Over a 24-month period, serum tumor necrosis factor (TNF) activity was determined in 289 horses with colic attributable to gastrointestinal tract disease. Serum TNF activity was quantitated by use of a modified in vitro cytotoxicity bioassay, using WEHI 164 clone-13 murine fibrosarcoma cells. Causes for colic, determined by clinical and laboratory evaluation, exploratory celiotomy, or necropsy included: gastrointestinal tract rupture (GTR); ileal impaction; small intestinal strangulating obstruction (SIO); proximal enteritis (PE); transient small intestinal distention; large-colon displacement; large-colon vovulus; large-colon impaction; colitis; small-colon obstruction; peritonitis; and unknown. Each diagnosis was placed into 1 of 3 lesion categories: inflammatory disorders (GTR, PE, colitis, peritonitis); strangulating intestinal obstruction (SIO, large-colon volvulus); and nonstrangulating intestinal obstruction (ileal impaction, transient small intestinal distension, large-colon displacement, large-colon impaction, small-colon obstruction, unknown). The prevalence of high serum TNF activity and/or mortality were evaluated. Differences were tested at significance level of P < 0.05. Approximately 20% of the 289 horses has serum TNF activity greater than that found in clinically normal horses (> 2.5 U/ml). Twenty-three horses (8%) had marked increase in serum TNF activity (greater than or equal to 10 U/ml) which was more prevalent among horses with SIO and PE than in horses of other diagnostic groups, except those with GTR. Mortality and marked increase in serum TNF activity were greater in horses with intestinal inflammatory disorders or strangulating intestinal obstruction than in horses with nonstrangulating intestinal obstruction. Similarly, a greater proportion of the horses that died had markedly high serum TNF activity than did horses that lived. Mortality of horses with serum TNF greater than or equal to 10 U/ml was greater than that of horses with serum TNF activity < 10 U/ml. Results indicate possible association between colic and serum TNF activity in horses and that high mortality may be associated with horses with markedly increased serum TNF activity.

The growth of the heart, relative to body weight, was measured by M-mode echocardiography in dogs during the first year of life. Echocardiographic measurements were obtained from 16 English Pointers at 1, 2, 4, and 8 weeks of age and at 3, 6, 9, and 12 months of age. Left atrial (LA), aortic (AO), left and right ventricular internal dimensions, interventricular septal and left ventricular wall thickness measurements increased in curvilinear fashion relative to increasing body weight. Least-squares regression analysis, performed on logarithmically transformed data, was used to develop power-law equations describing the relationship of echocardiographic measurements to body weight. Linear dimensions of the LA, AO, left and right ventricular internal dimensions and interventricular septal and left ventricular wall thickness changed proportionally to slightly differing exponential powers of body weight (BW), varying from 0.31 to 0.45 (BW(0.31) to BW(0.45)). Fractional shortening and the LA to AO ratio decreased slightly, but significantly, as body weight increased. Indexing echocardiographic measurements to BW(1/3) was more appropriate than indexing such measures linearly to body weight, offering a practical method for developing accurate normative graphs or tables for M-mode echocardiographic dimensions in growing dogs.

Ultrasonic pachymetry was used to measure central, superior peripheral, and temporal peripheral corneal thicknesses of 75 dogs (150 eyes) with normal corneas, anterior chambers, and intraocular pressure. Mean corneal thickness averaged over the 2 eyes, 3 locations, and 75 dogs was 562 +/- 6.2 micromole. The peripheral cornea was thicker on average than the central cornea by 49.43 +/- 8.45 micromole and this difference increased with age at 6.97 +/- 1.3 micromole/month of age. Mean corneal thickness changed with age (14.23 +/- 2.26 micromole/month), and weight (1.83 +/- 0.38 micromole/kg). Females had significantly thinner corneas (22.43 +/- 11.03 micromole than males) after adjusting for age and weight.

