A 10-month-old girl was transferred from an outside hospital with complaints of prolonged diarrhea, decreased activity, and eye wandering. About 9 days prior to admission, she developed multiple episodes of non-bloody, non-bilious emesis. The following day, she developed non-bloody diarrhea that continued for the next 5 days. Several days prior to admission, she started having decreased oral intake with minimal interest in food or drink. Her pediatrician saw her 1 day prior to presentation with concerns about vomiting, diarrhea, and poor intake. It was recommended that she try to use an oral glucose-electrolyte solution. Over the ensuing 24 hours, she only drank 6 oz of fluid and had decreased urine output. On the day of presentation, her grandmother, a nurse, went to check on her and felt that she was unresponsive and also noted “vertical nystagmus.”
A review of systems and family history was unremarkable. Her development was normal. Immunizations were up to date.
On exam, she was awake and irritable and would occasionally look around while lying in bed. Her temperature was 37.4ºC, respiratory rate was 52, and pulse was 120. Weight, length, and head circumference were in the 90th percentile. Head, ears, eyes, nose, and throat (HEENT) exam was unremarkable. Extraocular movements were intact. No nystagmus was noted. There were positive Kernig and Brudzinski signs as well as nuchal rigidity. Her lungs were clear. S1 and S2 were normal without murmur. Abdomen was soft without masses or hepatosplenomegaly. On neurologic exam, she was awake with her eyes open and cried with pain when moved. Deep tendon reflexes were 2+ bilaterally. Her tone was described as normal. She had decreased capillary refill and mottling. She had a few petechiae on her buttocks and calves. There was a scattered, blanching, macular red rash on her neck.
Initial laboratory evaluation showed hemoglobin of 10.6 g/dL; white blood cell (WBC) count of 6,700/mm3 with 55% neutrophils, 3% immature neutrophils, and 43% lymphocytes; and platelet count was 108,000/mm3. Toxic vacuolizations and Dohle bodies were present on the complete blood count. Prothrombin time was 13.4 seconds, partial thromboplastin time was 29 seconds, fibrinogen was 329 mg/dL, and d-dimer was greater than 22 mcg/mL (elevated). Lumbar puncture revealed slightly cloudy, non-xanthochromic fluid, 504 red blood cells/mm3, and 21 WBC/mm3 (52% neutrophils, 34% lymphocytes). Cerebrospinal fluid (CSF) glucose was less than 2 mg/dL, and CSF protein was 415 mg/dL. Gram’s stain revealed numerous extracellular bacteria and gram-positive cocci in pairs. Ultimately, cultures of the CSF and the blood grew Streptococcus pneumoniae.
Robert Listernick, MD, moderator: How does the diarrhea fit in?
Tina Tan, MD, pediatric infectious disease physician: Most likely it doesn’t. Unlike upper respiratory infections, viral gastroenteritis should not predispose a child to pneumococcal disease.
Dr. Listernick: What’s the role of neuroimaging prior to lumbar puncture?
Mark Wainwright, MD, PhD, pediatric neurologist: In a febrile child in whom meningitis is suspected, neuroimaging prior to lumbar puncture is indicated in the presence of focal neurologic findings suggestive of a focal central nervous system lesion. Lumbar puncture is contraindicated in a child with focal neurologic findings.
Dr. Listernick: Let’s interpret the results of the lumbar puncture.
Rebecca Reindel, MD, pediatric infectious disease physician: Clearly she has bacterial meningitis. The CSF glucose is very low, the protein is high and the Gram’s stain is positive. What we found remarkable was the very low WBC count in the face of striking CSF infection. The concentration of visible bacteria on the Gram’s stain was truly incredible. We wondered about the possibility of an immunologic abnormality that would blunt the inflammatory/WBC response.
