Midrange-proadrenomedullin as a marker for mortality and morbidity in COPD
Due to the COVID-19 crisis measures the PhD defence of Maaike Zuur-Telgen will take place online without the presence of an audience.
Maaike Zuur-Telgen is a PhD student in the department of Research Methodology, Measurement and Data Analysis (OMD). Her supervisors are prof.dr. J.A.M. van der Palen from the faculty of Behavioural Management and Social Sciences and prof.dr. H.A.M. Kerstjens from the University of Groningen.
Chronic Obstructive Pulmonary Disease (COPD) is defined as a preventable and treatable disease characterized by persistent respiratory symptoms and airflow limitation that is due to airway and/or alveolar abnormalities, usually caused by significant exposure to noxious particles or gases, especially tobacco smoke. COPD is also a heterogeneric and complex disease with multiple subtypes. The prediction of mortality is relevant to identify patients in whom adjustment of care may be appropriate. There is growing interest in biomarkers to assess disease heterogeneity and predict progression. One possible biomarker is midrange-proadrenomedullin (MR-proADM), the more stable precursor of adrenomedullin (ADM). ADM has immune-modulating, metabolic, and vascular actions. The main objective of this thesis was to see if midrange-proadrenomedullin (MR-proADM) is associated with morbidity and mortality in COPD patients.
In chapter two MR-proADM was measured in a subgroup of 181 patients from the COMIC (“Cohort of Mortality and Inflammation in COPD” ) study with a plasma sample in stable state. We showed that high levels of MR-proADM in stable state are associated with a 3-fold increased risk of mortality when corrected for potential confounding variables. The threefold-increased risk is highly relevant, especially because the absolute mortality risk in this study was high, with up to 23% not surviving 2 years. MR-proADM could be used in the process of clinical decision making. To make a clinical decision it is important to choose the right cut-off value of MR-proADM. Until now, the cut-off value for MR-proADM levels in patients with COPD was chosen to be the median of the population. In chapter two we also tried to find the best cut-off point in stable state for MR-proADM to predict mortality. The best cut-off point depends on the intended use and the population under study.
An important step in biomarker development next to the discovery of the biomarker is to validate this biomarker to be sure that the test is reliable, reproducible and of adequate sensitivity and specificity for the proposed use. In chapter three we responded to the letter to the editor of Dr Khorfan who suggested to validate the cut-off value in another group of patients. We agreed with the importance of validation and explained our next steps in further research.
In chapter four the cut-off values mentioned in chapter two in stable state and the cut-off value as previously studied by Stolz et al. were validated in another subgroup of 545 patients of the COMIC study. This validation study confirmed that high MR-proADM levels, measured in patients with COPD in stable state, were associated with a higher mortality risk. We showed a 1.9 and 2.2-fold increased risk of 2-year mortality associated with MR-proADM levels measured in stable state, with a cut-off value of 0.71 and 0.75 nmol/l, respectively. The absolute mortality risk in this population was considerable with 13% not surviving two years. From this we concluded that MR-proADM is a valid and clinically relevant biomarker when measured in stable state.
In this thesis we also studied procalcitonin (PCT) as potential biomarker for mortality prediction. PCT is a marker of systemic bacterial infection, PCT is elevated at hospitalization for an AECOPD and is associated with poorer in-hospital prognosis, greater length of hospital stay and need for intensive care unit (ICU). In our study we wanted to confirm the association of PCT with mortality in stable state. In chapter two we showed that high levels of PCT measured in stable state and at hospitalization for AECOPD were not significantly associated with survival. Because our findings were in line with the results of a previous study by Stolz et al., we did not qualify PCT as an important biomarker for mortality prediction in COPD in both stable state and at hospitalization for AECOPD.
