SEARCH

SEARCH BY CITATION

Keywords:

  • Albumin;
  • Central nervous system;
  • C-Reactive protein;
  • α2-Macroglobulin

Abstract

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

Background: Measurement of concentrations of acute-phase proteins (APPs) is used as an aid in the diagnosis of a variety of diseases in animals.

Objective: To determine the concentration of APPs in dogs with steroid responsive meningitis-arteritis (SRMA) and other neurologic diseases.

Animals: One hundred and thirty-three dogs with neurologic diseases, 6 dogs with sepsis, and 8 healthy dogs were included in the study. Thirty-six dogs had SRMA (31 of which had monitoring), 14 dogs had other meningoencephalitides (ME), 32 had disk disease (IVDD/DLSS), 26 had tumors affecting the central nervous system (TCNS), and 25 had idiopathic epilepsy (IE).

Methods: Prospective, observational study: C-reactive protein (CRP), α2-macroglobulin (AMG), and albumin concentrations were determined in the serum or plasma. CRP was also measured in the cerebrospinal fluid.

Results: Serum CRP was significantly higher in dogs with SRMA (inline image= 142 μg/mL ± 75) and sepsis (inline image= 114 μg/mL ± 67) in comparison with dogs with other neurologic diseases (inline image= 2.3–21 μg/mL; P < .001). There was no significant difference detected in AMG between groups. Serum albumin concentration was significantly lower (P < .01) in dogs with SRMA (inline image= 3.2 g/dL ± 0.41) than in other groups (inline image= 3.6–3.9 g/dL). Serum CRP concentration of SRMA dogs correlated with alkaline phosphatase levels (r= 0.515, P= .003).

Conclusions and Clinical Importance: CRP concentrations in serum are useful in diagnosis of dogs with SRMA. Serum CRP could be used as a monitoring parameter in treatment management of these dogs.

Acute-phase proteins (APPs) are part of the acute-phase reaction, a phylogenetically old component of the nonspecific innate immune system.1 The acute-phase response is generated by a variety of systemic diseases, such as inflammation, neoplasia, or trauma.1 The measurement of these proteins has been used for diagnostic purposes during the last 5 decades in humans and increasingly in veterinary medicine.1,2 For example, the determination of the APP C-reactive protein (CRP) concentration in serum is used to differentiate between viral and bacterial infections in humans, especially in patients with meningitis and pneumonia.3,4 CRP is a major APP in humans and dogs, with a significant increase within an acute-phase response.3,5 Indeed, a markedly elevated serum CRP concentration was found in infectious and immune-mediated diseases in dogs.6,7

The increase of one major APP is not disease specific. Therefore, it is recommended to measure an APP profile including positive and negative APPs, to improve the specificity of this diagnostic finding in respect of a special disease.2,8

α2-macroglobulin (AMG) is increased in experimental inflammation in mice9 and in neoplastic disease in humans,10 whereas AMG concentration is significantly diminished in patients with sepsis.11 Experimentally induced inflammation in dogs causes a reduction of AMG.12 Albumin is known to decrease in acute-phase response in dogs.2,5,12

Steroid responsive meningitis-arteritis (SRMA) is one of the most commonly diagnosed meningitides in dogs.13 The disease consists of an inflammation of the meninges and meningeal arteries.14,15 The diagnosis is confirmed by typical clinical signs such as cervical rigidity, fever, leucocytosis, and pleocytosis with polymorphonuclear cells in the cerebrospinal fluid (CSF), and elevated IgA levels in serum and CSF.16,17 Treatment is monitored by repetitive CSF examinations.16 The disadvantage of this diagnostic procedure is the need for general anesthesia for CSF collection. Therefore, it would be useful to find a less invasive method for monitoring the efficacy of treatment in dogs with SRMA.

The purpose of this study was to determine concentrations of APPs in serum and CSF of dogs with SRMA in comparison with dogs with other neurologic diseases, and to evaluate their usefulness for diagnosis and during monitoring of treatment efficacy of SRMA.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

Study Population

The study population was 133 dogs with varying CNS diseases, and 6 dogs with a diagnosis of sepsis, that were a positive control group, which were examined to the Department of Small Animal Medicine and Surgery, University of Veterinary Medicine Hannover, Germany, between September 2004 and August 2007. Additionally, 8 healthy Beagles served as negative controls (animal experiment number 33.42502/05-12.05). The dogs were categorized as healthy on the basis of a complete physical examination, electrocardiography,a and blood biochemistry.

