Hematochezia and thrombocytopenia in a 3-day-old infant: Congenital adenoviral infection

1 Case report

This full term female newborn baby presented at 3 days of life with bloody stools, thrombocytopenia and elevated C-reactive protein. She was born to a 33-year-old mother at 39 5/7 weeks of gestation with a birthweight of 3.5 kilograms at a regional hospital in the state of Wyoming. Maternal prenatal laboratories were within normal limits, screening for GBS was negative. Prenatal medications included prenatal vitamins and Zofran. The mother denied alcohol, tobacco, or recreational drug use. She had 2 living healthy children. The mother had no documented adenoviral infection either during pregnancy or at delivery, although she reported to be sick with fatigue, headache, abdominal pain with vomiting, and diarrhea approximately 10 days before the date of delivery. Her symptoms lasted 3 to 4 days with full recovery. There were no other sick contacts.

This baby was born following elective induction, artificial rupture of membranes, a vacuum extraction that failed, and a high forceps delivery. Apgars were 7 at 1 minute and 8 at 5 minutes. The baby initially remained with her mother but was brought into the well-baby nursery where she was found to have hypoglycemia at 4 hours of life. She was started on Dextrose 10 intravenous fluids plus breast milk and formula feedings. At that time the infant was also started on ampicillin and gentamicin for rule out sepsis. Blood cultures remained negative for 48 hours. On her third day of life she had an episode of frank bloody stool. Oral feedings were stopped and an abdominal X-ray was obtained that was negative for intraabdominal pathology. A complete blood count showed a white count of 7.9 with 20 polymorphonuclear and 4 bands, hematocrit was 54%, and platelet count was low at 80,000/uL. C-reactive protein was elevated at 7.3 mg/dL. Decision was made to transfer the patient to Primary Children’s Hospital in Salt Lake City, UT for further management. The baby remained in isolation throughout her hospital stay. Physical exam and vital signs upon admission to Primary Children’s Hospital neonatal intensive care unit was only remarkable for mild jaundice. The baby had no respiratory distress and did not require supplemental oxygen during the hospital course. Abdominal exam was also benign and she had positive bowel sounds. The anus was patent with no evidence of fissures. Due to the history of bloody stools and thrombocytopenia, viral etiology was suspected and stool viral studies were sent. The baby remained n.p.o on IV fluids from the time of admission until day of 5 when Adenovirus was detected in the stool sample that had been sent upon admission. Other laboratories included screening for Rotavirus and Enterovirus in stool which were negative. CMV in urine was also negative. Liver function demonstrated a Direct Bilirubin of 1 mg/dL on the day of admission that increased to 1.3 mg/dL on day of life 7. AST was 53 U/L and ALT was 17 U/L on the day of admission and were not rechecked. Prothrombin Time (PT) and partial thromboplastin time (PTT) were 16.2 seconds and 42.4 seconds respectively. Renal function panel was also within normal limits.

The thrombocytopenia persisted and on day of life 4 platelet count was as low as 63,000/uL. However, the infant did not have any further gastrointestinal hemorrhage. The patient did have intermittent emesis that resolved by the time of discharge.

The remainder hospital course was significant for slow reintroduction of oral feedings and completion of 7 days of Ampicillin and Gentamicin. CRP decreased to 0.8 mg/dL by day of life 8 and platelet count rapidly normalized (125,000/uL by day of life 8 and 245,000/uL by day of life 12).

The infant was discharged home on full oral feedings consisting of ad lib breast feeding and supplementation with Similac Total Comfort on day of life 13. Weight at time discharge was 3560 grams.

2 Discussion

Neonatal adenovirus infection is infrequent, but in most cases the outcome of the infection is fatal [1–3]. Previous reports have shown a mortality rate greater than 50% for infants with pneumonia and greater than 85% for infants with disseminated disease [4, 5]. In this case report we describe the hospital course of a 3 day old newborn patient presenting with bloody stools and thrombocytopenia that remained clinically stable and had a benign course following adenovirus infection.

Adenoviruses are associated with a broad spectrum of clinical presentations in adult and children. There are at least 55 serotypes of Adenovirus subdivided in seven species (A to G) that can cause disease in humans [6]. More than 80% of these infections occur in children younger than 4 years of age, typically resulting in a self-limited disease affecting the gastrointestinal and respiratory tracts. In developed countries serotypes 40 and 41 (in the past called fastidious enteric adenoviruses) make up to 20% of hospitalizations for diarrhea among children less than 2 years of age [7]. However, older children and adults can also become infected, with or without symptoms. Incubation typically takes between 3 and 10 days, with symptoms of illness lasting for up to 7 days (longer than for other enteric viral pathogens). Rare manifestations of Adenovirus infections include hepatitis, hemorrhagic colitis, pancreatitis, hemorrhagic cystitis, nephritis, or encephalitis [2]. Infants, neonates, and preterm infants are susceptible to severe disease, often with extrapulmonary dissemination [8, 9].

