This research was performed at the Clinic of Small Animal Medicine, LMU University of Munich, Veterinaerstrasse 13, 80539 Munich, Germany. Some of the results were presented at the ACVIM Congress, Montreal, Canada, June 3–6, 2009.
Prognostic Factors in Cats with Feline Panleukopenia
Version of Record online: 12 OCT 2010
Copyright © 2010 by the American College of Veterinary Internal Medicine
Journal of Veterinary Internal Medicine
Volume 24, Issue 6, pages 1271–1276, November/December 2010
How to Cite
Kruse, B.D., Unterer, S., Horlacher, K., Sauter-Louis, C. and Hartmann, K. (2010), Prognostic Factors in Cats with Feline Panleukopenia. Journal of Veterinary Internal Medicine, 24: 1271–1276. doi: 10.1111/j.1939-1676.2010.0604.x
- Issue online: 3 NOV 2010
- Version of Record online: 12 OCT 2010
- Submitted November 23, 2009; Revised July 6, 2010; Accepted July 29, 2010.
- Feline parvovirus infection;
- Feline parvovirus;
Background: Feline panleukopenia is a highly contagious and often lethal disease.
Objective: The purpose of the study was to identify prognostic factors for survival of cats with panleukopenia.
Animals: Between 1990 and 2007, 244 cats were diagnosed with panleukopenia in the Clinic of Small Animal Medicine, LMU University of Munich, Germany. Diagnosis was established by electron microscopy, polymerase chain reaction of feces or blood, antigen ELISA of feces, pathognomonic histopathological lesions at necropsy, or some combination of these procedures.
Methods: Medical records of each cat were evaluated retrospectively.
Results: Survival rate was 51.1%. No significant correlation was found between outcome and living conditions, age, vaccination status (unvaccinated versus one or more vaccines administered), or severity of clinical signs. However, of the vaccinated cats, none had received a vaccine later than 12 weeks of age as a kitten. Nonsurvivors had significantly lower leukocyte and thrombocyte counts at presentation compared with survivors. The relative risk of death for patients with <1,000/μL leukocytes was 1.77 times as high as in patients with a leukocyte count of 1,000–2,500/μL (P= .038), and 1.85 times as high as in patients with >2,500/μL leukocytes (P= .001). The likelihood of a fatal outcome was higher when serum albumin concentration was <30 g/L or serum potassium concentration <4 mmol/L.
Conclusions and Clinical Importance: Vaccination strategies that do not include vaccination of kittens beyond 12 weeks of age may not be adequate to prevent panleukopenia. Leukopenia, thrombocytopenia, hypoalbuminemia, and hypokalemia are negative prognostic factors in cats with panleukopenia.
disseminated intravasal coagulopathy
feline leukemia virus
feline immunodeficiency virus
polymerase chain reaction
Statistical Package for the Social Sciences
Feline panleukopenia, caused by the single-stranded DNA-containing feline parvovirus (FPV), is a highly contagious disease affecting all members of Felidae.1 Severity of clinical signs depends on age, immune status, and concurrent infections.2 Clinical disease ranges from subclinical infection to a peracute syndrome with sudden death. Typical initial signs include fever, lethargy, and anorexia.3 Affected cats initially may present for vomiting and, with lower frequency, develop watery to hemorrhagic diarrhea. Patients die from complications associated with secondary bacterial infection, sepsis, dehydration, and disseminated intravasal coagulopathy (DIC).4 The highest morbidity and mortality occur in kittens up to 12 months of age.5,6 Mortality is 25–90% in acute panleukopenia and up to 100% in peracute infections.7 Compared with dogs, limited data have been published concerning prognostic factors for survival. In canine parvovirus (CPV) infection, marked leukopenia, lymphopenia, monocytopenia, and eosinopenia are indicators of a poor prognosis.8 In addition, dogs have a worse prognosis if their serum cortisol concentrations fail to normalize.9 In addition, serum thyroxine10 and cholesterol11 concentrations are significantly lower in nonsurvivors. Systemic inflammatory response syndrome (SIRS) develops more frequently among nonsurviving dogs4 and concurrent viral, bacterial, or parasitic infection may worsen the prognosis.12 Young dogs also have a worse prognosis.13 Despite ongoing research in FPV infection, limited data are available on factors influencing outcome, including living conditions, transmission of infection, vaccination status, and prognostic factors in naturally infected cats with feline panleukopenia. Therefore, the aims of this study were to provide data concerning history, signalment, clinical, and laboratory findings of cats with feline panleukopenia and to assess clinical outcome and prognostic factors for survival.
