The use of flow cytometry for immunophenotyping lymphoproliferative disorders in cats: a retrospective study of 19 cases

Authors


Abstract

Flow cytometric immunophenotyping is a useful step in the diagnosis of lymphoproliferative malignancies in human and veterinary medicine. The purpose of this study was to assess the usefulness of this technique for the diagnosis of lymphoproliferative disorders in cats. Nineteen cats were retrospectively enrolled in this study and allocated into two groups. Group 1 consisted of 13 cats with lymphoma, whereas group 2 consisted of 6 cats with non-neoplastic lymphoproliferative disorders. Fine-needle aspiration biopsies were analysed by flow cytometry in order to evaluate the immunophenotype. Flow cytometric analysis identified a neoplastic lymphoid population in 12 of the 13 cats of group 1, confirming the diagnosis of lymphoma and further characterizing it. The six cats in group 2 showed a mixed lymphoid population, which was not suggestive of a neoplastic disorder. Flow cytometry is a valuable and powerful tool for refining the diagnosis of feline lymphoproliferative disorders.

Introduction

Despite being the most common haematopoietic tumour in cats,[1] feline lymphoma remains a diagnostic challenge. It is well documented that flow cytometric immunophenotyping combined with fine-needle aspiration cytology (FNAC) enhances diagnostic accuracy in human and canine lymphoma and enables its subclassification[2-4]; therefore, invasive and costly diagnostic procedures (i.e. surgical biopsy) can sometimes be avoided. Flow cytometry is a method commonly used in human medicine for the diagnosis of non-Hodgkin lymphoma. In the last decade, flow cytometric immunophenotyping has also become a routine diagnostic tool in veterinary medicine, especially in the canine species.[4] In other species, the use of this technique is limited by the restricted panel of antibodies available and by the unclear prognostic role of the immunophenotype in lymphoproliferative disorders.[5, 6]

Cats are prone to a wide range of lymphoma types. Louwerens et al. showed that reduction in the prevalence of feline leukaemia virus (FeLV) over the last decades triggered a change in the presentation and epidemiology of feline lymphoma. Despite the decline in FeLV-associated lymphoma (mostly mediastinal form), the number of new lymphoma cases has been steadily increasing over the last few years mostly because of an increased incidence of non-retroviral-associated lymphomas (mainly alimentary and extranodal forms).[7]

In 2002, the World Health Organization (WHO) adapted a new classification system of feline and canine lymphoma from human medicine. The new scheme, apart from morphological criteria, clinical features, anatomical location, molecular and genetic abnormalities, takes into consideration the immunophenotype of the tumour.[8] Although further analyses are needed, the prognostic value of the new system is considered to be high.

Morphology of neoplastic cells is not always predictive of their immunophenotypic pattern, and therefore flow cytometric immunophenotyping should be used to determine it.[9] Flow cytometry may also help in distinguishing between reactive and neoplastic conditions by the identification of a single antigenically homogeneous population or an abnormal lymphoid population, thereby facilitating the diagnosis of the disease.[4]

In most of the previous reports, the immunophenotype of lymphoma in cats was determined by immunohistochemical and immunocytochemical techniques. Reports on the application of flow cytometric immunophenotyping in cats are scant. Roccabianca et al. used flow cytometry to characterize 21 cases of large granular lymphocyte lymphoma.[10] The technique has also been applied in other studies to reach a definitive diagnosis of chronic lymphocytic leukaemia (CLL) and assess its phenotypic pattern,[11] to aid the differentiation between thymoma and lymphoma,[12] to confirm the diagnosis of chronic eosinophilic leukaemia,[13] to establish and characterize feline lymphoma cell lines[14] or to evaluate the effects of dexamethasone treatment on the immune system of cats.[15]

At present, there are no published reports on the diagnostic use of flow cytometry for immunophenotyping feline lymphoma.

The objective of the study was to evaluate flow cytometric immunophenotyping as a method for refining the diagnosis of lymphoproliferative disorders in cats.

Materials and methods

Samples from 22 feline patients were submitted to the Central Diagnostic Service, Department of Veterinary Medicine, University of Cambridge (Cambridge, UK) between January 2012 and December 2013 for flow cytometric immunophenotyping of FNA samples.

