Interleukin‐6 and thrombopoietin concentrations in dogs with carcinoma with and without thrombocytosis

Abstract Background Carcinoma‐associated thrombocytosis involves tumor production of mediators such as interleukin‐6 (IL‐6) and thrombopoietin (TPO) that increase thrombopoiesis and may play a role in tumor evasion and metastasis. Carcinoma‐associated thrombocytosis is described in people, but has not been described in dogs. Hypothesis/Objectives Evaluate the concentrations of IL‐6 and TPO in dogs diagnosed with carcinoma with or without thrombocytosis. We hypothesized that IL‐6 and TPO concentrations would be higher in dogs with carcinoma compared to healthy dogs, and that IL‐6 and TPO concentrations would be higher in dogs with carcinoma and thrombocytosis when compared to dogs with carcinoma and normal platelet counts. Animals One‐hundred sixteen dogs: 63 with carcinoma and 53 healthy control dogs. Methods Complete blood count was performed in all dogs, and they were stratified for sub‐group analysis based on the presence or absence of thrombocytosis (platelet count > 500 103/µL). Serum TPO and IL‐6 concentrations were measured by ELISA. Results of selected numeric variables were compared using Wilcoxon rank sum tests for pairwise comparisons. A value of P < .05 was considered significant. Results Twelve of the dogs with carcinoma (12/63, 19.0%) and none of the healthy control dogs (0%) had thrombocytosis. Thrombopoietin concentrations (median [range]) were significantly higher in dogs with carcinoma when compared to controls (87.42 pg/mL [0 to >600] vs 15.99 pg/mL [0 to >600], P < .001). Interleukin‐6 concentrations (median [range]) were not different between dogs with carcinoma and healthy control dogs (9.70 pg/mL [0‐181.53] vs 3.03 pg/mL [0‐280.77], P = .15). In dogs with carcinoma, the TPO and IL‐6 concentrations were not different between dogs with thrombocytosis and dogs with normal platelet count. Conclusions and Clinical Importance Thrombopoietin concentrations were significantly higher in dogs with carcinoma, regardless of platelet count. Thrombopoietin is likely to be 1 of multiple factors that can impact platelet number, production, and consumption in dogs with carcinoma.

K E Y W O R D S biomarkers, carcinoma, thrombocytosis

| INTRODUCTION
In health, platelets respond to injury, and provide hemostasis as well as a scaffolding for angiogenesis and vessel repair. 1,2 Platelets also have a role in other important pathways such as modulating the immune response, microbial resistance, and cellular regeneration. 1,2 Consumption of platelets leading to thrombocytopenia is described in a number of inflammatory and neoplastic disease processes. 3 Additionally, thrombocytosis has been described in a number of disease states, including endocrine diseases, heart disease, infectious disease, and cancers. [3][4][5][6] Expanded knowledge of platelets and their interactions in disease states has led to investigation into maladaptation of platelet synthesis and function. One maladaptation is paraneoplastic thrombocytosis, in which tumor production of inflammatory mediators or stimulatory molecules leads to an increase in platelet count, which may aid in metastasis and tumor evasion of the immune system. 2,6,7 Several different tumor types in people have been associated with paraneoplastic thrombocytosis, many of which are carcinomas resulting in the term carcinoma-associated thrombocytosis. [5][6][7][8][9][10][11][12][13][14][15][16] Carcinoma-associated thrombocytosis is rarely described in dogs. 4,17,18 Ultrastructure studies in people have shown tumor emboli distant from primary tumors surrounded by platelets, suggesting that tumors have the ability to activate and aggregate platelets. 6 Further evaluation of platelet interaction with tumor cells has shown that platelets can form a cloak around tumor cells to protect them from the immune system. 6,7,19 Additionally, activated platelets stimulate angiogenesis, bringing additional blood flow to tumors. Malignancies are thought to interact with platelets through numerous mechanisms including tumor expression of platelet receptors and tumor secretion of thrombin and tumor-specific cytokines, specifically interleukin-6 (IL-6). 6,7,20,21 In carcinoma-associated thrombocytosis, it is hypothesized that tumor production of IL-6 directly leads to the generation of hepatic thrombopoietin (TPO), which stimulates megakaryocytic generation of platelets. 20,[22][23][24][25][26][27][28][29] Interleukin-6 concentrations are increased in people with ovarian, pulmonary, and urothelial carcinoma. [5][6][7][8][9][10][11][12][13][14][15][16]20,21 Carcinoma-associated thrombocytosis has been associated with and predictive of neoplastic progression in human patients with ovarian cancer, and is an independent prognostic factor in endometrial, pancreatic, esophageal, and pulmonary carcinoma. [5][6][7][8][9][10][11][12][13][14][15][16]20,21 Thrombocytosis has been evaluated retrospectively in dogs and cats, and neoplasia has been identified as a primary or concurrent diagnosis in 17% to 55% of these patients. 3,4,17,18 In dogs with cancer and thrombocytosis, carcinoma was the most common tumor type diagnosed. Carcinoma-associated thrombocytosis has not been investigated in dogs. Carcinoma-associated thrombocytosis represents a novel mechanism for tumor evasion in dogs. There is continued interest in exploring inflammatory mediators or tumor biomarkers that could improve prognostication.
Our aim was to evaluate concentrations of IL-6 and TPO in dogs with carcinoma with respect to their platelet count, and to compare the findings to those in normal, healthy dogs. Our hypothesis was that concentrations of IL-6 and TPO would be significantly higher in dogs with carcinoma when compared to healthy dogs. We further hypothesized that dogs with carcinoma and concurrent thrombocytosis would have higher concentrations of IL-6 and TPO when compared to dogs with carcinoma and normal platelet counts.

