Trajectories of Host-Response Subphenotypes in Patients With COVID-19 Across the Spectrum of Respiratory Support

Abstract

BackgroundHospitalized patients with severe COVID-19 follow heterogeneous clinical trajectories, requiring different levels of respiratory support and experiencing diverse clinical outcomes. Differences in host immune responses to SARS-CoV-2 infection may account for the heterogeneous clinical course, but we have limited data on the dynamic evolution of systemic biomarkers and related subphenotypes. Improved understanding of the dynamic transitions of host subphenotypes in COVID-19 may allow for improved patient selection for targeted therapies.Research QuestionWe examined the trajectories of host-response profiles in severe COVID-19 and evaluated their prognostic impact on clinical outcomes.Study Design and MethodsIn this prospective observational study, we enrolled 323 inpatients with COVID-19 receiving different levels of baseline respiratory support: (1) low-flow oxygen (37%), (2) noninvasive ventilation (NIV) or high-flow oxygen (HFO; 29%), (3) invasive mechanical ventilation (27%), and (4) extracorporeal membrane oxygenation (7%). We collected plasma samples on enrollment and at days 5 and 10 to measure host-response biomarkers. We classified patients by inflammatory subphenotypes using two validated predictive models. We examined clinical, biomarker, and subphenotype trajectories and outcomes during hospitalization.ResultsIL-6, procalcitonin, and angiopoietin 2 persistently were elevated in patients receiving higher levels of respiratory support, whereas soluble receptor of advanced glycation end products (sRAGE) levels displayed the inverse pattern. Patients receiving NIV or HFO at baseline showed the most dynamic clinical trajectory, with 24% eventually requiring intubation and exhibiting worse 60-day mortality than patients receiving invasive mechanical ventilation at baseline (67% vs 35%; P < .0001). sRAGE levels predicted NIV failure and worse 60-day mortality for patients receiving NIV or HFO, whereas IL-6 levels were predictive in all patients regardless of level of support (P < .01). Patients classified to a hyperinflammatory subphenotype at baseline (< 10%) showed worse 60-day survival (P < .0001) and 50% of them remained classified as hyperinflammatory at 5 days after enrollment.InterpretationLongitudinal study of the systemic host response in COVID-19 revealed substantial and predictive interindividual variability influenced by baseline levels of respiratory support. Hospitalized patients with severe COVID-19 follow heterogeneous clinical trajectories, requiring different levels of respiratory support and experiencing diverse clinical outcomes. Differences in host immune responses to SARS-CoV-2 infection may account for the heterogeneous clinical course, but we have limited data on the dynamic evolution of systemic biomarkers and related subphenotypes. Improved understanding of the dynamic transitions of host subphenotypes in COVID-19 may allow for improved patient selection for targeted therapies. We examined the trajectories of host-response profiles in severe COVID-19 and evaluated their prognostic impact on clinical outcomes. In this prospective observational study, we enrolled 323 inpatients with COVID-19 receiving different levels of baseline respiratory support: (1) low-flow oxygen (37%), (2) noninvasive ventilation (NIV) or high-flow oxygen (HFO; 29%), (3) invasive mechanical ventilation (27%), and (4) extracorporeal membrane oxygenation (7%). We collected plasma samples on enrollment and at days 5 and 10 to measure host-response biomarkers. We classified patients by inflammatory subphenotypes using two validated predictive models. We examined clinical, biomarker, and subphenotype trajectories and outcomes during hospitalization. IL-6, procalcitonin, and angiopoietin 2 persistently were elevated in patients receiving higher levels of respiratory support, whereas soluble receptor of advanced glycation end products (sRAGE) levels displayed the inverse pattern. Patients receiving NIV or HFO at baseline showed the most dynamic clinical trajectory, with 24% eventually requiring intubation and exhibiting worse 60-day mortality than patients receiving invasive mechanical ventilation at baseline (67% vs 35%; P < .0001). sRAGE levels predicted NIV failure and worse 60-day mortality for patients receiving NIV or HFO, whereas IL-6 levels were predictive in all patients regardless of level of support (P < .01). Patients classified to a hyperinflammatory subphenotype at baseline (< 10%) showed worse 60-day survival (P < .0001) and 50% of them remained classified as hyperinflammatory