Meningoencephalitis Due to Listeria monocytogenes in a Pregnant Rhesus Macaque (Macaca mulatta)

Abstract

We here report a spontaneous case of meningoencephalitis due to Listeria monocytogenes in an adult primiparous rhesus macaque (Macaca mulatta) during an outbreak of listeriosis in an outdoor enclosure. Clinical signs included tremors, abnormal posture, and altered mental status. Hematology and analyses of cerebrospinal fluid were consistent with bacterial infection. Pure cultures of L. monocytogenes were recovered from the placenta–abortus, cerebrospinal fluid, and brain tissue. The macaque did not respond to treatment and was euthanized. Histopathologic examination of the brain revealed acute meningoencephalitis. This case represents an unusual clinical and pathologic presentation of listeriosis in a nonhuman primate in which the dam and fetus both were affected.

Listeria monocytogenes is a ubiquitous, facultative anaerobic, intracellular gram-positive coccobacillus. This bacterium is found in diverse environments including (but not limited to) soil, water, plant matter, food items, and the intestinal tract of mammalian hosts.^15,18 The organism is environmentally resistant, being able to survive in dried media for several months and in moist soil for up to a year.¹⁵ L. monocytogenes is the causative agent of listeriosis, a bacterial infection that has a worldwide distribution and affects a wide range of mammals and birds, including human beings.

In people, L. monocytogenes is a relatively uncommon foodborne pathogen; its abilities to survive food processing and grow in cold conditions allow it to persist in appropriately stored or refrigerated foods.² In people, listeriosis occurs both sporadically and as large outbreaks,¹⁸ generally comprising 3 separate syndromes with clinical manifestations ranging from mild to life-threatening.³⁵ The most common form is seen in immunocompetent, nonpregnant adults as a febrile gastroenteritis.^2,18,21 The other 2 forms, which occur in fetuses and immunocompromised patients, are more severe.^19,21 In pregnant women, maternal listeriosis is asymptomatic or causes mild, flu-like symptoms, but the bacterium's ability to cross the placenta and the blood–brain barrier of the fetus results in neonatal septicemia, meningitis, abortion, and stillbirth.¹⁶ In elderly and immunocompromised patients, septicemia and meningoencephalitis are life-threatening manifestations of literiosis.²⁶ The worldwide case fatality rate varies widely among countries, sometimes exceeding 50% despite what is considered to be appropriate antibiotic therapy.¹⁸ In 2009, the Centers for Disease Control reported 524 cases of listeriosis in the United States, which were associated with a 19% resulting in death.⁴

In ruminants, listeriosis is also known as ‘circling disease’ and ‘silage disease.’^8,18,21 Foodborne infection with L. monocytogenes is well described, and many studies have shown that spoiled silage may be a source of listeria outbreaks.^8,18 Rhombencephalitis and diffuse meningoencephalitis are the most recognized forms of the infection in ruminants; sporadic abortion is reported also.²² Clinical signs of listeria encephalitis in cattle, sheep, and goats are characterized by unilateral or bilateral brainstem dysfunction and cranial nerve deficits. In sheep and goats, the course of the disease is acute, but the disease in cattle has a more chronic progression, with neurologic manifestations that can last 4 to 14 d.^1,22

In rabbits, infection with L. monocytogenes is characterized by abortion in pregnant does or sudden death; neurologic signs are rare.¹ In poultry, an acute form with septicemia and sudden death occurs in adults, in contrast to a subacute–chronic form, with encephalitis, in the young.⁶

The literature on L. monocytogenes in nonhuman primates is sparse^5,11,17,33 and more recently limited to experimental infection of pregnant animals. In pregnant rhesus macaques (Macaca mulatta), experimental infection during the last trimester of gestation can cause stillbirth with no other clinical signs.^23,24 In our colony, however, infection with L. monocytogenes is endemic. Every year, several spontaneous abortions or stillbirths in our outdoor colony are caused by infection of the dam with this organism. Culture of L. monocytogenes from both the abortus–fetus and placenta are well documented. As described in the literature,^23,24 the dams in our colony do not demonstrate any clinical signs prior to the abortion or stillbirth.

