Gastric devascularization in a neonate with massive upper GI bleeding: a case report and literature review

Abstract

Severe upper gastrointestinal (UGI) bleeding in children is an uncommon but severe condition, with neonatal cases being rare. Herein, we described a 3-day-old boy with massive UGI bleeding due to diffuse hemorrhagic gastritis that was treated via gastric devascularization with a good outcome. We reviewed the literature on massive upper gastrointestinal bleeding in neonates retrieved from the PubMed database since 1977 and integrated the cases they reported. Based on the literature and our experience, gastric devascularization can serve as an alternative method for neonatal diffuse hemorrhagic gastritis.

Introduction

Upper gastrointestinal (UGI) bleeding refers to bleeding originating from the esophagus, stomach, and duodenum proximal to the ligament of Treitz, presenting with hematemesis and/or melena. Massive UGI bleeding in children is a rare but serious condition, with neonatal cases being particularly uncommon [1]. Its causes vary, including stress gastritis, gastric/duodenal ulcers, esophagitis, nasogastric tube trauma, gastric hyperplastic polyps, vascular malformations, gastrointestinal duplications, coagulopathy, prenatal stress, and cow’s milk protein allergy. During pregnancy, maternal use of dexamethasone, indomethacin, tolazoline, sulindac, or ketorolac is also considered a factor for neonatal UGI bleeding [2–4]. Previous studies indicate that conservative treatment, such as the use of proton pump inhibitors (PPIs) and histamine-2 (H2) receptor antagonists, can be initially attempted in children with upper gastrointestinal bleeding [1, 2, 5]. If conservative treatment fails to control the bleeding, exploratory procedures must be undertaken [2]. Nowadays, gastrointestinal endoscopy has become a preferred method for investigating major gastrointestinal bleeding. It can not only identify the cause of bleeding but also provide therapeutic intervention simultaneously. However, when fiberoptic endoscopy fails to achieve hemostasis or is unavailable, surgical exploration remains a crucial treatment option [2, 4, 6]. Herein, we reported a neonate presenting with hematemesis and melena due to diffuse hemorrhagic gastritis that was successfully treated via surgical stomach devascularization with a good outcome.

Case report

A male baby was born vaginally at 36⁺² weeks to a 30-year-old second-gravid mother who had conceived spontaneously. The antenatal period was uneventful. The mother was healthy during pregnancy and did not receive any medications, such as antiplatelet agents, antithrombotic drugs, or indomethacin. The Apgar score was 10 at one minute after birth. Eighteen hours later, the baby was transferred to the pediatric department for further treatment due to poor response and suspicion of hypoxic-ischemic encephalopathy (HIE). The birth weight of this patient at admission was 2.9 kg (Appropriate for Gestational Age, AGA), with a body temperature of 36.1 °C, a heart rate of 140 beats per minute, and a respiratory rate of 42 breaths per minute. Physical examination revealed poor responsiveness, cyanosis of the lips, and decreased muscle tone in the limbs, with no other significant abnormalities noted. Oxygen inhalation, keeping warm, fluid replacement, and medications nourishing brain cells (Vitamin B6, Coenzyme A, adenosine triphosphate) were given. On his second day of life, the baby presented with vomiting of coffee-ground-like contents. Nasogastric tube insertion, gastric lavage, intravenous infusion of Vitamin K, and cimetidine injection were given for suspicion of upper gastrointestinal hemorrhage. Laboratory tests showed that the white blood cell count (WBC) was 20.8 × 10⁹/L, the red blood cell count(RBC) was 5.53 × 10¹²/L, the hemoglobin level (Hb) was 220 g/L, the hematocrit was 59.3%, and the platelet count was 215 × 10⁹/L. On the third day of his life, the baby vomited fresh blood and passed a large amount of bloody stool. One day later, his red blood cell count decreased to 4.50 × 10¹²/L, hemoglobin 59 g/L, the hematocrit 45.3%, platelet count 162 × 10⁹/L, prothrombin time (PT) 28.6 s, partial thromboplastin time (APTT) greater than 180 s, thrombin time (TT) greater than 120 s. His heart rate was 124 beats per minute. Blood transfusion (160 ml concentrated red blood cells and 100 ml fresh frozen plasma), gastric lavage with cold normal saline, thrombin administration via nasogastric tube, intravenous infusion of the H2-receptor antagonist (cimetidine injection), and intravenous use of hemocoagulase were given but failed to control the bleeding. Dopamine was given intravenously at a dose of 2.5 mg. An urgent operation was indicated. Laparotomy was performed via an upper abdominal midline incision. The operative finding showed that a large amount of blood accumulated in the ileum and colon; petechial hemorrhage was observed on the gastric wall, and no ulcers were seen in the duodenum. Subsequently, after performing a gastrotomy of 3 cm in length on the anterior wall of the gastric greater curvature, diffuse active bleeding with a large amount of blood clots on the congested and edematous gastric mucosa was found. Acute diffuse hemorrhagic gastritis was confirmed. The area of oozing blood was extensive, and pressure failed to stop the bleeding. Ligation of the right gastroepiploic artery, right gastric artery, left gastric artery, and branches of the left gastroepiploic artery was performed. The short gastric arteries were preserved. No further active bleeding from the gastric mucosa was observed, and the blood supply to the stomach was sufficient. A two-layer suture with 4 − 0 absorbable sutures was used to close the gastric incision. 100 ml of concentrated red blood cells and 60 ml of fresh frozen plasma were transfused. On the first postoperative day, his RBC and Hb levels elevated to 4.19 × 10¹² and 128 g/L, respectively. The infant experienced episodes of respiratory arrest at the 7th and 16th hour postoperatively, primarily manifesting as cyanosis, bradycardia, and a decrease in oxygen saturation (maintained at 70%−80%). Cardiopulmonary resuscitation was successfully performed. Breastfeeding was initiated on the third postoperative day, and the baby tolerated feeding well. The patient was discharged home on the 14th postoperative day. During the one-year follow-up period, his body weight was 10 kg. No recurrent UGI bleeding occurred.

