Serum C-reactive protein and iron levels following gonadectomy are not modified by perioperative administration of robenacoxib to dogs

Abstract

The aim of this study was to evaluate the perioperative effects of robenacoxib on serum C-reactive protein (CRP) and iron concentrations in dogs undergoing gonadectomy. In a prospective, blinded, controlled clinical trial, 60 healthy dogs were randomly assigned to receive preoperative subcutaneous injection of either robenacoxib [2 mg/kg body weight (BW)], meloxicam (0.2 mg/kg BW), or saline (0.04 mL/kg BW), followed by oral administration over 72 h (robenacoxib: 2 to 4 mg/kg BW; meloxicam: 0.1 mg/kg BW; saline: gelatin capsules). Blood samples were taken before surgery and 12, 24, 48, 72 h, and 7 d after surgery. Pain scores were assessed via the short-form Glasgow Composite Pain Scale over 72 h postoperatively. C-reactive protein (CRP) and iron serum levels increased and decreased (P < 0.01, both), respectively, after surgery and returned to baseline within 1 wk. No differences were observed among treatments (P > 0.05) or based on surgery/gender (P > 0.05). Pain assessment revealed a higher incidence of treatment failure in saline (6 females versus 2 and 1 female in robenacoxib and meloxicam, respectively). In conclusion, robenacoxib and meloxicam had no influence on postoperative CRP or iron in dogs, which suggests that these nonsteroidal anti-inflammatory drugs (NSAIDs) do not have a relevant effect on these biomarkers.

Introduction

The assessment of acute-phase proteins (APPs) can be used as a rapid screening test for inflammation, as they are more sensitive than other biomarkers, such as leukocytosis and neutrophilia (1), even when clinical signs are absent (2). C-reactive protein (CRP) is one of the APPs released by the liver in response to inflammation and is thought to be less affected by stress factors such as excitement, fear, and handling (3).

The characterization of changes in serum CRP in dogs undergoing surgery and its relevance in assessing postoperative inflammation to monitor recovery from surgery has been reported. In dogs, CRP may predict infectious complications after tibial plateau levelling osteotomy (4), as well as postoperative complications after ovariohysterectomy in bitches with pyometra (5). Furthermore, serum CRP levels may actually match the magnitude of the surgical insult (1,6,7).

Although administering nonsteroidal anti-inflammatory drugs (NSAIDs) for perioperative pain management in dogs is common practice, their effects on CRP are limited and contradictory. Furthermore, NSAID treatment may interfere with the normal pattern of acute-phase response by altering CRP and iron levels and thus limiting its diagnostic value.

No effects of either perioperative meloxicam or carprofen (preferential COX-2 inhibitor) on CRP levels in dogs undergoing ovariohysterectomy were reported in one study (8). In humans, postoperative flurbiprofen, a non-selective NSAID, reduced CRP levels after thoracotomy (9), although a variable effect has been observed in patients with rheumatoid arthritis, depending on the NSAIDs considered (COX-2 selective/non-selective) (10). This suggests an individual pattern of response to specific NSAIDs.

Serum iron has been used as a biomarker of acute inflammation in humans and other animal species (11,12). Recent work has reported decreased iron levels following ovariectomy and ovariohysterectomy in dogs (7), a mechanism that might enhance the organism’s nonspecific resistance to bacterial infection (11).

Robenacoxib has recently been introduced for the relief of perioperative pain and inflammation in dogs and cats. Robenacoxib has a higher selectivity for inhibition of COX-2 versus COX-1 than meloxicam (13) and is regarded as having a wide margin of safety. However, the effects of robenacoxib on CRP or iron levels have not been characterized.

We hypothesized that preoperative treatment with robenacoxib might modify CRP and iron levels in dogs undergoing gonadectomy. The aim of the present study was to determine whether robenacoxib might modify concentrations of serum CRP and iron in dogs undergoing gonadectomy and to compare its effects to the known lack of effects of meloxicam on CRP levels.

Materials and methods

This study was approved by the Institutional Ethics Committee (University of Extremadura; reference number 42/2014). Informed consent was obtained from owners before dogs were included in the study.

