This was a prospective observational cohort study of client-owned dogs with sepsis admitted to the Cornell University Hospital for Animals. Dogs were eligible for enrollment if they weighed >5 kg, had a documented clinical syndrome associated with systemic infection (such as pyometra, septic peritonitis or pneumonia) and satisfied ≥2 systemic inflammatory response syndrome (SIRS) criteria, specifically: hypo- or hyper-thermia, temperature <37.8 or >39.4°C (<100.0 or >102.9°F); tachycardia, heart rate >140 bpm; tachypnea, respiratory rate >20 bpm; leukopenia or leukocytosis, <6 × 10³/μL or >16 × 10³/μL or >3% band neutrophils (3, 40). When bacterial cultures were not performed, or were negative, sepsis was confirmed by alternative means e.g., confirmation of gastrointestinal content leakage, surgical lesion identification, or determination through radiographic criteria. Dogs were ineligible if they had sepsis due to viral disease e.g., parvovirus or fungal disease e.g., candidiasis. To protect enrolled animals from potential complications associated with venipuncture, dogs with severe anemia, coagulopathy, or thrombocytopenia (Hb <5 g/dL; PT or aPTT >150 % normal; platelets <30 × 10³/μL) were excluded. Dogs not expected to live more than 12 h were also ineligible. Dogs were enrolled with written informed client consent. The local Institutional Animal Care and Use Committee approved the study protocol (Cornell IACUC Protocol #2014-0053). Healthy dogs were recruited from staff-owned pets and were eligible for the study if they weighed >5 kg, were aged between 1 and 9 y, had no chronic or recent illness, and had received no medications other than preventative healthcare (e.g., parasiticides) in the preceding 3 months. Dogs were classified as healthy based on history, physical examination, and the results of complete blood count and serum biochemistry profile results.

Primary clinicians determined all aspects of case management. Signalment and physical examination findings at hospital admission were recorded. Blood gases, electrolytes and lactate concentrations were measured immediately after sample collection with a point-of-care device (RapidPoint 500, Siemens Healthcare, Malvern, PA). Complete blood counts (CBC) (ADVIA 2120, Siemens Healthcare) with clinical pathologist review and serum biochemistry profiles (Cobas C501, Roche Diagnostics, Indianapolis, IN) were analyzed immediately whenever possible, but always within 48 h of collection. Mentation score, blood glucose, albumin and lactate concentrations and platelet counts were used to calculate the acute patient physiologic and laboratory evaluation illness severity score (APPLE_fast) (41, 42). Outcome status at discharge was recorded as survived, died, or euthanized. Blood samples were collected at study entry, and prior to administration of any antithrombotic medications, into evacuated tubes (Vacutainer, BD and Co, Franklin Lakes, NJ) containing no-additive (for serum biochemistry analyses), 3.2% sodium citrate (1:9 ratio) (for coagulation testing) and K₂-EDTA (for complete blood counts). Citrate plasma was prepared from whole blood by centrifugation for 10 min at 1,370 g (Ultra-8V Centrifuge, LW Scientific, Lawrenceville, GA). Plasma was transferred into polypropylene freezer tubes (Polypropylene Screw-Cap Microcentrifuge Tubes, VWR, Radnor, PA) with some plasma deliberately left in each tube to minimize the risk of cell contamination and frozen at −80°C pending batch analysis.

Determination of antiplasmin activity (AP), antithrombin activity (AT), D-dimer concentration, fibrinogen concentration (Clauss fibrinogen) and plasminogen activity was performed using an automated instrument with mechanical and spectrophotometric endpoint detection modes (STA Compact Max, Diagnostica Stago, Parsippany, NJ). Plasma AP and AT were measured with the manufacturer's synthetic chromogenic substrate kits (Stachrom Antiplasmin and Stachrom AT III, Diagnostica Stago). The AP assay is configured with a plasmin substrate and human plasmin reagent such that residual cleavage of the substrate is inversely proportional to AP in the test plasma. The AP standard curve was derived from a human plasma standard, with results reported as percentage of the human standard. The AT assay was modified using a pooled canine plasma standard (prepared at the Coagulation Laboratory from 20 healthy dogs). The pooled canine plasma had an assigned value of 100% AT and results were reported as percentage of the canine standard. Plasma D-dimer concentration was measured using a quantitative, turbidimetric immunoassay and the manufacturer's human D-dimer calibration standard (HemosIL D-dimer and Calibrator, Instrumentation Laboratory, Lexington, MA). Fibrinogen concentrations were measured via the Clauss method using a human thrombin reagent (STA Fibrinogen 100 U/mL, Diagnostica Stago) and the canine plasma standard. The fibrinogen content of the plasma standard was determined by gravimetric method (43). Concentrations of D-dimer and fibrinogen were reported as ng/mL and mg/dL, respectively. Plasminogen activity was measured based on cleavage of a chromogenic plasmin substrate (S-2251, Diapharma, West Chester, OH) following sample incubation with urokinase (Prospec, East Brunswick, NJ), as previously described (44). The assay was modified by an initial acidification/neutralization step and results were reported as the percentage plasminogen activity of a pooled canine plasma standard with an assigned value of 100%.