Several case reports of newborn infants with localized and disseminated infection have been published up to date [10–13]. In 2015 Ronchi et al., published the largest cohort of newborn infants including all neonates with adenoviral infection collected retrospectively at 2 hospitals in Texas during a 17-year period [1]. In this report the authors combined the 28 patients that were diagnosed during the 17-year-period in Texas with 72 patients collected in 38 reports published since 1939 when the first histopathologic description of findings suggestive of neonatal adenoviral pneumonia was reported. Similar to the previous reports, adenoviral infection in the Texas cohort was associated with significant morbidity and mortality. Clinical presentation ranged from upper respiratory tract infection to fulminant disease with pneumonia, apnea, hypotension, and hepatitis. Mortality in patients with disseminated disease was 80% in this case series. Infants less than 14 days of age at presentation were more likely to have disseminated disease (44% vs 12%, P = 0.004) and death (48% vs 8%; P < 0.001). The most common clinical presentation of adenovirus infection was respiratory abnormalities in 76% (75/98) of patients. Pneumonia was present in 50% of these infants and 71% died. Compared with localized disease, disseminated infection was associated with respiratory failure requiring more significant respiratory support such as mechanical ventilation or ECMO (54% [19/35] vs 29% [18/63]; P = 0.02)

In neonates the potential sources of adenovirus infection include vertical and horizontal transmission (from mother or other family members) [3, 14]. Most commonly neonates acquire the infection horizontally as demonstrated in the review by Ronchi et al. in which the mean age of presentation was 16±8 days [1]. The presence of maternal symptoms prior to or shortly after delivery and the development of symptoms in the newborn infant within the first few days of life suggest possible in utero transmission. The newborn in our case report had Adenovirus 40 and 41, most likely acquired via vertical fashion since the mother had become symptomatic approximately 10 days prior to delivery. Furthermore, there were no ill contacts following delivery. In utero transmission of Adenovirus has been demonstrated by detection of adenovirus DNA by PCR in amniotic fluid of a pregnancy complicated by non-immune hydrops and fetal tachyarrhythmia [15].

It is not well understood what the contributing factors are in regards to the severity and wide variety of presentations of adenovirus infection in the neonates. It has been proposed that the disparity in severity may be serotype dependent: it would appear that infection with subgroup D strains is mostly associated with only transient conjunctivitis, whereas fatal infection appears to have been mainly associated with serotypes of subgroup B, possibly due to their predilection for respiratory tissues [16]. However, Faden et al. reported that adenovirus type 30 caused conjunctivitis in 8 of 21 infected infants during an outbreak in a neonatal intensive care unit and it was the only clinical finding in 7 of the infants. But pneumonia occurred in 8 infants with the same serotype, and in 7 of these infants it was the sole presentation. Six of the infants with pneumonia died [17].

Differential diagnoses of hematochezia in the newborn period once necrotizing enterocolitis has been ruled out (negative abdominal X-rays findings) should include viral etiology. A recent survey showed that only 34% NICU of practitioners would consider stool viral testing as a part of hematochezia work up [18]. However, early diagnosis of viral etiology can prevent prolonged antibiotic treatment and unnecessary fasting. Currently available Adenovirus testing includes culture, polymerase chain reaction (PCR) assays and Adenovirus-specific enzyme immunoassay antigen immunoassay (EIA) [19].

Differential diagnosis of thrombocytopenia in an otherwise well looking newborn include fetal and neonatal alloimmune thrombocytopenia (FNAIT). The suspicion of FNAIT is typically raised when a newborn (typically a first-born child) develops widespread skin petechiae and/or bruising shortly after birth and blood tests show severe thrombocytopenia (platelet count <50,000/uL). Intracranial hemorrhage secondary to FNAIT has an incidence of 13–21%. Extracranial hemorrhage sites include gastrointestinal and pulmonary bleeding.

The patient presented in this case report was not first child, did not present with petechiae/bruising and the platelet count was never <50,000/uL, making FNAIT an unlikely etiology [20].

The patient described in this report remained in isolation with contact and droplet precautions (routine use of gowns, gloves, and mask) from the time of admission until discharge home. Initial isolation was part of our admission protocol but following the detection of adenovirus in the stool, she remained in isolation to comply with infection control measures. No other NICU patients or personnel taking care of this patient showed evidence of Adenovirus infection. The report by Faden et al., highlights the importance of instituting infection control protocols to avoid dissemination of Adenovirus to vulnerable premature infants and immunocompromised patients in the NICU setting. Adenoviruses are very stable agents with prolonged shedding period of time. These factors may account for the ease in nosocomial dissemination from droplets, respiratory tract secretions, fecal-oral route, contaminated surfaces and hands, and equipment.

Lastly, medical management of Adenovirus infection is mainly supportive measures. Antiviral agents, like Cidofovir, are reserved for patients who are immunocompromised or have disseminated disease [1].

In conclusion, neonatal adenovirus infection, which is frequently disseminated and generally fatal as described above, should be considered in the differential diagnosis of neonatal sepsis and pneumonia but also in patients that present with hematochezia, thrombocytopenia and/or other less specific signs and symptoms of viral illness.