Time-action profiles of insulin detemir in normal and diabetic dogs

Insulin detemir is the first member of a new class of long-acting soluble insulin analogues capable of maintaining the basal level of insulin in humans. In this preliminary study, we investigated the time-action profiles of insulin detemir in normal and diabetic dogs since the use of insulin detemir in canines has yet to be determined. Eight animals were used in our study (three normal and five insulin dependent diabetic dogs). Time-action profiles of insulin detemir were monitored in normal dogs using an artificial pancreas apparatus under euglycemic condition. Blood sampling was performed at 2 h intervals post feeding, with insulin administration, in insulin dependent diabetic dogs. Time-action profiles of insulin detemir, in normal dogs, demonstrated that insulin detemir is a long-lasting preparation similar to what has been observed in humans. A pronounced peak was detected at 8–10 h while the glucose-lowering effect lasted for over 24 h after insulin injection, thus illustrating its longer prolonged peak activity time. Furthermore, intensive glycemic control was achieved with insulin detemir in insulin dependent diabetic dogs, using a lower dosage than NPH insulin and insulin glargine therapeutic doses. Our results indicate that insulin detemir has a greater effect than either NPH insulin or insulin glargine in canines, requiring a lower dose than either insulin preparation. However, using insulin detemir also carries a higher risk of inducing hypoglycemia as compared to either NPH insulin or insulin glargine.     

Comparison of time-action profiles of insulin Glargine and NPH insulin in normal and diabetic dogs

Intermediate insulin injections are commonly used for glycemic control in insulin dependent diabetic dogs acting as a replacement for natural insulin. Neutral Protamin Hagedorn (NPH) insulin and insulin glargine are two types of injectable insulin preparations commonly used in humans. In our study, we investigated the time-action profiles of both aforementioned insulin preparations in normal dogs in order to determine whether co-administration of NPH and glargine would be of benefit to insulin dependent diabetic dogs as it is for humans suffering from insulin dependent diabetes. Time-action profiles of NPH insulin and insulin glargine in normal dogs demonstrated a clear difference between both insulin preparations confirming that NPH insulin is an intermediate-acting preparation whereas insulin glargine is a long-lasting preparation. In addition, co-administration of NPH insulin and insulin glargine resulted in tight glycemic control as compared to NPH insulin alone in insulin dependent diabetic dogs. However, co-administration result in hypoglycemia at the dosages tested.     

Effects of chronic obesity and weight loss on plasma ghrelin and leptin concentrations in dogs

The objective of this study was to evaluate, in dogs, the effects of obesity and weight loss on plasma total ghrelin and leptin concentrations. Twenty-four Beagle dogs, 12 control lean and 12 obese dogs of both genders and aged between 1 and 9 years, were used for the experiments. Mean body weight was 12.7+/-0.7 kg for the lean group and 21.9+/-0.8 kg for the obese group. The trial was divided into three phases. During phase 1, all 24 Beagle dogs were fed a maintenance diet. During phase 2, the obese dogs were submitted to a weight loss protocol with a high protein-low energy diet. The weight loss protocol ended once dogs reached optimal body weight. During phase 3, the dogs that were submitted to the weight loss protocol were maintained at their optimal body weight for 6 months. Plasma total ghrelin, leptin, insulin and glucose concentrations were measured to evaluate the effects of obesity and weight loss on these parameters in dogs. Body weight, body condition score, thoracic and pelvic perimeters, and ingested food amounts were also recorded during the study. Obese dogs demonstrated a significant decrease in plasma ghrelin and a significant increase in plasma leptin and insulin concentrations when compared with control dogs. During weight loss, significant increases in plasma total ghrelin and glucose and significant decreases in plasma leptin and insulin were observed. The increase in plasma ghrelin concentrations seemed to be transient. Body weight and the morphometric parameters correlated positively with leptin concentrations and negatively with total ghrelin concentrations. These results suggest that ghrelin and leptin could play a role in dogs in the adaptation to a positive or negative energy balance, as observed in humans.     