Discussion

Relevant articles in the international literature, retrieved from PubMed from 1977 to the present, were identified using the keywords gastrointestinal hemorrhage, upper GI bleeding, massive hemorrhage, massive transfusion, newborn, hematemesis, endoscopy, hematochezia, and peptic ulcer. We reviewed 23 articles and identified 29 cases of neonatal massive UGI bleeding [1–23]. We integrated the cases reported in these articles by evaluating their clinical manifestations, comorbidities, diagnoses, and treatments (Table 1).

Table 1.

Demographics, clinical features, risk factors, and outcome in neonates with massive UGI bleeding (n = 29)

Variables	Number of cases (n)	Percent (%)
Sex
Male	19	65.52%
Female	9	31.03%
NA	1	3.45%
Gestational age
Full term	24	82.76%
premature	3	10.34%
NA	2	6.90%
Birth weight
Normal range	18	62.07%
Low birth weight	7	24.14%
Very low birth weight	3	10.34%
NA	1	3.45%
Age at onset of bleeding (day)
0–3	23	79.31%
4–7	2	6.90%
> 7	4	13.79%
Symptomatolgy
hematemesis	24	82.76%
Fresh blood on the nasogastric tube	2	6.90%
Melaena	3	10.34%
Vital signs
Pulse > 120 beats per minute	14	48.28%
NA	14	48.28%
Low blood pressure	15	51.72%
NA	14	48.28%
Hematological investigations
Low hematocrit	12	41.38%
NA	17	58.62%
Hemoglobin < 150 g/L	16/19	55.17%/65.52%
NA	10	34.48%
Elevated WBC count	3	10.34%
Normal	10	34.48%
NA	16	55.17%
Coagulation studies
Decreased platelet count	2	6.90%
Within normal limits	17	58.62%
NA	10	34.48%
Prothrombin time (PT)
Normal	21	72.41%
Elongated	1	3.45%
NA	7	24.14%
Activated partial thromboplastin time (APTT)
Normal	12	41.38%
Elongated	7	24.14%
NA	10	34.48%
Other risk factors
Prenatal psychosocial maternal stress	2	6.90%
Birth asphyxia	2	6.90%
Drug (antenatal indomethacin exposure and stress)	1	3.45%
Diaphragmatic hernia repair with a prosthetic graft patch	1	3.45%
Heart surgery	1	3.45%
Diagnosis
Gastric ulcer	8	27.59%
Dieulafoy lesions (DL)	7	24.14%
Diffused hemorrhagic gastritis	5	17.24%
Duodenal ulcer	4	13.79%
Multiple gastric ulcers; duodenal ulcer	1	3.45%
Active hepatic arterial bleeding into the duodenum	1	3.45%
Massive esophageal and gastric varices	1	3.45%
Non-neoplastic non-hamartomatous gastric polyp	1	3.45%
Extensive gastric necrosis with numerous perforations	1	3.45%
Comorbidity
Hypoplastic left heart syndrome	1	3.45%
Diaphragmatic hernia	1	3.45%
Patent ductus arteriosus (PDA)	1	3.45%
Atrial septal defect, ventricular septal defect, and PDA	1	3.45%
Pyloric atresia	1	3.45%
Respiratory distress, necrotising enterocolitis (NEC)	1	3.45%
Total anomalous pulmonary venous drainage (TABVD), esophageal atresia with tracheoesophageal fistula	1	3.45%
Respiratory distress; NEC༛subglottic stenosis༛bronchopulmonary dysplasia	1	3.45%
Treatments
Antacid therapy
Histamine-2 (H2) blockers	9	31.03%
Proton pump inhibitors (PPI)	5	17.24%
Octreotide	1	3.45%
Blood transfusion	26	89.66%
