Neonatal Sepsis Caused by Streptococcus gallolyticus Complicated with Pulmonary Hypertension: A Case-Report and a Systematic Literature Review

Streptococcus gallolyticus (S. gallolyticus) has been linked to the development of infections in adults; however, in neonates S. gallolyticus sepsis is very rare and resembles Group B Streptococcal infections. In this case report, we present the case of a full-term neonate who developed early-onset sepsis due to S. gallolyticus. A systematic review of the literature was also conducted. The neonate had good APGAR scores at 1′ and 5′. At 5 h postnatally, the neonate developed poor feeding and respiratory distress. She received oxygen in a head box, and a complete blood count and biochemistry, blood, CSF and body surface cultures were obtained. Empiric intravenous antibiotics (ampicillin and tobramycin) were initiated, and she was transferred to a tertiary NICU for further treatment. The neonate was mechanically ventilated and received dopamine and colloid fluids for circulatory support. A cardiology consultation revealed pulmonary hypertension on day one. S. gallolyticus was isolated in the blood culture. Central nervous system ultrasonography, brainstem auditory evoked potentials, and a second cardiology evaluation were normal on day three. Clinical and laboratory improvement was noted on day three, and the baby was discharged after a 12-day hospitalization. Follow-up visits were scheduled for reevaluation.


Introduction
Streptococcus gallolyticus (S. gallolyticus) is a gram-positive coccus, which belongs to group D Streptococci, frequently found as part of the human intestinal flora, previously known as Streptococcus bovis (S. bovis) [1]. S. gallolyticus has been linked with infective endocarditis and meningitis in adults. Colon cancer is recognized as a risk factor for the development of bacteriemia, resulting in the aforementioned infections [2]. However, S. gallolyticus is a rare cause of neonatal infection, both in full-and pre-term neonates, and its clinical presentation resembles that of group B Streptococcal (GBS) infections [1]. We present the case of a full-term female neonate who developed early-onset sepsis due to S. gallolyticus, complicated with pulmonary hypertension. We also report all cases of neonatal infection caused by this infectious organism found in the literature.

Materials and Methods
Our systematic literature review was conducted until 27 October 2022 and followed the protocol proposed by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). We performed a Pubmed and Scopus search using the following keywords: "neonate", "newborn", "infant", "Streptococcus bovis", "Streptococcus gallolyticus". An additional manual electronic search was carried out to identify reports not found in our initial search. We used predetermined inclusion and exclusion criteria to find all relevant articles. Reports describing cases of neonatal infection by S. gallolyticus were eligible to be included in our review, and there was no restriction on publication year. Reports referring to cases of S. gallolyticus infection in any age group other than neonates, reports not including identification of S. bovis to the subspecies level, reviews, as well as studies not published in English were excluded from our study. Two authors (KK, MT) independently screened titles and abstracts of the retrieved studies for possible inclusion in the review and then reviewed the articles in depth. Any disagreement was resolved by a third author (RS).

Results
Our systematic search retrieved 245 articles. Duplicates were removed, and the remaining 178 studies were screened for eligibility. Ninety-one records were excluded during the process of title and/or abstract evaluation, and six reports were not accessible in full text to the authors; therefore, eighty-one studies were assessed in full text for possible inclusion in our review. Finally, 26 reports fulfilled the selection criteria and were thus included in our study. The study selection process is depicted in the respective flowchart ( Figure 1).

Materials and Methods
Our systematic literature review was conducted until 27 October 2022 and followed the protocol proposed by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA). We performed a Pubmed and Scopus search using the following keywords: "neonate", "newborn", "infant", "Streptococcus bovis", "Streptococcus gallolyticus". An additional manual electronic search was carried out to identify reports not found in our initial search. We used predetermined inclusion and exclusion criteria to find all relevant articles. Reports describing cases of neonatal infection by S. gallolyticus were eligible to be included in our review, and there was no restriction on publication year. Reports referring to cases of S. gallolyticus infection in any age group other than neonates, reports not including identification of S. bovis to the subspecies level, reviews, as well as studies not published in English were excluded from our study. Two authors (KK, MT) independently screened titles and abstracts of the retrieved studies for possible inclusion in the review and then reviewed the articles in depth. Any disagreement was resolved by a third author (RS).

