A previously healthy, fully immunized 7-year-old boy was taken to the emergency department in Colorado in June for headache, photosensitivity, irritability, and fevers up to 38.5°C, all lasting 6 days. He had no relevant sick contacts, travel, food and/or water or animal and/or vector exposures.
On arrival, he was afebrile, alert, and appropriately interactive, with abdominal tenderness. Pertinent laboratory results included a peripheral white blood cell count of 12.7 103/μL and a sodium level of 125 mmol/L, with values for the remainder of electrolytes, transaminases, C-reactive protein, and lactate all within normal limits. A computed tomography scan of the abdomen and pelvis with contrast and a computed tomography scan of the head without contrast were unremarkable. He received an intravenous (IV) normal saline bolus of 20 mL/kg. Eighteen hours later, he was transferred to our PICU because of increasing irritability, with suspicion for central nervous system (CNS) infection. On arrival to the ICU, the child was confused and intermittently following commands, prompting lumbar puncture and empirical initiation of IV ceftriaxone (50 mg/kg per dose, every 12 hours), vancomycin (20 mg/kg per dose, every 6 hours), and acyclovir (10 mg/kg per dose, every 8 hours).
Cerebrospinal fluid (CSF) studies were notable for 245 white blood cells per microliter with lymphocytic predominance (84%), 232 red blood cells per microliter, elevated protein level of 171 mg/dL, and normal glucose level of 41 mg/dL. Opening pressure was 28 cm H2O. No organisms were seen on Gram-stain. CSF was sent for bacterial culture, enterovirus polymerase chain reaction (PCR), herpes simplex virus (HSV) PCR, and West Nile virus immunoglobulin G and immunoglobulin M. Vancomycin, ceftriaxone, and acyclovir were continued while awaiting bacterial culture and HSV PCR results.
Additional workup included negative respiratory pathogen PCR panel results, negative blood culture results, and an initial serum creatinine level of 0.4 mg/dL. His hyponatremia was attributed to syndrome of inappropriate antidiuretic hormone, so fluids were restricted to an 80% maintenance rate until sodium normalized the following day. He had a normal routine EEG and no clinical seizures. After 1 day of hospitalization, the patient was answering questions appropriately and had a reassuring examination. He received 2 doses of IV ketorolac for analgesia. On hospital day 2, the patient’s creatinine level increased nearly threefold to 1.16 mg/dL, consistent with acute kidney injury (AKI). A vancomycin trough collected before the fourth dose was noted to be 18.2 μg/mL; vancomycin was discontinued the following day, after 6 total doses. Negative HSV PCR testing results returned 16 hours after CSF sample submission. Acyclovir was discontinued once these results were reported, after a total of 2 doses. Ceftriaxone was continued until CSF cultures revealed no growth at 48 hours (5 doses total). No causative pathogen was identified. He received IV fluids, and serum creatinine level trended downward over 8 serial checks to 0.65 mg/dL on hospital day 5 when he was discharged at baseline.
In Nonneonatal Patients on Ceftriaxone, Could We Reduce the Empirical Use of Vancomycin and Acyclovir for Suspected CNS Infections?
Out of concern for third-generation cephalosporin-resistant pneumococcus and untreated HSV, administration of the nephrotoxic triad of vancomycin, ceftriaxone and acyclovir to children with suspected meningoencephalitis has been routine practice for decades while awaiting CSF testing results.1 However, given the dramatic epidemiological changes driven by pneumococcal vaccines and development of rapid molecular CSF testing for both pneumococcus and HSV, we may now be able to safely limit the empirical use of vancomycin and acyclovir added to empirical ceftriaxone in many cases.
American Academy of Pediatrics guidelines since 2000 have recommended vancomycin, in addition to ceftriaxone, as empirical treatment of suspected bacterial meningitis to treat the potential of resistant Streptococcus pneumoniae, which is continued while monitoring for growth on CSF cultures and antimicrobial susceptibilities when pneumococcus is identified.2 Since that time, pneumococcal conjugate vaccines have greatly reduced the incidence of invasive pneumococcal disease and ceftriaxone resistance. After the success of the 7-valent vaccine, pneumococcal conjugate vaccine provides additional protection against strains such as 19A, which have higher rates of intermediate susceptibility or resistance to cephalosporins.3 The Center for Disease Control and Prevention Active Bacterial Core surveillance reveals decreasing third-generation cephalosporin-resistance among US invasive pneumococcal isolates from almost 20% in 1997 to 0.2% in 2018, with 3% intermediate and no resistant isolates from meningitis cases from 2011 to 2013.4–6
Despite stably low rates of invasive HSV infections in pediatric patients aged >2 months among 15 Pediatric Health Information System hospitals from 1999 to 2012, use of empirical acyclovir in cases of suspected HSV encephalitis rose from 7.6% to 15.6%, with only 0.2% of CSF HSV PCR tests sent resulting positive.7 Most children who received acyclovir with ultimately negative HSV PCR results lacked concerning findings for HSV encephalitis, such as altered mental status, seizures, or CSF pleocytosis.
