Correlation of synovial leptin with body habitus in the healthy and osteoarthritic canine stifle
Stephanie Kleine1, Sherry Sanderson2, Clint George3, Ira Roth1, Robert Gogal3, MaryAnn Thaliath1, Steven Budsberg1
1 Department of Small Animal Medicine and Surgery, University of Georgia, Athens,Ga, 2Department of Comparative Physiology and Pharmacology,
University of Georgia, Athens, Ga, 3Department of Veterinary Biosciences and Diagnostic Imaging, University of Georgia, Athens, Ga
Email: sakleine@gmail.com
INTRODUCTION: Naturally occurring cranial cruciate ligament (CrCL) insufficiency and progressive osteoarthritis (OA) is a common identity in the
canine population and can serve to provide useful information to understand both canine and human knee joint disease. Additionally, there is a high
prevalence of obesity in dogs that often parallels human obesity. While alterations in joint loading may be important in the progression of OA,
biomechanical changes alone do not explain the link between OA and obesity in non-load bearing joints. Leptin, an adipocytokine released from white
adipose tissue, has been shown to be proportional to adipose mass in humans and canines, with serum leptin being higher in obese compared to lean patients.
Serum leptin, in humans, has been associated with radiographic progression of knee OA and markers of cartilage damage. Additionally, an association
between synovial fluid (SF) leptin and pain in osteoarthritic knees has been reported. While in humans the link between obesity and OA is more extensively
studied, there is very little evidence of this relationship in dogs. The primary goal of this study is to evaluate the relationship between serum and SF leptin
and body condition score, body fat mass, and body fat percentage in dogs. The secondary goal was to characterize the relationship between SF leptin and
markers of synovial inflammation, such as IL-1β.
METHODS: Forty-eight client-owned dogs, 19 healthy controls and 29 with CrCL injury and OA, were enrolled in the study. Dogs in the OA group were
presented for surgical repair of CrCL deficiency. Any dog with evidence of systemic disease, based on physical exam and routine hematology was excluded
from the study. Informed owner consent was obtained prior to enrollment and all procedures were approved by the Clinical Research Committee. Sample
Collection – Dogs in the control group were sedated with dexmedetomdine (5mcg/kg I.V.) and a single femortibial joint was randomly selected for
arthrocentesis. Synovial fluid was collected using aseptic technique and joint lavage was not performed. For dogs in the OA group, SF was obtained intraoperatively
from the affected limb prior to arthrotomy. Blood was collected via standard jugular venipuncture. All dogs were fasted for 8 hours.
Morphometric Measures – Body condition score, BCS, was determined on a standard 9-point scalea by 3 individual observers, who were unaware of the
other raters’ score. Dogs were then divided into to 2 groups, optimal weight (BCS 4-5) and overweight (BCS >5). Additionally trunk and limb measurements
were obtained to calculate body fat mass (g) and body fat percentage, as previously reported1. Osteoarthritis Severity Scoring- In the OA group, radiographs
were scored for severity of OA by a radiologist utilizing a previously described scoring system2. Owners were required to fill out a diet history and a clinical
metrology questionnaire (Liverpool Osteoarthritis in Dogs, LOAD). Finally intra-operative meniscal status was documented. Sample Analysis - Serum and
SF leptin concentration was measured with a canine specific leptin ELISAb. IL-1β was also determined in serum and SF via a canine IL-1β ELISAc. Data
Analysis – Student t-tests were used to test for differences SF leptin means due to CCL status, presence of IL-1β in serum or SF, meniscal status (in the OA
group only), and the potential confounders non-steroidal anti-inflammatory (NSAID) use and dietary Ω-6:Ω-3. Linear regressions were used to test for
association of synovial leptin with body condition score, body fat percent, serum leptin concentration, LOAD, and radiographic score (for OA group only).
Multiple linear regression was used to adjust the linear models for the association of synovial leptin and the primary variables of interest for confounding
factors with p<0.10. Multiple linear regression was also used to adjust the linear model for the association of synovial leptin and OA status for body
condition score.
RESULTS SECTION: SF leptin was significantly lower than serum leptin (p<0.01) and there was a 5.8ng/mL change in SF leptin for every 10ng/ml
increase in serum leptin. Overweight dogs had significantly higher concentration of SF leptin than those that were optimal body weight (p<0.01). SF leptin
concentration was associated with both BCS and body fat % (p<0.01). For every 1unit increase in BCS there was a 1.7ng/mL increase in SF leptin and for
every 10% increase in body fat, there was a 2.7ng/mL increase in SF leptin. Synovial fluid leptin was significantly higher in dogs with OA by a mean
amount of 2.8 (95% CI 0.6-5.1) but this difference decreased and became no longer significant when adjusted for differences in body condition score
between OA groups (p=0.25). There was a weak association between SF leptin and LOAD; for every 1ng/mL increase in leptin a 10unit increase in LOAD
was observed (p=0.03). Also, SF leptin increased 0.5ng/mL for every 1unit increase in dietary Ω-6:Ω-3. There was no significant association between SF
leptin and radiographic score (p=0.24), SF IL-1β (p=0.64) or NSAID use (p=0.55).
DISCUSSION: This is the first study to document the presence of leptin in canine SF. As expected, SF leptin was positively correlated with serum leptin and
measures of body habitus, including BCS and body fat %. However, SF leptin levels in dogs with CrCL insufficiency were not significantly different from
control dogs. Still, there was weak positive association between SF leptin and LOAD, in dogs with OA. Also, more objective measures, such as kinetic gait
analysis or activity monitoring, may better detect pain and dysfunction than LOAD. As all these dogs had CrCL insufficiency at the time of data collection,
the pain and dysfunction from joint instability may have masked small changes due to chronic OA and contributed to the weak nature of this link. Unlike
human studies, SF leptin was not associated with radiographic severity of OA. Additionally, contrary to previous studies reporting increased chondrocyte
secretion of IL-1β in the presence of leptin, there was no correlation between SF leptin and IL-1β. One limitation to this study was that all dogs in the OA
group had CrCL insufficiency for greater than 1 month prior to enrollment in the study. Given the chronicity, IL-1β concentrations may have been slowly
decreasing prior to collection of SF. The results of this study did not correlate leptin levels with severity of OA, or local concentrations of IL-1β. In the
canine model of OA, obesity related OA severity may be related to other adipokines or may purely be biomechanical in nature. Further studies are needed in
both people and dogs to determine the role of leptin in OA.
SIGNIFICANCE/CLINICAL RELEVANCE: Serum and synovial fluid leptin are correlated with body habitus in canines; however leptin was only weakly
associated with LOAD and not associated with radiographic severity of OA or synovial inflammation as measured by IL-1β.
a. Body Condition System, Nestle Purina, St. Louis, MO, USA
b. Canine Leptin ELISA, EMD Millipore, Darmstadt, Germany
c. Canine IL-1β ELISA, ThermoFischer Scientific, Waltham, Ma, USA
1. Witzel et al, 2014, J Am Vet Med Assoc.
2. Torres et al, 2016, Vet Comp Orthop Traumatol.

1239 - Correlation Of Synovial Leptin With Body Habitus In Healthy And Osteoarthritic Canine Stifles