Results
Our systematic search retrieved 245 articles. Duplicates were removed, and the remaining 178 studies were screened for eligibility. Ninety-one records were excluded during the process of title and/or abstract evaluation, and six reports were not accessible in full text to the authors; therefore, eighty-one studies were assessed in full text for possible inclusion in our review. Finally, 26 reports fulfilled the selection criteria and were thus included in our study. The study selection process is depicted in the respective flowchart ( Figure 1).  A total of 26 articles reported 66 cases . For reasons of completeness, we decided to include in our review three cases of neonatal infection caused by S. bovis biotype II, since, according to the bibliography, neonatal infections caused by S. gallolyticus species (i.e., S. bovis biotype II/2) are much more common than neonatal infections caused by S. infantarius (i.e., S. bovis biotype II/1). We analyzed these 66 cases and the present case. Characteristics of the cases describing neonatal infection caused by S. gallolyticus and S. bovis biotype II are summarized in Tables 1 and 2. Preterm and term newborns were equally represented among the cases included in this review. Twenty-eight cases (49.5%) referred to preterm neonates (GA 26-36 +6 ), and twenty-eight (49.5%) to term neonates. One neonate was post-term, while gestational age was not reported in 10 cases. Regarding the type of delivery, there were 14 cesarean sections (35%), 26 vaginal deliveries (65%), and, in 27 cases, the delivery type was not recorded. Data regarding the length of rupture of membranes were only available in 17 cases; among them, there was only 1 case of prolonged rupture of membranes (preterm premature rupture of the membranes, PPROM). Similarly, details concerning maternal perineal and vaginal cultures were provided in only 18 cases; the majority was GBS negative (12/18, 67%), 3 cases were of unknown GBS status (3/18, 16.5%), 2 cases were GBS positive (2/18, 11%) and 1 case (1/18, 5.5%) was positive for gamma Streptococci that did not belong to Group D. Peripartum antibiotics were used in only 2 cases; one was for premature spontaneous labor at 32 weeks and one for PPROM. However, in most cases (54 out of 67), relevant information was not reported. The time of onset of symptoms was specified in 43 cases. There were 17 cases (39.5%) with early-onset sepsis (data reported in Table 1) and the remaining (60.5%) referred to late-onset sepsis (data reported in Table 2). Among the 67 cases reported, respiratory distress, poor feeding, fever and lethargy were the symptoms most frequently described, followed by seizures, apnea, and symptoms from the gastrointestinal tract. Meningitis was by far the most common complication, being reported in 40 out of 53 cases (75%) (in 14 cases, complications were not listed). Bacteremia complicated with meningitis was reported in 47% (8/17 cases) of early-onset sepsis and in 61.5% (16/26 cases) of late-onset sepsis. Complications from the nervous system also included five cases of ventriculitis (7%), two cases of grade III intraventricular hemorrhage (IVH) (3%), three cases of subdural effusion (4.5%) and one case of hydrocephalus requiring ventriculoperitoneal bypass (1.5%). Regarding the respiratory tract, two cases with persistent pulmonary hypertension were described (3%) (including the present case), one case with pulmonary hemorrhage (1.5%) and one case with pneumonia (1.5%). In almost half of the cases (32 out of 67 cases), S. gallolyticus was identified to the subspecies level; in 28 cases (87.5%), S. gallolyticus subspecies pasteurianus was isolated, and in 4 cases (12.5%) S. gallolyticus subspecies gallolyticus. Penicillin, or a penicillin-derivative, alone or in combination with another agent (usually cephalosporin or aminoglycoside) was the definitive treatment of choice in 41 out of 56 cases (73%). Among 67 cases, only three neonates (4.5%) died. Abbreviations: CFA, cellular fatty acid; CSF, cerebrospinal fluid; GBS, group B streptococci; IV, intravenous; IVH, intraventricular hemorrhage; MALDI-TOF, matrix-assisted laser desorption ionization-time of flight mass spectrometry; N/A, not available; PFGE, pulsed-field gel electrophoresis; PCR, polymerase chain reaction; PPHN, persistent pulmonary hypertension of the newborn; RFLP, restriction fragment length polymorphism. * This study included 23 infants, 15 of whom were neonates. ** Subspecies of S. bovis were identified in 10 cases. Not specified if the data of S. gallolyticus identification refers to neonates or to infants. *** This study included 15 infants, 9 of whom were neonates. **** S. bovis was not classified into subspecies in one case. Not specified if the data of S. gallolyticus identification refers to neonates or to infants. ***** Not specified if cases of meningitis refer to neonates or to infants.      Abbreviations: CSF, cerebrospinal fluid; GBS, group B streptococci; IV, intravenous; IVH, intraventricular hemorrhage; MALDI-TOF, matrix-assisted laser desorption ionization-time of flight mass spectrometry; N/A, not available; PFGE, pulsed-field gel electrophoresis; PCR, polymerase chain reaction; RFLP, restriction fragment length polymorphism. * This study included 23 infants, 15 of whom were neonates. ** Subspecies of S. bovis were identified in 10 cases. Not specified if the data of S. gallolyticus identification refers to neonates or to infants. *** This study included 15 infants, 9 of whom were neonates. **** S. bovis was not classified into subspecies in one case. Not specified if the data of S. gallolyticus identification refers to neonates or to infants. ***** Not specified if cases of meningitis refer to neonates or to infants.

