Evaluation of secondary haemostasis in canine leishmaniasis

Secondary haemostasis was evaluated in 26 dogs with leishmaniasis and 10 normal dogs by measurements of modified one-stage prothrombin time (m-OSPT), activated partial thromboplastin time (APTT), thrombin time, fibrinogen concentration and fibrin degradation products. There were no significant differences between the groups in the m-OSPT, fibrinogen concentration, or levels of fibrin degradation products. The APTT was significantly (P = 0.006) longer in the infected dogs than in the control group, and in infected dogs with alanine aminotransferase (ALT) activities > 50 U/litre. There was a significant linear regression between ALT and APTT. Thrombin time was significantly (P = 0.003) longer in the infected dogs than in the normal dogs. There were no significant differences between the thrombin times of sick dogs with different levels of creatinine or activities of ALT, or between male and female dogs, whether diseased or normal.     

Correlation of haemostatic abnormalities with tumour stage and characteristics in dogs with mammary carcinoma

Sixty female dogs with untreated mammary carcinoma, comprising equal numbers of dogs in tumour stages I to IV, were evaluated for haemostatic abnormalities using the following tests: platelet count, prothrombin time, activated partial thromboplastin time, thrombin time, plasma activity of factor V, VIII and X, plasma concentration of fibrinogen, fibrin monomers and fibrinogen degradation products, and plasma antithrombin III activity. Two-thirds of all dogs had one or more haemostatic test abnormality of which the likelihood and frequency was increased in those with stage III and IV neoplasia. Haemostatic abnormalities were more frequently observed in dogs which had mammary tumours with distant metastases, extended tumour necrosis, inflammatory carcinomas, tumours fixed to underlying structures, or tumours in which there was penetration of the tumour capsule by tumour cells. As in humans with mammary carcinoma, these haemostatic abnormalities might be used as prognostic indicators, but their clinical importance remains unknown.     

Hemostatic changes in dogs with naturally occurring sepsis

Sepsis is a frequent source of morbidity and mortality in critically ill patients. The goal of this case control study was to measure hemostatic changes in dogs with naturally occurring sepsis. Blood was collected within 24 hours of admission from 20 dogs that fulfilled the criteria for sepsis. Sepsis was defined as histologic or microbiological confirmation of infection and 2 or more of the following criteria: hypo- or hyperthermia, tachycardia, tachypnea, or leukopenia, leukocytosis, or > 3% bands. Culture and sensitivities were performed on appropriate samples from all septic dogs. Twenty-eight control dogs were enrolled on the basis of normal results of physical examination, CBC, serum biochemistry, and coagulation profile. Plasma samples were analyzed for prothrombin time (PT), partial thromboplastin time (PTT), fibrin(ogen) degradation products (FDP), D-dimer (DD) concentrations, antithrombin (AT) activity, and protein C (PC) activity. Data were compared between groups by chi-square or independent t-tests. PC (P < .001) and AT (P < .001) activities were significantly lower in dogs with sepsis compared to controls. Dogs with sepsis had significantly higher PT (P = .007), PTT (P = .005), D-dimer (P = .005), and FDP (P = .001) compared to controls. Platelet counts were not significantly different between groups. Ten of the 20 septic dogs (50%) died, but no association was identified between any of the measured variables and outcome. These findings are consistent with previous studies in animals with experimentally induced disease and in clinical studies of humans. On the basis of these results, further investigation of the role of AT and PC in canine sepsis is warranted.     

Frequent respiratory tract infections in the canine model of X-linked ectodermal dysplasia are not caused by an immune deficiency

As in many human patients with X-linked hypohidrotic ectodermal dysplasia (XHED), XHED dogs are at an increased risk for pulmonary disorders. Localized immune system defects had been suspected previously in affected dogs because of frequent infections and unexpected deaths due to opportunistic respiratory tract infections. Experiments were designed to examine systemic and localized humoral and cellular responses, development and function of T cells, and thymic morphology. All dogs used in these experiments were clinically healthy at the time of examination and their immune responses were compared to normal littermates. Serum immunoglobulin concentrations differed somewhat between normal dogs and dogs affected with XHED but they were all within normal ranges. The XHED dogs responded appropriately to vaccination with tetanus toxoid suggesting normal systemic B and plasma cell function. Thymic morphology was compared between normal and affected dogs and T cells were assessed for functionality. Numbers and phenotypes of T and B cells in blood and thymus of affected dogs were within normal limits suggesting normal development of T cells. Cytotoxic and phagocytic ability of macrophages and neutrophils was also normal in affected dogs. In contrast, the secretory IgA concentrations found in affected dogs were significantly higher than in normal dogs, while lacrimal secretions were significantly decreased. These results suggest a compensatory mechanism for secretory IgA, so that the total amount equals that in normal dogs. The results presented in this study indicate that the XHED dogs have a relatively intact immune system and suggest that the same is true for humans with the homologous form of XHED.     

