Association Between the Level of Plasma B-type Natriuretic Peptide and in-hospital Mortality in Patients With Acute Exacerbation of COPD Without a History of Coronary Ischemic Heart Disease: A Retrospective Study

Previous studies suggested that plasma B-type natriuretic peptide (BNP) level was often elevated in patients with acute exacerbations of chronic obstructive pulmonary disease (AECOPD) and was associated with increased mortality. However, most studies did not consider the fact that conditions such as coronary ischemic heart disease can also increase BNP level. Therefore, we aimed to explore the association between BNP level and in-hospital mortality in patients with AECOPD without a history of coronary ischemic heart disease.


Page 3/18
Chronic obstructive pulmonary disease (COPD) is a global public health challenge because of its high prevalence and the related disability and mortality. [1] COPD has become the third leading cause of death worldwide and is estimated to become the source of the seventh highest burden worldwide in 2030. [2] Prevalence was higher in men than in women, and in individuals aged ≥40 years, it was 13.7% in China; [1] 6.4% of Americans reported receiving a diagnosis of COPD. [3] Plasma brain natriuretic peptide and N-terminal fragment of pro-B-type natriuretic peptide (BNP; NT-pro-BNP) are cardiac neurohormones that are synthesized and released from ventricular myocytes in response to myocardial stretching. [4] It had been proven that BNP and NT-pro-BNP are diagnostic and prognostic biomarkers of congestive heart failure [5] and established that BNP and NT-pro-BNP are independent predictors of death and cardiovascular events in patients with stable coronary artery disease. [6] Furthermore, they were found to be elevated with right ventricular overload and to be associated with a poor prognosis in patients with pulmonary arterial hypertension and pulmonary thromboembolism. [7][8][9] Previous studies have shown that BNP and NT-pro-BNP are probably useful in predicting the outcomes of COPD. A prospective, single-center study on acute exacerbation of COPD (AECOPD) suggested that elevated NT-pro-BNP level was associated with an increase in 30-day mortality. [10] Another prospective study showed that NT-pro-BNP level was associated with in-hospital mortality and 1-year mortality in patients with exacerbations of COPD; however, in their studies, about 35.9% of patients had congestive heart failure. [11] Stolz et al. found that BNP accurately predicted the need for intensive care unit (ICU) treatment, [12] and BNP was an independent predictor of requirement for noninvasive ventilation, mechanical ventilation, and noninvasive ventilation failure in patients with AECOPD along with preserved left ventricular function. [13] However, some studies have found that NT-pro-BNP level was not associated with the requirement for noninvasive positive pressure ventilation, mechanical ventilation, or in-hospital mortality in patients with AECOPD without underlying left ventricular dysfunction. [14] We aimed to assess the association between BNP level and in-hospital mortality or requirement for invasive mechanical ventilation in patients hospitalized for AECOPD without a history of coronary ischemic heart disease.

Methods
Ethics Committees of Jinan Central Hospital reviewed and approved this retrospective study protocol in accordance with the precepts established by the Declaration of Helsinki (the assurance number 2020-137-01), and waived the need for obtaining informed consent from patients.

Data Sources
All data were obtained from electronic patient les and medical data intelligence platform of Jinan Central Hospital. The medical data intelligence platform is administered by Jinan Central Hospital, and all local in-patient and out-patient information was registered in it since January 2009.

Study Patients and Covariates
In this retrospective cohort study, patients who were diagnosed with AECOPD using International Statistical Classi cation of Diseases and Related Health Problems, Nineth Revision (ICD-9 codes) between January 2017 and December 2019 were included (explicit ICD-9 codes J44.0, J44.1 and J44.9). Patients with a medical history of coronary ischemic heart disease and advanced renal disease were excluded. Patients were excluded if the discharge diagnosis included acute asthma, bronchiectasis, bronchial lung cancer, and acute myocardial infarction. If patients were admitted more than once during the study period, only the rst admission was included in the analysis. We obtained medical history and physiological variables that were recorded on admission. These data were used to compile two prognostic scores: CURB65 and BAP65 (BUN level > 25 mg/dL, Altered mental status, Pulse > 109 beats/min, Age≥65 years). [15][16] If laboratory parameters were measured more than once during the inhospital stay, data obtained from the rst measurement within 24 h of admission alone were collected. The plasma BNP was a continuous variable. In this retrospective cohort study, there were 2 outcomes of interest, primary outcome, in-hospital mortality; secondary outcomes, (1) a composite outcome of inhospital mortality or invasive mechanical ventilation and (2) invasive mechanical ventilation.