Discussion
This is a case of early-onset neonatal sepsis due to S. gallolyticus, with rapid clinical deterioration leading to mechanical ventilation and administration of inotropes, complicated by the development of pulmonary hypertension, with finally benign clinical course and discharge after a 12-day hospitalization.
S. gallolyticus belongs to group D Streptococci in the S. bovis-S. equinus complex. The taxonomy and nomenclature of group D Streptococci were modified in 2003 [28]. S. bovis, based on its ability or inability to ferment mannitol, was designated biotype I (mannitol fermentation-positive) or biotype II (mannitol-fermentation-negative). Biotype II was further divided into biotype II/1 and biotype II/2 depending on further biochemical characteristics. These strains are now known as S. gallolyticus subsp. gallolyticus (biotype I), S. infantarius (biotype II/1) and S. gallolyticus subsp. pasteurianus (biotype II/2). Proper identification is usually accomplished with a combination of various biochemical tests (e.g., API 20 Strep and Rapid ID 32 Strep test system, Vitek 2, matrix-assisted laser desorption ionization-time of flight mass spectrometry-MALDI TOF) with molecular tests (e.g., 16S rRNA and sodA gene sequencing, pulsed-field gel electrophoresis-PFGE) [1]. Based on our literature search, biochemical testing is initially applied to identify bacteria at the species level. However, enzymatic reaction interpretation is not always definite, and this could lead to the misidentification of the isolated strain. Therefore, to minimize the risk of misidentification, various phenotypic tests should be used to verify the results. Moreover, whenever available, genotypic identification of the isolated pathogen with the utilization of molecular tests should be preferred in order to confirm the identity of the isolated strains and to further identify them to the subspecies level. Gene sequencing and PFGE are the methods more widely used to identify the isolated microorganism [29,30] reliably.
In adults, the most common source of infection includes the gastrointestinal and hepatobiliary tracts. S. gallolyticus subsp. gallolyticus is mainly associated with colonic neoplasia and endocarditis in adults, while S. infantarius is linked to non-colonic cancers [31]. Due to the epidemiological correlation between S. gallolyticus infection and colon cancer, it is strongly recommended that all adults with S. gallolyticus bacteremia are evaluated with colonoscopy [27]. No gastrointestinal disease was known in the mother of the present case; a colonoscopy was scheduled after discharge from the maternity hospital. Recent studies also suggest that contact with infected animals or contaminated food promotes fecal colonization of healthy adults with S. gallolyticus, indicating a potential zoonotic origin [32].
The pathophysiology of neonatal infection caused by S. gallolyticus species remains unclear. Based on our review of the literature, S. gallolyticus subsp. pasteurianus is the most common pathogen among S. gallolyticus species in neonates. Similarly to GBS, it is believed that colonization of the vagina and vertical transmission to the fetus via an ascending route, especially in case of prolonged rupture of membranes, or transmission to the fetus during parturition, could lead to neonatal colonization and probably infection [33]. In the present case, the exact contagious route was not clarified since no pathogens were isolated in the vaginal swab at 35 weeks of gestation, and there was no prolonged rupture of membranes. Fikar and Levy reported a case of neonatal meningitis caused by S. bovis, whose mother had positive vaginal and rectal swabs for the same microorganism [34]. Additionally, Binghuai et al. described a case of intrauterine infection complicated by postpartum bacteremia caused by S. gallolyticus subsp. pasteurianus in a mother whose infant tested positive for the same infectious agent in ear and throat swabs [35]. This neonate remained asymptomatic, contrary to the cases reported by Sasi et al. [27]. In their study, four out of five neonates born to mothers with intrauterine infection and postpartum bacteremia caused by S. gallolyticus subsp. gallolyticus had symptoms compatible with sepsis and/ or meningitis. However, no microorganism was isolated from neonatal blood and/ or CSF cultures. Abnormalities of the immune system may further predispose neonates to infections, including those caused by S. gallolyticus, as described by Koh et al. and Mettananda et al. [4,19]. Moreover, as stated by Saegeman et al. [2] and Floret et al. [8], nosocomial transmission may play a role in some cases of neonatal infections by S. gallolyticus species; this underlines the importance of adherence of medical staff with health protocols and raises awareness of healthcare-associated infections.
