High Rate of Failure in Treatment of Imported Schistosomiasis
Marie Helleberg, MD, Department of Infectious Diseases, Rigshospitalet, Blegdamsvej 9, DK-2300 Copenhagen Ö, Denmark. E-mail: firstname.lastname@example.org
Background. There is an increasing number of imported cases of schistosomiasis in Europe, but there are only few studies on the efficacy of praziquantel for the treatment of schistosomiasis in non-endemic settings.
Methods. Patients treated for schistosomiasis in 2003 to 2008 were offered reexamination with serology, eosinophil count, IgE, microscopy of 24 h urine samples and/or rectal biopsies >3 months after treatment. All patients had been treated with at least one dose of praziquantel 40 to 60 mg/kg >12 weeks after exposure and had not been reexposed to schistosomiasis after treatment.
Results. Twenty-eight traveler (15 tourists and 13 expatriates) and two immigrants were reexamined after treatment. Viable ova were detected in six traveler (20%). Ova were found in 5/23 (22%) rectal biopsies and in 2/12 (17%) urine samples. Treatment failure was suspected in a symptomatic patient who 2 years after treatment had eightfold rise in antibody titer and elevated IgE but no detectable ova in urine or rectal biopsies. Additional 13 patients had one or more parameters, which could indicate persistent infection.
There were no significant differences in eosinophil count, IgE or, change in antibody titer between patients with versus without detectable ova after treatment.
Conclusions. In traveler with a low parasite burden, assessment of treatment results can be difficult because of the low sensitivity of microscopy and persistence of antibodies for several years after treatment. We found a high rate of treatment failure among traveler, indicating that nonimmune patients may need more than the recommended single day of treatment for eradication of parasites. Until more sensitive and specific methods for detection of persistent, active infection are available, repeated treatment should be considered in patients with continuous symptoms or other indications of treatment failure even when viable ova cannot be detected by microscopy.
Schistosomiasis is transmitted by skin contact with contaminated freshwater (ie, swimming, fishing, or rafting) inhabited by snails carrying the parasite and can be transmitted even after brief exposure to freshwater in endemic areas. In European countries there is an increasing number of imported cases because of migration, international travel, and adventure tourism.1
The gold standard for the diagnosis of schistosomiasis is the detection of viable ova by microscopy of urine, feces and/or, tissue biopsies. In traveler, who usually only have a low parasite burden, ova are often not detectable and diagnosis relies on serology,2 which in patients with detectable ova has demonstrated good sensitivity for S. mansoni but somewhat lower sensitivity for other species.3 Antibodies can be detected for several years after treatment of the infection, and assessment of treatment effectiveness in traveler can be difficult.4,5
Praziquantel has been used to treat schistosomiasis for more than 25 years and is still the drug of choice.6 The mechanism of action of praziquantel is unknown, but one effect of praziquantel might be disruption of the surface membrane of schistosomes and exposure of antigens that can be attacked by antibodies, implying that efficacy of treatment depends on immunity of the host.7 Most studies on the efficacy of praziquantel are conducted in endemic areas, and data regarding the efficacy of treatment in non-endemic settings are limited. Here, we present an evaluation of treatment outcome in patients treated for schistosomiasis at the Department for infectious diseases at Copenhagen University Hospital in 2003 to 2008 and review previous reported studies of treatment results in non-endemic areas.
Materials and Methods
Study population: In 2003 to 2008, a total of 49 patients were treated for schistosomiasis at Copenhagen University Hospital. All patients were treated with at least one dose of praziquantel 40 to 60 mg/kg more than 12 weeks after exposure. At the time of the present study 11 of the 49 patients had been reexamined for ova at least 3 months after treatment; the remaining 38 patients, who had not been reexamined or examined with blood samples only, were offered follow-up examination by microscopy of 24 h urine samples and/or rectal mucosa biopsies and measurement of eosinophil count, IgE, and Schistosoma serology. Of the 38 patients 19 were reexamined and 19 did not respond to the invitation. None of the patients had been reexposed to freshwater in Schistosoma endemic areas after treatment.
Serology: Serum samples were examined by an indirect hemagglutination assay (Cellognost-Schistosomiasis, Dade-Gehring, Marburg, Germany) and/or by immunofluorescence antibody test with measurement of antibody titer against gut associated antigen (GAA) and membrane bound antigen (MBA) at Statens Serum Institute, Denmark.
