Article Text
Abstract
Objectives To evaluate the potential of soluble cluster of differentiation 146 (sCD146) in the detection and grading of congestion in patients with acute dyspnoea.
Design Subanalysis of the prospective observational Lithuanian Echocardiography Study of Dyspnoea in Acute Settings (LEDA) cohort.
Setting Two Lithuanian university centres.
Participants Adult patients with acute dyspnoea admitted to the emergency department.
Methods Congestion was assessed using clinical and sonographic parameters. All patients underwent sCD146 and N-terminal pro-B-type natriuretic peptide (NT-proBNP) testing.
Results The median value of sCD146 concentration in the study cohort (n=437) was 405 (IQR 315–509) ng/mL. sCD146 was higher in patients with peripheral oedema than in those without (median (IQR) 472 (373–535) vs 400 (304–501) ng/mL, p=0.009) and with pulmonary rales than in those without (439 (335–528) vs 394 (296–484) ng/mL, p=0.001). We found a parallel increase of estimated right atrial pressure (eRAP) and sCD146 concentration: sCD146 was 337 (300–425), 404 (290–489) and 477 (363–572) ng/mL in patients with normal, moderately elevated and high eRAP, respectively (p=0.001). In patients with low NT-proBNP, high sCD146 distinguished a subgroup with a higher prevalence of oedema as compared with patients with low levels of both biomarkers (76.0% vs 41.0%, p=0.010). Moreover, high sCD146 indicated a higher prevalence of elevated eRAP, irrespective of NT-proBNP concentration (p<0.05).
Conclusion sCD146 concentration reflects the degree of intravascular and tissue congestion assessed by clinical and echocardiographic indices, with this association maintained in patients with low NT-proBNP. Our data support the notion that NT-proBNP might represent heart stretch while sCD146 rather represents peripheral venous congestion.
- Echocardiography
- Heart failure
- Adult cardiology
- Ultrasound
- VASCULAR MEDICINE
Data availability statement
Data are available upon reasonable request. The data presented in this study are available on request from the corresponding author. The data are not publicly available due to patient privacy.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
This study investigated important relationships between circulating soluble cluster of differentiation 146 (sCD146) and clinical and echocardiographic markers of tissue, intravascular and intracardiac congestion, in the context of a prospective observational study cohort.
The sample size is limited as it was not always possible to perform a high-quality echocardiography study on admission.
When assessing the association between blood biomarkers and congestion, we relied on clinical and echo-derived markers of volume overload, instead of invasively measured intracardiac filling pressures.
We were also unable to test how fast sCD146 level decreases after diuretic administration and, more importantly, we did not assess whether sCD146-guided decongestion could improve clinical outcome.
Introduction
Assessment of clinical congestion is challenging, particularly in an acute setting as it can be present in the vascular system (intravascular congestion) or the interstitium (tissue congestion), although the majority of patients will have a mix of both.1 Endothelial soluble cluster of differentiation 146 (sCD146) is a cell adhesion molecule that is secreted in the intercellular junction of endothelial cells mediating interaction with other cells or extracellular matrix.2 sCD146 is involved in the control of vessel integrity, while its release is potentially dependent on endothelial cell stretch.3 This makes sCD146 a potential marker for congestion.4
This study was designed to examine the levels of sCD146 in a cohort of patients with acute dyspnoea due to multiple causes and its relationship with clinical signs and sonographic markers of congestion as well as cardiac morphology and function.
Methods
Study design
This is a subanalysis of the Lithuanian Echocardiography Study of Dyspnoea in Acute Settings (LEDA, NCT03048032). LEDA was a prospective observational multicentre study performed in two Lithuanian university centres in collaboration with a research protocol of the international Global Research on Acute Conditions Team network. The enrolment period, inclusion/exclusion criteria, data collection and diagnosis adjudication have been previously described.5 Briefly, adult patients with acute dyspnoea admitted to the emergency department at two Lithuanian university centres were enrolled; patients with acute coronary syndrome (ACS) occurring during the first 4 hours of admission were excluded. Demographic data, baseline medication, comorbidities, clinical signs and laboratory parameters at admission were recorded. Clinical signs of congestion included peripheral oedema and pulmonary rales.
