HMN 2026: How Destructive meningitis cases raise concerns about emerging pathogen

Penn State College of Medicine is reporting on two cases of destructive infant meningitis linked to Paenibacillus infection, documenting severe neurologic injury and raising concerns about diagnosis and treatment. Reports from Uganda first linked these organisms to destructive neonatal infections, including cases with abnormal accumulation of cerebrospinal fluid, seizures, and extensive brain injury. Similar cases have since been recognized in multiple U.S. states and routine antibiotic regimens for infant bacteremia and meningitis may be inadequate when Paenibacillus is involved.

Organisms within the genus are commonly identified as part of the soil microbiome. Historically regarded as an uncommon human pathogen, increasing recognition of neonatal cases has begun to reshape that perception.

In the study, “Paenibacillus dendritiformis as a Cause of Destructive Meningitis in Infants,” published in NEJM Evidence, researchers described two infants with devastating neurologic manifestations associated with Paenibacillus infection along with laboratory work that re-identified the organism at the species level.

Severe neurologic presentations

One case involved a 2-month-old female infant born at 26 weeks’ gestation. Respiratory distress and seizures prompted blood culture testing, which identified gram-negative rods. Blood and cerebrospinal fluid cultures grew what they thought was Paenibacillus thiaminolyticus with no additional pathogens detected. Brain imaging revealed progressive hydrocephalus, encephalomalacia, and abscess formation requiring placement of a ventriculoperitoneal shunt.

Continuous-infusion meropenem was administered for 8 weeks. Vancomycin and rifampin were added midway through treatment in response to persistent cerebrospinal fluid abnormalities.

Thiamine supplementation began four days after symptom onset to offset the bacteria’s enzymatic destruction of thiamine, the depletion of which can cause brain tissue damage. At 8 months of age, neurologic function remained significantly impaired, with preserved eye contact and smiling but inability to feed orally, sit unsupported, or roll independently.

A second infant, previously reported in Minnesota, presented at 37 days of age following preterm birth at 33 weeks’ gestation. Poor feeding and unresponsiveness led to hospitalization. Blood and cerebrospinal fluid cultures were again identified as P. thiaminolyticus with imaging demonstrating liquefactive meningoencephalitis, where brain tissue begins to dissolve into a viscous liquid.

Treatment included intravenous ampicillin and ventriculoperitoneal shunt placement. Clinical deterioration progressed to feeding difficulties and seizures, culminating in death at 11 months of age.

Diagnostic complexity

Initial lab testing for both infants incorrectly identified Paenibacillus thiaminolyticus, with later whole-genome sequencing of isolates confirming the species as Paenibacillus dendritiformis in both cases.

Paenibacillus species present additional diagnostic challenges at the microscopy level. Gram-variable staining characteristics may delay recognition, particularly when initial classification suggests gram-negative organisms.

Genomic analysis detected multiple features associated with pathogenicity and antimicrobial resistance. Identified genes encoded the type IV pilus operon, several ?-lactamases, vancomycin resistance determinants, and thiaminase 1. Previous work has implicated the type IV pilus as a virulence factor in neonatal paenibacilliosis.

Therapeutic uncertainty

Optimal antimicrobial therapy remains undefined. Whole-genome sequencing identified multiple ?-lactamase genes in the isolates, enzymes capable of breaking down ?-lactam drugs before they can disrupt bacterial cell-wall synthesis.

Genomic detection of vancomycin resistance determinants introduces another layer of complexity. While vancomycin resistance genes have also been detected in Ugandan isolates, the U.S. strains tested as phenotypically susceptible to vancomycin, initially suggesting to a clinician that the treatment might work. It would not then be until whole-genome analysis or a nonresponsive outcome that any resistance would be revealed.

Cases associated with more favorable outcomes have involved meropenem combined with thiamine supplementation. Thiamine destroying enzymes produced by both P. thiaminolyticus and P. dendritiformis offer a biologically plausible path linking infection to injury. Depletion of thiamine within brain tissue may contribute to tissue destruction alongside direct microbial effects.

Unclear mode of transmission

Both infants described were born preterm and required neonatal intensive care. Environmental reservoirs, including soil and water sources, have been proposed but appear unlikely explanations in many U.S. cases. Past Ugandan observations noted associations with rainfall and proximity to large bodies of water, though applicability to U.S. case settings remains uncertain.

Paenibacillus infection represents a clinically significant and potentially under recognized cause of severe neurologic injury, requiring heightened awareness among clinicians who care for young infants. Antibiotic regimens commonly used for neonatal bacteremia and meningitis may not be able to provide adequate countermeasures.

Early recognition, broader antimicrobial consideration, and timely neurosurgical consultation are all critical to improving outcomes.

The content is provided for information purposes only.