Clinical validation of circulating immune complexes for use as a diagnostic marker of canine leishmaniosis

Introduction Canine leishmaniosis (CanL) is a systemic disease that affects dogs. When multiplication of the parasite cannot be controlled, dogs consistently show high levels of antigen and IgG antibodies, which lead to the formation of circulating immune complexes (CIC). Timely intervention to reduce the parasite load and CIC levels is crucial for preventing irreversible organ damage. However, a diagnostic test to quantify CIC levels is currently lacking. Methods In this real-world study, we aimed to examine the performance of a new ELISA to measure CIC levels in dogs naturally infected with Leishmania infantum. Thirty-four dogs were treated according to their clinical condition and followed for 360 days. Before (day 0) and after treatment (days 30, 90, 180, 270, and 360), all dogs underwent a physical examination, and blood samples were obtained for CBC, biochemical profile, serum protein electrophoresis and IFAT. Serum PEG-precipitated CIC were determined by ELISA. Results Our results indicate higher CIC levels in dogs in advanced disease stages showing higher antibody titres (p < 0.0001, r = 0.735), anemia (p < 0.0001), dysproteinemia (p < 0.0001), and proteinuria (p = 0.004). Importantly, dogs responding well to treatment exhibited declining CIC levels (p < 0.0001), while in poor responders and those experiencing relapses, CIC were consistently elevated. CIC emerged as a robust discriminator of relapse, with an area under the curve (AUC) of 0.808. The optimal cut-off to accurately identify relapse was an optical density of 1.539. Discussion Our findings suggest that declining CIC levels should be expected in dogs showing a favorable treatment response. Conversely, in dogs displaying a poor response and recurrent clinical relapses, CIC levels will be high, emphasizing the need for vigilant monitoring. These findings suggest that CIC could serve as a valuable biomarker for disease progression, treatment efficacy, and relapse detection in CanL. Our study contributes to enhancing diagnostic approaches for CanL and underscores the potential of CIC as a complementary tool in veterinary practice. As we move forward, larger studies will be essential to confirm these findings and establish definitive cut-offs for clinical application.

Participants were dogs brought to the hospital by their owners over the period May 2021 to April 2022 for evaluation after a first diagnosis of CanL, or follow-up for check-ups or due to a clinical relapse.Dogs were considered eligible if naturally infected with L. infantum [positive serology for Leishmania by indirect fluorescent antibody test (IFAT >1:100, positive cytology and/or positive PCR result)] and their disease was then graded as LeishVet stages I-IV.Before enrolment, we informed all dog owners about the study protocol, including the option to withdraw their dogs from the study at any time, and obtained their written consent.

Study protocol
Dogs were included in the study on day 0 (Visit 1 enrolment pre-treatment) and thereafter followed for 12 months.Appropriate treatment was started shortly after enrolment.Visits to our service were then scheduled for post-treatment onset days 30, 90, 180, 270, and 360 (Figure 1).Dogs underwent a thorough physical examination in each visit.Data were compiled on clinical signs including signs of CIC deposition (Figure 2).Each clinical sign was graded from 0 to 3 (low to high severity) to give an overall clinical score for each animal (maximum 68) (Table 1).
At each visit, blood samples were collected to determine serum CIC concentrations by PEG ELISA (11) (Figure 1).On days 0, 30, 180, and 360, routine laboratory tests were also performed to determine the clinical condition of the dog, which included complete blood count (CBC), biochemical profile (creatinine, urea, and alanine aminotransferase), serum protein electrophoresis, and anti-Leishmania immunofluorescence antibody test (IFAT) (15).Serial dilutions from 1:50 to 1:6400 were used and seropositivity defined by a cut-off ≥1:200.Urinalysis, including urine density, sediment analysis and protein/creatinine ratio (UPC), were also undertaken on days 0, 30, 180, and 360 on urine samples obtained preferably by cystocentesis, or by free catch into a sterile container.Quantitative PCR was also performed to confirm the infection on day 0, and to quantify parasite load at 6 and 12 months post enrolment.Based on physical examination and clinical-pathological findings, dogs were assigned to LeishVet clinical disease stages I-IV (3).

