Cerebrospinal fluid analysis in horses, cattle, and sheep diagnosed with rabies: A retrospective study of 62 cases

Abstract

Objectives

This study aimed to characterize the findings in cerebral spinal fluid (CSF) analysis of horses, cattle, and sheep diagnosed with rabies.

Animals

The study included 62 animals (horses, cattle, and sheep) diagnosed with rabies at a referral hospital.

Methods

This was a retrospective study using medical records from large animals with neurological signs and confirmed positive direct immunofluorescence test for rabies from 2003 to 2020. The results of CSF analysis are presented descriptively.

Results

Cerebral spinal fluid samples (N = 67) from 62 animals (31 horses, 24 cattle, and 7 sheep) were retrospectively evaluated. Of these 3 species, 28% (19/67) showed increased protein concentration, whereas 58% (39/67) presented mononuclear pleocytosis. In total, 37% of the samples (25/67) had protein concentration and total nucleated cell count within the reference range.

Conclusions and clinical relevance

Cerebral spinal fluid from animals diagnosed with rabies was either normal or characterized by mild mononuclear pleocytosis and hyperproteinorrachia.

Rabies is a severe and fatal neurologic disease that affects animals and humans (1). The definitive diagnosis of rabies is confirmed post-mortem (1). The gold-standard test is direct immunofluorescence antibody on central nervous tissue (1). A PCR, mouse inoculation test, and indirect immunofluorescence of brain tissue can also be performed (1). As the clinical signs of animals with rabies are non-specific (1,2) ante-mortem tests that support the diagnosis would be of great value.

A cerebral spinal fluid (CSF) cytological analysis provides valuable information that can help condense and prioritize the list of differential diagnoses and aid in the establishment of treatment and prognosis (3,4). To our knowledge, current data on changes in the CSF caused by the rabies virus is limited, but mononuclear/lymphocytic pleocytosis seems to be the predominant finding (2,3,5–7). However, the degree of pleocytosis varies among cases.

Although several techniques have been studied to develop a reliable ante-mortem test for rabies in other species, especially humans (8,9), these methodologies have not yet been extrapolated to large animal medicine. Improving the current knowledge of the CSF characteristics of large animals with rabies, therefore, could be of value to guide medical decisions, especially due to the zoonotic risk associated with rabies infection.

The aim of this study was to retrospectively describe CSF cytology results from horses, cattle, and sheep with a confirmed rabies diagnosis (direct immunofluorescence). Our hypothesis is that the results would show a mild mononuclear/lymphocytic pleocytosis.

Medical records (2003–2020) from the Large Animal Clinic at the Sao Paulo State University (Unesp — Botucatu) were reviewed. The inclusion criteria were any horse, cattle, sheep, or goat, regardless of sex or age, with a post-mortem positive diagnosis of rabies (direct immunofluorescence and/or mice inoculation). Animals without CSF collected or with a CSF red blood (cell) count (RBCC) ≥ 2000 cells/μL were excluded. Although it has been reported that RBCC ≤ 13 200/μL does not significantly alter protein concentration and total nuclear cell count (TNCC) (10), a lower limit was used to limit any microscopic or macroscopic influence. The following information was collected from each medical record: species, breed, age, sex, days of evolution until hospitalization, and the CSF analysis.

The CSF was collected aseptically from the atlanto-occipital (AO) and/or lumbosacral (LS) space (4) using sterile spinal needles (0.9 × 120 mm or 0.9 × 152 mm). Samples were placed in sterile glass tubes for analysis within 1 h. The CSF analysis was performed as previously described (11). Analysis consisted of an initial macroscopic evaluation for color and clarity. Protein determination was performed using a biochemical kit (Bioprot U/LCR — Bioclin — Belo Horizonte, MG, Brazil). The RBCC and TNCC were conducted in a Neubauer’s chamber. Differential cell counts were obtained through cytocentrifugation. The cells were stained with Diff-Quick (Diff-Quick — Laborclin, Pinhais, PR, Brazil) and the cell differential was performed on a 1000×-oil objective.

The normal CSF reference ranges of this report were based on previously published data (12). Normal TNCC (cells/μL) range was 0 to 6, 0 to 3, and 0 to 5 for horses, cattle, and sheep, respectively. The protein concentration (mg/dL) was considered normal between 5 to 100, 20 to 40, and 8 to 70 for horses, cattle, and sheep, respectively.