at 5 days after enrollment. Longitudinal study of the systemic host response in COVID-19 revealed substantial and predictive interindividual variability influenced by baseline levels of respiratory support. Take-home PointsStudy Question: We examined the trajectories of host-response profiles in severe COVID-19 and evaluated their prognostic impact on clinical outcomes.Results: Lung epithelial injury plasma biomarker sRAGE (soluble receptor of advanced glycation end products) predicted adverse outcome in patients supported by noninvasive respiratory methods earlier during the inpatient course of COVID-19, whereas systemic inflammation measured by plasma IL-6 levels was predictive across all time points and regardless of level of support. By synthesis of host-response biomarkers into subphenotypes, patients classified to a hyperinflammatory subphenotype either at baseline or during follow-up showed markedly worse survival than their counterparts classified to a hypoinflammatory subphenotype.Interpretation: Longitudinal study of host response in severe COVID-19 demonstrated prognostic interindividual variability influenced by baseline levels of respiratory support. Study Question: We examined the trajectories of host-response profiles in severe COVID-19 and evaluated their prognostic impact on clinical outcomes. Results: Lung epithelial injury plasma biomarker sRAGE (soluble receptor of advanced glycation end products) predicted adverse outcome in patients supported by noninvasive respiratory methods earlier during the inpatient course of COVID-19, whereas systemic inflammation measured by plasma IL-6 levels was predictive across all time points and regardless of level of support. By synthesis of host-response biomarkers into subphenotypes, patients classified to a hyperinflammatory subphenotype either at baseline or during follow-up showed markedly worse survival than their counterparts classified to a hypoinflammatory subphenotype. Interpretation: Longitudinal study of host response in severe COVID-19 demonstrated prognostic interindividual variability influenced by baseline levels of respiratory support. SARS-CoV-2 has infected > 676 million individuals and led to > 6.8 million deaths worldwide,1World Health OrganizationWHO coronavirus (COVID-19) dashboard. 2023. World Health Organization website.https://covid19.who.int/Date accessed: March 12, 2023Google Scholar with > 1 million deaths in the United States2Centers for Disease Control and PreventionCOVID data tracker. 2023. Centers for Disease Control and Prevention website.https://covid.cdc.gov/covid-data-tracker/#datatracker-homeDate accessed: March 12, 2023Google Scholar as of March 2023. Extensive research has shown that a dysregulated inflammatory response against the virus develops in patients with COVID-19 with severe illness requiring hospitalization, often leading to acute respiratory failure with parenchymal lung damage and impaired gas exchange.3Xu Z. Shi L. Wang Y. et al.Pathological findings of COVID-19 associated with acute respiratory distress syndrome.Lancet Respir Med. 2020; 8: 420-422Abstract Full Text Full Text PDF PubMed Scopus (6170) Google Scholar Current care consists of two main elements: (1) provision of appropriate respiratory support (invasive or noninvasive options) to improve gas exchange and work of breathing, and (2) delivery of timely and effective antiviral and immunomodulatory therapies4Tomazini B.M. Maia I.S. Cavalcanti A.B. et al.Effect of dexamethasone on days alive and ventilator-free in patients with moderate or severe acute respiratory distress syndrome and COVID-19: the CoDEX Randomized Clinical Trial.JAMA. 2020; 324: 1307-1316Crossref PubMed Scopus (812) Google Scholar,5Gordon A.C. Mouncey P.R. Al-Beidh F. et al.REMAP-CAP InvestigatorsInterleukin-6 receptor antagonists in critically ill patients with Covid-19.N Engl J Med. 2021; 384: 1491-1502Crossref PubMed Scopus (1143) Google Scholar to curtail the aberrant inflammatory response. The provision of the first main element of care, appropriate respiratory support, is dynamic and responsive to clinical changes at the bedside. Provision of the second main element of care, antiviral and immunomodulatory agents, is based largely on cross-sectional assessments of respiratory failure severity and crude biomarkers that are available clinically (eg, C-reactive protein levels for anti-IL-6 treatment initiation). However, the systemic inflammatory response in severe COVID-19 is complex, with multiple pathways involved and differences compared with non-COVID ARDS.6Bain W. Yang H. Shah F.A. et al.COVID-19 versus non-COVID-19 acute respiratory distress syndrome: comparison of demographics, physiologic parameters, inflammatory biomarkers, and clinical outcomes.Ann Am Thorac Soc. 2021; 18: 1202-1210Crossref PubMed Scopus (73) Google Scholar Extensive research in non-COVID ARDS has shown replication validity of distinct host-response subphenotypes (eg, hyperinflammatory and hypoinflammatory), potentially offering new opportunities for targeted therapeutics.7Drohan C.M. Nouraie S.M. Bain W. et al.Biomarker-based classification of patients with acute respiratory failure into inflammatory subphenotypes: a single-center exploratory study.Crit Care Explor. 