Four methods of evaluating renal function were performed in 6 cats anesthetized with halothane in oxygen. Glomerular filtration rate (GFR) was measured simultaneously in each cat by exogenous creatinine clearance (ECC), bolus inulin clearance, and 99mTc(Sn)-diethylene-triaminepentaacetic acid (DTPA) clearance determined by 2 different methods. In the first DTPA clearancemethod (DTPA-1), we measured radioactivity in serial blood specimens to construct plasma disappearance curves for calculation of GFR. In the second DTPA clearance method (DTPA-2), we used serial external head counts of radioactivity and a single blood specimen to construct plasma disappearance curves for calculation of GFR. Bolus inulin clearance was calculated from plasma disappearance curves using a 1-compartment open pharmacokinetic model (IN-1) and a 2-compartment open pharmacokinetic model (IN-2). Glomerular filtration rates were measured over 3 hours, for creatinine and DTPA methods, and over 4 hours for the inulin methods. The GFR obtained with the reference method (ECC) was 2.56 +/- 0.61 ml/min/kg of body weight (mean +/- SD). Values for GFR determined by ECC and DTPA-1 were significantly correlated (r = 0.852; P less than or equal to 0.05). Correlationbetween ECC and DTPA 2 was not as good (r = 0.783; P less than or equal to 0.10), but the 2 DTPA methods significantly correlated with one another (r = 0.897; P less than or equal to 0.05). Regardless of the method of calculation, bolus inulin clearance was poorly correlated with ECC (IN-1: r = 0.538, P greater than or equal to 0.10; IN-2: r = 0.430, P greater than or equal to 0.10) and DTPA-1 IN-1: r = 0.601, P greater than or equal to 0.10; IN-2: r = 0.625, P greater than or equal to 0.10). The 2 methods of calculating inulin clearance were highly correlated (r = 0.927; P less than or equal to 0.01). The DTPA clearance calculated from directly measured plasma disappearance curves (DTPA-1) comparedfavorably with ECC as an estimate of GFR and appears to be a safe, reliable, and less invasive method of determining GFR incats.

The purpose of this study was to compare the thermodilution technique for estimation of cardiac output with the indocyanine green dye dilution technique at flows between 10 and 39 L/min in halothane-anesthetized horses. The estimation of area of dye dilution cardiac output curves was made by using the fore-'n-aft (FA) triangle method. This shorthand technique was compared with logarithmic exponential extrapolation and summation (extrapolated area), using 64 cardiac output curves. Then, 256 simultaneous thermodilution measurements were compared with dye dilution measurements calculated by use of the FA technique. Forty milliliters of iced 0.9% NaCl solution containing 15 mg of indocyanine green dye was used as the indicator. This was delivered in < 1 second to the right atrium, using a power injector. A thermistor positioned in the pulmonary artery detected the thermal indicator. Blood was withdrawn from the carotid artery through a densitometer cuvette to measure the dye concentration. The FA estimations of area were higher than those determined by use of extrapolated area. A multiplicative adjustment of 0.837 was estimated. On average, thermodilution estimates of cardiac output exceeded the adjusted FA determinations. Using a weighted linear regression, we determined the following calibration adjustment: thermal dilution cardiac output/1.048 = indocyanine green dye dilution cardiac output.

Naturally acquired gram-negative bacterial intramammary infections (n = 160) were studied in 99 cows over a 2-year period. Escherichia coli, Klebsiella spp, Serratia spp, Enterobacter spp, and unidentified gram-negative bacteria were isolated from 28.8, 39.4, 9.4, 5.0, and 11.2%, respectively, of infected mammary glands. A majority (61%) of intramammary infections were first detected during the nonlactating period. Gram-negative bacteria isolated during the first half of the nonlactating period were predominantly Klebsiella spp, Serratia spp, and Enterobacter spp. Onset of E coli intramammary infections was more prevalent during the second half of the nonlactating period and during the first 7 days of lactation. The majority (59%) of infections were <28 days in duration, but Klebsiella spp and Serratia spp infections were of significantly (P <0.05) greater duration than infections with E coli. The greatest percentage (47%) of gram-negative bacterial intramammary infections were first detected during the summer.