Of the dogs examined, 36 had SRMA, 14 had other meningoencephalitides (ME), 32 had intervertebral disk disease (IVDD) or degenerative lumbosacral stenosis (DLSS), 26 had primary or secondary tumors affecting the central nervous system (TCNS), and 25 had idiopathic epilepsy (IE). Thirty-one of the SRMA dogs were also examined for monitoring treatment efficacy. The treatment schedule used was as described.16 The monitoring period was 1–28 months (mean: 5.4 months, 1–10 controls, respectively). Repeat examinations were performed 4, 8, 12 and 16 weeks after the 1st examination. Five dogs with SRMA suffered a relapse of SRMA during the period of study.

SRMA was diagnosed on the basis of characteristic clinical signs, CSF cell count (≥ 8/3/μL) with neutrophilic pleocytosis, and IgA concentrations ≥0.2 μg/mL in CSF and ≥100 μg/mL in serum.17 In dogs with other ME, IVDD or DLSS, tumors affecting the CNS, and IE, diagnostic procedure consisted of physical and neurologic examination, complete blood examinations, magnetic resonance imaging, and examination of CSF. In cases of canine distemper virus—encephalitis, histopathology, and immunohistochemistry were performed. Furthermore, diagnosis was confirmed by surgery in 23 cases with IVDD and DLSS, and histopathology in 12 dogs with tumors affecting the CNS (2 meningiomas, 2 hemangiosarcomas and 1 nephroblastoma, malignant histiocytosis, ganglioneuroma, anaplastic ependymoma, anaplastic astrocytoma, malignant blastoma, osteosarcoma, or metastatic adenocarcinoma each). Dogs with sepsis were diagnosed on the basis of clinical signs, blood work, and positive bacterial culture.

Sample Collection and Handling

Blood was collected by venipuncture. A complete blood cell count, blood biochemistry analysis (including serum-alkaline phosphatase, total protein, and albumin) was performed using routine methods,b and in dogs with suspected SRMA, serum IgA concentration (enzyme linked immunosorbent assay, ELISA17) was also determined.

CSF was collected under general anaesthesia by suboccipital puncture with the dog in lateral recumbency and was analyzed for total nucleated cell count, cell differentiation, total protein concentration, IgA, and CRP. CSF was considered normal with a total nucleated cell count <8 cells/3/μL, a total red blood cell count <12,000 cells/3/μL,18 a total protein concentration <25 mg/dL, and an IgA content <0.2 μg/mL.

For measurement of CRP concentration, serum and CSF samples were separated into aliquots. For measurement of AMG, citrated plasma (9 parts of blood are mixed with 1 part of 0.11 mol/L tri-sodium citrate solution, centrifuged for 20 minutes at 2,000 ×g) was used as sample material. All samples were stored frozen at −20 °C until analysis. The numbers of samples measured are described in Table 1.

Table 1.   Number of samples measured for each variable in each disease.
VariableSRMA (n = 36)ME (n = 14)IVDD/DLSS (n = 32)TCNS (n = 26)IE (n = 25)Normal (n = 8)Sepsis (n = 6)
  1. SRMA, steroid responsive meningitis-arteritis; ME, other meningoencephalitides; IVDD/DLSS, intervertebral disk disease/degenerative lumbosacral stenosis; TCNS, tumors affecting the central nervous system; IE, idiopathic epilepsy; s-CRP, C-reactive protein in the serum; CSF CRP, C-reactive protein in the cerebrospinal fluid; AMG, α2-macroglobulin; Alb, albumin.

s-CRP361232262586
CSF CRP351332262583
AMG171132212586
Alb311430232486

CRP

CRP was measured in serum and CSF samples with a commercially available ELISA kit.c As recommended by Cerón et al,2 a laboratory-specific reference range was created by measuring serum samples of the 8 healthy Beagles. The detection limit was determined by making a serial dilution of serum of a known concentration. The lowest concentration with an optical density (OD), which was twofold over the background OD was considered to be the detection limit of the assay (DL: 0.1 μg/mL).

