Prognostic Value of Secretoneurin in Patients with Acute Respiratory Failure: Data from the FINNALI Study.
Secretoneurin (SN) is a 33-amino acid peptide belonging to the chromogranin-secretogranin (granin) protein family (2). We have previously demonstrated the potential of the granin proteins as prognostic biomarkers in patients with cardiovascular (CV) disease (3-6), including a recent study that identified SN as a potent prognostic biomarker in patients with acute heart failure (7). We have also found circulating SN concentrations to be high in patients with myocardial dysfunction and that the patients with the highest SN concentrations had the worst prognosis after ventricular arrhythmia-induced cardiac arrest (7). The potential of SN as a prognostic biomarker after cardiac arrest has been validated by other groups (8). More recently, we reported that SN adds to established risk indices in 2 separate cohorts of patients with infectious disease, including patients with infections in the FINNish Acute Lung Injury (FINNALI) Study (9). Hence, SN seems to complement established CV biomarkers, which is reflected in only moderate correlations between SN and N-terminal-pro-B-type natriuretic peptide (NT-proBNP) and high-sensitivity troponin T concentrations (7, 10), and no significant correlations between SN and norepinephrine concentrations (7, 11). The potent prognostic information from SN measurements could relate to the association between SN and cardiomyocyte [Ca.sup.2+] handling (7), which is pathophysiology not reflected by established risk markers (12). In addition, SN has also been found to influence additional myocardial pathophysiology (13). As circulating SN concentrations are also increased by renal dysfunction and correlate with cortisol concentrations in ICU patients (9), SN could represent an index of CV, renal, and neuroendocrine dysfunction, processes that are all expected to impact clinical outcomes in ARF patients and especially in CV-related ARF. Accordingly, in this study, we hypothesized that SN would provide additional prognostic information to established risk indices in a prospective cohort of critically ill patients with ARF.
Materials and Methods
This is a substudy of the FINNALI Study, a multicenter prospective, observational, epidemiological study that included all patients with ARF admitted to Finnish ICUs during 8 weeks in 2007. The main results of the FINNALI epidemiology study have been reported previously (14). In short, the FINNALI epidemiological study was designed to assess the quality of care for patients with ARF in Finland, and the recruiting ICUs covered 97% of the Finnish population (4.3 million inhabitants). In total, 2670 admissions in 25 Finnish ICUs were recorded and the FINNALI epidemiology study included adults (age [greater than or equal to] 16 years) who received any form of positive airway pressure (invasive or noninvasive ventilatory support) for more than 6h(n = 958; Fig. 1).No prespecified treatment protocol was provided to the sites. Accordingly, any treatment restrictions during the ICU stay were solely determined by the treating physicians, and these physicians had no knowledge of SN concentrations. Terminally ill patients or palliative patients were not included in the FINNALI epidemiology study. The Local Ethics Committees approved the study, and the ethics committees waived the need for informed consent for data registration for the FINNALI epidemiological study. In contrast, informed consent from the patient or a close relative was required before any blood sampling could be performed in the FINNALI laboratory substudy, which was designed to study established and novel biomarkers in ARF.
Of the 958 patients included in the total FINNALI epidemiological cohort, we obtained informed consent and subsequent blood samples on ICU admission from 584 patients (61%). We have recently published the results for SN in the patients with pneumonia and/or sepsis in the FINNALI Study (n = 94) (9); thus, these patients were excluded from the current study. Accordingly, 490 patients with ARF were included in the current substudy, and we also had blood samples from 401 patients after 48 h (82% of the cohort). We collected demographic and clinical data daily by an electronic daily case report form according to the routine data set of the Finnish ICU quality consortium (14), although additional information was not available owing to the multicenter design of the study. The attending physicians classified all ICU admissions according to the Acute Physiology and Chronic Health Evaluation (APACHE) III diagnosis. Based on this classification, we classified the patients as either CV- or non-CV-related ARF, which is analogous to the strategy previously used by other groups (15). Patients admitted after cardiac surgery and patients admitted with acute heart failure were classified together as CV-related ARF, whereas patients classified as non--CV-related ARF included the following etiologies: gastrointestinal, trauma, neurological, intoxication, respiratory, or miscellaneous. We prospectively calculated Simplified Acute Physiology Score (SAPS) II and Sequential Organ Failure Assessment (SOFA) scores after 24 h in all patients. In the FINNALI Study (14), the primary end point was all-cause mortality after 90 days, and the mortality data were obtained from Statistics Finland (www. stat.fi). For this substudy, we also have included 30-day and 12-month mortality as secondary end points.
Serum samples were obtained by venipuncture or from an indwelling arterial catheter and centrifuged at approximately 1500g for 15 min before long-term storage at -80[degrees]C. The plasma concentration of SN was measured by an in-house SN RIA, as previously described (10, 11, 16), and the measurements were finalized on August 25, 2010. We recently demonstrated that storage up to 8.5 years does not appear to influence SN concentrations when stored at -80 [degrees]C (9). The limit of detection for the SN RIA in plasma is 50 pmol/L, and the assay has a CV of 9% in the lower range (110 pmol/L) and 4% in the upper range (380 pmol/L). We measured NT-proBNP concentrations by using a commercially available assay (proBNP II, Roche Diagnostics) (17).