Three patients were excluded from the study because of low cellularity of the aspirates, which prevented reliable cytometric analysis. One had lymphoma; the other two had been diagnosed with non-neoplastic lymphoproliferative disorders.

The laboratory results and the clinical history of the 19 cats were retrospectively reviewed. According to these, the feline patients were divided into 2 groups: group 1 (n = 13) cats with previously diagnosed lymphoma and group 2 (n = 6) cats with non-neoplastic lymphoproliferative disorders. The final diagnosis was made on the basis of final clinical diagnosis that included all the clinical information. FNAC was the primary diagnostic test used for diagnosis in majority of the cases in both groups (n = 15); other tests included histology (n = 2), cavity fluid analysis (n = 2), immunohistochemistry (IHC) (n = 1) and polymerase chain reaction (PCR) for antigen receptor rearrangement (PARR) (n = 1).

FNA samples were obtained using conventional techniques and placed in 1-mL tube containing Streck Cell Preservative (Streck, Omaha, NE, USA) and shipped by post in the dark at room temperature, as suggested by the manufacturer. All samples were analysed within 24 h from the time of collection.

Flow cytometric measurements were performed on the submitted samples using monoclonal antibodies for the following extracellular antigens: CD4 (T-helper cell marker, clone vpg39, AbD Serotec, Kidlington, UK), CD5 (pan-T lymphoid marker, clone FE1.1B11, AbD Serotec), CD8α/β (T-cytotoxic and NK cell marker, clone vpg9, AbD Serotec) and CD21-RPE (B-lymphoid marker, clone CA2.1D6, AbD Serotec).[16] The staining was performed according to the protocol that is routinely used by the laboratory to perform immunophenotyping of canine lymphomas and leukaemias.

A total of 6 × 10[6] cells were required to perform the analysis using all the above-mentioned antibodies. Briefly, aliquots of samples were placed in tubes and primary antibodies were added. Samples were then incubated for 20 min in the dark at 4 °C. After washing, samples labelled with non-conjugated primary antibodies were incubated for 20 min with corresponding secondary polyclonal antibodies conjugated with fluorescein isothiocyanate (FITC): rabbit anti-mouse (AbD Serotec) or rabbit anti-rat (AbD Serotec). BD FACS Lysing Solution (BD Bioscience, San Jose, CA, USA) was used to lyse erythrocytes in whole blood samples. Finally, samples were washed prior to cytometric analysis.

Stained cells were analysed with a FACSCalibur flow cytometer and CellQuest software (BD Biosciences). Calibration of the machine was performed using CaliBRITE beads and AutoCOMP software (BD Biosciences), and 10 000 cells were acquired for each analysis. Corresponding isotype controls were used to set the gates.

Lymphoid cells were gated on the basis of their light scattering properties. The expression of surface antigens was analysed in the gated population. Identification of a predominant antigenically homogeneous lymphoid population (>90% of the lymphoid cells) was considered diagnostic for lymphoma. A sample was considered negative for a certain antigen when the antigen was expressed by less than 10% of lymphocytes in the gated population of cells. Simultaneously, the phenotypic pattern was assessed. Detecting a mixed lymphoid population (CD4+, CD5+, CD8+, CD21+ cells present in variable proportions) was suggestive of a non-neoplastic lymphoproliferative disorder.

Results of other tests that were necessary for the diagnosis and staging of lymphoma were provided by the clinicians in charge of the feline patient [i.e. haematology, clinical biochemistry, cytological and histological examination, IHC, PARR, FeLV and feline immunodeficiency virus (FIV) testing (serology), and imaging (ultrasound, magnetic resonance imaging)].

Results

Enrolled cats were of various breeds: Domestic Shorthair (DSH) (n = 15), Domestic Longhair (DLH) (n = 1), British Shorthair (n = 1), Bengal (n = 1), Persian (n = 1); sex: male neutered (n = 10), female neutered (n = 9); age: from 6 months to 16 years, median 8.5 years. Flow cytometric analysis was requested by referring veterinarians based on clinical and laboratory signs suggestive of lymphoid neoplasia: peripheral lymphadenopathy (n = 5), abdominal (n = 10), mediastinal (n = 2) or retrobulbar mass (n = 1), lymphocytosis (n = 1), thickened intestinal wall (n = 1), enlarged kidneys (n = 1) and pleural effusion (n = 3), as well as on the basis of prior cytological and histological evaluation of these masses.