CIC isolation and quantification
A modified precipitation method with polyethylene glycol (PEG; Sigma-Aldrich, St. Louis, MO, United States) was used to separate free antigens and antibodies from CIC (11).PEG-precipitated CIC were pelleted by centrifugation, then reconstituted in 0.01 M phosphate-buffered saline (PBS) and stored at −80°C for further use.CIC were quantified using a Leishmania-specific ELISA method as described by Parody et al. (11).All samples were tested in duplicate, and the mean value was recorded.A mean OD of 0.274 + 3 standard deviations of negative samples was considered as the cut-off value; any sample exhibiting absorbance above the cut-off value was considered positive.

Treatment
The treatment for each dog was determined by one of the authors (a veterinarian specialist in infectious diseases) and was based on clinical history, physical examination, and laboratory findings.Clinically healthy infected dogs or those at stage I received no treatment or received only allopurinol for 6-12 months.Dogs in stages II or III were treated with a combination of meglumine antimoniate (MGA) (50 mg/kg subcutaneously q12h for 28 days) and allopurinol (10 mg/kg orally q12h for 6-12 months).When considered necessary, prednisone (0.5 mg/kg orally q12h for 3-4 weeks) was added to the treatment regimen in dogs in stage III to control signs of CIC deposition (10).Proteinuric dogs were treated following International Renal Interest Society Guidelines for CKD in dogs (16).
Relapses were also recorded and a second course of leishmanicide was administered if needed.Disease relapse was defined as significant Study protocol.The clinical study consisted of six visits, enrolment (pre-treatment), after completing 30 days of treatment, and four follow-up visits.IFAT, indirect fluorescent antibody test; CBC, complete blood count; UPC, urinary protein/creatinine ratio; CIC, circulating immune complexes; qPCR, quantitative polymerase chain reaction.The clinical response to treatment was assessed by examining clinical records, changes produced in clinical scores and laboratory variables over time, and the number of relapses produced during the study period.Good responders were dogs that showed improvement in clinical signs and laboratory abnormalities and no relapses in the 12 months following treatment onset.Dogs were considered poor responders if they experienced one or more relapses during the study or showed no improvement in clinical-pathological variables by the end of the study.

Statistical analysis
All statistical tests were performed using the software packages SAS (SAS Institute Inc., Cary, NC, United States) version 9.4, and SPSS Statistics version 28 (SPSS Inc., Chicago, IL, United States).Graphs were created using GraphPad Prism 10.Significance was set at p ≤ 0.05.Some variables were categorized for analysis: age (2-4 years: young adult, 5-8 years: mature adult, ≥9 years: senior) (17), IFAT titer according to cut-off values established by the laboratory (<200 negative, 200-400 low positive, 800-1,600 medium positive, >1,600 As CIC values within groups were not normally distributed (Kolmogorov-Smirnov test, p < 0.001), non-parametric tests were used.The Wilcoxon test was used to compare categorical variables between two established groups and the Kruskal-Wallis test with Bonferroni correction when there were more than two groups.The association between CIC levels and the clinical signs from the scoring system used in this study was evaluated using the Chi-Square test, specifically analyzing the 2 × 2 contingency table formed by the variables.From this analysis, relative risks (RR) for each clinical sign were calculated, accompanied by their respective 95% confidence intervals (CI).These measures serve to quantify the strength and direction of the association between CIC and the clinical signs, providing valuable insights into their relationship.
To assess changes in CIC levels over time, we used the Friedman test for repeated measures also with Bonferroni correction.Spearman's correlation test was used to examine possible correlations between CIC levels and continuous variables.We also assessed the capacity of CIC level to accurately identify relapse by calculating the area under the receiver operating characteristic (ROC) curve (AUC).