We present the data as medians [quartile (Q) 1-3] or as absolute numbers and percentages. Continuous variables were compared with the Mann-Whitney U-test for nonparametric data and paired by the Wilcoxon Signed Ranks Test. Categorical data were compared by the [chi square] test or the Fisher exact test, as appropriate. Correlations were calculated by Spearman rank correlation. We transformed white blood cell count (WBC), serum creatinine, SN, and NT-proBNP concentrations by the natural logarithm before regression analysis owing to a right-skewed distribution (assessed by the Kolmogorov-Smirnov test). Prognostic utility was visualized by Kaplan-Meier plots according to SN Qs and compared by the log-rank test. Follow-up time was calculated from ICU admission to time of death or censored after 90 days. We explored variables available on admission that were related to time to death in 90 days in CV- and non--CV-related ARF by univariate and multivariate Cox proportional hazards regression analysis. Variables included in the univariate models were age, sex, body mass index, smoking status, previous CV disease, diabetes mellitus, chronic obstructive pulmonary disease, surgery/nonsurgery, heart rate, mean arterial blood pressure, and NT-proBNP and SN concentrations. We included univariate predictors of mortality in the multivariate Cox regression models using a backward selection strategy (i.e., all variables were included into the model and removed one by one if P > 0.10). Prognostic accuracy was assessed by calculating the area under the curve (AUC) with 95% CIs. The AUCs are from nested models and presented as recently recommended (18). Reclassification was performed by calculating the category-free net reclassification index (NRI) (19) with the basic model built on the clinical variables that were significantly associated with mortality in multivariate Cox regression analysis. P values <0.05 were considered significant for all analyses. Statistical analyses were performed with SPSS[R] Statistics Version 21 for Windows, MedCalc Statistical Software version 14.10.2 (MedCalc for Windows), and the statistical programming language R (R Development Core Team, 2008).
The patients included in this SN substudy of the FINNALI Study (n = 490) were comparable to the patients included in the main epidemiological FINNALI Study (n = 958) in relation to age, sex, disease severity, and mortality during follow-up (see Table 1 in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol62/ issue10). In total, 209 patients (43%) were classified as hospitalized with CV-related ARF, and 281 patients (57%) were classified as hospitalized with non-CV-related ARF (Fig. 1). We found that several demographic and clinical variables differed between patients admitted with CV-related and non--CV-related ARF (Table 1). Only 19 patients (9%) hospitalized with CV-related ARF and 23 patients (8%) with non--CV-related ARF did not receive invasive ventilatory support at any time point during the admission. Among the 209 patients admitted with CV-related ARF, 113 patients (54%) were hospitalized with ARF after surgery and 96 patients (46%) were hospitalized with ARF after acute heart failure.
SN CONCENTRATIONS IN PATIENTS WITH ARF
We found no difference in admission SN concentrations between ICU patients with CV-related ARF and non-CV-related ARF: 120 (Q1-3 95-167) vs 120 (95-148) pmol/L, P = 0.61 (Table 1). SN concentrations on ICU admission in the total cohort (n = 490) correlated with age (r = 0.18, P < 0.001), respiratory rate (r = 0.16, P < 0.001), body mass index (r = 0.12, P = 0.006), and creatinine (r = 0.54, P < 0.001) and NT-proBNP concentrations (r = 0.34, P < 0.001) (see online Supplemental Table 2). As assessed by multivariate linear regression analysis, determinants of high SN concentration on ICU inclusion in CV-related ARF were nonoperative causes for admission (P < 0.001) and higher creatinine concentrations (P < 0.001), and these variables explained 42% of the variance in SN concentrations ([r.sup.2] = 0.42) (see online Supplemental Table 3A). In patients with non-CV-related ARF, only high creatinine concentrations (P < 0.001) were associated with increasing SN concentrations, and creatinine explained 45% of the variance in SN concentrations (see online Supplemental Table 3B).
ADMISSION SN CONCENTRATIONS AND MORTALITY IN CV- AND NON-CV-RELATED ARF
In total, 64 patients with CV-related ARF and 68 patients with non--CV-related ARF died during 90-day follow-up, and mortality rates were comparable between CV- vs non--CV-related ARF: 31% vs 24%, P = 0.11 (Table 1). Most deaths occurred during the index hospitalization [104 of 132 deaths (79%)]. Treatment was restricted in 61 patients (12%) of whom 32 patients were hospitalized with CV-related ARF (15%) and 29 patients with non-CV-related ARF (10%) (P = 0.11 between the groups). Among the 209 patients hospitalized for CV-related ARF, the mortality rate during 90-day follow-up for the patients hospitalized after surgery was 15% (17/ 113 patients), and the mortality rate for patients with acute heart failure was 49% (47/96 patients).
Baseline characteristics of 90-day survivors and nonsurvivors are presented in Table 2. Admission SN concentrations were higher in 90-day nonsurvivors compared to survivors for both patients with CV-related ARF [148 (117-203) vs 108 (87-143) pmol/L, P < 0.001] and non-CV-related ARF [139 (115-184) vs 113 (91-139) pmol/L, P < 0.001] (see Table 2 and online Supplemental Fig. 1). The associations between admission SN concentrations stratified into Qs and mortality during 90-day follow-up in CV-related and non-CV-related ARF are presented in Fig. 2 (P < 0.001 by the log-rank test for both). Among patients with CV-related ARF, 4 out of 51 patients (8%) with admission SN concentrations in the 1st Q died during 90-day follow-up compared to 25 out of 52 patients (48%) with admission SN concentrations in the 4th Q. Analogously, among patients with non-CV-related ARF, 9 out of 67 patients (13%) with 1st QSN concentrations died during 90-day follow-up compared to 29 out of 70 patients (41%) with 4th QSN concentrations. Admission SN concentrations were also associated with time to death during 90 days in univariate Cox regression analyses in both CV- and non-CV-related ARF (Table 3). Adjusting for the other variables associated with 90-day mortality in multivariate analyses, admission SN concentrations were associated with time to death during 90 days in CV-related ARF, whereas the association between SN concentrations and mortality in non-CV-related ARF was attenuated and no longer significant (Table 4).
