Skip to content
Publicly Available Published by De Gruyter February 22, 2021

Soluble fms-like tyrosine kinase-1: a potential early predictor of respiratory failure in COVID-19 patients

José L. Eguiburu-Jaime, Aitor Delmiro, Antonio Lalueza, Pedro L. Valenzuela, José M. Aguado, Carlos Lumbreras, Joaquín Arenas, Miguel A. Martín, Alejandro Lucia and Elena A. López-Jiménez

To the Editor,

A main feature of the most severely affected patients with coronavirus disease 2019 (COVID-19) is hypoxemic respiratory failure from acute respiratory distress syndrome (ARDS). The presence of intussusceptive (‘splitting’) angiogenesis and other vascular features has been reported to distinguish the lungs of patients who died from respiratory failure due to COVID-19 from those of individuals who died from influenza [1]. Angiogenesis is chiefly regulated by vascular endothelial growth factor (VEGF) and its associated family members, whose proangiogenic activity is mediated through their engagement with VEGF receptor 1 (VEGFR1, also known as ‘fms-like tyrosine kinase 1’ or Flt1) and VEGFR2. Another member of the VEGF family, placental growth factor (PlGF), stimulates endothelial healing and recruitment of mononuclear bone marrow cells as well as microvascular angiogenesis through engagement with Flt1, which is expressed in many tissues including the lung [2]. Alternative splicing of Flt1 pre-mRNA creates the soluble form of this receptor, known as ‘sFlt1’, which binds and antagonizes VEGF and PlGF signal [3]. Excess levels of sFlt1 induce endothelial dysfunction in hypoxia and contribute to cardiovascular disease [4].

In this single-center prospective cohort study, we aimed to analyze the potential role of sFlt-1 and PlGF, together with other ‘traditional’ biomarkers, as predictors of respiratory failure in patients hospitalized with COVID-19.

We enrolled patients hospitalized with COVID-19 at Hospital 12 de Octubre (Madrid, Spain) between 03/12/2020–04/20/2020. The study was approved by the local Ethics Committee (reference#20/222 and 20/117) and adhered to the Declaration of Helsinki. Patients’ main characteristics along with their progress and complications during hospitalization were extracted from electronic medical records. COVID-19 was confirmed via SARS-CoV-2 real-time reverse transcription polymerase chain reaction (nasopharyngeal swab or sputum samples). Respiratory failure was defined as the presence of an arterial oxygen partial pressure/fractional inspired oxygen ratio ≤200 mmHg or the need for mechanical ventilation (non-invasive positive pressure ventilation including high-flow nasal cannula oxygen, or invasive mechanical ventilation). Respiratory failure could occur at any time during hospital admission.

Blood samples were obtained within the first 48 h of hospitalization. Plasma sFlt1/PlG1 concentrations were measured using automated immunoassays (Cobas e601, Roche Diagnostics, Risch-Rotkreuz, Switzerland) following manufacturer’s instructions. Other major blood biomarkers associated with inflammation and/or COVID-19 were also determined (e.g., lactate dehydrogenase (LDH), C-reactive protein (CRP), lymphocyte count, albumin) as reported elsewhere [5].

To determine the clinical utility of blood biomarkers for prediction of respiratory failure (vs. not suffering this condition) in patients hospitalized with COVID-19, we used receiver operating characteristic (ROC) curve analysis. The sensitivity, specificity, positive and negative predictive values for the computed cutoff values were also calculated.

One hundred eleven hospitalized patients with confirmed COVID-19 were enrolled, of whom 52 (46.8% of total) developed respiratory failure 1–9 days following blood extraction (median, 2.5 days). Thirty-four (31% of total) of the patients died, most (n=32) having developed respiratory failure.

Significant differences were found for most blood biomarkers between patients who developed respiratory failure and those who did not (Table 1). On the other hand, four biomarkers (sFlt1, LDH, CRP and lymphocyte count) had area under the curve values >0.7 to predict the occurrence of respiratory failure in patients hospitalized with COVID-19, with sensitivity and specificity values ranging from 75.0 to 88.2% and from 52.8 to 75.4%, respectively, and with sFlt1 showing the best predictive accuracy (Figure 1).

