Comparison of Clinical and In ammatory Parameters in Dogs with Pyometra Before and After Ovariohysterectomy

This study aimed to identify potential biomarkers of canine pyometra and their correlations with clinical parameters. First, 90 dogs with pyometra and 26 healthy female dogs were compared. Then, paired samples (before and after ovariohysterectomy) from 22 dogs with pyometra and 9 healthy controls from the initial cohort were compared. Concentrations of acute inflammatory proteins, C-reactive protein (CRP) and serum amyloid A (SAA), and cell-free DNA (cfDNA), were significantly higher in dogs with pyometra than in clinically healthy dogs. Cell-free DNA was the most sensitive biomarker for systemic inflammation, based on the receiver operating characteristic curve analysis (area under the curve = 0.959). In addition, cfDNA and CRP were significantly associated with inflammation and organ injury-related clinical parameters. Following the surgical removal of the inflamed uterus, interleukin-6 (IL-6), high-mobility group box 1 (HMGB1), and procalcitonin (PCT) significantly decreased, whereas changes in CRP, SAA, and cfDNA were not significant. These findings indicate that cfDNA, CRP, and SAA are potential clinical biomarkers of systemic inflammation in dogs with pyometra and PCT, IL-6, and HMGB1 are potential biomarkers of clinical recovery.


Background
Pyometra is characterised by the accumulation of uterine pus and may be compounded by opportunistic bacterial infections. Pyometra can result in systemic in ammatory response syndrome (SIRS), as well as sepsis and organ damage, and therefore, dogs with pyometra are used as a natural model of sepsis [1]. Surgical treatment by ovariohysterectomy (OHE) removes the source of in ammation. It could thus be used as a model for therapeutic monitoring of canine sepsis and to identify diagnostic biomarkers to better understand the pathophysiology of both human and veterinary sepsis.
The objective of this study was to identify potential biomarkers for screening and therapeutic monitoring of naturally occurring canine sepsis caused by pyometra, based on the differences in clinical and in ammatory parameters between female dogs with pyometra and healthy ones and on the changes in these parameters after surgical correction of pyometra.

Comparison of clinical data and in ammatory parameters between pyometra and healthy dogs
The clinical data, including haematologic and biochemical data, showed signi cant differences between the subjects with pyometra and healthy controls (Table 1). Among the haematologic variables, the mean leukocyte, neutrophil, and monocyte counts were signi cantly higher (P < 0.01), while erythrocyte counts, haematocrit levels, and haemoglobin concentrations were signi cantly lower (P < 0.01) in the dogs with pyometra than in the healthy subjects. The total protein, globulin, alanine aminotransferase (ALT; P < 0.05), and alkaline phosphatase (ALP; P < 0.01) concentrations were signi cantly higher in the dogs with pyometra than in the healthy ones.
The concentrations of in ammatory parameters in the subjects with pyometra and healthy dogs are presented in Fig. 1. In the dogs with pyometra, the serum CRP, SAA, and cfDNA concentrations were signi cantly higher (P < 0.01) than those in the healthy subjects. However, the IL-6, PCT, and HMGB1 concentrations did not show signi cant differences.
Using ROC curve analysis to evaluate the usefulness of each biomarker for discrimination between the subjects with pyometra and the healthy ones (Fig. 2), cfDNA was found to be the best predictive biomarker (area under the curve = 0.959), with a combined sensitivity of 97% and a speci city of 80% for differentiating dogs with pyometra, followed by plasma CRP, with a combined sensitivity of 88% and a speci city of 100%.
Additional categories were assessed to obtain more detail about the differences in in ammatory parameters between subgroups of the dogs with pyometra that were clinically worse (above or below the reference intervals for clinicopathologic parameters) (Table 3). Among in ammatory parameters, differences were detected in cfDNA and CRP, and both biomarkers were associated with several signi cant in ammation-related clinical parameters 2.3. Comparison of clinical data and in ammatory parameters depending on surgical correction After surgical correction, dramatic changes were observed in the clinical data (Table 4). In the dogs with pyometra, the post-OHE leucocyte and neutrophil counts were signi cantly higher (P < 0.01), whereas erythrocytes, haematocrit, and haemoglobin were signi cantly lower (P < 0.01) than the respective pre-OHE levels. In the biochemical panel, the post-OHE total protein, globulin, cholesterol, pH, and potassium concentration values were signi cantly lower (P < 0.01) than the pre-OHE ones, as were the glucose, BUN, creatinine, albumin, ALP, and ionised calcium concentrations (P < 0.05).
There were no differences between the pre-and post-OHE values of CRP, SAA, and cfDNA in the dogs with pyometra (Fig. 3).