Materials and Methods
Between 1990 and 2007, 244 cats with a diagnosis of panleukopenia were identified from the records of the Clinic of Small Animal Medicine, LMU University of Munich, Germany. Inclusion criteria for this study were a definitive diagnosis by positive electron microscopy (n = 182), positive parvovirus antigen (canine parvoviral SNAP ELISAa) in feces (n = 6), positive polymerase chain reaction (PCR)14 of feces (n = 1) or blood (n = 8), or histopathological lesions consistent with FPV at necropsy (n = 88). Cats were excluded if they had received a vaccination against panleukopenia within 3 weeks before admission. Data concerning signalment (breed, age, sex), living conditions (multi- versus single-cat household, indoor versus outdoor cat), and vaccination status were obtained. Patients were categorized as “unvaccinated” if they had never received any vaccine against panleukopenia. Cats were categorized as “one or more vaccines administered” if they had received one or more vaccines against panleukopenia during any time of their lives. Clinical signs were recorded: presence of diarrhea versus absence of diarrhea; presence of hemorrhagic diarrhea versus nonhemorrhagic or no diarrhea; presence of vomitus versus absence of vomitus; and presence versus absence of fever, anorexia, apathy. Furthermore, laboratory variables (CBC, serum biochemistry profile, feline immunodeficiency [FIV], and leukemia virus [FeLV] status), and clinical outcome were recorded. To evaluate potential prognostic factors, the following variables were compared between survivors and nonsurvivors: age, breed, sex, living conditions, vaccination status, clinical signs, FIV and FeLV status, and laboratory variables at presentation.
Statistical analysis was performed using SPSS 16.0 (http://www.spss.com) and StatCalc (http://www.epiinfo.com). Differences in continuous variables of illness characteristics (leukocyte count, bands, lymphocytes, platelets, hematocrit, and serum concentrations of potassium, albumin, bilirubin, glucose, and ALT activity at day 1 as well as age of patients) between cats with different disease outcome (survivors versus nonsurvivors) were analyzed using the nonparametric test Mann-Whitney U-test, because most of the data were not normally distributed. Associations between categorical variables of illness characteristics (breed, sex, housing conditions, vaccination status [unvaccinated versus one or more vaccines administered], diarrhea [presence versus absence], hemorrhagic diarrhea [presence versus nonhemorrhagic or no diarrhea], presence versus absence of vomitus, anorexia, apathy, fever) and disease outcome (survivors versus nonsurvivors) were tested using a Pearson's Chi-square test. If an expected number in one of the cells in the contingency table was <5, the Fisher's exact test was used. For variables with significant differences between survivors and nonsurvivors, an ROC analysis was performed to identify the best cut-off value. These variables then were recoded into categorical variables, and relative risks were calculated. Differences were considered significant at P < .05.
Ninety percent of the cats (219/242) diagnosed with panleukopenia were domestic shorthair cats, 5.5% were Persians (13/242), 2.4% were mixed breeds (6/242), and 0.8% were Siamese (2/242). In addition, 1 each of Burmese and British Shorthair was included. There were 59.5% males (45/142 neutered) and 40.5% females (15/95 spayed). Median age was 4 months (range, 2 weeks to 14.5 years). Most cats (n = 174) were <1 year of age, 56.7% of cats were <6 months, and 10.7% were >5 years. Of the 184 cats in which vaccination status was known, 60.3% (111/184) were never vaccinated. The other 39.7% (73/184) had received at least 1 vaccine. Of these, 79.5% (58/73) had only received 1 vaccine before the age of 12 weeks; 11.0% (8/73) had received 2 vaccines before the age of 12 weeks. Another 9.5% (7/73) had received 2 vaccines before the age of 12 weeks and in addition had received several vaccines on a yearly basis. None of the cats had received a vaccine later than 12 weeks of age as a kitten.
