A Proposed Cytological Grading Scheme for Canine Cutaneous Mast Cell Tumor Aspirates that Correlates with High and Low Tier Histological Grades
M. Camus1; E. Driskell2; M. Ilha3; J. Koehler4; H. Priest5; P. Rakich6; P. Krimer6
Introduction
A new histologic grading system using two tiers (high and low) was proposed to replace the 3-tier Patnaik grades for canine mast cell tumors (MCTs).1 This new system is based on cytomorphologic characteristics of neoplastic cells, including karyomegaly, multinucleation, bizarre nuclei, and > 7 mitotic figures/10 HPF. All features can be examined cytologically, unlike the Patnaik grading scheme, and fine needle aspirates from MCTs may provide the same information in a quicker, cheaper, and less invasive manner.
Objectives
To determine if the new 2-tier grading scheme can be applied to fine-needle aspirates from MCTs.
Methods
Paired cytologic and histologic samples from 152 canine MCTs were blindly graded high or low by 3 board-certified anatomic pathologists. Three board-certified clinical pathologists evaluated cytology specimens using the same criteria and reported an intuitive grade. A kappa statistic was used to measure agreement between average histological grade and various proposed cytological grading schemes, as well as intuitive impression.
Results
The cytological grading scheme that agreed best with histology (kappa = 0.725 ± 0.085) classified a cytological sample as high grade if it was 'poorly granulated' or it had two grading characteristics (mitotic figures, multinucleated cells, binucleated cells, bizarre nuclei, or > 50% anisokaryosis), with 88.24% sensitivity, 94.07% specificity, positive likelihood ratio of 14.89 and negative likelihood ratio of 0.13. Interestingly, intuitive impression was extremely similar: kappa = 0.700 ± 0.089, sensitivity 82.35%, specificity 94.81%, positive likelihood ratio 15.88 and negative likelihood ratio 0.19.
Conclusions
Cytology is a useful indicator of histological grade of canine mast cell tumors.