AMG

The concentration of AMG was measured using a commercial chromogenic substrate assay kit.d

Diluted plasma is mixed with excess trypsin, and AMG becomes complexed with trypsin. The remaining, noncomplexed trypsin is inhibited with soybean trypsin inhibitor. The trypsin complexed with AMG is still able to cleave small substrates and releases p-nitroaniline (pNA) from a suitable chromogenic peptide substrate. The pNA concentration, which is measured photometrically, is proportional to the AMG concentration.19

The test was performed manually using a 96-well polystyrene microtiter plate. The test procedure followed the method description of the manufacturer with the following exceptions: Test samples were diluted 1 : 80 instead of 1 : 160 due to the lower AMG activity in canine plasma when compared with human plasma. The lower sample dilution was also considered for the preparation of standards. Dilutions of a canine pooled plasma from equal aliquots from the citrated plasma of 100 healthy dogs were used as standards. Standard curves were prepared on the basis of the results of standards with the following activities: 200, 150, 125, 100, 75, 50, 25, and 0%.

Statistical Analysis

Arithmetic means and medians were calculated using routine descriptive statistical procedures.

Statistical evaluation was performed by computer software.e Data were analyzed using the 1-way-ANOVA approach with grouping of dogs as classification factor and mean values (CRP, AMG, albumin) as dependent variables. The effect of age was tested as a covariate. Levene's test was used to assess the equality of variance in different samples. The P-value was highly significant (P<.001), therefore the null hypothesis (no difference between neurological diseases) of equal variances was rejected and the level of significance of the ANOVA was lowered to 1%. R2 proved that the total variance of parameters can be explained by difference of groups of diseases; R2 indicates the goodness of model fit. Multiple comparisons between the groups of diseases were performed by tests following Scheffé for unbalanced models. An error probability of .01 (P) was used as the significance level for ANOVA and Scheffé's test.

An unpaired t-test to compare concentrations of serum CRP and alkaline phosphatase between SRMA dogs with and without glucocorticosteroid pretreatment and between dogs with relapse and unremarkable recovery during monitoring of treatment efficacy of SRMA was performed.

Analyses of associations of blood and CSF values (serum CRP, CSF CRP, CSF total nucleated cells, CSF neutrophils, CSF erythrocytes, CSF total protein, blood leucocytes, serum total protein, serum albumin, serum alkaline phosphatase, CSF IgA, serum IgA) within the SRMA group were performed by calculating Pearson's correlation coefficients with a significance level of P < .05.

Boxplots were used to visualize minimal and maximal values, median of the values. The box contains the middle 50% of the sample values.

Results

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

Breeds, Age, Sex, and Weight

In the presented SRMA population Jack Russell Terriers were affected in a proportion similar to Boxers and Beagles (Table 2).16,20

Table 2.   Breed distribution of 133 dogs with neurological diseases.
BreedSRMA (n = 36)ME (n = 14)IVDD/DLSS (n = 32)TCNS (n = 26)IE (n = 25)
  1. SRMA, steroid responsive meningitis-arteritis; ME, other meningoencephalitides; IVDD/DLSS, intervertebral disk disease/degenerative lumbosacral stenosis; TCNS, tumors affecting the central nervous system; IE, idiopathic epilepsy.

Beagle511
Boxer51 
Jack Russell Terrier431
Dachshund5
Westhighland White Terrier412 
Golden Retriever222
Labrador211
Cocker Spaniel21
Doberman2
German Shepherd Dog13
Mix667104
Further breeds (n = 1)Bernese Mountain Dog, German Pinscher, Nova Scotia Duck Tolling Retriever, Griffon, Weimaraner, Brandlbracke, Miniature Pinscher, German Shorthair Pointer, German Wirehaired PointerAiredale TerrierBernese Mountain Dog, Russian Terrier, Malinois, Bobtail, Small Munsterlander, Cairn Terrier, Hovawart, Miniature Schnauzer, Briard, Fila BrasileiroSaluki, Irish Wolfhound, Barsoi, Rottweiler, Greyhound, Border Collie, Giant Schnauzer, Small Munsterlander, Mastino, Appenzeller Mountain Dog, Bullmastiff, MalinoisPyrenean Mountain Dog, Cairn Terrier, Eurasier, Great Dane, Scottish Terrier, Australian Shepherd, Groenendal, Hovawart, Swiss Mountain Dog, Gordon Setter, Brandlbracke, Magyar Viszla, Yorkshire Terrier, Border Collie, French Bulldog, German Wirehaired Pointer