Dr. Listernick: I understand, but I have to say that in the old days when bacterial meningitis was rampant, we occasionally saw children who had perfectly normal CSF values whose cultures grew S. pneumoniae or Haemophilus influenzae. We’ll come back to the question of immunodeficiency later. Let’s talk about initial antibiotic treatment.
Dr. Tan: Given the concern for resistant strains of pneumococcus, we recommend empiric treatment with ceftriaxone at meningitic doses (50 mg/kg every 12 hours) and vancomycin. Interestingly, the use of pneumococcal conjugate vaccine (PCV) 7 and PCV13 has controlled the rise of some of the more resistant pneumococcal serotypes. In Chicago, penicillin resistance has dropped considerably, to less than 15%, and ceftriaxone resistance is less than 5%.
Dr. Listernick: I’d like to say a word about the timing of antibiotic administration.
Robert Tanz, MD, general academic pediatrician: Certainly, you would like to start antibiotics expeditiously. However, past textbooks and experts stated that antibiotics needed to be administered within 30 minutes of presentation to the emergency department. We identified 90 children with bacterial meningitis in our institution and at the University of Chicago; the average time of administration of antibiotics was 90 minutes. Clearly, reality didn’t live up to the “standard of care” of experts that was not evidence based.
Ram Yogev, MD, pediatric infectious disease physician: One small but important point. The initial dose of intravenous antibiotics should never be “pushed.” There is both laboratory and clinical evidence that the massive release of endotoxin and subsequent production of cytokines can lead to shock and systemic inflammatory response syndrome.
Dr. Listernick: Just before I started training in the late 1970s, I believe the recommendation was to repeat a lumbar puncture at 24 hours into therapy to assess response. Are there circumstances when a repeat lumbar puncture is warranted now?
Dr. Tan: I would consider repeat lumbar puncture if the patient was not responding clinically or if I suspected the presence of a resistant organism. In addition, there are some infectious disease physicians who recommend repeat lumbar puncture in neonates who have gram-negative meningitis.
Dr. Listernick: How did you manage this child’s fluids in the first 24 hours of hospitalization?
Leah Harris, MD, pediatric intensive care physician: Maximizing cerebral protective measures is key in children who have bacterial meningitis. As such, I believe that in the intensive care unit setting, the only appropriate intravenous fluid is isotonic saline. This helps maintain cerebral perfusion pressure in the face of increased intracranial pressure. We could also have a long discussion as to how much dextrose should be in the solution. We definitely would like to avoid hyperglycemia or hypoglycemia in patients with any injury to the central nervous system.
Dr. Listernick: When I was a resident in the era of bacterial meningitis, we spent a great deal of time worrying about the development of the syndrome of inappropriate antidiuretic hormone (SIADH). I was taught that SIADH is found in euvolemic patients in whom one measures an inappropriately elevated urine osmolarity in the face of decreasing serum osmolarity in the absence of renal, adrenal, or thyroid disease. In these three conditions, the body may be unable to excrete a free water load.
Dr. Harris: We can monitor children’s urine output hourly in the intensive care unit. It’s not a good idea to restrict fluids in children with meningitis because they already may be hypovolemic due to vomiting, fever, or decreased oral intake prior to presentation. Further dehydrating a patient may lead to venous thromboses. Rather, we use isotonic saline and frequently monitor electrolytes and urine output.
Dr. Listernick: What neuroprotective measures can be employed?
Dr. Wainwright: There are multiple mechanisms of brain injury in bacterial meningitis, including cytotoxic edema due to neuronal injury, that lead to disruption of the blood-brain barrier. Dehydration may lead to arterial or venous strokes. Fluid resuscitation may lead to a greater passage of fluid across the disrupted blood-brain barrier, exacerbating cerebral edema. Simple neuroprotective measures include maintenance of euglycemia, adequate hydration, control of body temperature, and elevation of the head of the bed.
Dr. Listernick: Is there a role for corticosteroids?
Dr. Wainwright: The latest edition of one child neurology textbook recommends the use of dexamethasone in children with known or suspected bacterial meningitis.
