Abstract
Background
Adrenomedullin is a vasoactive hormone with potentially prognostic and therapeutic value, which mainly has been investigated in intensive care unit (ICU) settings. The triaging in the emergency department (ED) of patients to the right level of care is crucial for patient outcome.
Objectives
The primary aim of this study was to investigate the association of bioactive adrenomedullin (bio-ADM) with mortality among sepsis patients in the ED. Secondary aims were to investigate the association of bio-ADM with multiple organ failure (MOF), ICU admission and ED discharge.
Methods
In this prospective observational cohort study, adult sepsis patients in the ED (2013–2015) had blood samples collected for later batch analysis of bio-ADM. Odds ratios (OR) with 95% confidence interval (CI) for bio-ADM were calculated.
Results
Bio-ADM in 594 sepsis patients was analyzed of whom 51 died within 28 days (8.6%), 34 developed severe MOF, 27 were ICU admitted and 67 were discharged from the ED. The median (interquartile range) bio-ADM was 36 (26–56) and 63 (42–132) pg/mL among survivors and non-survivors, respectively, 81 (56–156) pg/mL for patients with severe MOF and 77 (42–133) pg/mL for ICU admitted patients. Each log-2 increment of bio-ADM conferred an OR of 2.30 (95% CI 1.74–3.04) for mortality, the adjusted OR was 2.39 (95% CI 1.69–3.39). The area under the receiver operating characteristic curve of a prognostic mortality model based on demographics and biomarkers increased from 0.80 to 0.86 (p = 0.02) when bio-ADM was added. Increasing bio-ADM was associated with severe MOF, ICU admission and ED discharge with adjusted ORs of 3.30 (95% CI 2.13–5.11), 1.75 (95% CI 1.11–2.77) and 0.46 (95% CI 0.32–0.68), respectively.
Conclusion
Bio-ADM in sepsis patients in the ED is associated with mortality, severe MOF, ICU admission and ED discharge, and may be of clinical importance for triage of sepsis patients in the ED.
Introduction
Background
Sepsis is a life-threatening condition which comes in a variety of shapes and severities, affecting millions of people worldwide [1]. In spite of improvements in recent years, the mortality of the most severe form of sepsis, septic shock, is still unacceptably high, up to 38% in North America and Europe [2].
The success rate of treating sepsis is time-sensitive, a short time to recognition and treatment is fundamental for outcomes, exemplified by the recommendation to consider one-hour bundles [3].
Identification of patients with sepsis in the emergency department (ED) is difficult and triaging patients to the correct level of care is a challenge. Biomarkers may be of help in identifying and stratifying sepsis according to severity of disease. An optimal biomarker in the ED setting should thus offer a method to distinguish individuals who can return home from those at high risk of developing multiple organ failure (MOF), thereby guiding clinicians to ensure patients an adequate level of care.
Adrenomedullin
Adrenomedullin (ADM) is a hormone produced by a variety of different cell types and was first derived from pheochromocytoma nearly three decades ago [4]. ADM has homeostatic and regulating effects on renal, immunological, endocrine and cardiovascular systems [4–7]. The effects of ADM on blood vessels include vasodilation [8] and stabilization of the barrier function of endothelial cells maintaining adequate permeability [9, 10]. ADM is typically elevated in patients with the metabolic syndrome [11], heart failure [12–15], chronic kidney failure [16–18] as well as in unselected critically ill patients [19, 20].
There are two predominant methods to measure ADM in peripheral blood. One is based on a part of the pre-cursor pro-hormone of ADM–mid regional pro-adrenomedullin (MR-proADM) [21], while the other measures bioactive adrenomedullin (bio-ADM) directly [22]. Few studies have described the correlation between measured MR-proADM and bio-ADM [22–24]. Although MR-proADM shows prognostic value in disease, it has no known action by itself, which makes the measurement of bio-ADM more attractive and clinically relevant. A median bio-ADM concentration of 20.7 pg/mL with 43 pg/mL as the 99th percentile among 200 healthy subjects has been reported [22].
Adrenomedullin in sepsis
Several studies have reported a strong association between elevated ADM levels and mortality, severity of illness and need for organ support in sepsis patients, using either of the two methods [19, 22, 25–31], proposing ADM to be a predictive biomarker in sepsis. Our group has recently reported that bio-ADM may be a specific marker of sepsis in a general intensive care unit (ICU) population [19].
In addition, the potential of modulating the ADM hormonal system has gained interest since exogenous infusion of ADM in animal models of sepsis has been shown to improve outcomes [10, 32, 33], which has led to the hypothesis that an increment of intravascular bio-ADM may be of therapeutic value in sepsis [9]. This has led to studies of the non-neutralizing anti-ADM antibody Adrecizumab in humans [23, 24, 34]. The formation of Adrecizumab-ADM complexes generates elevated intravascular bio-ADM concentrations where ADM can exert its endothelium-stabilizing effects [9, 35]. The increase of bio-ADM, on the other hand, is not accompanied by an elevation of MR-proADM suggesting a redistribution of ADM rather than an increased synthesis [23, 24]. The clinical implication of the use of Adrecizumab in sepsis is yet unanswered, but clinical trials to investigate this are planned [36].
While most of the studies describing ADM in sepsis are derived from ICU settings, similar findings have been found in populations originating in the ED. Studies performed on infected patients in the ED have reported MR-proADM to have a higher association with mortality and ICU admission compared to other commonly used biomarkers and clinical scores [37–39]. Further, a combination of MR-proADM with clinical scores and other biomarkers in order to improve prognostic accuracy has also been proposed [40–43].