>calculated blood volume	4	13.79%
Fresh frozen plasma	8	27.59%
Platelet/albumin	1/3	3.45%/10.34%
NA	3	10.34%
Vaspressor	2	6.90%
Non-operative treatment	13	44.83%
Endoscopic hemostatic clip for DL	2	6.90%
Endoscopic 1:10,000 epinephrine injection for DL	2	6.90%
Endoscopic monopolar thermocoagulation For DL	1	3.45%
Endoscopic gastric polyp resection	1	3.45%
Surgical management	9	31.03%
Gastric devascularization	2	6.90%
Spurting arterial vessel ligation for DL	1	3.45%
Simple bleeding arterial ligation and close of the ulcer	1	3.45%
Vagotomy and pyloroplasty	1	3.45%
Total gastrectomy and oesophagojejunostomy	1	3.45%
Repairing the hepatic artery (aneurysm)	1	3.45%
Evacuation of the hematoma, pyloroplasty, and gastrostomy	1	3.45%
Laparotomy, gastric membranous diaphragm excision, pyloroplasty	1	3.45%
Complications related to treatment
Repeat bleeding of hepatic artery aneurysm on postoperative day 6	1	3.45%
Occurrence of massive bleeding on postoperative day 4	1	3.45%
Stomach without emptying	1	3.45%
Sepsis	2	6.90%
Cardiorespiratory depression caused by antacid therapy related hypermagnesemia	1	3.45%
Outcomes
Alive	22	75.86%
Parents refused for further treatment and withdraw all the life supports due to unsatisfactory improvement and poor prognosis	1	3.45%
Death (possible cause)
Uncontrolled bleeding from esoghageal varices	1	3.45%
Uncontrolled massive bleeding from gastric ulcers	1	3.45%
Coagulase-negative staphylococcal sepsis	1	3.45%
Respiratory complications from MRSA	1	3.45%
Respiratory insufficiency	1	3.45%
Unknown outcome (Hypoplastic left heart syndrome and dyspnea)	1	3.45%

Abbreviation: APTT Activated partial thromboplastin time, DL Dieulafoy lesions, H2 Histamine-2, MRSA Methicillin-resistant Staphylococcus aureus, PPI Proton pump inhibitors, NA Not available, NEC Necrotising enterocolitis, PDA Patent ductus arteriosus, PT Prothrombin time, TABVD Total anomalous pulmonary venous drainage, UGI Upper gastrointestinal.

Description of the table: We conducted a statistical analysis of the 29 children included in our literature review, covering gender, gestational age, birth weight, age at onset of bleeding, symptoms, vital signs, blood test results (including hemoglobin, hematocrit, white blood cell count, coagulation parameters), diagnosis, comorbidities, managements, and outcome. We sought to find better treatment options for these children, utilizing beneficial therapeutic approaches.

Gastrointestinal bleeding in neonates and infants under 1 year of age has unique etiologies. The common cause of gastrointestinal bleeding in healthy neonates is swallowing maternal blood (from cracked nipples) [1, 9, 17, 24–26]. Therefore, to differentiate the source of bleeding, the Apt-Downey test can be applied. The mother of our patient did not have any nipple fissures, which ruled out the possibility of the bleeding originating from the mother.