2021; 3e0518Crossref PubMed Google Scholar, 8Calfee C.S. Delucchi K. Parsons P.E. Thompson B.T. Ware L.B. Matthay M.A. Subphenotypes in acute respiratory distress syndrome: latent class analysis of data from two randomised controlled trials.Lancet Respir Med. 2014; 2: 611-620Abstract Full Text Full Text PDF PubMed Scopus (835) Google Scholar, 9Sinha P. Delucchi K.L. Thompson B.T. McAuley D.F. Matthay M.A. Calfee C.S. Latent class analysis of ARDS subphenotypes: a secondary analysis of the statins for acutely injured lungs from sepsis (SAILS) study.Intensive Care Med. 2018; 44: 1859-1869Crossref PubMed Scopus (182) Google Scholar, 10Sinha P. Spicer A. Delucchi K.L. McAuley D.F. Calfee C.S. Churpek M.M. Comparison of machine learning clustering algorithms for detecting heterogeneity of treatment effect in acute respiratory distress syndrome: a secondary analysis of three randomised controlled trials.EBioMedicine. 2021; 74103697Abstract Full Text Full Text PDF PubMed Scopus (17) Google Scholar Such biomarker-based subphenotypes also have been described in COVID-19 ARDS and may allow better targeting of immunomodulatory interventions. Enhanced understanding of the dynamic variability of the longitudinal systemic inflammatory response in patients with COVID-19 across the spectrum of respiratory failure severity may help to improve prognostication and patient selection for timely interventions. In this prospective, observational study spanning the first 2 years of the SARS-CoV-2 pandemic, we collected longitudinal data in two independent cohorts of inpatients with COVID-19 requiring different levels of respiratory support. We characterized the clinical, biomarker, and subphenotype trajectories in COVID-19 and examined whether host-response biomarkers and subphenotypes had different prognostic value on patient outcomes depending on the baseline level of respiratory support. Detailed methods are provided in e-Appendix 1. We prospectively enrolled hospitalized patients with COVID-19 in two independent, prospective cohort studies within the University of Pittsburgh Medical Center Health System (see e-Appendix 1 for details): the Acute Lung Injury Registry and Biospecimen Repository enrolled critically ill patients with COVID-19 hospitalized in ICUs,6Bain W. Yang H. Shah F.A. et al.COVID-19 versus non-COVID-19 acute respiratory distress syndrome: comparison of demographics, physiologic parameters, inflammatory biomarkers, and clinical outcomes.Ann Am Thorac Soc. 2021; 18: 1202-1210Crossref PubMed Scopus (73) Google Scholar and the COVID INpatient Cohort enrolled moderately ill inpatients with COVID-19 hospitalized in dedicated inpatient wards.11Al-Yousif N. Komanduri S. Qurashi H. et al.Inter-rater reliability and prognostic value of baseline Radiographic Assessment of Lung Edema (RALE) scores in observational cohort studies of inpatients with COVID-19.BMJ Open. 2023; 13e066626Crossref PubMed Scopus (2) Google Scholar We enrolled patients after admission to the hospital and obtained informed consent from the patients or their legally authorized representatives under study protocols STUDY19050099 and STUDY20040036 approved by the University of Pittsburgh Institutional Review Board. We collected baseline blood samples on enrollment (day 1) and at follow-up intervals (days 5 and 10) for those who remained hospitalized, and we measured host-response biomarkers. From the electronic medical record, we extracted data on demographics, comorbid conditions, vital signs, and laboratory test results at baseline, as well as immunomodulatory treatments received during hospitalization. We broadly classified baseline respiratory support in four ordinal categories of increased intensity, referred to as clinical groups: (1) low-flow oxygen (LFO), that is, patients with a conventional nasal cannula or oxygen mask; (2) noninvasive ventilation (NIV) or high-flow oxygen (HFO), that is, patients receiving either NIV (continuous or bilevel positive airway pressure) or humidified, heated HFO delivered via nasal cannula or mask; (3) invasive mechanical ventilation (IMV) via