Evolution of Clinical Phenotypes of COVID-19 Patients During Intensive Care Treatment: An Unsupervised Machine Learning Analysis.

BACKGROUND: Identification of clinical phenotypes in critically ill COVID-19 patients could improve understanding of the disease heterogeneity and enable prognostic and predictive enrichment. However, previous attempts did not take into account temporal dynamics with high granularity. By including the dimension of time, we aim to gain further insights into the heterogeneity of COVID-19. METHODS: We used granular data from 3202 adult COVID patients in the Dutch Data Warehouse that were admitted to one of 25 Dutch ICUs between February 2020 and March 2021. Parameters including demographics, clinical observations, medications, laboratory values, vital signs, and data from life support devices were selected. Twenty-one datasets were created that each covered 24 h of ICU data for each day of ICU treatment. Clinical phenotypes in each dataset were identified by performing cluster analyses. Both evolution of the clinical phenotypes over time and patient allocation to these clusters over time were tracked. RESULTS: The final patient cohort consisted of 2438 COVID-19 patients with a ICU mortality outcome. Forty-one parameters were chosen for cluster analysis. On admission, both a mild and a severe clinical phenotype were found. After day 4, the severe phenotype split into an intermediate and a severe phenotype for 11 consecutive days. Heterogeneity between phenotypes appears to be driven by inflammation and dead space ventilation. During the 21-day period, only 8.2% and 4.6% of patients in the initial mild and severe clusters remained assigned to the same phenotype respectively. The clinical phenotype half-life was between 5 and 6 days for the mild and severe phenotypes, and about 3 days for the medium severe phenotype. CONCLUSIONS: Patients typically do not remain in the same cluster throughout intensive care treatment. This may have important implications for prognostic or predictive enrichment. Prominent dissimilarities between clinical phenotypes are predominantly driven by inflammation and dead space ventilation.

Topology and redescriptions detect multiple alternative biological pathways from clinical phenotypes.

Biological pathways play a crucial role in the properties of diseases and are important in drug discovery. Identifying the logical relationships among distinctive phenotypic clusters could reveal possible connections to the underlying pathways. However, this process is challenging since clinical phenotypes are often available through unstructured electronic health records. Moreover, in the absence of a standardized questionnaire, there could be bias among physicians toward selecting certain medical terms. In this article, we develop an efficient pipeline to address these challenges and help practitioners to reveal the pathways associated with the disease. We use topological data analysis and redescriptions and propose a pipeline of four phases: (1) pre-processing the clinical notes to extract the salient concepts, (2) constructing a feature space of the patients to characterize the extracted concepts, (3) leveraging the topological properties to distill the available knowledge and visualize the extracted features, and finally, (4) investigating the bias in the clinical notes of the selected features and identify possible pathways. Our experiments on a publicly available dataset of COVID-19 clinical notes testify that our pipeline can indeed extract meaningful pathways.

Guillain-Barré syndrome after SARS-CoV-2 infection in an international prospective cohort study.