Assessing prognostic value by ROC analyses, the AUC for SN to predict 90-day mortality in CV-related ARF patients was 0.72 (95% CI, 0.65-0.79) compared to AUC 0.64 (0.56-0.72) for NT-proBNP. In contrast, the AUC of SN to predict 90-day mortality in non-CV-related ARF patients was 0.67 (0.60-0.75) and the AUC of NT-proBNP was 0.77 (0.71-0.83). Assessing prognostic accuracy in patients with CV-related ARF after surgery separately (n = 113), the AUC of SN to predict 90-day mortality was 0.72 (0.63-0.80), also in this subgroup. As tested by the NRI, SN improved risk prediction in CV-related ARF on top of surgical admission (NRI = 0.32, 95% CI, 0.04-0.59, P = 0.03), which was also associated with time to death in multivariate Cox analysis (Table 4). SN downgraded the risk in patients with CV-related ARF who had a favorable outcome (see online Supplemental Fig. 2). The optimal admission SN concentration to predict time to death in 90 days for patients with CV-related ARF was 122 pmol/L [sensitivity 73% (61%-84%), specificity 63% (55%-71%), and positive likelihood ratio 2.01 (1.5-2.6)]. Admission SN concentrations over 122 pmol/L on ICU admission were associated with increased mortality after adjustment for SAPS II [hazard ratio (HR) 1.89 (1.06-3.39), P = 0.03] and SOFA score [HR 2.56 (1.41-4.64), P = 0.002] (see online Supplemental Table 4). The optimal cutoff of SN to predict 90-day mortality in non--CV-related ARF was 114 pmol/L [sensitivity 78% (66%-87%), specificity 51% (44%-58%), and positive likelihood ratio 1.60 (1.3-1.9)].
Mortality rates for 30-day follow-up were 29% (60/ 209 patients) in CV-related ARF and 19% (53/281 patients) in non-CV-related ARF. In patients with CV-related ARF, the AUC of admission SN concentrations to separate 30-day survivors from nonsurvivors was 0.73 (0.66-0.78), the AUC of admission NT-proBNP concentrations was 0.64 (0.57-0.71), and admission SN concentrations were associated with 30-day mortality in multivariate Cox analysis, which also included other risk indices (see online Supplemental Table 5). Corresponding ROC-AUC values for 30-day mortality in patients with non-CV-related ARF were 0.67 (0.61-0.72) for admission SN concentrations and 0.76 (0.71-0.81) for admission NT-proBNP concentrations.
During 12 months of follow-up, 74 patients (35%) with CV-related ARF died and 83 patients (30%) with non--CV-related ARF died. The ROC-AUC for 1-year mortality in CV-related ARF was 0.72 (0.66-0.78) for admission SN concentrations, 0.73 (0.67-0.78) for admission NT-proBNP concentrations, and admission SN concentrations were associated with 1-year mortality in multivariate Cox analysis (see online Supplemental Table 6). ROC-AUC values to separate 1-year survivors from nonsurvivors in non-CV-related ARF were 0.65 (0.59 0.70) for admission SN concentrations and 0.73 (0.670.78) for admission NT-proBNP concentrations.
SN CONCENTRATIONS AFTER 48 H AND MORTALITY IN CV- AND NON-CV-RELATED ARF
In the subgroup of patients with serial blood sampling (n = 401), SN concentrations were lower after 48 h [median 107 (Q1-3 88-139) pmol/L] compared to SN concentrations on ICU admission [median 120 (95155) pmol/L, P < 0.001]. We found no difference in SN concentrations at 48 h in the patients with CV-related and non-CV-related ARF: 113 (88-140) vs 105 (88 135) pmol/L, P = 0.27. Analogous to the results on ICU admission, SN concentrations measured after 48 h in the ICU were higher in nonsurvivors compared to survivors for CV-related ARF [117 (92-153) vs 111 (88-139) pmol/L, P = 0.04] and non-CV-related ARF [121 (90158) vs 104 (87-126) pmol/L, P = 0.02] (Table 2). The AUC for SN measured after 48 h to predict mortality in CV-related ARF was 0.56 (95% CI, 0.46-0.65), and the AUC of SN to predict mortality in non--CV-related ARF was 0.61 (0.52-0.70).
The main result of this study is that SN provides additional prognostic information when measured on ICU admission in ARF, and especially in the subgroup of patients with CV-related ARF. Our data also demonstrate that circulating SN concentrations are not increased in patients with CV-related ARF compared to patients with non--CV-related ARF, but rather that the potential of SN as a novel biomarker seems linked to improved risk prediction.
There is a need for improved risk prediction in ARF, and one potential strategy could be to identify subgroups of patients based on circulating biomarkers. Accordingly, a number of biomarkers have previously been tested in the FINNALI Study for risk prediction (17, 20-25), including the well-established CV biomarker NT-proBNP. This work found underlying CV disease to be an important determinant of NT-proBNP concentrations, and also demonstrated that admission NT-proBNP concentrations provide prognostic information concerning mortality after 90 days (17). The prognostic utility of NT-proBNP in ARF also has been validated in other cohorts (26, 27). Potential mechanisms underlying the association between cardiac biomarkers and clinical outcomes in patients with respiratory failure include hypoxemia (28), increased proinflammatory cytokine (29), and catecholamine concentrations (30), which influence myocardial function and increase mortality (26, 27). In addition, a substantial proportion of ARF patients have established CV disease as either directly causing the respiratory failure or contributing to respiratory failure. NT-proBNP can be useful in this setting, but because NT-proBNP concentrations are known to be influenced by volume status and primarily reflect myocardial stretch (31), additional biomarkers are warranted.