The cats were divided into two groups. Group 1: cats with lymphoma (Table 1). Cats aged from 6 months to 16 years. The median age of cats in this group was 8.5 years. They were 8 males and 5 females – 10 DSH, one British Shorthair, one Bengal and one Persian. Group 2: cats with non-lymphoproliferative disease (Table 2). They were from 1 to 13 years old. The median age of cats in this group was 8 years. They were 2 males and 4 females – five DSH and one DLH.

Table 1. Characteristics (age, sex, breed, sample type, lymphoma subtype, phenotype, diagnostic tests used for the definitive diagnosis) of cats with lymphoma (group 1)a
No.AgeSexBreedSample analysedLymphoma subtypePhenotype of lymphomaDefinitive diagnosis based on
 CD4CD5CD8α/βCD21
  1. FN, female neutered; MN, male neutered; UTD, unable to determine; N/d, not determined.

  2. a

    The percentage of lymphocytes positive for a certain antigen is presented in brackets.

17MNDSHIntestinal massAlimentary lymphomaB cell+ (95%)Intestinal mass FNAC
212MNDSHIntestinal massAlimentary lymphomaB cell+ (92%)Intestinal mass FNAC
38FNDSHGastric massAlimentary lymphomaB cell+ (96%)Intestinal mass FNAC
412FNDSHIntestinal massAlimentary lymphomaB cellN/dN/d+ (95%)Intestinal mass FNAC
515MNPersianIntestinal massAlimentary lymphomaB cellN/dN/d+ (91%)Intestinal mass FNAC
69MNDSHMesenteric lymph nodeAlimentary lymphomaT cellN/d+ (93%)N/dLymph node FNAC
716FNDSHSpleenMulticentric lymphoma/leukaemiaT cell+ (94%)+ (95%)Spleen and liver FNAC
82MNDSHMesenteric lymph nodeMulticentric lymphomaT cell+ (94%)+ (93%)Lymph node FNAC
910MNBSHIntestinal massMulticentric lymphomaB cellN/dN/d+ (90%)Intestinal mass FNAC
100.5MNDSHPeripheral lymph nodeMulticentric lymphomaUTD+ (31%)+ (53%)+ (15%)+ (32%)Lymph node histology and IHC (B cell multicentric lymphoma)
111MNDSHPleural fluidMediastinal lymphomaB cell+ (90%)PARR of pleural fluid
121FNDSHPleural fluidMediastinal lymphomaT cell+ (99%)+ (93%)Mediastinal mass FNAC and pleural fluid analysis
135FNBengalRetrobulbar massExtranodal lymphoma – retrobulbarB cellN/dN/d+ (92%)Retrobulbar mass FNAC
Table 2. Characteristics (age, sex, breed, sample type, reason for performing immunophenotyping, diagnostic tests used for the definitive diagnosis, disease recognized, phenotype) of cats with non-neoplastic lymphoproliferative disorders (group 2)a
No.AgeSexBreedSample analysedReason for performing immunophenotypingLymphoproliferative disorder excluded byDisease recognizedPhenotype of lymphocytes
 CD4CD5CD8α/βCD21
  1. FN, female neutered; MN, male neutered; N/d, not determined.

  2. a

    The percentage of lymphocytes positive for a certain antigen is presented in brackets.