Rectal biopsies: Two biopsies from the rectal mucosa were obtained by proctoscopy and were examined under a microscope as a crush preparation between two slides at 3 × 10 magnification.
Urine: 24 h urine samples were filtered under vacuum; the filter was stained with ninhydrine and examined under a microscope.
Feces: Two samples of feces were concentrated using the formol-ether technique and examined by microscopy. Feces was examined at the first consultation but not at follow-up because of the low sensitivity of this method compared to microscopy of rectal biopsies.
Definitions: Treatment failure was defined by the finding of viable ova in rectal biopsies or urine >3 months after treatment. Viability of the ova was confirmed by finding a well-defined fully developed miracidium in unfixed fresh ova by microscopy, using a high-power objective. Microscopy was performed by a laboratory assistant, who has several years of experience in parasitology.
Statistical analyses: Differences between groups were analyzed by Mann-Whitney ranksum test using Stata 9.2 software.
This study was conducted as part of quality control of treatment of schistosomiasis in our department and was approved by the Danish Data Protection Agency.
In 2003 to 2008, 49 patients were treated for schistosomiasis; 10 were immigrants, 19 were tourists and 20 were expatriates. Of these, 43 patients had been exposed to freshwater in Eastern- or Southern parts of Africa, 5 in West Africa and 1 in Brazil. Patients were diagnosed by detection of ova in urine (8), feces (1) or in tissue biopsies (10) or by serology only (31). Ova of Schistosoma haematobium were found in 10 patients, S. mansoni in 6, and 2 patients had coinfection of S. haematobium and S. mansoni.
Of the 30 patients, who were reexamined after treatment, 2 were immigrants, 15 were tourists, and 13 were expatriates. At the time of diagnosis 26 of the 30 patients were examined by serology and tested positive. Ova were detected in 12 of 30 patients (40%) (urine: 6, feces: 1, and tissue biopsies: 6). There were seven cases of S. haematobium, three of S. mansoni and two with coinfection of S. haematobium and S. mansoni.
Patients were reexamined 3 to 36 months after treatment (mean 16 months). At follow-up, viable ova were detected in 6 of 30 patients (20%), ova had also been detected at the time of diagnosis in these patients. Of these six patients, four were tourists and two were expatriates; four had S. mansoni, one S. haematobium and one had coinfection of S. mansoni and S. haematobium (Table 1). Viable ova were detected in 5 of 23 (22%) rectal biopsies and in 2 of 12 (17%) urine samples. Treatment failure was suspected in a symptomatic patient who 2 years after treatment had eightfold rise in antibody titer and elevated IgE but no detectable ova in urine or rectal biopsies.
Table 1. Summary of exposure, treatment, and findings in six patients with treatment failure
|1||M||22||Lake Victoria, Jun.07||May 08: 60 mg/kg||Rectal snip, Oct.08||S. mansoni||None|
|2||F||34||Lake Victoria, 98+00||Oct.03: 40 mg/kg||Rectal snip, Jan.04||S. mansoni||None|
|Jan.04: 60 mg/kg||Rectal snip, Apr.04|
|Apr.04: 60 mg/kg||Rectal snip, Oct.04|
|3||F||63||Madagascar, Nov.06||Feb.07: 60 mg/kg||Rectal snip, Mar.09||S. mansoni||Loose stools|
|4||M||60||The Nile, Jun.04||Nov.05: 40 mg/kg||Rectal snip, Mar. 06||S. mansoni||Abdominal pain|
|5||F||40||Lake Malawi, 96||Feb.06: 40 mg/kg||Urine, Dec.06||S. haematobium||None|
|Mar.05: 60 mg/kg||Serology, Sep.05|
|6||F||32||Rwanda 03-05||Sep.05: 60 mg/kg||Rectal snip, Jun.06||S. haematobium + S. mansoni||None|
|Jun.06: 60 mg/kg||Urine, Jul.07|
There were no significant differences in IgE, eosinophil count, or presence of symptoms between patients with versus without detectable ova at follow-up. IgE was elevated in 1 of 4 patients with, and in 8 of 18 patients without detectable ova (p = 0.6). Eosinophil count was elevated in 0 of 4 patients with, and in 6 of 19 patients without detectable ova (p = 0.2). Symptoms (abdominal pain, flank pain, loose stools and/or, blood in stool) were reported by 2 of 6 patients with, and by 5 of 24 patients without detectable ova (p = 0.6). At follow-up serology was positive in 19 of 25 patients. Titers had been measured by the same method twice in 17 patients and had increased in 2 patients (3 months and >2 years after treatment, respectively), were unchanged in 4 patients and had decreased in 11 patients. In one patient, in whom titers had decreased since the time of diagnosis, viable ova were detectable at follow-up.
