Introduction

Proteomics is a methodological approach that allows for the analysis of many proteins simultaneously and has been successful in identifying many novel disease biomarkers [1]. Proteomic methodologies have been previously used in varying contexts, such as discovering biomarkers of diabetic nephropathy and identifying novel diagnostic markers of cancer [2, 3]. Plasma proteomics is advantageous as it only uses a small volume of blood to study hundreds and sometimes thousands of proteins, and can identify changes in protein expression that may occur with age and disease [4]. Proteomics is not limited to analysis of blood samples, and enables the use of biological fluids such as saliva and urine, and tissue samples (e.g. tumours) [5]. Due to the small volume required for analysis, plasma proteomics has become increasingly popular and has enabled investigations of plasma proteins in vulnerable populations such as in paediatrics, as well as in critically ill patients, where blood may be scarce and not readily available for research purposes [4].

Preterm birth is the leading cause of death among the paediatric population globally [6]. With major technological advances in neonatal care over the last few decades, there has been an increase in survival of infants born preterm (< 37 weeks’ gestation), in particular those born extremely preterm (< 28 weeks’ gestation) [7]. Despite the technological advances that have improved survival in these vulnerable populations, preterm birth is associated with significant morbidities including intraventricular haemorrhage (IVH), necrotising enterocolitis (NEC), bronchopulmonary dysplasia (BPD), and neurosensory impairments [8].

Within the last decade proteomics has enabled researchers to identify predictive biomarkers of NEC in preterm infants using buccal swabs [9]. More specifically, plasma proteomics has previously identified proteins that may play a role in the development of retinopathy of prematurity [10]. However, to date there has been limited research into plasma protein biomarkers in predicting other outcomes in preterm infants. Consequently, a scoping review was conducted to understand the current state of knowledge in this space, and to identify knowledge gaps that could be addressed by future studies. A preliminary search of MEDLINE, PubMed, JBI Evidence Synthesis and Embase was conducted and did not identify any current systematic reviews or scoping reviews on this topic. Thus, this review is novel and will make a significant contribution to the understanding and knowledge in the use of proteomics in preterm infants.

Review question

The following research question was formulated using the PCC (Population, Concept, Context) framework: What is the existing proteomic evidence of blood biomarker research in the setting of prematurity?

Methods

Study design

This scoping review was conducted based on the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) Checklist [11].

Search strategy

The following three electronic databases: Medline, Embase and PubMed were searched on the 24th September 2020 for all peer-reviewed studies. An additional search for grey literature was conducted using the OpenGrey and GreyLit databases. The specific search terms used for each database are detailed in Appendix A. In summary, studies included in this review were identified using the search terms [‘preterm’ OR ‘premature’) AND [‘proteome’ OR ‘protein-analysis’] AND [‘blood-protein’ OR [biomarker’], as well as including derivatives of these terms. Studies identified in this review were limited to those written in the English language and conducted in humans only. Studies retrieved using these search terms and parameters were screened by two authors (NL and TC), initially focusing on the eligibility of the studies’ titles and abstracts using the following inclusion and exclusion criteria.

Selection criteria

Inclusion criteria: (I) infants born preterm (< 37 weeks), (II) blood proteome assessed, (III) primary research, (IV) English language and (V) human study.

Exclusion criteria: (I) infants born at term or post-term (≥ 37 weeks), (II) proteome of other biological samples (e.g. saliva or urine) assessed, (III) case report, review, conference abstract or editorial correspondence and (IV) animal studies.

Data extraction and charting

Studies that were chosen for full-text assessment were assessed by NL and TC and with any discrepancies and uncertainties, a third reviewer (VI) was to assess the studies. Data extracted included publication year, disease/outcome assessed, aims, study population, comparative groups, proteomic methodology, protein-pathway analysis, key findings and study limitations. The detailed assessment for each critically reviewed study is presented in Table 1.

Table 1 Summary of included studies in the scoping review of proteomics in setting of prematurity
Full size table

Results

The initial search identified 678 studies using the scoping review search strategy, with an additional 22 studies identified using the grey literature search. After the removal of duplicates, 462 publications remained for title and abstract screening. A vast majority of studies (n = 444, 96%) were excluded due to not fulfilling the inclusion criteria or having no relevance to the topic of prematurity and blood biomarker discoveries. Eighteen studies underwent full-text review, with three studies excluded because they did not primarily investigate biomarkers of disease and outcomes. One study of children born preterm did not collect samples at birth and one study presented data in brief report, which did not include any proteomic data. Figure 1 illustrates the article screening and selection process, following the PRISMA guidelines (Fig. 2).

Fig. 1
figure 1

Summary of the study selection process for the scoping review

Full size image
Fig. 2
figure 2

Blood proteomic studies identified were primarily conducted in the setting of LOC/NEC (23%, 3 studies) and ROP (15%, 2 studies)

Full size image

Description of included studies

A total of thirteen studies met the inclusion criteria for this scoping review and are summarised in Table 1. Eleven of the thirteen included studies investigated proteins and their associations with known outcomes of prematurity. The participant gestational age at birth ranged from < 23 to 37 weeks, with sample sizes varying from 4 to 77 participants. Most studies used a tandem Mass Spectrometry method (MS/MS) to analyse the proteins of interest [10, 12,13,14,15]. Three of the fourteen studies also conducted protein validation and completed this task using protein microarray and immunoassay techniques [10, 16, 17]. Approximately half of the studies (n = 7, 47%) were completed using plasma samples (Fig. 3). The proteins identified as proteins of interest across the 13 studies included in this scoping review, with reference to the specific study/ies are summarised in Table 2.