Sixty adult client-owned dogs (30 males and 30 females) of various breeds admitted for elective gonadectomy were enrolled in the study. Dogs were clinically examined to ensure their behaviors were suitable for the experimental protocol and were classified according to the American Society of Anesthesiologists’ (ASA) criteria as having a physical status 1 or 2. Exclusion criteria included pre-existing disease or prior administration of anti-inflammatory drugs within the 2 preceding weeks. The dogs were admitted to the hospital 24 h before surgery. During this period, the observer, who assessed pain, fed and exercised the dogs every 8 h. All dogs were fasted overnight before anesthesia but had free access to water.

Treatment groups

A randomized block design was used with the same number of dogs in each group (n = 20) and the same distribution of males and females (n = 10:10) among the groups. The researcher responsible for administering the treatments (VLR) blindly selected a card from the corresponding bag (male, female) containing 30 cards, 10 assigned to each group (robenacoxib, meloxicam, or saline). Once selected, the card was sealed in an envelope and stored until the collection period was completed.

All treatments were first administered by subcutaneous (SQ) injection after 20 min of preanesthetic medication. Dogs were administered either: i) Robenacoxib (Onsior 20 mg/mL; Novartis Animal Health UK, Basel, Switzerland), 2 mg/kg body weight (BW), followed by oral administration of robenacoxib tablets (Onsior 5, 10, 20, 40 mg; Novartis Animal Health UK), 2 to 4 mg/kg BW, PO, q24h for 72 h, according to the manufacturer’s instructions; ii) Meloxicam (Metacam 0.5 mg/mL; Boehringer Ingelheim Vetmedica, Germany), 0.2 mg/kg BW, followed by meloxicam tablets (Metacam 1 mg; Boehringer Ingelheim Vetmedica), 0.1 mg/kg BW, q24h in food for 72 h, in fractions to adjust for the weight of the dogs; or iii) Saline [Placebo, sodium chloride (NaCl) 0.9%; B Braun Vet Care, Barcelona, Spain], 0.04 mL/kg BW, followed by empty gelatin capsules given with food, q24h over 3 postoperative days. The incidence of vomiting, diarrhea, or any other adverse events in each group was documented.

Surgical procedure

A catheter was placed in the cephalic vein and each animal received acepromazine (Equipromazina 5 mg/mL; Fatro Iberica, Barcelona, Spain), 0.025 mg/kg BW and buprenorphine (Buprex 0.3 mg/mL; RB Pharmaceuticals, Hull, United Kingdom), 0.02 mg/kg BW intravenously (IV). After premedication, dogs were kept in a quiet environment for 20 min and then received a single dose of ampicillin (Britapen 500 mg; Laboratorio Reig Jofré, Barcelona, Spain), 11 mg/kg BW, IV and the assigned treatment (meloxicam, robenacoxib, or saline), as previously described.

The investigator responsible for the anesthetic procedure (V.V.) was blinded to the drug treatment. Anesthesia was induced with propofol (Vetofol 10 mg/mL; Esteve Pharmaceuticals, Barcelona, Spain) to effect, until orotracheal intubation was carried out 30 min after premedication. A cuffed orotracheal tube was placed and anesthesia was maintained with isoflurane (Isovet; Piramal Healthcare UK, Morpeth, United Kingdom) in 100% oxygen delivered via a circle system with a fresh gas flow of 1 L/min. The vaporizer setting was adjusted to maintain a surgical plane of anesthesia as judged by eye position, jaw tone, and lack of response to noxious stimuli. Intravenous fluid therapy was provided with Ringer’s Lactate solution (Lactato-RingerVet; B Braun Vet Care) infused at 10 mL/kg BW per hour.

Surgery started 20 min after anesthetic induction when the animals were properly instrumented, depth of anesthesia was achieved, and the surgical area clipped and aseptically prepared. Open ovariohysterectomy via a midline approach and prescrotal orchidectomy with an electrothermal bipolar vessel-sealing device (LigaSure Atlas Tissue Fusion; Covidien, Boulder, Colorado, USA) were carried out by the same experienced surgeon. The suspensory ligament, the ovarian and uterine artery and vein, the uterine body, and the spermatic cord were sealed with the device. The linea alba, vaginal tunic, and subcutaneous tissue were sutured with a simple continuous pattern and the skin was closed with an intradermic pattern.

