Canine leishmaniosis is a severe parasitic disease due to various species of diphasic protozoan parasites that belong to the genus Leishmania. It is transmitted by sandflies of the genus Phlebotomus in the Old World and Lutzomyia in the New World (Baneth 2006). Canine leishmaniosis on the Mediterranean basin is essentially caused by Leishmania infantum. Several drugs and various dosage regimens have been used for the treatment of canine leishmaniosis with questionable efficacy; amongst them, and until recently, only meglumine antimoniate (Glucantime®) was registered for the treatment of canine leishmaniosis (Bourdoiseau and Denerolle 2000). Allopurinol (a purine analogue with leishmanistatic effects; not registered as a veterinary drug) is frequently used in combination with pentavalent antimonials in the initial treatment phase, and then on its own as a maintenance therapy to control relapses. Other anti-leishmanial treatment options include amphotericin B, aminosidine, pentamidine, metronidazole, ketoconazole and fluoroquinolones, however these are not registered for use in veterinary medicine, or if registered (e.g. aminosidine), there are severe concerns about their side effects. In addition, these treatments have questionable efficacy and in some instances are cost-prohibitive (Noli and Auxilia 2005; Miró et al. 2008a). Currently, most authors point out therapeutic failure, relapses and toxicity related to the use of the treatment strategies available (Denerolle 1996; Baneth and Shaw 2002). In addition, resistances to pentavalent antimonials (including the most frequently used in veterinary medicine Glucantime®) were observed during the past few years in human and canine patients suffering from leishmaniosis (Gramiccia et al. 1992; Denerolle 1996; Pouyol 2004). Based on these observations, there is a need for new therapeutic protocols based on the use of new molecules which are easy to administer and less toxic. The antiprotozoal activity of phospholipid analogues has been demonstrated in the past decade (Croft et al. 2003). Miltefosine was originally developed as an anti-cancer agent in humans (Sindermann and Engel 2006) and for the treatment of human visceral leishmaniosis (Jha et al. 1999; Sundar et al. 1998, 2000, 2006; Bhattacharya et al. 2004).

It has recently been registered for the oral treatment of canine leishmaniosis in several European countries. In the present Good Clinical Practice compliant, multicentric, controlled and randomised field trial performed in France and in Spain (completed in 2004), the efficacy and safety of miltefosine were compared to meglumine antimoniate.

Materials and methods


One hundred and nineteen dogs were selected from client-owned dogs presented to different veterinarians. Among the 119 included cases, 29 were lost to follow-up or withdrawn from analysis of the efficacy results because of protocol deviations. Therefore, 90 dogs of various breeds were considered as interpretable for the efficacy analysis. The clinical and laboratory baseline values for dogs included in the trial indicated a homogenous population of dogs, with no statistically significant difference (p > 0.05) between the two treatment groups (Tables 1 and 2), indicating that the dogs in both groups are comparable. Out of the 90 enrolled dogs, 69 (76.7%) were pet dogs, 17 (18.9%) were hunting dogs and four (4.4%) had other living conditions. Out of the 29 cases excluded from the efficacy analysis, only seven were excluded from safety analysis, therefore 112 dogs were interpretable for safety analysis. All enrolled animals exhibited clinical signs of leishmaniosis following a natural infection by L. infantum (Table 3), confirmed by positive laboratory results for leishmaniosis (Table 4), and a serological negative result for ehrlichiosis (cut-off < 1/80 by Immuno-fluorescence Antibody Test (IFAT; Sainz et al. 1996). Following the non-inclusion criteria, dogs did not clinically exhibit severe renal insufficiency (creatinaemia <1.5 mg/dl, lack of compatible clinical signs), hepatic (values for ALT > 200 IU/l and ALP > 500 IU/l) or cardiac insufficiency, or a pustular cutaneous form related to bacterial super-infections and therefore requiring concomitant antibiotic treatment. The dogs were not pregnant or lactating females, and had not been treated for leishmaniosis during the 2 months prior to inclusion.