Coagulopathy, thrombocytopenia, and stress gastritis or ulcers are common causes of gastrointestinal hemorrhage [5, 27, 28]. Among the cases we have documented, diffuse hemorrhagic gastritis and peptic ulcers were found in 18 cases (62.07%). Neonatal peptic ulcers are categorized into primary and secondary types, with secondary ulcers being more common and often caused by stress or medications, such as physiological stresses like shock, respiratory failure, hypoglycemia, cow milk protein intolerance, major surgery, and sepsis. Perinatal stress (such as asphyxia, prenatal medication use, prolonged labor, use of obstetric instruments, or psychosocial stress) can also lead to ulcers [1–3, 9, 11, 19, 28–30]. These stressors can disrupt gastric mucosal homeostasis, resulting in ulcer formation. Among the 29 reported cases, 4 were associated with these stressors. Gastrointestinal bleeding in neonates due to vitamin K deficiency may be related to the mother’s use of antibiotics or vitamin K antagonists during pregnancy [2, 6, 25]. In our case, the mother had no history of medication use during pregnancy, and the neonate was administered vitamin K after birth and following upper gastrointestinal bleeding, which helps reduce the risk of bleeding caused by vitamin K deficiency.

If a healthy neonate suddenly experiences acute gastrointestinal bleeding, Dieulafoy lesions (DL), gastrointestinal vascular malformations, or Meckel’s diverticulum may also need to be considered [8, 13, 18, 23, 26, 28]. Among the 29 reported cases, 4 were attributed to uncommon etiologies of hemorrhage.

Massive upper gastrointestinal bleeding typically manifests as hematemesis, melena, or hematochezia. Hematemesis can be either bright red blood or a coffee-ground-like substance. When hemoglobin is digested by digestive enzymes, it results in melena. However, if UGI bleeding is too rapid, hematochezia may appear. It has been reported that among every 10 patients with hematochezia, one may have upper gastrointestinal bleeding [12, 17, 25–31]. Among these 29 cases, 24 (82.76%) patients presented with hematemesis, 3 (10.34%) with melena, and 2 (6.90%) with fresh blood on the nasogastric tube. Our patient presented with both hematemesis and melena.

Acute gastric peptic ulcers can be managed conservatively [25]. A gastric pH below 2.5 is one of the risk factors for stress ulcers and gastrointestinal bleeding [1, 27]. Proton pump inhibitors (PPIs) and histamine-2 (H₂) receptor antagonists can significantly increase gastric pH and improve UGI bleeding in neonates [2, 6, 26, 29]. Additionally, PPI can reduce the rebleeding rate, blood transfusion requirements, and hospital stay for children with UGI bleeding caused by gastric or duodenal ulcers [1, 2, 5, 24, 25, 27–39, 31]. Among these 29 cases, 9 (31.03%) patients were treated with H₂ receptor antagonists (including our case) and 5 (17.24%) with PPI; Octreotide is a somatostatin analog that has been proven to alleviate gastrointestinal bleeding in children and is safe and effective for severe non-arterial gastrointestinal bleeding [25, 28, 29, 31]. Among these 29 cases, 1 (3.45%) patient received octreotide.

Massive UGI bleeding typically leads to poor blood perfusion of terminal organs, resulting in organ damage [28]. Infusion of red blood cells can maintain hemoglobin levels to transport oxygen to tissues. Neonates have a lower blood volume compared to adults, and massive UGI bleeding can rapidly worsen their condition. More importantly, the measured hemoglobin concentration lags behind the actual hemoglobin level, making it unreliable in the presence of active bleeding [25, 28]. Therefore, some studies suggest that early blood transfusion should be considered in cases of massive UGI bleeding [25, 26, 28, 29]. Among these 29 cases, 12 (41.38%) had low hematocrit, 16 (55.17%) had hemoglobin < 150 g/L, and 26 (89.66%) received transfusions. Among them, 4 (13.79%) received a transfusion volume exceeding their calculated blood volume, and 8 (27.59%) were transfused with fresh frozen plasma. Our patient received a transfusion of 260 ml of packed red blood cells, which exceeded his calculated blood volume (240 ml).