During surgery, end-tidal carbon dioxide (CO₂), end-tidal isoflurane, heart rate, respiratory rate, invasive arterial blood pressure, pulse oximetry, and esophageal temperature were continuously monitored (S/5 Compact Monitor; Datex Ohmeda, Madison, Wisconsin, USA). Expired gases were sampled from the orotracheal tube connector; a lead II electrocardiography was also monitored.

At the end of surgery, the vaporizer was turned off and the orotracheal tube remained in place until the dog recovered the swallowing reflex. The cephalic vein catheter was left in place for 24 h to administer rescue analgesia. The dogs were placed in a quiet environment and continuously observed until complete recovery. The time from premedication to the start of surgery and the surgery time were calculated.

Serum CRP and iron

Blood samples were collected from the cephalic vein of all dogs in 2 tubes containing EDTA-K₃ (2 mL, BD Vacutainer K2 EDTA; Becton Dickinson, Madrid, Spain) and a clotting activator (3.5 mL, Vacuette CAT; Greiner Bio-One, Madrid, Spain) 30 min before preanesthetic medications (8:00 AM, baseline) and 12, 24, 48, 72 h, and 1 wk after surgery. Hematological variables were determined within 6 h after sample collection with an automated analyzer (Mindray BC-5300; Spinreact, Barcelona, Spain). Serum was prepared by 10-minute centrifugation (200 × g) of full blood, frozen at −20°C, and stored until analysis.

Serum CRP concentrations were analyzed by turbidimetric immunoassay for measuring human CRP, which was previously validated for dogs (14). Iron concentrations were measured with a commercial kit (Spinreact) and clinical chemistry analyzer (Quantum Saturno 100 Vet; Quantum Vet Diagnostics, Lancashire, England). All laboratory tests were carried out according to the manufacturer’s instructions by trained laboratory staff.

Pain assessment

The short-form of the Glasgow Composite Measure Pain Scale (CMPS-SF) was applied by an experienced person (VV) blinded to the drug treatment. Pain values were obtained before treatment (baseline) and at 2, 4, 6, 12, 24, 48, and 72 h after NSAIDs were administered. The surgical wound was always palpated with a conscious effort to apply the same amount of pressure. If an animal showed a score of 6/24 or higher, rescue analgesia with buprenorphine (Buprex 0.3 mg/mL; RB Pharmaceuticals) was provided at 0.02 mg/kg BW, IV and dogs were further assessed for pain up to 72 h. Additional data from these animals were recorded but not considered for the statistical analysis of pain scores. Dogs were discharged 3 d after surgery when the final assessments were completed.

Statistical analysis

A sample size of 9 to 10 dogs was considered adequate to detect an expected difference of 1 mg/L CRP or 35 μg/dL of iron, with a standard deviation of 0.79 mg/L and 26.33 μg/dL (15), respectively, a power of 80%, and an alpha value of 5%. The normality of the data was checked with the Shapiro-Wilk test. Data are reported as mean ± standard deviation (SD) or median (25th and 75th percentiles) for parametric and non-parametric data as appropriate. Differences in demographic data among groups were analyzed using the 1-way analysis of variance (ANOVA), Kruskal-Wallis test, and Pearson’s Chi-square test.

As the CRP and iron levels were not normally distributed, a logarithmic transformation was applied to normalize the distribution before statistical analysis. A repeated measures (time) ANOVA with 2 factors (treatment × surgery/gender) was carried out. The within subjects Bonferroni correction post-hoc test was used to compare the studied times. The CMPS-SF scores were compared using the Kruskal-Wallis test, followed by a multiple pairwise Mann-Whitney U-test. This test was not done when rescue analgesia was provided. The number of dogs requiring rescue analgesia was compared among treatments with Fisher’s exact test. Differences were considered significant when P < 0.05. Statistical analyses were carried out with commercially available software (SPSS version 22; IBM, Armonk, New York, USA).

