Klainbart S, Agi L, et al. Clinical, laboratory, and hemostatic findings in cats with naturally occurring sepsis. J Am Vet Med Assoc 2017;251:1025-34.
Sepsis may be considered a subset of the systemic inflammatory response syndrome (SIRS) associated with infection, and infectious agents associated with sepsis can include bacteria, viruses, fungi, protozoa, or helminths. Even in human patients, where generally every available diagnostic and therapeutic modality is employed to identify and treat sepsis, morbidity and mortality are still considerable. Mortality rates from sepsis in people are reported to range from 28-50%. There is also a significant body of information regarding sepsis in dogs, but less is known about the diagnosis and treatment of sepsis in cats and, as is typical of the feline species, signs of sepsis may be more subtle and harder to identify than in canines or humans, which may lead to delayed treatment and greater morbidity and mortality for cats with sepsis.
Although there is still no consensus regarding the diagnosis of SIRS in cats, current criteria for the diagnosis of SIRS in this species are defined as the presence of two or more of the following clinical signs: rectal temperature > 39.7 degrees C (>103.5 degrees F) or <37.8 degrees C (<100 degrees F); heart rate > 225 beats/minute or < 140 beats/minute; respiratory rate > 40 breaths/minute or hyperventilation as defined by decreased PaCO2; white blood cell count (WBC) >19,500/microliter or <5,000/microliter, or the presence of a left shift on the leukogram. For the purposes of the current study, a left shift was defined as a band neutrophil count > 1000 cells/microliter, or 10% of all neutrophils present in neutropenic patients.
Abnormalities in hemostasis and blood coagulation are well-recognized phenomena in laboratory animals, dogs, and humans with sepsis. All of these species often demonstrate hemostatic derangements associated with SIRS associated with sepsis as well as other causes, due to disruption of the normal balance between the prothrombotic, antithrombotic, and fibrinolytic pathways associated with the blood coagulation cascade.
In septic canines, PT (prothrombin time) and aPTT (activated partial thromboplastin time), which are measures of the competency of the tissue factor ("extrinsic") and common coagulation pathways, and the contact activation ("intrinsic") and common coagulation pathways, respectively, may be increased over their respective reference ranges. Likewise, also in septic dogs, fibrin[ogen] degradation products (FDPs; indicators of plasmin activity; plasmin lyses fibriongen and fibrin ) and D-dimer (a type of FDP that indicates recent fibrinolysis) levels will also be higher than their reference ranges. Unless they improve, subnormal antithrombin and protein C (a clotting inhibitory protein) activities in septic dogs are predictors of death.
In contrast, there is little information regarding the behavior of such hemostatic parameters in cats with sepsis/SIRS. In this prospective, observational, clinical study of 31 cats with sepsis (18 males, 13 of which were neutered and 5 sexually intact, and 13 females, 10 spayed and 3 intact) and 33 healthy control cats, the investigators' goal was to learn more about the clinical and laboratory findings in cats with naturally occurring sepsis, especially those associated with hemostasis, and to identify potential association of changes in these parameters with patient outcomes.
Median age of the sepsis group was 4.4 years (range, 0.3-13.0 years) and of the control group, 5.1 years (range, 0.7-14.5 years). The septic group included 26 domestic shorthair and 5 purebred cats, while the control group was composed of all domestic shorthair or domestic longhair cats. Those cats with sepsis included 10 cats with pyothorax, 7 with septic peritonitis, 5 with panleukopenia virus infection, 5 with bite wounds, 3 with abscesses and diffuse cellulitis, and one with pyometra. Clinical signs observed in the septic cats included dehydration (n = 21), lethargy (n = 21), anorexia (n = 18), mucosal pallor (n =15), and dullness (n = 15). Although the median rectal temperature for the sepsis group was normal at 38.1 degrees C (100.6 degrees F), almost half (14/31) were hypothermic, with rectal temperatures of < 37.8 degrees C (<100.0 degrees F). Seven of the cats were hyperthermic (rectal temperature > 39.7 degrees C, or > 103.5 degrees F). Leukocytosis was identified in 12/31 (39%) of the septic cats, and another 12/31 were found to be leukopenic. A left shift was present in 19/31 (61%) of the septic cats.
All 31 of the cats met at least two out of four of the feline sepsis criteria; 3/31 (10%) met all four criteria; 14 (45%) met three of the criteria, and 14 (45%) met two of the criteria. All of the septic cats had at least one leukogram abnormality. Infection was confirmed in the septic cats with one or more of the following findings: positive bacterial culture (9/31); cytologic examination (22/31); histopathologic examination (5/31); fecal parvovirus antigen serology (5/31). Positive bacterial cultures in the septic cats demonstrated Pasteurella multocida (5/31), Fusobacterium spp. (2/31), Prevotella spp. (1/31), and a mixed bacterial population (1/31).
