Prognostic factors in surgical excised single cutaneous mast cell tumor in 110 dogs: a monoistitutional study

Cutaneous mast cell tumor (cMCT) is a common skin cancer in dog. In the last 7 years several studies have updated prognostic role of some clinical and pathological variables. Based on these updates we want to analyze the impact of clinical and pathological variables in a monoinstitutional cases series of single excised cMCT in dogs. Single cMCT underwent wide excision were retrospectively included in this study if these variable were all available: signalment, weight, tumor dimension and ulceration, first presentation or local recurrence, WHO clinical stage and substage, Patnaik and Kiupel histological grading, status of surgical margins, and type of adjuvant therapies (if administered). The endpoint was relapse of tumor (local and/or distant). The prognostic effect of each variable was evaluated by Cox model. For covariates measured on a continuous scale the potential non‐linear relationship between covariate values and logarithm of hazard (or sub‐distribution hazard) was investigated considering regression splines. The Wald test was used to evaluate the statistical significance. Model results were reported as estimated hazard ratios (HR) and 95% confidence intervals (95% CI). Progression free survival curves for each variable were traced by the Kaplan–Meier method. A total of 110 dogs were included. Most commonly breeds were: mix breeds (27.7%), boxer (15.45%), retriever (13.64). Male and female were equally represented. At diagnosis median age was 8.5 years (1-15 years) and median weight was 25.27 kg (2-47 kg). Tumors were located on trunk in 66 cases, distal limb in 18 cases, head and neck in 17 cases and on perineal region in 9 cases. Tumor median dimension was 2 cm (0.3-17.5 cm); the 91% was at first presentation and 9% a local recurrence; 28% were ulcerated; clinical stage I, II and III were 68.2%, 5.4% and 26.4%, respectively. Regarding histopatologic results: Patnaik grade I, II and III were respectively 14.55%, 79.09% and 6.37% of cases; Kiupel low grade was 84.55% and high grade was 15.45% of cases; surgical margins was clean, infiltrated and clean but close in 69.09, 29.09% and 1.82% of cases, respectively. Relapse was observed in 16 cases (local relapse in 14 and distant metastasis in 2) with a median time to relapse of 114 days (13-815 days). Overall median progression free survival was not reached (13-2500 days). The variable statistically associated with relapse were: weight (14-25 kg; P=0,04), tumor dimension (>2.5 cm; p<0.01) recurrence (p<0.001), ulceration (p<0.001), clinical stage (p<0.01), Patnaik grade III (p<0.001), Kiupel high grade (p<0.001), margin at risk (p<0.001). The results confirmed the prognostic role of several clinical and pathological variables previously documented and in particular clinical stage, Kiupel grading system and status of surgical margins.