Results

A total of 60 dogs was enrolled in the study and none were excluded or failed to complete the study. At the end of the study, a female dog (#9) mistakenly allocated to PG was administered robenacoxib by the investigator in charge. This dog was not excluded but was assigned to RG. Since 10 dogs were enrolled per treatment group and a minimum of 9 dogs were involved in the study, no further dogs were required. Therefore, the total number of dogs was as follows: 11 females/10 males in robenacoxib, 9 females/10 males in saline, and 10 females/10 males in meloxicam. The postoperative adjusted dose of robenacoxib was 2.9 ± 1.0 mg/kg BW.

There were no differences in age, weight, packed cell volume (PCV), creatinine, urea, and albumin levels, breed distribution, time from premedication to start of surgery, and surgery time among the 3 groups (Table I). The same anesthetist carried out all procedures, managed them in the same way, and maintained a similar depth of anesthesia and cardiovascular stability throughout the procedure. No adverse events were recorded in any of the groups throughout the observation period.

Table I.

Data from 60 dogs undergoing gonadectomy and treated with either robenacoxib (n = 21), meloxicam (n = 20), or saline (n = 19). Data are expressed as mean ± standard deviation or median (25th and 75th percentiles).

	Treatment group
Variable (unit)	Saline	Robenacoxib	Meloxicam	P-value
Body weight (kg)	19 (12 to 25)	15 (8 to 23)	24 (12 to 29)	0.196
Age (mo)	18 (12 to 48)	24 (12 to 48)	18 (12 to 33)	0.493
Temperature
Baseline (°C)	38.4 ± 0.5	38.3 ± 0.8	38.2 ± 0.7	0.756
Post-surgery (°C)	37.5 ± 0.9	37.6 ± 0.8	37.3 ± 0.8	0.736
Packed cell volume
Baseline (%)	44 ± 8	45 ± 6	44 ± 5	0.838
7 d post-surgery (%)	46 ± 7	44 ± 6	44 ± 5	0.772
Urea
Baseline (mg/dL)	37.2 (30.4 to 40.7)	35.5 (28.4 to 45.3)	33.1 (29.9 to 46.5)	0.962
7 d post-surgery (mg/dL)	51.5 (38.4 to 61.1)	47.8 (35.7 to 65.5)	44.0 (34.8 to 61.8)	0.870
Creatinine
Baseline (mg/dL)	1.0 (0.8 to 1.1)	1.0 (0.9 to 1.2)	0.9 (0.8 to 1.2)	0.724
7 d post-surgery (mg/dL)	1.0 (0.9 to 1.4)	1.0 (0.9 to 1.2)	1.1 (0.8 to 1.3)	0.844
Albumin
Baseline (g/dL)	3.6 ± 0.3	3.7 ± 0.4	3.6 ± 0.2	0.506
7 d post-surgery (g/dL)	3.5 ± 0.4	3.4 ± 0.3	3.4 ± 0.2	0.815
Times
From premedication to surgery (min)	45 (39 to 49)	45 (42 to 48)	45 (42 to 46)	0.681
Surgery time (min)	20 (9 to 33)	18 (11 to 30)	16 (11 to 34)	0.915

C-reactive protein

Serum CRP increased postoperatively compared to baseline, peaking 24 to 48 h postoperatively and returning to baseline values after 7 d (P > 0.05) in all treatment groups. There were no differences in changes in serum CRP among surgeries (P = 0.823) or treatment groups (P = 0.411) at any time point (Figure 1).

Figure 1.

Mean ± SEM concentration of CRP from 60 dogs (30 females and 30 males) treated with either robenacoxib (n = 21), meloxicam (n = 20), or saline (n = 19) at each time point [0 (baseline), 12, 24, 48, 72, and 168 h] after gonadectomy. C-reactive protein (CRP) levels were not significantly different among gender or treatment.

^a Significant differences in saline compared with baseline value (P < 0.05).

^b Significant differences in robenacoxib compared with baseline value (P < 0.05).

^c Significant differences in meloxicam compared with baseline value (P < 0.05).

Iron

Serum iron levels decreased postoperatively from baseline, reaching a minimum 12 to 24 h postoperatively (Figure 2). After a week, serum iron levels returned to baseline values. There were no differences in serum iron changes among surgeries (P = 0.351) or treatment groups (P = 0.450).

Figure 2.