SN is a novel biomarker that we previously have found to provide incremental prognostic information to established risk indices and NT-proBNP in patients with acute heart failure and after ventricular arrhythmia-induced cardiac arrest (7). In addition, we have also demonstrated a direct association between SN and cardiomyocyte [Ca.sup.2+] handling via [Ca.sup.2+]/calmodulin dependent kinase II 8 (CaMKIIS) inhibition, thus linking SN directly to important CV pathophysiology (32). In addition to myocardial production, SN is also produced and released from neuroendocrine cells and circulating levels can be increased by prolonged, strenuous physical activities (11). We found that circulating SN concentrations were higher in patients with renal failure, possibly due to impaired clearance. However, the influence of renal function on SN concentrations cannot account for the prognostic utility of SN alone, because SN has also previously been found to be associated with mortality after adjusting for indices of renal function (7, 9). Thus, we believe the potential of SN as a biomarker in ARF could relate to circulating SN concentrations functioning as an index of important information from various organ systems, like the heart, kidneys, and the neuroendocrine system. This was recently also demonstrated in critically ill patients with infections (9), and we now add to previous data by demonstrating that SN provides incremental prognostic information to established risk indices in patients with CV-related ARF. Furthermore, patients with SN concentrations >122 pmol/L on ICU admission had an increased risk of mortality after adjustment for SOFA and SAPS II scores. Of note, because patients with ARF may die from several causes (e.g., respiratory failure, renal failure, myocardial dysfunction/cardiac arrest), a single CV biomarker like SN will not provide optimal discrimination for death reflected by AUCs for 90-day mortality for both SN and NT-proBNP in our study. Pertinent to this point, the AUC of admission SN concentrations to predict mortality in patients with CV-related ARF in the FINNALI Study was similar to the AUC previously reported for SN to predict mortality in patients with acute heart failure (7). Moreover, the AUC of admission SN concentrations to predict 90-day mortality was also similar in patients with CV-related ARF after surgery; thus, SN seems to have potential as a prognostic biomarker across different populations with myocardial dysfunction. As other groups have previously found ^-blocker therapy to be associated with reduced mortality in ARF patients (15), it is possible that SN could identify the individuals with CV-related ARF at highest risk who would benefit from cardioprotective therapy. We found SN concentrations to decrease from ICU admission to 48 h, which is analogous to the results previously reported in patients admitted to the ICU after ventricular arrhythmia-induced cardiac arrest (7) and in critically ill patients with infections (10). However, based on ROC values, SN concentrations after 48 h in ARF patients provided more limited prognostic information compared to admission SN concentrations, which has also previously been reported for other cohorts (7, 10); thus, SN seems to provide especially strong prognostic information when measured early after hospital admission.
This study has several strengths and limitations. A strength of our study is the multicenter design, which makes our results representative for the general population of ICU patients with ARF. Pertinent to this point, we demonstrate in online Supplemental Table 1 that the patients included in this SN substudy of the FINNALI Study were comparable for key clinical variables to the patients that were included into the total FINNALI Study cohort. Our study also has some limitations. First, the heterogeneity of the ARF patient population is a general limitation, but we attempted to standardize patient classification by employing the APACHE III diagnostic system. Second, the FINNALI Study also lacks cardiac imaging and detailed information concerning cardiac function, volume status, and cause of death. However, relating to the cause of death, because the majority of deaths occurred during the index hospitalization, we find it plausible that a high number of deaths in CV-related ARF patients were related to myocardial pathology. We also lack information on the specific therapy in the patients with CV-related ARF, time from onset of symptoms to blood sampling, and we do not have data on additional end points like nonfatal myocardial infarction, heart failure, or recurrent ARF. Accordingly, there is a need for additional studies in larger and more contemporary cohorts with detailed clinical phenotyping to validate our results.
We found SN on ICU admission to provide incremental prognostic information to established risk indices in patients with CV-related respiratory failure, including in patients with CV-related ARF after surgery. Accordingly, SN seems to have potential as a prognostic biomarker across different populations with myocardial dysfunction. In contrast, SN did not improve risk assessment in non-CV-related ARF.
Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors' Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the author disclosure form. Disclosures and/or potential conflicts of interest:
Employment or Leadership: T. Omland, CardiNor AS.
Consultant or Advisory Role: None declared.
Stock Ownership: M. Stridsberg, CardiNor AS; G. Christensen, CardiNor AS; T. Omland, CardiNor AS.
Honoraria: None declared.
Research Funding: G. Christensen, the K.G. Jebsen Foundation; T. Omland, the K.G. Jebsen Foundation and the Research Council of Norway; H. Rosjo, Akershus University Hospital, Lorenskog, Norway, and the K.G. Jebsen Foundation. Expert Testimony: None declared.
Patents: M. Stridsberg, G. Christensen, T. Omland, and H. Rosjo are partners in a patent filed by the University of Oslo regarding the use of secretoneurin as a biomarker in patients with cardiovascular disease and critical illness.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, and final approval of manuscript.
Acknowledgments: We are grateful for the contributions by all participating centers in FINNALI to this study. We would also like to acknowledge the expert contribution by Inger Olsson to the SN analyses.