The role of prostaglandin F2 alpha (PGF2 alpha) in embryonic loss following induced endotoxemia was studied in mares that were 21 to 44 days pregnant. Thirteen pregnant mares were treated with a nonsteroidal anti-inflammatory drug, flunixin meglumine, to inhibit the synthesis of PGF2 alpha caused by Salmonella typhimurium endotoxin given IV. Flunixin meglumine was administered either before injection of the endotoxin (group 1, -10 min; n = 7), or after endotoxin injection into the mares (group 2, 1 hour, n = 3; group 3, 2 hours, n = 3); 12 pregnant mares (group 4) were given only S typhimurium endotoxin. In group 4, the secretion of PGF2 alpha, as determined by plasma 15-keto-13,14-dihydro-PGF2 alpha concentrations, was biphasic, initially peaking at 30 minutes followed by a second, larger peak approximately 105 minutes after the endotoxin was given IV. When flunixin meglumine was administered at -10 minutes, synthesis of PGF2 alpha was inhibited for several hours, after administration of flunixin meglumine at 1 hour, the second secretory surge of PGF2 alpha was blocked, and administration of the drug at 2 hours did not substantially modify the secretion of PGF2 alpha. Plasma progesterone concentrations were unchanged after endotoxin injections were given in group 1. In group 2, progesterone values decreased < 2 ng/ml and remained low for several days. In group 3 and group 4, progesterone concentrations decreased to values < 0.5 ng/ml by 48 hours after endotoxin injections were given. Pregnancy continued in all 7 mares in group 1; in group 2, pregnancy continued in 2 of 3 mares; and in group 3, none of the 3 mares was pregnant by 4 days after endotoxin administration. The abortifacient effect of endotoxemia in mares < 2 months pregnant is mediated indirectly through the secretion of PGF2 alpha; compromised luteal activity and inadequate progesterone secretion are the primary causes of fetal death. Although flunixin meglumine administration can be used to prevent loss of pregnancy in cases of endotoxemia, it must be initiated at an early stage of the endotoxemia, that is, when clinical signs are often not yet apparent.

The antebrachiocarpal and tarsocrural joints of 10 adult horses were randomly assigned to 1 of 4 groups. Groups were formulated and were treated as follows: group 1, control (arthrocentesis only); group 2, buffered lactated Ringer solution; group 3, 10% dimethyl sulfoxide (DMSO; w/v) in lactated Ringer solution; and group 4, 30% DMSO (w/v) in lactated Ringer solution. Joints were lavaged once with the respective solution. Prior to lavage and on days 1, 4, and 8 after lavage, all horses were evaluated for lameness and joint effusion; synovial fluid total and differential WBC counts, synovial fluid total protein concentration, and mucin clot quality were determined. Horses were euthanatized on day 8, and joints were evaluated grossly, histologically, and histochemically. Significant difference was not observed in effect of lactated Ringer solution, 10% DMSO, and 30% DMSO on any measured variable. At 24 hours after treatment, significant (P < 0.05) difference in synovial fluid WBC numbers and total protein concentration was detected between control and treated joints. Eighty percent of lavaged joints had effusion 24 hours after treatment, compared with 30% of control joints. Gross, histopathologic, or histochemical differences were not detected between treated and control joints. Results of the study indicate that buffered lactated Ringer, 10% DMSO, and 30% DMSO solutions induce similar inflammatory changes in articular structures and significantly greater inflammatory reaction than does arthrocentesis alone.