The main cause of acute dyspnoea was adjudicated by three cardiologists in their respective centres. Final diagnoses were classified as acute heart failure (AHF) or non-AHF; the latter included chronic obstructive pulmonary disease (COPD)/asthma exacerbation, pulmonary embolism, pulmonary/non-pulmonary infections, cancer, ACS (which occurred later in the admission) and others.
This analysis includes a subgroup of patients who had their sCD146 levels measured at admission and cardiac ultrasound performed within 48 hours of admission.
Ultrasound examination
Echocardiographic measurements of cardiac chambers, ventricular systolic and diastolic function and inferior vena cava (IVC) diameters were obtained by experienced operators using System Vivid 4 (GE Healthcare, Israel), System Vivid 7 and 9 machines (GE Healthcare, Norway) and Philips EPIQ 7 (Koninklijke Philips, the Netherlands) according to the guidelines for cardiac chamber quantification.6 Right atrial pressure (RAP) was estimated based on the IVC size and collapsibility: normal–IVC diameter <2.1 cm and collapsing ≥50%, moderately elevated–IVC diameter ≥2.1 cm and collapsing ≥50% or IVC diameter <2.1 cm and collapsing <50%, high–IVC diameter ≥2.1 cm and collapsing <50%.
Focused right-sided parameters, including right ventricular (RV) basal diameter, tricuspid annular plane systolic excursion (TAPSE), RV fractional area change and peak systolic velocity of tricuspid annulus were obtained when feasible. Furthermore, images for the RV deformation assessment were obtained from apical four-chamber view as described by Rudski et al.7 Offline speckle tracking analysis was performed by the two-dimensional strain software in the EchoPac (V.110.1.2 GE Healthcare, Norway) and QLAB (V.9.0, Koninklijke Philips, the Netherlands). Both the entire RV strain (including basal, mid, and apical segments of the RV free wall and interventricular septum) and RV free wall strain (excluding septal segments) were calculated. Left-sided parameters related to intracardiac congestion included E/e’ ratio, left atrial volume index (LAVI) and left ventricular diastolic diameter (LVDD). Lung ultrasound was performed with a phased array transducer scanning in four chest sites bilaterally.8 The total number of B-lines was recorded as the final result.
Biomarkers
Blood samples were taken within 4 hours of presentation, frozen at −80°C and sent to the INSERM UMR942 institute (Paris, France) for centralised measurements of sCD146 (BioCytex, Marseille, France), N-terminal pro-B-type natriuretic peptide (NT-proBNP) and high-sensitive troponin-T (hs-TnT) (Roche Diagnostics, Mannheim, Germany).
Statistical analysis
Values were expressed as counts and frequencies for qualitative variables and as means and SDs or medians with IQR for quantitative variables, depending on the distribution.
All study parameters were compared between two groups based on the sCD146 median value for a more detailed description of the biomarker in the acute dyspnoea cohort. To assess the relationship between sCD146 and cardiac morphology and function, sCD146 concentration was compared between the groups based on terciles of echocardiographic parameters.
The possible value of combining sCD146 and NT-proBNP in assessing congestion was examined by dividing the patients into four subgroups on the basis of the median values of the biomarkers: (1) both biomarkers below the median, (2) only NT-proBNP above the median, (3) only sCD146 above the median and (4) both biomarkers above the median. Congestion parameters (presence of peripheral oedema and estimated RAP (eRAP)) were compared between the groups.
The χ2 test was used to compare categories. We used adjusted residuals with the Bonferroni correction for post hoc tests following the χ2 test. The means of continuous non-parametric variables were compared using the Mann-Whitney U or Kruskal-Wallis H test when appropriate. For group comparison, Dunn’s tests with the Bonferroni correction were performed.
