To date, the use of xenodiagnosis to assess the efficacy of treatment, measured as the infection capacity of treated dogs naturally infected with L. infantum, has only been reported in four studies including 4, 2, 36 and 10 dogs respectively [20, 26, 28, 29]. The present survey along with the study by Ribero et al. (2008)  was conducted on a sufficiently large sample size for valid conclusions to be drawn.
As may be observed in Figures 1 and 2, the most effective of our treatment regimens in terms of clinical improvement was the combined use of antimonials and allopurinol (Group A). Notwithstanding, we and other authors have noted that in some measure all treatments are able to improve the clinical condition of dogs [20, 26, 28, 29].
By examining the capacity of the dogs to infect sand flies, we observed a considerable reduction in the infectivity of the dogs in response to the three treatments tested. Only 7.7% of the entire study population was able to infect sand flies after the second month of treatment. This finding suggests that similarly treated dogs will not play a major role in promoting or maintaining new foci of disease spread.
The liposome formulation of meglumine antimoniate has been described to significantly diminish the capacity of dogs to infect Lutzomyia longipalpis five months after the end of treatment . In prior work, we observed a lack of infectivity towards sand flies of 6 dogs treated with meglumine antimoniate and allopurinol for at least 4 months post-treatment onset . In early studies, treatment with meglumine antimoniate was found to reduce the percentage of sand flies becoming infected after feeding on treated dogs [26, 29].
The bone marrow culture results indicate that regardless of a favorable clinical outcome of treatment, 30.8% of the dogs still showed the presence of the parasite in some of the check ups. These animals can be described as not parasitologically cured, or carriers of the parasite.
In the survey conducted in Brazil, bone marrow parasite burden was significantly reduced 4 days after the end of treatment with meglumine antimoniate . In addition, it was observed that five of six dogs treated with meglumine antimoniate plus allopurinol still harbored parasites in the spleen 10 months after treatment .
It should be highlighted that in the treatment Group C of the present study, 80% of the dogs remained infected at the end of the study. This would appear to point to the ineffective nature of treatments based solely on allopurinol since, despite showing clinical improvement, these still infected animals could experience disease recrudescence. However, xenodiagnosis of the dogs in Group C indicated no evidence of the spread of the parasite to the phlebotomine sand flies. This finding has significant epidemiological implications in that although allopurinol may not be an effective shock treatment, it could perhaps be used in longer treatment maintenance courses (6-12 months). Notwithstanding, this should be done under supervision given the known long-term adverse effects of this agent (nephrolithiasis due to xanthine) [34, 35].
The bone marrow culture and xenodiagnostic findings indicate that by Day 180 after the onset of treatment (A, B or C), 81% of the dogs had been "cured" or the parasite had been eliminated, although in some dogs (11.1 to 25%) it persisted for a variable length of time.
In conclusion, our results highlight the need to treat sick dogs since these represent a serious epidemiological risk. Despite not being a simple method, the xenodiagnostic approach used here emerged as a useful tool to assess the infection capacity of dogs treated with new drugs and/or new treatment regimens. A reduction in infectivity to sand flies in response to treatment is a key factor to consider in control programs designed to eradicate active foci of canine leishmaniosis. In endemic areas there is a plenty of evidence indicating the effectiveness of repellents against sand flies in reducing the spread of Leishmania infection.