Statistical Analysis
Continuous variables with normal distribution are presented as mean ± standard deviation (SD). Nonnormally distributed variables are presented as median (interquartile range [IQR]). Categorical variables are presented as percentage. Differences between groups were compared using the Kruskal-Wallis test for continuous variables or Wilcoxon-Mann-Whitney Test for categorical variables. BNP was log 2 transformed. Outcomes were compared using cox regression, both unadjusted and adjusted. Adjusted models included the following covariates: age; sex; smoking status of never, ex-smoker, and current smoking; drinking status of never, ex-drinker, and current drinker; past medical history of hypertension, diabetes, cerebral vascular disease, renal dysfunction, atrial brillation, and malignancy; variables with p ≤ 0.1 in univariate analysis (CURB65, blood neutrophil count, D-dimer level, arterial partial pressure of O 2 [PaO 2 ], and arterial partial pressure of CO 2 [PaCO 2 ]). Subgroup analyses (by sex, age and missing of lung function) were performed, and differences in HRs between subgroups were tested by Wald tests for statistical interaction. Statistical analysis was performed using Packages R (The R Foundation, Vienna, Austria); a statistical signi cance level of 0.05 was used.

Results
In this retrospective cohort study designed to examine the association of BNP level with the outcomes of patients with AECOPD, medical records of 722 patients hospitalized with AECOPD were reviewed, and those on 300 patients were included. Patient selection is shown in the Figure 1. The demographic and baseline characteristics of all the included patients are shown in Table 1. Among the patients, 63.33% were men. The mean age of patients was 76.88 ± 9.53 years. Sixteen patients died during hospitalization, 29 patients were assisted with invasive mechanical ventilator, and 22 patients were assisted with noninvasive positive pressure ventilation. The duration of hospital stay was 12.04 ± 6.35 days. BNP levels of 300 (100%) patients were available. Its median was 65.50 pg/ml (IQR, 30.00, 181.50). C-reactive protein level, ndings of arterial blood gas test, and lung function data were missing in the study, but the missing data was processed according to the median value. After median replacement of missing data, there was no difference in the ndings of multivariable cox regression models (Table 1s).
In univariate cox regression analysis, we used all baseline variables of the entire cohort to assess factors associated with the primary outcome and the secondary outcomes ( Table 2, Table 2b and Table 2c). The in-hospital mortality was 5.33% (16/300). The secondary outcome of in-hospital mortality or invasive mechanical ventilation was observed 10.67% (32/300). Table 3 shows the unadjusted and adjusted HRs derived from the Cox regression analysis. As in the case of the primary outcome of in-hospital mortality, a higher log 2 BNP value was signi cantly associated with a higher risk of the combined end point of in-hospital mortality or invasive mechanical ventilation in unadjusted analysis (HR, 1.42 [95% CI, 1.19-1.71]), and this association was still signi cant after adjusting for age, sex, smoking status, drinking status, past medical history, CURB65, blood neutrophil count, D-dimer level, PaO 2 , and PaCO 2 (HR, 1.36 [95% CI, 1.07-1.74]).
The secondary outcome of invasive mechanical ventilation was observed in 9.67% (29/300) of patients. Table 3 shows the unadjusted and adjusted HRs derived from the Cox regression analysis. As in the case of the primary outcome of in-hospital mortality, a higher log 2 BNP value was signi cantly associated with a higher risk of invasive mechanical ventilation in unadjusted analysis ( In this study, among patients with AECOPD, subgrouping was performed according to (1) sex, (2) age, and (3) missing lung function data were examined. The results were similar to the overall results, and all tests for interaction with these covariates were not statistically signi cant (P >0.05) ( Table 4).