Only 10 of 30 patients did not have any sign of treatment failure defined as detectable ova in urine or rectal biopsies, elevated eosinophil count or IgE, and increase in antibody titer or symptoms. Patients with detectable viable ova or suspected treatment failure were given another course of praziquantel and offered follow-up examination after 3 to 6 months.
When reviewing results of treatment in imported cases of schistosomiasis, we found treatment failure, defined by the finding of viable ova, in 6 of 30 patients (20%) and further 14 patients showed one or more signs of failure of treatment.
Assessing outcome of treatment in traveler with only light infections and limited excretion of ova is difficult. Antibodies are detectable for several years after treatment and in traveler titer may peak 6 months after treatment.5 In the present study we found that a decrease in titer was not a reliable marker of success of treatment as viable ova were found in biopsies of the rectal mucosa of a patient, in whom titer had decreased. Eosinophil count and IgE are neither sensitive nor specific.2 Continuous elevation of eosinophil count and IgE can either be caused by treatment failure or by a number of other parasitic or nonparasitic diseases. Among our limited number of patients we found no association between eosinophil count and IgE and detection of viable ova at follow-up.
Examination of tissue biopsies and large samples of urine seems to be the most sensitive methods for the detection of viable ova after treatment, but presumably sensitivity is not higher than when these methods are used at initial diagnosis, when ova are detected in only <50% of traveler with positive serology.2,5,8,9 Until more sensitive and specific methods for assessment of treatment results are available, repeated treatment should be considered in patients with symptoms or other indications of treatment failure even when ova are not detectable. Alternatively, given the low toxicity of praziquantel, repeated treatment of all nonimmune patients after 1 to 3 months might be reasonable.
In a recent study by Wichmann et al. polymerase chain reaction (PCR) for the detection of parasite DNA in plasma samples demonstrated high sensitivity and specificity in diagnosis and assessment of treatment results among traveler.10 Further clinical studies of this method are needed.
Previous studies reporting results of treatment of schistosomiasis in traveler are summarized in Table 2. Only studies reporting results of examination for ova at follow-up are included. Generally rates of treatment failure were high. Additionally, several case reports indicate that failure in treatment of schistosomiasis in traveler is not uncommon.11–17 The study by Whitty et al., including 550 traveler, found a low rate of parasitological treatment failure compared with other studies.8 In that study, biopsies were not performed and ova were only searched for in feces and small urine samples, which could have compromised sensitivity.
