Systemic inflammatory response syndrome (SIRS) in children is defined by the presence of two out of four clinical criteria, including elevated or depressed leukocyte count, irregular body temperature, tachy- or bradycardia and elevated respiratory rate. Abnormal leucocyte count or temperature are obligatory features1. Infections, major surgeries, and hyperinflammatory syndromes are the most frequent triggers of SIRS [1, 2]. In patients with complex underlying diseases, specific infectious symptoms are often absent due to immunosuppression or overlap with features of the underlying disease. In these cases, SIRS is often treated with a variety of anti-infective medications, which may confound diagnosis and yield false negative results. Results also remain negative in cases of noninfectious inflammation, lack of method sensitivity, preanalytical errors, or non-cultivable/fastidious germs. Discrimination of different causes of SIRS is challenging especially in immunocompromised patients since opportunistic infections have to be taken into consideration and usually need to be addressed by targeted diagnostic procedures.

The clinical course of SIRS depends on timely initiation of adequate diagnostics and therapy. The infectious work-up routine includes culture-based methods, nucleic-acid-based technologies (NAT), polymerase chain reaction (PCR), and antigen assays, which all have limitations. Culture-based methods are time-consuming and prone to pre-analytic errors (e.g., contamination, delayed processing, or insufficient specimen) [3]. Targeted NAT and antigen assays are usually faster but are limited to the targeted pathogens and give no or incomplete information on antibiotic resistance. PCR is limited as it can only analyze predefined microbes.

Despite proper conventional infectious workup including application of the above-mentioned tests, the causes of SIRS in a complex pediatric case may not be identified. Recently, next generation sequencing (NGS) of free microbial desoxyribonucleic acid (DNA) in blood was evaluated in adults with sepsis [4]. This method increased the number of identified pathogens compared to standard diagnostic care. In contrast to conventional PCR-based results, NGS is not limited to the identification of a predefined list of suspected species [5].

In this study we sought to investigate the diagnostic value of NGS in children with SIRS.

In this single-center retrospective study (from November 2020 to August 2021) we evaluated NGS analysis of serum samples from seven pediatric patients with prolonged fever and SIRS. One patient was analyzed twice, in two different hospitalizations due to severe SIRS. SIRS was diagnosed based on the criteria defined at the International paediatric sepsis consensus conference [1]. Two out of four criteria applied and one was abnormal temperature or leucocyte count (core temperature $>$38.5 °C or 36.0 °C, tachycardia with heart rate $>$2 standard deviations (SD) above normal for age or otherwise unexplained elevation; mean respiratory rate $>$2 SD above normal for age or mechanical ventilation for an acute pulmonary process; leukocyte count elevated or depressed for age or $>$10% immature neutrophils [1]). The following parameters were collected: patient demographics, clinical findings, and laboratory test results including pathogen diagnostics. Patient demographics are presented as median and range, as well as direct descriptive values, as shown in Tables 1,2.

This retrospective analysis was approved by the local Ethics Commission of the University of Duisburg-Essen (21-10180-BO).

In our university children’s hospital, we performed eight NGS analyses in seven children with SIRS to identify potential pathogens. NGS was performed in four male and three female patients ages 8 to 16 years (median age 10 years). Four patients had underlying hematopoietic malignancies (one acute lymphoblastic leukemia (ALL), one acute myeloid leukemia (AML), two lymphoma); one patient had cystic fibrosis, one respiratory failure and fever in the context of granulomatosis with polyangiitis, and one suffered from Pediatric Inflammatory Multisystem Syndrome temporally associated with SARS-CoV-2 infection (PIMS-TS).

Median duration of fever was 7 days (0–18 days) before NGS was performed. In one patient fever persisted even longer over several weeks. We performed NGS in one patient without a fever due to rapid clinical deterioration despite receiving both broad-spectrum antibiotics and antifungal therapy. NGS results were available within 48 hours after taking blood in five NGS analyses (patient 1 (once), 2–4,6), within three Days for one (patient 5) and four days in another two analyses (patient 1 (once), 7). NGS yielded positive results in five out of eight analyses (63%). In six out of eight analyses (patient 1–5,7) results of NGS were verified by standard/established diagnostic care but were received faster (median 1.8 days (range: –2 – $>$7 days)) in six out of these eight analyses (patient 1–4,6,7). NGS results led to a change of therapy in five out of eight analyses (patient 2,4–7). In two patients (patient 2,4) with negative NGS results, antibiotic treatment was discontinued without recurrence of SIRS and in three patients’ antimicrobial therapy was adjusted due to NGS. In one patient (patient 5) HHV-3 could be detected and antiviral therapy with acyclovir was initiated. In another patient (case 6) two different strains of Pseudomonas spp. could be detected leading to additional treatment with tobramycin, after a variety of ineffective treatments. The patients’ clinical condition slowly improved, fever discontinued and the patient completely recovered. All NGS analyses are summarized in Tables 1,2.