Performance of sCD146 and NT-proBNP in the prediction of IVC dilation and reduced collapsibility, B-lines and cardiac abnormalities were assessed conducting a receiver operating characteristic (ROC) analysis using the area under the ROC curve (AUC). Change in two AUCs was tested by DeLong et al.9
The analysis was carried out using IBM SPSS Statistics for Windows, V.23.0 (IBM Corp) and MedCalc for Windows, V.14.8 (MedCalc software). All tests were two sided and a p value of <0.05 was considered significant.
Patient and public involvement
None.
Results
Study population
sCD146 measurements were available for 437 out of 1455 LEDA participants (30%). There were some differences in the baseline characteristics among patients included in the present subanalysis and the entire LEDA population (online supplemental table 1). The included patients were more likely to be dyspnoeic because of AHF and ACS, had higher plasma NT-proBNP and hs-TnT concentration and were more often on diuretics.
Supplemental material
The baseline characteristics of the study population are shown in table 1. The patients were elderly (median age 70 (61–78) years) and 270 (68.1%) were male patients. AHF was the dominant cause of acute dyspnoea. The frequency of cardiovascular and non-cardiovascular comorbidities was high.
sCD146 in the study population
The median value of sCD146 concentration in the study cohort was 405 (315–509) ng/mL. The comparison of patients with plasma sCD146 concentration above and below the median is shown in table 1. Patients with sCD146 levels above the median were older and more frequently diagnosed with chronic heart failure, atrial fibrillation, valvular heart disease and anaemia before admission. Pulmonary rales at admission and chronic diuretic treatment were more common in patients with elevated sCD146: 58.9% vs 48.2% (p=0.036) and 53.0% vs 43.1% (p=0.044), respectively. Likewise, patients with sCD146 levels above the median had a higher plasma concentration of NT-proBNP, hs-TnT and creatinine (p<0.001 for all, see table 1). Regarding the cause of acute dyspnoea, the highest median value of sCD146 was found in patients with AHF: 441 (344–541) ng/mL, while the lowest was measured in patients diagnosed with COPD/asthma: 271 (220–363) ng/mL (online supplemental table 2).
sCD146 and clinical signs of congestion
sCD146 concentration was higher in patients with oedema, as compared with those without oedema (472 (373–535) vs 400 (304–501) ng/mL, p=0.009) (documented data about the presence or absence of peripheral oedema were available for 161 (37%) patients). Of note, the median sCD146 concentration in patients without a documented oedema status (394 (306–498)) was like the concentration measured in patients with no oedema on admission (p=0.93).
Likewise, sCD146 concentration was higher in patients with pulmonary rales than in those without rales (439 (335–528) and 394 (296–484) ng/mL, p=0.001, data about the presence or absence of rales were available for 404 (92%) patients). Again, the median sCD146 concentration in patients without a documented rales status (384 (273–480)) was similar to the concentration detected in patients with no rales (p=0.79). Echocardiographic parameters in the total study population and the subgroups with sCD146 levels higher or lower than the median are presented in online supplemental table 3.
sCD146 and echocardiographic parameters
We revealed a significant association between the concentration of sCD146 and several functional and structural ultrasound markers, including signs of intravascular and intracardiac congestion.
First, the association with a dilated IVC and respiratory variations in IVC was observed (complete data on the size and collapsibility of IVC were available for 276 (63%) patients). As for terciles of IVC diameter at expiration and IVC collapsibility, the more IVC was dilated and the less collapsible, the higher was the concentration of sCD146 (see figure 1A,B and online supplemental table 4). We also found a parallel increase in eRAP and plasma sCD146 concentration: sCD146 was 337 (300–425), 404 (290–489) and 477 (363–572) ng/mL in patients with normal, moderately elevated and high eRAP, respectively (p=0.001, figure 1C).