Quantitation of TAFI activity was performed as previously reported, with minor modifications (18). Briefly, canine TAFI activity was analyzed using a commercial kinetic chromogenic assay kit (Pefakit TAFI, Pentafarm, Basel, Switzerland). The assay uses a thrombin-thrombomodulin complex reagent to activate TAFI in the test plasma. Activated TAFI then acts on a synthetic chromogenic TAFI substrate. Plasma samples were diluted 1:2 in 0.9% sodium chloride prior to analysis. Diluted plasma samples were combined with the thrombin-thrombomodulin reagent in a 1:10 ratio in a 96-well microtiter plate and incubated for 3 min at 37°C before addition of the synthetic substrate. Upon addition of the substrate, the absorbance at 405 nm of each well was monitored every 10 s for 5 min in an automated plate reader (Cytation 1, BioTek Agilent, Santa Clara, CA). All measurements were run in duplicate. The activity of TAFI was expressed as percent activity of a pooled human plasma provided as a calibrator.

Fibrin clot formation and lysis over time in patient plasma was evaluated in the OHP assay (38, 39). The assay was configured with paired reaction mixtures containing test plasma and thrombin to generate a coagulation curve, and test plasma with thrombin and tPA to generate a fibrinolysis curve (39, 45). The assay was performed as previously described (45). In brief, coagulation and lysis reactions were performed in flat-bottom 96-well microtiter plates containing seventy-five microliters of plasma and a buffer containing bovine alpha-thrombin (final concentration 0.05 U/mL). The lysis reactions also contained human recombinant tPA (final concentration 350 ng/mL). After addition of the buffer containing coagulation and fibrinolysis activators to the test plasma, absorbance at 405 nm was measured every minute for 60 min and plotted over time to visualize changes in turbidity related to fibrin formation and degradation. The parameter overall coagulation potential (OCP) was defined as the area under the coagulation curve. This parameter is a measure of the rate and amount of fibrin formation. The OHP parameter was defined as the area under the curve in the lysis reaction containing thrombin and tPA, and thus depends on both fibrin formation and fibrinolysis. The overall fibrinolysis potential (OFP) parameter is derived from the OCP and OHP values. The OFP is calculated as the relative difference in area between the coagulation and lysis curves: OFP% = (OCP – OHP)/OCP × 100.

