Gaps: paediatric cardiology and cardiac surgery

Though serum biomarkers are used daily in inpatients with CHD, the landscape of biological insight from currently available biomarkers is notably incomplete. With advancing omics platforms and decreasing expense of advanced biological tests, many additional biomarkers are likely yet to be discovered. For example, necrotising enterocolitis continues to be a disease in both preterm neonate and CHD populations; despite its association with significant morbidity and mortality, few diagnostic biomarkers are available clinically. As such, physicians continue to struggle predicting which infants will develop necrotising enterocolitis, highlighting the need for unbiased target discovery. ^{Reference Sugimoto, Kuwata and Kurishima2} Outside of necrotising enterocolitis, many infants develop feeding issues, despite being relatively “normal” feeders before developing heart failure symptoms or undergoing cardiac surgery using cardiopulmonary bypass. Despite occasional use of prealbumin as a marker of nutritional status as well as other vitamin and micronutrient levels in combination with weight gain, we have limited ability to measure or predict the nutritional status of infants and children to guide supplementation and encourage growth.

All these known and yet unknown biomarkers may inform diagnosis, management, and prognosis in both the inpatient and outpatient settings. The frequency of testing is determined based upon the relative degree of patient illness in relation to the utility of the biomarker and inconvenience of the sample acquisition. Blood sampling may require as little as a few drops of blood to more than 6 ml of blood per test. Nevertheless, serum biomarker tests require blood to be sampled, which can be challenging to accomplish in paediatric populations. Many paediatric and CHD patients have limited venous access from repeat venipunctures and frequent labs. Many develop a significant fear of needles, making blood draws even more challenging. Additional challenges are experienced in the outpatient setting, where families can be located hours away from a specialised centre, or a local laboratory drawing site may be staffed by personnel with limited experience with children, especially children with heart disease. The average newborn has about 80–100 ml/kg of circulating blood volume, meaning that at most, a 3 kg newborn would have about 300 ml of blood. With this perspective, each 0.5–2 ml blood draw can quickly add up. A multicentre study of 30 paediatric ICUs showed that blood draws accounted for 73% of daily blood loss, with a median volume loss per blood draw of 2 ml/draw, a mean of 3 blood draws per day, and a median daily loss of 5 ml/day. ^{Reference Bateman, Lacroix and Boven3} Together, these clinical limitations, coupled with the recent acceleration of new low-input technologies, yield fertile ground for the discovery and implementation of new biomarkers using non-traditional biofluids and limited sample quantities.

Gaps: translational research

Translational research for the detection of paediatric diseases and for biomarker discovery is limited by several factors related to sample acquisition and stability. In an ideal paediatric case, biomarker detection can occur with patient biofluids that are easily and frequently obtained, molecularly stable or able to be stabilised in an outpatient environment, and accessible for longitudinal studies. For sample collection, there are two primary considerations: 1) type of biofluid to be collected and 2) type of analyte to be measured. From least to most invasive, potential biofluids for cardiovascular biomarker detection include saliva, sweat, urine, faeces, interstitial fluid, and blood, each with distinct advantages and disadvantages ^{Reference Rao, Mandal and Patolsky4} in paediatric cardiovascular disease detection (see Graphical Abstract).

Saliva is a non-invasive biofluid that has a moderate correlation to systemic biomarker levels; however, it is susceptible to contamination and degradation, requiring careful collection and analysis techniques. While wearable sweat-based sensors are an emerging field of research, their application capacity is mostly limited to minerals, metabolites, and small molecules with intra- and inter-patient variability and low systemic correlation. Urine biomarker levels have a strong correlation with systemic levels, and urine has the advantages of being easy to collect, non-invasive, and containing certain biomarkers at higher levels than plasma. Urine, however, reflects a cumulative physiology due to delayed clearance and can be challenging to collect without faecal contamination in infants and incontinent children. Faecal matter has potential for the diagnosis of primarily GI-related biomarkers but can also be useful in inflammatory diseases. Stool has been shown to have high rates of false positives and is affected by exogenous factors such as diet. Interstitial fluid can be accessed via continuous microneedle collection-based monitoring systems, the most common usage being continuous glucose monitoring. However, limitations include high consumable costs, skin adhesive reactions, and biofouling of the microneedle site, which induces a localised immune response and reduces the accuracy of target detection. Blood provides a reliable means of monitoring; however, its collection is invasive. Blood has a broad dynamic range of analyte levels, limiting differentiation between subtle disease stages.

The primary biomarker types—DNA, protein, and small molecules, such as metabolites and lipids—vary by source, stability, and the fidelity of their reflection of disease status. While metabolites are the most stable, their intensity in circulation can come from both exogenous and endogenous sources. Protein biomarkers offer a functional insight into active endogenous disease processes but are initially unstable without biochemical manipulation. DNA offers insight into the host’s heritable contribution to disease stage but is a static metric, can degrade easily, and is expensive to analyse. ^{Reference Aronson and Ferner5} The remainder of this perspective will therefore focus on protein-based biomarkers for the paediatric population.

