg production.21 Reduced fecundity of schistosome females has been observed in T cell-deprived animals.21 This would result in the reduced egg excretion observed in coinfected persons.
The dependence of schistosome egg excretion on the immune system could also explain the statistically significant lower proportion of heavy infections, prior to treatment, in subjects coinfected with HIV and 15 years old. Conversely, the 10–14-year-old HIV-infected group had just acquired HIV, so that the depletion and dysfunction of the CD4+ T cells would not have been as far enough advanced to reduce egg excretion significantly.
Another indication that the pathologic effects of S. haematobium infection are less marked in coinfected subjects was that the prevalence of a reported history of hematuria prior to treatment was lower in subjects coinfected with HIV than in subjects without HIV infection. If the size of the egg granuloma in subjects coinfected with HIV is smaller, there will be less blood loss. The reduced blood loss was probably the reason that the sensitivity and positive predictive value of reported hematuria in relation to heavy infection were lower in the group coinfected with HIV. This has important practical implications. Questionnaire-based methods provide rapid means of assessing the magnitude of urinary schistosomiasis in a given community in the absence of microscopy.22 Since these methods are based on observations of hematuria, there is a need to be carefully validated for use in communities where HIV is prevalent. We should also be cautious in the interpretation of the result regarding history of hematuria as there was no significant difference in the proportion of visible hematuria between the two groups. The discrepancy in the results obtained for reported and for visible hematuria is an interesting finding that requires further investigation.
An important conclusion of the study was that the efficacy of praziquantel in treating S. haematobium infections in subjects coinfected with HIV was not impaired. This confirmed the results of an earlier, small study with S. mansoni/HIV coinfection in Kenya.4 A probable immunologic explanation is that, as has been demonstrated in experimental animals, it is the antibody-mediated immune response, which is not affected by HIV infection, that is primarily responsible for the immune-dependent action of praziquantel in human schistosomiasis, and not the cell-mediated immune response.8
We also found no statistical difference between the HIV-infected and HIV-uninfected groups in either the intensity of S. haematobium infection reappearing after 6 and 12 months or the time to the reappearance of the infection. Reappearance of infection might be due to reinfection from outside, or to recrudescence of the infection if some worms might have not been killed by the drug but severely impaired, ceasing temporally egg production and recovering later. Infection from outside is not likely to have been different in the two groups as the analysis of exposure data collected during the baseline survey has shown. The absence of any difference between the two groups therefore suggests that the destruction of adult schistosome worms by a single dose of praziquantel at baseline was as efficacious in subjects coinfected with HIV as in those without HIV infection. Had the efficacy been affected as a result of a deficient immune response so that some adult worms had remained alive, we would have expected the rate of S. haematobium resurgence to be higher in subjects coinfected with HIV than in subjects without HIV infection. We would also have expected subjects coinfected with HIV to have a heavier intensity of infection reappearing after treatment than subjects without HIV infection. The results also indicate that resistance to reinfection with S. haematobium after praziquantel treatment in subjects coinfected with HIV was not impaired. In S. haematobium infection, resistance to reinfection after chemotherapy is antibody mediated.23,24 Several studies have shown a very strong correlation between high levels of IgE and resistance to reinfection with S. haematobium after chemotherapy.25,26 Infection with HIV probably did not prevent the production of sufficient IgE to mediate resistance to reinfection.
The persistence of reported history of hematuria, dysuria, and abdominal pain even in the absence of infection in both exposure groups at 3, 6, and 12 after treatment implies that regression of S. haematobium pathology lags behind decrease in intensity of S. haematobium infection after treatment in subjects with or without HIV coinfection.27 This in turn shows that the development of S. haematobium-related pathology following a single dose of praziquantel in subjects coinfected with HIV is similar to subjects without HIV infection.
We conclude from this study that a single oral dose of praziquantel (40 mg/kg of body weight) is still very effective in treating infection with S. haematobium in persons coinfected with HIV (without symptoms and signs of AIDS/HIV disease), and therefore that praziquantel can be used for effective community-level control of schistosomiasis in areas where HIV is also prevalent. Similarly, resistance to S. haematobium reinfection after chemotherapy is not impaired in subjects coinfected with HIV. However, the results also indicate that the questionnaire method for identifying communities at high risk of S. haematobium infection may underestimate the prevalence of infection in areas with a high prevalence of HIV.
Received July 3, 2002. Accepted for publication July 1, 2003.
Acknowledgments: We thank the populations of Kawama and Milopa villages for participating in the study; Drs. E. Kafwembe, R. Banda, and C. Banda for their support and assistance in the execution of the fieldwork; the nurses at Kakoso and Chawama clinics for their assistance; and the teachers at Kawama and Milopa schools for their help. We also thank Dr. Rosemary Musonda for facilitating the HIV testing of the blood samples, Dr. M. Stehata and Mr. J. Chiima for their assistance in selecting the study villages, and A. Katenga for her tireless dedication to the logistics of the study. Special thanks are given to Drs. Mushaukwa Mukunyandela and S. Wayling for their support of the study. We are also grateful to Jennifer Jenkins for critically reading the manuscript and very helpful suggestions.
Financial support: This work was supported by a World Health Organization (WHO)/TDR grant.
Authors’ addresses: Victor Mwanakasale, MacDuff Ziba, and Arthur Ernest, Tropical Diseases Research Centre, PO Box 71769, Ndola, Zambia, Telephone: 260-097-804-740, Fax: 260-2-621-112, E-mail: vicsale{at}zamtel.zm. Penelope Vounatsou and Marcel Tanner, Swiss Tropical Institute, PO Box, CH 4002 Basel, Switzerland, Telephone: 41-61-284-8282, Fax: 41-61-271-7951, E-mails: Penelope.Vounatsou{at}unibas.ch and Marcel.Tanner{at}unibas.ch. Thomas Y. Sukwa, World Health Organization/Africa Regional Office, Harare, Zimbabwe.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Butrous, H. A. Ghofrani, and F. Grimminger Pulmonary Vascular Disease in the Developing World Circulation, October 21, 2008; 118(17): 1758 - 1766. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