Discussion
In this study, we investigated the clinical data and in ammatory parameters in dogs with pyometra at presentation, before and after surgical correction. Increases in the WBC counts, ALT and ALP activities, and concentrations of cholesterol and amylase and decreases in the RBC counts, haematocrit, hyperglobulinaemia, and hypoalbuminaemia were statistically signi cant in the dogs with pyometra compared to healthy dogs. Furthermore, signi cantly increased CRP, SAA, and cfDNA concentrations were found in dogs with pyometra compared to healthy dogs. After surgical correction of pyometra via OHE, the WBC and RBC counts, cholesterol, albumin, globulin concentrations, and ALP activity, as well as the IL-6, HMGB1, and PCT levels signi cantly decreased compared to pre-OHE.
In the dogs with pyometra, leucocytes and especially neutrophils and monocytes were signi cantly increased, whereas the RBC counts and haematocrit were decreased. Myeloid hyperplasia, especially neutrophilia, contributes to leucocytosis, and decreased iron utilisation in myeloid cells of the bone marrow by the APPs lactoferrin and hepcidin in response to systemic in ammation can reduce erythropoiesis [13][14][15]. Hyperproteinaemia, hyperglobulinaemia, and hypoalbuminaemia were also statistically signi cant in the dogs with pyometra vs. healthy controls, owing to an increased synthesis of APPs and antibodies in response to bacterial infection and in ammation. Decreased albumin levels may suggest a negative APP function [16] and are commonly associated with an in ammatory disease.
Organ damage is common during systemic in ammation in dogs [17]. There were signi cant increases in ALT, ALP activities, and cholesterol concentration in the subjects with pyometra in this study. Increased liver enzyme activities and cholesterol concentrations were also observed in intrahepatic cholestasis, which is associated with endotoxaemia and hepatocellular damage caused by diminished circulation and cellular hypoxia [16]. Renal injuries could be found in canine endotoxemia [18], and increases in creatinine concentrations, although not statistically signi cant, were found in this study. A decreased glomerular ltration rate due to endotoxaemia, may contribute to increases in amylase and lipase levels. Endotoxaemia may contribute to renal injury, but disease severity may affect the results. All dogs with pyometra survived in this study, which may indicate a lower disease severity, and therefore renal damage that might be statistically insigni cant. Dogs with vaginal discharge were selected in this study, which could help with an early diagnosis before septic insult. Surgical correction decreased the RBC indices, total protein, albumin, and globulin levels, while increasing the sodium and chloride levels. Fluid administration during a surgical procedure may result in haemodilution and contribute to changes in electrolytes.
The CRP and SAA concentrations were signi cantly higher in the dogs with pyometra than in the healthy ones. CRP and SAA are the major APPs in dogs, and these results are consistent with those of previous reports on dogs with SIRS and pyometra [3] or chronic in ammation [19][20][21]. We also found a strong positive correlation between CRP and SAA. Thus, combined monitoring of the CRP and SAA biomarkers may be clinically advantageous for diagnosing in ammation.
However, cfDNA was the most discriminative among the in ammatory parameters tested, based on the ROC curve analysis. Recent studies have suggested that the presence of cfDNA is due to necrosis, apoptosis, and the release of neutrophil extracellular traps (NETs), which are networks of extracellular bres, primarily composed of DNA from neutrophils and binding pathogens. These bres are released as a result of an innate immune mechanism initiated by neutrophils [22,23], and thus, higher concentrations of cfDNA in dogs with pyometra may be due to tissue damage, resulting from systemic in ammation and NET formation. Besides, we found that the cfDNA concentrations, unlike the CRP and SAA levels, were correlated with the WBC counts, one of the SIRS criteria. The elevation of WBCs, especially neutrophils, is a typical sign of in ammation, suggesting that serum cfDNA concentrations may re ect in ammatory processes in dogs with sepsis.