Studies measuring bio-ADM in the ED are sparse. Two recent studies have described bio-ADM in ED populations but patients presented with either acute heart failure or dyspnea [14, 44]. The original paper presenting bio-ADM [22], however, analyzed bio-ADM in patients with suspected sepsis in the ED. In the present study, our aim was to investigate the prognostic capability of bio-ADM in a large sepsis cohort in the ED.
Objectives
We hypothesized that increasing levels of bio-ADM in sepsis patients in the ED were associated with subsequent severity of sepsis and increased mortality.
The primary aim of this study was to investigate the association of bio-ADM with 28-day mortality. Secondary aims were to I) assess whether bio-ADM could improve the prognostic precision of a mortality prediction model, II) compare the prognostic properties of bio-ADM with other commonly used biomarkers, and III) investigate the association of bio-ADM with a) severe MOF, b) ICU admission among patients with no limitations of care and c) ED discharge.
Material and methods
Study design and setting
This single center prospective observational cohort study was performed in the ED of Skåne University Hospital in Malmö, Sweden. With a catchment population of 400000, the hospital has approximately 85000 emergency visits per year.
Both oral and written consent was obtained by the patients or by their next of kin after they had the opportunity to read and review a written description of the study design and purpose. If a patient at inclusion had a decreased level of consciousness, consent was obtained retrospectively. This consent procedure and the study as a whole, was approved by the Regional Ethical board in Lund (DNR 2013/635).
The STROBE guidelines were followed [45].
Participants
Between December 2013 and February 2015, patients 18 years or older, seeking care during office hours (Monday to Friday, 6 AM to 6 PM) in the ED, were screened for inclusion by trained research nurses. The inclusion criteria were based on the sepsis definition at the time [46]: suspected infection in addition to two or more systemic inflammatory response syndrome (SIRS) criteria. Inclusion criteria were: 1) a body temperature lower than 36°C, or higher than 38°C, or self-reported fever/chills within 24 hours preceding the ED visit, 2) a respiratory rate higher than 20 breaths/min, 3) a heart rate higher than 90 beats/min. White blood cell count was not part of the inclusion criteria due to unavailability at the time of screening.
The study size was not predefined and consisted of a convenience sample of patients included during the study period.
Variables
The primary outcome was 28-day mortality. Secondary outcomes were number of failing organ systems, ICU admission and ED discharge. Failing organ systems, defined in S1 Table, were registered up to 48 hours after presentation at the ED and trichotomized into 1) no organ failure, 2) intermediate organ failure (one to three failing organ systems) and 3) severe MOF (four or more failing organ systems). ICU admission was registered during the entire follow-up time. Furthermore, a prognostic baseline model including covariates with significantly different distribution in relation to 28-day mortality, and three commonly used biomarkers, lactate, C-reactive protein (CRP) and creatinine was created to investigate whether the addition of bio-ADM improved the model. Premorbid comorbidities were registered and classified as shown in S2 Table.
Data sources
Patient demographics and comorbidities were systematically and prospectively collected from medical records which were reviewed by infectious disease physicians. Site of infection and type of ward, if admitted to the hospital, were recorded.
Biomarkers
Blood was drawn peripherally within one hour of presentation to the ED. All biomarkers except for bio-ADM were analyzed routinely in the certified hospital laboratory. For the analysis of bio-ADM plasma ethylenediaminetetraacetic acid plasma samples were frozen within 2 hours and stored at -80˚C until later batch analysis. Measurements of bio-ADM was undertaken at the laboratory of SphingoTec GmbH in Hennigsdorf, Germany in June 2018 as described elsewhere [47].
Statistics
For all hypotheses tests, we considered p-values <0.05 as significant. Group comparisons of continuous variables were performed using Wilcoxon rank-sum test (Mann-Whitney U test) for two groups. If there were more than two groups to be compared, Kruskal-Wallis rank sum test was used, and if significant, a comparison with pairwise Wilcoxon test, with Holm´s procedure for adjustment for multiple testing was performed. Differences in proportions were assessed using Pearson’s X2 test. Medians were reported with their corresponding interquartile ranges (IQR). Uni- and multivariable binary logistic regression was used to analyze outcomes. Covariates in the multivariable binary logistic regression analyses were included if they were significantly differently distributed in relation to the primary outcome. The results of the regression analyses were reported as odds ratios (OR) with 95% confidence intervals (CI). The regression models were evaluated with the Hosmer-Lemeshow goodness-of-fit test with ten groups, and only models resulting in non-significant tests were reported [48]. Body mass index (BMI) was stratified according to underweight (<18.5), normal (18.5–25), overweight (25–30) and obese (>35) prior to inclusion in the multivariable binary logistic regressions, with the normal group as reference. If a parameter, due to skewness, needed transformation, the base 2 logarithm was used. The difference in Kaplan-Meier curves was evaluated with the log-rank test [49]. Areas under the receiver operating characteristic curve (AUROC) were calculated [50]. Differences in AUROCs were tested with the method of DeLong et al [51]. Admissions with missing data were excluded from calculations. If a variable had missing values (MV) these were specified. R Studio version 1.2.1335 was used as statistical software.
Results
Participants
Inclusion criteria were met by 647 patients. Due to missing data 50 patients were excluded and bio-ADM was analyzed in 597 patients. Of these, three additional patients had missing mortality follow up data leaving 594 subjects to be included in the study, see Fig 1.
Fig 1. Patient flowchart according to inclusion eligibility, referral after assessment in the emergency department and 28-day mortality.
In total 53 patients were excluded due to missing plasma and missing outcome data as 28-day mortality, organ failure and ICU admission. ED: emergency department; bio-ADM: bioactive adrenomedullin; ICU: intensive care unit.
Demographics
Demographics including age, sex, comorbidities and site of infection are shown in Table 1. Non-survivors were generally older, had a lower BMI and a higher burden of cardiovascular disease. Further, non-survivors more often had a decision on limitation of care. The site of infection varied, non-survivors had a higher rate of pulmonary infections, whereas infections refrained to the upper respiratory tract and urinary sites were more common among survivors.