Furthermore, the sCD146 level was significantly related to the average E/e’ ratio—a marker of left ventricular diastolic filling (n=202 (46%)) (see figure 2A and online supplemental table 4). A parallel rise, though non-significant, was observed in sCD146 concentration with the increasing number of B-lines (n=77 (18%)) (figure 2B and online supplemental table 4), (p=0.06). The sCD146 concentration proportionally increased with the larger values of LVDD (figure 2C) and LAVI (figure 2D) (p<0.001 for both).
In addition, the sCD146 concentration corresponded to the markers of left and right ventricular function. Figure 2E,F shows an incremental increase of sCD146 levels in terciles of left ventricular ejection fraction (LVEF) and TAPSE. Similarly, the sCD146 concentration paralleled the decrease in the RV strain, both the free wall and the entire right ventricle (figure 2G,H).
sCD146 and NT-proBNP
In patients with high NT-proBNP, oedema was equally present in both high and low sCD146 subgroups (p=0.99). However, in patients with low NT-proBNP, high sCD146 distinguished a subgroup of patients with a significantly higher prevalence of oedema, as compared with patients with low levels of both biomarkers (76% vs 41%, p=0.010) (figure 3A). Furthermore, high sCD146 indicated a higher prevalence of elevated eRAP, irrespective of NT-proBNP concentration (p<0.05, figure 3B).
We also compared the value of sCD146 and NT-proBNP and the combination of both biomarkers in predicting the presence of congestion and cardiac dysfunction. AUCs for sCD146 and NT-proBNP were similar for IVC diameter, IVC collapse and the number of B-lines (p>0.05, figure 4 and online supplemental table 5). However, sCD146 better predicted the presence of peripheral oedema than NT-proBNP (AUC 0.63 (0.51–0.70) vs 0.51 (0.43–0.59), p=0.009), whereas NT-proBNP was a better predictor of LVEF and rales than sCD146 (AUC 0.76 (0.71–0.80) vs 0.63 (0.58–0.68), p<0.001 and AUC 0.66 (0.61–0.70) vs 0.60 (0.55–0.64), p=0.049, respectively). The AUC values for echocardiographic parameters are detailed in online supplemental table 5.
Discussion
The present study reveals an important relationship between circulating sCD146 and clinical and echocardiographic markers of tissue, intravascular and intracardiac congestion, which is maintained in patients even with lower NT-proBNP.
Clinical congestion is a result of a complex pathophysiological process which in the case of heart failure starts with a gradual increase in filling pressures and culminates with extravascular fluid accumulation.10 Yet, congestion is the main reason for hospitalisation in an AHF setting, as the majority of patients show up with signs and symptoms of volume overload.10 It is however difficult to assess the level of congestion, and accurate detection and grading are therefore crucial for optimal medical care.
Along with a significant reduction in the length of AHF-related inpatient stays, the readmission rate remained unchanged during the past four decades,11 meaning that a large proportion of patients with AHF are readmitted for acute decompensation within a year.12 The deleterious effect of congestion in decompensated heart failure on outcome has been identified in other clinical scenarios, including end-stage renal disease,13 COPD14 and COVID-19 infection.15 Lung ultrasound is being increasingly recognised as a useful tool to assess pulmonary congestion and pleural effusion in acute dyspnoea,15–18 but may not be accessible in all facilities. Despite obvious harm, congestion management often remains suboptimal. This means that many patients remain wet and possibly experience overhydration-related events.
