Severe Acute Respiratory Syndrome Coronavirus 2 in Infants Younger Than 90 Days Presenting to the Pediatric Emergency Department: Clinical Characteristics and Risk of Serious Bacterial Infection.

OBJECTIVES: There are scant data on severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection in infants younger than 90 days. This study was designed to characterize COVID-19 presentation and clinical course in this age group and evaluate the risk of serious bacterial infection. METHODS: Data on all SARS-CoV-2-polymerase chain reaction-positive infants presenting to the pediatric emergency department (PED) were retrospectively collected, followed by a case-control study comparing those infants presenting with fever (COVID group) to febrile infants presenting to the PED and found to be SARS-CoV-2 negative (control group). RESULTS: Of the 96 PCR-positive SARS-CoV-2 infants who met the inclusion criteria, the most common presenting symptom was fever (74/96, 77.1%) followed by upper respiratory tract infection symptoms (42/96, 43.8%). Four (4.2%) presented with symptoms consistent with brief resolved unexplained event (4.2%).Among the febrile infants, the presenting symptoms and vital signs were similar in the COVID and control groups, with the exception of irritability, which was more common in the control group (8% and 26%; P < 0.01). The SARS-CoV-2-positive infants had decreased inflammatory markers including: C-reactive protein (0.6 ± 1 mg/dL vs 2.1 ± 2.7 mg/dL; P < 0.0001), white blood cell count (9.3 ± 3.4 × 10 9 /L vs 11.8 ± 5.1 × 10 9 /L; P < 0.001), and absolute neutrophils count (3.4 ± 2.4 × 10 9 /L vs 5.1 ± 3.7 × 10 9 /L; P < 0.001). The rate of invasive bacterial infection was similar between groups (1.4% and 0%; P = 0.31). No mortality was recorded. Although not significantly different, urinary tract infections were less common in the COVID group (7% and 16%; P = 0.07). CONCLUSIONS: The SARS-CoV-2 infection in infants aged 0 to 90 days who present to the PED seems to be mostly mild and self-limiting, with no increased risk of serious bacterial infection.

Characterization of Factors Affecting the Detection Limit of EGFR p.T790M in Circulating Tumor DNA.

OBJECTIVE: Circulating tumor DNA is a promising noninvasive tool for cancer monitoring. One of the challenges in applying this tool is the detection of low-frequency mutations. The detection limit of these mutations varies between different molecular methods. The aim of this study is to characterize the factors affecting the limit of detection for epidermal growth factor receptor p.T790M mutation in circulating tumor DNA of patients with lung adenocarcinoma. METHODS: DNA was extracted from plasma samples of 102 patients. For sequencing the DNA, we used 2 different next-generation sequencing-based platforms: Ion Torrent Personal Genome Machine (56 cases) and Roche/454 (46 cases). Serially diluted synthetic DNA samples carrying the p.T790M mutation were sequenced using the Ion Torrent Personal Genome Machine for validation. Limit of detection was determined through the analysis of non-hot-spot nonreference reads, which were regarded as sequencing artifacts. RESULTS: The frequency of the non-hot-spot nonreference reads was higher in Ion Torrent Personal Genome Machine compared to Roche/454 (0.07% ± 0.08% and 0.03% ± 0.06%, respectively, P < .001). We found that different base type substitutions occur with different frequency. Since the base substitution leading to p.T790M mutation is C>T transition, its frequency was used to determine the limit of detection for the assay. Based on the C>T non-hot-spot nonreference allele frequency, we found that the limit of detection is 0.18% in Ion Torrent Personal Genome Machine and 0.1% in Roche/454. Based on these values, 48% and 56% of the cases were positive for T790M mutation in Ion Torrent Personal Genome Machine and Roche/454 groups, respectively. Agreement between duplicates was 76% in Ion Torrent Personal Genome Machine and 72% in Roche/454. Using serially diluted synthetic DNA samples carrying the p.T790M mutation, we could identify mutations with allele frequency of 0.18% or more using the Ion Torrent Personal Genome Machine, supporting our approach to determine the detection limit. CONCLUSION: Both the sequencing platform and the specific nucleotide change affect the limit of detection and should therefore be determined in the validation process of new assays.