Advanced glycation end products and sorbitol in blood from differently compensated diabetic dogs

Canine diabetes mellitus (DM) is a common metabolic disorder with long term complications, most of which are caused by glycosylation of structural proteins, decreases in antioxidant concentrations, altered osmotic balance and hypoxia due to impaired oxygen transport. Previous studies have demonstrated that under hyperglycemic conditions canine erythrocytes undergo swelling, probably due to activation of the polyol pathway. The present work aimed to assess the plasma concentration of advanced glycation end (AGE) products, stable Amadori-products generated by non-enzymatic glycosylation of proteins and the intracellular concentration of sorbitol, produced by the activation of polyol pathway in 34 blood samples from diabetic dogs and in 14 controls. AGE products were significantly higher (p<0.01) in plasma from diabetic dogs compared with control animals. The sorbitol concentration in erythrocytes was also significantly higher in diabetic dogs and, in particular, in poorly compensated animals and in dogs with ketonuria. In five cases that were analysed before and after clinical improvement, sorbitol concentration was found to correlate with improvement. These results suggest that non-specific glycosylation is increased and that the polyol pathway is activated in diabetic dogs in a manner that is proportionate to the severity of disease. Moreover, the concentration of AGE products and sorbitol may be useful for monitoring the onset of diabetic complications and assessing the most appropriate therapeutic approaches for management of canine DM.     

Effect of octreotide on plasma concentrations of glucose, insulin, glucagon, growth hormone, and cortisol in healthy dogs and dogs with insulinoma

The inhibitory effect of the somatostatin analogue octreotide on the secretion of insulin could be used in the treatment of insulinoma. However, current information on the effectiveness of octreotide in dogs is conflicting. Therefore, the endocrine effects of a single subcutaneous dose of 50 microg octreotide were studied in healthy dogs in the fasting state (n=7) and in dogs with insulinoma (n=12). Octreotide did not cause any adverse effects. In healthy dogs in the fasting state, both plasma insulin and glucagon concentrations declined significantly. Basal (non-pulse related) GH and ACTH concentrations were not affected. A slight but significant decrease in the plasma glucose concentrations occurred. Dogs with insulinoma had significantly higher baseline insulin concentrations and lower baseline glucose concentrations than healthy dogs in the fasting state. Plasma glucagon, GH, ACTH, and cortisol concentrations did not differ from those in healthy dogs. Baseline plasma insulin concentrations decreased significantly in dogs with insulinoma after octreotide administration, whereas plasma concentrations of glucagon, GH, ACTH, and cortisol did not change. In contrast to the effects in the healthy dogs, in the dogs with insulinoma plasma glucose concentrations increased. Thus, the consistent suppression of plasma insulin concentrations in dogs with insulinoma, in the absence of an suppressive effect on counter-regulatory hormones, suggests that further studies on the effectiveness of slow-release preparations in the long-term medical treatment of dogs with insulinoma are warranted.     

Comparison of whole blood and plasma potassium concentrations in dogs using the Reflovet system

Background: The Reflovet system is designed for chemical analysis of whole blood. However, plasma or serum is recommended for potassium analysis because of possible interference from RBC potassium. Because RBC potassium concentration is low in most canine erythrocytes, however, there should be little or no interference.
Objective: The objective of this study was to compare potassium results obtained in whole blood and in plasma from dogs using the Reflovet system.
Methods: Blood samples were collected from 104 dogs into lithium-heparin tubes. The potassium concentration was measured in whole blood, and subsequently the PCV was measured. Samples were centrifuged and the potassium concentration was measured in plasma. Comparisons were made using Deming's regression and Bland-Altman difference plots.
Results: There was very good correlation between results of potassium measurements in whole blood and plasma ( r = 0.93). Potassium values were moderately lower in whole blood: Potassium_blood= 0.912 × Potassium _plasma+ 0.119. Hemolysis had a negligible effect on the results, but the difference increased with the PCV value. In more than 90% of samples, the difference between the 2 measurements was ≤ 0.3 mmol/L.
Conclusion: There is only a negligible difference in most cases between potassium values in canine plasma and whole blood using the Reflovet system.     