Of the never vaccinated cats, 54.1% were nonsurvivors (60/111). Of the cats that had only received 1 vaccine before the age of 12 weeks, 58.6% were nonsurvivors (34/58). Of the cats that had received 2 vaccines before the age of 12 weeks, 12.5% were nonsurvivors (1/8). Of the cats that had received 2 vaccines before the age of 12 weeks and in addition had received several vaccines on a yearly basis, 71.4% were nonsurvivors (5/7). Housing conditions, laboratory results, and clinical abnormalities are presented in Tables 1 and 2.
|Urban private household||21||18.1|
Nine cats with negative blood PCR had positive electron microscopy results, and 3 cats with negative fecal antigen ELISA results had positive electron microscopy results and findings consistent with FPV at necropsy. Concerning the outcome, 51.1% (n = 118) of cats were able to be discharged and 48.9% (n = 113) died or were euthanized. Median time of hospitalization in survivors was 7 days and median time to death or euthanasia was 2 days.
The leukocyte count at presentation was significantly lower in nonsurvivors compared with survivors (Table 3). The relative risk of death for patients with leukocyte counts <1,000/μL was 1.77 times as high as for patients with leukocyte counts of 1,000–2500/μL (P= .038), and 1.85 times as high as for patients with leukocyte counts >2,500/μL (P= .001). According to the ROC analysis, the best cut-off for low platelets as a prognostic indicator was 135,000/μL. The risk of a lethal outcome for patients with thrombocyte numbers <135,000/μL at presentation was 2.60 times as high as for patients with platelet counts ≥135,000/μL. Additional cut-offs identified in the ROC analysis were 30 g/L serum albumin concentration and 4 mmol/L serum potassium concentration. The relative risk of death for patients with serum albumin concentration <30 g/L (P= .006) and serum potassium concentration <4 mmol/L (P < .001) was 2.90 and 2.48 times as high, respectively, compared with cats above these cut-offs.
|Albumin||32.0 g/dL||17.0–44.4 g/dL||27.1 g/dL||15.6–39.8 g/dL||.014|
|Potassium||4.71 mmol/L||2.80–6 mmol/L||4.05 mmol/L||2.5–9.4 mmol/L||.010|
All of the other laboratory variables (bands, lymphocytes, hematocrit, and concentrations of bilirubin, glucose, and ALT activity at day 1, and FIV and FeLV status) were not associated with disease outcome. Furthermore, age, breed, sex, vaccination status (unvaccinated versus one or more vaccines administered), living conditions (eg, indoors versus outdoors; single versus multicat household), and clinical signs (presence versus absence of diarrhea, presence of hemorrhagic diarrhea versus nonhemorrhagic or no diarrhea, presence versus absence of vomitus, anorexia, apathy, fever) were not associated with disease outcome.
This study demonstrates that leukocyte and thrombocyte counts as well as serum albumin and potassium concentrations at presentation are prognostic indicators in cats with panleukopenia, whereas vaccination status, age, clinical signs, and housing conditions are not.
The findings regarding signalment reported in this study are similar to those of previous studies of canine parvovirosis.15 Most affected cats were <1 year of age: median age was 4 months and 56.7% of the affected cats were <6 months of age. However, 25.3% were >1 year of age and 10.7% were > 5 years of age. This finding is interesting in light of the perception that older cats are protected either by widespread vaccination or earlier subclinical infection. In 1 study, the prevalence of antibodies against FPV in young domestic cats from Costa Rica was 92.8%; only 16.5% of them were vaccinated previously.16 Reasons for the high infection rate in older cats in this study may include inadequate immunization and lack of virus exposure in the environment (eg, high percentage of indoor cats, resulting in a lower immunity). In contrast to Costa Rica, the present study was performed in Munich, a German city in which many cats are kept strictly indoors.
A study from the United Kingdom reported that 25% of all kitten mortality was due to FPV.6 In CPV infection, the older the dog, the more favorable the prognosis.13 In this study, there was no significant correlation between age and severity of clinical signs and outcome. Based on this study, young cats are more susceptible to feline panleukopenia, but do not have an increased risk of death, which contradicts the idea that the highest morbidity and mortality occur in kittens up to 12 months of age.5,6
In this study, indoor cats were overrepresented compared with outdoor cats, and 14.5% had no contact with other cats indicating that indirect transmission is an important mode of infection. Transmission can occur via contaminated clothing, cages, and insect vectors.17 In this study, there was no significant difference in housing conditions (indoors versus outdoors) between unvaccinated cats and cats that were administered ≥1 vaccines. It can be assumed that natural contact with parvovirus is less likely in indoor cats, thus their chance to build protective immunity because of inapparent infection is decreased.18
In this study, 39.7% of cats with panleukopenia had received ≥1 vaccines and despite this fact developed the disease. However, none of the cats that received vaccines would be considered to be completely vaccinated according to current guidelines, which advise administration of vaccines every 3–4 weeks from 6 weeks of age through at least 16 weeks of age.19,20 The lack of protective antibody titers in the face of vaccination may be because of interference with maternal antibodies, which can last up to 20 weeks of age.5,19,20–22 Lack of or incomplete vaccination is a significant risk factor for development of CPV enteritis.15 Additional reasons for an inadequate increase in protective antibody titers in this population might be vaccination of immunocompromised or subclinically infected patients, administration of less immunogenic inactivated vaccines, or infection with new parvovirus strains not neutralized by vaccine-induced protective antibodies.