| Study design
All dogs diagnosed with carcinoma had blood collected as part of their diagnostic evaluation with excess serum used for the evaluation of inflammatory and thrombopoietic markers. Healthy control dogs had blood collected after physical examination. All blood samples were collected by peripheral venipuncture into 2 tubes: a lavender top tube anticoagulated with EDTA and a serum tube. Lavender top tubes were used for a CBC and blood smear analysis and performed by medical technologists using a commercial analyzer (Advia 2021I, Siemens Medical Solutions Inc, Malvern, Pennsylvania). Serum was separated by centrifugation at 3000g for 10 minutes and transferred to a cryovial for storage in an À80 C freezer.
Based on CBC results, dogs were stratified for subgroup analysis based on the presence or absence of thrombocytosis. Thrombocytosis was defined as a platelet count ≥500 Â 10 3 /μL. Serum IL-6 concentrations were measured using a commercial ELISA (Quantikine ELISA Canine IL-6 Immunoassay, R&D Systems Inc, Minneapolis, Minnesota).
Interkeukin-6 concentrations were measured using a quantitative sandwich enzyme immunoassay technique based on color change.
The intensity of color change then was compared to a standard curve using a microplate reader (Synergy HT multi-detection microplate reader, Biotek U.S., Winooski, Vermont). The detection range for canine IL-6 is 6.1 to 2000 pg/mL. 30 The stability of serum IL-6 at À80 C has been shown to be 6 months. 31 Serum TPO concentrations were measured using a commercial ELISA kit (Canine TPO ELISA kit, mybiosource.com, San Diego, California), or more specifically by quantitative double-antibody sandwich technique resulting in colorimetric change and comparison to a standard curve. The detection range for canine TPO is 15.6 to 1000 pg/mL. Serum TPO stability at À80 C is 3 months. 32 Serum samples were batched and run at 3-month intervals so as to preserve stability of IL-6 and TPO.

| Statistical analysis
A statistical power analysis was performed for sample size estimation, based on prior evaluation of IL-6 and TPO concentrations in people with carcinoma-associated thrombocytosis. 5 In this study, IL-6 and TPO were present at significantly increased mean concentrations (

| RESULTS
One hundred sixteen dogs were included in the study: 63 were diagnosed with histologically confirmed carcinoma and 53 dogs were healthy controls. Originally, 55 healthy dogs were included, but in analysis it was identified that 2 were Beagle mixed breed dogs, and these dogs subsequently were excluded. The dogs with carcinoma were significantly older than those in the control group (P < .01; T A B L E 1 Demographic information and median [range] platelet count for the dogs with carcinoma, subclassified based on the presence or absence of thrombocytosis (defined as a platelet count ≥500 Â 10 3 /μL), as well as demographic information and median [range] platelet count of the group of healthy control dogs

| DISCUSSION
Serum TPO concentrations and platelet count were increased in dogs with carcinoma compared to healthy dogs. This finding suggests carcinoma-associated thrombopoiesis, but thrombocytosis was noted infrequently. In our study, thrombocytosis was noted in 20% of dogs with carcinoma. This frequency is similar to what has been observed in people, in which 10% to 50% of patients with carcinoma have concurrent thrombocytosis. 5,[12][13][14][15][16] Thrombopoietin is a cytokine that is mainly generated in the liver and works in an endocrine manner to stimulate megakaryocytes in the bone marrow, spleen, and other extramedullary sites to increase numbers of platelets. [22][23][24][25][26][27][28] In health, TPO concentration is mediated by platelet load. 22,[26][27][28] Platelets have substantial capacity to absorb, metabolize, and destroy TPO, thus regulating the amount of TPO in circulation. [26][27][28] Neoplastic manipulation of TPO is not completely understood, but it is hypothesized that carcinoma stimulates the pro- Interleukin-6 is produced by carcinomas in people, and is hypothesized to stimulate excess production of TPO. 5,7 In our study, IL-6 concentrations were not increased in dogs with carcinoma, and no difference was found in IL-6 concentration when dogs with carcinoma were stratified by platelet count. In dogs, IL-6 is reliably increased in models of sepsis and systemic inflammatory response syndrome, but IL-6 concentrations tend to normalize within 6 days of an inciting inflammatory event, whereas other acute inflammatory markers such as C-reactive protein and tumor necrosis factor α (TNF-α) remained increased for 14 to 21 days after the inciting event. 34  Mean platelet volume was significantly increased in the carcinoma group when compared to the control group. Mean platelet volume is a commonly used morphological parameter that is determined using a machine-calculated platelet volume. [35][36][37][38] In health, platelet volume has been shown to be inversely proportional to platelet count. 39 Mean platelet volume has been shown to be a significant indicator of inflammation in various types of cancer in people, both independent of and when compared to platelet count. 35,36 However, in inflammatory conditions, MPV can be discordant with platelet count, as was the case in our patient population. 37,38 In people, measurement of immature platelet fraction can help differentiate among causes of increased destruction or consumption of platelets and decreased production of platelets. [35][36][37][38] This approach may be a more direct method for evaluation of factors affecting platelet count in dogs that could be evaluated in future studies.
Our study identified an increase in TPO in dogs with carcinoma, but an association with thrombocytosis was not observed. This disparity is likely a consequence of small sample size and overrepresentation of cer- Additionally, comorbid diseases in these dogs likely impacted platelet count, both positively and negatively, but future studies will be necessary to understand if these diseases affect TPO concentration.
Given that thrombocytosis creates a favorable environment for neoplastic cells, it represents an intriguing and important abnormality to investigate. Thrombopoietin is increased in dogs with carcinoma regardless of platelet count, but is likely only 1 component of a more complex pathophysiology in the tumor environment of carcinoma in dogs.

ACKNOWLEDGMENT
No funding was received for this study.