A study was conducted to determine the immune (increased antibody) and protective (reduced colony-forming units) responses induced in mice by a: (i) single vaccinal inoculation, using various concentrations of Brucella cell surface protein (BCSP) or lipopolysaccharide (LPS); (ii) primary inoculation, using various concentrations of BCSP, followed by a secondary inoculation, using a standard concentration of BCSP; and (iii) primary inoculation, using 1 concentration of BCSP or LPS, followed by a secondary inoculation, using various concentrations of BCSP or LPS. Four weeks after the primary inoculation, mice were challenge exposed with approximately 1 X 10(4) colony-forming units of Brucella abortus strain 2308 and all mice were euthanatized at 6 weeks. Reduced splenic weights and reduced colony-forming units in the spleens of vaccinated mice, compared with nonvaccinated mice, were the criteria of protection. Increase in serum IgM and IgG was defined as immunity. Both BCSP and LPS induced protective and immune responses that were proportional to the dose given up to an optimal limit. However, concentrations higher than optimal decreased the protective and immune responses. This was true for mice given either 1 or 2 vaccinal inoculations. Enhanced secondary protective responses were seen only when suboptimal doses were used in the primary inoculation. Excessive or optimal doses in the secondary inoculations prevented or obscured the protectiveness and immunity by primary inoculations. The protective effects appeared to be additive when suboptimal doses were used in the primary and secondary inoculations. Inoculation of subimmunogenic doses induced a relative reduction in the antibody concentration after challenge exposure, compared with nonvaccinated mice. The overall results indicated that the protective responses induced by BCSP were probably attributable to LPS. The results also indicated a linear increase in protection and immune response corresponding to increasing doses up to an optimal dose, and this stoichiometric optimum may be achieved by the use of 1 or more vaccinal inoculations. However, once this optimum was obtained, additional amounts of BCSP or LPS cause perturbation of both the protective and serologic responses.

Canine and human platelets (washed 4 times in a solution containing EDTA, prostaglandin E1, and theophylline to prevent release of alpha-granule constituents) were lysed by being frozen and thawed in the presence of detergent. Radioelectroimmunoassay for von Willebrand factor (vWf) in 5 human platelet lysates produced precipitin rockets, shaped like those produced from vWf in plasma from healthy human beings, and indicated that the mean von Willebrand factor antigen (vWf:Ag) content in platelets from healthy human beings was 526 +/- 87 human U/10(12) platelets. Radioelectroimmunoassay for vWf in platelet lysates from 17 healthy dogs with normal plasma vWf:Ag concentration produced precipitin rockets that looked different from those produced from canine plasma and indicated vWf:Ag content of 59 +/- 35 canine U/10(12) platelets. Inclusion of protease inhibitors in the lysing solution did not normalize the appearance of the precipitin rockets or substantially alter the measured platelet content of vWf:Ag. The array of vWf multimers revealed by sodium dodecyl sulfate-agarose gel electrophoresis of canine platelet lysates had a distinct appearance that differed from that of vwf in canine or human plasma and platelets; the intensity of the canine platelet vWf multimer bands was skewed, with relatively greater density in the lower molecular weight region and faint or undetectable multimer bands in the higher molecular weight region. Electrophoretograms with visible multimers in the high molecular weight region had vwf components that had higher molecular weight than did any vWf components in canine plasma. Radioelectroimmunoassay for fibronectin in these same canine platelet lysates indicated that the fibronectin content in platelets was 2.89 +/- 1.10 mg/10(12) platelets. An Airedale Terrier with type-I von Willebrand disease (vWd), but lacking clinical signs of vWd, had normal platelet content of vwf:Ag (28 +/- 12 canine U/10(12) platelets), whereas a German Shorthaired Pointer with moderately severe type-II vWd and a mildly affected Doberman Pinscher with type-I vWd had only a trace or undetectable amounts of vWf:Ag in their platelets. The concentration of vWf:Ag in platelet lysates from the Doberman Pinscher with vWd remained undetectable when the platelets were isolated from the Doberman Pinscher's blood mixed with citrated plasma from dogs with normal plasma vWf:Ag concentration. In all 3 dogs with vWd, platelet fibronectin content was within the normal range.