Assays relevant for protein-level detection of biomarkers can be split into two major categories—1) immunoassays, or those using antibodies to detect protein targets such as ELISA and proximity labelling, and 2) direct sequencing methods, such as mass spectrometry, each having distinct advantages and limitations. Immunoassays use simplified instrumentation and sample preparation for rapid clinical interpretation. However, accurate measurement of isoforms or epitomes, particularly those with post-translational modifications, is a significant challenge for immunoassays. Similarly, off-target binding potentially leads to limited assay accuracy and reliability, with low correlation to more acute quantification techniques such as mass spectrometry proteomics. Direct sequencing methods such as mass spectrometry-based protein assays can screen a higher number of proteins more accurately, yet sample preparation and data analysis are more complicated, with expensive instrumentation required. These two methods can be used symbiotically: mass spectrometry for initial biomarker identification and validation, and immunoassays for clinically translatable detection.

Protein quantification, whether performed by immunoassays or direct sequencing methods, is challenging in the context of salivary and faecal samples due to the presence of proteins from multiple species—that is, the host (human), animal protein (food sources), and oral and gut microbiome (bacterial proteins). This cross-species problem may lead to off-target effects of antibody binding in immunoassays. In mass spectrometry-based assays, the protein sequence similarity leads to challenges in accurate quantification. In the most common proteomic method, proteins are first digested into peptides prior to direct sequencing by mass spectrometry. In samples of mixed species, such as saliva or faeces, cross-species peptides often only differ by a single amino acid, complicating downstream protein identification and quantification without sufficient sequence coverage. Careful consideration of collection, stabilisation and storage, preparation, acquisition design, and analytical techniques must be applied for each biofluid and biomarker subclass.

Proposed solutions: clinical

Utilising the state-of-the-art advancements in biomarker development and discovery may improve the practice of paediatric cardiology and cardiac surgery. Saliva, stool, and urine are constantly produced and discarded. Yet, these are biological “gold mines” of diverse biomarkers that can be obtained noninvasively and at the point of care, whether at the hospital bedside or in the comfort of each patient’s home. Because of its abundance and availability, this huge supply of noninvasively obtained biofluid will facilitate longitudinal monitoring. For biomarker research and discovery, non-invasive biofluids may be easier to request from families than blood draws. Once the biomarkers are identified, devices designed specifically for paediatrics can be developed to acquire samples, potentially allowing the specimen to be collected and the test to be run at point-of-care. These devices may come in infant-friendly form factors, such as pacifiers, garments to collect sweat, or diapers to collect urine or stool.

With a better definition of clinically relevant biomarkers and detailed evaluation of post-translational modifications, there will be more opportunities to monitor specific biomarkers of health and disease. Currently, clinicians integrate information from multiple non-specific biomarkers combined with additional clinical features to interpret a patient’s clinical condition. For example, in neonatal sepsis, multiple inflammatory biomarkers such as CD11b, CD64, IL-6, IL-8, procalcitonin, and C-reactive protein may be combined to determine risk for sepsis, yet none have been shown to be sensitive or specific enough in isolation or together to avoid starting antibiotics when sepsis is considered. ^{Reference Eichberger, Resch and Resch6} Initial work to identify biomarkers that are not dependent on inflammation and are more specific to the presence of bacteria in the blood now allows physicians to avoid starting antibiotics when sepsis is absent. A similar concept can be applied to necrotising enterocolitis. During necrotising enterocolitis, the intestinal epithelium becomes ischaemic and can break down. Ideally, there would be a test to predict the high risk of necrotising enterocolitis before it occurs. For instance, a daily saliva or stool test could demonstrate early signs of intestinal stress or ischaemia, limiting the plan for feed advancement but avoiding the typical 3–7-day treatment of NPO and antibiotics that is currently followed when clinical concern for necrotising enterocolitis is present.

Shifting from blood to other biofluids for testing may allow transition of other commonly monitored biomarkers, such as drug levels for medications commonly used in both paediatric and adult settings, such as digoxin, lidocaine, and tacrolimus, or anticoagulation monitoring (i.e., INR for Coumadin management). This may enable more frequent and easier monitoring for high-risk medications. Many therapeutic drug levels are performed in reference labs; with point-of-care testing, results could be achieved in a more clinically relevant timeframe. Less phlebotomy for chronically ill children may also translate into less needle fear, fewer painful procedures, fewer venous thromboembolisms from indwelling lines, and more available veins when phlebotomy or IV access is needed.