Discussion
Among patients diagnosed with exacerbations of COPD, this study found signi cant association between log 2 BNP and in-hospital mortality, a composite outcome of in-hospital mortality or mechanical ventilation, and mechanical ventilation, after adjusting for baseline demographics, comorbidities and other hematological data; results of the subgroup analysis were stable; and this suggests that BNP can be used as an important biomarker of prognosis in patients with COPD exacerbations.
A previous prospective study by Chang et al. found that elevated NT-pro-BNP levels (>220 pmol/L) were associated with increased 30-day mortality, and this nding is consistent with that of our study. [10] Another study by Hai et al. found that NT-pro-BNP level (≥551.35 ng/l) was associated with in-hospital mortality and 1-year mortality in patients with exacerbations of COPD, but in their study, about 35.9% of patients had congestive heart failure. [11] A systematic review and meta-analysis also found that elevated NT-pro-BNP values were associated with all-cause in-hospital mortality in patients with and without exacerbation of COPD. [17] Stolz et al. found that BNP was not associated with 6-month or 2-year mortality in patients with exacerbations of COPD, but BNP accurately predicted the need for ICU treatment, and the need for mechanical ventilation was the most common reasons for ICU admission. [12] This is con rmed in our study, we also found that the association between log 2 BNP and invasive mechanical ventilation was strong in multivariate analysis. This was consistent with the nding of previous studies in which BNP was an independent predictor of higher noninvasive ventilation requirement, mechanical ventilation use, and noninvasive ventilation failure in patients with AECOPD with preserved left ventricular function, but they did not assess the association of BNP and in-hospital mortality. [13] Another previous retrospective study by Muhammad et al. demonstrated that log-transformed NT-pro-BNP levels were positively associated with echocardiographically estimated right ventricular systolic pressure; however, NT-pro-BNP was not associated with the need for noninvasive positive pressure ventilation/mechanical ventilation or with in-hospital mortality in patients with AECOPD without underlying left ventricular dysfunction. [18] Plasma brain natriuretic peptide and NT-pro-BNP are cardiac neurohormones that are synthesized and released from ventricular myocytes in response to myocardial stretch. [4] It was established that NT-pro-BNP is a diagnostic and prognostic biomarker of congestive heart failure, [5] and it has also been found to be a signi cant and independent predictor of death and cardiovascular events in stable coronary artery disease. [6] In our study, we had excluded patients with a past medical history of coronary ischemic heart disease; however, we had not obtained the echocardiographic data; therefore, a small number of patients with undiagnosed heart failure may have been included. The meta-analysis performed by Pavasini et al. suggested that having a history of heart failure did not in uence the relationship between elevated NTpro-BNP levels and long-term mortality. [18] The pathophysiological processes underlying the elevation of BNP level in patients with exacerbations of COPD are unknown. Hypoxia often occurs in severe COPD, especially in patients with AECOPD. In our research, oxygen tension on arterial blood gas measurement was strongly associated with in-hospital mortality (P=0.008), composite outcome of in-hospital mortality or mechanical ventilation(P=0.009), and mechanical ventilation(P=0.014) in univariate analysis. Hypoxia can cause pulmonary hypertension and right ventricular dysfunction by pulmonary vasoconstriction. [7] Concentration of NT pro-BNP was signi cantly higher in those with a higher pulmonary hypertension than in those with a lower pulmonary hypertension, and they found that NT pro-BNP was an excellent discriminator of right-ventricular impairment but not of cor pulmonale. [8] Previous studies have found that BNP and NT-pro-BNP levels were elevated in patients with pulmonary thromboembolism, and BNP may be a useful determinant of the short-term outcome. [9] Coronary artery ischemia may be more likely to occur in patients with AECOPD. Gavin et al found that exacerbations of COPD increase the risk of myocardial infarction. [14] This nding may be related to hypoxic ischemia of myocardium caused by hypoxia. Hypoxic myocardial ischemia leads to a decrease in myocardial contractility, which increases the burden on the heart and increases BNP or NT-pro-BNP secretion. In our study, 12 patients had acute myocardial infarction during hospitalization and were excluded.
There are some limitations to our research. First, our study was observational; no causal inference can be obtained, and relationships should be interpreted as associations. Second, COPD speci c diagnostic codes for the identi cation of patients were adopted. Spirometry data of some patients were not available completely. However, we reviewed the electronic patient les and medical data intelligence platform of Jinan Central Hospital and found that they were hospitalized many times and inhaled maintenance drugs of COPD for a long time; in subgroup analysis, irrespective of whether data on the lung function were missing or not, the results were consistent. Third, although our study was a retrospective single-center trial, it was closer to real-world clinical practice, and the patient treatments by clinicians were not in uenced by the study. Considering that no multicenter study has been performed so far, a multi-center prospective study is urgently needed to further explore the prognostic e ciency of BNP for AECOPD.

Conclusions
This study suggests that a higher log 2 BNP value was signi cantly associated with in-hospital mortality and invasive mechanical ventilation in patients with AECOPD, and BNP is a promising prognostic marker of in-hospital mortality and invasive mechanical ventilation in patients with AECOPD. These ndings justify that a multi-center prospective study is urgently needed to further explore the prognostic e ciency of BNP for AECOPD.
Abbreviations AECOPD, Acute exacerbations of chronic obstructive pulmonary disease; BAP65, BUN level > 25 mg/dL, Altered mental status, Pulse > 109 beats/min, Age≥65 years; BNP, B-type natriuretic peptide; COPD, Chronic obstructive pulmonary disease; CURB65, Confusion, Urea > 7mmol/L, Respiratory rate≥30/min, Blood pressure systolic < 90 mmHg or diastolic <60 mmHg and age > 65 years; ICD-9, International Statistical Classi cation of Diseases and Related Health Problems, Nineth Revision; ICU, Intensive care unit; SD, Standard deviation Declarations Ethics approval and consent to participate Ethics Committees of Jinan Central Hospital reviewed and approved this retrospective study protocol.

Consent for publication
Not applicable.

Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.