Outcome of dogs with diabetic ketoacidosis: 127 dogs (1993-2003)

The aim of this study was to retrospectively describe the outcome of 127 dogs with naturally occurring diabetic ketoacidosis (DKA) and to examine the association between outcome of canine DKA and clinical and clinicopathologic findings. Eighty-two (65%) dogs were diagnosed with DKA at the time of initial diagnosis of diabetes mellitus (DM). Eighty-seven dogs (69%) had one or more concurrent disorders diagnosed at the time of hospitalization. Commonly identified concurrent conditions included acute pancreatitis (52, 41%), urinary tract infection (21, 20%), and hyperadrenocorticism (19, 15%). Dogs with coexisting hyperadrenocorticism were less likely to be discharged from the hospital (P = .029). Of 121 treated dogs, 89 dogs (70%) survived to be discharged from the hospital, with a median hospitalization of 6 days. Nonsurvivors had lower ionized calcium concentration (P < .001), lower hematocrit (P = .036), lower venous pH (P = .0058), and larger base deficit (P = .0066) than did survivors. Time from admission to initiation of subcutaneous insulin therapy was correlated with lower serum potassium concentration (P = .0056), lower serum phosphorus concentration (P = .0043), abnormally high white blood cell count (P = .0060), large base deficit (P = .0015), and low venous pH (P < .001). Multivariate analysis showed that base deficit was associated with outcome (P = .021). For each unit increase in the base deficit, there was a 9%) greater likelihood of discharge from the hospital. In conclusion, the majority of dogs with DKA were not previously diagnosed with DM. Concurrent conditions and electrolyte abnormalities are common in DKA and are associated with length of hospitalization. Survival was correlated to degree of anemia, hypocalcemia, and acidosis.     

Changes in fetal mannose and other carbohydrates induced by a maternal insulin infusion in pregnant sheep

Background

The importance of non-glucose carbohydrates, especially mannose and inositol, for normal development is increasingly recognized. Whether pregnancies complicated by abnormal glucose transfer to the fetus also affect the regulation of non-glucose carbohydrates is unknown. In pregnant sheep, maternal insulin infusions were used to reduce glucose supply to the fetus for both short (2-wk) and long (8-wk) durations to test the hypothesis that a maternal insulin infusion would suppress fetal mannose and inositol concentrations. We also used direct fetal insulin infusions (1-wk hyperinsulinemic-isoglycemic clamp) to determine the relative importance of fetal glucose and insulin for regulating non-glucose carbohydrates.

Results

A maternal insulin infusion resulted in lower maternal (50%, P < 0.01) and fetal (35-45%, P < 0.01) mannose concentrations, which were highly correlated (r² = 0.69, P < 0.01). A fetal insulin infusion resulted in a 50% reduction of fetal mannose (P < 0.05). Neither maternal nor fetal plasma inositol changed with exogenous insulin infusions. Additionally, maternal insulin infusion resulted in lower fetal sorbitol and fructose (P < 0.01).

Conclusions

Chronically decreased glucose supply to the fetus as well as fetal hyperinsulinemia both reduce fetal non-glucose carbohydrates. Given the role of these carbohydrates in protein glycosylation and lipid production, more research on their metabolism in pregnancies complicated by abnormal glucose metabolism is clearly warranted.     

Plasma aldosterone concentrations and plasma renin activity in healthy dogs and dogs with hyperadrenocorticism

The mean (se) basal plasma aldosterone concentrations were significantly lower in 31 dogs with pituitary-dependent hyperadrenocorticism (PDH) (75 [9] pmol/litre) than in 12 healthy dogs (118 [14] pmol/litre), whereas in five dogs with hyperadrenocorticism due to an adrenocortical tumour they were significantly higher (205 [109] pmol/litre). The mean basal renin activity was not significantly different between the dogs with PDH (303 [48] fmol/litre/second), the dogs with an adrenocortical tumour (141 [63] fmol/litre/second), and the control dogs (201 [25] fmol/litre/second). At three and four hours after the intravenous administration of 0.1 mg/kg dexamethasone, the concentrations of aldosterone decreased significantly to about 60 per cent of their initial values in the control dogs but did not change in the dogs with PDH or an adrenocortical tumour. In the dogs with PDH the renin activity increased significantly after the administration of dexamethasone.     