Table 1 Animal characteristics at baseline (D0) indicating a homogenous population of dogs at D0 (p > 0.05)
Table 2 Laboratory parameters before (D0) and after treatment (D42) in both treatment groups, Group M and Group G
Table 3 Clinical parameters evaluated for efficacy assessment. Total clinical score = ∑ individual score for each parameter
Table 4 Inclusion criteria

During the phlebotome season (main vector of L. infantum in the study sites), an external anti-parasitic association of permethrin and pyriproxyfen (Duowin® Spray, Virbac) was administered regularly. A concomitant treatment with an anti-seborrhoeic shampoo (Sebocalm®, Virbac) or an antiseptic and keratoplastic benzoyl peroxide shampoo (Paxcutol®, Virbac) was permitted in case of non-pustular cutaneous forms or pyoderma, respectively. The only authorised anthelmintic treatments were pyrantel pamoate (Canex® or Strongid®, Pfizer) and praziquantel (Droncit®, Bayer Pharma). Anti-emetic and anti-diarrhoea treatments with metopimazine (Vogalene®, Boehringer Ingelheim) or dypirone (Estocelan®, Boehringer Ingelheim) were allowed in case of gastrointestinal adverse effects. Associated treatments likely to modify the immune response (corticosteroids, serum, transfusions), vaccination and all drugs used for the treatment of canine leishmaniosis were forbidden.


The dogs were randomly allocated into two treatment groups, according to a randomisation table. The trial could not be blinded due to the different routes of administration between antimonials and miltefosine.

Sixty dogs were orally treated with miltefosine (Milteforan®, Virbac) at a dosage of 2 mg/kg per day for 28 days (group M) and 59 dogs received subcutaneous injections of meglumine antimoniate (Glucantime®, Merial) at a dosage of 100 mg/kg once daily (25 dogs) or twice daily in two equally divided doses (34 dogs) for 28 days (group G).

Clinical follow-up and samplings

Dogs were observed five times during 6 weeks. During a pre-inclusion visit, the case suspected of clinical leishmaniosis was selected, and samples (blood for serological, haematological and biochemical examinations [parameters examined outlined in Table 2] and a bone marrow aspirate in order to detect the presence of L. infantum amastigotes by smear examination and by nested-PCR) were collected for all dogs (Cruz et al. 2002). Three bone marrow smears were performed for each dog at pre-inclusion and D42 to assess the parasitological efficacy (Cruz et al. 2002). The qualitative nested-PCR was solely used as a diagnostic tool, given its high sensitivity.

In the whole study, all bone marrow smears were assessed by the same pathologist, although the quality of the samples was dependent on the practitioners’ sampling techniques. The microscopic examination of bone marrow smears was performed by the same highly qualified person to ensure consistency in the technique and results obtained.

After positive results for leishmaniosis and negative results for ehrlichiosis, the case was included in the trial, randomly allocated to either treatment group M or G, treated for 28 days, and followed up for 6 weeks from D0 to D42, with re-checks every 14 days. At each visit, the dog was clinically observed and scored; a clinical score was calculated by adding the points given to each of the 27 clinical parameters monitored (Table 3). On D42 samples (blood and a bone marrow aspirate) were collected again. Animals treated with systemic corticosteroids, which exhibited significant signs of intolerance to the product or severe renal insufficiency, which were not administered the entire treatment, or whose owners withdrew their consent, were excluded during the trial.

Data analysis

The efficacy of the two tested products was primarily assessed on D42 by comparing the percentage reduction of the clinical score between the two groups. The following parameters were considered secondary and compared between groups: time course of clinical scores, mean rectal temperature of animals over time, and on D42, clinical result (percentage of dogs with a reduction of clinical scores classified as excellent >80%, good = 60%–80%, fair = 40%–60%, and nil <40% reduction), mean percentage increase in the albumin/globulin ratio, percentage of animals with a decrease in IFAT titres of at least two dilutions, percentage of dogs with a negative result for Leishmania parasites on examination of bone marrow smears on D42, mean bodyweight of the animals, mean of each haematological and biochemical value, and percentage of animals with abnormal haematological or biochemical values.

Criterion of evaluation for a good tolerance of the products was the percentage of dogs with product-related adverse effects. The number of dogs in each treatment group with abnormal levels of urea, creatinine and hepatic enzymes before and after treatment was also taken into account. Quantitative parameters were compared between groups using a one-way ANOVA or a Kruskal–Wallis test according to assumption validity. Qualitative parameters and proportions of dogs were compared between groups using a Chi-square test or a Fisher’s exact test. The time course of the clinical score between D0 and D42 was compared using a mixed effect model with group and time as fixed effect, and subject as a random one. The difference between the administered dosage and the theoretical one was tested in each group with a Wilcoxon signed rank test. The last observation carried forward method was used for missing final data: the last observation was retained until the last record. The threshold of significance for all statistical analyses was fixed at α = 0.05 (5%).


Efficacy assessment

Clinical efficacy

A continuous, steady and significant improvement of clinical scores in both treatment groups was observed throughout the study (p < 0.0001). On D42, the clinical score was reduced by 51.1% and 63.4% in groups M and G respectively, with no statistical differences between groups (Fig. 1). The distribution of the clinical results on D42 was not significantly different between groups, with 24 (52%) and 30 (68%) of the dogs presenting an excellent or good assessment in groups M and G, respectively.