Table 1. Demographics and outcomes of the sepsis cohort and comparisons between 28-day non-survivors and survivors.
| Baseline characteristics | Sepsis cohort | Non-survivors | Survivors | p-value |
|---|---|---|---|---|
| Number, n (% of Sepsis cohort) | 594 (100) | 51 (8.6) | 543 (91.4) | |
| Age in years, median (IQR) | 73 (61–82) | 80 (73–88) | 72 (59–82) | <0.001 |
| Female sex, n (%) | 289 (48.6) | 22 (43.1) | 267 (49.2) | 0.50 |
| Body mass index (MV = 27), median (IQR) | 25.7 (22.5–29.6) | 24.0 (21.7–27.9) | 25.8 (22.6–30) | 0.05 |
| Comorbidities | ||||
| Cardiovascular disease (MV = 2), n (%) | 229 (38.7) | 316 (60.8) | 198 (36.6) | 0.001 |
| Respiratory disease (MV = 2), n (%) | 140 (23.6) | 18 (35.3) | 122 (22.6) | 0.06 |
| Neurological disease (MV = 1), n (%) | 98 (16.5) | 9 (17.6) | 89 (16.4) | 0.98 |
| Renal disease, n (%) | 45 (7.6) | 5 (9.8) | 40 (7.4) | 0.72 |
| Cancer (n = 591), n (%) | 165 (27.9) | 20 (40) | 145 (26.8) | 0.13 |
| Immunodeficiency (MV = 9), n (%) | 32 (5.5) | 6 (12) | 26 (4.9) | 0.07 |
| Diabetes (MV = 1), n (%) | 114 (19.2) | 15 (29.4) | 99 (18.2) | 0.08 |
| Psychiatric disorder (MV = 2), n (%) | 63 (10.6) | 4 (8) | 59 (10.9) | 0.69 |
| None of those listed above, n (%) | 146 (24.6) | 4 (7.8) | 142 (26.1) | 0.006 |
| Limitation of care (MV = 5), n (%) | 90 (15.3) | 24 (47.1) | 66 (12.3) | <0.001 |
| Site of infection | ||||
| Pulmonary, n (%) | 199 (33.5) | 24 (47.1) | 175 (32.2) | 0.02 |
| URTI, n (%) | 52 (8.8) | 0 (0) | 52 (9.6) | 0.05 |
| Urinary, n (%) | 129 (21.7) | 4 (7.8) | 125 (23) | 0.03 |
| Bone and joint, n (%) | 7 (1.2) | 1 (2.0) | 6 (1.1) | 1 |
| SSTI, n (%) | 58 (9.8) | 7 (13.7) | 51 (9.9) | 0.36 |
| Gastrointestinal, n (%) | 23 (3.9) | 1 (2.0) | 22 (4.1) | 0.78 |
| Other, n (%) | 76 (12.8) | 7 (13.7) | 69 (12.7) | 0.87 |
| No confirmed infection, n (%) | 50 (8.4) | 7 (13.7) | 43 (7.9) | 0.24 |
| Outcomes | ||||
| No organ failure, n (%) | 278 (46.8) | 8 (15.7) | 270 (49.7) | <0.001 |
| Intermediate organ failure (1–3), n (%) | 282 (47.5) | 30 (58.8) | 252 (46.4) | 0.12 |
| Severe MOF (≥4), n (%) | 34 (5,7) | 13 (25.4) | 21 (3.9) | <0.001 |
| ICU admission, n (%) | 27 (4.5) | 7 (13.7) | 20 (3.7) | 0.003 |
| Discharged from ED, n (%) | 67 (11.3) | 1 (2.0) | 66 (12.2) | 0.05 |
| Biomarkers | ||||
| Bio-ADM pg/mL, median (IQR) | 38 (27–60) | 63 (42–132) | 36 (26–56) | <0.001 |
| Lactate (MV = 25) mmol/L, median (IQR) | 1.7 (1.3–2.7) | 2.1 (1.3–3.0) | 1.7 (1.2–2.6) | 0.11 |
| CRP (MV = 7) mg/L, median (IQR) | 72 (25–160) | 100 (51–178) | 69 (23–156) | 0.04 |
| Creatinine (MV = 5) μmol/L, median (IQR) | 87 (68–120) | 105 (79–160) | 85 (68–117) | 0.006 |
Data regarding general characteristics, comorbidities, site of infection, outcomes and biomarkers are presented. Non-survivors were compared to survivors, and the p-values refer to that comparison. Proportions (%) are within their subgroups unless otherwise specified. IQR: interquartile range; MV: missing values, URTI: upper respiratory tract infection; SSTI: skin and soft tissue infection; ED: emergency department; MOF: multiple organ failure; ICU: intensive care unit; bio-ADM: bioactive adrenomedullin; CRP: C-reactive protein
Outcomes
Fifty-one patients (8.6%) died within 28 days, of whom 25 patients (4.2%) died within 7 days. Among 316 patients who developed organ failure (53.2%), 34 patients (5.7%) developed severe MOF as shown in Table 1. Twenty-seven patients (4.5%) were admitted to the ICU. Just over every tenth patient (11.3%) was discharged directly from the ED. One of them, the only 28-day non-survivor in the group, was offered admission to the ICU but declined and was discharged to palliative care at home after discussion with the patient and the patient´s family.
Bio-ADM
Levels of bio-ADM ranged 8–813 pg/mL and were logarithmically transformed due to skewness.