During the winter to spring of 2011, one of our outdoor housing enclosures experienced an outbreak of listeriosis. This outside corral housed 100 rhesus macaques in a social group that included 42 reproductive females. Of these reproductive females, 37 (88%) were confirmed pregnant by abdominal palpation or ultrasonography or both. From January 2011 to May 2011, 19 (51%) stillbirths and neonatal deaths (in infants younger than 3 d) were reported in this enclosure; 13 (68%) of these tissues (placenta, 3; fetal lungs, 8; fetal peritoneum, 2) were culture-positive for L. monocytogenes. In all cases except the one presented here, the dam did not manifest any clinical signs prior to or after the delivery of a stillborn or premature birth with neonatal death.

Here we describe an unusual case of listeriosis in a primiparous pregnant female rhesus macaque that manifested severe neurologic impairment and intrautero death of the fetus.

Case Report

A pregnant female rhesus macaque (Macaca mulatta; age, 3 y 9 mo; weight, 4.5 kg) was presented to the hospital in March 2011 for neurologic deficits and abnormal behavior. This animal had no previous medical history. She was colony-bred and maintained in a large conventional outdoor breeding colony. All animals in this enclosure were maintained according to the Guide for the Care and Use of Laboratory Animals. Protocols for the maintenance and breeding of rhesus macaque colonies were approved by the University of California IACUC. The macaque had not undergone any experimental procedures or manipulation prior to presentation.

On initial presentation, the macaque was sternally recumbent and unable to assume a normal posture. She demonstrated fasciculation of her facial muscles and was obtunded (decreased level of consciousness, lack of responsiveness to stimuli to the environment). The macaque was sedated with midazolam (0.1 mg/kg IM; APP Pharmaceuticals, Schaumburg, IL) and ketamine (10 mg/kg IM; Butler Animal Health Supply, Dublin, OH). A complete physical exam, whole-blood analysis (NOVA-CCX Stat Analyzer, Nova Biomedical, Waltham, MA), hematology, serum biochemistry profile, and abdominal ultrasonography were performed.

Clinical examination under sedation revealed the presence of nystagmus. Tremors were still present even under sedation; diazepam (0.5 mg/kg IV; Hospira, Lake Forest, IL) was administered to control these tremors. The macaque's temperature, heart rate, and respiratory rate were within normal limits. Abdominal palpation confirmed the presence of a term pregnancy. On vaginal examination, the external cervical os of the uterus was completely dilated, and the internal cervical os was dilated approximately 2 cm. Ultrasonography of her abdomen showed the presence of a nonviable fetus with a biparietal measurement of 48.9 mm, which is considered full-term, according to the predicted values for biparietal diameter for rhesus macaques.¹² There was an appropriate amount of amniotic fluid in the 4 quadrants of the uterus, and the 2 placental disks appeared intact and well adhered to the uterine wall. No particulates were observed in the amniotic fluid. Her whole-blood analysis showed mildly decreased ionized calcium (0.92 mmol/L) and mild hyponatremia (129.8 mmol/L).⁹

Hematology results were consistent with bacterial infection. The WBC count was 6.9 × 10³/µL, plasma protein was 7.7 mg/dL, fibrinogen was elevated at 500 mg/dL. There was a moderate left shift with 19% bands, and the segmented neutrophils (69%) showed moderate toxic granulation.⁹

The macaque was treated immediately with antimicrobial therapy comprising cefazolin (25 mg/kg IV; Steri-Pharma, Syracuse, NY) and enrofloxacin (5 mg/kg IM; Bayer HealthCare, Shawnee Mission, KS), isotonic fluid therapy (lactated Ringer solution with 2.5% dextrose; 30 mL/kg IV hourly; Baxter HealthCare, Deerfield, IL), and calcium gluconate (4.4 mg/kg IV; 10% Ca²⁺ Gluconate, APP Pharmaceuticals). Her whole-blood analysis was repeated at 2 h after treatment; the ionized calcium (1.04 mmol/L) and sodium (135.9 mmol/L) levels had normalized. The tremors and altered mental status persisted.