Study population
Of the 44 dogs enrolled, 23 were male and 21 female.Fourteen dogs were classified as young adults (2-4 years), 18 as mature adults (5-8 years), and 11 as senior (mean age 6.6 years; 95% CI: 1.6-14.64).During the study, three dogs died of a comorbidity (2 lymphoma, 1 carcinoma), and seven were lost to follow-up as owners failed to return for the re-visits.Even though only 34 dogs completed the 12 months of the study, data for all 44 dogs enrolled were included in the analysis until the loss of follow-up.
The CIC levels were significantly higher in young adult dogs compared to senior (p < 0.001), and low negative correlation was found between CIC level and age (r = −0.273).No differences in CIC levels were detected between mongrels and pure breeds (p = 0.107) or between male and female dogs (p = 0.416).

CIC and clinical scores, and LeishVet stages
The mean clinical score awarded to the 34 dogs that completed the 12 months of follow up gradually decreased throughout the study (Figure 3), and clinical scores and CIC levels showed positive correlation (Figure 3) (r = 0.5135, p < 0.0001).Dogs showing clinical signs were also found to have a greater risk of having elevated CIC levels.Table 2 displays the relative risks (RR) for various clinical signs evaluated in association with the presence of CIC in dogs.Interestingly, 10/44 dogs (22.72%) had a clinical score of five or less, but high CIC levels (≥2 OD).
At enrolment, 15 dogs were classified as LeishVet stage I, 20 as stage II, 7 as stage III and 2 as stage IV.Owing to the small number of animals at stage IV, data for stages III and IV were combined for analysis.Dogs at stage I showed significantly lower CIC levels than those classified as having CanL stages II and III (p < 0.0001) (Figure 4).No differences were found between dogs staged as II or III (p = 0.485).Moreover, stage I dogs consistently maintained low CIC levels throughout the study (Figure 4).In stage II patients, CIC levels decreased after treatment and remained stable, whereas in stage III patients, they increased after 90 and 360 days of treatment (Figure 4).

CIC and laboratory findings
Leukopenia was the most frequent abnormality detected in the CBC, occurring in 27.78% of dogs, followed by thrombocytopenia (22.62%) and non-regenerative anemia (18.23%).However, significant differences in CIC levels were only found for anemic dogs (p < 0.0001) (Figure 5), which not only had higher CIC levels but also experienced a gradual increase in these levels as early as 6 months after treatment Repeated measures were treated as independent measures for relative risk calculation.CI, confidence interval.
Among the biochemical variables measured, creatinine was found elevated in only 8.57% of dogs, whereas alanine aminotransferase (ALT) was increased in 11.49%.No significant differences were found in CIC levels between dogs with normal or elevated creatinine (p = 0.789, r = −0.327)or ALT (p = 0.434, r = −0.216)levels.
Hyperproteinemia was detected in 37.22% of visits.Dogs displaying hyperproteinemia had significantly higher CIC levels (p < 0.0001) (Figure 6).At visit 3, which was the first follow up visit after the end of treatment, both groups showed reduced CIC levels.However, in dogs with dysproteinemia, these levels subsequently increased, while dogs without dysproteinemia consistently maintained lower CIC levels (Figure 6).Additionally, we found moderate positive correlation between CIC and total protein levels (r = 0.533, p < 0.0001).
Hypoalbuminemia was present in 22.02% of patients.Dogs with hypoalbuminemia displayed significantly more elevated CIC levels (p = 0.02) (Figure 7), and low negative correlation was found between CIC and albumin levels (r = −0.306).
Among the globulins, gamma globulin was the most affected with hypergammaglobulinemia present in 31.1% of animals, followed by hyperbetaglobulinemia (9.82%).Dogs with these abnormalities showed significantly higher CIC levels (p < 0.0001) (Figure 8), and we found positive correlation between CIC and beta (r = 0.444) and 10.3389/fvets.2024.1368929 Frontiers in Veterinary Science 08 frontiersin.orggamma (r = 0.720) globulin levels.Elevated alpha-1 and alpha-2 globulins were noted in 2.48 and 8.18% of dogs, respectively, but no differences in CIC levels were detected between dogs with normal or high blood alpha-1 globulins (p = 0.960).Dogs with elevated alpha-2 globulins had higher CIC levels (p = 0.022).No correlation was found between CIC levels and alpha-1 and alpha-2 globulins (r = 0.021; r = −0.022,respectively).Among all variables examined, dysproteinemia was the most common laboratory finding, detected in 41.66% of animals.Dogs with dysproteinemia showed significantly higher CIC levels (p < 0.0001), and high negative correlation was found between CIC level and the A/G ratio (r = −0.618).When we assessed the effect of the level of dysproteinemia on CIC level (Figure 9), we found no differences between dogs with mild (p = 1.0), moderate (p = 0.419) or severe (p = 0.121) dysproteinemia.Throughout the study, CIC levels remained below 1.5 OD in dogs without dysproteinemia, while levels were consistently over 2.5 OD in those with severe dysproteinemia (Figure 9).
A large proportion (36.18%) of dogs showed a normal UPC ratio.Borderline proteinuria and proteinuria were recorded in 26.32 and 36.18% of the patients, respectively.Dogs with borderline proteinuria (UPC = 0.2-0.5)had higher CIC levels than dogs without proteinuria (UPC > 0.5) (p = 0.035, r = 0.172).However, no differences in CIC were found between dogs with borderline proteinuria or proteinuria (p = 0.69).