Familial po1ycystic kidnev disease in bull terriers

Polycystic kidney disease was observed in eight related bull terriers. Two dogs died suddenly, two were presented for haematuria, while the remainder were outwardly normal but detected during ultrasonographic screening programmes. A definitive diagnosis was made at either necropsy or using renal ultrasonography, although affected dogs also had abnormal urinalyses with haematuria, proteinuria, cast formation and sometimes bacterial infection. Valvular heart disease was detected clinically or at necropsy in all the affected dogs. The presence of polycystic kidney disease in several related dogs suggests that this condition is familial. The disease has clinical and pathological similarities with autosomal dominant polycystic kidney disease of humans.     

Muscular dystrophy in female dogs

The most common form of muscular dystrophy in dogs and humans is caused by mutations in the dystrophin gene. The dystrophin gene is located on the X chromosome, and, therefore, disease-causing mutations in dystrophin occur most often in males. Therefore, females with dystrophin deficiency or other forms of muscular dystrophy may be undiagnosed or misdiagnosed. Immunohistochemistry was used to analyze dystrophin and a number of other muscle proteins associated with muscular dystrophy in humans, including sarcoglycans and laminin alpha2, in muscle biopsy specimens from 5 female dogs with pathologic changes consistent with muscular dystrophy. The female dogs were presented with a variety of clinical signs including generalized weakness, muscle wasting, tremors, exercise intolerance, gait abnormalities, and limb deformity. Serum creatine kinase activity was variably high. One dog had no detectable dystrophin in the muscle; another was mosaic, with some fibers normal and others partly dystrophin-deficient. A 3rd dog had normal dystrophin but no detectable laminin alpha2. Two dogs could not be classified. This study demonstrates the occurrence of dystrophin- and laminin alpha2-associated muscular dystrophy and the difficulty in clinical diagnosis of these disorders in female dogs.     

MITRAL ANNULUS MOTION AS DETERMINED BY M-MODE ECHOCARDIOGRAPHY IN NORMAL DOGS AND DOGS WITH CARDIAC DISEASE

M-mode echocardiography was used to assess apical mitral annulus motion (MAM) in 103 normal dogs and 101 dogs with cardiac disease, to obtain information on systolic left ventricular long axis function. In normal dogs, a close relationship was found between MAM and body weight (r = 0.80, P < 0.001). There was a weak correlation between MAM and heart rate (r = -0.25, P < 0.05), but no correlation between MAM and age or left ventricular shortening fraction (P > 0.05). Mean MAM (95% confidence intervals) were established for normal dogs of differing body weight, and were 0.70 cm (0.65 to 0.75) in dogs < 15 kg, 1.08 cm (1.03 to 1.13) in dogs weighing 15 to 40 kg, and 1.51 cm (1.21 to 1.81) in dogs > 40 kg. "Cut-off" values to define decreased MAM for normal dogs of differing body weight were 0.45 cm (dogs < 15 kg), 0.80 cm (dogs 15-40 kg), and 1.20 cm (dogs > 40 kg). In dogs with cardiac disease, median MAM was normal in mitral valve endocardiosis or aortic stenosis, but significantly decreased (P < 0.05) in dilated cardiomyopathy. All dogs with mitral valve endocardiosis (n = 54) or aortic stenosis (n = 26) had MAM above the above-mentioned "cut-off" values, suggesting normal or increased left ventricular longitudinal systolic shortening, whereas 81% (17/21) of dogs with dilated cardiomyopathy had MAM below the "cut-off" value, indicating decreased long axis systolic function. It is concluded that MAM may be used to evaluate systolic left ventricular long axis performance in dogs and may add useful information on global left ventricular contraction dynamics.     