(1.) Vincent JL, Sakr Y, Ranieri VM. Epidemiology and outcome of acute respiratory failure in intensive care unit patients. Crit Care Med 2003; 31:S296-9.
(2.) Bartolomucci A, Possenti R, Mahata SK, Fischer-Colbrie R, LohYP, Salton SR. The extended granin family: structure, function, and biomedical implications. Endocr Rev 2011; 32:755-97.
(3.) Estensen ME, Hognestad A, Syversen U, Squire I, Ng L, Kjekshus J, et al. Prognostic value of plasma chromogranin a levels in patients with complicated myocardial infarction. Am Heart J 2006; 152:927 e1-6.
(4.) Jansson AM, Rosjo H, Omland T, Karlsson T, Hartford M, Flyvbjerg A, Caidahl K. Prognostic value of circulating chromogranin a levels in acute coronary syndromes. Eur Heart J 2009; 30:25-32.
(5.) Omland T, Dickstein K, Syversen U. Association between plasma chromogranin a concentration and long-term mortality after myocardial infarction. Am J Med 2003; 114:25-30.
(6.) Rosjo H, Masson S, Latini R, Flyvbjerg A, Milani V, La Rovere MT, et al. Prognostic value of chromogranin a in chronic heart failure: data from the GISSI-heart failure trial. Eur J Heart Fail 2010; 12:549-56.
(7.) Ottesen AH, Louch WE, Carlson CR, Landsverk OJ, Ku rola J, Johansen RF, et al. Secretoneurin is a novel prognostic cardiovascular biomarker associated with cardiomyocyte calcium handling. J Am Coll Cardiol 2015; 65: 339-51.
(8.) Hasslacher J, Lehner GF, Harler U, Beer R, Ulmer H, Kirchmair R, et al. Secretoneurin as a marker for hypoxic brain injury after cardiopulmonary resuscitation. Intensive Care Med 2014; 40:1518-27.
(9.) Rosjo H, Stridsberg M, Ottesen AH, Nygard S, Christensen G, Pettila V, et al. Prognostic value of secretoneurin in critically ill patients with infections. Crit Care Med [Epub ahead of print 2016 July 13].
(10.) Rosjo H, Stridsberg M, Florholmen G, Stenslokken KO, Ottesen AH, Sjaastad I, et al. Secretogranin II; a protein increased in the myocardium and circulation in heart failure with cardioprotective properties. PloS One 2012; 7:e37401.
(11.) Rosjo H, Opstad PK, Hoff JE, Godang K, Christensen G, Stridsberg M, Omland T. Effect of short- and long-term physical activities on circulating granin protein levels. Regul Pept 2013; 185:14-9.
(12.) Anderson ME. Will secretoneurin be the next big thing? J Am Coll Cardiol 2015; 65:352-4.
(13.) Albrecht-Schgoer K, Schgoer W, Holfeld J, Theurl M, Wiedemann D, Steger C, et al. The angiogenic factor secretoneurin induces coronary angiogenesis in a model of myocardial infarction by stimulation of vascular endothelial growth factor signaling in endothelial cells. Circulation 2012; 126:2491-501.
(14.) Linko R, Okkonen M, Pettila V, Perttila J, Parviainen I, Ruokonen E, et al. Acute respiratory failure in intensive care units. FINNALI: a prospective cohort study. Intensive Care Med 2009; 35:1352-61.
(15.) Noveanu M, Breidthardt T, Reichlin T, Gayat E, Potocki M, Pargger H, et al. Effect of oral beta-blocker on short and long-term mortality in patients with acute respiratory failure: Results from the BASEL-II-ICU Study. Crit Care 2010; 14:R198.
(16.) Stridsberg M, Eriksson B, Janson ET. Measurements of secretogranins II, III, V and proconvertases 1/3 and 2 in plasma from patients with neuroendocrine tumours. Regul Pept 2008; 148:95-8.
(17.) Okkonen M, Varpula M, Linko R, Perttila J, Varpula T, Pettila V, FINNALI Study Group. N-terminal-pro-BNP in critically ill patients with acute respiratory failure: a prospective cohort study. Acta Anaesthesiol Scand 2011; 55:749-57.
(18.) Demler OV, Pencina MJ, D'Agostino RB, Sr. Misuse of DeLong test to compare AUCs for nested models. Stat Med 2012; 31:2577-87.
(19.) Pencina MJ, D'Agostino RB, Sr., Steyerberg EW. Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers. Stat Med 2011; 30:11-21.
(20.) Hastbacka J, Linko R, Tervahartiala T, Varpula T, Hovilehto S, Parviainen I, et al. Serum MMP-8 and TIMP-1 in critically ill patients with acute respiratory failure: TIMP-1 is associated with increased 90-day mortality. Anesth Analg 2014; 118:790-8.
(21.) Jalkanen V, Yang R, Linko R, Huhtala H, Okkonen M, Varpula T, et al. SuPAR and PAI-1 in critically ill, mechanically ventilated patients. Intensive Care Med 2013; 39:489-96.
(22.) Linko R, Karlsson S, Pettila V, Varpula T, Okkonen M, Lund V, et al. Serum zinc in critically ill adult patients with acute respiratoryfailure. Acta Anaesthesiol Scand 2011; 55:615-21.
(23.) Linko R, Pettila V, Kuitunen A, Korhonen AM, Nisula S, Alila S, et al. Plasma neutrophil gelatinase-associated lipocalin and adverse outcome in critically ill patients with ventilatory support. Acta Anaesthesiol Scand 2013; 57:855-62.