A benign clinical course is usually reported for S. gallolyticus infections in neonates. Based on our search of the literature, the clinical presentation of S. gallolyticus neonatal infection cannot be easily distinguished from GBS infection, with signs and symptoms including sepsis and metabolic acidosis, respiratory distress, apnea and poor overall activity being prominent. Meningitis, which may present with seizures, constitutes a common complication, especially of late-onset sepsis, though it seems to have a favorable outcome. Despite the increased frequency of concomitant bacteremia and meningitis, the neonate of the present case did not have any signs or laboratory evidence of meningitis during its clinical course. Additional complications of the central nervous system described in the literature include ventriculitis, [18,24,26] hydrocephalus [17] and intraventricular hemorrhage (IVH) (two cases with grade III IVH, though one refers to an extremely preterm neonate and hence IVH could be either a complication of the infection or due to underlying prematurity) [13,26]. Complications from the respiratory tract seem to be rare: two cases with pulmonary hypertension have been described (including the present case), [20] one case of pulmonary hemorrhage (referring to a preterm neonate) and one case of pneumonia [27]. Even though only a couple of cases of persistent pulmonary hypertension (PPHN) secondary to S. gallolyticus infection have been described thus far, PPHN has been clearly described in the literature as a complication of neonatal sepsis [36]. Release in the bloodstream of bacterial endotoxins, together with inflammatory mediators (including endothelin-1 and thromboxane, which constitute potent vasoconstrictors) and various cytokines, seem to contribute to the pathophysiology of PPHN associated with neonatal sepsis. [37,38] Additionally, although S. gallolyticus has been clearly linked with cases of endocarditis in adult patients, to our knowledge, only one case has been reported in the neonatal population; this patient recovered without sequelae [20]. The overall prognosis is encouraging, with three neonatal deaths by S. gallolyticus being recorded in the literature out of 67 cases of neonatal infection (including one extremely premature and extremely low birth weight neonate) and 2 cases complicated by severe neurologic sequelae [2,17,25,26]. Postmortem examination findings of the deceased neonates were not described in the literature. Procalcitonin constitutes a promising diagnostic marker of neonatal sepsis, being able to discriminate cases of infection versus non-infectious inflammatory disorders; it could be thus speculated that immunohistochemical techniques, with anti-procalcitonin antibodies, among others, could contribute to the postmortem diagnosis of neonatal sepsis, in cases of diagnostic ambiguity [39,40].
S. gallolyticus, in the majority of cases, is sensitive to penicillin and to penicillin derivatives, ref. [5,22] and this was also the case for the neonatal infection caused by S. gallolyticus hereby presented. However, we should mention two reports by Khan et al. [7] and Klatte et al. [9] referring to two cases of neonatal meningitis caused by S. pasteurianus relatively resistant to penicillin. Therefore, de-escalation of antibacterial therapy after accurate identification of the causative agent and antimicrobial susceptibility testing is legitimate to minimize the risk of multi-drug resistant pathogens. According to our search of the literature, the duration of therapy is variable depending on the clinical course, ranging from a minimum of 5 days [27] to a maximum of approximately 50 days (one case with liver abscess formation [23] and one case with late central nervous system complications) [14].
There are some limitations in our systematic review. In many cases, documentation regarding perinatal history was incomplete, including type of delivery, duration of rupture of membranes, maternal perineal and vaginal swabs and use of prophylactic peripartum antibiotics; consequently, the impact of these perinatal parameters on the pathogenesis of neonatal infection by S. gallolyticus cannot be precisely estimated. Additionally, there may be cases of infection by S. gallolyticus that have not been captured. Specifically, there are cases of neonatal infection by S. bovis in the literature that have not been identified at the subspecies level or cases that may have been misclassified as other Streptococcus species, mainly before the currently available biochemical and molecular techniques. Therefore, the true prevalence of S. gallolyticus infection and its subspecies remains unknown.

Conclusions
In conclusion, the occurrence of S. gallolyticus infection in the neonate of the present case report indicates that S. gallolyticus may be considered a novel and emerging pathogen in the field of neonatal sepsis. Increased clinical suspicion may be warranted in sepsis cases resembling GBS infections, even in the presence of negative GBS vaginal swab cultures. Precise identification of the pathogens is necessary to detect possible epidemiological changes. More studies are needed to clarify the epidemiology, pathogenesis and possible sequelae of neonatal infections caused by S. gallolyticus species. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.