Hysterotomy and collection of cerebrospinal fluid were performed later that day. The macaque was sedated (ketamine, 10 mg/kg IM; Butler Animal Health Supply; atropine, 0.05 mg/kg IM, Baxter HealthCare), intubated, and maintained on isoflurane (1.5% to 2%; Piramal Critical Care, Bethlehem, PA). A sample of grossly turbid CSF was submitted for cytology and culture. A fetectomy was performed, and the abortus and placenta were submitted for histopathology and culture. Tremors subsided while the macaque was under general anesthesia but reoccurred postoperatively. She was treated with dexamethasone (0.13 mg/kg IM once; APP Pharmaceuticals), oxymorphone (0.3 mg/kg IM every 8 h; DMS Pharmaceuticals, Greenville, NC), midazolam (0.1 mg/kg IM, APP Pharmaceuticals) cefazolin (25 mg/kg IM every 8 h; Steri-Pharma), enrofloxacin (5 mg/kg IM every 12 h; Bayer HealthCare), and continuous intravenous infusion of diazepam (1.5 mg/kg hourly then 0.75 mg/kg hourly; Hospira). The macaque did not respond to treatment and was euthanized with pentobarbital (0.25 g/kg IV; Vortech Pharmaceuticals, Dearborn, MI) at 48 h after presentation.

The results of the serum biochemistry analysis, evaluation of cerebrospinal fluid, and culture of placenta, cerebrospinal fluid, and brain ventricles were received after the animal's death. The serum biochemistry showed hyponatremia (131 mmol/L), normal corrected calcium level (10.4 mg/dL: [9.2 + 0.8 × {4 to 2.5}]), hypoproteinemia (5.9 g/dL), hypoalbuminemia (2.5 g/dL), mild elevation of AST (93 U/L), elevated CPK (3338 U/L), and elevated LDH (1811 U/L).⁹

CSF results were compatible with bacterial meningoencephalitis, with an elevated nucleated cell count (5.6 × 10³/µL) and hyperproteinorrhachia (480 mg/dL). Cytology revealed 64% neutrophils, 19% lymphocytes, and 17% monocytes, with rare RBC and bacteria.^25,29 Cultures of the placenta, cerebrospinal fluid, and brain ventricles were positive for L. monocytogenes.

Gross pathology of the bidiscoid placenta revealed the presence of multifocal to coalescing gray to tan discolored areas on the cut surface that contrasted with the normal reddish brown placenta (Figure 1). Histopathology revealed a multifocal mixed inflammatory infiltrate consisting mainly of neutrophils and macrophages that separated and effaced the parenchyma over approximately one third of the section. Placental villi were either separated or completely replaced by the inflammatory infiltrate. Intensely eosinophilic areas of necrosis with karyolysis and nuclear pyknosis occurred separately or were admixed with neutrophils and macrophages (Figure 2).

Figure 1.

Placenta: gross, multifocal necrotizing placentitis.

Figure 2.

Placenta: multifocal necrosis admixed with degenerate inflammatory cells. (A) Zone of necrosis. (B) Unaffected tissue. Hematoxylin and eosin stain; scale bar, 500 µm.

The only gross lesion evident on necropsy of the adult female macaque was approximately 15 mL of clear fluid in the pericardial sac. A full set of tissues was processed for routine histopathology. On histologic examination, sections of the brain showed severe ependymitis and choroiditis (Figure 3) with mild to moderate meningitis (Figure 4). The ependymal epithelial cell lining of the lateral ventricles was largely disrupted and lost, and the underlying white matter was edematous and variably infiltrated with mixed inflammatory cells (predominantly neutrophils). The leukoencephalitis in the white matter underlying the ependyma was secondary to extension of inflammation through the damaged ependymal lining into the underlying neural tissue. The choroid plexus was expanded by a mixed infiltrate of lymphocytes, neutrophils, and macrophages. Blood vessels within the choroid plexus were congested, and small multifocal areas of acute hemorrhage were present. The epithelial cells of the choroid plexus were multifocally lost or detaching in the area where the choroid plexus was expanded most severely. The lumen of the ventricles contained cellular debris, detached epithelial cells, and moderate numbers of neutrophils, many of which were degenerate. The meninges were expanded multifocally by few to moderate numbers of mixed inflammatory cells, principally macrophages and neutrophils, with smaller numbers of lymphocytes and plasma cells. The inflammation in the meninges was much less severe than that in the ependyma and choroid plexus.

