INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Hydatid disease, caused by infection in humans or animals with Echinococcus granulosus, poses a significant economic and public health problem in many temperate and tropical areas of the world (4, 11, 15, 16, 21-23). Sheep, the natural intermediate host for the parasite, provide an excellent model of hydatid disease for drug trials, as the disease process in sheep closely resembles that seen in humans (19). In both species, Echinococcus larvae settle principally in the liver or lungs, although cysts have been documented in the bone, brain, heart, spleen, kidney, and orbit. In humans, as larval cysts begin to enlarge 5 to 30 years after infection, signs consistent with a slowly expanding neoplasm occur. Cysts occasionally rupture with pyrexia, urticaria, anaphylaxis, and/or abdominal seeding (25). Despite the relatively rapid growth of cysts (12), infected sheep are rarely symptomatic, and diagnosis is generally made at slaughter.

In human hydatid disease, perioperative chemotherapy to reduce cyst viability remains an important adjuvant to surgical removal of cysts (18). Chemotherapy will likely prove a useful supplement to the more recently developed puncture-aspiration-injection-reaspiration (PAIR) therapy (1, 27), as well as in cases of pulmonary or multiple hepatic cysts or in other instances where ultrasound is impractical. In cases of inoperable cysts, multiple cysts, or recurrent disease or in developing countries where prevalence is high and surgery and/or PAIR is cost prohibitive, chemotherapy alone is invaluable (19). Effective chemotherapy would also be of great benefit in the treatment of Echinococcos multilocularis (13).

Albendazole is currently the benzimidazole of choice for perioperative prophylaxis and treatment of inoperable cases of hydatid disease in humans (5, 14, 18). However, prolonged, repeated, high doses of albendazole are often necessary and result in cure rates approaching only 30% (13). No effective cestode larvicide exists for the treatment of Echinococcus in sheep; little research has been pursued in this area, as the control of the disease at the level of the primary host Canidae is generally more cost effective and practical (25). However, recent work in swine with cysticercosis (10) suggests that oxfendazole, another benzimidazole, may be effective in the treatment of cestodes in sheep and, ultimately, in humans.

This project evaluated the efficacy and safety of oxfendazole against hydatid disease in a sample of naturally infected sheep. The 14-consecutive-day treatment mimicked a typical treatment regimen of benzimidazoles in humans. The weekly and monthly dosing schemes served as potential future alternative schemes for the treatment of humans, perhaps with equal efficacy and fewer side effects than the daily treatment regimen currently in vogue.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

This study was approved by the ethical review boards of the Universidad Peruana Cayetano Heredia and The Johns Hopkins University.

Tupac Amaru, a sheep ranching cooperative located in the central Peruvian Andes (17), was selected to provide sheep naturally infected with E. granulosus. A total of 215 Tupac ewes greater than 4 years of age were chosen at random from 400 sheep destined for slaughter between February and May 1996. Ewes were destined for culling due to poor reproductive performance and/or tooth disease and loss.

All sheep were identified by an individual number and assigned to one of five groups. Sheep were individually weighed prior to dosing on a large, commercial livestock scale to the nearest 0.1 kg; appropriate dosages were administered via a Hauptner automatic oral dose syringe (Nasco) calibrated to 0.1 ml. In the first phase of the study, 57 sheep were treated with either oral oxfendazole (Synthatec, 22.5% solution) or oral placebo at 30 mg/kg of body weight/day for 14 consecutive days. Sheep from these daily dosing groups were monitored twice daily for side effects including lethargy, anorexia, and diarrhea during the initial 14 days of the study. Sheep were slaughtered at a cooperative abattoir in two groups at 86 and 93 days (12 and 13 weeks) following the initial dosing. All sheep were pastured in normal range conditions throughout the study.

In the second phase of the study, 158 sheep were assigned to one of three groups and were treated with either oral oxfendazole or oral placebo at 30 mg/kg once per week. One group received oxfendazole once per week for 7 to 11 weeks, and a second group received oxfendazole at weeks 0 and 4 and placebo at weeks 1 to 3 and 5 to 11 to simulate a monthly dosing schedule. The third group served as a control for both groups, receiving placebo once per week for 7 to 11 weeks. Due to slaughterhouse constraints, sheep from the second phase of the study were slaughtered in 5 groups, at 51 to 79 days (7 to 11 weeks) following initial dosing. Final doses for each group were scheduled to correspond with recommended drug withdrawal periods prior to slaughter.