1413FNDSHMesenteric lymph nodeGeneralised lymphadenomegalyLymph node FNACReactive lymphoid hyperplasiaMixed+ (32%)+ (51%)+ (22%)+ (42%)
159MNDSHPeripheral lymph nodeLymphadenomegalyLymph node FNACReactive lymphoid hyperplasiaMixed+ (33%)+ (65%)+ (37%)+ (28%)
161FNDLHPeripheral lymph nodeGeneralised lymphadenomegalyLymph node FNACReactive lymphoid hyperplasiaMixed+ (21%)+ (53%)+ (26%)+ (44%)
174MNDSHMesenteric lymph nodeThickening of intestinal wall and mesenteric lymphadenomegalyIntestine and lymph node FNACReactive lymphoid hyperplasiaMixedN/d+ (29%)N/d+ (57%)
1810FNDSHSpleenRecurrence of IMHA, suspicion of multicentric lymphomaSpleen histologyIMHAMixed+ (57%)+ (85%)+ (18%)+ (10%)
197FNDSHPleural fluidChylothoraxPleural and abdominal fluid analysisIdiopathic chylothoraxMixed+ (30%)+ (55%)+ (26%)+ (24%)

Adequate tissue FNAs (n = 19) were analysed. All cats in the study were tested for FeLV and FIV and were negative.

In group 1: six feline patients presented with alimentary lymphoma, three with multicentric lymphoma, two with mediastinal lymphoma, one with extranodal lymphoma and one with multicentric lymphoma/leukaemia. Flow cytometric analyses enabled identification of cell lineage in 12 out of 13 lymphoma patients (Table 1). The immunophenotype of cat #1 (intestinal mass) is shown in Fig. 1A.

  1. Eight B-cell neoplasms (CD21+) – alimentary lymphoma (n = 5), multicentric lymphoma (n = 1), mediastinal lymphoma (n = 1), extranodal lymphoma – retrobulbar (n = 1);
  2. Four T-cell neoplasms (CD4 + CD5+ or CD5 + CD8+) – alimentary lymphoma (n = 1), multicentric lymphoma/T-cell leukaemia (n = 1) (Fig. 2), multicentric lymphoma (n = 1), mediastinal lymphoma (n = 1).
Figure 1.

Representative plots of immunophenotyping by flow cytometry. (A) FNA of an intestinal mass, alimentary B-cell lymphoma, CD4− (0%), CD5− (0%), CD8α/β− (0%), CD21+ (95%) (cat #1). (B) FNA of a peripheral lymph node, multicentric lymphoma, CD4+ (31%), CD5+ (53%), CD8α/β+ (15%), CD21+ (32%) (failure in detection of a predominant antigenically homogeneous lymphoid population) (cat #10). (C) FNA of a peripheral lymph node, reactive lymphoid hyperplasia, CD4+ (21%), CD5+ (53%), CD8α/β+ (26%), CD21+ (44%) (cat #16). *Red line (isotype control), green line (test sample).

Figure 2.

Splenic FNA from a cat with multicentric T-cell lymphoma/T-cell leukaemia (cat #7). High numbers of large lymphoid cells with abundant cytoplasm, a round nucleus, fine granular chromatin, containing prominent multiple round nucleoli. (Wright's-Giemsa stain, ×50 objective).

The results of flow cytometric immunophenotyping in one feline patient with multicentric lymphoma (cat #10) were inconclusive showing a mixed population of T and B cells, which made the determination of the immunophenotype impossible (Fig. 1B). Histological and immunohistochemical examination revealed the presence of a dense monomorphic population of large neoplastic cells (identified as B cells on the basis of IHC with CD79a antibodies) with areas of residual normal lymphoid tissue. Lymphoma was classified as diffuse large B-cell lymphoma (Table 1).

Two feline patients from group 1 – one with multicentric lymphoma/T-cell leukaemia (cat #7) and one with mediastinal B-cell lymphoma (cat #11) – were reassessed after treatment when complete remission was achieved. The analyses were performed on FNAs sampled from the same site as previously. Both samples were adequately cellular. An antigenically mixed lymphoid population was identified in the gated population of cells suggesting resolution of the disease.

Final diagnoses for all the cats in group 2 were as follows: reactive lymphoid hyperplasia (n = 4) (Fig. 1C), systemic inflammatory response secondary to immune-mediated haemolytic anaemia (IMHA) (n = 1) and idiopathic chylothorax (n = 1). In this group, immunophenotypic analyses did not suggest a neoplastic lymphoproliferative disease in any of the animals. Flow cytometry revealed the presence of a mixed population of lymphoid cells (both T and B cells). In all the cases, the diagnosis was supported by other diagnostic tests (Table 2).