Plasma L-carnitine concentration in healthy dogs and dogs with hepatopathy

BACKGROUND: L-Carnitine has an essential role in lipid metabolism. Disturbances of L-carnitine metabolism can influence the energy supply of the organism. L-Carnitine is synthesized exclusively in the liver. Hence, we hypothesized that liver disease can influence L-carnitine metabolism. OBJECTIVES: The goal of this study was to compare plasma L-carnitine concentrations in dogs with different liver diseases of differing severity with the plasma L-carnitine concentrations of healthy dogs. METHODS: Sixteen dogs with inflammatory liver disease and 12 dogs with liver neoplasia were included in the study. Liver disease was diagnosed by clinical chemistry, ultrasonography, and histology of liver biopsy specimens. L-Carnitine concentration was measured in plasma samples using mass spectrometry, and compared among groups using unpaired Student's t-tests. RESULTS: Compared with healthy controls (24.4 +/- 8.4 micromol/L), the plasma L-carnitine concentration in dogs with liver disease (44.2 +/- 23.7 micromol/L) was significantly higher (P<.0001). The difference in L-carnitine concentration between dogs with moderate (n=8; 33.6 +/- 13.7 micromol/L) and severe (n=8; 57.4 +/- 22.9 micromol/L) hepatitis was also significant (P=.02). No difference in plasma L-carnitine concentration was found between dogs with hepatitis and those with liver tumors. CONCLUSIONS: Liver disease in dogs was accompanied by elevated plasma L-carnitine concentration. The severity of hepatitis appears to influence L-carnitine concentration.     

Effect of hemodialysis on plasma amino acid concentrations in healthy dogs

OBJECTIVE: To characterize the effect of maintenance hemodialysis on plasma amino acid concentrations and to quantitate free amino acid losses into the dialysate during hemodialysis in healthy dogs. ANIMALS: 8 healthy adult dogs. PROCEDURE: Five dogs received hemodialysis treatments 3 times per week for 4 weeks. Plasma amino acid concentrations were evaluated once per week for 4 weeks in each of the 5 dogs prior to hemodialysis (time 0), 90 minutes during hemodialysis, and immediately after hemodialysis (180 minutes). Total free amino acid concentrations and plasma amino acid concentrations (time 0, 90 minutes, and 180 minutes) in the dialysate were evaluated in 3 dogs that received 1 hemodialysis treatment. RESULTS: Significant time versus week interactions with any plasma amino acid were not detected; however, significant decreases in all plasma amino acid concentrations measured were detected at the midpoint of dialysis (46 +/- 2%) and at the end of each dialysis session (38 +/- 2%). Mean (+/- SEM) total free amino acid loss into the dialysate was 2.7 +/- 0.2 g or 0.12 g/kg of body weight. CONCLUSIONS AND CLINICAL RELEVANCE: Hemodialysis is associated with significant alterations in plasma amino acid concentrations and loss of free amino acids into the dialysate. Loss of amino acids into the dialysate, coupled with protein calorie malnutrition in uremic patients, may contribute to depletion of amino acid stores.     

Serum and skin IgA concentrations in normal and atopic dogs

Objective To compare serum and skin surface IgA concentrations from atopic and normal dogs.
Procedure IgA concentrations in sera and skin washings of 20 clinically normal dogs that had no history of pruritus or skin disease were compared to those obtained in 20 dogs with a diagnosis of atopy determined by history, clinical examination and positive intradermal skin test.
Results There was no significant difference in the mean serum IgA concentration in normal dogs (252 ± 187 mg/L) versus atopic animals (314 ± 327). When skin washings from all sites in both groups were compared, atopic dogs had significantly greater concentrations of IgA in their skin washings than normal dogs as evaluated by an enzyme-linked immunoassay (P < 0.001). However, there was no significant difference between the individual sites of the skin washings of atopic and normal dogs.
Conclusion IgA concentrations of skin washings in atopic dogs were greater than in normal dogs. Further investigations need to determine if the greater concentrations were caused by nonspecific inflammation or by secretion of allergen-specific IgA onto the skin surface.     

Characterization of Escherichia coli isolated from healthy dogs

Five month old dogs from a Midwestern research kennel occasionally developed bloody diarrhea after shipment to other facilities. As previous diagnostic efforts failed to reveal any potential pathogens in feces from normal and diarrheic dogs, Escherichia coli was investigated for select virulence properties that may contribute to the occurrence of bloody diarrhea. Fecal swabs from 52 healthy dogs were examined for E. coli. Two hundred and sixty E. coli-like colonies were screened by PCR for the attaching and effacing (eae) gene, Shiga toxin (stx) genes, and the heat-stable enterotoxin type A (sta) gene. One hundred forty two of the 260 E. coli-like colonies (54.6%) from 43 dogs were eae or sta positive; and 60 of the eae and/or sta positive isolates were examined further. Among the 60 isolates, 23 (38.3%) possessed the eae gene, 32 (53.3%) possessed the sta gene, and five (8.3%) possessed both eae and sta genes (eae⁺/sta⁺). Of the 60 isolates, six sta⁺ and one eae⁺/sta⁺ isolates were hemolytic. When examined in the suckling mouse assay, five of six sta⁺ isolates and three of four eae⁺/sta⁺ isolates gave gut-to-remaining carcass ratios ≥0.083, indicating expression of heat-stable enterotoxin. These enterotoxin-producing isolates belonged to serogroups O42, O170, and O-negative.     