Table 1:

Baseline characteristics and main results by group.

No respiratory failureRespiratory failurep-Value between groups
n=59n=52
Age, years71 [59.5, 80.5]74 [62.0, 81.2]0.304
Sex
Male31 (52.5%)37 (71.1%)0.053
Female28 (47.5%)15 (28.9%)0.053
Signs and symptoms on admission
Days with symptoms7 [4, 8]5 [4, 7]0.027
Cough48 (81.4%)39 (75.0%)0.491
Dyspnea27 (45.8%)36 (69.2%)0.021
Myalgia24 (40.7%)15 (28.8%)0.234
Diarrhoea19 (32.2%)16 (30.8%)1.000
Expectoration12 (20.3%)11 (21.1%)1.000
Vomit9 (15.2%)4 (7.7%)0.251
Bronchospasm2 (3.4%)3 (5.8%)0.664
Vital signs on admission
Systolic blood pressure, mmHg130 [121, 144]125 [116, 138]0.116
Heart rate, beats/min87 [79, 98]89 [74, 97]0.453
Body temperature, °C37.4 [36.9, 38.0]37.8 [37.3, 28.4]0.011
Respiratory rate, breaths/min18 [16, 20]21 [18, 23]0.001
Arterial oxygen saturation, %96 [93, 98]92 [89, 94]<0.001
PaO2 (mmHg)/FiO2 (%), mmHg438 [345, 471]305 [263, 433]0.012
Comorbidities
Hypertension29 (49.2%)34 (65.4%)0.124
Influenza vaccination (last year)31(52.5%)30 (57.7%)0.703
Dyslipidemia19 (32.2%)23 (44.2%)0.240
Obesity16 (27.1%)18 (34.6%)0.416
Diabetes15 (25.4%)14 (26.9%)1.000
Malignancies7 (11.9%)9 (17.3%)0.433
Chronic heart failure6 (10.2%)9 (15.2%)0.405
Acute myocardial infarction6 (10.2%)6 (11.5%)1.000
Chronic obstructive pulmonary disease1 (1.7%)10 (19.2%)0.003
Asthma7 (11.9%)3 (5.8%)0.331
Severe nephropathies4 (6.8%)6 (11.2%)0.501
Liver diseases3 (5.1%)7 (11.9%)0.185
Transplant6 (10.2%)4 (6.8%)0.748
Treatments on admission
Antihypertensive drugs28 (53.8%)32 (54.2%)0.252
ACE inhibitors17 (32.7%)13 (22.0%)0.675
ARB10 (16.9%)14 (23.7%)0.250
Other1 (1.7%)5 (8.5%)0.097
ASA11 (18.6%)10 (16.9%)1.000
Anticoagulants7 (11.9%)12 (23.1%)0.136
Steroids5 (8.5%)6 (10.2%)0.753
Treatments during hospitalization
Lopinavir/ritonavir32 (54.2%)37 (71.1%)0.079
Hydroxychloroquine55 (93.2%)47 (90.4%)0.737
Tocilizumab2 (3.4%)11 (21.1%)0.006
Antibiotics58 (98.3%)51 (98.1%)1.000
Steroids16 (27.1%)36 (69.2%)<0.001
Scores
SOFA1.29 ± 1.822.28 ± 1.630.005
qSOFA0.28 ± 0.450.71 ± 0.70<0.001
Charlson comborbidity index3.41 ± 2.134.39 ± 2.300.024
Exitus3 (5.1%)33 (63.5%)<0.001
eGFR (CKD-EPI, mL/min per 1.73 m2)20.3 [15.1, 26.1]24.0 [18.9, 30.5]0.046
Blood biochemical variables
sFlt1, ng/L88.2 [77.6, 98.6]118.8 [98.3, 139.4]<0.001
LDH, U/L287 [250, 348]404 [342, 476]<0.001
CRP, mg/L44.3 [22.0, 100.4]105.9 [70.8, 177.9]0.001
Lymphocyte count (×109/L)1.1 [0.9, 1.4]0.8 [0.6, 1.0]0.003
AST/ALT1.35 [1.03, 1.68]1.56 [1.26, 1.86]0.018
PlGF, ng/L84.6 [58.9, 97.5]71.4 [49.8, 87.0]0.022
Albumin, g/L37.6 [34.0, 39.8]35.9 [32.7, 38.0]0.066
RDW, %13.7 [13.4, 14.5]14.1 [13.6, 17.3]0.143
Ferritin, µg/L761 [336, 1,250]1,016 [549, 2,183]0.219
sFlt1/PlGF4.24 [3.55, 6.00]4.77 [3.83, 6.17]0.219