Dogs with SRMA had a median age of 10 months (25–75% quantile range, 8–14.5 months), and were younger than the dogs with other neurologic diseases (other ME: median 42.5 months, 21–74.3 months; IVDD/DLSS: 87 months, 44.5–116 months; tumors affecting the CNS: 96 months, 70–115 months; IE: 39 months, 21–68.5 months, respectively). The SRMA group consisted of 13 females, 1 spayed female, and 22 males. The median weight of the SRMA group was 19 kg (25–75% quantile range 10–27 kg; other ME: 11.6 kg, 8.1–19 kg; IVDD/DLSS: 25.7 kg, 9–39.2 kg; tumors affecting the CNS: 27.5 kg, 16.8–35.6 kg; IE: 28.5 kg, 12.8–37.5 kg, respectively).

Serum CRP Concentrations

The laboratory-specific reference range for serum CRP was 0.3–0.9 μg/mL. Serum CRP was significantly higher in dogs with SRMA than in dogs with other neurologic diseases and negative controls (meanSRMA: 142 μg/mL ± 75,P<.001), but did not differ from values of dogs with sepsis (meanSepsis: 114 μg/mL ± 67, P= .941; Fig 1). Within the group of dogs with SRMA, CRP was not significantly different between dogs pretreated with glucocorticosteroids and dogs that were not treated with glucocorticosteroids before diagnosis (meanpretreated=163 μg/mL ± 98, meannon-pretreated=131 μg/mL ± 55, P= .242).

image

Figure 1.  Serum C-reactive protein (CRP) concentrations in the groups of dogs at the time of diagnosis (steroid responsive meningitis-arteritis: SRMA n = 36; other meningoencephalitides: ME n = 12; intervertebral disk disease/degenerative lumbosacral stenosis: IVDD/DLSS n = 32; tumors affecting the central nervous system: TCNS n = 26; idiopathic epilepsy: IE n = 25; healthy controls: normal n = 8, dogs with sepsis: sepsis n = 6); boxplots: median of values, minimal and maximal values, boxes contain the middle 50% of sample values. Groups SRMA, sepsis differed significantly from other groups (P < .001; ★).

Download figure to PowerPoint

There was no effect of age on serum CRP values.

CSF CRP Concentrations

The healthy Beagles had CRP concentrations in CSF that were below the detection limit. Therefore, a nondetectable CRP concentration (<0.1 μg/mL) in the CSF was considered to be normal. All the dogs of the groups of IVDD/DLSS and IE had nondetectable CRP values in the CSF. Six dogs had undetectable CRP in CSF. In 29 dogs, the values ranged from 0.19 up to 13.6 μg/mL. In SRMA, CRP values in the CSF were significantly higher than in the groups disk disease, CNS tumors, and IE (meanSRMA=1.59 ± 0.48 μg/mL, meanIVDD/DLSS =0 μg/mL, P<.01, meanTCNS=0.07 μg/mL ± 0.04, P = .01, meanIE= 0 μg/mL, P<.01; Fig 2). Because of high intragroup differences (81%), only 19% of variance could be explained by disease grouping. In dogs with various CNS tumors (n=26), CRP in the CSF was nondetectable (n=22), with the exception of 2 meningiomas and 2 hemangiosarcomas (1.03, 0.23 μg/mL and 0.53, 0.1 μg/mL). It was remarkable, that in 2 of 3 dogs with sepsis, the concentration was also equal to or above the detection limit (0.1 and 0.14 μg/mL).

image

Figure 2.  C-reactive protein (CRP) concentrations in the cerebrospinal fluid (CSF) in the groups of dogs at the time of diagnosis (steroid responsive meningitis-arteritis: SRMA n=35; other meningoencephalitides: ME n = 13; intervertebral disk disease/degenerative lumbosacral stenosis: IVDD/DLSS n=32; tumors affecting the central nervous system: TCNS n = 26; idiopathic epilepsy: IE n = 25; healthy controls: normal n = 8, dogs with sepsis: sepsis n = 3); boxplots: median of values, minimal and maximal values, boxes contain the middle 50% of sample values. Groups with different letters (A, B) are significantly different at P≤ .01.

