An experimental study to compare inflammatory response due to liquid or gas joint distension in horses submitted to arthroscopy

PURPOSE: To assess comparatively the inflammatory response that follows CO2 or Ringer’s lactate joint capsular distension of horses submitted to experimental arthroscopy METHODS: Each animal was submitted to a bilateral tarsocrural arthroscopy employing gas distention in one joint and fluid distention in the contralateral joint. Synovial fluid was evaluated at 0, six, 12, 24 and 48 hours post-operative. RESULTS: The use of CO2 for arthroscopy causes an acute and mild synovitis alike to the liquid capsular distension, showing similar synovial fluid increase of leukocytes, TP, and TNF-α. Although synovial fluid PGE2 content was higher in joints submitted to CO2 distension, lower levels of hemoglobin and leukocytes oxidative burst after surgery indicates that CO2 arthroscopy decreased intraarticular bleeding and activation of infiltrating leukocytes. CONCLUSIONS: The use of CO2 for arthroscopic examination causes acute and mild synovitis that is similar to the effects caused by the liquid capsular distension. CO2 also seems to decrease intra-articular bleeding and activation of leukocytes.


Introduction
Although the use of carbon dioxide had been described since 1921 in the human literature 1 liquid distension remains the most commonly protocol used in arthroscopy for joint capsule distension. Compared to fluid arthroscopy, the use of CO 2 provides easier evaluation of intra-articular anatomical structures, increased joint space and visualization and also improves identification and removal of bone fragments 2 . Additionally, intra-articular bleeding does not cause the same amount of impaired vision as in liquid distension because blood will just follow gravity to the deepest point of the joint 2 .
Other advantages that gas distension using CO 2 offers for operative arthroscopy is the possibility of using CO 2 laser, microarthroscopy, in vivo staining of the synovial membrane and transplantation of chondrocytes and/or cancellous bone grafting 2 .
Besides all this described advantages, the use of this technique remains controversial: catastrophic reports of fatal embolization 2 , increased pain 3 and a tendency to cause some irritation to the joint in humans 4 were described as complications of this technique.
The inflammatory response to arthroscopy using fluid capsular distension in horses was described as minimal, transient and frequently overwhelmed by its benefits to treatment 5 . However none of these effects has been investigated when employing CO 2 during arthroscopic examination. Therefore, the aim of this study was to assess comparatively the inflammatory response that follows CO 2 joint capsular distension to that of Ringer's lactate solution distension.

Methods
Nine health adult horses (weighing between 350 and 400kg) of mixed breed and sex, with no history or signs of tarsal joint diseases, normal findings on physical and lameness examination, tarsal radiographs, complete blood count, and synovial fluid analyses were studied. One tarsus of each horse was randomly assigned to gas distension (GD) and the opposite tarsus underwent the liquid distension (LD). During the experimental period the animals were kept in individual, 3 x 4 m stalls. Each animal was fed a total of 4kg of pelleted concentrate, twice a day; while grass hay and water were offered ad libitum. Bedding consisted of wood shavings and was changed daily. The study protocol was approved by the Animal Care and Use Committee. All of the horses were cared for according to standards and guidelines and were adopted by private individuals upon conclusion of the study.

Surgical procedure
Horses were anesthetized, positioned in dorsal recumbency, and both hocks (tarsocrural joint) aseptically prepared for surgery. Synovial fluid samples were collected immediately before liquid distension of the joint for arthroscope insertion (M0).
After distension of the tarsocrural joint with approximately 60 mL of Ringer's lactate solution, the arthroscope portal was made on the dorsal aspect of the dorsomedial pouch of the tarsocrural joint.
A 19-gauge needle was inserted into the dorsolateral pouch and used to ensure a suitable location for the instrument portal. Joint After the 30 minutes of distension, the excess liquid was drainage from both joints and skin portals were closed with nonabsorbable suture. Sterile bandages were applied.

Postoperative care
Horses received ceftiofur (2.2mg/kg intravenously) before induction of anesthesia and every 24 hours for two days.
Because the aim of the study was to assess the inflammatory response, no NSAIDs were used after surgery but pain was successfully managed using the opioid petidine (4mg/kg intramuscularly) every 12 hours for two days.