In the dogs with pyometra, abnormal clinical parameters were strongly correlated with the cfDNA, CRP, and SAA concentrations. Correlations between the concentrations of these parameters and disease severity have been previously reported for CRP, SAA [24], and cfDNA [25]; however, there were no signi cant differences in these parameters before and after surgical correction. Moreover, the CRP level was the only in ammatory parameter that increased after surgery in healthy dogs, consistently with previous studies [12,19,26]. Surgical insult such as a laparotomy could affect the CRP concentration, even in healthy dogs, making CRP an unspeci c biomarker for therapeutic monitoring. A more representative controlled study should be performed in the future to clarify the signi cance of this parameter.
When the infection source is removed, the concentrations of several biomarkers should decrease relative to those of endotoxins. Thus, we considered changes in the concentrations of in ammatory parameters as biomarkers for therapeutic monitoring. The study showed that most dogs with pyometra had lower concentrations of all cytokines following surgical treatment. These ndings agree with the results of the comparison between dogs with pyometra and healthy ones. Although not all in ammatory parameters were fully accounted for by the treatment effects, IL-6, HMGB1, and PCT emerged as potentially relevant biomarkers for therapeutic monitoring. The baseline concentrations of these molecules were not signi cantly higher in the dogs with pyometra than in the healthy controls but decreased following the removal of the in amed uterus, thus indicating their usefulness in the postoperative period in dogs with pyometra [9]. IL-6 and HMGB1 levels usually increase in canine endotoxaemia and systemic in ammation [5,7]; however, changes in their levels were not statistically signi cant in our study. IL-6 is a major proin ammatory cytokine released during an early phase of endotoxaemia [17], whereas HMGB1 is a late mediator, released by damaged or killed cells [27]. Because the analysed pyometra cases did not develop severe sepsis or organ damage, the peak concentrations of these in ammatory parameters might not have been persistent, or no peak levels were reached. However, the concentrations did signi cantly drop after surgical correction, and thus, IL-6 and HMGB1 could be useful biomarkers for therapeutic monitoring.
Both human and animal studies have suggested that an increase in the PCT concentration may be associated with serious infections and endotoxaemia [28,29] and demonstrated that PCT acted as an APP in dogs. In humans, PCT is used for monitoring during treatment and as a prognostic marker [30]. Although PCT did not discriminate subjects with pyometra from healthy dogs, its concentration was signi cantly lower after surgical correction. Moreover, there were signi cant correlations between PCT levels and several types of clinical data, such as haematocrit, monocyte counts, BUN, total calcium, albumin, globulin, and ALP concentrations. Thus, application of PCT in clinical practice as a relevant surrogate biomarker for organ dysfunction and the severity of disease may be suggested.
Our study has several potential limitations, one of which is that the identi cation and assessment of the severity of pyometra were subjective. Additionally, the diagnosis of pyometra was con rmed based on gross ndings, making it di cult to distinguish sepsis and SIRS, as gross con rmation of infection may or may not correspond to systemic illness [31]. However, the high frequency of clinical signs, abnormalities in blood tests, and persistent suppurative purulence suggested a systemic in ammatory response as a possible explanation of abnormal clinical criteria. Finally, all subjects survived until discharge, and thus, we could not compare parameters between the survival and non-survival groups. Further studies are necessary to validate the prognostic indicators.

