Serratia marcescens and its role in Pink breast milk and upper GI bleeding in an infant: a Tanzanian case report

Abstract

Background

Serratia marcescens, a gram-negative bacterium once considered non-pathogenic, is now recognized as a cause of hospital-acquired infections, particularly in immunocompromised patients with outbreaks reported in neonatal intensive care units. In newborns, it can result in severe illness, including sepsis and meningitis. It is well recognized by the production of pink pigment found on drains and other equipment, including breast pumps.

Case report

We report a case of a two-month-old previously healthy infant with upper gastrointestinal bleeding, whose mother had pink-stained expressed breast milk. Cultures of the breast milk and gastrointestinal aspirates of the infant grew S. marcescens. The infant received proton pump inhibitors, blood transfusions, and antibiotics, with successful resolution of the bleeding.

Conclusion

S. marcescens should be considered a potential cause of gastrointestinal bleeding in infants at risk. Although previously known to cause outbreaks in vulnerable infants, pathogenicity in healthy infants is also possible. Maintaining strict hygiene of feeding and breast milk pumping equipment is important to prevent infections in infants.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-025-10862-1.

Introduction

Upper gastrointestinal bleed (UGIB)- defined as gastrointestinal bleed above the ligament of Treitz- is one of the most urgent gastrointestinal emergencies in pediatric population. It is uncommon in healthy children, occurring in 1 in 10,000 children [1]. Common causes of UGIB in infants are milk protein sensitivity, Mallory Weis tear, and oesophagitis [2]. Infections except H. pylori are rarely identified as the cause of UGIB, however, opportunistic infections such as cytomegalovirus, herpes, and fungal infections have been identified to cause haematemesis in infants [2].

We present a case of UGIB in an infant caused by erosive gastritis with isolation of Serratia marcescens from gastrointestinal secretions. S. marcescens is a gram-negative Enterobacteriaceae previously considered non-pathogenic but has recently been associated with a variety of infections, including gastroenteritis and necrotizing enterocolitis in newborns [3]. Like other enteric bacterial pathogens, it possesses various virulence factors, including fimbriae for adherence to epithelial cells, the RssAB-FlhDC-ShlBA pathway- a major pathogenesis pathway for hemolytic activity and biofilm attachment, quorum sensing systems, and a range of secreted enzymes, such as phospholipase, which is similar to that produced by Helicobacter pylori and can cause gastric cellular damage [4].

This species is commonly found in the environment and is known to colonize hospital instruments and breast milk pumps [5], which contributes to the high incidence of nosocomial infections associated with it. It is frequently responsible for the pinkish discoloration and biofilm [6–10]. The compound responsible for the red or pink coloration is prodigiosin, a member of the prodiginine family of linear and cyclic tripyrrolic compounds [11].

Case

A previously well, two-month-old female infant presented with a one-day history of sudden onset of regurgitation of blood-stained saliva. This was noticed as blood staining around the mouth and on the napkin by the mother after breastfeeding. The episode was also associated with passing black stool twice. There were no reports of irritability, crying, refusal to feed, vomiting, or fever.

She was delivered at term by cesarean section to non-consanguineous parents with no family history of bleeding disorders. She had received vitamin K at birth and was immunized according to the national vaccination program. The infant was on exclusive breast milk feeding, directly at the breast and with expressed breast milk (EBM), which was pumped using a hand pump. Mother had no mastitis or infection.

On examination, the infant appeared playful and was not pale. Her oral mucosa was clear, with no signs of bleeding, and no petechiae or purpura were noted. The rest of her examination was unremarkable.

A diagnosis of UGIB was established, with a differential diagnosis including late-onset hemorrhagic disease of the newborn, coagulation disorder, and H.pylori infection.

The infant was admitted to the pediatric ward for monitoring and workup. A complete blood count (CBC) showed hemoglobin of 9.6 g/dl (normal 9.4–13.0 g/dl), platelets of 632 × 10⁹(normal: 150–450 × 10⁹), and white blood cells (WBC) of 10.83 × 10⁹ (normal 10–26 × 10⁹). Activated partial thromboplastin time (aPTT) was 31.7 s (normal: 30–40 s), while prothrombin time (PT) was 10.1 s (normal: 11.0–13.0 s), and international normalization ratio (INR) was 0.92 (normal: 0.9–1.15). C-reactive protein (CRP) was 1.12 mg/l (normal: 0.5–5.5). Liver function test (LFT) was normal, and stool for occult blood and H. Pylori was negative. An abdominal ultrasound was normal.

