Heart failure biomarkers: The importance of cardiac specificity Piek, Arnold
DOI:
10.33612/diss.146698618
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2020
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Piek, A. (2020). Heart failure biomarkers: The importance of cardiac specificity. University of Groningen. https://doi.org/10.33612/diss.146698618
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter 9
202
SUMMARY AND DISCUSSION
In this thesis, we investigated heart failure (HF) biomarkers at three different levels: 1) Biomarker biology, 2) Clinical performance of novel markers, and 3) Biomarkers as pharmacological treatment targets. Previously, the majority of biomarker research was performed in patient cohorts. These clinical studies are ideal to determine the diagnostic and prognostic quality of novel markers and can, when performed thoroughly as suggested in a scientific statement by the American Heart Association (AHA)1, provide high quality information. However, by studying biomarkers in human plasma samples only, uncertainties concerning biological function, tissue origin and cardiac and/or disease specificity remain. Therefore, despite major efforts of many researchers, the HF biomarker field has not progressed significantly and nowadays only two markers, including natriuretic peptides and troponin, are used clinically2,3. In this thesis, by combining biomarker studies in clinical HF cohorts with bioinformatics analyses and HF animal models, we aimed to provide more insight into the cardiac specificity and tissue origin of HF biomarkers.
Plasma biomarkers of myocardial fibrosis in HF
Ideally, the blood plasma level of a HF biomarker reflects a specific pathological process in the myocardium just before or during HF development. Myocardial fibrosis is a key pathological process in cardiac remodeling, and the search for a biomarker reflecting its severity and/or to distinguish between different types of fibrosis is still ongoing. For a better understanding of myocardial fibrosis, in chapter 2 we reviewed the current literature on this topic. Though fibrosis is a mechanism of the heart to cope with stress (e.g. maintain myocardial integrity after myocardial infarction or cope with increased wall stretch due to hypertension), fibrosis also increases cardiac wall stiffness, promotes the development of arrhythmias and limits nutrient and oxygen supply to the myocardium. This leads to increased stress on the healthy myocardium, resulting in pathological cardiomyocyte hypertrophy and, ultimately, may cause cardiomyocyte cell death. In turn, dead cardiomyocytes are replaced by fibrosis to ensure tissue integrity, which further increases stress on the healthy myocardium. Taken together, we postulate the existence of a vicious cycle of myocardial fibrosis and cell death. Substances involved in either cardiomyocyte death or fibrosis formation are not only present locally, but also might end up in the circulation and in that case could serve as biomarkers. Upon cardiomyocyte death, for example during myocardial infarction, cardiac specific troponins are passively released from the cardiomyocyte and can be measured in the blood plasma4. Since these death cardiomyocytes need to be replaced by fibrosis to maintain cardiac integrity, circulating levels of troponin do not only correspond with the amount of death cardiomyocytes and hence the size of the infarct5-7, but most likely also directly correspond to the degree of replacement fibrosis. Besides, other substances involved in fibrosis formation could also mark for the degree of these processes. Previously, suggestions have been made and investigated and include amongst others galectin-3 (Gal-3) and soluble suppression of
203 tumorigenicity 2 (sST2). Studies in HF animal models showed that both Gal-3 and sST2 are involved myocardial fibrosis and in HF patients plasma levels were associated with HF severity and outcome8-12. However, though mentioned in the HF guidelines, their clinical value is still under debate2,3.
