Introduction

There is a paucity of published information regarding the diagnostic utility of cytology specimens. The purpose of this study was to characterise the proportion of cytology specimens submitted to a commercial veterinary laboratory that yielded diagnostically useful information. This is not a study into cytological accuracy, which has been well documented elsewhere.

Materials and Methods

1206 cytological submissions from July to August 2007 were retrospectively collated. Samples that had additional tests performed (e.g., protein analysis, cell counts) that may have affected the cytologist's interpretation of the microscopic examination were excluded. The remaining 945 samples were classified into four categories:

 Category 1--Extremely useful: diagnosis achieved e.g., mast cell tumour

 Category 2--Very useful: cytological diagnosis achieved e.g., neoplasia, inflammation etc.

 Category 3--Useful: cytological diagnosis not possible, but a description can be made which yields some information

 Category 4--Unacceptable: sample rejected due to insufficient cellularity/extreme haemodilution etc.

Results

All samples were analysed by one of four cytologists. A diagnosis was reached for 552 samples (58%). A further 212 (22%) of samples yielded some useful information. 181 (19%) of samples were rejected (of which 75% due to low cellularity, 12% to non-representative sample, 8% to sample degeneration and 5% to haemodilution).

Of the samples in Category 1, the diagnoses achieved were: lipoma--105 (48%), mast cell tumour--49 (22%), epidermal inclusion/follicular cyst--17 (8%), lymphosarcoma--16 (7%), histiocytoma--8 (4%), sebaceous hyperplasia/adenoma--6 (3%), perianal adenoma--5 (2%) and anal sac adenocarcinoma--4 (2%). Other diagnoses included haemangiopericytoma, hepatocellular carcinoma, leishmaniasis, melanoma, multiple myeloma, plasmacytoma, thyroid adenoma and transitional cell carcinoma.

Conclusion

The vast majority of samples were diagnostically useful. The most common cause for rejection was low cellularity. This suggests than in-house evaluation to assess cellularity prior to submission would increase the diagnostic utility of cytology.

Figure 1. Data sorted according to origin of sample ('Miscellaneous' includes sites yielding fewer than 20 samples).