Table 2 Proteins identified in the studies included in this scoping review
Full size table
Fig. 3
figure 3

Sample types used in the identified studies were primarily conducted using plasma (47%, 7 studies) and serum (40%, 6 studies)

Full size image

Retinopathy of prematurity (ROP)

Two studies investigated the outcome associated with prematurity, ROP [10, 18]. ROP is seen most commonly among infants born very preterm (< 32 weeks’ gestational age) or < 1250 g birth weight. Abnormal blood vessel development occurs in the retina in response to oxygen exposure, which can lead to retinal detachment and blindness in severe cases [18]. Currently there is no existing method to predict the occurrence of ROP in infants born preterm or born with a low birth weight and all high-risk infants are routinely screened. Hence, a proteomic approach was adopted to identify underlying biomarkers of the disease [10, 18]. Several biomarkers of the complement and inflammatory system were identified in infants who developed ROP [10]. Lynch et al. identified mitochondrial Superoxide dismutase (MnSOD), an antioxidant located in the mitochondria, as a potential therapeutic target for significant ROP [18].

Bronchopulmonary dysplasia (BPD) and pulmonary vascular disease (PVD)

Two of the thirteen included studies investigated plasma proteins and their association with BPD [15, 19]. BPD is a chronic lung disease that affects infants born preterm [20]. Arjaans et al. implemented the use of a SOMAscan proteomic assay, whereas Zasada et al. utilised MS/MS approach to identify key biomarkers of BPD. Both studies identified several proteins that may be used in future diagnosis of BPD as well associations between severity and disease prognosis [15, 19]. Wagner et al. investigated plasma proteins and their association with the pathogenesis of PVD, a term used to describe abnormal function and vascular growth of the lungs. They identified 18 proteins that were associated with PVD, including proteins associated with growth factors, angiogenesis and the extracellular matrix [21]. The protein analysis conducted by Wanger et al. also identified proteins of several different biological process pathways (e.g. Tissue Inhibitor of Metalloproteinases 3 (TIMP-3) used in platelet degradation and Bone proteoglycan II, involved in degradation of the extracellular matrix (ECM)) that may be associated with PVD.

Necrotising enterocolitis (NEC) and late onset sepsis (LOS)

Two of the thirteen studies examined biomarkers for NEC and LOS [12, 16]. Ng et al. investigated biomarkers for the early diagnosis of NEC among preterm infants. Ng et al. investigated their samples with a variety of proteomic methods, which included matrix-assisted laser desorption/ionisation (MALDI-ToF), 2D Gel-Electrophoresis (2DGE). The results of the discovery component of the study were validated using commercially available immunoassay kits and protein microarrays. Ng et al. identified a des-arginine variant of serum amyloid A (SAA) and Proapolipoprotein CII (Pro-apoC2) as very promising biomarkers of late-onset septicaemia and NEC [16]. Stewart et al. investigated the serum and metabolome of preterm infants with NEC and LOS longitudinally with a LC- MS/MS technique. Among all patient groups investigated the proteins and metabolite were comparable, with 12 proteins (e.g. Serum Amyloid A-2 and Haptoglobin) associated with NEC and LOS diagnosis [12]. Interestingly, the only protein common across the two studies was SAA [12, 16].

Gestational age and signalling pathways

Suski et al. completed several studies [13, 14] investigating plasma proteome changes in preterm infants comparing gestational ages [13] and malfunctioning proteins in various signalling pathways [14]. Utilising a tandem MS approach they were able to identify proteomic changes across varying gestational ages for several pathways which include; coagulation, inflammation, complement activations and immunomodulation [13, 14]. Suski et al. also observed Complement C3, Factor V and Complement C4-A were associated with gestational age [13]. LRG1 was the only common protein identified across the two studies [13, 14].

Discussion

In this scoping review we identified 13 primary studies that used proteomics to identify blood protein biomarkers in the setting of prematurity that used either plasma or serum as the sample which was analysed. It is important to note that studies conducted in serum cannot be directly compared to studies conducted in plasma as these are two entirely different samples. Unlike plasma which is prepared only via centrifugation, Preparation of serum entails formation and removal of a blood clot activating not only coagulation proteins but also changing the concentration of inflammatory proteins, a scenario that reflects the manipulation itself and not the physiological setting. Similarly, a cord-blood sample is different to the blood sample collected from babies at birth, due to differences in the vasculature of the umbilical cord and blood vessels within the newborn. Our findings indicate that the focus of research in the setting of blood protein biomarkers in the setting of prematurity focused on several diseases, such as ROP, BPD, LOS and NEC. However, there has been a lack of research focusing into other outcomes known to be associated with preterm birth such as cerebral palsy, intraventricular haemorrhage, or hypertension. To our best knowledge, none of the findings from the studies included in our scoping review have been translated into the clinical setting. Blood proteomic studies within this population may reflect a lack of collaboration between clinicians and proteomic experts, as well as difficulty in accessing samples from premature babies, factors that could be overcome, particularly in research institutes associated with tertiary hospitals [22].

Limitations of current published studies

The main limitation of the studies included in this review are the small sample sizes represented in those studies. Future studies should be adequately powered, and a shift of the primary focus from not only understanding mechanism of disease, but also on  identifying proteins that are associated with outcomes or disease and which can be used in the clinical setting to improve outcomes for premature infants.

Conclusions

This scoping review  identified a paucity of evidence around biomarker discoveries in the population of preterm infants. Several proteomic methods, including tandem mass spectrometry, immunoassays, and MALDI-TOF MS, have been used to identify biomarkers for various outcomes (e.g. ROP and BPD) associated with preterm birth. This review identifies the need for future research focusing on biomarkers to understand the possible mechanisms related to preterm birth, as well as to identify predictive protein biomarkers for complications or long-term sequelae associated with preterm birth, such as intraventricular haemorrhage and hypertension.