No significant association between vaccination status and outcome or duration of hospitalization could be identified. This is an unexpected finding because it was assumed that patients that received at least 1 vaccine would at least be more likely to recover if they developed the disease. In this retrospective study, however, there were no cats that would be considered to have been adequately vaccinated according to current guidelines. Therefore it is possible that vaccination based on current guidelines19,20 may result in better outcomes or less severe clinical signs in infected cats.
In this study, 69.3% of patients had diarrhea and 62.7% had vomiting. However, 30.7% never developed diarrhea and 37.3% never had vomiting. Some of these cats may have had a peracute form of the disease and died before gastrointestinal signs occurred. However, 10.2% never developed diarrhea or vomiting despite a prolonged course of illness and 54.2% had a favorable outcome. Interestingly, 34.2% of the cats never developed leukopenia during the course of their disease, and 15 patients had neither gastrointestinal signs nor leukopenia. In these cats, a parvovirus infection was suspected either because of known contact with a cat with FPV or the disease was diagnosed at necropsy. Cats with normal or increased leukocyte counts may have had leukopenia before testing or alternatively may have been tested before leukopenia occurred. A peracute infection also could have been present in these cats. Although cats were excluded if vaccinated against panleukopenia in the 3 weeks before admission, a vaccine-induced false-positive result for FPV cannot be totally excluded in cats with unknown vaccination history that were not diagnosed at necropsy.
Bloody diarrhea may develop as consequence of severe enteritis with serious mucosal damage, DIC, or coinfections with other organisms. In CPV infections, giardiasis exacerbates the clinical syndrome.23 In kittens, coexisting salmonellosis has been reported.2 Hemorrhagic diarrhea, a typical sign of CPV enteritis, was only present in 14.1% of our cats. Canine intestinal crypts may be more susceptible to viral destruction and serious mucosal damage compared with feline enterocytes. Potentially, local intestinal immune defense (eg, Peyer's patches) may be more effective in cats than in dogs.
Interestingly, in this study, no clinical sign was significantly related to outcome as described in dogs.24 However, previous studies have identified clinical predictors for lethal outcome in dogs and cats (eg, presence of SIRS,4 sepsis, and hypothermia7).
FPV and CPV infection lead to bone marrow suppression typically resulting in leukopenia.25 This study confirms that severity of leukopenia parallels that of clinical illness and risk of death, similar to results of previous studies in dogs.8 In most dogs, neutrophil nadir and most severe clinical illness occur concurrently.26 Presumably death of cats also would occur during this severe state of clinical illness and neutrophil nadir. Experimental infections of cats using CPV-2b caused only slight leukopenia but marked lymphopenia, similar to CPV infection in dogs.27 Of 137 cats, 38.7% had lymphopenia. In contrast to total leukopenia, lymphopenia was not associated with disease outcome. Lymphopenia with concomitant neutropenia was observed more often than solitary lymphopenia (27.7% versus 11.0%). Possibly, some cats of this study were infected with CPV rather than with FPV. Ideally, it should have been investigated whether CPV instead of FPV was the infectious agent in some cases. However, samples from the cats were no longer available.
In this study, the 2nd most commonly observed hematologic abnormality, exclusive of leukopenia, was thrombocytopenia attributed to megakaryocyte destruction or increased consumption caused by DIC.13 DIC generally is associated with a high case fatality rate.28 In most cats, thrombocytopenia was not prominent (median in survivors, 260,000/μL; median in nonsurvivors, 195,000/μL at presentation). However, a lower thrombocyte count was significantly correlated with negative outcome. This finding is in contrast to a study of dogs with CPV that found no significant association between thrombocytopenia and outcome.11 Furthermore, in a study of hypercoagulability in dogs with CPV, all dogs had hypercoagulability, but none had measurable D-Dimers or thrombocytopenia.29 A low thrombocyte count also may reflect severity and more advanced stage of disease because of severe viral megakaryocyte destruction, and thus may explain its role as a negative prognostic factor as detected in the present study. Additional studies would be necessary to investigate if parvovirus has more cell tropism for megakaryocytes in cats compared with dogs or if cats have a higher risk for developing DIC in severe systemic infectious diseases compared with dogs.