Mean ± SEM concentration of iron from 60 dogs (30 females and 30 males) treated with either robenacoxib (n = 21), meloxicam (n = 20), or saline (n = 19) at each time point [0 (baseline), 12, 24, 48, 72, and 168 h] after gonadectomy. Iron levels were not significantly different among gender or treatment.

^a Significant differences in saline compared with baseline value (P < 0.05).

^b Significant differences in robenacoxib compared with baseline value (P < 0.05).

^c Significant differences in meloxicam compared with baseline value (P < 0.05).

Pain scores

The CMPS-SF scores are shown in Figure 3. Nine female dogs required rescue analgesia: 6, 2, and 1 in saline, robenacoxib, and meloxicam groups, respectively (saline versus robenacoxib and meloxicam was P < 0.05, both groups). In saline, 1 dog required 2 analgesic rescues (at 6 and 10 h) and another dog (same group), required 3 analgesic rescues (at 4, 9, and 13 h). There were significant differences in CMPS-SF between sexes at 2 h (P = 0.024), 4 h (P = 0.003), and 6 h (P = 0.048). Pain scores were increased in saline at 4 h (females) and 6 h (male) versus robenacoxib (P = 0.034 and 0.012, respectively) and meloxicam (P = 0.023 and 0.042, respectively). Preoperative CMPS-SF scores were zero in all cases.

Figure 3.

Pain scores determined with the short-form Glasgow Composite Measure Pain Scale (CMPS-SF) for female and male dogs undergoing gonadectomy and treated with either robenacoxib, meloxicam, or saline. Females: n-value per group = 11, 10, and 9 dogs in robenacoxib, meloxicam, and saline, respectively, unless otherwise stated (figures indicate the n-value per group and time point). Males: the n-value is always 10 in all groups. Values are expressed as median ± interquartile range (IQR).

^a Significant differences in saline compared with robenacoxib and meloxicam (P < 0.05).

Discussion

The perioperative administration of either robenacoxib or meloxicam had no relevant effects on serum CRP or iron levels in response to tissue injury after gonadectomy in male and female dogs.

In humans, a higher concentration of CRP has been reported in women than in men (16). This effect is mainly due to differences in the distribution of subcutaneous (SQ) and abdominal fat between sexes, with a greater accumulation of subcutaneous fat in women (16). Such a difference has not been observed in dogs, in which only ageing has been associated with an increased abdominal-to-SQ-fat ratio (17). Also differences in CRP levels between sexes have not been demonstrated in dogs (18).

A previous report has already shown that meloxicam and carprofen do not affect CRP serum levels in dogs undergoing elective ovariohysterectomy (8). Our results confirmed this lack of effect of meloxicam, but also of robenacoxib. C-reactive protein is released by the liver, induced by cytokines such as interleukin-6 (IL-6) and tumor necrosis factor alpha (TNF-α) (19), which in turn are produced after surgery. The anti-inflammatory and analgesic effects of NSAIDs are mediated by blocking the cyclooxygenase (COX) enzymes, thus reducing release of prostaglandin (20). Therefore, although both mechanisms mediate the inflammatory response, they are not necessarily linked and surgically induced cytokine release is not affected by concurrent administration of NSAIDs (21), which would explain the lack of effect on CRP.

In humans, however, administering flurbiprofen to patients undergoing thoracotomy showed reduced IL-6 and CRP levels 4 d postoperatively (9). Since CRP peaked at 24 to 72 h in humans, it remains unclear whether flurbiprofen might have influenced the acute phase response triggered by surgery. In dogs undergoing orthopedic surgery, elevated CRP levels beyond the fourth postoperative day have been associated with the presence of complications (22).

Serum CRP values also rapidly increase after surgery in dogs, peaking 24 h postoperatively (1,2,5,8). The CRP response is thought to match the degree of surgical trauma (6,23). The observed increase in CRP produced by gonadectomy using a vessel-sealing device (2.4- to 4-fold) was similar to that reported after laparoscopic procedures (1.5- to 5-fold) in dogs (6,23,24), whereas a higher increase has been observed after conventional surgical techniques (14- to 37-fold) (1,2,5,7). However, no differences in CRP levels have been demonstrated when different surgical procedures were compared, such as hemilaminectomy versus ovariohysterectomy (25) or ovariohysterectomy versus ovariectomy (7). These differences in CRP levels do not provide a reliable quantitative trauma value (25).