Bio-ADM and mortality
Non-survivors had higher levels of bio-ADM than survivors, 63 (42–132) pg/mL versus 36 (26–56) pg/mL, see Table 1. Dividing the patients into quartiles based on levels of bio-ADM a significant separation between the corresponding Kaplan-Meier curves for 28-day mortality, was observed, see Fig 2. The association of bio-ADM with 28-day mortality showed a univariate OR of 2.30 (95% CI 1.74–3.04), which remained significant after adjustments, 2.39 (95% CI 1.69–3.39), see Table 2.
Fig 2. Kaplan-Meier curve according to quartiles of bio-ADM and 28-day mortality.
The range of bio-ADM (pg/mL) was for Quartile 1: <27; Quartile 2: 27–38; Quartile 3: 38–60; Quartile 4: >60. The p-value was derived from the log-rank test. bio-ADM: bioactive adrenomedullin.
Table 2. Odds ratios for bio-ADM from uni- and multivariate binary logistic regression analyses for primary and secondary outcomes.
| Univariate | Multivariate | ||||||
|---|---|---|---|---|---|---|---|
| Primary outcome | OR | 95% CI | p-value | Primary outcome | OR | 95% CI | p-value |
| 28-day mortality | 2.30 | 1.74–3.04 | <0.001 | 28-day mortality (MV = 29) | 2.39 | 1.69–3.39 | <0.001 |
| Secondary outcome | OR | 95% CI | p-value | Secondary outcome | OR | 95% CI | p-value |
| Severe MOF | 3.22 | 2.26–4.59 | <0.001 | Severe MOF (MV = 29) | 3.30 | 2.13–5.11 | <0.001 |
| ICU admission (MV = 5) | 2.21 | 1.50–3.24 | <0.001 | ICU admission (MV = 27) | 1.75 | 1.11–2.77 | 0.02 |
| ED discharge | 0.41 | 0.29–0.56 | <0.001 | ED discharge (MV = 29) | 0.46 | 0.32–0.68 | <0.001 |
The odds ratio for bio-ADM was calculated on a base 2 logarithmic scale. Multivariate included covariates bio-ADM, age, known cardiovascular disease, BMI, urinary, URTI and pulmonary site of infection. The outcome ICU admission was only calculated among patients with no limitations of care (n = 499). bio-ADM: bioactive adrenomedullin; BMI: body mass index; ICU: intensive care unit; URTI: upper respiratory tract infection; OR: odds ratio; CI: confidence interval; MV: missing values; ED: emergency department.
A baseline mortality prediction model including age, previous cardiovascular disease, BMI, URTI, urinary or pulmonary infection site and routine biomarkers (CRP, lactate, creatinine) resulted in an AUROC of 0.80, which significantly improved with the addition of bio-ADM to an AUROC of 0.86 (p = 0.02), see Fig 3.
Fig 3. Receiver operating characteristics curves for mortality predictive models.
Baseline model with covariates age, known cardiovascular, BMI, URTI, urinary and pulmonary site of infection, C-reactive protein, lactate and creatinine. The additive value of bio-ADM is shown in Baseline + bio-ADM. The p-value is derived from the DeLong’s test for comparison between the two AUROCs. BMI: body mass index; URTI: upper respiratory tract infections; bio-ADM: bioactive adrenomedullin; AUROC: area under the receiver operating characteristic; CI: confidence interval.
Bio-ADM and other biomarkers
The receiver operating characteristics curves with corresponding AUROCs for lactate, CRP, creatinine and bio-ADM in relation to 28-day mortality are shown in Fig 4. Bio-ADM had a significantly higher AUROC than lactate, CRP and creatinine.
Fig 4. Receiver operating characteristics curves for the biomarkers bio-ADM, lactate, CRP and creatinine corresponding to 28 day mortality.
Only patients with all four biomarkers analyzed were included (n = 562). P-values are derived from the DeLong’s test for comparison with the AUROC of bio-ADM. bio-ADM: bioactive adrenomedullin; CRP: C-reactive protein; AUROC: area under the receiver operating characteristic curve.
Bio-ADM and organ failure
Bio-ADM concentrations among patients without organ failure, 31 (21–44) pg/mL, intermediate organ failure, 45 (31–72) pg/mL, and severe MOF, 81 (56–156) pg/mL, are shown in Fig 5. A significant separation between the groups was seen (p<0.001).
Fig 5. Boxplots showing levels of bio-ADM according to number of failing organ systems.
P-values are derived from the pairwise Wilcoxon test. **: p<0.001 bio-ADM: bioactive adrenomedullin.
ORs from uni- and multivariate regressions for bio-ADM for the development of severe MOF were 3.22 (95% CI 2.26–4.59) and 3.30 (95% CI 2.13–5.11), respectively, see Table 2.
Bio-ADM and ICU admission
Patients admitted to the ICU had significantly higher levels of bio-ADM, 77 (42–133) pg/mL, than patients not admitted to the ICU, 41 (28–61) pg/mL, and patients discharged from the ED, 26 (19–32) pg/mL (p<0.001). Fig 6 shows the distribution of bio-ADM according to patient referral after assessment in the ED. The distribution was significantly separated between the groups (p<0.001). There was a significant association between ICU admission and increasing levels of bio-ADM, both before and after adjustment, see Table 2.
Fig 6. Boxplots showing levels of bio-ADM according to patient referral after assessment in the emergency department.
P-values are derived from the pairwise Wilcoxon test. *: p<0.05, **: p<0.001. bio-ADM: bioactive adrenomedullin; HDU: high dependency unit; ICU: intensive care unit.
Bio-ADM and ED discharge
The median bio-ADM among patients discharged from ED was 26 (19–32) pg/mL, significantly lower than the corresponding median of 41 (28–63) pg/mL among patients admitted to a hospital ward or admitted to the ICU, 73 (41–130) pg/mL (p<0.001).