Discussion

This study showed that flow cytometric immunophenotyping is a useful and efficient method for aiding the diagnosis of lymphoproliferative disorders and evaluation of its phenotypic pattern in cats.

In most cases, the diagnosis of lymphoma is more challenging in cats than that in other species because of the presence of specific benign hyperplastic lymph node syndromes unique to this species (i.e. idiopathic peripheral lymphadenopathy, plexiform vascularisation of lymph nodes and peripheral lymph node hyperplasia of young cats).[17-19] Confirming the diagnosis of neoplasia composed of well-differentiated lymphocytes (small cell lymphoma/CLL) may also be difficult.[20]

FNAC or biopsy histology alone may not be sufficient for making a definitive diagnosis of feline lymphoma.[21] The possibility of evaluating the architecture of the lymphoid tissue together with the cellular morphology makes histopathology potentially superior to cytology for the diagnosis of lymphoproliferative disorders.[3] On the other hand, histological sample collection requires sedation and/or general anaesthesia of the animal, making the procedure potentially more risky and costly. For that reason, many pet owners are reluctant to agree to these procedures. Furthermore, the above-mentioned approach is time-consuming and is not always sufficient for making the definitive diagnosis. In humans and dogs, cytology of FNA samples of a lymph node or mass combined with immunophenotyping often gives a more certain diagnosis.[2-4] Flow cytometric immunophenotyping may complement results of other tests and may also allow avoidance of more invasive procedures in cats.

The main advantages of flow cytometry include high specificity, quantitative evaluation of cells, very high sensitivity in detecting the expression of antigens, quantification of surface antigen expression and short sample processing time. The technique can be used for a wide variety of sample types (e.g. FNAs, peripheral blood and effusions) and can more thoroughly characterize the phenotypic pattern of haematopoietic tumours than IHC or immunocytochemistry because of the availability of a wider panel of antibodies, and simultaneous evaluation of multiple antigens.[2, 3]

As shown, flow cytometry can be used to assess the phenotypic pattern and simultaneously confirm the diagnosis of lymphoma in cats. In this study, flow cytometry identified a predominant antigenically homogeneous lymphocyte population in 12 out of 13 feline patients in group 1 with confirmed lymphoma. In the other cat in group 1 as well as in all cats in group 2, the result of immunophenotyping indicated a mixed cell population. The immunophenotypic analysis, repeated in two patients of group 1 during complete remission of lymphoma, confirmed the resolution of the disease following treatment.

The sensitivity of the method in this study might have been influenced by the tissue architecture. According to Morse et al., it is not uncommon in human medicine that the cells of interest are not represented in the specimen in an adequate proportion. The architecture of the tissue is lost during sample processing; therefore, the presence of small foci of neoplastic lymphocytes surrounded by a residual population of normal lymphoid cells might be missed.[22] Furthermore, the distinction between neoplastic and residual non-neoplastic lymphoid cells is impossible solely on the basis of single-colour flow cytometry, which is a limitation of this study. It can be a reason for failure in the diagnosis of lymphoma by flow cytometry in cat #10. In some feline patients, histological and immunohistological examination may provide more accurate results. Moreover, as feline lymphoma is mostly located in internal organs, the sampling procedures for histology, cytology and immunophenotyping is frequently more difficult in cat than in other species. Obtaining non-diagnostic samples (mostly because of their low cellularity) is a common problem, and three feline patients had to be excluded from the study because of low sample cellularity. Furthermore, poor sampling technique may also affect the accuracy of the method.

The lack of predominant antigenically homogenous samples in group 2 is supportive of the high specificity of this method when compared with FNA, which is in agreement with the literature.[2]

Abnormal co-expression of both B- and T-cell markers or loss of the expression of one or more markers is a common finding in dogs and humans.[23, 24] Similar finding has not yet been fully described in cats, but it is also expected to be present in this species. For lymphoma cats (group 1), all the antigens were tested for 8 out of 13 cats; there was no evidence of aberrant expression of antigens by neoplastic cells although it might have been present but not detected because of the limited number of available antibodies.