1 mg of intramuscular vitamin K was administered, and intravenous Esomeprazole infusion at 1.2 mg/hr was started while continuing breastfeeding. On admission day, the infant passed black stool once. Thereafter, no further bleeding was noted, and the infant was discharged on the third day on esomeprazole tablet 5 mg twice a day.

Three days after discharge, she was readmitted due to regurgitating up ring-like clots of blood (Fig. 1), which the mother noticed after breastfeeding. She also passed a dark, red-colored soft stool (Fig. 2), and the mother noticed that expressed breast milk stored in the bottle, which had remained at room temperature for 2 h, had turned pink (Fig. 3).

Fig. 1.

Ring-like blood clot from the mouth

Fig. 2.

Fresh blood in the diaper

Fig. 3.

Feeding bottle with pink milk

On assessment, the baby was tachycardic at 185 bpm (normal: 100-160 bpm), moderately pale, not dehydrated, with a soft abdomen and no organomegaly. The infant was admitted to the pediatric high dependency unit, kept nil orally, with a nasogastric tube (NGT) inserted for drainage. IV dextrose saline was administered at a maintenance rate of 22 ml/h, and IV esomeprazole was restarted.

CBC showed severe anemia with hemoglobin of 5.2 g/dl, WBC of 10.46 × 10⁹/l, platelets of 419 × 10⁹/l. The APTT, PT, and INR remained within normal. CRP was 7.05 mg/L. Stool, nasogastric aspirate, and discolored breastmilk from the bottle were cultured and grew a smooth, medium-sized, pink-red, and non-fermenting colony on Mac Conkey agar plate, as shown in Fig. 4. The colonies showed gram-negative rods on gram stain. The identification and Antimicrobial Susceptibility Testing (AST) procedure was performed by using the Vitek 2 compact equipment (bioMérieux, France) and AST interpreted according to the Clinical Laboratory Standards Institute guidelines (CLSI), and the pathogen isolated was Serratia marcescens sensitive to Aminoglycosides, Carbapenems, Cephalosporins, Trimethoprim/Sulfamethoxazole, and Fluoroquinolones. A multiplex PCR to identify biofilm-associated genes in S. marcescens was not done due to the unavailability of the test locally.

Fig. 4.

Serratia Colonies

Post-admission, the infant passed melena twice and frank blood once, associated with hematemesis and restlessness. The infant was transfused with packed red blood cells twice and given tranexamic acid infusion. A second dose of Vitamin K was also administered, and the infant was transferred to the Paediatric Intensive Care Unit.

An oesophagogastroduodenoscopy (OGD) revealed blood clots on antral mucosa and multiple corpus erosive lesions with superficial ulceration and blood clots (Fig. 5). A tissue sample was not collected due to the unavailability of a pediatric-size probe. IV amikacin (15 mg/kg once daily) and IV cefotaxime (50 mg/kg every 6 h) were initiated to treat the Serratia infection. Fresh hand-expressed breast milk was collected for culture, but no growth was observed.

Fig. 5.

Antral hemorrhage (shown by black arrow) and blood clots (shown by blue arrow), erosion (shown by red arrow)

One day post-OGD, the infant had one episode of melena and had stopped regurgitating blood. Tranexamic acid infusion was stopped, and feeding with 10 mL of EBM every 3 h was initiated and well tolerated. Esomeprazole infusion was continued for the next 4 days while oral feeds were gradually increased. Breastfeeding was initiated after 3 days. The infant had no further episodes of melena stools or regurgitation of blood; antibiotics were switched to oral ciprofloxacin (10 mg/kg every 12 h) and oral cefixime (4 mg/kg every 12 h). The infant was discharged on the 6th day, with hemoglobin of 10.4 g/dl and CRP of 0.61 mg/l.

On discharge, the mother was educated on breastfeeding equipment hygiene; previously, the mother would soak the bottles, pumping unit, and cleaning brush in hot water for about 30 min, as opposed to the general guideline of submerging breastfeeding equipment in rolling boiling water for at least 10 min. Upon follow-up, the infant has been doing well, with no recurrence of melena or blood-stained regurgitant.

Discussion

This is a case of a previously healthy infant who presented with UGIB with isolation of Serratia marcescens from GI secretions as well as expressed breast milk. This highlights the possible role of Serratia marcescens as a cause of UGIB in a healthy infant. S. marcescens was once thought to be non-pathogenic, but recent findings indicate that it is responsible for a range of systemic infections, predominantly in immunocompromised individuals [12]. It has been linked to conditions such as endocarditis, intra-abdominal infections, meningitis, osteomyelitis, soft tissue and ocular infections, pneumonia, gastrointestinal diseases, and necrotizing enterocolitis, contributing to significant morbidity and mortality [4, 5, 12–14].