  1. Data are mean [interquartile range] for age and vital signs and symptoms on admission, median [interquartile range] for days with symptoms, mean ± standard deviation for scores, and n (%) for the rest of variables. p-Values corresponds to the comparisons between groups with the Mood ´s median test or with the χ2 or Fisher’s test (for proportions). Significant p-values (<0.05) are in bold. ACE, angiotensin converting enzyme; ARB, angiotensin-II receptor blockers; ALT, alanine transaminase; ASA, acetylsalicylic acid; AST, aspartate transaminase; CKD-EPI, chronic kidney disease epidemiology collaboration equation; CRP, C-reactive protein; FiO2, fractional inspired oxygen; eGFR, estimated glomerular filtration rate; LDH, lactate dehydrogenase; PaO2, arterial oxygen partial pressure; PlGF, placental growth factor; RDW, red blood cell distribution width; ROC, receiver operating characteristic; sFlt1, soluble fms-like tyrosine kinase 1; SOFA, sequential organ failure assessment.

Figure 1: Area under the curve (AUC), sensitivity, specificity, cut off points and Youden Index (J) values for the biomarkers with the highest AUC (>0.70) to predict occurrence of respiratory failure in patients hospitalized with COVID-19.CRP, C-reactive protein; LDH, lactate dehydrogenase; sFlt1, soluble fms-like tyrosine kinase 1.

Figure 1:

Area under the curve (AUC), sensitivity, specificity, cut off points and Youden Index (J) values for the biomarkers with the highest AUC (>0.70) to predict occurrence of respiratory failure in patients hospitalized with COVID-19.

CRP, C-reactive protein; LDH, lactate dehydrogenase; sFlt1, soluble fms-like tyrosine kinase 1.

Endothelial disease is an essential part of the pathological response to severe COVID-19, which leads to respiratory failure, multiorgan dysfunction and thrombosis [1]. Because the vascular endothelium depends on proangiogenic factors, excess release of antiangiogenic factors (e.g., sFlt1) is a possible cause of the endothelial dysfunction observed in patients with COVID-19. Yet, sFlt1 and PlGF (an anti and proangiogenic factor, respectively) have received scant attention in the context of this disease. Giardini et al. [6] reported higher values of sFlt1 and sFlt1/PlGF ratio in patients with COVID-19–associated pneumonia than in healthy controls. Dupont et al. [7] reported high sFlt1 circulating levels in patients severely affected with COVID-19, finding a correlation between sFlt1 and an endothelial dysfunction biomarker, soluble vascular cell adhesion molecule-1. On the other hand, the finding of a potential role of angiogenic processes in patients with COVID-19 opens the door for new therapeutic approaches. Low-dose aspirin (60–100 mg/day) inhibits the expression of sFlt1 in hypoxia-induced human cytotrophoblasts [8]. It can be thus hypothesized that aspirin administration might diminish the hypoxemic respiratory failure due to COVID-19–associated ARDS in critically ill patients. In addition, sFlt1 values could guide the timing of steroid administration.

To our knowledge, this is the first study with sFlt1 as an early predictor of respiratory failure in the context of COVID-19. Our findings suggest that this biomarker might identify a subgroup of hospitalized patients with this condition who are at higher risk of developing respiratory failure, and could thus help physicians to better triage these patients.