This difficulty is prevalent in the setting of acute dyspnoea, given often similar clinical presentations despite a diverse spectrum of aetiologies.19–21 Delayed identification leads to delays in initiation of decongestive therapies in acute dyspnoea, although early diuretic administration has been related to a better outcome in AHF.22
Recently, sCD146—a promising blood biomarker for congestion—has been introduced. Initially, sCD146 was identified as a marker for tumour progression and metastasis formation in human melanoma.23 Later, it was discovered that sCD146 is involved in the control of vessel integrity and angiogenesis,2 24–27 especially in different tumour pathogenesis, including paediatric leukaemia,28 breast cancer,29–31 melanoma32 and other types of cancer.33 34 Later on, Gayat et al showed that sCD146 concentration correlates well with weight gain and the size of IVC in patients with AHF.35 The relationship between venous stretch and sCD146 was demonstrated in an elegant experimental study by Arrigo et al.3 The authors measured the level of sCD146 in patients with heart failure at baseline and after 90 min of unilateral forearm venous congestion, and documented a rapid release of sCD146 in response to congestion-mediated venous stretch that may reflect systemic congestion in chronic heart failure. Another study revealed a relationship between sCD146 and pulmonary congestion in the early phase of ACS, which was independent from the severity of myocardial cell necrosis.36
The present study extends the accumulated data by revealing a close association between the levels of sCD146 and clinical and echocardiographic evidence of congestion as well as cardiac morphology and function in an unselected population of acute dyspnoea.
We found that median sCD146 concentration in patients with different aetiologies of acute dyspnoea differs, with patients with AHF having the highest concentrations, and patients with bronchial asthma (BA)/COPD having the lowest. Due to the modest sample size, we did not investigate this association further.
Regarding the left heart and pulmonary congestion, we were able to confirm a definite link between sCD146 concentration and the presence of pulmonary rales as well as impairment of diastolic filling reflected by elevated E/e’. We also showed that the greater the level of sCD146, the larger the left atrium and left ventricle, and the lower the LVEF. There was a parallel rise, though non-significant, in sCD146 with the increasing number of B-lines that requires further testing in a larger population.
Concerning the right ventricle and systemic congestion, the present results demonstrate a strong relationship between sCD146 concentration and clinical (peripheral oedema) as well as ultrasound signs of right-sided congestion. A stepwise increase in sCD146 concentration is demonstrated along with rising eRAP, consistent with intravascular volume overload. In addition, elevated sCD146 concentration was related to impaired RV function, as evidenced by a reduction in RV strain and TAPSE. The present findings indicate that sCD146 concentration proportionally increases not only with increasing congestion but also in parallel with decreasing biventricular function and progressing cardiac enlargement. Cardiac dysfunction is a likely cause of water retention in some cases, and sCD146 can reflect them both, that is, the cause and the consequence.
We have previously shown in dialysis patients that sCD146 rapidly follows patient hydration independently from the levels of natriuretic peptides. Most haemodialysis subjects with low BNP but high sCD146 have been shown to be overhydrated.37 The present study further confirms the significant additional value of sCD146 in detecting congestion in patients with acute dyspnoea with relatively low NT-proBNP. For instance, subjects with high sCD146 and low NT-proBNP frequently had elevated eRAP. By contrast, sCD146 had a minor additional value in patients with high NT-proBNP that likely carry both prominent intracardiac and peripheral congestion. However, in patients with relatively low NT-proBNP, sCD146 was able to discern a subgroup of patients with a high prevalence of peripheral oedema. Altogether, these data support the notion that natriuretic peptides are more likely to represent cardiac stretch in the context of heart failure, while sCD146 reflects hypervolaemia and peripheral venous stretch in acute dyspnoea.
Clinical significance
The present study indicates that NT-proBNP—an established diagnostic biomarker—may not accurately reflect systemic congestion in some scenarios. This observation is in line with some previous findings showing that BNP-guided congestion management is not superior to a conventional clinical approach.38 On the other hand, sCD146 presents a potential to be used as a congestion biomarker, and we hypothesise that its role in decongestion guidance might be superior to NT-proBNP, given the venous origin of its release.3 This may be particularly helpful in patients with acute dyspnoea not related to heart failure who may have hypervolaemia and may need diuretics despite low values of natriuretic peptides. These cases could potentially include patients with renal and hepatic failure, or subjects with severe malnutrition-related volume overload. The present data further suggest that the goal of diuretic treatment might be to reduce both NT-proBNP and sCD146. Our study was held in two academic facilities with immediate access to a cardiologist experienced in echocardiography, but we assume that in other settings many patients do not have or have delayed access to cardiac ultrasound. This means that sCD146 may assist emergency physicians, internists and intensivists in prompt congestion detection, leading to faster medical care. Also, this reliable congestion biomarker has a potential to fill the current gap of knowledge regarding congestion detection and grading, especially in emergency and primary care where modalities such as advanced echocardiography and lung ultrasound are often unavailable.