CIC and IFAT titres
The IFAT titres ranged from negative to 1:6400, with the cut-off set at >1/100.Significant differences in CIC levels were observed among groups showing negative, low, medium, or high antibody titres (p < 0.0001) (Figure 10), and there was strong positive correlation between CIC levels and IFAT titres (r = 0.735; p < 0.0001).Further, throughout the study, CIC levels remained consistently low in clinically healthy infected dogs returning negative IFAT results (Figure 10).
The CIC levels were significantly higher in dogs treated with meglumine antimoniate (p = 0.0036) or allopurinol (p = 0.015) and in those with proteinuria treated with benazepril (p = 0.0039).Dogs treated with meglumine antimoniate or allopurinol or that received benazepril to control proteinuria were almost two times more likely to have high CIC levels (Table 4).In contrast, no differences were detected according to treatment with nucleotides (p = 0.0519) or prednisone (p = 0.2312).

CIC and treatment response
To assess the impacts of treatment in sick dogs, we also examined CIC levels over time.As the data were not normally distributed, we employed the non-parametric Friedman test.Nine of the 44 dogs, each missing one visit, were excluded from this part of the analysis.Our results show that mean CIC levels gradually decreased after treatment and remained relatively stable throughout the study period (Figure 11).Nevertheless, significant differences in CIC levels were only detected between the first (D0) and last visit (D360) (Friedman test p = 0.022) (Figure 11).
Of the 44 dogs initially included in our study, 28 were classified as good responders, 11 as poor responders and 5 could not be classified,

CIC and relapses
Relapses were detected in 23.18% of the animals.At inclusion, 18 of the 44 dogs had had a previous diagnosis of CanL and were referred for treatment due to clinical relapse.Over the 12-month follow-up period, six dogs had one relapse and two had two relapses (Table 5).Dogs experiencing relapse had significantly higher CIC levels compared to dogs not experiencing relapse (p < 0.0001) (Figure 13); these were elevated consistently in all visits (OD > 2) (Figure 13).