Potential harms of prolonged empirical acyclovir and vancomycin use include excess health care resource usage and adverse medication effects, namely nephrotoxicity. The aforementioned study revealed that children who received acyclovir for suspected HSV encephalitis had longer hospital stays by 2.1 days, on average.7 Additionally, renal dysfunction has occurred in approximately one-third of hospitalized children receiving IV acyclovir and led to renal failure in 4%.1 The nephrotoxicity of the empirical triad of vancomycin, ceftriaxone, and acyclovir is often compounded by the use of IV contrast for imaging studies and nonsteroidal antiinflammatory drugs for analgesia in cases of suspected CNS infection. In PICU patients, development of AKI was found to carry a 10% risk of developing chronic kidney disease within 1 to 3 years of hospitalization.8
We conducted a retrospective analysis of 659 immunocompetent, nonneonatal pediatric patients at our children’s hospital with suspected meningitis and/or encephalitis from December 2016 to January 2018. S pneumoniae was detected in 3 (0.45%) CSF samples, and all were susceptible to penicillin and ceftriaxone. Approximately half of these patients received IV antibiotics, and 31% of antibiotic regimens included vancomycin. An analogous study of HSV testing in our institution found HSV was detected in 0.15% of CSFs tested, whereas 13% of children with suspected CNS infection were started on IV acyclovir for a median of 5 doses.9
Could Rapid Molecular Diagnostics for Pneumococcus and HSV in CSF Enable More Targeted Antimicrobial Therapy?
Newly available rapid molecular CSF tests provide an opportunity to more promptly and effectively tailor antimicrobial treatment regimens. The FilmArray Meningitis Encephalitis Panel enables rapid molecular detection of 14 common causative infectious agents of meningitis and/or encephalitis from CSF, including S pneumoniae and HSV-1 and HSV-2, with high sensitivity and specificity.10 Test turnaround time can be within 2 hours when run on-demand, thereby decreasing the period of diagnostic uncertainty. Specifically, with rapid negative testing for pneumococcus on the Meningitis Encephalitis Panel, empirical vancomycin use could be reduced or avoided. Similarly, in cases without high clinical suspicion for HSV and rapid negative testing for HSV, empirical use of acyclovir could be potentially reduced or avoided. Exceptions to this would include cases with antibiotic pretreatment, in which sensitivity of CSF testing for pneumococcus may be decreased, or cases in which empirical vancomycin is warranted to cover resistant Gram-positive infection outside the CSF. It is important to note that in cases with high suspicion for HSV encephalitis based on clinical features (such as temporal lobe findings, lateralized periodic discharges on EEG, or aphasia), acyclovir should be continued while repeat lumbar puncture for repeat HSV testing is pursued, given lower sensitivity of CSF testing early in the course.11
In the case presented, the patient received 2 days of therapy with the nephrotoxic triad of vancomycin, ceftriaxone, acyclovir, in addition to IV contrast and nonsteroidal antiinflammatory drugs, which likely led to AKI necessitating serial blood draws to monitor renal function and prolonged length of stay. With rapid negative testing for pneumococcus and HSV in this case, the empirical use of vancomycin and acyclovir could have been reduced, decreasing the risk of AKI. This case reveals the common harms of prolonged empirical use of broad, nephrotoxic, empirical antimicrobial agents when CNS infection is suspected. Although molecular diagnostic testing comes at a cost, these tests could bend the value curve by enabling decreased administration of unnecessary empirical nephrotoxic medications and shortened hospitalizations.
By rapidly reducing the likelihood of the already uncommon CNS infections of cephalosporin-resistant S pneumoniae and HSV, molecular diagnostics provide an opportunity to reduce or avoid the empirical use of vancomycin and acyclovir in immunocompetent, nonneonatal pediatric patients with suspected meningitis or encephalitis. Adverse effects from antimicrobial therapy, such as nephrotoxicity, and unnecessary venipuncture would be reduced, and costs related to drug usage, drug monitoring, and hospital length of stay could be saved.
Dr Ackley primarily conceptualized and designed the manuscript and drafted the initial manuscript; Dr Messacar provided substantial contributions to the conceptual design of the manuscript; Drs Tchou, Press, Parker, Dominguez, and Gaensbauer critically reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: Dr Messacar received funding from the National Institutes of Allergy and Infectious Diseases (grant K23AI28069). Funded by the National Institutes of Health (NIH).
POTENTIAL CONFLICT OF INTEREST: Dr Dominguez has received grant support from BioFire and Pfizer and is a consultant for Biofire and DiaSorin Molecular; the other authors have indicated they have no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.