Figure 3.

Brain. (A) Choroid plexus is expanded by mixed inflammation and congestion. (B) Ependymitis with edema and inflammation extending into the underlying white matter. Hematoxylin and eosin stain; scale bar, 500 µm.

Figure 4.

Brain: focal meningial expansion by mixed inflammatory cells. Hematoxylin and eosin stain; scale bar, 500 µm.

Discussion

We present, to our knowledge, the first case of spontaneous meningoencephalitis due to L. monocytogenes in an immunocompetent adult macaque. In primates, the most common etiologic agent isolated during bacterial meningitis is Streptococcus pneumonia. Other agents including Klebsiella pneumonia, Pasterella multocida, Haemophilus influenza, and Neisseria meningitides are reported also.¹⁰

The histopathologic lesions in the brain of the macaque we describe here most closely resemble those seen in human cases of listerial meningoencephalitis. In both humans and macaques, CNS lesions due to L. monocytogenes are primarily superficial, involving the meninges, choroid plexus, and ependyma. In contrast, CNS lesions in cattle with listeriosis are usually located deep in the parenchyma and traditionally are described as microabscesses in the brainstem, sometimes with secondary involvement of the meninges.^14,15 These focal lesions do not expand much and remain small, and the surrounding parenchyma remains unchanged, suggesting that this organism is not highly toxigenic to the CNS of cattle.

The difference in lesion distribution among species most likely reflects the route of infection. In humans, infection of the brain occurs via a hematogenous route, after entry through the gastrointestinal tract and unrestricted proliferation in the liver.¹⁴ A similar scenario most likely occurred in this macaque. The choroid plexus is an important portal of entry into the brain via the hematogenous route,¹⁸ and this macaque had marked inflammation in the choroid plexus. In contrast, brainstem lesions in cattle are thought to occur after the organism gains access through injury to the oral mucosa, and the organism is passed centripetally along the cranial nerve to the medullary centers. In cattle, there is a correlation between Listeria encephalitis and trigeminal neuritis; and organisms have been demonstrated to be present in the myelinated axons of the trigeminal nerve.^14,15

The selection of the appropriate antibiotic for the treatment of listeriosis is critical. Due to the severity of her illness, our macaque was treated empirically with broad-spectrum antibiotics effective against most common organisms, before a final diagnosis was established. The failure of the antimicrobial therapy in this case is linked to our selection of an antibiotic that is ineffective against L. monocytogenes. In vitro, L. monocytogenes is usually resistant to cephalosporins and first-generation quinolones.² Cephalosporins have minimal to no affinity for listerial penicillin-binding proteins 3 and 5² and have poor penetration through the blood–brain barrier. Protein 3 is essential in the late stage of peptidoglycan synthesis, and its antibiotic-associated inactivation causes cell death.²⁷ The resistance of L. monocytogenes to quinolones is explained by the inherent resistance of the bacterium to nalixidic acid.² In people, the drugs of choice for the treatment of invasive listeriosis include ampicillin, penicillin, and amoxicillin, with or without an aminoglycoside (that is, gentamycin).^2,3,13,27,35 In the event of allergy to penicillins, an alternative treatment with trimethropine–sulfamethoxazole is recommended.³ In people, prolonged (21 d) massive doses of ampicillin or penicillin (more than 6 g daily) is considered to be standard-of-care therapy for acute meningitis due to L. monocytogenes.²⁷