The lungs and livers from sheep were collected at slaughter, were individually identified, and were maintained in a cooler filled with lukewarm water until cyst analysis was performed within the next 6 h. All cysts that were found upon examination of parallel 1-cm slices through the parenchyma of the lungs and liver were examined. Number, location, diameter, type, and description of cysts were recorded. Normal cysts were found to be filled with clear fluid and to have a distinct white membrane and one or more separate chambers. Calcified cysts were hard and nodular with at least one internal chamber or had calcified, chalky deposits in the cyst wall. Degenerated cysts had blackened parynchemal adventitial layers and one or multiple chambers filled with turbid yellow-to-grayish purulent material, with or without a wrinkled degraded membrane.

Cyst fluid was extracted from each cyst with a syringe and was gravity sedimented in a test tube in an incubator (35°C) for a minimum of 0.5 h. Sediment from each cyst was observed under a light microscope for protoscoleces, which were then tested for their ability to exclude 1% aqueous eosin. Two hundred protoscoleces from each cyst were counted and characterized as dead (red) or live (clear). For those cysts in which protoscoleces were not immediately apparent, the contents of the cyst and cyst membrane were finely chopped and treated with 0.2% pepsin in balanced salt solution (adjusted to pH 2.0) and incubated and examined as above.

All dosing, evaluation of side effects, and viability testing were masked as to treatment group. The placebo consisted of fine corn flour (Maizina) diluted at 200 g per gallon, which adequately imitated oxfendazole in color and consistency and did not effect rumen metabolism. To provide further assurance of masking, all treatments were prepared by a noninvestigator and were administered from opaque gallon containers, labeled only with a randomly assigned group letter. At slaughter, livers and lungs from each individual were randomly assigned a new number by a noninvestigator in order to mask the letter group origin of the organs throughout the viability studies. Dosing codes were not released to investigators until initial analysis of all data was completed.

Data from the two control groups were combined and analyzed as a single unit. The unit of analysis was the individual animal; cysts were also analyzed independently of the animal of origin. For all analyses, cysts with protoscoleces were defined as fertile, cysts with live protoscoleces were defined as viable, and cysts with 100% dead protoscoleces were defined as nonviable. Viability calculations were performed using only fertile cysts as the denominator. Sheep were deemed cured if all cysts were nonviable; even a single viable protoscolex amid multiple nonviable cysts disqualified the individual sheep from the cured group. Sheep in which 40 to 99% of protoscoleces were dead were arbitrarily termed improved, and sheep in which less than 40% of protoscoleces were dead were arbitrarily defined as unchanged. All degenerated cysts were included in the viability analysis as nonviable. Viability analyses were performed both including and excluding these degenerated cysts, and results did not vary significantly.

For the analyses of individual sheep, the aggregate percent of dead protoscoleces from all cysts of each individual was calculated and then averaged with the results from all individuals in the group to obtain the mean aggregate percent of dead protoscoleces for each group. When using the individual cyst as the unit of analysis, the mean percent of dead protoscoleces was calculated directly across groups, irrespective of the number of cysts per animal. Means were evaluated using the Kruskal-Wallis test; where overall differences were found by the Kruskal-Wallis test, differences between treatment groups were compared by use of the Mann-Whitney U test with a Bonferroni adjustment for multiple comparisons. The chi-square test was used to compare tabulated percentages. Logistic regression models were computed using fertile cysts, nonviable cysts, degenerated cysts, and calcified cysts as dependent variables. Treatment group, cyst location, and cyst size were explored as potential independent variables. All analyses and logistic regressions were completed in SPSS and EPIINFO. For all statistical analyses, a significance level (P) of less than 0.05 was used to reject the null hypothesis.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Of the 215 ewes enrolled in the study, 9 (4%) were lost to follow-up, with no significant differences between groups. A total of 15 additional animals died during the study. In the weekly and monthly treatment groups, two (4%) and three (6%) enrolled animals died, respectively; these death rates did not vary significantly from the control group (4%). However, death rates were dramatically increased in the daily treatment group, with seven (24%) animal deaths. Four of the seven deaths in the daily group occurred during the initial 14-day dosing period. Two of these four animals died from acute fibrinous pneumonia and pleuritis, and one animal died secondary to a dosing gun injury with extensive pharyngeal necrosis. The fourth sheep died 6 days after initial dosing, exhibiting severe, progressive ataxia, opisthotonos, lateral strabismus, severe dyspnea, and hyperresponsiveness to stimuli. A deep laceration of the medial head of the gastrocnemius from a shearing injury 13 days prior was debrided and cleaned. Despite treatment with parenteral penicillin and supportive care, the sheep died with a presumptive diagnosis of tetanus. No additional gross abnormalities were found upon necropsy; histopathological test results were not available on this individual. The remaining 11 deaths occurred between days 17 and 49 and were evenly distributed among all groups. As these deaths occurred after animals had been returned to highland pasture, necropsies were not carried out on these animals. With these noted exceptions, no side effects, such as lethargy, anorexia, or diarrhea, were observed in any of the remaining sheep.