Encapsulating peritoneal sclerosis associated with abnormal liver development in a young dog

A 6-month-old male miniature pinscher dog developed chronic ascites, formation of fibrous membrane covering the abdominal organs, and numerous adhesions between the intestinal loops. The membrane and adhesions were surgically removed twice, but the dog died 1 month after initial presentation. Necropsy revealed recurrence of the membrane and intestinal adhesions. The fibrous membrane was composed of a thick layer of mature collagenous connective tissue, covered by immature collagenous connective tissue with mild lymphoplasmacytic infiltration, prominent neovascularization, and fibrin exudation. Similar fibrotic lesions were observed in the serosa of the liver, spleen, stomach and intestines. These findings are consistent with encapsulating peritoneal fibrosis reported in humans and dogs. The dog also had a maldeveloped liver, which is characterized by disorganized hepatic lobules and disarranged hepatic cords.     

Increased renal vascular resistance in dogs with hepatic disease

Doppler ultrasound is a non-invasive technique that can be used to estimate vascular resistance by calculation of resistive index (RI) and pulsatility index (PI). Liver disease may increase renal RI and PI, and in humans with liver disease the indices are monitored to attain prognostic information. Systemic hypertension has been found in dogs with hepatic disease and is also related to increased renal vascular resistance in humans. The aim of this study was to examine renal vascular resistance increases in dogs with hepatic disease and to ascertain whether these may be related to blood pressure increases and biochemical parameters. Twenty dogs with hepatic disease were evaluated. The mean renal RI, PI, and systolic blood pressure were significantly higher than in normal animals. A positive correlation was found between the indices and alkaline phosphatase but not with systolic blood pressure. It is concluded that renal vascular resistance may increase in dogs with hepatic disease and in this study was above the limit value in 50% of the animals.     

Clinical, biochemical, and molecular aspects of Glanzmann's thrombasthenia in humans and dogs

Glanzmann's thrombasthenia (GT) is an inherited, intrinsic platelet function defect that involves the platelet glycoprotein complex IIb-IIIa, also known as the fibrinogen receptor and the integrin alphaIIbbeta3. The defect was originally described by Dr. Glanzmann in humans in 1918 as a bleeding disorder that differed clinically from other known coagulopathies. Over the decades that followed, researchers determined the biochemical and molecular basis for the disease in humans. Otterhounds with thrombasthenic thrombopathia, described in the 1960s, were the only animal model that closely resembled the disease described in humans until 1996. At that time, a Great Pyrenees dog was identified with unequivocal clinical and biochemical features of Type I GT The cDNA encoding for glycoproteins IIb and IIIa were sequenced in normal dogs in 1999, allowing for identification of specific mutations causing Type I GT in both Otterhounds and Great Pyrenees dogs. Knowing the molecular basis for Type I GT in dogs as well as the cDNA sequences in normal dogs should enhance the understanding of structure/function relationships of the alphaIIbbeta3 integrin and provide an excellent animal model for studies aimed at correction of GT in humans. The following review focuses on the structure and function of this platelet receptor and reviews the molecular, biochemical, and clinical aspects of Glanzmann's thrombasthenia in humans and dogs.     

Use of tissue plasminogen activator for intraocular fibrinolysis in dogs.

Fibrin clots were induced in eyes of dogs by injection of autogenous citrated plasma into the anterior chamber. Twenty-four hours after clot formation, 0.01 ml of tissue plasminogen activator at a concentration of 1 microgram/100 microliters (group 1, n = 5) or 25 micrograms/100 microliters (group 2, n = 5) was injected into 1 anterior chamber of each dog. The contralateral eye served as a nontreated control. Serial photographs were taken of the fibrin clots after intracameral injection of tissue plasminogen activator. Computerized morphometric analysis was then used to evaluate changes in cross-sectional surface area of the fibrin clot. Significant (P less than 0.001) fibrin-clot lysis was detected in treated eyes of group-2 dogs, but was not found in treated eyes of group-1 dogs. A mean decrease of greater than 90% in clot surface area was detected by 120 minutes after injection in treated eyes of group-2 dogs.     

Escherichia coli as a pathogen in dogs and cats.