It is believed that even a relatively small number of S. marcescens can lead to infection in healthy individuals; studies have shown that ingesting as few as 10 to 100 S. marcescens bacteria can result in gastroenteritis [15]. Symptoms of gastrointestinal infection caused by S. marcescens may include fever, chills, diarrhea, and abdominal cramps [15, 16]. This patient exhibited progressively worsening UGIB, which was confirmed to be due to erosive gastritis through an OGD. Although these symptoms have not been previously documented, the presence of positive cultures for S. marcescens in gastric aspirates, along with negative results for Helicobacter pylori, indicates that S. marcescens is the likely cause.

The diagnosis of S. marcescens is confirmed by isolating the organism through the culture of appropriate clinical specimens, such as blood, sputum, urine, or aspirated body fluids. In our case, we cultured samples from gastrointestinal aspirates and milk from the suspected contaminated feeding equipment. Both cultures yielded gram-negative rods, which were subsequently identified as S. marcescens. However, the culture from the fresh hand-expressed milk by the mother did not show any growth of organisms. This indicated that only the breast pump and the bottle used for collecting and storing the milk were colonized by the bacterium.

There have been multiple reports of Serratia species being isolated from breast pumps [17–21], including one case where a mother informed her obstetrician that the tubing of her breast pump had turned bright pink. She used the expressed milk to feed her twin infants and showed no signs of mastitis [22]. Unlike this infant, neither of the cases reported a sick infant, except for some reports of increased fussiness during feeding. In this case, the infant was breastfeeding and bottle-feeding with milk expressed using an electric breast pump. The mother observed that the milk in the bottle, which was initially white, later turned pink, indicating that both the breast pump and the bottle were contaminated with S.marcescens [23]. Breast pump equipment should be washed using warm water (30◦C) and washing liquid after every use and disinfected at least once a day using either boiling water at 100◦C for 10 min, steam sterilizing in an electric or microwave sterilizer, or cold water and sterilizing solution such as 1% hypochlorite or 70% ethanol for 30 min [24, 25].

Antibiotics are employed to treat infections caused by S. marcescens, which typically show susceptibility to various classes of antibiotics, including aminoglycosides and cephalosporins [26]. The microbial sensitivity testing conducted in our laboratory also indicated sensitivity to aminoglycosides, cephalosporins, and fluoroquinolones. Serratia species are resistant to beta-lactamase, narrow-spectrum cephalosporins including cefazolin, and macrolide [27]. A similar resistance pattern was seen in our case.

Conclusion

This case demonstrates a rare but serious presentation of UGIB in an infant, where Serratia marcescens was identified as the possible cause of erosive gastritis leading to bleeding. Although previously considered harmless, it is now recognized as a significant pathogen, particularly in infant and neonatal care. It is crucial to provide new mothers with appropriate resources and education on recognizing the signs and symptoms of bacterial infections. Additionally, demonstrating proper cleaning and disinfection techniques for breast pump equipment would be beneficial in preventing infections. Early intervention and appropriate treatment are crucial for managing infections and ensuring the recovery of affected infants. Further research is needed to confirm the role of Serratia species in erosive gastritis.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

aPTT

Activated partial thromboplastin time

AST

Antimicrobial Susceptibility Testing

CBC

Complete Blood Count

CRP

C-Reactive Protein

EBM

Expressed Breast Milk

H.Pylori

Helicobacter Pylori

INR

International Normalisation Ratio

IV

Intravenous

IM

Intramuscular

LFT

Liver function test

NGT

Nasogastric Tube

NPO

Nil per oral

OGD

Oesophagogastroduodenoscopy

PPI

Proton Pump Inhibitor

PT

Prothrombin Time

UGIB

Upper Gastrointestinal Bleeding

Author contributions

R.O: writing and editing, prepared figuresM.M: Writing J.S: Writing M.B: reviewingH.S: Writing.M.E: reviewingN.W: reviewingM.D.N: prepared figures 5 M.N: supervising, reviewing and editingAll authors reviewed the manuscript.

Funding

This report was not funded.

Data availability

Not applicable.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

Not applicable.

Clinical trial number

Not applicable.

Consent for publication

Written informed consent for publication of clinical details and clinical images was obtained from the parents of the infant and is available on request.