Another well studied biomarker in COPD is fibrinogen. Fibrinogen is accepted as a prognostic biomarker in COPD by the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA). Besides a blood clothing factor, fibrinogen is also an acute phase reactant. In chapter five we described that fibrinogen (as a continuous variable), measured in the COMIC cohort in stable state, is a predictor for mortality in COPD. In the multivariate Cox regression with fibrinogen (as a continuous variable) a doubling of fibrinogen levels was associated with a 2.39 (95% CI 1.41-4.05) fold increased risk of dying, with a C-statistic of 0.73.
Because COPD is a complex and heterogeneous disease, one single biomarker is unlikely to fully represent this. In chapter five we therefore combined MR-proADM with fibrinogen in order to optimize mortality prediction. We found that adding MR-proADM to fibrinogen increases the accuracy of the short term mortality prediction model (one year follow-up) both relevantly and significantly. However, the combined model was not superior in predicting short term mortality compared to a model with solely MR-proADM.
Several multicomponent indices of clinical variables are available. Although these index scores are multicomponent scores, the systemic component of COPD is not sufficiently reflected in these scores. In chapter six we combined MR-proADM (cut-off value of 0.87 nmol/l) with the ADO (Age, Dyspnea and airflow Obstruction) the updated ADO and the BOD (BMI, airflow Obstruction and Dyspnea). In a pooled assessment of two large European prospective observational cohort studies (COMIC and the PROMISE-COPD study) adding MR-proADM to the indices ADO, updated ADO and BOD improved mortality prediction of all three indices. The additive predictive value of MR-proADM to both the ADO, updated ADO and BOD index reflects the idea that systemic components are insufficiently reflected in the mentioned multicomponent indices. In conclusion the ADOA index and the updated ADOA index have a numerically higher predictive power than the BODA index.
COPD is characterized by acute exacerbations of COPD (AECOPDs), which can accelerate the already existing gradual decline in lung function and increase the risk of death. Furthermore, AECOPDs are associated with increased risk of hospitalization, lower quality of life, and increased health-care costs. The aim of the study as described in chapter seven was to evaluate the association of stable-state MR-proADM with future severe AECOPD and community acquired pneumonia in COPD patients. In our pooled analysis of the COMIC and the PROMISE-COPD study MR-proADM, measured in stable state, was shown to be an important biomarker for the occurrence of a future severe AECOPD (HR 1.30 (95% CI 1.01-1.68)). A high level (≥ 0.87 nmol/l) of MR-proADM, measured in stable state, was associated with a 30% increased risk of a severe AECOPD. With 25% of the patients having a high stable-state MR-proADM level and with 34% (cumulative proportion) of the patients having at least one severe AECOPD during the 3 years of follow-up, our results are clinically significant.
Interestingly, the study showed that a previous hospitalisation for an exacerbation (time to next hospitalization for AECOPD) within the 3 years of follow-up is stronger associated with the risk of exacerbation than MR-proADM (HR 1.90 (95% CI 1.51-2.40).
In chapter eight, the general discussion, we summarize our findings and provide a wider context. This chapter is concluded with future perspectives. The most important ones are summarized below.
In this thesis we conclude that MR-proADM is a valid and clinically relevant biomarker when measured in stable state. Although we believe this is an important conclusion we do not see it being used in practice yet. Although MR-proADM has been shown to be associated with mortality, we should now focus on how predicting mortality may lead to improved patient care. To further establish the usefulness of MR-proADM for stratifying risk we need to know its behaviour over time. Because COPD is an unstable disease and markers may vary over time for the same patient, it is interesting to know how the association and predictive value of MR-proADM with mortality and morbidity behaves when serial longitudinal measurements within one patient are made during the course of the disease. Use of a biomarker for selecting the right treatment is also an exciting future area. In the current thesis we show how this biomarker is related to disease mortality and morbidity. Modifying treatment type or intensity depending on the outcome would be the next step.
Finally, before MR-proADM may be used to monitor disease progression and response to therapy, further studies will need to address whether MR-proADM levels change with treatment (pharmacologic and/or nonpharmacologic) within COPD.