Uni- and multivariate logistic regression analyses showed an inverse association of increasing levels of bio-ADM and ED discharge, see Table 2.
Discussion
To our knowledge, this is the largest study to date investigating bio-ADM as a prognostic biomarker in patients with sepsis in the ED. Our data show that high levels of bio-ADM in the ED are associated with mortality, development of severe MOF and referral to intensive care. Moreover, we found that bio-ADM adds important prognostic information to the commonly used prognostic factors age, comorbidities, site of infection and routine biomarkers, and that low levels of bio-ADM are related to less severe disease and discharge from the ED.
Our study suggests that bio-ADM is of potential clinical use for early stratification of unselected sepsis patients in the ED. Alongside with the first study describing bio-ADM [22] and recent reports on possible applications of bio-ADM in patients with dyspnea [44] as well as heart failure [14], our data show that bio-ADM is a potentially important clinical biomarker in the ED. Whether these results are generalizable to a broader unselected ED population remains unknown and needs to be addressed in future studies. However, reports where MR-proADM was measured in broader ED populations show promising results [42, 52].
We found a strong association between bio-ADM in the ED and mortality, which remained after adjustments for known prognostic factors. Similar findings have been described in previous studies for both septic [19, 29–31] and non-septic [19, 20] patients treated in the ICU, but not as clearly among septic patients in the ED [22]. The prognostic ability of bio-ADM to predict mortality by itself was modest in the present study, but superior to three commonly used biomarkers, lactate, CRP and creatinine. Importantly, a baseline prediction model was improved when bio-ADM was added, indicating strong additional prognostic properties for bio-ADM. Our findings resemble results from a study in a similar setting where ADM was analyzed using the MR-proADM method. In that study, Scheutz et al. reported an improvement of a predictive model with an increased AUROC from 0.79 to 0.84, when MR-proADM was added [52].
The highest levels of bio-ADM in our study were found among patients admitted who developed severe MOF. Rising levels of bio-ADM were associated with increasing number of failing organ systems in sepsis patients. These results are in line with previous findings that septic patients with high levels of bio-ADM in the ICU had an increased need of organ support [19, 20, 29–31].
Interestingly, in the present ED cohort the median bio-ADM of 73 (41–130) pg/mL in the group of patients admitted to the ICU was similar to the distribution of bio-ADM in an ICU sepsis population where the median bio-ADM was 74 (42–145) pg/mL [19]. This is the first report to describe that bio-ADM is predictive of ICU admission in a sepsis cohort in the ED, which is a novel finding. Due to known variations in the availability of ICU beds across countries, this may however not be generalizable to other hospital environments [53].
The patients discharged from the ED in our cohort had low levels of bio-ADM with levels close to those in healthy subjects [22]. There were some extreme outliers within the group, making a clear threshold of bio-ADM difficult to identify. To our knowledge, no previous study has reported levels of bio-ADM in patients with sepsis discharged from ED.
Strengths and limitations
This large prospective observational cohort study affirms previous findings from ICU settings and demonstrates the potential applicability of bio-ADM in the ED setting. Furthermore, all patient records in this study were thoroughly revised by infectious disease physicians to assure correct diagnoses. Also, this study included patients with limitations of care.
This study has several limitations. First, we only enrolled participants during office hours which may have led to a selection bias. Second, we were confined to admission samples, making it impossible to analyze dynamic changes and how these could correlate with outcomes. Third, this was a single-center study why generalizability of our results to other hospital settings may be limited. Finally, the study was initiated when sepsis was defined by the Sepsis-2 criteria and thus SOFA score was not recorded.
Conclusions
Bio-ADM in sepsis patients in the ED is associated with mortality, MOF, ICU admission and ED discharge. Bio-ADM exceeds the prognostic properties of routine biomarkers as lactate, CRP and creatinine and may be of clinical importance for triage of sepsis patients in the ED.
Supporting information
(DOCX)
(DOCX)
Acknowledgments
We would like to thank all the staff of the ED, Skåne University Hospital in Malmö, without whom this study would not have been possible. Further, we are grateful to SphingoTec GmbH who, with no influence on the interpretation of the results, performed the analysis of bio-ADM free of charge.
Data Availability
Due to ethical and legal restrictions related to the Swedish Biobanks in Medical Care Act (2002:297) and the Personal Data Act (1998:204) regarding deposition of data, data are available upon request via the authors. Lund University, represented by the authors, is the authority obliged to follow Swedish legislation and can be contacted at registrator@lu.se.
Funding Statement
OL, HF: Regional Research Support in Region Skåne (https://sodrasjukvardsregionen.se/) MR, HF: Governmental funding of Clinical Research within the Swedish National Health Service (ALF) (https://www.med.lu.se/intramed/styrning_organisation/ekonomi_alf/alf) OM: Swedish Heart- and Lung Foundation, grant nr: 20180278 (https://www.hjart-lungfonden.se/); Swedish Research Council, grant nr: 2018–02760 (https://www.vr.se). JS: employed by SphingoTec GmbH, the manufacturer of the bio-ADM assay. (https://sphingotec.com/) The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
References
- 1.World Health Organisation Improving the prevention, diagnosis and clinical management of sepsis 2017, April 13. Available from: http://apps.who.int/gb/ebwha/pdf_files/WHA70/A70_13-en.pdf?ua=1.
