Viral metagenomic sequencing in a cohort of international travellers returning with febrile illness.

BACKGROUND: Diagnosis of infections in returning international travellers can be challenging because of the broad spectrum of potential infectious etiologies potentially involved. Viral metagenomic next-generation sequencing (mNGS) has the potential to detect any virus present in a patient sample and is increasingly being used for difficult to diagnose cases. The aim of this study was to analyze the performance of mNGS for viral pathogen detection in the clinical setting of international travellers returning with febrile illness. METHODS: Thirty-eight serum samples from international travellers returning with febrile illness and presenting at the outpatient clinic of the Leiden University Medical Center in the Netherlands in the time period 2015-2016 were selected retrospectively. Samples were processed for viral metagenomic sequencing using a probe panel capturing all known vertebrate viruses. Bioinformatic analysis was performed using Genome Detective software for metagenomic virus detection. Metagenomic virus findings were compared with viral pathogen detection using conventional methods. RESULTS: In 8 out of the 38 patients (21%), a pathogenic virus was detected by mNGS. All viral pathogens detected by conventional assays were also detected by mNGS: dengue virus (n=4 patients), Epstein-Barr virus (n=2), hepatitis B virus (n=1). In addition, mNGS resulted in additional pathogenic findings in 2 patients (5%): dengue virus (n=1), and hepatitis C virus (n=1). Non-pathogenic viruses detected were: GB virus C (n=1) and torque teno viruses (n=3). High genome coverage and depth using capture probes enabled typing of the dengue viruses detected. CONCLUSIONS: Viral metagenomics has the potential to assist the detection of viral pathogens and co-infections in one step in international travellers with a febrile syndrome. Furthermore, viral enrichment by probes resulted in high genome coverage and depth which enabled dengue virus typing.

Feasible alternatives to DBS in the retrospective diagnosis of congenital cytomegalovirus infection.

BACKGROUND: Retrospective diagnosis of congenital cytomegalovirus (cCMV) infection may be challenging mainly because of the high variable sensitivity of PCR on dried blood spots (DBS) samples. OBJECTIVES: To compare cytomegalovirus (CMV) viral load (VL) values in different samples obtained at birth from infants with cCMV infection. To evaluate dried umbilical cord (DUC) samples as an alternative to DBS. STUDY DESIGN: Saliva and/or urine, peripheral blood (PB), and DBS from 16 infants with confirmed cCMV infection were collected at birth. CMV VL were determined by DNA extraction and real-time polymerase chain reaction (rt-PCR). In two cases, VL was determined from DUC samples. RESULTS: Six (37.5 %) of the 16 infants were symptomatic, and 10 (62.5 %) were asymptomatic. The CMV VL found in saliva (median: 1,958,525 [IQR: 597,683-3,483,843] IU/mL) and in urine (median: 691,865 [IQR: 188,489.5-3,175,696] UI/mL) were both higher than those found in PB (median: 1115 [IQR: 364-4,002] IU/mL), p: 0.0001). Symptomatic infants presented 100 % of detectable VL in PB and 50 % in DBS. Asymptomatic infants showed 75 % of detectable VL in PB and 30 % in DBS. The VL in DUC were 22,341, 9754 IU/mL and 994 IU/mL. CONCLUSIONS: When VL was detectable in PB, the values were lower than in saliva or urine, in both symptomatic and asymptomatic cases of cCMV. The low sensitivity in DBS samples could be due to low blood volume content, making CMV VL undetectable even when using optimised extraction and PCR protocols. In our limited experience, DUC could play a complementary diagnostic role when DBS VL is undetectable.

SARS-CoV-2 infection in cancer patients undergoing active treatment: analysis of clinical features and predictive factors for severe respiratory failure and death.

AIM: Previous studies have suggested a more frequent and severe course of novel coronavirus SARS-CoV-2 infection in cancer patients undergoing active oncologic treatment. Our aim was to describe the characteristics of the disease in this population and to determine predictive factors for poor outcome in terms of severe respiratory distress (acute respiratory distress syndrome [ARDS]) or death. PATIENTS AND METHODS: Patients consecutively admitted for SARS-CoV-2 infection were prospectively collected, and retrospective statistical analysis was performed. Univariate and multivariate analyses were performed to assess potential factors for poor outcomes defined as ARDS or death. RESULTS: Sixty-three patients were analysed, and 34 of them developed respiratory failure (70% as ARDS). Lymphocytes/mm3 (412 versus 686; p = 0.001), serum albumin (2.84 versus 3.1); lactate dehydrogenase (LDH) (670 versus 359; p < 0.001) and C-reactive protein (CRP) levels (25.8 versus 9.9; p < 0.001) discriminate those that developed respiratory failure. Mortality rate was 25%, significantly higher among ARDS, neutropenic patients (p = 0.01) and in those with bilateral infiltrates (44% versus 0%; p < 0.001). Multivariate logistic analyses model confirmed the predictive value of severe neutropenia (odds ratio [OR] 16.54; 95% confidence interval [CI] 1.43-190.9, p 0.025), bilateral infiltrates (OR 32.83, CI 95% 3.51-307, p 0.002) and tumour lung involvement (OR 4.34, CI 95% 1.2-14.95, p 0.02). CONCLUSION: Cancer patients under active treatment admitted for SARS-CoV-2 infection have worse outcomes in terms of mortality and respiratory failure rates compared with COVID-19 global population. Lymphopenia, LDH, CRP and albumin discriminate illness severity, whereas neutropenia, bilateral infiltrates and tumour pulmonary involvement are predictive of higher mortality.