Table 2. Summary of studies reporting results of treatment of schistosomiasis in non-endemic settings
|Helleberg et al. 2009||30||East Africa (25), West Africa (4), Brazil (1)||40–60 mg/kg * 1–3||24 h urine and rectal biopsies||6/25 (24%)||9/22 (41%)||7/30 (23%)||6/30 (20%)||S. mansoni (5).|
|S. haematobium (2)|
|Duus et al.5||98||Tropical Africa (93), North Africa (3), Asia (2)||40 mg/kg * 1–2||24 h urine, feces and rectal biopsies||22/98 (22%)||10/90 (11%)||ND||3/20 (15%)||ND|
|Grandiére-Pérez et al.18||18||Mali||40 mg/kg * 1||Unine and feces||4/18 (22%)||ND||2/18 (11%)||10/18 (56%)||S. haematobium|
|Silva et al.19||26||Mozambique||40 mg/kg * 1||24 h urine * 3 + cystoscopy with biopsy||ND||ND||ND||9/26 (34%)||S. haematobium|
|Roca et al.20||14||West/central Africa (12), Egypt (1), Brazil (1)||40 mg/kg * 1/4†||Feces * 3||ND||ND||ND||0/14||None|
|Whitty et al.8||550||Africa||40 mg/kg * 1/3||Terminal urine and feces||19%||ND||ND||7/550 (1.3%)||ND|
|Rabello et al.21||18||Brazil||50–60 mg/kg * 1||Feces||1/4 (25%)||ND||8/18 (44%)||4/18 (22%)||S. mansoni|
|Jelinek et al.22||62||Africa (58), Amazone (2), Mekong (1), Iraq (1)||40 mg/kg * 1||Feces and urine||2/62 (3%)||ND||ND||4/62 (6.5%)||S.mansoni|
|Visser et al.23||28||Mali||40 mg/kg * 1–2||10 mL urine, feces||ND||ND||0/8||8/22 (36%)||S. mansoni (7)|
|S. haematobium (1)|
It could be debated if low cure rates should raise concern as only few patients had symptoms and the symptoms were mild. However, we believe that elimination of parasites is important even in asymptomatic patients because symptoms often develop several years after exposure,24 and at that time it may not be acknowledged that they are caused by schistosomiasis. Another concern is the risk of development of severe neurological complications, such as seizures, ataxia, acute transverse myelitis, or subacute myeloradiculopathy because of the inflammatory response of the host to deposition of ova in the brain or spinal cord.25
There are several possible explanations for the high rate of failure in treatment of schistosomiasis in traveler. Praziquantel has only limited effect against schistosomules,26 and if the infection is treated in the invasive phase, low cure rates can be expected. Grandiére-Pérez et al. studied the efficacy of praziquantel in patients recently exposed to schistosomiasis and found that early treatment could prevent development of symptoms of acute schistosomaisis but failed to prevent chronic schistosomiasis in 17 of 18 patients.18 Doherty reports treating 16 patients with Katayama syndrome, 7 patients required further courses of praziquantel because of continuing symptoms, persisting eosinophilia and/or a subsequent rise in the antibody titer.27 In a study conducted by Rabello et al. patients were treated day 26 to 57 postexposure and at follow-up, viable ova were found in fecal samples from 4 of 18 patients,21 in spite of the fact that the sensitivity of microscopy of feces is relatively low, when the parasite burden is low. Roca et al. reports treating 14 traveler of whom 4 had Katayama syndrome and received a 3d course of praziquantel 40 mg/kg. Treatment was repeated after 3 to 4 weeks. All patients were considered cured as ova could not be detected in feces 3 months after treatment.20
Drug resistance could be a cause of the observed high rate of treatment failure, but even though some studies have shown that schistosomes in certain areas have reduced sensitivity to praziquantel, clinically significant drug resistance has not been documented.28 In the studies summarized in Table 2, treatment failure occurred among traveler, who had acquired the infection in many different areas where occurrence of drug resistance has not been suspected.
Studies conducted in endemic areas have generally shown higher cure rates than those found among traveler.29 This could be because of the use of less sensitive methods (ie, Kato-Katz technique for fecal samples) when assessing results of treatment in endemic areas. Another explanation might be that host-immunity is an important factor for the efficacy of praziquantel in the treatment of schistosomiasis.7,30 The finding, that in endemic areas cure rates are higher in adults than in children,31 further supports this hypothesis. Repeated doses of praziquantel might improve treatment outcome in nonimmune traveler. Whitty et al. found that treatment failure was less common in patients treated for 3 days versus those treated 1 day, but the difference was not statistically significant, possibly owing to the overall low rate of treatment failure documented.8 To our knowledge there are no prospective, clinical studies comparing the efficacy of different regiments of praziquantel in treatment of the chronic phase of imported schistosomiasis. Given the high rate of treatment failure among traveler, such studies are needed.
We found a high rate of failure in treatment of imported schistosomiasis, indicating that traveler may need more than single day treatment with praziquantel for eradication of parasites.
Until more sensitive and specific methods for assessment of treatment results are available, repeated treatment should be considered in patients with continuous symptoms or other indications of treatment failure even when viable ova are not detectable. Alternatively, given the low toxicity of praziquantel, repeated treatment of all nonimmune patients after 1 to 3 months might be reasonable. Clinical studies comparing the efficacy of different regiments of praziquantel in treatment of imported schistosomiasis are needed.
Declaration of Interests
Both authors state they have no conflicts of interest to declare.