Clinical performance of a commercially available thymidine kinase 1 assay for diagnosis of lymphoma in 42 hospitalized horses (2017‐2020)

Abstract Background Antemortem definitive diagnosis of lymphoma in horses is often difficult. Thymidine kinase 1 (TK1) assay is a potentially useful biomarker for lymphoma in horses. Hypothesis/objectives To report the clinical performance of a commercially available TK1 assay for diagnosis of lymphoma in horses. We hypothesized that there would be no association between serum TK1 activity and a diagnosis of lymphoma in horses. Animals Forty‐two hospitalized horses, 14 with a definitive diagnosis of lymphoma, 4 with other neoplasia, and 24 with inflammatory disease. Methods Retrospective medical record review, groups were compared via Kruskal‐Wallis and Mann‐Whitney tests, and logistic regression was performed. Results Median (range) TK1 was 3 U/L (0.4‐17.7 U/L) in horses with lymphoma and 3.9 U/L (0.8‐94 U/L) in horses without lymphoma (P = .59). There was no significant difference in total protein between horses with and without lymphoma (6.6 g/dL [5.5‐8.3 g/dL] vs 6.6 g/dL [4.7‐10.4 g/dL]; P = .83). There was no significant difference in fibrinogen between horses with and without lymphoma (447 [100‐1364] mg/dL vs 433 [291‐2004] mg/dL; P = .47). On logistic regression, serum TK1 activity was not associated with a diagnosis of lymphoma (odds ratio, 0.97; 95% confidence interval, 0.9‐1.05, P = .48). Conclusion and Clinical Importance Serum TK1 values were not predictive of lymphoma diagnosis in this cohort of horses.


| INTRODUCTION
Lymphoma is a common malignant neoplasia in horses, with a reported prevalence as high as 14% in some cohorts. 1 Common clinical signs associated with lymphoma in horses are vague and variable and can include fever, lymphadenopathy, weight loss, colic, ventral or peripheral edema, cutaneous lesions, diarrhea, and/or inappetence. 2 Antemortem diagnosis of equine lymphoma often remains a diagnostic challenge for clinicians.
Because of the invasive nature of obtaining biopsies to confirm a diagnosis of lymphoma, an accurate serum biomarker would be an attractive alternative. An assay for such a biomarker, thymidine kinase 1 (TK1), has been validated for use in horses. 3 Abbreviation: TK1, thymidine kinase 1.
Thymidine kinase 1 is a cytosolic enzyme involved in pyrimidine salvage during DNA synthesis. 4 Thymidine kinase 1 is upregulated during rapid cell proliferation and serum activity of this enzyme correlates with levels of cell proliferation within neoplasms. Increased serum TK1 activity is seen associated with leakage of the enzyme from dying cells during replication and increased activity of TK1 is associated with hematologic neoplasia in humans. 5,6 Serum TK1 activity is used as a marker of lymphoma in dogs and cats. 4,7,8 The use of TK1 has expanded in other species, including monitoring progression of disease and remission in dogs, cats, and humans. [9][10][11][12] Serum TK1 activity in horses can be transiently elevated with inflammatory conditions and is consistently elevated in horses with lymphoma. 3 However, despite its clinical utility in other species, as well as the recent study suggesting its use in horses, our clinical impression has been that TK1 is not a clinically useful marker for lymphoma in horses.
Therefore, the objective of this study was to report the results of a commercially available thymidine kinase assay in a cohort of hospitalized horses in which lymphoma was suspected or diagnosed. We hypothesized that there would be no association between serum TK1 activity and a diagnosis of lymphoma in horses.

| Data collection
Cases were identified by search of electronic medical records from 2017 to 2020 at the George D Widener Hospital for Large Animals at the University of Pennsylvania's New Bolton Center. Case inclusion criteria consisted of (a) horses >6 months of age with (b) serum thymidine kinase assay (VDI Labs, Simi Valley, California) and (c) cytologic or histologic diagnosis of lymphoma attempted, and/or necropsy results (where applicable).
Additional information gathered from medical record review included signalment, presenting complaint, total protein, fibrinogen, diagnostics performed (biopsy, cytology, and/or necropsy), if lymphoma was identified, and any other neoplasia or underlying disease that was identified in the course of hospitalization. If lymphoma was identified, the tumor location and cell type was characterized as previously described. 2 Horses were dichotomized as either lymphoma (based on histopathologic diagnosis) or as not lymphoma (based on a final diagnosis obtained and lack of confirmation of lymphoma on histopathologic or cytologic analysis).

