Cronobacter sakazakii Infections in Two Infants Linked to Powdered Infant Formula and Breast Pump Equipment

United States, 2021 and 2022

Julia C. Haston, MD; Shanna Miko, DNP; Jennifer R. Cope, MD; Haley McKeel; Cynney Walters; Lavin A. Joseph, MS; Taylor Griswold, MS; Lee S. Katz, PhD; Ashley A. Andújar, MHSA; Laura Tourdot, MPH; Joshua Rounds, MPH; Paula Vagnone; Carlota Medus, PhD; JoAnn Harris, MD; Robert Geist, MPH; Daniel Neises, MPH; Ashley Wiggington, MPH; Trey Smith, MS; Monica S. Im, MS; Courtney Wheeler; Peyton Smith, MS; Heather A. Carleton, PhD; Christine C. Lee, PhD

Disclosures

Morbidity and Mortality Weekly Report. 2023;72(9):223-226. 

In This Article

Abstract and Introduction

Introduction

Cronobacter sakazakii, a species of gram-negative bacteria belonging to the Enterobacteriaceae family, is known to cause severe and often fatal meningitis and sepsis in young infants. C. sakazakii is ubiquitous in the environment, and most reported infant cases have been attributed to contaminated powdered infant formula (powdered formula) or breast milk that was expressed using contaminated breast pump equipment.[1–3] Previous investigations of cases and outbreaks have identified C. sakazakii in opened powdered formula, breast pump parts, environmental surfaces in the home, and, rarely, in unopened powdered formula and formula manufacturing facilities.[2,4–6] This report describes two infants with C. sakazakii meningitis reported to CDC in September 2021 and February 2022. CDC used whole genome sequencing (WGS) analysis to link one case to contaminated opened powdered formula from the patient's home and the other to contaminated breast pump equipment. These cases highlight the importance of expanding awareness about C. sakazakii infections in infants, safe preparation and storage of powdered formula, proper cleaning and sanitizing of breast pump equipment, and using WGS as a tool for C. sakazakii investigations.

In September 2021, a state public health department reported an infant with C. sakazakii infection to CDC (patient A). In February 2022, a physician reported another case to CDC and a different state health department (patient B). CDC was invited to participate in the investigations and distributed case report forms to obtain detailed feeding information about the two cases. Patient isolates and environmental samples were sent to CDC, state public health laboratories, and other federal agencies for isolation, identification, and WGS analysis of C. sakazakii.

The first case occurred in September 2021 in a full-term (gestation of 40 weeks, 1 day) male infant (patient A) born to a healthy mother who had an uncomplicated pregnancy and spontaneous vaginal delivery. At age 14 days, the infant was evaluated at hospital A for fever, irritability, and excessive crying, oral candidiasis (thrush), and diaper dermatitis. Before his illness, he was fed both expressed breast milk and powdered formula. A lumbar puncture was performed, and C. sakazakii was isolated from the cerebrospinal fluid (CSF). The infant was admitted to the hospital and treated with intravenous antibiotics for 21 days; he made a full recovery with no apparent long-term sequelae (Table).

A C. sakazakii isolate from the CSF was analyzed by WGS. Samples of opened powdered formula and an opened bottled water container used for formula preparation in the patient's home were cultured in gram-negative enrichment broth followed by selective media. C. sakazakii isolates were recovered from both the powdered formula and the water container. WGS* was performed and identified two distinct strains of C. sakazakii (Figure). Sequence reads were cleaned, assembled,§ and analyzed using high-quality single nucleotide polymorphism (SNP) analysis; results showed that the patient isolate was closely genetically related to an isolate from the powdered formula (0 SNPs apart).** A second isolate from the same can of powdered formula was closely genetically related to an isolate from the water container (within 4 SNPs).†† These two strains were not related to one another (>50,000 SNPs apart). Federal partners tested unopened powdered formula from the same lot as the powdered formula consumed by the patient; C. sakazakii was not detected.

Figure.

Maximum likelihood phylogeny from high quality single nucleotide polymorphism* analysis†,§,¶ of Cronobacter sakazakii patient A and B isolates and environmental isolates — United States, 2021 and 2022
Abbreviations: CSF = cerebrospinal fluid; SNP = single nucleotide polymorphism;//= skip mark.
*SNP distances guide genetic relatedness between strains.
Scale bar indicates the genomic distance of samples to right of skip mark.
§Values at branch nodes indicate bootstrapping from 100 iterations. C. sakazakii isolates from patients A and B, in addition to environmental samples, form three distinct clusters in a maximum likelihood phylogeny.
Skip mark indicates a truncation to capture the relationship between clusters.

The second case occurred in February 2022 in a preterm (gestation of 30 weeks, 6 days), hospitalized male infant (patient B) born by cesarean delivery because of breech presentation and the mother's worsening preeclampsia. At the time of illness onset, he was being treated in the neonatal intensive care unit for complications of prematurity but was stable, feeding, growing, and breathing without respiratory support. He was fed expressed breast milk fortified with liquid human milk fortifier primarily through an orogastric tube before becoming ill. At age 20 days, he experienced apneic and bradycardic episodes, temperature elevation, and a requirement for respiratory support. Seizures developed the following day. C. sakazakii was isolated from blood and CSF cultures. Despite treatment with intravenous antibiotics and repeat negative blood cultures, the patient died 13 days after illness onset.

Expressed milk samples, breast pump parts from two separate devices (one used in the hospital and one used in the mother's home), and liquid human milk fortifier samples from three lots were cultured in gram-negative enrichment broth followed by selective media; C. sakazakii was recovered from the breast pump parts used in the home. An interview revealed that these breast pump parts were cleaned in a household sink, sanitized, and sometimes assembled while still moist. No bacteria were recovered from the expressed milk samples, liquid human milk fortifier samples, hospital breast pump parts, or unopened powdered formula from the hospital. C. sakazakii isolates from the patient's blood and CSF, and those isolated from environmental samples, were analyzed by WGS,§§ which indicated that the patient's CSF and blood isolates were closely genetically related to the isolate recovered from the home breast pump (0–1 SNPs).

*Genomic DNA was extracted and sequenced on MiSeq instruments (Illumina) following Nextera XT or DNA Prep library preparation.
Sequence reads were cleaned using run_assembly_trumClean.pl. https://github.com/lskatz/CG-Pipeline.
§Cleaned reads were assembled using SPAdes (version 3.14.0) https://github.com/ablab/spades.
Lyve-SET (version 1.1.4f) was used for high-quality SNP analysis on reads https://github.com/lskatz/lyve-SET.
**C. sakazakii ATCC 29544 was used as the reference genome.
†† C. sakazakii isolated from the open formula (2021EL-1052c) was used as the reference genome.
§§Methods for WGS analysis were identical to the investigation for patient A.

processing....