Certain strains of Escherichia coli behave as pathogens in dogs and cats causing gastro-intestinal and extra-intestinal diseases. Among the five known groups of diarrhoeagenic E. coli, namely enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC), enteroinvasive E. coli (EIEC), shiga-toxin producing E. coli (STEC) and enteroaggregative E. coli (EAggEC), only EPEC and ETEC were clearly associated with enteric disease in young dogs. ETEC isolates from diarrhoeic dogs were found to be positive for the heat-stable enterotoxins STa and STb but negative for heat-labile enterotoxin (LT). Canine ETEC were found to be different from those of other animals and humans by their serotypes, production of alpha-haemolysin and adhesive factors and by the production of uncharacterized types of enterotoxins by some ETEC. Canine EPEC could be distinguished from EPEC of humans or other animals by their serotypes and by the eae-protein intimin which mediates intimate adherence of EPEC to intestinal mucosa cells. STEC were occasionally isolated from faeces of healthy and diarrhoeic dogs but their role in canine diarrhoea is not yet well known. EIEC and EAggEC were not reported to occur in dogs or cats. Very little is known on diarrhoegenic E. coli in cats and further epidemiological investigations on this subject are needed. Besides its role in gastro-intestinal infections, E. coli can cause infections of the urogenital tract and systemic disease in dogs and cats. Extra-intestinal pathogenic E. coli strains from dogs and cats belong to a limited number of serotypes and clonal groups and are frequently found as a part of the normal gut flora of these animals. Many of these E. coli strains carry P-fimbriae and produce alpha-haemolysin and a necrotizing cytotoxin (CNF1). Some of the frequently isolated types of extra-intestinal pathogenic E. coli from dogs, cats and humans were found to be highly genetically related but showed differences in their P-fimbrial adhesins which determine host specificity. Transmission of extra-intestinal and enteral pathogenic E. coli between dogs and humans was reported. Further research is needed, however, to determine the role of dogs and cats as transmission vectors of pathogenic E. coli strains to other animals and humans.     

VARIATIONS IN MAGNETIC RESONANCE VENOGRAPHIC ANATOMY OF THE DORSAL DURAL VENOUS SINUS SYSTEM IN 51 DOGS

Variations in intracranial dural venous sinus anatomy have been widely reported in humans, but there have been no studies reporting this in dogs. The purpose of this retrospective study was to describe variations in magnetic resonance (MR) venographic anatomy of the dorsal dural venous sinus system in a sample population of dogs with structurally normal brains. Medical records were searched for dogs with complete phase contrast, intracranial MR venograms and a diagnosis of idiopathic epilepsy. Magnetic resonance venograms were retrieved for each dog and characteristics of the dorsal dural sinuses, symmetry of the transverse sinuses and other anatomic variations were recorded. A total of 51 dogs were included. Transverse sinus asymmetry was present in 58.8% of the dogs, with transverse sinus hypoplasia seen in 39.2%, and aplasia in 23.5% of dogs. For 70.6% of dogs, at least one anatomic variation in the dorsal sagittal sinus was observed, including deviation from the midline (33.3%) and collateral branches from either the dorsal sagittal sinus or dorsal cerebral veins (54.9%). In 5 dogs (9.8%) a vessel was also identified running from the proximal transverse sinus to the distal sigmoid sinus, in a similar location to the occipital sinus previously reported in children. Findings from this study indicated that, as in humans, anatomic variations are common in the intracranial dural venous sinus system of dogs. These anatomic variations should be taken into consideration for surgical planning or diagnosis of cerebrovascular disease.     

Non-immunoglobulin defense factors in canine saliva and effects of a tooth gel containing antibacterial enzymes

Selected innate: non-immunoglobulin defense factors in canine saliva were characterized and quantitated. The samples from dogs showed increased pH, higher lysozyme and salivary peroxidase activities, but lower hypothiocyanite concentration and myeloperoxidase activity when compared with human saliva. Secondly, a 1-month clinical pilot study was performed using a commercial tooth gel to determine acute and long-term changes in salivary host defense proteins. Daily application of the tooth gel did not substantially affect the concentrations of these factors in dogs with normal salivation. Our results suggest that canine saliva may be similar to human saliva, comprised of both immune and non-immune antimicrobial factors. However, as in humans, oral administration of antimicrobial proteins as reported here does not seem to benefit dogs with normal saliva secretion. Products such as the tooth gel evaluated in this study may benefit dogs with xerostomia or other clinical conditions causing decreased saliva production.     