Continuous data (e.g., dog characteristics, physical examination findings and clinicopathologic values) were assessed for normality using the D'Agostino Pearson test and appropriate descriptive statistics calculated. Comparisons between groups were performed using unpaired t-tests with Welch's correction for normally distributed data or the Mann-Whitney U test when data were non-parametric. Correlations between coagulation parameters were evaluated using Spearman's correlation coefficients, associated P-values and scatterplots. Strength of correlation was assessed as follows: <0.5 weak, 0.5–0.6 mild, 0.6–0.7 moderate, 0.7–0.8 strong, 0.8–0.9 very strong, 0.9–1.0 excellent. The overall pattern of coagulation values in dogs with sepsis were compared with those in healthy dogs using radar plots (Excel for Mac, Microsoft, Redmond, WA). Coagulation variable data for each dog was ratioed against the midpoint of the corresponding reference interval. The within-group means of these fold values corresponding to the degree of divergence from the reference interval were then overlaid on a hexagonal reference chart to allow visual comparison of dogs with sepsis with healthy controls. Using comparisons against the relevant reference interval, the coagulation disturbances in each dog with sepsis was classified as hyperfibrinolytic, hypofibrinolytic, mixed disturbance, or no disorder. Dogs satisfying 3/5 of the following criteria: low TAFI, low fibrinogen, high D-dimer, low AP, and low plasminogen were classified as hyperfibrinolytic, most consistent with a bleeding risk. Dogs satisfying 3/5 of the following criteria: high TAFI, high fibrinogen, high AP, low AT, and high plasminogen were classified as hypofibrinolytic, most consistent with a thrombotic risk. Dogs with disturbances characteristic of more than one type of disturbance were categorized as having a mixed disorder. Dogs with 4/6 parameters within the reference interval were classified as no disorder. The association between coagulation status classification (hypofibrinolytic versus other) and outcome was assessed with Fisher's exact test. Statistical analyses were performed using commercial software (Prism 9 for macOS, GraphPad, La Jolla, CA) with alpha set at 0.05. No post-hoc corrections were made for multiple comparisons because all between group comparisons were based on a priori hypotheses.

A total of 40 dogs were enrolled; 20 dogs with sepsis and 20 healthy controls. The 20 dogs with sepsis had a variety of different diseases, specifically four dogs had abscesses or cellulitis, three dogs had peritonitis, three dogs had pneumonia, three dogs had pyometra, and two dogs had mastitis. Other causes included anaplasmosis, gastroenteritis (with bacteremia), osteomyelitis, pyothorax and urosepsis (all n = 1). Of the 20 dogs, three were euthanized for disease severity prior to discharge, the remainder survived to discharge, equivalent to a 15% case fatality rate. Of the 17 dogs that survived to hospital discharge, 16 dogs were alive at day 28, with 1 dog lost to follow up, equivalent to a 16% 28-day case fatality rate. Demographic characteristics, initial assessments and clinicopathologic variables are summarized in Table 1. Dogs had been treated with a variety of medications prior to study enrolment, summarized in Table 2. Positive cultures were obtained in 64% (9/14) dogs for which culture samples were submitted. Bacterial organisms cultured from the dogs included Escherichia coli (n = 4), Staphylococcus pseudintermedius (n = 2), Actinomyces canis, Bacteroides spp., Clostridium perfringens, Enterococcus faecium, Fusobacterium sp., Microbacterium phyllosphaerae, Mycoplasma sp., Peptostreptococcus sp., and Pseudarthrobacter sp. (all n = 1).

All coagulation and fibrinolysis test results are summarized in Table 3. Compared to healthy controls, dogs with sepsis had significantly lower AT, higher AP and TAFI, and higher concentrations of clottable fibrinogen and D-dimer. No difference in plasminogen activity was observed between the two groups (Figure 1). Control samples on two OHP assay plates generated inappropriately low OHP values, indicating an assay error. Study samples quantitated on these plates were excluded from subsequent analyses and lack of sample volume precluded repeating these tests. Comparison of OHP assay results for the remaining septic dogs (n = 9) and controls (n = 10), revealed that septic dogs had significantly greater OHP, OCP and smaller OFP compared to healthy controls (Figure 2). Only two pairs of variables were associated with a Spearman correlation coefficient >0.7 (strong or better). The activity of the fibrinolysis inhibitor, TAFI, was negatively correlated with the extent of fibrinolysis, measured as % OFP (r_s −0.817, P = 0.011, Figure 3A), and the OCP was positively correlated with illness severity as assessed by APPLE_fast score (r_s 0.729, P = 0.033, Figure 3B). A single graphical display of the 6 individual tests revealed that fibrinogen and D-dimer concentrations in the septic dog group demonstrated the most profound deviations from reference intervals (Figure 4). No significant differences between the test values of septic dog survivors and non-survivors were observed (Supplementary material S1). Similarly, none of the coagulation variables were significantly associated with duration of hospitalization. Classification of dogs with sepsis based on the number of test parameters outside of their corresponding reference intervals suggested that seven dogs had a mixed disorder, seven dogs had no disorder, five dogs appeared hypofibrinolytic, and one dog appeared hyperfibrinolytic. There was no association between hypofibrinolytic status and non-survival to hospital discharge, relative risk 1.56 (0.97–4.07), P = 0.140.