Dexamethasone as an adjunctive therapy for the treatment of bacterial meningitis is no longer controversial. Several studies have shown that the use of steroids in pneumococcal meningitis and other community-acquired bacterial meningities is associated with significant reduction in mortality and neurologic sequelae, without increased adverse effects.^7,30 In experimental studies of bacterial meningitis, corticosteroid therapy decreased the inflammatory response at the level of the subarachnoid space and subsequently correlated with a favorable outcome.³¹ Short-term therapy with corticosteroids (that is, 10 mg dexamethasone every 6 h for 4 d) is considered standard-of-care in people with acute bacterial meningitis.⁷

In humans, the role of host immunity during pregnancy has been linked with impaired responses to infectious agents. For example, pregnant women have increased susceptibility to respiratory viruses, including influenza and malaria.²⁸ The immunologic mechanisms which maintain pregnancy have not been elucidated fully, but our understanding of the role of the immune system has evolved from immunosuppression to active immunotolerance of fetal cells during pregnancy.²⁸ Several local and systemic maternal–fetal immunologic tolerance interventions have been postulated: lack of expression of MHC class I or II from the trophoblast, immunoprivileged uterine site, expression by the trophoblast of complement regulatory protein, uterine decidual, and placental cell production of cytokines that redirect the immune system from a Th1 response to Th2 response, hormonally mediated effects on thymus and on B cells, and change in the circulating and local T-cell subsets.³⁴ Despite the numerous changes in immunologic mechanisms that take place during pregnancy, Listeria-associated meningoencephalitis rarely occurs in pregnant women.¹⁶ Instead pregnant women with listeriosis usually experience mild flu-like illness followed by fetal loss.^16,32 In the vast majority of cases, listeria causes abortion after septicemia and subsequent hematogenous spread to the gravid uterus. Although miliary abscesses and meningitis often develop in the fetus,^16,21 the mother does not develop CNS infection. Therefore, pregnancy-associated immunomodulation alone is unlikely to explain the invasive form of listerial infection in the pregnant macaque we have presented. The immunologic status of this particular animal was not assessed further; however, routine virologic surveys of our conventional rhesus macaque colony over the last 5 y have not detected any positive animals for SIV or simian retrovirus.

L. monocytogenes displays various strain-specific differences in virulence.² Of the 12 known serovars of LM, 3 (serovars 1/2a, 1/2b, and 4b) account for 90% of human and animal cases worldwide.^18,32 The virulence of this organism correlates with the presence of virulent gene clusters in its genome.²⁰ These genes (hly, inlA, inlB, plcA, plcB, and ActA) encode listerial proteins that enable the survival and promote the pathogenicity of the bacteria in the host cell. These determinants of virulence include hemolysin (hly) and phospholipase (plcA, plcB), which are involved in the disruption of phagosomal membranes; internalins (inlA, inlB), which are required for the uptake of bacteria into nonphagocytic cells; and ActA, which confers intracellular mobility to the bacteria by enhancing actin polymerization.^20,32 We did not determine the specific strain of L. monocytogenes in the current case to determine its virulence.

As reported in experimental studies in nonhuman primates,^23,24 we suspect that oral exposure to L. monocytogenes occurs in the outdoor enclosure in which the presented macaque was housed. The source of infection was not investigated because we are well aware of the ubiquitous nature of the organism and the macaque's access to potentially contaminated soil, plant matter, water, and animal feces.

After the review of this case and the epidemiologic health assessment of this macaque's enclosure, several herd population interventions were implemented at the animal facility. The veterinary staff recommended the placement of several elevated structures in the enclosure to remove the macaques from the ground and the drainage of any standing water. In addition, a canvas canopy was installed in this enclosure to partially shield the cage from rain.

In conclusion, we here describe the first spontaneous case of meningoencephalitis due to L. monocytogenes in an immunocompetent pregnant rhesus macaque. Further assessment of the immune status of the infected animal and strain-specific virulence of the bacterium are warranted to establish whether this animal was particularly susceptible to invasive infection. Even though a rare condition, listeriosis should be included in the differential diagnosis of bacterial meningoencephalitis in rhesus macaques and appropriate antibiotic therapy should be instituted.