| Laboratory and clinical methods
Whole blood was collected from the jugular vein of all horses to evaluate serum TK1 activity. The serum was collected, frozen, and shipped for subsequent analysis by a commercially available TK1 assay (Veterinary Diagnostics Institute, Simi Valley, California). The LIASION TK assay (DiaSorin, Stillwater, Minnesota) is an indirect, modified 2 step, competitive chemiluminescence immunoassay for the quantitative determination of serum TK1, which has been validated for use in monitoring lymphoma in dogs as a nonradiometric assay. 8,13,14 The laboratory reference range reports that a result >3.3 U/L is consistent with a diagnosis of lymphoma, while evaluation of a radioimmunoassay method in horses found a result of >2.7 U/L to be consistent with a diagnosis of lymphoma. 3

| RESULTS
Forty-two horses met the inclusion criteria, with a median age of 13 years (range, 1-28 years). There were 13 mares, 23 geldings, and

| DISCUSSION
In this retrospective evaluation of 42 hospitalized horses, there was no association between serum TK1 activity and a diagnosis of lymphoma. This is in contrast to the results previously reported in horses, dogs, and cats. 3 The horses within our sample diagnosed with other neoplasia had a similar median serum TK1 when compared to those diagnosed with lymphoma. This could be owed to the non-specific nature of the cytosolic enzyme which is not limited to hematopoietic cells. This is evident in hepatocellular damage in viral hepatitis, where death of hepatocytes can spill TK1 into the circulation. 16 Two horses with hepatitis in our study had TK1 values of 7.0 U/L and 10.9 U/L, while a third horse with a hepatic abscess had a TK1 of 6.9 U/L. Given the non-specific nature of the enzyme, it has been utilized in humans in 15 types of malignancies, both solid tumors and hematopoietic. 12 Within our sample, no significant difference was found between horses diagnosed with any form of neoplasia and those who did not, despite a variety of tumor types being represented in the sample. In contrast, dogs with leukemia and lymphoma showed marked increases in TK1 activity, when compared to dogs with inflammatory disease, healthy controls, and those in remission. 17 However, in another study, dogs diagnosed with B-cell lymphoma did have higher TK1 activity than those with T-cell lymphoma. 10 The most common immunophenotype identified in horses is T-cell rich large B-cell lymphoma, as demonstrated in previous reports of equine lymphoma. 2,17,18 Given this variation between immunophenotypes in canines, it could be that this specific immunophenotype in horses does not reliably cause elevations in serum TK1 activity. Therefore, another immunophenotype or neoplasm could cause more reliable elevations in TK1 activity in horses. However, it is difficult to support that assertion from our sample, as only 2 immunophenotypes were reported: 1 T-cell rich large B-cell and 1 B-cell lymphoma.
Another possible explanation for the lack of association in our sample is a lack of understanding of the temporal trends of TK1 in horses with lymphoma. In contrast, serum TK1 has been utilized as a diagnostic in human oncology since the 1980s, and therefore this value has been well-established throughout the course of disease, even preceding the diagnosis of neoplasia. In humans, TK1 elevation is associated with numerous comorbidities such as obesity, fatty liver, and pre-cancerous lesions. Patients within this cohort also had a 3-to 5-fold risk of developing cancer in the subsequent 6 years. 19 It is therefore reasonable that horses without a histopathologic diagnosis of lymphoma that displayed elevated TK1 in our sample could represent a pre-cancerous group. However, because our sample represents horses at a single time point, this precludes the recognition of trends over time. It also should be noted that several horses within our sample had cutaneous topography, which in dogs have been noted to have less marked TK1 activity. 10 Some equine lymphoma subtypes might have a more indolent course, leading to inherently lower serum TK1.
Because of the retrospective nature of this study, the study was limited by small sample size, the information available within the medical record, lack of a validated assay, and a lack of follow-up. The lack of validation of the TK1 assay used in this study presents a major limitation.
The commercially available TK1 assay used in this study (chemiluminescence immunoassay) is a different method from the method previously evaluated in horses (radio enzyme immunoassay). 3 In a previous study in dogs, the chemiluminescence assay had good correlation with the radioimmunoassay, but similar validation has not been performed in horses. 14 Further research comparing the 2 methods in horses is warranted. The horses reflected in the study cohort likely represent selection bias, as the TK1 assay was submitted by clinicians based on a clinical indication or suspicion of lymphoma, and no normal horses were included in this study. techniques. This is amplified in large animal species where it has been reported that even despite exhaustive attempts to obtain an antemortem diagnosis of lymphoma, 62% could only reach the diagnosis of intestinal lymphoma at postmortem. 21 While the availability of diagnostics has improved since the publication of that study, the clinical experience in the pursuit of a diagnosis of lymphoma in horses is still difficult for clinicians. In conclusion, this study showed no association between a diagnosis of lymphoma and serum TK1 activity in a sample of 42 hospitalized horses.

ACKNOWLEDGMENT
No funding was received for this study.

CONFLICT OF INTEREST DECLARATION
Authors declare no conflict of interest.

OFF-LABEL ANTIMICROBIAL DECLARATION
Authors declare no off-label use of antimicrobials.

INSTITUTIONAL ANIMAL CARE AND USE COMMITTEE (IACUC) OR OTHER APPROVAL DECLARATION
Authors declare no IACUC or other approval was needed.

HUMAN ETHICS APPROVAL DECLARATION
Authors declare human ethics approval was not needed for this study.