In this study, anemia was not associated with negative outcome, an unexpected finding considering that gastrointestinal blood loss, coinfections (4.8% of cats were infected with FeLV), serious bone marrow suppression, or sepsis-associated anemia of inflammatory disease30 represent severe and serious complicating factors. Because of the relatively long life span of erythrocytes, marked anemia is less common in CPV and FPV, unless intestinal blood loss is severe, but nonregenerative anemia can be seen with FPV infection.31,32 Mild anemia may have been masked by severe dehydration in this study.
Hypoalbuminemia was the most frequent biochemical abnormality, resulting from decreased protein intake and leakage into the gastrointestinal tract because of severe mucosal lesions. In dogs, hypoalbuminemia (<20 g/L) represents a risk factor for negative outcome in enteropathies.33 Serum albumin concentration <30 g/L at presentation was associated with a negative outcome in this population of cats, and can be explained by the association between hypoalbuminemia and decreased plasma colloid osmotic pressure, which diminishes effective perfusion at the capillary level, possibly leading to DIC, organ failure, and death.
A serum potassium concentration <4 mmol/L at presentation was associated with negative outcome. On the contrary, low serum potassium concentration is not associated with outcome in CPV infection.34 However, dogs with CPV had significantly lower serum concentrations of potassium than did controls in another study.35 Hypokalemia can be explained by anorexia, vomiting, increased gastrointestinal potassium losses, fluid therapy, or possible refeeding syndrome, and most likely reflects the severity of the enteritis.
Limitations of this study include inaccuracy in history that is associated with any retrospective study. Exact details (eg, types of vaccines given) could not be obtained from incomplete medical records. Because most patients were treated with a very similar treatment protocol, an effect of therapy on disease outcome could not be evaluated. One major limitation is that false positive diagnoses by PCR or ELISA cannot be excluded in all cases. Positive PCR or ELISA results can occur after vaccination for approximately 3 weeks.36,37 Although cats were excluded from this study when it was known that vaccines against panleukopenia had been administered within 3 weeks before admission, some of these cats may have been inadvertently included because the date of vaccination was not known in all cats. Finally, in some patients, complete blood analysis could not be performed on the day of presentation (eg, presentation on weekends or death before blood sampling). Coagulation profiles and fibrin degradation products were not performed in every patient, potentially resulting in underestimation of DIC.
In conclusion, feline panleukopenia represents an infectious disease with a high mortality rate. Age, living conditions, and clinical signs were not associated with illness severity and outcome whereas leukocyte and thrombocyte counts as well as serum albumin and potassium concentrations did represent prognostic factors in feline panleukopenia. Vaccination strategies that do not include vaccination of kittens beyond 12 weeks of age may not be adequate to prevent panleukopenia.
a IDEXX, Ludwigsburg, Germany.
This work was performed at the Clinic of Small Animal Medicine, LMU University of Munich, Germany, without additional funding. The authors thank Dr J. McGill, currently Internal Medicine Resident at University of Georgia, Athens, Georgia, for her input concerning English language and editorial matters.
- 1Viral diseases. Panleukopenia. In: HolzwurmJ, ed. Diseases of the Cat: Medicine and Surgery. Philadelphia, PA: WB Saunders; 1987:182–193.
- 7Feline panleukopenie. In: HartmannK, HeinJ, eds. Infektionskrankheiten der Katze. Hannover, Germany: Schlütersche; 2008:87–98.,
- 12CVT update: Diagnosis and treatment of parvovirosis. In: KirkRW, ed. Current Veterinary Therapy XIII. Philadelphia, PA: WB Saunders; 2000:629–632.
- 13The leucocyte status of surviving dogs as compared to dogs dying from parvovirus infections. Tierarztl Prax 1987;15:409–415.,
- 27Canine Parvovirus. Recent Advances in Canine Infectious Diseases, 2000. http://www.ivis.org.