Synovial fluid sample collection and analysis
Synovial fluid specimens were collected aseptically from the dorsomedial pouch of the tarsocrural joint at time 0, six, 12, 24 and 48 hours after surgery in heparinized syringes.
Synovial fluid color and clarity were evaluated subjectively immediately after collection, and total protein, hemoglobin and leukocytes concentrations were determined by use of routine clinicopathologic methods (biuret technique, photometric detection of cyanmetahemoglobin and manual count). As described previously 6 color of synovial fluid was graded as yellow, orange or red, and numerical values of 1 to 3, respectively, were assigned. Clarity was graded as clear, hazy, cloudy or opaque, and numeric values of 1 to 4, respectively were assigned. Differential leukocyte counts were made by counting 100 cells on direct smears that were air dried and stained with Methanol-May Grunwald-Giemsa.
Synovial fluid samples for analysis of total protein, hemoglobin and TNF-α were collected in microtubes and centrifuged, and the cell free supernatant was stored at -80°C until assay.

Tumor necrosis factor-α bioassay
The bioassay for TNF-α was adapted from the assay described by Santos et al. 8

Quantification of PGE 2
The PGE 2 concentration was determined by use of commercially available human PGE 2 EIA Kits (Cayman, EUA) according manufacturer's instructions.

Statistical analysis
Statistical analyses were performed using computer software 1 . For synovial fluid total protein, hemoglobin concentrations and leucocytes count, Kruskal-Wallis nonparametric tests followed by Dunn post test was used to evaluate differences between distension protocols and among different times of collection. Pairwise comparisons were made at each time point for synovial fluid PGE 2 , TNF-α and oxidative burst production, using Friedman two-way analysis of variance by ranks. Results are expressed as means ± SEMs. Synovial fluid PGE 2 concentration was expressed as a ratio, compared with values for baselines samples. A paired t test was used to evaluate differences between distension protocols. A significance level of p≤0.05 was assigned for all tests.

Color and clarity of synovial fluid
After six hours (M6), all specimens were similarly cloudy and red in color. Although all samples remained cloudy in every moment, after M6, the liquid from joints of group GD started to became orange in color differentially from group LD where all samples remained red until M48.

Synovial fluid WBC count
Significant differences were not detected in relation to number of cell between groups but in the kinetic cell migration group GD it was observed a more acute increase in WBC count compared to it control values at M6 and M12 (p<0.01 to M0). The group LD only reached its peak at M12 (p<0.01 compared to M0).
After M12, the WBC count decreased similarly in both groups but remained significant higher than M0 values at M48 (p<0.05) ( Figure 1).  Only at group GD a tendency to return to normality was noted by an increase in the presence of mononuclear cells because at M48 the percentage of mononuclear cells was significant higher (p<0.01) than it percentage at M6 (Table 1). to liquid distension showed an increase in concentration reaching its peak at M24 (p<0.001 compared to M0) and remained significant higher than it control levels at M48 (p<0.05) ( Table 2). Quantification of leukocyte oxidative burst   At the joints submitted to gas distension, the oxidative burst by PMN increased four fold to it control levels after the surgical procedure and remained significantly high until M48 (p<0.05 to control levels).
The oxidative burst by macrophages increased after the surgical procedure in both groups.

PGE 2 production
Synovial fluid PGE 2 ratio was significantly increased over time in joints of both groups ( Figure 4) compared with M0.
Although no statistically different, the joints submitted at gas distension had peak ratio of 11 ± 3 times of the control levels versus 5.1 ± 1 times greater than the control levels from the joints from LD. These peak ratios were observed at 12 hours (M12).
Gas distension also had a significant decrease to its peak values (p<0.05) when compared to 12 hours after the surgical procedure.

TNF-α production
TNF-α concentration was increased two folds at six hours after the surgical procedure in both groups in a similar way but not statistically significant compared to its baseline values. At 12 hours both groups TNF-α concentration had already returned to control values ( Figure 5). Our results regarding the oxidative burst demonstrate that significantly higher oxidative burst observed in this experiment by the joints of group LD suggests that the joint distension protocol using Ringer lactate's can potentially cause a bigger oxidative damage to joint when compared to gas distension. Apparently the joints submitted to CO 2 distension suffered a higher inflammatory stimulus to the synovial membrane compared to the joints submitted to liquid distension, because PGE is considered one of the principals' indicators of active synovial inflammation 14 and this joints also had a higher and more acute increase in the WBC.
The peak of TNF probably preceded our first sampling after surgery at six hours because at this moment we verified the maximum level of TNF in both groups but were not significant higher than baseline values. The maximum levels of TNF in joint fluid preceded the infiltration of leukocytes and coincide with the data reported by Pettipher and Salter 15 that TNF is produced from the resident joint tissues such as the synovial lining cells rather than infiltrating neutrophils or monocytes. Apparently the TNF values at our study were not influenced by repeated arthrocentesis similarly as reported by Cornelissen et al. 16 .

Conclusion
The use of CO 2 for arthroscopy examination in horses induces transient synovitis similar to the one observed after Ringer's lactate solution and is safe to be used in horses.