(24.) Okkonen M, Lakkisto P, Korhonen AM, Parviai-nen I, Reinikainen M, Varpula T, et al. Plasma cell-free DNA in patients needing mechanical ventilation. Crit Care 2011; 15:R196.
(25.) de Kretser DM, Bensley JG, Pettila V, Linko R, Hedger MP, Hayward S, et al. Serum activin A and B levels predict outcome in patients with acute respiratory failure: a prospective cohort study. Crit Care 2013; 17:R263.
(26.) Bajwa EK, Januzzi JL, Gong MN, Thompson BT, Christiani DC. Prognostic value of plasma N-terminal probrain natriuretic peptide levels in the acute respiratory distress syndrome. Crit Care Med 2008; 36:2322-7.
(27.) Meyer B, Huelsmann M, Wexberg P, Delle Karth G, Berger R, Moertl D, et al. N-terminal pro-B-type natriuretic peptide is an independent predictor of outcome in an unselected cohort of critically ill patients. Crit Care Med 2007; 35:2268-73.
(28.) Khan AR, Birbach M, Cohen MS, Ittenbach RF, Spray TL, Levy RJ, Gaynor JW. Chronic hypoxemia increases ventricular brain natriuretic peptide precursors in neonatal swine. Ann Thoracic Surg 2008; 85:618-23.
(29.) Ma KK, OgawaT, de Bold AJ. Selective upregulation of cardiac brain natriuretic peptide at the transcriptional and translational levels by pro-inflammatory cytokines and by conditioned medium derived from mixed lymphocyte reactions via p38 MAP kinase. J Mol Cell Cardiol 2004; 36:505-13.
(30.) Rengo G, Pagano G, Parisi V, Femminella GD, de Lucia C, Liccardo D, etal. Changes of plasma norepinephrine and serum N-terminal pro-brain natriuretic peptide after exercise training predict survival in patients with heart failure. Int J Cardiol 2014; 171:384-9.
(31.) Omland T. Advances in congestive heart failure management in the intensive care unit: B-type natriuretic peptides in evaluation of acute heart failure. Crit Care Med 2008; 36:S17-27.
(32.) Anderson ME, Brown JH, Bers DM. CaMKII in myocardial hypertrophy and heart failure. J Mol Cell Cardiol 2011; 51:468-73.
Peder L. Myhre, [1,2] Anett H. Ottesen,  Marjatta Okkonen,  Rita Linko,  Mats Stridsberg,  Stale Nygard,  Geir Christensen,  Ville Pettila, [3,7] Torbjorn Omland,  and Helge Rosjo,  on behalf of the FINNALI Laboratory Study Group+
 Division of Medicine, Akershus University Hospital, Lorenskog, Norway and Center for Heart Failure Research, University of Oslo, Oslo, Norway; 2 Center for Clinical Heart Research, Oslo University Hospital Ulleval, Oslo, Norway; 3 Division of Intensive Care Medicine, Department of Anesthesiology, Intensive Care and Pain Medicine, University of Helsinki and Helsinki University Hospital, Helsinki, Finland; 4 Department of Medical Sciences, Uppsala University, Uppsala, Sweden; 5 Bioinformatics Core Facility, Oslo University Hospital and the University of Oslo, Oslo, Norway; 6 Institute for Experimental Medical Research, Oslo University Hospital, Ulleval, Oslo, Norway and Center for Heart Failure Research, University of Oslo, Oslo, Norway; 7 Department of Intensive Care Medicine, Bern University Hospital, University of Bern, Bern, Switzerland.
* Address correspondence to this author at: Division of Medicine, Akershus University Hospital, Sykehusveien 25, 1478 Lorenskog, Norway. Fax +47-67962190; e-mail: email@example.com.
([dagger]) complete list of the investigators who participated to the study is presented in the online Supplemental Appendix.
Received April 14, 2016; accepted June 23, 2016.
Previously published online at DOI: 10.1373/clinchem.2016.258764
 Nonstandard abbreviations: ICU, intensive care units; ARF, acute respiratory failure; Biomarker, biological marker; SN, secretoneurin; CV, cardiovascular; FINNALI, FINNish Acute Lung Injury; NT-proBNP, N-terminal-pro-B-type natriuretic peptide; APACHE III, Acute Physiology and Chronic Health Evaluation III; SAPS II, Simplified Acute Physiology Score; SOFA, Sequential Organ Failure Assessment; RIA, Q, quartile; WBC, white blood cell count; AUC, area under the curve; HR, hazard ratio; NRI, category-free net reclassification index; CaMKIIS, [Ca.sup.2+]/calmodulin dependent kinase II [delta]; MAP, mean arterial pressure.
Caption: FIG. 1. Flow chart of the study.
Caption: FIG. 2. 90-day survival curves in ARF stratified by SN Qs on ICU admission and etiology for ARF.