Download figure to PowerPoint

Plasma AMG Concentrations

No significant differences were detected between the groups (Fig 3). Results for dogs with SRMA or sepsis did not differ from healthy controls (meanSRMA= 109%± 20, meanSepsis= 86%± 54, meanNormal= 91%± 17). The age of the dogs had no effect on AMG values.

image

Figure 3.  Plasma α2-macroglobulin (AMG) concentrations in the groups of dogs at the time of diagnosis (steroid responsive meningitis-arteritis: SRMA n = 17; other meningoencephalitides: ME n = 11; intervertebral disk disease/degenerative lumbosacral stenosis: IVDD/DLSS n = 32; tumors affecting the central nervous system: TCNS n = 21; idiopathic epilepsy: IE n = 25; healthy controls: normal n = 8; dogs with sepsis: sepsis n = 6); boxplots: median of values, minimal and maximal values, boxes contain the middle 50% of sample values.

Download figure to PowerPoint

Serum Albumin Concentrations

Dogs with SRMA had albumin concentrations significantly lower than those in dogs with other ME, IVDD/DLSS, and tumors of the CNS (meanSRMA= 3.20 g/dL ± 0.412 g/dL, P<.01; Fig 4). Septic dogs had lower albumin concentrations but without significance because of a high standard deviation (meanSepsis= 3.16 g/dL ± 0.899, P≥ .07; Fig 4). The age of the dogs did not influence these results.

image

Figure 4.  Serum albumin concentrations in the groups of dogs at the time of diagnosis (steroid responsive meningitis-arteritis: SRMA n = 31; other meningoencephalitides: ME n = 14; intervertebral disk disease/degenerative lumbosacral stenosis: IVDD/DLSS n = 30; tumors affecting the central nervous system: TCNS n = 23; idiopathic epilepsy: IE n = 24; healthy controls: normal n = 8, dogs with sepsis: sepsis n = 6); boxplots: median of values, minimal and maximal values, boxes contain the middle 50% of sample values. Groups with different letters (A, B) are significantly different at P < .01.

Download figure to PowerPoint

Results of Blood and CSF Examinations in the SRMA Group

Dogs from the SRMA group had leucocytosis (n = 34), increased alkaline phosphatase (n = 23; medianleucocytes=26,620/μL, range 11,500–48,060, reference 6,000–12,000/μL; medianalkaline phosphatase= 231 U/L, range 82–553, reference < 150 U/L) activity, or both. The serum alkaline phosphatase activity in dogs pretreated with glucocorticosteroids was not higher than in dogs not pretreated with glucocorticosteroids (meanpretreated= 245 U/L ± 119, meannon-pretreated= 241 U/L ± 122, P= .925).

Dogs with SRMA had high numbers of nucleated cells with a neutrophilic pleocytosis within the CSF (mediantotal nucleated cells=1,775/3/μL, range 6–9,720, reference<8/3/μL; medianneutrophils=79%, range 34–100). The median of total protein concentration of the CSF was 66 mg/dL (range 7–1,333, reference<25 mg/dL). In many cases, the CSF was xanthochromic, suggestive of subarachnoid bleeding (medianerythrocytes=235/3/μL, range 0–76,800; reference<12,000/3/μL18). IgA levels in serum and CSF were high in this disease group (medianIgA serum=188 μg/mL, range 40–994, reference <100 μg/mL; medianIgA CSF=2 μg/mL, range 0.06–14.77, reference<0.2 μg/mL17).

Correlations in the SRMA Group

There was only a weak, nonsignificant correlation between the concentration of CRP in serum and CSF (r= 0.301, P= .084) in dogs with SRMA. There was no association between concentration of CRP in serum and other characteristics of the CSF, such as CSF cell count, CSF neutrophils, the total protein content of the CSF, or IgA. There was also no correlation between CRP in serum and total protein content or albumin in the blood. However, there was an association between CRP in serum and serum alkaline phosphatase (r = 0.515, P =.003; Fig 5). The CRP concentration in the CSF was significantly correlated to the number of erythrocytes in the CSF (r = 0.382, P =.026) and a weak correlation without significance could be found between CRP in the CSF and the total protein content of the CSF (r= 0.308, P= .072).

image

Figure 5.  Serum C-reactive protein (CRP) concentration and serum alkaline phosphatase activity correlated highly significant in dogs suffering from steroid responsive meningitis-arteritis (SRMA). Scatter plot of 31 paired measurements of serum alkaline phosphatase activity and serum CRP concentration, showing the regression line (solid) with 95% confidence intervals (dashed lines). The linear regression equation was y= 63.32 + 0.3409x, R2= 0.265, P= .003.