Corresponding author: Alejandro Lucia, MD, PhD, Research Institute of the Hospital 12 de Octubre (‘imas12’), Madrid, Spain; and Faculty of Sport Sciences, Universidad Europea de Madrid, 28670Villaviciosa de Odón, Madrid, Spain, Phone: (34) 661 39 31 01, E-mail:

Funding source: Spanish Ministry of Science and Innovation, Instituto de Salud Carlos III

Award Identifier / Grant number: COVID-19 research call COV20/00181

Funding source: Spanish Ministry of Economy and Competitiveness and Fondos Feder

Award Identifier / Grant number: PI18/00139

  1. Research funding: This study was supported by the Spanish Ministry of Science and Innovation, Instituto de Salud Carlos III (COVID-19 research call COV20/00181) co‐financed by the European Development Regional Fund “A way to achieve Europe”. Research by A.L. is funded by the Spanish Ministry of Economy and Competitiveness and Fondos Feder (grant PI18/00139). The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

  2. Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.

  3. Competing interests: Authors state no conflict of interest.

  4. Informed consent: Informed consent was obtained from all individuals included in this study.

  5. Ethical approval: The study was approved by the local Ethics Committee (reference#20/222 and 20/117) and adhered to the Declaration of Helsinki.

References

1. Ackermann, M, Verleden, SE, Kuehnel, M, Haverich, A, Welte, T, Laenger, F, et al.. Pulmonary vascular endothelialitis, thrombosis, and angiogenesis in Covid-19. N Engl J Med 2020;383:120–8. https://doi.org/10.1056/nejmoa2015432.Search in Google Scholar

2. Mazure, N, Brahimi-Horn, M, Pouysségur, J. Protein kinases and the hypoxia-inducible factor-1, two switches in angiogenesis. Curr Pharmaceut Des 2003;9:531–41. https://doi.org/10.2174/1381612033391469.Search in Google Scholar

3. Cindrova-Davies, T, Sanders, DA, Burton, GJ, Charnock-Jones, DS. Soluble FLT1 sensitizes endothelial cells to inflammatory cytokines by antagonizing VEGF receptor-mediated signalling. Cardiovasc Res 2011;89:671–9. https://doi.org/10.1093/cvr/cvq346.Search in Google Scholar

4. Hammadah, M, Georgiopoulou, VV, Kalogeropoulos, AP, Weber, M, Wang, X, Samara, MA, et al.. Elevated soluble fms-like tyrosine kinase-1 and placental-like growth factor levels are associated with development and mortality risk in heart failure. Circ Heart Fail 2016;9:1–6. https://doi.org/10.1161/circheartfailure.115.002115.Search in Google Scholar

5. Santos-Lozano, A, Calvo-Boyero, F, López-Jiménez, A, Cueto-Felgueroso, C, Castillo-García, A, Valenzuela, PI, et al.. Can routine laboratory variables predict survival in COVID-19? An artificial neural network-based approach. Clin Chem Lab Med 2020;58:e299–302. https://doi.org/10.1515/cclm-2020-0730.Search in Google Scholar

6. Giardini, V, Carrer, A, Casati, M, Contro, E, Vergani, P, Gambacorti-Passerini, C. Increased sFLT-1/PlGF ratio in COVID-19: a novel link to angiotensin II-mediated endothelial dysfunction. Am J Hematol 2020;95:E188–91. https://doi.org/10.1002/ajh.25882.Search in Google Scholar

7. Dupont, V, Kanagaratnam, L, Goury, A, Poitevin, G, Bard, M, Julien, G, et al.. Excess soluble fms-like tyrosine kinase 1 correlates with endothelial dysfunction and organ failure in critically ill COVID-19 patients. Clin Infect Dis 2020:ciaa1007. https://doi.org/10.1093/cid/ciaa1007 [Online ahead of print].Search in Google Scholar

8. Li, C, Raikwar, NS, Santillan, MK, Santillan, DA, Thomas, CP. Aspirin inhibits expression of sFLT1 from human cytotrophoblasts induced by hypoxia, via cyclo-oxygenase 1. Placenta 2015;36:446–53. https://doi.org/10.1016/j.placenta.2015.01.004.Search in Google Scholar

Received: 2021-01-14
Accepted: 2021-02-04
Published Online: 2021-02-22
Published in Print: 2021-06-25

© 2021 Walter de Gruyter GmbH, Berlin/Boston

Scroll Up Arrow