Hereditary spectrin deficiency in Golden Retriever dogs

Spectrin deficiency with increased erythrocyte osmotic fragility (OF) is a hallmark of hereditary spherocytosis, which is the most common congenital hemolytic anemia in humans of northern European ancestry. A radioimmunoassay revealed that erythrocyte spectrin concentration was 50-65% of normal in 5 adult Golden Retriever dogs, which had recovered from hemolytic anemia but whose OF had persistently remained increased. OF also was increased and spectrin concentration was decreased (60-73%) in 10 dogs of an apparently healthy family of 19 Golden Retrievers related to a proband. Pedigree analysis revealed autosomal dominant inheritance. In addition, OF was increased in 23 (17%) of 134 randomly chosen Golden Retrievers with nonhematologic diseases. In these Golden Retrievers, the spectrin concentration was decreased in 5 dogs with increased OF and within the reference range in 6 dogs with normal OF, indicating that in this population spectrin deficiency and increased OF are highly associated (P < .002). Considering these patients a representative sample of the Golden Retriever population in the Netherlands, spectrin deficiency may occur in 11.2-24.6% of Dutch Golden Retrievers (confidence level = 0.95). In blood smears, spherocytes were recognized only in dogs with immune-mediated anemia. At scanning electron microscopy, blood from spectrin-deficient Golden Retrievers showed slight crenation when fixed freshly but abundant echinospherocytes after 24 hours of incubation. We conclude that occult autosomal dominant spectrin deficiency occurs in dogs and is frequent in Dutch Golden Retrievers. It is not clear whether spectrin deficiency in Golden Retrievers may result in hemolytic anemia, as in humans.     

Resident forum abstractsEVALUATION OF THROMBOELASTOGRAPHY (TEG) IN NORMAL CATS

Objective: To establish normal parameters of thromboelastography (TEG) in healthy adult cats. Background: Thromboelastography (TEG) is an in vitro test of coagulation that has been shown to be useful in humans, dogs and select species to identify and quantify alterations of hemostasis (e.g., hypercoagulable and hypocoagulable states). It has also been demonstrated to be useful in monitoring effects of anticoagulant therapies. This test has not been evaluated in cats. Methods: Blood was collected from 25 clinically normal cats by venipuncture using a 21 gauge×3 1/2 inch butterfly catheter and syringe for medial saphenous or jugular venipuncture. A single 1.8 mL sample in 3.8% Sodium Citrate (9:1) was collected from each cat. Recalcified whole blood was analyzed 30 minutes following collection with the TEG® 5000 analyzer (Haemoscope, Niles, IL). Analysis temperature was 37.6°C. TEG parameters recorded included: R‐value (represents initial fibrin formation), K (time from R to standard fixed measure of clot firmness which represents contributions of platelets and fibrinogen), maximum amplitude (MA; represents absolute clot strength), and alpha angle (α; the slope of TEG tracing which represents rate of clot formation). The coagulation index (CI) was derived from the formula generated for humans to provide an overall assessment of whether the sample was hyper‐ or hypocoagulable. Results: Values for the 25 normal cat samples are reported as mean ±2 standard deviations. R=2.97; 1.23–4.72; K=1.54, 0.38–2.71; α=70.70, 57.76–83.65; MA=58.50, 45.26–71.74 and CI=2.27, 0.07–4.46. Compared to historical information obtained on normal dogs, cats have significantly shorter R and K and larger α, MA and CI. Conclusions: TEG does have reproducible performance when used to evaluate coagulation status in normal cats. Compared to dogs, normal cats favor a hypercoagulable state. Species‐specific normal values are necessary for interpretation of TEG results. This test bears potential value for use in future experimental and clinical work to investigate hemostasis in cats receiving anticoagulant therapies or in cats suffering from diseases such as cardiomyopathy which are thought to be associated with altered coagulation status.     