Therefore, in chapter 3, we investigated a possible third candidate, namely human epididymis protein 4 (HE4). HE4 is involved in fibrosis formation13-15 and in patients with acute HF circulating HE4 levels were increased and, moreover, correlated with HF severity and outcome16. In this clinical study, we investigated HE4 plasma levels in patients with chronic HF. Compared to healthy age and gender matched controls, patients with chronic HF showed increased levels of HE4. HE4 levels were independently associated with HF outcome (combined endpoint of HF hospitalization and all-cause mortality) and correlated with fibrosis-associated proteins, which corroborates the suggested role of HE4 in fibrosis formation. However, in our study HE4 levels also showed a clear association with renal function. This association with renal function was also found in several other studies17-20 and, based on studies in animal models of renal failure, a role for HE4 in renal fibrosis formation was suggested15,21. Possibly, since HF and renal dysfunction often co-exist, this could explain the increased levels observed in our study. Besides renal function, HE4 is also associated with several other diseases, including several types of cancer like ovarian cancer and lung cancer22,23. It is interesting that increased levels of HE4 were observed in both HF patients and in cancer patients since recently the field of cardio-oncology has gained increased attention. Most of the theories in this field are based on the fact that both cardiovascular diseases and oncological diseases share risk factors (e.g. smoking, diet, obesity, diabetes, hypertension, alcohol usage, inflammatory status)24. Proof of concept of the existence of a cardio-oncological axis was provided by an elaborate animal experiment, which showed that the presence of a failing heart promotes tumor growth peripherally, probably through secreted factors25. Herein, α-1-antichymotrypsin (Serpin A3), fibronectin, paraoxonase 1 (PON-1), ceruloplasmin and α-1-antitrypsin (Serpina A1) were identified as possible tumor growth stimulating factors, and circulating levels of these proteins were increased in patients with HF. Based on our analyses, HE4 levels were predominantly dependent on renal function and, moreover, in a mouse model of cardiac pressure overload (transverse aortic constriction (TAC)), left ventricular HE4 gene expression was not increased after both 4 and 8 weeks of TAC (unpublished data). Therefore, it is unlikely that increased cardiac HE4 secretion in HF peripherally affects tumor-associated processes. Nevertheless, since circulating HE4 levels are increased in both HF and oncological diseases, this suggests the existence of disease homology, or at least that similar processes are present in both diseases. Finally, only 0.1% of total body HE4 expression of HE4 is located in the heart (unpublished data). Taken together we can conclude that, though associated with cardiovascular syndromes like HF, HE4 is not a cardiac specific marker.
204
The cardiac specificity of HF biomarkers
The observation that, besides with HF, HE4 is also associated with several other diseases could possible also be true for other HF biomarkers. Therefore, in chapter 4, we reviewed the current literature on novel HF biomarkers with a focus on cardiac specificity. As discussed above, proteins and other substances involved in pathophysiological processes (e.g. processes of cardiac remodeling, endothelial dysfunction, cardiac strain, hypertrophy, and fibrosis) are potential HF biomarker candidates. However, these processes, especially fibrosis and inflammation, are also observed in other diseases of non-cardiac organs and tissues. Therefore, circulating levels of these proteins could also be increased in association to other diseases or to comorbidities of HF. Moreover, it is likely that these proteins are also expressed in other tissues besides the heart, which makes these proteins less cardiac specific and therewith less disease specific. An extensive literature search showed that most novel HF biomarkers are, besides with HF, also associated with many other diseases and/or organs. We postulated that this complicates their clinical use as HF biomarkers, since it reduces their cardiac and/or disease specificity. Therefore, besides determining clinical performance of HF biomarkers in HF patient cohorts, we proposed that preclinical studies in animals should be standard in the investigation of novel HF biomarkers. These kinds of studies allow thorough molecular analysis, including total body expression analysis to determine cardiac specificity, and provide the opportunity to make direct correlations between circulating levels, cardiac function and processes of cardiac remodeling. It is impossible to retrieve these kinds of data based solely on clinical studies, since tissue and organ samples, including cardiac samples, are often not available. Moreover, animal HF models can be studied in an extremely regulated setting and are therefore ‘clean models of disease not affected by confounding factors like amongst others age, comorbidities, non-cardiac diseases and use of medication. Together, we hypothesised that a lack of cardiac specificity is the reason that most suggested novel HF biomarkers do not make it to the clinic.
To provide evidence for our cardiac specificity hypothesis, in chapter 5 we used the preclinical approach suggested in the previous chapter to investigate the tissue origin HF biomarkers. An elaborate animal study was performed using three different HF mouse models, including a myocardial infarction (MI) mouse model, a TAC mouse model of cardiac pressure overload and an obesity/hypertension mouse model with characteristics of heart failure with preserved ejection fraction (HFpEF). In total, four HF biomarkers were investigated, including galectin-3 (Gal-3), growth differentiation factor 15 (GDF-15), tissue inhibitor of metalloproteinase 1 (TIMP-1) and, as a reference, atrial natriuretic peptide (ANP). All markers were studied extensively in clinical cohorts and are associated with HF severity and outcome12,26-28 and play a role in cardiac remodeling11,29,30. In this preclinical study, we observed increased cardiac gene expression and protein levels of all markers in association to HF, but this did not directly translate into altered circulating levels of Gal-3, GDF-15 and TIMP-1, and therefore did not reflect cardiac status or severity of cardiac remodeling. Moreover, all three markers were highly expressed in other organs and tissues.