Table 1. Characteristics of the patients stratified into CV -and non-CV-related ARF. (a) CV-related ARF Non-CV-related ARF (n = 209) (n = 281) Age, years 71(61-77) 58 (45-71) Male sex 141 (68%) 177 (63%) Body mass index, (c) 27 (23-29) 26 (23-30) kg/[m.sup.2] Current smoker (d) 39(19%) 84 (30%) Diabetes mellitus (c) 48 (23%) 42 (15%) Chronic obstructive 41 (20%) 45(16%) lung disease (c) Chronic heart 139(67%) 66 (24%) disease (c) Operative admission 113 (54%) 104 (37%) Heart rate on 83(72-93) 82 (70-96) inclusion, (c) 1/min MAP on inclusion, (c) 76 (66-84) 79 (68-88) mmHg Respiratory rate on 14(12-16) 14(12-17) inclusion, (c) 1/min SOFA score (c) 8(6-11) 7 (5-9) SAPS II score (c) 40(28-60) 43 (30-53) WBC, (c) 1 x 10.6(8.2-14.1) 10.2 (7.6-14.3) [10.sup.9]/L Creatinine on 0.96(0.76-1.56) 0.78 (0.64-1.20) inclusion, (e) mg/dL SN on inclusion, 120(95-167) 120(95-148) pmol/L NT-proBNP on 2013(771-6951) 688(179-3612) inclusion, ng/L SN at 48 h, (f) pmol/L 113(88-140) 105 (88-135) NT/proBNP at 48 3281 (1611-7066) 767 (222-2499) h, (f) ng/L 90-day mortality 64 (31%) 68 (24%) P value (b) Age, years <0.001 Male sex 0.31 Body mass index, (c) 0.49 kg/[m.sup.2] Current smoker (d) 0.001 Diabetes mellitus (c) 0.02 Chronic obstructive 0.30 lung disease (c) Chronic heart <0.001 disease (c) Operative admission <0.001 Heart rate on 0.91 inclusion, (c) 1/min MAP on inclusion, (c) 0.006 mmHg Respiratory rate on 0.44 inclusion, (c) 1/min SOFA score (c) <0.001 SAPS II score (c) 0.94 WBC, (c) 1 x 0.28 [10.sup.9]/L Creatinine on <0.001 inclusion, (e) mg/dL SN on inclusion, 0.61 pmol/L NT-proBNP on <0.001 inclusion, ng/L SN at 48 h, (f) pmol/L 0.27 NT/proBNP at 48 <0.001 h, (f) ng/L 90-day mortality 0.11 (a) Data presented as numbers (%) or median (interquartile range). (b) P value for comparison between CV-related ARF vs non-CV-related ARF. (c) <5% missing. (d) <10% missing. (e) <15% missing. (f) <20% missing. Table 2. Characteristics of 90-daysurvivors and nonsurvivors with patients stratified as CV-and non-CV-related ARF. (a) CV-related ARF Nonsurvivors Survivors (n = 64) (n = 145) P value Age 74(62-81) 69 (60-76) 0.03 Male sex 42 (66%) 99 (68%) 0.71 Body mass index 27 (24-30) 26 (23-30) 0.17 Current smoker 13(20%) 26 (18%) 0.66 Diabetes mellitus 13(20%) 35 (24%) 0.53 Chronic obstructive 17 (27%) 24(17%) 0.08 lung disease Chronic heart 48 (75%) 91 (63%) 0.08 disease Surgical admission 17 (27%) 96 (66%) <0.001 Heart rate on 79(64-93) 84(75-93) 0.01 inclusion MAP on inclusion 76(67-85) 75 (65-83) 0.67 Respiratory rate on 14(12-17) 14(12-15) 0.27 inclusion SOFA score 11 (8-14) 7 (6-9) <0.001 SAPS II score 50(41-64) 40 (29-50) <0.001 WBC, 1 x [10.sup.9]/L 12.0 (8.0-17.0) 10.1 (8.3-13.5) 0.41 Creatinine on 1.30 (0.83-2.04) 0.90 (0.72-1.24) <0.001 inclusion, mg/dL SN on inclusion, 148 (117-203) 108 (87-143) <0.001 pmol/L NT/proBNP on 3796 (1158-15922) 1593 (610-5089) <0.001 inclusion, ng/L SN at 48 h, pmol/L 117 (92-153) 111 (88-139) 0.04 NT/proBNP at 48 h, 4211 (955-15 455) 3233 (1721-6024) 0.02 ng/L Non-CV-related ARF Nonsurvivors Survivors (n = 68) (n = 213) Age 71 (60-77) 54(41-68) Male sex 40 (59%) 137 (64%) Body mass index 26(24-31) 26 (23-30) Current smoker 14(21%) 70 (33%) Diabetes mellitus 10(15%) 32 (15%) Chronic obstructive 13(19%) 32 (15%) lung disease Chronic heart 32 (47%) 34 (16%) disease Surgical admission 24 (35%) 80 (38%) Heart rate on 85(73-100) 82 (69-95) inclusion MAP on inclusion 77(67-91) 79 (68-88) Respiratory rate on 16(13-20) 14(12-16) inclusion SOFA score 9(6-112) 7 (4-9) SAPS II score 50 (40-65) 39 (29-51) WBC, 1 x [10.sup.9]/L 10.8(7.5-15.1) 10.1 (7.7-13.8) Creatinine on 1.17 (0.71-1.76) 0.77 (0.63-1.00) inclusion, mg/dL SN on inclusion, 139(115-184) 113(91-139) pmol/L NT/proBNP on 3567 (978-10594) 401 (113-1672) inclusion, ng/L SN at 48 h, pmol/L 121 (90-158) 104(87-126) NT/proBNP at 48 h, 3436(565-8140) 587 (172-1842) ng/L Non-CV-related ARF P value Age <0.001 Male sex 0.42 Body mass index 0.45 Current smoker 0.09 Diabetes mellitus 0.93 Chronic obstructive 0.46 lung disease Chronic heart <0.001 disease Surgical admission 0.74 Heart rate on 0.22 inclusion MAP on inclusion 0.73 Respiratory rate on 0.001 inclusion SOFA score <0.001 SAPS II score <0.001 WBC, 1 x [10.sup.9]/L 0.40 Creatinine on 0.001 inclusion, mg/dL SN on inclusion, <0.001 pmol/L NT/proBNP on <0.001 inclusion, ng/L SN at 48 h, pmol/L 0.02 NT/proBNP at 48 h, <0.001 ng/L (a) Data presented as numbers (%)or median (interquartile range). Table 3. Univariate associations between variables available on ICU admission and time to death during 90 days by Cox regression in patients with CV-and non-CV-related ARF. CV-related ARF (n = 209) HR 95% CI P Age, per 1-year 1.03 1.00-1.05 0.03 increase Female sex 1.10 0.66-1.85 0.71 Body mass index, per 1.03 0.99-1.06 0.13 1-kg/[m.sup.2] increase Current smoker 1.15 0.62-2.13 0.66 Diabetes mellitus 0.80 0.43-1.45 0.47 Chronic obstructive 1.52 0.87-2.66 0.14 lung disease Chronic heart 1.64 0.92-2.93 0.10 disease Operative admission 0.24 0.14-0.41 <0.001 Heart rate, on 0.98 0.97-1.00 0.01 inclusion, per 1/min MAP, on inclusion, 1.00 0.99-1.02 0.67 per mmHg Respiratory rate, on 1.04 0.98-1.10 0.21 inclusion per 1/min WBC on inclusion, 1.53 1.07-2.20 0.02 per unit increase, (a) 1 x [10.sup.9]/L SOFA score, per 1.23 1.22-1.43 <0.001 1-point increase SAPS II score, per 1.06 1.04-1.07 <0.001 1-point increase Creatinine on 2.50 1.59-3.91 <0.001 inclusion, per unit increase, (a) mg/dL SN on inclusion, per 3.12 2.02-4.82 <0.001 unit increase, (a) pmol/L NT-proBNP on 1.35 1.15-1.60 <0.001 inclusion, per unit increase, (a) ng/L Non-CV-related ARF (n = 281) HR 95% CI P Age, per 1-year 1.06 1.04-1.08 <0.001 increase Female sex 1.21 0.75-1.96 0.44 Body mass index, per 1.00 0.97-1.05 0.66 1-kg/[m.sup.2] increase Current smoker 0.60 0.33-1.09 0.10 Diabetes mellitus 0.96 0.49-1.96 0.91 Chronic obstructive 1.19 0.65-2.18 0.57 lung disease Chronic heart 3.56 2.21-5.73 <0.001 disease Operative admission 0.89 0.54-1.47 0.66 Heart rate, on 1.01 1.00-1.02 0.19 inclusion, per 1/min MAP, on inclusion, 1.00 0.99-1.02 0.93 per mmHg Respiratory rate, on 1.07 1.03-1.12 <0.001 inclusion per 1/min WBC on inclusion, 1.13 0.74-1.74 0.58 per unit increase, (a) 1 x [10.sup.9]/L SOFA score, per 1.18 1.1-1.27 <0.001 1-point increase SAPS II score, per 1.06 1.04-1.07 <0.001 1-point increase Creatinine on 2.31 1.53-3.49 <0.001 inclusion, per unit increase, (a) mg/dL SN on inclusion, per 4.09 2.27-7.34 <0.001 unit increase, (a) pmol/L NT-proBNP on 1.51 1.33-1.72 <0.001 inclusion, per unit increase, (a) ng/L (a) SN, NT-proBNP, creatinine, and WBC levels were transformed by the natural logarithm before analysis. Table 4. Multivariate associations between variables available on ICU admission and time to death during 90 days by Cox regression in patients with CV-and non-CV-related ARF. (a) HR 95% CI P CV-related ARF Operative admission 0.35 0.19-0.66 0.001 SN on inclusion, per 2.14 1.16-3.92 0.014 unit increase, (b) pmol/L WBC on inclusion, 1.50 0.90-2.48 0.12 per unit increase, (b) 1 x 1 09/L Age, per 1-year 1.02 0.99-1.04 0.17 increase Heart rate, on 0.99 0.98-1.01 0.42 inclusion, per 1/min Creatinine on 1.28 0.70-2.34 0.42 inclusion, per unit increase, (b) mg/dL NT-proBNP on 1.04 0.83-1.30 0.75 inclusion, per unit increase, (b) ng/L Non-CV-related ARF Age, per 1-year 1.04 1.02-1.06 <0.001 increase NT-proBNP on 1.20 1.008-1.43 0.04 inclusion, per unit increase, (b) ng/L Respiratory rate, on 1.05 1.00-1.11 0.08 inclusion, per 1/min SN on inclusion, per 2.14 0.91-5.05 0.08 unit increase, (b) pmol/L Creatinine on 1.21 0.64-2.28 0.56 inclusion, per unit increase, (b) mg/dL Chronic heart 1.20 0.66-2.18 0.55 disease (a) All variables significantly associated to outcome in the univariate model (Table 3) were included in the models. (b) SN, NT-proBNP, creatinine, and WBC levels were transformed by the natural logarithm before analysis.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Proteomics and Protein Markers|
|Author:||Myhre, Peder L.; Ottesen, Anett H.; Okkonen, Marjatta; Linko, Rita; Stridsberg, Mats; Nygard, Stale;|
|Date:||Oct 1, 2016|
|Previous Article:||Prognostic Utility of Galectin-3 for Recurrent Cardiovascular Events During Long-term Follow-up in Patients with Stable Coronary Heart Disease:...|
|Next Article:||A Nanoparticle-Lectin Immunoassay Improves Discrimination of Serum CA125 from Malignant and Benign Sources.|