Fig. 1
figure 1

Evolution over time of the mean total clinical scores

Between pre-inclusion and D42 the mean bodyweight did not show any significant difference between treatment groups (p = 0.2542), and the time course of the rectal temperature between D0 and D42 was similar in the two treatment groups (p > 0.05).

Parasitological efficacy and laboratory parameters

The parasitological efficacy, based on those dogs that were positive on D0 and became negative on bone marrow smear examination on D42 of 90% and 91.3% was achieved in groups M and G, respectively, and did not significantly differ between treatment groups (p = 1.0; Table 5). The percentage of dogs showing a decrease in serological IFAT titre did not correlate with the significant decrease in the clinical scores observed in both groups (only 9.1% of dogs in group M and 33.3% in group G).

Table 5 Parasitological results on bone marrow smears at D42 (analysis was only performed for dogs with positive bone marrow smears at pre-inclusion) demonstrating no significant difference between Group M and Group G (p > 0.05)

The mean percentages of increase in the albumin/globulin ratio between pre-inclusion and D42, the proportion of dogs with an albumin/globulin ratio that became higher than 0.7, the haematological values, and the proportion of dogs with, at least, one abnormal haematological or biochemical value on D42, did not show any significant difference between treatment groups (p > 0.05). On D42, there was a significant difference of the values for the total plasma protein (p = 0.0024) and albumin/globulin ratio (p = 0.0070) between the two treatment groups (Table 2), however, as stated above, there was no difference between the evolution of these parameters over time between the two groups.

Amongst the biochemical parameters monitored, a more significant percentage of dogs had increased levels of creatinine at the end of the trial compared to the baseline values at the beginning of the trial in group G (10.8%) compared to group M (0%; p = 0.0326). Similar results were observed for gamma-glutamyltransferase values (20% versus 2.4%; p = 0.0198). Both of these parameters were comparable between groups at D0 with p values > 0.05 (Table 6).

Table 6 Comparison of dogs with normal creatinine and GGT at D0, which became abnormal at D42

Safety assessment

Fifty-five and 57 dogs in groups M and G, respectively, were considered as interpretable for safety evaluation. The number of dogs with product-related adverse events was not significantly different (p = 0.0581) between the two groups [17 (30.9%) and nine (15.8%) in group M and G, respectively].

Out of the 17 dogs showing adverse events in group M, 17 (100%) had gastrointestinal signs (such as vomiting, diarrhoea, dysorexia or anorexia), two (11.8%) had general signs (such as lethargy, asthenia or depression) and one (5.9%) exhibited polyuria–polydipsia. Out of the nine dogs showing adverse events in group G, nine (100%) had gastrointestinal signs (such as vomiting, diarrhoea, dysorexia or anorexia), four (44.4%) had general signs (such as lethargy, asthenia, depression or weight loss) and three (33.3%) exhibited other clinical signs (such as local reaction at the injection site, aggravation of limb oedema and lameness).

The majority of clinical signs concerned the gastrointestinal tract in both treatment groups, and vomiting was the most frequently observed: ten (58.8%) dogs in group M and five (55.6%) dogs in group G. Dogs vomited at any time during the treatment period and only once per day for 1 or 2 days, and less frequently lasted for more than 3 days. Only two and three dogs in groups M and G, respectively, required symptomatic treatment.


Most of the clinical trials based on the therapeutic efficacy of different principles active against L. infantum natural infection in dogs show a good clinical improvement, and in some cases, a total remission of clinical signs (Valladares et al. 2001; Pasa et al. 2005; Oliva et al. 2006; Ikeda-García et al. 2007). Our results also demonstrate a clinical improvement of treated dogs in both groups.

Related to the clinicopathological findings in the treated dogs, we found that values from the CBC and biochemical profile in a variable period were steady until the end of the study. Similar results have been shown by different authors with the use of conventional drugs (Liste and Gascon 1995; Valladares et al. 2001; Koutinas et al. 2001; Pasa et al. 2005).

A decrease in the serological titres by IFAT results were not representative, as they did not appear to correlate with the clinical improvements observed during the trial period. This result is similar to previously published data, in which most authors consider this parameter unrepresentative of the clinical evolution of treated dogs (Solano-Gallego et al. 2001a; Pennisi et al. 2005). It appears that a decrease in Leishmania serological titres cannot be expected within the first few weeks after initiating treatment, and that it may not be a reliable parameter to monitor the initial response to treatment. A decrease in the Leishmania serological titres may potentially be observed over a longer time period, and this trial is limited due to a follow-up period of only 6 weeks. A further study with a longer follow-up is needed, in order to determine whether a decrease in Leishmania serological titres occurs over time with either treatment.