Complete fertility and viability data were collected on 19, 48, 47, and 77 animals in the daily, weekly, monthly, and control groups, respectively. In the control group, 71 (92%) sheep had at least one cyst. A total of 62 (81%) and 69 (90%) control sheep had at least one hepatic or pulmonary cyst, respectively. Fertile hepatic or pulmonary cysts were found, respectively, in 32 (52%) and 41 (59%) control sheep. Data was collected from 583 control group cysts, of which 282 (48%) had protoscoleces. A total of 275 (47%) and 308 (53%) cysts were respectively hepatic or pulmonary in origin; 136 (49%) and 146 (47%) hepatic and pulmonary cysts, respectively, were fertile. Complete fertility and viability data was available for an additional 106 treated animals with 804 cysts. The number of sheep with cysts, the total number of cysts per individual, the location of cysts, and the percentage of cysts with protoscoleces were not significantly different from those of the control group or between treatment groups at either the individual or cyst level, confirming that the randomization of groups was adequate.

The treatment results for individual sheep by group are shown in Table 1. All sheep in the daily group were either cured (88%) or improved (12%) following treatment. In the weekly and monthly treatment groups, respectively, 97 and 78% of animals were cured or improved, compared to 35% of animals cured or improved in the control group. The mean aggregate percent of dead protoscoleces in the daily, weekly, monthly, and control groups were 99, 93, 68, and 35%, respectively. The distribution of fertile and nonviable cysts and the mean percent of dead protoscoleces using the cyst as the unit of analysis are presented in Table 2. The number of cured sheep, the total number of nonviable cysts, the mean aggregate percent of dead protoscoleces, and the mean percent dead for cysts were significantly higher in all oxfendazole-treated groups than in the control group. The mean percentages of dead protoscoleces at both the individual sheep and cyst levels were significantly higher in both the daily and weekly groups than in the monthly group; the daily and weekly groups did not differ significantly from each other.

Logistic regression results are presented in Table 3. The odds of having fertile cysts did not differ between groups; however, pulmonary cysts were 1.4 times more likely to be fertile than hepatic cysts. The odds of having nonviable cysts in the daily, weekly, and monthly groups, respectively, were 131, 30, and 8 times that of the control group, after adjustment for location and size. Hepatic cysts were two times more likely to have nonviable cysts than were pulmonary cysts, after adjustment for group and size. For each 1-cm increase in cyst size, the odds of having protoscoleces present in the cyst increased by 7.5-fold and the odds of having a nonviable cyst decreased by 10%, after adjustment for group and location. There was no difference in protoscolex death rates between animals that were killed in the first week of slaughter and individuals slaughtered in the final period, 7 weeks later.

Logistic models indicated that the odds of having a degenerated cyst increased by 7.9-fold in the daily group and 4.7-fold in the weekly group, when adjusted for cyst location and size. The odds of having a degenerated pulmonary cyst were five times that of having a degenerated hepatic cyst, after adjustment for group and size. For each 1-cm increase in cyst size, the odds of finding a degenerative cyst increased by 1.5-fold, when adjusted for group and location. Degenerated cysts were rarely associated with live protoscoleces. Of the 77 degenerated cysts, no recognizable protoscoleces were recovered from 55 (71%) cysts; 19 (25%) cysts had 100% dead protoscoleces, and three (4%) had at least one viable protoscolex. The latter group is comprised of one sheep in the weekly group and two sheep in the control group. The percentages of sheep with degenerated cysts in the daily, weekly, monthly, and control groups were 66, 21, 16, and 10%, respectively.