In the wake of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic, an increasing number of patients with neurological disorders, including Guillain-Barré syndrome (GBS), have been reported following this infection. It remains unclear, however, if these cases are coincidental or not, as most publications were case reports or small regional retrospective cohort studies. The International GBS Outcome Study is an ongoing prospective observational cohort study enrolling patients with GBS within 2 weeks from onset of weakness. Data from patients included in this study, between 30 January 2020 and 30 May 2020, were used to investigate clinical and laboratory signs of a preceding or concurrent SARS-CoV-2 infection and to describe the associated clinical phenotype and disease course. Patients were classified according to the SARS-CoV-2 case definitions of the European Centre for Disease Prevention and Control and laboratory recommendations of the World Health Organization. Forty-nine patients with GBS were included, of whom eight (16%) had a confirmed and three (6%) a probable SARS-CoV-2 infection. Nine of these 11 patients had no serological evidence of other recent preceding infections associated with GBS, whereas two had serological evidence of a recent Campylobacter jejuni infection. Patients with a confirmed or probable SARS-CoV-2 infection frequently had a sensorimotor variant 8/11 (73%) and facial palsy 7/11 (64%). The eight patients who underwent electrophysiological examination all had a demyelinating subtype, which was more prevalent than the other patients included in the same time window [14/30 (47%), P = 0.012] as well as historical region and age-matched control subjects included in the International GBS Outcome Study before the pandemic [23/44 (52%), P = 0.016]. The median time from the onset of infection to neurological symptoms was 16 days (interquartile range 12-22). Patients with SARS-CoV-2 infection shared uniform neurological features, similar to those previously described in other post-viral GBS patients. The frequency (22%) of a preceding SARS-CoV-2 infection in our study population was higher than estimates of the contemporaneous background prevalence of SARS-CoV-2, which may be a result of recruitment bias during the pandemic, but could also indicate that GBS may rarely follow a recent SARS-CoV-2 infection. Consistent with previous studies, we found no increase in patient recruitment during the pandemic for our ongoing International GBS Outcome Study compared to previous years, making a strong relationship of GBS with SARS-CoV-2 unlikely. A case-control study is required to determine if there is a causative link or not.

Genomic monitoring of SARS-CoV-2 uncovers an Nsp1 deletion variant that modulates type I interferon response.

The SARS-CoV-2 virus, the causative agent of COVID-19, is undergoing constant mutation. Here, we utilized an integrative approach combining epidemiology, virus genome sequencing, clinical phenotyping, and experimental validation to locate mutations of clinical importance. We identified 35 recurrent variants, some of which are associated with clinical phenotypes related to severity. One variant, containing a deletion in the Nsp1-coding region (Δ500-532), was found in more than 20% of our sequenced samples and associates with higher RT-PCR cycle thresholds and lower serum IFN-ß levels of infected patients. Deletion variants in this locus were found in 37 countries worldwide, and viruses isolated from clinical samples or engineered by reverse genetics with related deletions in Nsp1 also induce lower IFN-ß responses in infected Calu-3 cells. Taken together, our virologic surveillance characterizes recurrent genetic diversity and identified mutations in Nsp1 of biological and clinical importance, which collectively may aid molecular diagnostics and drug design.

COVID-19 Clinical Phenotypes: Presentation and Temporal Progression of Disease in a Cohort of Hospitalized Adults in Georgia, United States.

BACKGROUND: The epidemiological features and outcomes of hospitalized adults with coronavirus disease 2019 (COVID-19) have been described; however, the temporal progression and medical complications of disease among hospitalized patients require further study. Detailed descriptions of the natural history of COVID-19 among hospitalized patients are paramount to optimize health care resource utilization, and the detection of different clinical phenotypes may allow tailored clinical management strategies. METHODS: This was a retrospective cohort study of 305 adult patients hospitalized with COVID-19 in 8 academic and community hospitals. Patient characteristics included demographics, comorbidities, medication use, medical complications, intensive care utilization, and longitudinal vital sign and laboratory test values. We examined laboratory and vital sign trends by mortality status and length of stay. To identify clinical phenotypes, we calculated Gower's dissimilarity matrix between each patient's clinical characteristics and clustered similar patients using the partitioning around medoids algorithm. RESULTS: One phenotype of 6 identified was characterized by high mortality (49%), older age, male sex, elevated inflammatory markers, high prevalence of cardiovascular disease, and shock. Patients with this severe phenotype had significantly elevated peak C-reactive protein creatinine, D-dimer, and white blood cell count and lower minimum lymphocyte count compared with other phenotypes (P < .01, all comparisons). CONCLUSIONS: Among a cohort of hospitalized adults, we identified a severe phenotype of COVID-19 based on the characteristics of its clinical course and poor prognosis. These findings need to be validated in other cohorts, as improved understanding of clinical phenotypes and risk factors for their development could help inform prognosis and tailored clinical management for COVID-19.