- 2.Vincent JL, Jones G, David S, Olariu E, Cadwell KK. Frequency and mortality of septic shock in Europe and North America: a systematic review and meta-analysis. Crit Care. 2019;23(1):196. doi: 10.1186/s13054-019-2478-6 ; PubMed Central PMCID: PMC6545004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Levy MM, Evans LE, Rhodes A. The Surviving Sepsis Campaign Bundle: 2018 Update. Crit Care Med. 2018;46(6):997–1000. doi: 10.1097/CCM.0000000000003119 . [DOI] [PubMed] [Google Scholar]
- 4.Kitamura K, Kangawa K, Kawamoto M, Ichiki Y, Nakamura S, Matsuo H, et al. Adrenomedullin: a novel hypotensive peptide isolated from human pheochromocytoma. Biochem Biophys Res Commun. 1993;192(2):553–60. doi: 10.1006/bbrc.1993.1451 . [DOI] [PubMed] [Google Scholar]
- 5.Zudaire E, Portal-Nunez S, Cuttitta F. The central role of adrenomedullin in host defense. Journal of leukocyte biology. 2006;80(2):237–44. Epub 2006/06/14. doi: 10.1189/jlb.0206123 . [DOI] [PubMed] [Google Scholar]
- 6.Samson WK, Resch ZT, Murphy TC, Vargas TT, Schell DA. Adrenomedullin: Is There Physiological Relevance in the Pathology and Pharmacology? News Physiol Sci. 1999;14:255–9. Epub 2001/06/08. doi: 10.1152/physiologyonline.1999.14.6.255 . [DOI] [PubMed] [Google Scholar]
- 7.Hinson JP, Kapas S, Smith DM. Adrenomedullin, a multifunctional regulatory peptide. Endocr Rev. 2000;21(2):138–67. doi: 10.1210/edrv.21.2.0396 . [DOI] [PubMed] [Google Scholar]
- 8.Kato J, Kitamura K. Bench-to-bedside pharmacology of adrenomedullin. Eur J Pharmacol. 2015;764:140–8. doi: 10.1016/j.ejphar.2015.06.061 . [DOI] [PubMed] [Google Scholar]
- 9.Geven C, Bergmann A, Kox M, Pickkers P. Vascular Effects of Adrenomedullin and the Anti-Adrenomedullin Antibody Adrecizumab in Sepsis. Shock. 2018;50(2):132–40. doi: 10.1097/SHK.0000000000001103 . [DOI] [PubMed] [Google Scholar]
- 10.Hippenstiel S, Witzenrath M, Schmeck B, Hocke A, Krisp M, Krull M, et al. Adrenomedullin reduces endothelial hyperpermeability. Circ Res. 2002;91(7):618–25. doi: 10.1161/01.res.0000036603.61868.f9 . [DOI] [PubMed] [Google Scholar]
- 11.Seissler J, Feghelm N, Then C, Meisinger C, Herder C, Koenig W, et al. Vasoregulatory peptides pro-endothelin-1 and pro-adrenomedullin are associated with metabolic syndrome in the population-based KORA F4 study. Eur J Endocrinol. 2012;167(6):847–53. doi: 10.1530/EJE-12-0472 . [DOI] [PubMed] [Google Scholar]
- 12.Ter Maaten JM, Kremer D, Demissei BG, Struck J, Bergmann A, Anker SD, et al. Bio-adrenomedullin as a marker of congestion in patients with new-onset and worsening heart failure. Eur J Heart Fail. 2019;21(6):732–43. doi: 10.1002/ejhf.1437 . [DOI] [PubMed] [Google Scholar]
- 13.Self WH, Storrow AB, Hartmann O, Barrett TW, Fermann GJ, Maisel AS, et al. Plasma bioactive adrenomedullin as a prognostic biomarker in acute heart failure. Am J Emerg Med. 2016;34(2):257–62. doi: 10.1016/j.ajem.2015.10.033 ; PubMed Central PMCID: PMC4753125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Molvin J, Jujic A, Navarin S, Melander O, Zoccoli G, Hartmann O, et al. Bioactive adrenomedullin, proenkephalin A and clinical outcomes in an acute heart failure setting. Open Heart. 2019;6(2):e001048. doi: 10.1136/openhrt-2019-001048 ; PubMed Central PMCID: PMC6615850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kremer D, Ter Maaten JM, Voors AA. Bio-adrenomedullin as a potential quick, reliable, and objective marker of congestion in heart failure. Eur J Heart Fail. 2018;20(9):1363–5. doi: 10.1002/ejhf.1245 . [DOI] [PubMed] [Google Scholar]
- 16.Sogbe-Diaz ME, Diaz-Lopez EE. [Adrenomedullin in the kidney: physiology and pathophysiology]. Invest Clin. 2016;57(1):66–76. . [PubMed] [Google Scholar]
- 17.Nishikimi T. Adrenomedullin in the kidney-renal physiological and pathophysiological roles. Curr Med Chem. 2007;14(15):1689–99. doi: 10.2174/092986707780830943 . [DOI] [PubMed] [Google Scholar]
- 18.McGregor DO, Troughton RW, Frampton C, Lynn KL, Yandle T, Richards AM, et al. Hypotensive and natriuretic actions of adrenomedullin in subjects with chronic renal impairment. Hypertension. 2001;37(5):1279–84. doi: 10.1161/01.hyp.37.5.1279 . [DOI] [PubMed] [Google Scholar]
- 19.Lundberg OHM, Lengquist M, Spangfors M, Annborn M, Bergmann D, Schulte J, et al. Circulating bioactive adrenomedullin as a marker of sepsis, septic shock and critical illness. Crit Care. 2020;24(1):636. doi: 10.1186/s13054-020-03351-1 ; PubMed Central PMCID: PMC7641835. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lemasle L, Blet A, Geven C, Cherifa M, Deniau B, Hollinger A, et al. Bioactive Adrenomedullin, Organ Support Therapies, and Survival in the Critically Ill: Results from the French and European Outcome Registry in ICU Study. Crit Care Med. 2020;48(1):49–55. doi: 10.1097/CCM.0000000000004044 . [DOI] [PubMed] [Google Scholar]
- 21.Morgenthaler NG, Struck J, Alonso C, Bergmann A. Measurement of midregional proadrenomedullin in plasma with an immunoluminometric assay. Clin Chem. 2005;51(10):1823–9. Epub 2005/08/16. clinchem.2005.051110 [pii] doi: 10.1373/clinchem.2005.051110 . [DOI] [PubMed] [Google Scholar]