Discussion
To our knowledge, this is the first study to clinically validate the use of CICs as biomarkers of the progression of canine leishmaniosis.Dogs in advanced stages of this disease show higher CIC levels (11) and a higher risk of multiple organ damage, which can significantly impair quality of life and dramatically reduce survival (6)(7)(8)29).Currently, diagnostic methods for quantifying CIC are lacking.Parody et al. ( 11) introduced a non-invasive Leishmania-specific method of measuring CIC levels in serum samples from dogs with leishmaniosis.This method was recently laboratory validated following the recommendations of the NIH Biomarkers group (12).The method's specificity and robustness suggest this new biomarker could be useful not only for a diagnosis of CanL but also for tracking disease progression and potentially assessing treatment efficacy (13).In the present real-world study, we sought to clinically validate the performance of this new biomarker.
In our study, 34 dogs naturally infected with L. infantum at different LeishVet stages were followed for 12 months.Our data show that CIC levels are significantly higher in young adult dogs than senior dogs (p < 0.001) while no differences were detected according to sex or breed.The influence of sex, breed, and age on susceptibility to acquiring the infection has been explored in numerous studies with contrasting results (30)(31)(32)(33)(34). Nonetheless, the authors of two epidemiologic studies involving a large number of dogs, which took into account the presence of clinical signs of CanL, described that the age distribution for affected animals was bimodal, with a first peak corresponding to animals of around 3 years of age and a second, less evident, peak representing animals around 8 years old (30,31).Age also seems to be an important risk factor associated with visceral leishmaniasis (VL) in humans according to two systematic reviews (35, 36) such that children are more susceptible to both infection and illness.Our results point to younger dogs controlling the disease less effectively and thus being more likely to have formed CIC.
Mean clinical scores progressively declined during the study and significant differences were detected in total clinical scores between visits.Moreover, we found positive correlation between clinical score and CIC levels (r = 0.5135, p < 0.0001).These results suggest that a decrease in clinical score will be accompanied by a drop in CIC levels.Nevertheless, some dogs had a low clinical score (≤5) despite having high levels of CIC (≥2 OD).The only clinical signs observed in these dogs were lameness, uveitis, skin ulcers, or polyuria/polydipsia.Hence, measuring CIC levels could be useful in dogs that do not have   ROC curve analysis of relapse (AUC = 0.8085, p < 0.0001).
electrophoresis has proven extremely useful for the diagnosis and monitoring of CanL and is routinely performed in clinical practice (41)(42)(43)(44).Indeed, several studies have shown that this technique may show abnormalities very early during the course of disease, even before the onset of overt clinical signs (21,41,42,45,46).In our study, we observed strong correlation between CIC and hematocrit (r = 0.524), total proteins (r = 0.507), beta-(r = 0.426) and gamma globulins (r = 0.673), and the A/G ratio (r = −0.647).Further, our results also show that dogs with non-regenerative anemia and/or dysproteinemia have significantly higher CIC levels.When we examined the kinetics of these immune complexes, we noted that dogs with anemia and/or dysproteinemia showed higher CIC levels throughout the study despite specific treatment.Taken together, our findings point to the high reliability of this new biomarker, as it correlates with other tools used for the diagnosis and clinical management of CanL.Moreover, they confirm that dogs with anemia or dysproteinemia are at a higher risk of CIC deposition in specific organs.
The IFAT is one of the most used quantitative serologic techniques for the detection of anti-Leishmania antibodies and is considered the gold standard test for Leishmania infection.Extensive research has identified correlation between IFAT results and clinical signs (47)(48)(49)(50)(51), becoming more evident in severe clinical forms of CanL (51).In human VL, high anti-Leishmania antibodies also correlate with disease progression (4).Studies have also shown that IgG antibodies not only fail to protect against this intracellular parasite but also contribute to disease progression by reversing the inflammatory cytokine profile of immune cells and inducing the production of high IL-10 levels via the receptor FcγRIII (4, 52).Here, significant differences in CIC levels were found among dogs showing negative, low, medium or high antibody titres, and high positive correlation was observed between CIC and IFAT results (r = 0.735).These findings are in line with those described by Parody et al. ( 11), who also noted high correlation between these two variables (r = 0.754).Thus, dogs with high antibody titres have significantly higher CIC levels and are thus at greater risk of their deposition in target organs.Mean CIC levels dropped after treatment and remained under 1.5 OD throughout the study, although significant differences were only found between the first and the last visits.Likely reasons for this are: (1) the small number of animals included in the study, and thus reduced statistical power when using the Friedman test with Bonferroni correction; and (2) in dogs showing a poor response to treatment, CIC levels can remain elevated for several months.In this study, sick dogs were treated with antimonials in combination with allopurinol, which is the standard treatment for CanL in Europe (3).Four dogs treated with antimonials, and one treated with allopurinol also received prednisone for 1 month to control clinical signs of CIC deposition.Dogs treated with antimonials or allopurinol, or those that received benazepril to control proteinuria, had significantly higher CIC levels and were up to two times more likely to have elevated CIC levels in serum.This is not surprising as, except for allopurinol, these medications are used in more advanced stages of the disease.Due to the small number of animals that received treatment with prednisone, no conclusions can be made regarding the kinetics of CIC levels in these dogs.Further studies are needed to examine the optimal use of this new biomarker in tailoring and monitoring corticosteroid use in dogs experiencing CIC deposition.
Good responders had significantly lower CIC levels compared to poor responders (p < 0.0001), which progressively decreased over the study course.In poor responders, although a CIC reduction was observed after treatment, levels progressively increased as early as 90 days post-treatment.These results suggest that measuring CIC levels before and after treatment could serve to assess treatment effectiveness.Dogs experiencing relapse after treatment had significantly higher CIC levels (p < 0.0001); over 2 OD in all visits.During the study, six dogs had one relapse and two had two relapses.Most relapses occurred in visit 6, 1 year after treatment.Nevertheless, three dogs had a relapse as early as at 6 months post-treatment.Studies have shown that the presence of immune complexes contributes to the establishment of chronic infections in murine models and human VL (4, 52).One key mechanism is their ability to stimulate macrophages and other immune cells to produce IL10.This cytokine reduces the expression of inducible nitric oxide synthase (iNOS) and the intracellular production of nitric oxide, which are both vital for eliminating intracellular Leishmania parasites.Moreover, IL-10 downregulates the Th1-associated IFN-γ response, which is essential to activate nitric oxide production within infected cells and eliminate the parasite.Consequently, dogs that relapsed had higher CIC levels, which may have been responsible for their inability to control the infection despite receiving a specific leishmanicidal treatment.
Importantly, our ROC curve analysis revealed that CIC were a good discriminator of relapse (AUC = 0.808) and that the optimal cut-off for accurately identifying relapse was 1.539 OD.This is relevant as the cut-off for positivity adopted in the study by Parody et al. (11) was 0.274 OD, which may be too low to be associated with clinical disease.In fact, CIC levels ranged between 0.8 and 1.0 OD in clinically healthy infected dogs, while sick dogs had levels between 1.9 and 3.1 OD.Further larger studies are needed to confirm our findings and establish definitive cut-offs to distinguish between healthy infected and sick dogs.
It is important to mention that our study did not include a formal sample size calculation due to the lack of prior data concerning expected effect sizes for the novel biomarker under investigation.However, our study yielded statistically significant results, providing valuable insights into the association between CIC clinical signs and clinicopathological parameters related to CanL.Moving forward, sample size calculations in future studies could enhance the robustness of our findings and contribute to a deeper understanding of CIC's role in the progression of the disease.