Conclusion
cfDNA, CRP, and SAA have the potential to be used as clinically appropriate biomarkers for screening systemic in ammation in canine pyometra, whereas PCT, IL-6, and HMGB1 may be useful biomarkers for the therapeutic monitoring.

Study design, ethical approval, and sampling
A prospective randomised in vitro analysis of several biomarkers of in ammation was based on continuous monitoring of blood samples before and after surgical OHE. The study was conducted with the approval of the Institutional Animal Care and Use Committee of Gyeongsang National University (approval number: IACUC GNU-190218-D0010).
Dogs with pyometra diagnosed during 2018-2019 were included in this study. In total, 116 female dogs of various breeds, weights, and ages were included, of which 90 cases had pyometra, and 26 cases were healthy controls. Con rmation of the diagnosis of pyometra was based on previously proposed de nitions [32,33]. The medical records of the dogs with pyometra, including the history, signalment, physical examination, clinical signs, and routine blood tests, were analysed.
Healthy, client-owned female dogs served as controls for comparison. Prior to surgery, dogs were classi ed as healthy based on unremarkable ndings in the history, physical examinations, and blood tests. Following surgery, consecutive samples from 22 dogs with pyometra and nine healthy controls were compared. Initial blood samples were collected at the time of presentation (pre-OHE), and then consecutive blood samples were collected within 3 days following surgical treatment (post-OHE). Samples were collected via venipuncture and stored in EDTA and lithium heparin tubes. The EDTA blood samples were analysed using an automatic haematology analyser. Heparinised plasma was stored at − 80 °C for analysis of in ammatory biomarkers.

Clinical and laboratory analyses
To assess organ injury and the overall clinical status of the dogs, acid-base balance, electrolyte concentrations, and serum biochemical parameters were analysed using a blood gas analyser (pHOx; Nova Biomedical, Waltham, MA, USA) and a Catalyst Dx chemistry analyser (IDEXX, Westbrook, ME, USA).
The measurement of cell-free DNA (cfDNA), previously validated for use in dogs [25], was performed using a benchtop uorimeter (Qubit 1.0; Life Sciences, Carlsbad, CA, USA) and the Qubit dsDNA HS assay kit, according to the manufacturer's instructions. were measured using commercial ELISA kits according to the manufacturers' instructions. The lower limits of detection were 7.5 ng/mL, 10 ng/mL, 62.5 pg/mL, 12.5 pg/mL, and 6.25 ng/mL, respectively. All biomarker measurements were performed in duplicate with appropriate controls.

Statistical analysis
All results are expressed as the mean ± SD. The signi cance of differences between the subjects with pyometra and healthy animals was calculated using the Mann-Whitney U-test. Linear correlation between two variables was tested using Pearson's correlation coe cient. For continuous variables, the Wilcoxon two-sample test was used to evaluate differences between the dogs with pyometra and healthy ones and between pre-and post-OHE subjects. Statistical signi cance was set at P < 0.05. The SPSS 25.0 software package (SPSS, Inc, Chicago, IL, USA) was used to perform all statistical analyses. All graphs were plotted using GraphPad Prism 8.0 (GraphPad Software, La Jolla, CA, USA). Abbreviations ALP, alkaline phosphatase; ALT, alanine aminotransferase; APP, acute-phase protein; BUN, blood urea nitrogen; cfDNA, cellfree DNA; CRP, C-reactive protein; HMGB1, high-mobility group box 1; NET, neutrophil extracellular trap; OHE, ovariohysterectomy; PCT, procalcitonin; ROC, receiver operating characteristic; SAA, serum amyloid A; SIRS, systemic in ammatory response syndrome; WBC, white blood cell.

Declarations Not applicable
Ethics approval and consent to participate The study was conducted with the approval of the Institutional Animal Care and Use Committee of Gyeongsang National University (approval number: IACUC GNU-190218-D0010), and written informed consents were obtained from all dog owners prior to sample collection.

Consent for publication
Not applicable.
Availability of data and materials Not applicable.

Competing Interests
The authors declare no con ict of interest.  Tables   Table 1 Mean variables for clinical data from pyometra and healthy female dogs.
Lactate (mmol/L) 2.16 ± 0.28 2.11 ± 0.31 0.869 All data are presented as mean ± SD. The P values represented the comparison the clinical data between the pyometra and healthy dogs.
Lactate (mmol/L) 2.52 ± 0.55 3.10 ± 1.10 0.655 All data are presented as mean ± SD. The P values represented the comparison the clinical data between the pre-and post-OHE pyometra dogs.