Download figure to PowerPoint

Monitoring of Treatment Efficacy in Dogs with SRMA

Serum CRP decreased markedly from the time of diagnosis until the 1st repeat examination in 30 of 31 dogs. The remaining dog had a relapse 4 weeks after diagnosis. The decrease of serum CRP of the clinically normal dogs was in the same relative dimension as the decline of total nucleated cell count in the CSF (Fig 6a). During monitoring of treatment efficacy, the concentration of serum CRP and the total nucleated cell count remained within the reference range (0.3–0.9 μg/mL; < 8 cells/3/μL).

image

Figure 6.  (a) Course of serum C-reactive protein (CRP) ▵ and CSF total nucleated cell count □ in dogs with steroid responsive meningitis-arteritis (SRMA) at the time of diagnosis and at monitoring of treatment efficacy; O: diagnosis, I: 4 weeks after diagnosis, II: 8 weeks, III: 12 weeks, and IV: 16 weeks; reference range CRP 0.3–0.9 μg/mL, total nucleated cell count < 8 cells/3/μL; values as means. (b) CRP in dogs with SRMA which made a full recovery inline image and dogs which showed a relapse within the observation period inline image. Serum CRP at the time of relapse ▴. O: diagnosis, I: 4 weeks after diagnosis, II: 8 weeks, III: 12 weeks, and IV: 16 weeks, reference range CRP 0.3–0.9 μg/mL; values as means.

Download figure to PowerPoint

There was no difference in the concentration of serum CRP between the dogs which recovered and the dogs which showed a relapse at the time of diagnosis and the 2nd control (P =.36, P =.49, respectively; Fig 6b), though there was a tendency toward higher serum CRP levels in the group of dogs, which showed a relapse later on, at the time of the 1st control (meanrecovery= 0.31 μg/mL ± 0.22, meanfuture relapse= 1.05 μg/mL ± 0.79, P= .063; Fig 6b). In 4 out of 5 cases a markedly increased serum CRP could be measured at the time of relapse (Fig 6b).

Discussion

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

In the present study a marked acute-phase response in dogs with SRMA could be shown supporting the hypothesis that SRMA is a systemic disease, or part of the systemic inflammatory response syndrome (SIRS).

CRP is an indicative parameter for the acute-phase response in dogs with SRMA, whereas AMG seems not to be a valuable APP in neurologic disorders in the dog. Thus, there is still no evidence, that AMG is changed in acute-phase response in naturally diseased dogs at all. Albumin levels—as a negative APP—tally with the findings of CRP levels, but no significant difference between SRMA dogs and negative controls could be found. Therefore, a decreased concentration of serum albumin could only support but not lead to a diagnosis of systemic inflammation.

Although the increases in the concentration of serum CRP in dogs suffering from SRMA was very impressive, other systemic inflammatory diseases such as sepsis cannot be differentiated using this test. This limitation was also found in other studies examining the diagnostic use of CRP in dogs.2

CRP in the CSF was high in the SRMA group. In this group, arteritis is expected and can lead to an increased risk of subarachnoid bleeding resulting in xanthochromic CSF.15,20,21 In fact, a correlation between CRP concentration in the CSF and the number of erythrocytes in the CSF could be shown (r= 0.382; P= .026), even though the erythrocyte count was still within the reference range. Hence, it is possible, that the elevated concentration of CRP in the CSF originates from subarachnoid bleeding. This corresponds to the findings in 2 dogs with sepsis, where CRP was also elevated in the CSF. In sepsis, a secondary CNS-vasculitis with increased permeability or a deranged function of the blood-brain barrier may occur and could explain an increase in CRP in the CSF.22

Another explanation for the increased CRP values in CSF samples could be an intrathecal production, eg by leucocytes in the meninges. A production of APPs by leucocytes was suspected in former studies.23,24 Determining a CRP index in accordance with the IgG index25 would be useful in further studies for deciding whether the CRP in the CSF originates from the blood or from intrathecal production. In addition to dogs with SRMA, CRP in the CSF was markedly increased in 4 dogs with tumors affecting the CNS, 2 with meningioma, and 2 with hemangiosarcoma. These are 2 types of tumors, where an alteration of the blood-brain barrier or hemorrhage might occur.26,27