endotracheal intubation; and (4) extracorporeal membrane oxygenation (ECMO) support. We recorded the clinical group trajectories starting from date of symptom onset to date of positive polymerase chain reaction (PCR) testing, as well as hospital or ICU admission. We followed up patients prospectively for the type(s) of respiratory support required during follow-up and recorded escalation or de-escalation of support. Because NIV and HFO often were in the as respiratory support to we patients as NIV or HFO for those who were supported by NIV or HFO at baseline but up requiring or or during hospitalization, vs NIV or HFO for those who required intubation and hospitalization. The outcome was 60-day survival from hospital and the secondary outcome was the of NIV or HFO vs a we measured plasma host-response biomarkers 1) and on four biomarkers with in COVID-19 (1) A. et plasma IL-6 and levels are associated with adverse clinical outcomes and in critically ill SARS-CoV-2 inflammatory response of SARS-CoV-2 Scholar a of approved immunomodulatory for (2) F. L. et value of C-reactive and in patients with 2020; Scopus Google Scholar as a biomarker for secondary (3) soluble receptor of advanced glycation end products A. Thompson B.T. et and injury in severe J Respir Care Med. PubMed Scopus Google S. M.A. et and prognostic in a prospective cohort 2021; Full Text Full Text PDF PubMed Scopus Google Scholar a biomarker for epithelial and (4) angiopoietin et admission levels of biomarkers as of mortality in critically ill COVID-19 2021; PubMed Scopus (73) Google L. N. et plasma levels of epithelial and in COVID-19 with lung after hospital PubMed Scopus Google Scholar a biomarker for From a of available plasma we also SARS-CoV-2 with Bain W. A. et acute respiratory syndrome coronavirus 2 is associated with coronavirus severity and clinical PubMed Scopus Google A. Shah F.A. et SARS-CoV-2 levels as a biomarker of respiratory SARS-CoV-2 infection in critically ill patients with PubMed Scopus Google Scholar We classified patients into host-response subphenotypes by two biomarker-based that had been via latent class (1) the by et C.M. Nouraie S.M. Bain W. et al.Biomarker-based classification of patients with acute respiratory failure into inflammatory subphenotypes: a single-center exploratory study.Crit Care Explor. 2021; 3e0518Crossref PubMed Google Scholar the using procalcitonin, soluble receptor and angiopoietin 2 levels using the for subphenotype classification and (2) the by et P. Delucchi K.L. McAuley D.F. C.M. Matthay M.A. Calfee C.S. and of algorithms to acute respiratory distress syndrome a secondary analysis of randomised controlled trials.Lancet Respir Med. 2020; 8: Full Text Full Text PDF PubMed Scopus Google Scholar the using IL-6, and soluble receptor 1 using the the of this in patients with 2 by characterized by elevated levels of biomarkers, was as whereas 1 was as We compared and respiratory support or subphenotypes with the and We of biomarker for We analysis by SARS-CoV-2 and immunomodulatory therapies. We examined the of biomarker levels time using against time from hospital admission with patient and of for time of respiratory support, as well as by biomarker levels follow-up intervals (days and 60-day we for time to from hospital as well as for time from hospital and of respiratory support at We for the and with the 60-day we for time from hospital and of respiratory support at for prognostic for biomarker or subphenotype by level of baseline respiratory support on 60-day we (eg, biomarker of respiratory in the and examined the (P < of the the outcome of NIV or HFO we a for clinical of NIV or HFO vs failure of and L. L. respiratory support for COVID-19 for and PubMed Scopus Google Scholar We all in for March and March we enrolled a of 323 patients with COVID-19 patients were and with a of baseline, we classified patients into the clinical of NIV or HFO and Patients with were often were and showed higher than the clinical Patients enrolled in this cohort with showed of compared with patients with COVID-19 hospitalized across the University of Pittsburgh Medical Center during a study et learning to the COVID-19 changes in clinical and mortality hospitalized Health PubMed Scopus (3) Google and or NIV of as of agents, against or or of as or from symptom from positive results for from subphenotype to University of Pittsburgh Medical Center hospital and or depending on and during the during are as or extracorporeal membrane HFO high-flow invasive mechanical low-flow NIV noninvasive polymerase chain as of agents, against or or of as or to University of Pittsburgh Medical Center hospital and or depending on and during