Clostridium difficile in faeces from healthy dogs and dogs with diarrhea

Background

This study was conducted to evaluate the faecal occurrence and characterization of Clostridium difficile in clinically healthy dogs (N = 50) and in dogs with diarrhea (N = 20) in the Stockholm-Uppsala region of Sweden.

Findings

Clostridium difficile was isolated from 2/50 healthy dogs and from 2/20 diarrheic dogs. Isolates from healthy dogs were negative for toxin A and B and for the tcdA and tcdB genes. Both isolates from diarrheic dogs were positive for toxin B and for the tcdA and tcdB genes. The C. difficile isolates from healthy dogs had PCR ribotype 009 (SE-type 6) and 010 (SE-type 3) whereas both isolates from dogs with diarrhoea had the toxigenic ribotype 014 (SE-type 21). One of the isolates from healthy dogs was initially resistant to metronidazole.

Conclusions

This study revealed presence of toxigenic C. difficile in faecal samples of diarrheic dogs and low number of non- toxigenic isolates in healthy dogs from Uppsala-Stockholm region in Sweden. However, more comprehensive studies are warranted to investigate the role of C. difficile in gastrointestinal disease in dogs.     

Doppler-derived deformation imaging in unsedated healthy adult dogs

Objectives

To evaluate feasibility, repeatability and reproducibility (R&R) of Doppler-derived deformation imaging (DI) in healthy adult dogs.

Animals, materials and methods

Forty-nine dogs underwent physical examination, ECG, blood pressure measurement and echocardiography. Doppler-derived DI parameters obtained from six selected Regions of Interest (ROI) within the left ventricle (LV) were: strain (S), systolic strain rate (SSR), strain rate E wave (SRE), and strain rate A wave (SRA). The averages of the six ROIs were calculated and labeled avS, avSSR, avSRE, and avSRA. Randomly selected dogs underwent two echocardiographic studies to evaluate intraoperator (n = 14) and interoperator (n = 17) variability.

Results

DI data were obtained in 87.2% of dogs and 77.2% of ROIs. Compared to controls, avSSR was significantly reduced in Doberman Pinchers (DP) and it was increased in dogs <30 kg, compared to dogs >30 kg. The intraoperator Coefficient of Variability (CV) for some ROIs was greater than 15% but for averaged measurements it was ≤ 5.0%. Interoperator CV varied widely but were all <15% for avSSR, avS, and avSRA.

Conclusions

The CV for Doppler-derived DI varied widely. Averaging values from multiple ROIs improved R&R. DI may help elucidate differences in LV mechanics between canine breeds.     

Plasma and urinary trypsinogen activation peptide in healthy dogs, dogs with pancreatitis and dogs with other systemic diseases

OBJECTIVE: To determine the specificity and sensitivity of plasma and urinary trypsinogen activation peptide (TAP) concentrations in diagnosing pancreatitis in dogs. DESIGN: Retrospective analysis of clinical cases. PROCEDURE: Dogs were classified into three groups: healthy animals, dogs with confirmed pancreatitis and dogs with nonpancreatic disease, which clinically or biochemically resembled pancreatitis. This last group was further subdivided into dogs with renal and those with nonrenal disease. The plasma and urinary TAP concentration was determined by a competitive enzyme immunoassay. Clinical cases additionally had serum trypsin-like immunoreactivity concentration measured, as well as radiography and ultrasound of the abdomen and further diagnostic procedures. Nonparametric analysis of variance (Kruskal-Wallis test) was performed using Statistix 4.0 program. RESULTS: There was a wide range of urinary TAP concentration in healthy dogs (mean 52.30 nmol/L, standard deviation 55.25) that made interpretation of urinary TAP concentrations difficult in the other groups. There was a narrow reference range for plasma TAP (mean 2.67 nmol/L, standard deviation 0.93). Plasma and urinary TAP concentrations, as well as urinary TAP to creatinine ratio, were all increased in dogs that died with necrotising pancreatitis. Values were not increased in mild, interstitial pancreatitis. Increased plasma TAP concentrations were also present in dogs with severe renal disease. CONCLUSION: Plasma TAP concentration is a good prognostic indicator in naturally occurring pancreatitis in dogs. The failure of TAP to increase in mild pancreatitis, and the increase present in severe renal disease, suggests its measurement has limited application as a sole diagnostic tool for canine pancreatitis. Further investigations are required in order to explain the large variability of urinary TAP concentration and the presence of circulating TAP in healthy dogs.     