If conservative treatment fails, exploration is necessary. Nowadays, gastrointestinal endoscopy has become a preferred method for investigating major gastrointestinal bleeding, as it can identify the specific cause of gastrointestinal bleeding in children, even in neonates [2, 4, 6, 23, 25–32]. It can also provide targeted treatment for specific etiologies, such as varices and ulcers [20, 25, 27, 28]. Among the 29 reported cases, 6 underwent endoscopic procedures. However, upper gastrointestinal endoscopy may not be suitable for neonates with massive UGI bleeding due to their severe conditions and technical difficulty.

If the above measures fail to control the massive UGI bleeding, surgical treatment should be required [8, 14, 26, 29]. Currently, there are several surgical options, including suture ligation of the bleeding points in the mucosa, vagotomy alone, vagotomy with pyloroplasty, radical subtotal gastrectomy, total gastrectomy, and gastric devascularization [2, 9, 14, 16]. Radical subtotal gastrectomy and total gastrectomy are not suitable for neonates and growing children. Vagotomy alone and vagotomy with pyloroplasty are feasible and do not have long-term adverse effects on the children’s growth and development. Diffuse massive bleeding may necessitate a total gastrectomy. The rebleeding rate after vagotomy is high, and the mortality rate associated with subtotal gastrectomy is also high. In children, especially in small babies, these procedures might have adverse effects on their nutrition, growth, and development [14, 16, 33, 34]. Studies have found that acute erosive-hemorrhagic gastritis is associated with gastric mucosa congestion rather than local ischemia. Experiments have confirmed that the total blood flow of the gastric mucosa immediately decreases after gastric devascularization [35]. This procedure has been applied in neonates, with satisfactory hemostasis effect and long-term outcome [14, 35]. Among the 29 reported cases, 9 underwent surgery. In our patient, intraoperative findings showed massive upper gastrointestinal bleeding caused by acute diffuse hemorrhagic gastritis. We chose gastric devascularization as an option with good results in a one-year follow-up period after surgery. In the reviewed literature, the cases using gastric devascularization [14, 35] are similar to our cases: bleeding could not be controlled after conservative treatment; operative findings confirmed the cause of bleeding was acute diffuse hemorrhagic gastritis; both cases achieved hemostasis, with no recurrent bleeding, with a good outcome. Gastric devascularization is not technically difficult, and when surgical exploration confirms acute diffuse hemorrhagic gastritis as the source of bleeding, gastric devascularization may be considered as an alternative to simple vagotomy or vagotomy with pyloroplasty for neonatal hemorrhagic gastritis [14]. Although gastric devascularization has yielded good results, it still has some potential risks, such as the possibility of gastric ischemia, necrosis, or rebleeding after surgery [9, 14, 33–35].

In conclusion, massive UGI bleeding in neonates is a rare but life-threatening condition. Hematemesis and/or melena accompanied by hemodynamic instability should raise a suspicion of UGI bleeding. Stress gastritis and ulcers are common causes. If the patient’s status is stable after blood transfusion, early endoscopic assessment might be an option for accurate diagnosis and precise management. In cases with uncontrolled bleeding, surgery remains a vital option for the management of massive UGI bleeding in neonates. However, prospective randomized clinical studies are needed.

Abbreviations

APTT

Activated partial thromboplastin time

DL

Dieulafoy lesions

H2

Histamine−2

MRSA

Methicillin−resistant Staphylococcus aureus

PPI

Proton pump inhibitors

NA

Not available

NEC

Necrotising enterocolitis

PDA

Patent ductus arteriosus

PT

Prothrombin time

TABVD

Total anomalous pulmonary venous drainage

UGI

Upper gastrointestinal

Authors’ contributions

XL: Data curation, Methodology, Software, Validation, Writing–original draft, Writing–review & editing. YH: Data curation, Supervision, Writing–original draft, Writing–review & editing. MH: Data curation, Supervision, Writing–original draft, Writing–review & editing. TF: Data curation, Methodology, Writing–original draft, Writing–review & editing. LG: Conceptualization, Supervision, Writing–original draft, Writing–review & editing.

Funding

None.

Data availability

Data is provided within the manuscript.

Declarations

Ethics approval and consent to participate

The ethics waiver was approved by the ethical committee of Binzhou Medical University Hospital due to a case report study. Written informed consent was obtained from the patient’s legal guardian for publication of this case report.

Consent for publication

Written informed consent was obtained from the patient’s mother (legal guardian) for the publication of this case report and any accompanying clinical details.

Competing interests

The authors declare no competing interests.