The inclusion of a surgical control group using the standard technique might have helped to determine the relationship between increased CRP serum values and surgical trauma in the present study. The increase in CRP levels after conventional ovariectomy (7) may be reduced by 2/3 when a vessel-sealing device is used (26). The latter technique allows the procedure to be carried out through a small-to-medium celiotomy, thus avoiding clamping the genital tract and requiring less time. As no ligatures are necessary, no suturing material is left in the abdominal cavity. In humans, a higher inflammatory response was shown after hernia surgery using mesh repair than using non-mesh repair (27).

The time required to return to baseline CRP values postoperatively remains unclear and may be 3 to 5 d (23,24) or up to 14 d (28) or more. Some studies have focused on the first 2 postoperative days, likely assuming the most relevant changes occurred during this time (6,7,25,29). Some reports suggest that CRP levels may be increased for longer periods of time despite no clinical signs of postoperative complications (2,5). A low profile but persistent inflammatory process may be responsible for maintaining these increased CRP levels when conventional techniques are carried out using suturing material such as silk (30), but may rapidly return to baseline values with less aggressive surgical techniques (26).

Serum iron decreased after gonadectomy regardless of NSAID treatment. Even minor surgery has been shown to have an impact on iron metabolism, which suggests that blood loss is not a relevant issue in determining changes in iron levels in blood (31). The inflammation-induced hypoferremia is thought to be a consequence of a cytokine-driven increase in hepcidin, which blocks iron transfer from tissues to extracellular fluids by binding to ferroportin, the sole known cellular iron exporter (32). Thus, iron is retained by the liver and spleen, while iron absorption from the duodenum is decreased (32). The increase of IL-6 is the main cause of hepcidin increases in response to an inflammatory state (33). Iron levels in blood may therefore be considered a suitable marker of inflammatory response in dogs (34), with low levels described in dogs with pyometra (35) and after ovariohysterectomy (7).

In humans, reduced serum iron concentrations may last up to 10 d after minor surgical procedures (laparoscopy, arthroscopy, and shoulder arthroplasty) or more than 30 d after major surgery (laminectomy and hip or knee replacement) (31). Our results suggest that such recovery may last no longer than 7 d in dogs undergoing gonadectomy. Although several studies have focused on evaluating iron levels as an indicator of inflammation in dogs (34,35), only 1 study evaluates the effect of ovariohysterectomy on iron levels over 24 h (7). Further studies with a longer timeline are needed to confirm the duration of reduced postoperative iron levels in dogs.

Dogs receiving robenacoxib or meloxicam required less rescue analgesia postoperatively than those administered a placebo rather than an analgesic. Robenacoxib is considered to be at least as effective as meloxicam for managing perioperative pain in soft tissue or orthopedic surgery (36,37).

An important limitation of this study was the fact that cytokine IL-6 was not assessed, which would have allowed us to determine if a lack of response of CRP was due to the inability of NSAIDs to block cytokine release. The method used for measuring CRP has some limitations, with several methods developed for canine CRP and polyclonal human antibody-based tests that may increase variability (38), as was found in this study. When these latter tests are used, the measured value is given in human CRP equivalents, which are different from canine CRP values. Furthermore, the concentration of the heterologous protein will often be proportionally underestimated at high concentrations (39), but not at lower concentrations as those found in the present study. Therefore, comparison of CRP levels with the results of other studies that used a different method should be done cautiously.

In conclusion, similar changes in serum CRP and iron levels after ovariohysterectomy and orchiectomy suggest that there is little difference in terms of acute-phase response triggered by both surgical stimuli. This is probably due to limited surgical damage derived from the use of a vessel-sealing device. The postoperative period showed increased pain in females and thus, between the 2 surgical techniques. Despite the effectiveness of the studied NSAIDS in controlling postoperative pain, the use of perioperative robenacoxib and meloxicam did not influence postoperative CRP or iron levels. This suggests that these drugs do not have a relevant effect on these 2 biomarkers of the acute-phase response triggered by gonadectomy.