Regression models also indicated that the daily and monthly groups had two times the odds of having calcified cysts when adjusted for location and size. Hepatic cysts were 1.7 times more likely to have calcified cysts than were pulmonary cysts, after adjustment for group and size. For each 1-cm increase in cyst size, the odds of finding a calcified cyst decreased by one-third. Protoscoleces were present in only 6% of calcified cysts. The percentages of sheep with calcified cysts in the daily, weekly, monthly, and control groups were 33, 44, 47, and 28%, respectively.

The mean baseline weight was 38.7 kg and did not vary significantly between groups. Mean weight gains in kilograms with standard errors (SE) in parentheses were 5.6 (0.9), 1.2 (0.8), 0.1 (0.7), and 0.8 (0.6) in the daily, weekly, monthly, and control groups, respectively. The daily group gained significantly more weight than all other groups, as determined by the Kruskal-Wallis test (P < 0.05). Linear regression indicated that the daily treatment group gained 3.2 kg for every 1-kg increase in the control group, while the monthly group lost 2.2 kg for every 1 kg gained by the control group (P < 0.0001) (data not shown). Baseline weight and weight gain were not associated with protoscolex death rates.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Naturally infected sheep were chosen as a model to assess the potential of oxfendazole in the treatment of human hydatid disease, as the disease process is similar in both species (19). In addition, the prevalence of hydatid disease in the Junin region of Peru is estimated to be 87% in sheep and 9.1% in humans (17). The dose of 30 mg/kg was selected based on the apparent efficacy of this dose of oxfendazole in a single treatment of pigs with cysticercosis (10); multiple doses were deemed necessary due to structural differences between the larval cysts or metacestodes of E. granulosus (hydatid) and Taenia solium (cysticercus). The dosing schemes were chosen to mimic the daily albendazole regimen currently in favor (13) while exploring the efficacy of staggered treatment regimens in the weekly and monthly dosing groups.

In this study, treatment with oxfendazole resulted in high protoscolex death rates in all treatment groups at both the individual and cyst levels. In the daily, weekly, and monthly groups, respectively, 100, 97, and 78% of sheep were either cured or improved following treatment, compared to 35% of control group animals cured or improved. The mean aggregate percent of dead protoscoleces in the daily, weekly, monthly, and control groups were 99, 93, 68, and 35%, respectively. These results are similar to those of a recent study by Blanton et al. wherein nine goats and four sheep were treated biweekly with 30 mg of oxfendazole per kg for 4 weeks. Postmortem protoscolex viability in that study indicated that 96% of protoscoleces in the treatment group were dead (7). In contrast to our study, wherein 32% of protoscoleces were also dead in the control animals, the study undertaken by Blanton et al. found no dead protoscoleces in the control group (7). Differences in Echinococcus strain, study population size, age of animals, or protoscolex examination protocol may have contributed to this difference.

Oxfendazole was twice as effective in hepatic cysts than in pulmonary cysts in this study. This is in contrast to cysts in humans, either where location is not a factor (26) or where the thinner-walled pulmonary cysts may respond more successfully to albendazole treatment (8). However, as in humans, larger cysts were more likely to contain viable protoscoleces (26), suggesting that oxfendazole may eliminate protoscoleces more quickly in smaller, younger cysts.