Hepatocytes are known to produce CRP.1,28 Therefore, it is not surprising that serum CRP and alkaline phosphatase correlated highly significantly. Both parameters are supposed to be induced by proinflammatory mediators, especially interleukin-1, in liver cells.29

An alternative explanation for a high CRP and a high alkaline phosphatase is pretreatment with glucocorticosteroids. It is well known, that in dogs alkaline phosphatase synthesis is induced by glucocorticosteroids.30 Furthermore, glucocorticosteroids generally enhance the stimulatory effects of cytokines on the production of APPs.31 Otherwise, production of cytokines is downregulated by glucocorticosteroids. Though Martínez-Subiela et al32 showed that CRP concentration in the blood of dogs is not affected by glucocorticosteroids, others supposed an association, although in pigs.33 In the underlying study population, no difference between dogs with pretreatment with glucocorticosteroids and dogs without pretreatment could be shown regarding the concentration of serum CRP and alkaline phosphatase (P= .242; P= .925). Thus, in the current study, glucocorticosteroids had no effect on CRP levels in the blood.

Besides its usefulness in diagnostics, measurement of CRP is used to monitor treatment success and as a predicting parameter for healing and treatment failure in human and veterinary medicine.34–37 In all these studies, a marked decline in the CRP concentration could be observed between the time of diagnosis and treatment control. This corresponds to the findings of the present study in SRMA: A perceptible decrease in serum CRP parallels the decline of CSF cell count after initiation of treatment. Therefore, measurement of serum CRP can be considered to be a useful monitoring parameter in SRMA. Up to now, remission of clinical signs and number of nucleated cells in the CSF have been the only parameters for monitoring the treatment schedule.16 As serum CRP concentrations decline in the same way as cells in the CSF during treatment and as both analytes increased again in cases of relapse in the dogs of this study, serum CRP concentrations could be considered as a valuable tool in diagnosing relapses, albeit other inflammatory diseases have to be ruled out.

For management of dogs with SRMA, it would be beneficial to have a laboratory parameter, which could predict a relapse before recurrence of clinical signs. At the 1st repeat examination for monitoring of treatment efficacy, serum CRP levels in dogs suffering from a relapse during the observation time showed a tendency toward higher CRP levels in comparison with dogs that made a full, unremarkable recovery. This difference was not observed anymore comparing these 2 groups 4 weeks later. CRP in the dog is known to increase significantly 4 hours after stimuli with a maximum peak 24 hours after surgical trauma38 and 48 hours after turpentine oil injection.5 In the latter study, the amount of serum CRP decreased markedly during the following days and reached nearly normal values after approximately 3 weeks. Therefore, measuring CRP values to monitor treatment of SRMA can only be an indicator for the current status of the disease, but clearly predicting relapses seems not to be feasible.

CRP was markedly elevated in serum and CSF of dogs with SRMA in contrast to dogs with other neurologic diseases. Hence, CRP seems to be a valuable parameter for SRMA in comparison with other neurologic diseases. However, other systemic diseases of the SIRS complex have to be ruled out. Serum albumin values support these findings as they were significantly lower in dogs with SRMA than in dogs with other neurologic diseases. Because albumin as a negative APP was also lower in septic dogs, the measurement of a combination of CRP and albumin does not help distinguishing these 2 groups of diseases. AMG values did not differ between groups of dogs with neurologic diseases. A combination of measuring CRP, albumin, and AMG as an APP profile unfortunately could not improve the diagnostic workup of SRMA in comparison with the determination of serum CRP alone. In conclusion, measuring CRP as an APP supports the diagnosis of SRMA very well, but cannot be used as a single diagnostic tool.

Footnotes

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

aMAC 5,000 marquette, GE Medical Systems, IT GmbH Freiburg, Germany

bHitachi 912, Boehringer-Mannheim, Germany

cTridelta development Ltd, Kildare, Ireland

dUnitest, Unicorn Diagnostics, Kent, UK

eSPSS, version 13.0 for Windows; SPSS Inc, Chicago, IL

Acknowledgments

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References

The authors thank Dr Veronika Stein, PhD, DiplECVN, Dr Irene Boettcher, DiplECVN, Dr Cornelia Bull, Dr Henning Schenk, PhD, and Dr Thilo von Klopmann for their support in collecting samples.

References

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Footnotes
  7. Acknowledgments
  8. References