Predictive Assessment (Odds Ratio) of the Therapeutic Response in Canine Leishmaniasis
*Dpto. Patología Animal II - Facultad de Veterinaria de Madrid, Avda. Puerta de Hierro s/n
Despite the extensive bibliography on canine leishmaniasis, few studies deal with the prognostic value that certain biochemical or hematological variables may have with regard to the therapeutic response of dogs with this disease. The aim of this study is to assess the influence these variables (particularly antibody titer, serum protein electrophoretogram, CBCc, urea and ALT), obtained at diagnosis, might have on the probability of three events related to the future of dogs to be treated: appearance of side-effects, attainment of clinical remission and administration of one or more than one treatment courses until remission. The multiple regression analysis is capable to select and quantify the influence of those variables that are predicted values for the future of the treated patients.
Sixty-one dogs of both sexes, different ages and breeds, diagnosed with leishmaniasis by IFAT, were included in this study. Dogs were treated with meglumine antimoniate SQ (150 mg/kg/day throughout two periods of 10 days, with a 10-day interlude between treatment periods) associated with allopurinol PO (10 mg/kg/12h, 45 days). Treatment courses were repeated until achieving clinical remission (disappearance of clinical signs, normalization of the serum protein electrophoretogram and drop in the antibody titre). The BMDP programme was used to perform the multiple regression analyses. The relative probability or odds ratio (OR) of a particular event (yes versus not), is represented by the equation: OR=ea0+a1x1+a2x2+...+apxp.
Fifteen of the 61 dogs developed side effects after starting therapy. Forty-eight of the 61 animals achieved remission. Only one treatment course was necessary in 34 dogs, two courses were needed in 13 and three in one animal. Among all the variables considered, the BMDPPR programme selected urea and ALT levels (p<0.05), followed by PMNC count (p=0.093) and antibody titre (p=0.053) as risk factors in the probability of achieving the remission: O.R=Pyes/Pno=e10.72-0.0853xurea-0.0302xALT-0.0004xPMNC-0.0005xab titre, and only alfa-globulin concentration as a risk factor in the probability of side-effects presentation (p=0.028): O.R=Pno/Pyes=e3.413-2.653xalfa-glo. In relation to the probability of administration of one or more treatment courses until achieving remission, the statistical program selected albumin/globulin ratio (p<0.01), followed by hemoglobin (p=0.084): O.R=Pyes/Pno=e-4.967+4.977xAlb/Glo+0.2972xHb.
Urea, ALT, PMNC count and antibody titre exhibit an inverse association with the probability of achieving remission. On the contrary, the concentration of alfa-globulins displays a direct association with the probability of side-effect presentation. Albumin/globulin ratio and hemoglobin display, as well, a direct association with the probability of administration of one treatment course to achieve remission. In the clinical practice the odds ratio is a useful tool for (statistically) predicting the evolution of the cases. Although the odds ratio mean values obtained in this study may be sufficiently representative, the confidence intervals are rather ample. Further studies are necessary in order to reduce these confidence intervals and also to assess the influence of different biochemical and hematological variables when administering other therapeutic protocols.