205 HF comorbidities, like lung congestion and obesity, had a major impact on their circulating levels. Only ANP showed cardiac specificity, and circulating levels adequately represented cardiac ANP production and showed a clear inverse association with cardiac function. Together, the data, as schematically depicted in Figure 1, provided evidence for our cardiac specificity hypothesis. Moreover, the study showed the importance of including preclinical studies in HF biomarker research, to provide a better understanding of HF biomarker biology. Our HF models allowed us to specifically determine cardiac contribution to HF biomarker blood plasma levels. The increased plasma levels of Gal-3, GDF-15 and TIMP-1 as observed in clinical association studies are probably the result of the contribution of peripheral organs in association to comorbidities and/or peripheral effects of the HF syndrome. Our results show that correlating circulating levels to specific processes of cardiac remodeling is, at least for the three investigated markers, impossible. Several studies in patients corroborate our data. For example, in a heart transplant study, compared to levels at pre heart transplantation, blood plasma levels of Gal-3 remained unaltered in patients post transplantation31. This while the purported Gal-3 producing entity, the failing heart, was fully replaced by a new heart. Moreover, post implantation of a left ventricular assistance device (LVAD), blood plasma levels of amongst others Gal-3, sST2 and GDF-15 improved, but so was true for the non-specific inflammation marker CRP32. The same applied to a study in which a prognostic proteomic panel of 17 biomarkers was developed of proteins which showed changes post heart transplant and should mark for improved function, but the proteins included in this panel were all associated with general processes like inflammation, coagulation and angiogenesis33. Therefore, these markers probably reflect an improvement in general health, but not specifically mark changes in cardiac remodeling. Unfortunately, we
Figure 1. The organ/tissue origin of heart failure biomarkers. In a schematic depiction the producing organs
are shown for a selection of biomarkers. ANP (as a representative of natriuretic peptides) is predominantly cardiac derived. Gal-3, GDF-15 and TIMP-1 are to some extent produced in the heart, but other organs and tissues are responsible for the majority of the production of these proteins. Thus, ANP is cardiac specific, and Gal-3, 15 and TIMP-1 are not cardiac specific. ANP=Atrial natriuretic peptide. Gal-3=Galectin-3. GDF-15=Growth differentiation factor 15. TIMP-1=Tissue inhibitor of metalloproteinase 1.
Gal- GDF-
206
were limited by the availability of commercial assays and antibodies and thus we were not able to measure other HF biomarkers. Nevertheless, probably these observations can be extrapolated to several other HF biomarkers, and it is likely that this concept also applies to biomarker research in non-cardiac diseases. Taken together, it seems that a high organ/tissue specificity (i.e. majority of total gene expression in the target organ) is an indispensable quality of a good biomarker, as is true for natriuretic peptides and troponins in the heart34,35, alanine aminotransferase (ALAT) in the liver36, pancreatic specific amylase and lipase in the pancreas37 and prostate specific antigen (PSA) in the prostate38. Proteins which are also (highly) expressed in other organs are more likely to be surrogate or general disease markers, at best.