Limitations
This study has several limitations. First of all, the sample size is limited as it was not always possible to perform a high-quality echocardiography study on admission. Also, some patients were excluded from different parts of the analysis due to missing values in the patient history, which is often the case in a real-life emergency setting. Further research is needed to determine if sCD146 could potentially help to distinguish cardiac and non-cardiac dyspnoea as well as to guide diuretic treatment in different conditions. We do not have the data of deceleration time on echocardiography, which could give additional information on diastolic function. When assessing the association between blood biomarkers and congestion, we relied on clinical and echo-derived markers of volume overload instead of invasively measured intracardiac filling pressures. Still, we do not believe that such a trial of invasive nature would have ever been carried out in an acute dyspnoea setting. We were also unable to test how fast the sCD146 level decreases after diuretic administration and, more importantly, if sCD146-guided decongestion would improve outcome.
Conclusion
To sum up, sCD146 concentration reflects the degree of intravascular and tissue congestion as assessed by clinical and echocardiographic indices. This association remains present in patients with low NT-proBNP. sCD146 better represents peripheral venous congestion, while NT-proBNP might better represent cardiac stretch.
Data availability statement
Data are available upon reasonable request. The data presented in this study are available on request from the corresponding author. The data are not publicly available due to patient privacy.
Ethics statements
Patient consent for publication
Ethics approval
The study was approved by the Lithuanian Bioethics Committee (No. L-15-01) and conducted in accordance with the Declaration of Helsinki. All participants provided their written informed consent.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
AM and JČ contributed equally.
Contributors Conceptualisation—AM and JČ. Methodology—AM, JČ and KČ-B. Software—AuM, KČB, RJ and JS. Validation—AM, JČ, LB, DV, MB, GZ, JB, DŽ and KS. Formal analysis—RJ, JS, AuM, KČ-B, AM and JČ. Investigation—VJ, IA-M, DK, LB, DV, MB, GZ, JB, DŽ, KS, AK, JČ and AM. Resources—MA, VJ, IA-M, DK, LB, DV, MB, GZ, JB, DŽ, KS, AK, JČ and AM. Data curation—RJ, JS, AuM, KČ-B, AM and JČ. Writing (original draft preparation)—RJ, JS, KČ-B, JČ and AM. Writing (review and editing)—RJ, JS, AuM, KČ-B, MA, VJ, IA-M, DK, LB, DV, MB, GZ, JB, DŽ, KS, AK, JČ and AM. Visualisation—JČ and AuM. Supervision—JČ and AM. Guarantor - JČ.
Funding KČ-B, AM, VJ, DK, LB, MB, DV and JČ have received financial support from the Research Council of Lithuania for the submitted work.
Competing interests AM received speaker’s honoraria from Orion, Otsuka, Philips, Roche and Servier; received fee as a member of advisory board and/or Steering Committee and/or research grant from Adrenomed, Epygon, Neurotronik, Roche, Sanofi and Sphyngotec. AM owns shares in S-Form Pharma. AK received speaker’s honoraria from Servier, Bayer, Berlin-Chemie-Menarini, Pfizer and KRKA. JČ received investigator and speaker fees from Sanofi, Amgen, Novartis, Roche Diagnostics, Servier, AstraZeneca and Boehringer Ingelheim. DŽ received speaker fees from Novartis and AstraZeneca. JS received personal fees from Novartis, Bayer, Boehringer Ingelheim and Servier. VJ received personal fees from Bayer, Boehringer Ingelheim, Servier and Pfizer. Other authors declared no conflict of interest.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.