Glucose tolerance and insulin response in diabetes mellitus of dogs

The low dose intravenous glucose tolerance test (IVGTT) and the insulin response to the glucose load were performed in a series of twenty–two diabetic dogs. All diabetic dogs were characterized by glucose intolerance as expressed by an abnormal half time (Tl/2) or fractional turnover rate (k) for glucose clearance. On the basis of the initial insulin level (Io), the insulin peak response (Ip) and the insulinogenic index (I/G), the dogs were classified into three types. Type I dogs were characterized by a low Io, low Ip and low I/G in response to glucose, similar to the juvenile form of diabetes in humans. Type II dogs were characterized by a normal or high Io, but also with a low Ip and a low I/G which are some of the features of the maturity onset form. Type III dogs were characterized by a normal Io and a normal or delayed response to glucose as seen in chemical diabetes. It is suggested that these types represent stages in the natural history of the development of diabetes mellitus in dogs.     

Sulphido-leukotriene production from peripheral leukocytes and skin in clinically normal dogs and house dust mite positive atopic dogs

Pathogenesis of canine atopy has not been completely elucidated. In humans, sulphido-leukotrienes (s-LT) play a role in atopy, and increased production of s-LT occurs in the skin and peripheral leukocytes after allergen challenge. The study population included 16 clinically normal and 13 atopic dogs. All atopic dogs had in common a positive reaction (4+) to the intradermal injection of house dust mite (allergen of reference). Blood samples and skin biopsies were collected. Sulphido-LT synthesis by peripheral leukocytes after stimulation was measured, and no statistically significant difference was found between clinically normal and atopic dogs. Sulphido-LT concentrations in skin samples from stimulated and unstimulated sites were measured, and no statistically significant difference was detected between clinically normal and atopic dogs or between lesional and nonlesional skin within the atopic group. Clinical signs of atopic dogs were graded by owners and no correlation was found between their severity and cutaneous concentrations of s-LT. In this study there was no increase in s-LT synthesis in atopic dogs.     

Glycosylated hemoglobin measurement in dogs and cats: implications for its utility in diabetic monitoring

The measurement of glycosylated hemoglobin (HbA₁) levels in humans is used to indicate the degree of long-term diabetic control. Using a commercially available kit for human HbA₁, values were obtained for normal and diabetic dogs and cats. The normal range established in dogs was broad and overlapped considerably with the range in diabetics. Under the assay conditions and with a limited number of diabetic animals, the test was not found to be of value for dogs or cats.     

Evaluation of Canine-Derived Fibrin Sealant as a Hemostatic Agent

The purpose of this study was to determine whether canine-derived fibrinogen concentrate applied with bovine thrombin was a safe and effective topical hemostatic agent. A canine liver biopsy model was selected to test this product. Cryoprecipitate was prepared from frozen canine plasma using two freeze/thaw/centrifugation cycles. Six healthy adult dogs (weighing more than 18 kg) were used in the fibrin sealant study, and an additional three dogs were used as controls for the liver biopsy. A 1 × 3 cm liver biopsy specimen was obtained, digital pressure was applied to reduce bleeding, and the fibrinogen concentrate was immediately sprayed on the bleeding surface simultaneously with bovine thrombin (1,000 IU/mL). The mean ± standard error of the mean (SEM) blood pressure at time of biopsy was 98 ± 9 mm Hg, and the rate of hemorrhage from the cut liver edge was 8.0 ± 1.1 mL/min. The total blood loss during fibrin sealant application was 37 ± 9 mL and total time for hemostasis was 5.5± 1.3 minutes. There was no additional hemorrhage after application of the fibrin sealant. In the three control dogs, fibrin sealant was not applied and only digital compression was used to decrease hemorrhage. Before digital compression, the rate of hemorrhage from the cut liver edge was 13.1 ±3.1 mL/min. Bleeding had not stopped after 10 minutes of compression and the mean postcompression rate of hemorrhage was 4.0 ± 2.6 mL/min. Signs of secondary bleeding after fibrin sealant was applied were not evident during the immediate postoperative period or over the next 14 days. The coagulation profile, alanine amino transferase (ALT), alkaline phosphatase (AP), and bile acids were not significantly different on day 14 from preoperative values, suggesting that the liver was only mildly affected. The dogs were euthanatized and gross and histological examinations of the biopsy site were performed on day 14. There were minimal or no adhesions at the biopsy site. In all dogs, the hepatic capsule was less than 1 mm thick and histological signs of secondary bleeding, hepatocellular damage, thrombosis, infection, or inflammation were not observed. In this study, canine-derived fibrin sealant was a safe and effective topical hemostatic agent.