The occurrence of viable residual protoscoleces in some cysts of certain treated sheep in our study necessitates review. It is likely that longer treatment and follow-up with this dose of oxfendazole would be effective in eliminating all viable protoscoleces. Studies have found that 100% of protoscoleces treated with intracyst hypertonic saline died, but only after 24 weeks (1). In swine with cysticercosis treated with oxfendazole, cyst death often took more than 4 weeks (9). Sheep in this study were followed for a maximum of 13 weeks. In addition, failure to kill all protoscoleces may be due to pharmacodynamics of the drug in ruminants, incomplete penetration of some cysts, and/or variability in the sensitivity of the parasite. The large number of degenerated cysts found in this study, some with dead protoscoleces and/or degenerative membranes, is of interest. Both the study undertaken by Blanton et al. and a recent study of intracyst hypertonic saline treatment found similar hepatic cyst lesions (1, 7). Surgical intervention studies have shown that structurally disorganized cysts are generally nonviable and resolve with time (6, 20). This study suggested that up to 23% of pulmonary cysts in treated animals may be degenerated, compared to 4% degeneration in control animals; pulmonary cysts had nearly five times the odds of hepatic cysts of being degenerated. The relative abundance of degenerated cysts in the oxfendazole-treated groups suggests that oxfendazole may be directly involved in the destruction process of cysts. The fact that small numbers of degenerated cysts were found in the control group in this study might suggest that degeneration is a feature of the normal immune response to a damaged cyst. Oxfendazole may cause changes in the cyst membrane which permit a loss of host tolerance and/or decrease the ability of the parasite to protect itself from the host immune response. It is possible that the direct effect of oxfendazole may not kill the parasite, but rather the collateral immune damage occurring after the drug dissipates kills the parasite over a period of weeks.

The role of oxfendazole in the destruction of cysts is also supported by the higher occurrence of calcified cysts in the daily and monthly groups. Microcalcification of cysts was also found in swine treated with oxfendazole for cysticercosis (10). Cyst wall calcification in six of seven sheep treated with intracyst hypertonic saline has been reported (1). The odds ratio of 2.0 for calcified cysts in the treated groups of this study is statistically significant, but not highly so. It is possible that higher rates of calcification would be noted with longer follow-up of studied animals, as calcification of cysts in treated swine is often not observed for 2 to 3 months following treatment with oxfendazole.

The large relative weight gain in the daily treated group may be partially explained by the death of less robust animals in this group. However, this hypothesis is difficult to substantiate, given the low number of animals in the daily group and the lack of relevant clinical parameters on those animals that died. These results do parallel unexplained findings in a recent study in which oxfendazole-drenched lambs (two doses of 4.5 mg/kg) gained significantly more weight than the controls with no difference in levels of growth-promoting hormones (24). Increased intestinal parasite loads in the controls relative to treatment groups do not explain these weight gain differences, because the daily group weight gain was four times that of any other treatment group; monthly or weekly dosing should have effectively controlled intestinal parasites in these groups.

The dramatic death rate (24%) in the daily treatment group remains a matter of concern. In addition, at least one animal treated with oxfendazole and tetracycline died of pneumonia in the study by Blanton et al. (7). In another study, 2 of 10 sheep died of acute septicemia or pneumonia with bone marrow depression following daily treatment with albendazole (19). Recent work with oxfendazole in parasite-free lambs suggests that oxfendazole may depress humoral and cellular immunity even at much lower doses (4.5 mg/kg at 0 and 28 days) (24). Although the deaths of two of the sheep can be attributed to trauma, the most likely cause of the remaining deaths in the daily group remains immunosuppression secondary to the stress of daily dosing and inclement weather conditions, exasperated by the potential immunosuppressive effect of oxfendazole.

Interspecies variation in the pharmacodynamics of oxfendazole may have accounted, in part, for the drug-associated toxicity found in the daily treatment group. Peculiarities of ruminant metabolism, as well as low feed intake due to poor local grazing and the condition of the teeth in these older ewes, may have contributed to prolonged, high plasma oxfendazole levels (2, 3, 19). As this effect is less important in monogastric animals such as swine or humans, dosing schemes and toxicity levels may be difficult to interpret between species. Now that it has been determined that oxfendazole is effective, the optimal dosage required in humans can be determined. Obviously, careful monitoring of leukocyte counts and hepatic profiles during human treatment would be of paramount importance.

This trial demonstrates the efficacy of oxfendazole (30 mg/kg) in the treatment of naturally infected sheep with hepatic and/or pulmonary hydatidosis. Clinical trials which determine appropriate, safe doses of oxfendazole in humans, in combination with the success of animal model drug trials, may promote testing of oxfendazole as an inexpensive treatment for cestode infections in humans. Finally, this study not only demonstrates that oxfendazole is effective in a weekly format, but it also implies that staggered benzimidazole therapy for human hydatidosis may produce results similar to those obtained by daily dosing regimens, with significant reductions in both cost and adverse effects.