Conclusion
Our findings indicate elevated serum CIC concentrations in both young adult dogs and/or those in advanced stages of CanL with anemia and/or dysproteinemia.Dogs showing a favorable response to treatment showed declining CIC levels.Conversely, dogs displaying a poor treatment response and recurrent clinical relapses consistently exhibited high CIC levels.In these dogs, vigilant monitoring is essential to enable prompt and targeted treatment and prevent irreversible organ damage due to CIC deposition.
While studies in larger populations of dogs are still needed to confirm our findings, the use of CICs as a complementary biomarker to track disease progression in CanL is promising.Further studies using this new biomarker may provide insight into specific organ tropism and relapse mechanisms in non-responders.Such studies will also improve our understanding of long-lasting cell-mediated immunity in resistant dogs, and aid in the development of Leishmania vaccines.

FIGURE 2
FIGURE 2 Clinical signs of CIC deposition in dogs with CanL.(A) Ulcer in the pinna associated with vasculitis.(B) Blindness associated with bilateral uveitis.

FIGURE 7 CIC
FIGURE 7    CIC levels recorded in dogs according to albumin level (nonparametric Wilcoxon test).Normal reference value albumin: 2.4-3.9 g/dL.

TABLE 1
Clinical scoring system used in this study.

TABLE 2
Relative risk of clinical signs associated with a significant risk of elevated circulating immune complexes (CIC).

TABLE 3
Treatment regimens of dogs included in this study.

TABLE 4
Relative risk of treatment regimens associated with a significant risk of elevated circulating immune complexes (CIC).

TABLE 5
Number of relapses detected during the study and their time points.