The results are reported descriptively and as medians and ranges. The graph (Figure 1) was generated using Microsoft Excel 16.54 (Microsoft Corporation/One Microsoft Way, Redmond, Washington, USA).

Figure 1.

Distribution of CSF protein concentration (mg/dL) and total nucleated cell count (cells/μL) of samples from horses (n = 31), cattle (n = 28), and sheep (n = 8) diagnosed with rabies. Results are shown as median and range. One sheep sample presented with a protein count of 220 mg/dL and was removed from the graph to facilitate visualization of the data.

The medical records search identified 62 animals (31 horses, 24 cattle, and 7 sheep) that fulfilled the inclusion criteria. No goats were diagnosed with rabies during this period. In animals with multiple CSF sample collections, the results of the first sample obtained were included in the study. Four cattle and 1 sheep had AO and LS collection performed within the same hour; both samples were included in the study. The total number of CSF cytological analyses included in this study was 67.

Of the 31 horses (16 females and 15 males), the median age was 20 mo (range: 1 mo to 30 y). Twelve animals were mixed breed, followed by 9 American Quarter Horses, 7 Mangalarga, 1 Thoroughbred, 1 Argentinian trotter, and 1 American miniature horse. Information regarding the clinical signs before hospitalization was available for 27 horses. The first clinical signs were observed a median of 2 d prior to presentation (range: hours to 7 d). The CSF collection was performed at a median of 1 d after hospitalization (range: hours to 3 d). Of the 31 equine CSF samples assessed, 6 were collected from the AO space and 2 from the LS space. The remaining 23 samples did not have a location of sample collection specified in the record. In total, 58% (18/31) were clear, and 42% (13/31) were xanthochromic. The median protein concentration was 51.7 mg/dL (range: 18.7 to 103.4 mg/dL) (Table 1, Figure 1). The TNCC median was 20 cells/μL (range: 0 to 115 cells/μL) (Table 1, Figure 1). The RBCC median was 81 cells/μL (range: 0 to 1198 cells/μL). In all samples, mononuclear cells were predominant (Table S1). Only 7% (2/29) of the samples had a protein concentration above the reference ranges, whereas 39% (12/31) of the samples had a pleocytosis. The protein concentration result was not available for 2 horses.

Table 1.

Cerebral spinal fluid protein concentration (mg/dL) and total nucleated cell count (cells/μL) of samples from horses (n = 31), cattle (n = 28), and sheep (n = 8) diagnosed with rabies. Results are shown as median and range.

Species	Protein (mg/dL)		TNCC (cell/μL)
	Median (Range)	RI (12)	Median (Range)	RI (12)
Horses	51.7 (18.7 to 103.7)	5 to 100	20 (0 to 115)	0 to 6
Cattle	43.3 (12.5 to 125.6)	20 to 40	9 (0 to 85)	0 to 3
Sheep	61.1 (23.0 to 220.6)	8 to 70	51 (14 to 109)	0 to 5

TNCC — Total nucleated cell count; RI — Reference interval.

Twenty-four cattle (10 male and 14 female) were included in this study. The age was available for 23 animals and the median was 1 y (range: 1 mo to 10 y). Fourteen animals were mixed breed, followed by 5 Nelore, 2 Angus, 1 Wagyu, 1 Holstein, and 1 Jersey. Historical information was available for 23 cattle. Clinical signs were present for a median of 2 d (range: hours to 5 d) before presentation. The median number of days after presentation that the CSF sample was collected was 1 d (range: hours to 10 d). In total, 28 CSF samples from 24 cattle were included. Four cattle had AO and LS CSF samples included. Of all samples, 4 were collected from the AO space and 7 from the LS space. The remaining 17 samples did not have a location of sample collection specified in the record. Only 1 out of 28 samples was xanthochromic, whereas all other ones were clear. The median protein concentration was 43.3 mg/dL (range: 12.5 to 125.6 mg/dL) (Table 1, Figure 1). The median TNCC was 9 cells/μL (range: 0 to 85 cells/μL) (Table 1, Figure 1). The median RBCC was 3 cells/μL (range: 0 to 640 cells/μL). In all samples, the predominant cell type was mononuclear (Table S2) and 71% (20/28) of the samples presented pleocytosis. Fifty percent of the samples (13/26) had increased protein concentrations, and 71% (20/28) of the samples had pleocytosis (12). The protein concentration result was not available for 2 cattle.