Proposed solutions: translational research

The search for biomarkers leveraging unbiased mass spectrometry approaches presents a critical opportunity for paediatric cardiovascular biomarker research. State-of-the-art mass spectrometry technology has advanced to high-throughput, high-sensitivity characterisation of hundreds to thousands of analytes for low-input samples, such as single cells. This shows promise for reduced sample collection capacity in paediatric patients. Current single-cell mass spectrometry-based analysis has improved multi-factorially because of low-loss sample preparation techniques, improvements in high-throughput chromatographic separation, and mass spectrometry instrumentation that is capable of high-dimensional evaluation of analytes such as ion mobility. ^{Reference Bennett, Stephenson, Rose and Darmanis7} Ion mobility works to evaluate analytes based on structural combination in addition to the atom composition, adding another degree of separation distinct from traditional mass-to-charge detection. The now common use of ion mobility increases the ability to detect low-abundant biomarker species within small amounts of complex samples. Further, ion mobility mass spectrometry can distinguish between analyte species of the same mass and charge. This allows for more nuanced analysis of biomarkers—such as separating distinct protein isoforms or post-translationally modified forms of proteins. For proteomics, single-cell mass spectrometry is remarkably fast and high-throughput; with the implementation of improved acquisition strategies—such as data-independent acquisition, where all analytes are sequenced regardless of intensity, and careful sample preparation—several thousand single-cell experiments or other low-input studies may be completed in one day to a depth of over 800 proteins per cell at costs of around 15 cents per cell. ^{Reference Leduc, Khoury, Cantlon, Khan and Slavov8} In a paediatric setting, this suggests that physicians could be provided with usable data from small amounts of samples, such as saliva, in real time. Recent advances in imaging mass spectrometry methods utilise array-style sample preparations and data visualisation, compatible with instruments already utilised clinically. This approach facilitates high-throughput sample detection in a targeted fashion relevant to clinical diagnostic settings once libraries have been built via traditional chromatographic methods.

In addition to instrumentation improvements, analytical strategies have also improved. High sequencing coverage facilitated by the above instrumentation improvements has allowed for complex spectra, normally convoluted by chimeric spectra, to be accurately resolved. This leads to the potential unbiased identification of more biomarkers in increasingly complex samples. This increase in sequence coverage to protein biomarkers also allows for multi-species contamination, as seen in saliva and faecal matter, to be delineated, isolating protein signatures coming solely from the patient. ^{Reference Cahalane, Meuris and Matamalas9} The forefront of target discovery is in multi-omic applications that can capture the working metabolites, lipids, and proteins in the attomole range, integrating benefits of each biomarker subtype as described above.

In paediatric settings, where CHD, cardiomyopathies, and abnormal developmental trajectories remain major causes of morbidity and mortality, modern low-input, high-sensitivity mass spectrometry and corresponding analytical strategies provide an opportunity to capture real-time cell-type–specific molecular signatures that are unique to paediatric stages of development and potentially cannot be extrapolated from adult studies. This opportunity is largely untapped and has the potential to serve as the basis for the next generation of patient monitoring.

Summary

This manuscript highlights the limitations of current blood-based biomarkers in paediatric cardiology and cardiac surgery and argues for a paradigm shift toward non-invasive, low-input biofluids combined with modern analytical technologies. It outlines clinical and translational gaps driven by invasive sampling, limited venous access, and an incomplete understanding of disease biology, particularly in fragile infants with CHD. Saliva, urine, stool, sweat, and interstitial fluid are underused, yet information-rich, sources for longitudinal monitoring, enabled by advances in mass spectrometry, ion mobility, and high-throughput proteomics. These novel technologies can not only be utilised in the detection of novel biomarkers in the paediatric population but also may aid in optimising methods for non-invasive, longitudinal monitoring in these patients. By focusing on protein-based biomarkers and unbiased discovery approaches, this perspective frames a roadmap for next-generation paediatric monitoring that is less invasive, more precise, and better suited to the unique developmental and clinical needs of children with cardiac disease. Implications for future research may include biomarker discovery in a clinical context, both inpatient and outpatient, where diagnostic and prognostic tools are grossly lacking (Summary Figure).

Summary Figure.

Summary Figure. Rationale and framework for biomarker discovery in CHD using alternative biofluids. Children with CHD require longitudinal, multi-domain monitoring to guide diagnosis, risk stratification, and management; however, current reliance on blood-based biomarkers is limited by invasiveness, sampling burden, and an incomplete representation of disease biology. This summary figure highlights the urgent need for expanded biomarker discovery across diverse clinical domains in CHD, including heart failure progression, feeding and nutritional status, necrotising enterocolitis risk, renal injury, pulmonary disease, neurologic injury, and systemic dysfunction. Alternative biofluids, such as saliva, urine, stool, sweat, and interstitial fluid, offer scalable, non-invasive, and repeatable sampling opportunities, enabling more frequent and outpatient-compatible assessment. These biofluids span a continuum of invasiveness and systemic correlation and contain complementary classes of analytes (DNA, small molecules, and proteins), each providing distinct biological insight. Advances in low-input, high-throughput analytical platforms, particularly mass spectrometry-based proteomics, enable unbiased, multiplexed biomarker discovery and characterisation, including isoform- and post-translational-level resolution. Together, this framework underscores a critical opportunity to identify novel, clinically meaningful biomarkers that support integrated, longitudinal, and precision monitoring strategies tailored to the unique physiology of paediatric patients with CHD.