Since most suggested novel markers, besides natriuretic peptides and cardiac troponin, appear to be not cardiac specific, the observations and results of the studies presented in the previous chapters urged us to further investigate cardiac specificity of biomarkers and to identify possible novel cardiac specific HF biomarker candidates. Therefore, in chapter 6 we designed a bioinformatics approach using an extensive RNAseq dataset containing gene expression in a large panel of organs and tissues. Herewith, the cardiac specificity of the total human genome and suggested novel HF biomarkers was determined. This analysis showed that, in accordance with our previous observations, natriuretic peptides are cardiac specific (> 99% specific), and markers like Gal-3, GDF-15 and TIMP-1 lack this specificity (< 10% specific). Based on our bioinformatics approach, two cardiac specific secreted protein encoding genes were identified. These included Dickkopf-3 (DKK3), which is approximately 44% cardiac specific, and bone morphogenic protein 10 (BMP-10), which is approximately 98% cardiac specific. For DKK3, assays were available that allowed DKK3 measurements. DKK3 gene expression and plasma concentrations were determined in three different HF models. In TAC and MI mice, cardiac DKK3 gene expression was increased, and in hearts of obesity/hypertension mice cardiac DKK3 gene expression was decreased. However, no changes in DKK3 plasma levels were observed in any of these models, indicating that there is no simple relationship between the cardiac DKK3 pool and the circulating DKK3 pool. DKK3 plasma levels were increased in HF patients, as compared to age and gender matched controls, and levels correlated with disease severity. However, levels were not independently associated with HF outcome and appeared to be more associated with kidney function, age and atrial fibrillation status. A HF biomarker can serve diagnostic and/or prognostic purposes. The absence of a simple relation between cardiac and circulating levels of DKK3, however, suggests circulating DKK3 will not correlate with degree of cardiac remodeling and is, hence, not suitable diagnostic HF biomarker. Moreover, the absence of an independent association with HF outcome implies DKK3 is not an independent prognostic biomarker. Thus far, our data showed that lack of cardiac specificity is probably the main reason why suggested HF biomarkers do not show clinical potential. However, though DKK3 is relatively cardiac specific and LV gene expression levels are increased in HFrEF models, there was no relation between LV DKK3 expression levels and circulating DKK3 levels.
207 Possibly, protein size plays herein a role since DKK3 is a relatively large protein (approximately 50kDa), which is significantly larger than the size of natriuretic peptides (approximately 4kDa), and this could hamper its passage over the tight cardiac endothelial barrier39. Based on our bioinformatics analysis, BMP-10 also seemed a possible HF biomarker candidate. Previous studies showed that BMP-10 plays a role in physiological an pathophysiological processes in the heart. During embryonic development, as shown by experiments in BMP-10 deficient mice, BMP-10 is essential for ventricular trabecularisation40,41. Also, in adult rats with hypertension induced cardiac hypertrophy, BMP-10 expression levels were increased42. Finally, a cardioprotective function was suggested for BMP-10, since cardiac fibrosis upon isoproterenol infusion was reduced in mice overexpressing BMP-10 and in mice administered with recombinant BMP-1043. Thus, BMP-10 is a cardiac specific secreted protein encoding gene which shows increased expression in association to cardiac disease, and therefore could be a HF biomarker. However, unfortunately we were not able to measure BMP-10 with adequate precision and reproducibility, a problem that has been described before and is likely to be caused by interference of BMP-9 with BMP-1044. Thus, for now it remains unknown whether BMP-10 could serve as a HF biomarker. Together, our data illustrate several important characteristics of a HF biomarker of cardiac remodeling: 1) Cardiac specificity, 2) Altered cardiac expression levels in disease, 3) Release towards the circulation, and 4) The availability of assays to determine plasma concentrations. Probably, very high cardiac specificity is necessary to be able to relate circulating levels of a biomarker to degree of cardiac remodeling, as is true for natriuretic peptides.