Seven sheep (6 female and 1 male), with a median age of 23 mo (range: 5 to 36 mo), were included. Four animals were Santa Ines, followed by 2 Dorper and 1 mixed breed. Clinical signs were present for a median of 3 d (range: 1 to 7 d) before presentation. After presentation, the median number of days until the CSF sample was collected was 0 d (range: hours to 2 d). One sheep had both AO and LS collections performed in the same hour, resulting in a total of 8 samples included for analysis in this study. One sample was collected from the AO space, 4 samples at the LS space, and 3 samples had no site of collection noted in the records. None of the samples were xanthochromic. The protein concentration was increased in 57% (4/7) (median of 61.1 mg/dL, range: 23 to 220.6 mg/dL) (Table 1, Figure 1). All samples presented pleocytosis (TNCC median of 51 cells/μL, range: 14 to 109 cells/μL) (Table 1, Figure 1). The RBCC median was 13 cells/μL (range: 0 to 1520 cells/μL). In all samples, mononuclear cells predominated (Table S3). Fifty-seven percent (4/7) of the samples presented with an increased protein concentration, and all samples had a pleocytosis (12).

Considering all samples from the 3 species in this report, 30% (19/63) had an increased protein concentration, whereas 58% (39/67) had a mononuclear pleocytosis. Twenty-six samples (38%) had protein concentrations and total nucleated cell counts within the reference range (12).

Five animals (1 sheep and 4 cattle) had CSF collected from both the AO and the LS spaces on the same day for comparison. Out of the 5 AO samples, 3 had increased TNCC, and 1 had increased protein concentration. Considering the LS samples, 3 had increased TNCC (same animals that had increased TNCC in the AO sample) and 3 had increased protein.

Interestingly, out of the 39 samples that had a pleocytosis, only 17 presented concomitant increased protein levels. Out of the 19 samples with increased protein, 17 also had concomitant pleocytosis.

This report describes that CSF cytological analysis of horses, cattle, and sheep with rabies may either present with mild mononuclear pleocytosis (58%) or be within the reference range (37%). This finding is congruent to previously published case reports and case series (2,3,5–7). In this group of cases, the CSF neutrophil count was mildly to moderately increased in a few cases; however, it was never the predominant cell type. This helps differentiate rabies cases from neurological diseases that cause marked suppurative inflammation (13), such as bacterial meningitis, and diseases that cause marked non-suppurative inflammation, such as bovine herpesvirus 5 meningoencephalitis in cattle (6). Solely based on CSF results, it might be difficult to differentiate rabies from other viral neurological diseases in equids, such as arbovirus diseases, as most also present with a mild to moderate mononuclear pleocytosis (14) and from leukoencephalomalacia, which presents with a mixed or mononuclear pleocytosis (15).

One of the main limitations of this study is the lack of information regarding the CSF site collection for 43 samples. In the 5 cases from which CSF was collected from both the AO and LS sites in this study, the TNCC was similar. In contrast, the protein concentration differed and was higher in 3 LS samples and 1 AO sample, respectively. Although the protein concentration was increased in LS compared to the AO samples in animals from which CSF was collected from both sites, mononuclear pleocytosis was the predominant finding in both samples, suggesting that cellular composition is less affected by sampling site than total protein concentration.

Due to the small sample size, it was not possible to differentiate species-specific changes. However, the 3 species appeared to demonstrate a similar pattern of abnormality. The only abnormality that was different among species was the high percentage of xanthochromic CSF in horses (42%) compared to cattle (4%) and sheep (0%). The median RBCC was also higher in horses (81 cells/μL) than in cattle (3 cells/μL) and sheep (13 cells/μL). If this change truly represents a specific equine response to the rabies virus, or is just a coincidence, remains undetermined. As the classification of color is a subjective evaluation and may present inter-individual variability, this information should be considered carefully before definitive conclusions are made.

In addition, no predilection for breed and age was detected in any of the 3 species. The number of females and males were evenly distributed in horses and cattle but not in sheep. This difference is believed to be due to the fact female sheep were over-represented in the hospital’s case load compared to male sheep.

Another limitation of this study is its retrospective nature, allowing for misclassification bias and confounding. As the sample was recruited by convenience (animals presented to the hospital), it might not represent the general population of animals with rabies in the field. Lastly, our small sample size limited the power of the study and the possibility of statistical analysis.

In conclusion, the study describes that CSF cytology of horses, cattle, and sheep with rabies is likely to have mild mononuclear pleocytosis or might be within the reference range. This result is especially important as there are currently no reliable ante-mortem tests for rabies available for large animals. Further studies with additional parameters and molecular identifications are warranted. CVJ