In this case series we present the results of NGS diagnostics and its impact on treatment of seven children with prolonged fever of unknown origin or SIRS. In five of eight NGS analyses, NGS yielded positive results (two bacterial, three viral, and one additional fungal) and led to treatment modification. In four positive NGS analyses, NGS results were later confirmed by standard procedures except in one case in which the patient rapidly improved after treatment modification according to the NGS result. This lack of confirmation may be culture-based methods is not uncommon for patients with prolonged aplasia and might be a benefit of the NGS. However, at this point in time, with such a small number of patients, we can neither document nor prove this. Future larger studies will be necessary. Two patients received specific testing for the pathogen identified by NGS (HHV-3, Mycobacterium chimaera). In three NGS analyses, no pathogen was identified. One of these patients suffered from PIMS-TS, which is caused by a sterile auto-inflammation. Retrospectively two cases could be explained by rising cells, as it occurs after chemotherapy-induced severe bone marrow-depression, as known in transplantation. In two NGS analyses the treatment was discontinued, sparing the patients possible side effects. Twenty-eight days after performing NGS only one patient suffered an ongoing infection and had to be treated at PICU. Four patients were without therapeutic anti-infective treatment. In seven patients, NGS provided a faster result than standard diagnostic work-up. Should future large prospective studies suggest further benefit for patients, a seven-day availability of NGS testing within a 24-hour turn-around would be desirable.

The present case series shows that NGS expands diagnostic possibilities in children with SIRS and comorbidities. These patients are at higher risk for infections by opportunistic pathogens and thus may particularly benefit from early, reliable identification of the causative pathogen. NGS can identify a wide variety of pathogens (bacteria, fungi and viruses) within a single sample, including pathogens that are difficult to cultivate or are slow-growing such as atypical mycobacteria, which are rarely examined [6]. However, 7 patients are a very small number and larger multicenter cohorts are needed to make valid statements.

By providing an individually tailored therapy based on the detected pathogens, adverse drug events or toxicities can potentially be reduced. Yet, these results need to be interpreted with care, as it remains unclear if a negative result in NGS is reliable, especially in patients with a localized infection like pneumonia or abscess. As NGS detects DNA, an important limitations are infections due to RNA-based Pathogenes. Only one if fit clinically, the diagnostics must be expanded. As a consequence, controlled prospective studies are urgently needed to investigate such uncertainties, especially the safety of discontinuation of anti-infective therapy based on a false negative NGS result.

There are several limitations to the NGS method. First, NGS diagnostics is currently reserved to a few hospitals for evaluation purposes. Depending on availability, it can take several days to obtain results. Moreover, there is no option for phenotypical resistance testing. High costs have to be taken into consideration as well and be balanced against benefits in prospective studies. Currently, NGS might be regarded a promising addition to conventional diagnostics but may not replace them [7]. Our NGS analyses illustrate its potential merit, especially in pediatric patients with a diagnosed SIRS or severe critically ill pediatric patients. A multicenter study on the value of NGS in pediatric sepsis is about to be launched and should provide more reliable data [8].

SCG collected all samples, as well as the clinical data and was with MS and CDS responsible for writing and editing this manuscript. The primary idea for collecting the data used within the manuscript was contributed by CDS as well as TB, who also contributed to the final changes. SCG and BD was responsible for further literature research contributing to editing the manuscript. NB, ET, SS and FS were responsible for providing clinical data as part of the patient care and contributed to editing the manuscript. PMR and SV were responsible for supervision of the laboratory process at the University hospital Essen and sequencing data. They also contributed to the manuscript by critically reviewing and editing. SG was responsible for supervision of the laboratory process at the NG and sequencing data. She also contributed to the manuscript by writing, critically reviewing and editing. PH collected sample data and contributed to editing the manuscript.

The study was approved by the Ethics Committee of the Medical Faculty of the University Duisburg-Essen (21-10180-BO) and conducted in accordance to the latest version of the Declaration of Helsinki. As the study presents a retrospective analysis with anonymous data the ethics committee waived the need for informed consent.

We thank all patients and families who participated in this case report. We thank René Hennig (Noscendo GmbH, Duisburg, Germany) for his excellent technical support required for data processing and data analysis, as well as for Cathy Bolger (Stanford, USA) for her critical review and excellent support in editing this manuscript. All sources of support, including pharmaceutical and industry support that require acknowledgment: Noscendo GmbH Duisburg, Germany; Department of Pediatrics I and III, University Hospital Essen, Germany; Institute of Medical Microbiology, University Medicine Essen, University Duisburg-Essen, Germany; Institute of Virology, University Medicine Essen, University Duisburg-Essen, Germany; Department of Anesthesiology and Intensive Care Therapy, University Hospital Essen, Germany.

This study was supported in part by Noscendo GmbH, Duisburg, which arranged the transport of serum samples collected at University Hospital Essen and the NGS analyses performed in their laboratory.

The authors declare no conflict of interest. Dr. Grumaz is a co-founder, employee, and shareholder of Noscendo GmbH. Petra Horvatek is an employee of Noscendo GmbH. No specific research funding was used for this research.