- 22.Marino R, Struck J, Maisel AS, Magrini L, Bergmann A, Di Somma S. Plasma adrenomedullin is associated with short-term mortality and vasopressor requirement in patients admitted with sepsis. Crit Care. 2014;18(1):R34. Epub 2014/02/19. doi: 10.1186/cc13731 . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Geven C, van Lier D, Blet A, Peelen R, Ten Elzen B, Mebazaa A, et al. Safety, tolerability and pharmacokinetics/pharmacodynamics of the adrenomedullin antibody adrecizumab in a first-in-human study and during experimental human endotoxaemia in healthy subjects. Br J Clin Pharmacol. 2018;84(9):2129–41. doi: 10.1111/bcp.13655 ; PubMed Central PMCID: PMC6089825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Laterre PF, Pickkers P, Marx G, Wittebole X, Meziani F, Dugernier T, et al. Safety and tolerability of non-neutralizing adrenomedullin antibody adrecizumab (HAM8101) in septic shock patients: the AdrenOSS-2 phase 2a biomarker-guided trial. Intensive Care Med. 2021. doi: 10.1007/s00134-021-06537-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Christ-Crain M, Morgenthaler NG, Struck J, Harbarth S, Bergmann A, Muller B. Mid-regional pro-adrenomedullin as a prognostic marker in sepsis: an observational study. Crit Care. 2005;9(6):R816–24. Epub 2005/12/17. doi: 10.1186/cc3885 [pii] ; PubMed Central PMCID: PMC1414007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Schuetz P, Christ-Crain M, Morgenthaler NG, Struck J, Bergmann A, Muller B. Circulating precursor levels of endothelin-1 and adrenomedullin, two endothelium-derived, counteracting substances, in sepsis. Endothelium. 2007;14(6):345–51. Epub 2007/12/18. 788541314 [pii] doi: 10.1080/10623320701678326 ; PubMed Central PMCID: PMC2430170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Guignant C, Voirin N, Venet F, Poitevin F, Malcus C, Bohe J, et al. Assessment of pro-vasopressin and pro-adrenomedullin as predictors of 28-day mortality in septic shock patients. Intensive Care Med. 2009;35(11):1859–67. Epub 2009/08/08. doi: 10.1007/s00134-009-1610-5 . [DOI] [PubMed] [Google Scholar]
- 28.Lundberg OH, Bergenzaun L, Ryden J, Rosenqvist M, Melander O, Chew MS. Adrenomedullin and endothelin-1 are associated with myocardial injury and death in septic shock patients. Crit Care. 2016;20(1):178. doi: 10.1186/s13054-016-1361-y ; PubMed Central PMCID: PMC4899903. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Caironi P, Latini R, Struck J, Hartmann O, Bergmann A, Maggio G, et al. Circulating Biologically Active Adrenomedullin (bio-ADM) Predicts Hemodynamic Support Requirement and Mortality During Sepsis. Chest. 2017;152(2):312–20. doi: 10.1016/j.chest.2017.03.035 . [DOI] [PubMed] [Google Scholar]
- 30.Mebazaa A, Geven C, Hollinger A, Wittebole X, Chousterman BG, Blet A, et al. Circulating adrenomedullin estimates survival and reversibility of organ failure in sepsis: the prospective observational multinational Adrenomedullin and Outcome in Sepsis and Septic Shock-1 (AdrenOSS-1) study. Crit Care. 2018;22(1):354. doi: 10.1186/s13054-018-2243-2 ; PubMed Central PMCID: PMC6305573. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kim H, Hur M, Struck J, Bergmann A, Di Somma S. Circulating Biologically Active Adrenomedullin Predicts Organ Failure and Mortality in Sepsis. Ann Lab Med. 2019;39(5):454–63. doi: 10.3343/alm.2019.39.5.454 ; PubMed Central PMCID: PMC6502946. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Ertmer C, Morelli A, Rehberg S, Lange M, Hucklenbruch C, Van Aken H, et al. Exogenous adrenomedullin prevents and reverses hypodynamic circulation and pulmonary hypertension in ovine endotoxaemia. Br J Anaesth. 2007;99(6):830–6. Epub 2007/10/27. aem295 [pii] doi: 10.1093/bja/aem295 . [DOI] [PubMed] [Google Scholar]
- 33.Temmesfeld-Wollbruck B, Brell B, David I, Dorenberg M, Adolphs J, Schmeck B, et al. Adrenomedullin reduces vascular hyperpermeability and improves survival in rat septic shock. Intensive Care Med. 2007;33(4):703–10. Epub 2007/02/24. doi: 10.1007/s00134-007-0561-y . [DOI] [PubMed] [Google Scholar]
- 34.Karakas M, Jarczak D, Becker M, Roedl K, Addo MM, Hein F, et al. Targeting Endothelial Dysfunction in Eight Extreme-Critically Ill Patients with COVID-19 Using the Anti-Adrenomedullin Antibody Adrecizumab (HAM8101). Biomolecules. 2020;10(8). doi: 10.3390/biom10081171 ; PubMed Central PMCID: PMC7465983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Voors AA, Kremer D, Geven C, Ter Maaten JM, Struck J, Bergmann A, et al. Adrenomedullin in heart failure: pathophysiology and therapeutic application. Eur J Heart Fail. 2019;21(2):163–71. doi: 10.1002/ejhf.1366 ; PubMed Central PMCID: PMC6607488. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Laterre PF, Mebazaa A, editors. Safety and Tolerability of non-neutralizing Adrenomedullin antibody Adrecizumab (HAM8101) in septic shock patients The AdrenOSS-2 phase 2a biomarker-guided trial. ESICM LIVES; 2021. 05/10; Virtual. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Saeed K, Wilson DC, Bloos F, Schuetz P, van der Does Y, Melander O, et al. The early identification of disease progression in patients with suspected infection presenting to the emergency department: a multi-centre derivation and validation study. Crit Care. 2019;23(1):40. doi: 10.1186/s13054-019-2329-5 ; PubMed Central PMCID: PMC6368690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Gonzalez Del Castillo J, Wilson DC, Clemente-Callejo C, Roman F, Bardes-Robles I, Jimenez I, et al. Biomarkers and clinical scores to identify patient populations at risk of delayed antibiotic administration or intensive care admission. Crit Care. 2019;23(1):335. doi: 10.1186/s13054-019-2613-4 ; PubMed Central PMCID: PMC6819475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Haag E, Gregoriano C, Molitor A, Kloter M, Kutz A, Mueller B, et al. Does mid-regional pro-adrenomedullin (MR-proADM) improve the sequential organ failure assessment-score (SOFA score) for mortality-prediction in patients with acute infections? Results of a prospective observational study. Clin Chem Lab Med. 2021;59(6):1165–76. doi: 10.1515/cclm-2020-1566 . [DOI] [PubMed] [Google Scholar]
- 40.Mearelli F, Barbati G, Casarsa C, Giansante C, Breglia A, Spica A, et al. The Integration of qSOFA with Clinical Variables and Serum Biomarkers Improves the Prognostic Value of qSOFA Alone in Patients with Suspected or Confirmed Sepsis at ED Admission. J Clin Med. 2020;9(4). doi: 10.3390/jcm9041205 ; PubMed Central PMCID: PMC7230329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Travaglino F, De Berardinis B, Magrini L, Bongiovanni C, Candelli M, Silveri NG, et al. Utility of Procalcitonin (PCT) and Mid regional pro-Adrenomedullin (MR-proADM) in risk stratification of critically ill febrile patients in Emergency Department (ED). A comparison with APACHE II score. BMC Infect Dis. 2012;12:184. doi: 10.1186/1471-2334-12-184 ; PubMed Central PMCID: PMC3447640. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Graziadio S, O’Leary RA, Stocken DD, Power M, Allen AJ, Simpson AJ, et al. Can mid-regional pro-adrenomedullin (MR-proADM) increase the prognostic accuracy of NEWS in predicting deterioration in patients admitted to hospital with mild to moderately severe illness? A prospective single-centre observational study. BMJ Open. 2019;8(11):e020337. doi: 10.1136/bmjopen-2017-020337 ; PubMed Central PMCID: PMC6278796. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Saeed K, Legramante JM, Angeletti S, Curcio F, Miguens I, Poole S, et al. Mid-regional pro-adrenomedullin as a supplementary tool to clinical parameters in cases of suspicion of infection in the emergency department. Expert Rev Mol Diagn. 2021;21(4):397–404. doi: 10.1080/14737159.2021.1902312 . [DOI] [PubMed] [Google Scholar]
- 44.Bronton K, Wessman T, Gransbo K, Schulte J, Hartmann O, Melander O. Bioactive adrenomedullin a prognostic biomarker in patients with mild to moderate dyspnea at the emergency department: an observational study. Intern Emerg Med. 2021. doi: 10.1007/s11739-021-02776-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453–7. doi: 10.1016/S0140-6736(07)61602-X . [DOI] [PubMed] [Google Scholar]
- 46.Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med. 2013;41(2):580–637. Epub 2013/01/29. doi: 10.1097/CCM.0b013e31827e83af . [DOI] [PubMed] [Google Scholar]
- 47.Weber J, Sachse J, Bergmann S, Sparwasser A, Struck J, Bergmann A. Sandwich Immunoassay for Bioactive Plasma Adrenomedullin. J Appl Lab Med. 2017;2(2):222–33. doi: 10.1373/jalm.2017.023655 . [DOI] [PubMed] [Google Scholar]
- 48.Assessing the Fit of the Model. Applied Logistic Regression2000. p. 143–202.
- 49.Harrington DP, Fleming TR. A class of rank test procedures for censored survival data. Biometrika. 1982;69(3):553–66%@ 0006–3444. doi: 10.1093/biomet/69.3.553 [DOI] [Google Scholar]
- 50.Tom F. An introduction to ROC analysis. Pattern Recognition Letters. 2006;27(8):861–74. 10.1016/j.patrec.2005.10.010. FAWCETT2006861. [DOI] [Google Scholar]
- 51.DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44(3):837–45. . [PubMed] [Google Scholar]
- 52.Schuetz P, Hausfater P, Amin D, Amin A, Haubitz S, Faessler L, et al. Biomarkers from distinct biological pathways improve early risk stratification in medical emergency patients: the multinational, prospective, observational TRIAGE study. Crit Care. 2015;19:377. doi: 10.1186/s13054-015-1098-z ; PubMed Central PMCID: PMC4625457. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Rhodes A, Ferdinande P, Flaatten H, Guidet B, Metnitz PG, Moreno RP. The variability of critical care bed numbers in Europe. Intensive Care Med. 2012;38(10):1647–53. doi: 10.1007/s00134-012-2627-8 . [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX)
(DOCX)
Data Availability Statement
Due to ethical and legal restrictions related to the Swedish Biobanks in Medical Care Act (2002:297) and the Personal Data Act (1998:204) regarding deposition of data, data are available upon request via the authors. Lund University, represented by the authors, is the authority obliged to follow Swedish legislation and can be contacted at registrator@lu.se.