In the diagnostic algorithm of human, non-Hodgkin's lymphoma cytology is initially recommended for the assessment of the cellular morphology, followed by complementary techniques such as immunophenotyping.[25] As new criteria had been incorporated in the classification of lymphoma including immunophenotypic, genotypic and clinical features, the histological architectural pattern (historically the gold standard for lymphoma diagnosis and classification) is not indispensable for the diagnosis of lymphoid neoplasm. In human medicine, FNAC and immunophenotyping are sufficient for accurate diagnosis and classification of lymphoma in most of the cases. Both cytology and flow cytometric immunophenotyping require a short preparation time, small sample size, moderate cellularity and can be applied to cells in suspension, such as body fluids. It has also been shown that flow cytometry improves the sensitivity and diagnostic accuracy of cytology in humans.[2-4]

Flow cytometry may also become an important adjunct to cytology in cats. Immunophenotyping can often resolve an equivocal cytological diagnosis. By identification of a predominant antigenically homogenous lymphoid population, immunophenotyping allows the distinction between lymphoma and lymph node hyperplasia, which is particularly common in cats, or between small cell lymphoma and non-neoplastic conditions in which the morphological features make the distinction almost impossible on cytology alone.

Although molecular analyses were not performed in most of the feline patients enrolled, it should be noted that they are recommended when the results of previously performed tests are inconclusive. Clonal lymphocyte populations can be identified with high sensitivity by PARR.[26] Molecular analysis of clonality is considered to be a useful adjunct to other diagnostic tools. According to Kiupel et al., histopathological and cytological examination and immunophenotyping should precede PARR, and the results of immunophenotyping should be considered superior to the results of the PCR assay when assessing lymphocyte cell lineage.[27] The diagnostic superiority of immunophenotyping by flow cytometry over PARR and its high correlation with IHC (94%) in the phenotyping of lymphoma in dogs was recently reported.[28]

Confirming the diagnosis of alimentary lymphoma may be particularly difficult by any of the currently available methods. It has been suggested that inflammatory bowel disease (IBD) can progress to lymphoma.[7] Therefore, the distinction between severe lymphocytic-plasmacytic enteritis and early alimentary lymphoma is often impossible when relying solely on the traditional techniques.[29] The results of the histopathological examination of endoscopically obtained intestinal samples are frequently misleading and agreement on diagnosis between pathologists can be poor.[20] A diagnostic algorithm for distinguishing lymphoma from IBD has been proposed. Waly et al. suggested that immunolabelling can aid the diagnosis of feline alimentary lymphoma especially when the results of other tests (i.e. histopathology and cytology) are equivocal. A stepwise strategy was recommended – histopathological examination is followed by immunophenotyping. If the diagnosis still remains uncertain, a PCR detection of clonality should be performed. This approach may enhance the diagnostic accuracy in alimentary lymphoma patients.[30] Flow cytometry is also a very useful tool that facilitates distinguishing mediastinal lymphoma from thymoma.[12]

Although further studies comparing different diagnostic strategies in different lymphoma subtypes are warranted, taking into considerations the results of this study and the literature data, we recommend performing flow cytometric analysis in conjunction with cytological examination prior to other time-consuming, invasive and costly tests such as histopathology.

It should also be noted that its routine application for the identification of cats in remission is not recommended as the residual population of neoplastic cells cannot be currently identified by single-colour flow cytometry.

In the literature, data on the predominant phenotype of some of the subtypes of feline lymphoma are still ambiguous. It has been suggested that the prevalence of the phenotype in different lymphoma subtypes may change depending on the geographical location and may be linked with genetic factors and FeLV strain differences.[7, 21, 31, 32] In this study, the number of cats with different types of the disease was too few for the determination of predominant cell lineage.

Limitations of this study are the few number of feline patients analysed and the restricted availability of antibodies reacting with feline lymphoid antigens. In future studies, simultaneous detection of multiple antigens should also be performed as it would allow a more thorough evaluation of the samples and could facilitate the differentiation between neoplastic and non-neoplastic residual lymphoid cells.