Dr. Listernick: The first, and one of the largest, randomized prospective studies on the efficacy of dexamethasone in H. influenzae meningitis showed that hearing loss can be reduced. The study randomized 200 patients to receive either placebo or dexamethasone 0.15 mg/kg every 6 hours for 4 days, with the first dose given 20 minutes before antibiotics. The incidence of moderate-to-severe hearing loss dropped from 15% in the placebo group to 3% in the dexamethasone group. However, it should be noted that this study was performed prior to the era of the H. influenza vaccine.
Dr. Yogev: We performed a similar multicenter study looking at the efficacy of dexamethasone in preventing hearing loss in pneumococcal meningitis, but we couldn’t identify a difference in hearing loss between the two groups. There are even compelling animal models that show that dexamethasone decreases antibiotic levels in the CSF.
Dr. Harris: Interestingly, the mechanism of action of 3% sodium chloride intravenous infusions is not clear, and it has been postulated that hypertonic saline may have an anti-inflammatory effect.
Dr. Listernick: Families will often ask about their child’s prognosis when they arrive. What can we tell them?
Matthew Friedman, MD, pediatric intensive care physician: Despite all of the changes in intensive care therapies and antibiotics, these numbers haven’t changed much in decades. If one looks at all cases of pediatric bacterial meningitis, there’s an approximately 5% to 10% incidence of sensorineural hearing loss, 10% incidence of at least one gross neurologic handicap (ie, cerebral palsy) and 10% to 20% incidence of subtler neuropsychologic deficits compared with sibling-matched controls.
Dr. Yogev: The one predictive factor of poor neurologic outcome found in multiple studies has been low CSF glucose.
Dr. Listernick: What has been her course?
Dr. Friedman: Her neurologic exam changed quickly during the first 24 hours. Fairly soon after admission, she developed contralateral 6th and 7th nerve palsies and her right pupil became less reactive than it had been. Her mental status deteriorated further, but we found that on electroencephalogram she was having nonconvulsive seizures. Mental status improved once anticonvulsant therapy was initiated and she became more interactive.
Dr. Listernick: She was fully immunized against pneumococcus. Why did she get severe invasive pneumococcal disease?
Dr. Reindel: There are multiple possibilities. Most likely, this is a non-vaccine strain of pneumococcus. Secondly, because no vaccine is 100% effective, this may simply be an example of vaccine failure. Third, she may have a specific humoral immunodeficiency syndrome, such as common variable immunodeficiency. Finally, she may be asplenic or functionally asplenic. We were concerned about one of these latter possibilities because of the sheets of pneumococci that were seen on Gram’s stain of the CSF accompanied by a minimal inflammatory response.
Dr. Listernick: How did you proceed?
Dr. Reindel: We performed abdominal ultrasonography that revealed polysplenia (ie, multiple small splenules). However, when we looked at her blood smear, we could not find any Howell-Jolly bodies, which would have confirmed functional asplenia. She must have some splenic function. There was disagreement among our division members as to whether this polysplenia could be the sole cause of her overwhelming infection. She certainly will receive daily penicillin prophylaxis.
Dr. Listernick: Are you considering performing an immunologic evaluation?
Dr. Reindel: First, we are going to see if this was a non-vaccine strain. If it was not, we will probably perform quantitative immunoglobulins, and flow cytometry to quantify B and T cells, as well as complement levels. In addition, there could be defects in the innate immune system, such as mutations of the gene coding for nuclear factor-kappa B essential modulator (NEMO) or deficiency of interleukin-1 receptor-associated kinase 4 (IRAK4). These latter conditions should be suspected in patients who have: 1) recurrent invasive disease with encapsulated organisms; 2) low or absent antibody titers to these organisms following immunization; and 3) little or no evidence of inflammatory response during these infections (eg, no fever, normal inflammatory markers).
Dr. Listernick: Thank you, everyone.