the in a new are as or extracorporeal membrane HFO high-flow invasive mechanical low-flow NIV noninvasive polymerase chain Patients receiving at the time of enrollment showed a time from COVID-19 positive results and onset of followed by patients receiving and NIV or HFO, of COVID-19 compared with patients receiving in COVID-19 of illness at time of enrollment was also supported by differences in plasma with patients receiving NIV or HFO the levels potentially earlier of SARS-CoV-2 infection with higher all we that patients requiring respiratory support at time of enrollment worse 60-day survival We examined the clinical group trajectories starting from baseline to level of respiratory support required during The NIV or HFO group showed the most clinical group with 24% requiring escalation to or patients who required escalation from NIV or HFO to or showed the 60-day mortality compared with the of the In baseline IL-6, procalcitonin, and angiopoietin 2 increased with higher level of support from to whereas for the patients receiving showed the levels compared with the differences levels of support in for time from symptom onset as well as in 5 and 10 By of biomarker levels from time of we different trajectories for the of time by of respiratory support (P < for for all four biomarkers. for we trajectories for of time of from hospital for patients receiving NIV or HFO, and that sRAGE is a biomarker that earlier in the COVID-19 The two and showed in baseline under the with and of patients classified to the hyperinflammatory Patients classified to the hyperinflammatory subphenotype showed higher and worse (P < patients with available follow-up from days 1 to 5 were for the hypoinflammatory subphenotype and transitions by the and but for the 1 hyperinflammatory with 50% of patients as hypoinflammatory on 5 by In of biomarker levels 60-day and we that showed higher levels of all four biomarkers at baseline compared with but IL-6 and angiopoietin 2 levels persistently were elevated in across all three time points We examined whether baseline level of support the biomarkers and 60-day mortality in that for biomarker and clinical group We that the prognostic for and sRAGE were by baseline respiratory support, whereas the prognostic for IL-6 and angiopoietin 2 were by respiratory support level In exploratory we compared biomarker levels of patients receiving NIV or HFO with vs noninvasive support and that the group showed higher sRAGE and levels (P < .0001) We for the outcome of noninvasive vs failure for clinical and that higher baseline sRAGE levels were independent of noninvasive failure subphenotypes by the two also were predictive of after for the different baseline levels of respiratory support. We that patients classified to the hyperinflammatory subphenotype at baseline showed higher for compared with patients classified as hypoinflammatory for the and the patients who were classified as hyperinflammatory by worse outcomes than patients classified as hypoinflammatory by as well as those with In for the outcome of 60-day we that the prognostic of subphenotype were by level of baseline respiratory support we that patient receiving was classified as hyperinflammatory subphenotype by the classification transitions by 5 were also prognostic of outcomes patients who to 5 and were Patients who or as hyperinflammatory subphenotype by 5 showed worse mortality P compared with patients who were classified as hypoinflammatory by 5 hypoinflammatory from baseline or of the baseline hyperinflammatory classification by the We demonstrated distinct clinical and biomarker trajectories that predicted patient outcomes in a prospective, observational study of hospitalized patients with COVID-19 across the spectrum of illness biomarker with clinical outcomes on the level of respiratory support at time of a biomarker of epithelial was predictive of outcome patients receiving NIV or HFO, whereas IL-6 was predictive patients receiving or sRAGE levels during hospitalization, whereas biomarkers showed or of host-response profiles with showed of the hyperinflammatory subphenotype in patients with COVID-19, but patients classified in the hyperinflammatory subphenotype showed markedly worse independent of the level of respiratory support. The clinical revealed that patients at different levels of support in the may different of SARS-CoV-2 infection and the host response. The plasma in patients receiving NIV or HFO that patients may in a of replication than patients requiring invasive support. Such have for treatment by targeting and based on the evolution of the From a care we the clinical of patients enrolled in the NIV or HFO most patients