Finally, it is currently very difficult to assess the parasitological cure of dogs suffering from leishmaniosis, and to compare the parasitological results from different authors in published clinical trials due to the different techniques that are used, the different drugs that are used to treat dogs, and the multiple different protocols used. Multiple methods of assessment have been documented; however, there is no standardised consensus on the ideal method to use. In addition, multiple variables can interfere, such as different sampling techniques used by veterinarians, the transport of samples, and the ability to assess the sample (whether performed in-house or sent to a laboratory for examination).

A limiting factor of this study is that it did not evaluate the quantitative parasitic load of the dogs. However, at the time of this study, real-time quantitative PCR techniques were not yet recognised, controlled and standardised in veterinary medicine. When this clinical trial was initiated in 2002, nested-PCR was the best diagnostic tool combined with parasite evidence from bone marrow aspirates to establish a parasitological diagnosis. Due to the extreme sensitivity of the nested-PCR analytical method, it was used as a confirming diagnostic tool at the beginning of the trial. At the time of the trial, bone marrow smear examination was indeed the most widely used and rapid method for the detection of Leishmania parasites in the field. It is also one of the most reliable tests since it allows the direct observation of parasites on bone marrow smears in a simple way, leading to an excellent specificity. Its sensitivity is also satisfactory when the sampling has been performed correctly (Alvar et al. 2004). In this study, all smears were observed by the same qualified and specialised pathologist, and the number of dogs with negative bone marrow smears at the end of the trial was very satisfactory in both treatment groups. We have to assume that in the majority of clinical studies performed with dogs suffering from canine leishmaniosis, the dogs remain positive, even though they show a good clinical response (Solano-Gallego et al. 2001b; Saridomichelakis et al. 2005; Moreira et al. 2007). So, in conclusion, dogs can achieve a clinical cure, but they remain parasitologically positive (Riera et al. 1999; Alvar et al. 2004; Gomes et al. 2008). This has been confirmed in more recent studies using real-time PCR assays in dogs treated with miltefosine (Manna et al. 2008b; Miró et al. 2008b) and meglumine antimoniate (Manna et al. 2008a).

Until 2005, the reference protocol for the treatment of canine leishmaniosis was the combination of meglumine antimoniate and allopurinol. The use of meglumine antimoniate as a monotherapy has become unfavourable due to its considerably high relapse rate (Noli and Auxilia 2005). The benefits of allopurinol lie not only in its synergistic effect but in the prevention of relapses, when it is used for a long period of time (Denerolle and Bourdoiseau 1999).

In the last 3 years, different trials have been carried out to evaluate the efficacy of miltefosine (alone and in combination with allopurinol; Manna et al. 2008a, b; Miró et al. 2008b, accepted and presented as a supporting original study at the 6th World Congress of Veterinary Dermatology and under review to be published in Veterinary Dermatology Journal). Miltefosine has been classified as a second-line treatment therapy for canine leishmaniosis (Miró et al. 2008a), as at the time of the extensive review there was no published data on the use of miltefosine in the treatment of canine leishmaniosis. The present results are very promising, and further publications of clinical studies should enable miltefosine to be considered as one of the first-line drugs for the treatment of canine leishmaniosis.

Although clinical studies for evaluating the efficacy of miltefosine in the treatment of natural canine leishmaniosis are few (Manna et al. 2005, 2008a, b; Miró et al. 2005), this study clearly demonstrated that miltefosine has clinical and parasitological efficacy similar to the control group in the treatment of natural canine leishmaniosis.

In addition, miltefosine is the first oral anti-leishmanial drug with a high degree of safety and efficacy for the treatment of visceral leishmaniosis in human beings: studies done in children have shown the safety and efficacy of miltefosine (2.5 mg/kg daily for 28 days) and it was established with a cure rate (94%) similar to that seen in adults (Sundar et al. 2006).

Moreover, miltefosine appears to exert a lower impact on parameters related to liver and kidney function than meglumine antimoniate in dogs, considering that a higher number of dogs had abnormal values at the end of the trial in the meglumine antimoniate group. With a new mode of action based on a direct anti-parasitic activity not dependent on a functional immune system, an ease of use related to its oral administration, and a low toxicity, miltefosine seems to meet all the requirements of a new alternative drug in the control of canine leishmaniosis.