Historical data were available for 27/31 septic cats, and of these, owners first identified clinical signs of illness in 13/27 up to two days prior to presentation; in 12/27, clinical signs of illness were first noticed 3-7 days prior to presentation; and in 2/27, clinical signs of illness were identified 8-14 days prior to presentation.
Results of blood smear evaluation were available for 26/31 septic cats: 23/26 (88%) had neutrophil cytoplasmic toxic changes (toxic changes were marked in 11/23); left shift (19/26; 73%; regenerative in 8/19, and degenerative in 11/19); lymphopenia (24/30; 80%); eosinopenia (20/27; 74%); leukocytosis (12/31; 39%); leukopenia (12/31; 39%); monocytosis (16/30; 53%); and reactive monocytes (7/26; 27%). Of the 19 cats which had a left shift, 8/19 had metamyelocytes and 7/19 had myelocytes. Abnormalities of the erythron in the septic cats included increased red blood cell distribution width (17/31; 55%); microcytosis (17/31; 55%); metarubricytosis (11/31; 35%), and anemia (9/31; 29%; always nonregenerative). Nonregenerative anemia has also been identified in septic cats in other studies and is most likely associated with chronic inflammation. Thrombocytopenia was present in 9/26 (35%) of the septic cats.
Serum biochemical abnormalities in the septic cats included: hyponatremia (30/31; 97%); hypochloridemia (24/31;77%); hypoproteinemia (22/30; 73%); hypertriglyceridemia (22/30; 73%); elevated creatine kinase activity (20/30; 67%); hypocalcemia (low total serum calcium, 19/30; 63%); hyperbilirubinemia (17/31; 55%); low urea (14/30; 47%); high AST activity (13/30; 43%); and hypocarbia (11/29; 38%). The high serum CK and AST activity levels in the septic cats were considered to be due to muscle damage, either directly from trauma such as bite wounds, or secondary to muscle ischemia, hypotension, shock, and hypoxia leading to rhabdomyolysis. Rhabdomyolysis in the septic cats was not verified by serum or urine myoglobin concentrations, however.
Abnormalities in hemostatic parameters were also found in the septic cats. The septic cats had significantly longer aPTT, lower median total protein C and antithrombin levels, and higher median D-dimer concentrations than the controls. However, although PT was often higher in cats with sepsis, the median PT did not differ significantly between case cats and control cats. Of the 31 septic cats, 22 had laboratory assessments of parameters required to evaluate all three categories of concern for identification of disseminated intravascular coagulation (DIC): clotting factor consumption, inhibitor consumption, and increased fibrinolysis. Four of those 22 cats were diagnosed with DIC, a syndrome of severe imbalance between blood clotting mechanisms and fibrinolytic processes. During the 30-day study period, a diagnosis of DIC was not associated with death in the cats in which it was diagnosed.
In the septic group, 19/30 (63%) survived the 30-day study period. Of the 11 that died, four were euthanized at the owners' request due to deterioration. Among those case cats who died, factors positively associated with death included monocytosis; severe toxic cytoplasmic changes in neutrophils; hypothermia; high serum concentrations of urea, phosphorus, creatinine, and bilirubin; low total protein C activity; and hyponatremia. Eosinopenia and presence of pleural effusion were negatively associated with death. Neither neutrophilia nor neutropenia was associated with death in the study cats. However, final multivariate analysis determined that none of the foregoing associations were significant. Also, there were no differences in mortality rates among the septic cats based on the underlying cause of the sepsis.
The limited size of the study cohort and the relatively low mortality rate of the patients decreases the power of the statistical analyses of the data. Because of this, the authors recommend further investigation of larger numbers of septic cats to determine if concurrent neutrophil cytoplasmic toxicity and left shift is a negative prognostic factor in these animals. Also, laboratory evaluation in the patients was only performed at the time of initial hospitalization. Serial measurements to evaluate the trends over the entire disease course and their association with recovery or death would be more useful and enlightening, but obtaining multiple blood samples from critically ill patients, many of whom are anemic, is neither ethical nor practical. Other parameters recommended by the authors for evaluation in septic cats include arterial blood gas measurements, blood lactate concentration, and viscoelastic coagulation testing, which is a more sensitive method of detecting early hypercoagulability than conventional blood coagulation testing in critical care patients. [PJS]
Declue AE, Delgado C, et al. Clinical and immunologic assessment of sepsis and the systemic inflammatory response syndrome in cats. J Am Vet Med Assoc 2011;238:890-7.