receiving NIV or HFO were supported patients for this showed the 60-day mortality of all We that patients with a NIV or HFO showed markedly higher sRAGE levels at the time of baseline compared with those with a (P < .0001) findings the that assessments of sRAGE levels in cohorts of patients with NIV or HFO support may in with assessments as to patients may from of a noninvasive vs those who The dynamic trajectories of sRAGE level new into prognostic a of epithelial sRAGE level to with COVID-19 showed a with the patients receiving markedly sRAGE levels in patients receiving may that patients injury and of sRAGE into the also is by the levels of sRAGE in patients for NIV or HFO in are to and may However, we have of and patients receiving different levels of respiratory support, and we for a sRAGE level and lung we also that sRAGE levels a that was different from the biomarkers and with A. Thompson B.T. et and injury in severe J Respir Care Med. PubMed Scopus Google Scholar that patients earlier in the course showed higher plasma SARS-CoV-2 is that sRAGE also may replication and lung injury in earlier of COVID-19 L. et has and prognostic value in hospitalized patients with PubMed Scopus Google Scholar In patients with non-COVID plasma sRAGE levels have been associated with worse in gas exchange and and Yang L. et evolution of in ARDS and with clinical a prospective cohort study in 2020; PubMed Scopus Google et changes time in the of lung are associated with survival in 2020; Full Text Full Text PDF PubMed Scopus Google Scholar plasma sRAGE levels may to epithelial injury and have results the as to whether sRAGE levels a dynamic of patient or lung examined in with validated from cohorts with non-COVID ARDS and respiratory failure prognostic across the spectrum of COVID-19 We two different that different of biomarkers to their prognostic soluble receptor 1 and whereas the IL-6 levels and the angiopoietin 2 and we of the hyperinflammatory subphenotype (< 10%) by but patients were classified as hyperinflammatory by patients showed the were from baseline to for the hypoinflammatory but patients classified as hyperinflammatory demonstrated dynamic with 50% of them classified as hypoinflammatory on follow-up by for Patients to the hyperinflammatory subphenotype on follow-up or those who persistently were classified as hyperinflammatory showed worse outcomes compared with patients classified as supported the of subphenotypes in non-COVID K. Ware L.B. Parsons P.E. Thompson B.T. Calfee C.S. of ARDS subphenotypes time in two randomised controlled 2018; PubMed Scopus Google Scholar but data in patients with COVID-19 with acute respiratory failure the for better understanding of the prognostic value and of study has data a care may of we enrolled patients from different and inpatient from three different in of enrollment was of in consent from legally authorized patient we obtained dedicated for research and with the after blood Yang L. et subphenotypes prognostic in patients with or at for acute respiratory distress Care Med. PubMed Scopus Google Scholar related to of clinical plasma of the were obtained in the hospital course, and biomarker levels may have been influenced by of immunomodulatory was limited by either of clinical and from the hospital or of of the longitudinal data and within the of COVID-19 respiratory failure requiring hospitalization. We to patient trajectories based on of COVID-19 as of testing, symptom and hospitalization. we patients receiving NIV or HFO into a we of positive (NIV) vs ventilation on host immune and injury biomarkers. the of > patients in the of the clinical were for subphenotype in follow-up is outcome and on of findings to patients with ARDS from non-COVID-19 the observational of study for prognostic assessments of biomarkers and subphenotypes on outcomes of the to for predictive and treatment effect heterogeneity by Calfee C.S. Thompson B.T. and for sepsis and acute respiratory distress syndrome clinical J Respir Care Med. PubMed Scopus Google Scholar Longitudinal of the systemic host response in hospitalized patients with COVID-19 revealed substantial and prognostic interindividual was influenced by baseline levels of respiratory support. studies the prognostic value of biomarkers and subphenotypes in COVID-19 and acute respiratory failure to for clinical illness trajectory, of and respiratory support predictive with of patients for enrollment in clinical may allow for better targeting of host and improved outcomes in K. is supported by the University of Pittsburgh Clinical and COVID-19 and the of Health and W. is supported by the of is supported by the of Health

Topics & Concepts