Analytical performances of d-ROMs test and BAP test in canine plasma. Definition of the normal range in healthy Labrador dogs

An high level of ROS (Reactive Oxygen Species), due to an increased production of oxidant species and/or a decreased efficacy of antioxidant system, can lead to oxidative stress, an emerging health risk factor involved in the aging and in many diseases, including inflammatory, infectious and degenerative disorders, either in humans or in animals. In the last years some assays panels have been developed to globally evaluate the oxidative balance by means of the concomitant assessment of ROS production and antioxidant system capability. In this report, the validation trials of d-ROMs (Reactive Oxygen Metabolites— derived compounds) and BAP (Biological Antioxidant Potential) tests in canine specie are described and also the specific referral ranges are calculated in a Labrador population. The results of linearity, precision and accuracy trials show that both tests exhibit good to excellent analytical performances. The possibility of measuring oxidative stress in vivo with simple, cheap and accurate tests, d-ROMs test and BAP test, provides for the veterinarians a very suitable tool to monitor oxidative stress and to correctly choice of eventual antioxidant supplementations in diseases proven related to oxidative stress in animals and particularly in dogs. Further studies will be useful to confirm this possibility.     

Endostatin concentrations in healthy dogs and dogs with selected neoplasms

Endostatin prevents angiogenesis and tumor growth by inhibiting endothelial cell proliferation and migration. The purpose of this study was to determine serum endostatin concentrations in 53 healthy dogs and in 38 dogs with confirmed malignant neoplasms. Endostatin concentration was determined with a competitive enzymatic immunoassay (EIA) with rabbit polyclonal antibody generated against a recombinant canine endostatin protein. Both the presence of cancer and increasing age were associated with increased serum concentration of endostatin. Endostatin concentration in healthy dogs was 87.7 +/- 3.5 ng/mL. Upper and lower limits of the reference range for serum endostatin concentration in healthy dogs were 60 and 113 ng/mL. Dogs with lymphoma (LSA) and hemangiosarcoma (HSA) had endostatin concentrations of 107 +/- 9.3 ng/mL. In conclusion, this study demonstrates that endostatin can be quantified in dogs and that endostatin concentrations are high in dogs with HSA and LSA.     

Serial blood lactate concentrations in systemically ill dogs

BACKGROUND: Lactate concentration often is quantified in systemically ill dogs and interpreted based on human data. To our knowledge, there are no published clinical studies evaluating serial lactate concentrations as a prognostic indicator in ill dogs. OBJECTIVES: Our objective was to perform a prospective study, using multivariate analysis, to determine whether serial lactate concentrations were associated with outcome in ill dogs requiring intravenous fluids. METHODS: Eighty sick dogs had lactate concentrations evaluated, using an analyzer that measures lactate in the plasma fraction of heparinized whole blood, at 0 hours and 6 hours after initiation of treatment. Severity of illness and outcome (survivor, nonsurvivor) were determined by reviewing the patient's record 2 weeks after admission. Lactate concentrations, age, body weight, gender, and severity of illness were evaluated using multivariate analysis to determine their effects on outcome. RESULTS: Dogs with lactate concentrations greater than the reference interval at 6 hours were 16 times (95% confidence interval = 2.32-112.71 times, P <.01) more likely not to survive compared to dogs with lactate concentrations within the reference interval. Lactate concentrations above the reference interval at 0 hours were not significantly related to outcome. However, hyperlactatemia that did not improve by > or = 50% within 6 hours was significantly associated with mortality (P = .024). CONCLUSION: Dogs with a lactate concentration higher than the reference interval at 6 hours were more likely not to survive. These results indicate an association between lactate concentration and outcome and emphasize the importance of serial lactate concentrations in evaluating prognosis.