The gold standard used for making the final diagnosis in this study is the final clinical diagnosis that incorporates all the clinical information and is likely to have a very high diagnostic accuracy. The limitation of this study is that the diagnostic tests used for the evaluation of feline patients (mostly FNAC) could have led to underdiagnosis of lymphoma in group 2. Cytological examination is sometimes not sufficient for reaching the final diagnosis in patients with lymphoproliferative diseases. Careful consideration of each patient should also be undertaken to minimize false negative results. Considering that all cytological examinations in this study were performed by an experienced clinical pathologist and clinical history and outcome were thoroughly studied, the risk of underdiagnosing lymphoma in group 2 is low.

Furthermore, the retrospective nature of the study and lack of consistency in the diagnostic tests used are also considered as limitations of this study.

In human flow cytometry, a cut-off of 20% in the gated region is considered to define positivity for a specific marker.[3, 4] Although such guidelines have not yet been developed in veterinary medicine in a study on canine lymphoma, using a normal lymph node as a standard, a value of 60% (of all the harvested cells) was proposed and used to characterize the phenotypic patterns of canine lymphoma (either of B- or T-cell lineage).[33] As it is unclear how these values were determined, they are both disputable.[33]

Until now in veterinary diagnostics, only a cut-off value for cytometric assessment of bone marrow infiltration in canine large B-cell lymphoma has been proposed.[34] To the authors knowledge, there are ongoing studies that aim to establish guidelines for flow cytometric analysis in veterinary medicine, although establishing general guidelines may be challenging as normal tissue specimens, such us peripheral blood, bone marrow, lymph node, spleen, liver or intestinal mass FNA, have a variable percentage of lymphocyte subsets. It should also be emphasized that the expression of some antigens in certain tissues in humans and dogs, for instance CD34 is considered significant outside of the bone marrow even when a relatively small percentage of cells is positive. As for that a precise interpretation of complex flow cytometric data may be difficult and should always be performed by an experienced operator.[35]

As recommended by other authors in this study, the expression of the antigens was evaluated in conjunction with light scattering properties (which are consistent with size and internal cellular complexity) of the lymphoid cells, and the results of flow cytometric immunophenotyping are presented as quantitative data, as showing only the percentage of positive cells may lead to spurious interpretation of the results.[33, 36]

The diagnosis of lymphoma in this study was made when an antigenically homogeneous population of lymphocytes comprising >90% of all lymphocytes was identified in the gated region (CD4 + CD8+, CD5 + CD8+ or CD21+). Gated lymphocytes were considered negative for a certain antigen when it was expressed by less than 10% of the cells. We propose a cut-off value of 90% of lymphocytes for the definition of a predominant antigenically homogeneous population in an aspirate of a lymphatic tissue based on high specificity and high sensitivity of the method used in this study, although further studies evaluating the usefulness of different approaches are needed.

In humans and dogs, a multicolour approach and a higher number of antibodies generated against lymphocyte surface molecules enable recognition of small populations of neoplastic lymphoid cells with aberrant expression of molecules even when a reactive population of residual lymphocytes is present, which may not be identified on cytological examination. It is expected that further development of the technique will also allow a more thorough analysis in cats.

The phenotype has not been established as a prognostic indicator in cats, although further studies with higher numbers of feline patients under standardized treatment conditions should be performed to confirm this.[5]

All fine-needle aspirates were collected into a preservative. Although Streck Cell Preservative has not yet been validated for use in cats, it is routinely used by the authors for collection and shipment of feline and canine specimens for flow cytometric immunophenotyping. The authors' own experience shows that Streck Cell Preservative does not alter the expression of feline antigens or canine antigens.[37]

Authors' preliminary studies (data not shown) indicate that flow cytometric immunophenotyping is also useful in the diagnosis and characterization of non-FeLV-associated feline leukaemia, although further studies are warranted.

Flow cytometric immunophenotyping may become a routinely used tool for confirming the diagnosis of feline lymphoma and its subtypes classified under the WHO scheme. Eventually, it may enable early assessment of the recurrence of feline lymphoma and a more accurate staging of the disease.

Acknowledgements

The authors acknowledge the referring veterinarians from Queen Mother Hospital for Animals – Royal Veterinary College, Small Animal Teaching Hospital – University of Liverpool, Millennium Way, North Downs Specialist Referrals, Culverden Veterinary Group, Pet Vaccination Centre, Powell Torrance Diagnostic Services and CTDS laboratory.

Ancillary