Previous studies investigated DKK3 as a biomarker in association to cardiovascular disease and renal disease, and in chapter 6 we observed associations between DKK3 plasma concentrations and several cardiovascular risk factors. Therefore, in chapter 7 the associations of DKK3 with cardiovascular risk factors, cardiovascular disease (CVD) and chronic kidney disease (CKD) were investigated in the Prevention of Renal and Vascular ENd-stage Disease (PREVEND) cohort. This is a large general population cohort designed to study cardiovascular risk factors, renal function, and both prevalent and new-onset CVD and CKD45. Similar as to in the clinical cohort described in chapter 6, in this general population cohort DKK3 plasma concentrations were associated with several cardiovascular risk factors. However, neither independent associations of DKK3 with prevalent CVD and CKD, nor with new-onset CVD and CKD were observed. Whilst previous work showed that plasma DKK3 levels are inversely associated with atherosclerosis46, in our study no independent association with either prevalent or new-onset CVD was observed. Though our study was larger and therewith better powered than this previous study, our subjects were also on average younger, and since DKK3 is strongly associated with age47 this could possibly explain the observed differences. Moreover, previous studies showed that urinary DKK3 concentrations are independently associated with future loss of renal function in patients with pre-existing CKD, and in patients about to undergo major cardiac surgery48,49. Since
208
urinary DKK3 originates almost exclusively from renal tubules, this makes DKK3 a highly specific marker for renal tubular disease48-51. On the other hand, the blood plasma DKK3 pool is dictated by multiple organs and tissues, and hence plasma DKK3 is probably not a specific but a more generic marker. The distinct difference in tissue specificity between urinary DKK3 and blood plasma DKK3 nicely illustrates the main message of this thesis and is an excellent example of the fact that biomarkers should be organ/tissue/disease specific. As a next step, it would be interesting to determine urinary DKK3 concentrations in the PREVEND cohort, consisting predominantly of non-CKD subjects. This could provide information on whether urinary DKK3 can predict for new-onset kidney disease in the general population.
Specific targeting of non-cardiac specific HF biomarkers
As said, most novel HF biomarkers are associated with general processes like inflammation and fibrosis. Whilst we showed that for diagnostic purposes and patient stratification these markers are inferior, targeting these markers as a means of HF therapy is still possible. Currently, according to guidelines, HF patients are treated with either an angiotensin converting enzyme inhibitor (ACE-inhibitor) or angiotensin receptor blocker (ARB), combined with a beta blocker and diuretics2,3. These treatment options are effective for patients with heart failure with reduced ejection fraction (HFrEF), but have not improved outcome in patients with HFpEF. It is suggested that HFpEF is the result of a systemic inflammatory state caused by HF comorbidities like obesity, diabetes and hypertension52. We hypothesized that systemically targeting the inflammation associated with these comorbidities could therefore have protective effects throughout the body, including the heart. Therefore, in chapter 8 we aimed to target this systemic inflammation by inhibiting myeloperoxidase (MPO), a non-cardiac specific inflammatory enzyme associated with obesity, hypertension, diabetes and HF53-59, using the novel oral available MPO-inhibitor AZM198. This was investigated in an obesity/hypertension mouse model that displays characteristics of HFpEF. In this model, MPO blood plasma levels were clearly increased in obese/hypertensive mice, and we showed AZM198 can reduce both MPO activity and circulating MPO blood plasma levels. MPO inhibition resulted in reduced weight gain, attenuated adipose tissue inflammation, and reduced the degree of non-alcoholic steatohepatitis (NASH) in the obese/hypertensive mice, and this was independent of food intake. Though the severity of these cardiovascular risk factors and systemic inflammation were attenuated, the HFpEF phenotype of this model was not relieved by AZM198 treatment. Still, we postulate that attenuating the severity of cardiovascular risk factors by prolonged AZM198 treatment could have value in cardiovascular risk reduction. Since MPO is not tissue specific, the effects of targeting MPO are also multitude. Besides attenuating obesity and liver steatosis, as shown by our study and MPO-knockout studies60,61, targeting MPO also improves glucose intolerance60 and endothelial dysfunction62, and, though not observed in our study, can attenuate cardiac remodeling63,64. Therefore, in general it can be concluded that in systemically targeting non-cardiac specific HF biomarkers beneficial systemic effects can possibly be observed.
209 However, this also warrants the need for thorough checking for detrimental systemic side-effects alongside the beneficial side-effects in these kinds of therapies.
FUTURE PERSPECTIVES
The impasse in HF biomarkers research: More complex than cardiac specificity alone In this thesis, we focused on organ and tissue specificity of biomarkers, and concluded that a lack of cardiac specificity of HF biomarkers explains why most novel HF biomarkers fail to provide clinical value. This is explained by the fact that a lack of cardiac specificity hampers the possibility to correlate circulating biomarker levels to degree of cardiac remodeling. Indeed, cardiac specificity of markers is important, and the high relative cardiac expression of natriuretic peptides65 and cardiac troponin4 seems to be the major reason for their success. Interestingly, whilst being clearly cardiac enriched/specific, the biomarker candidate DKK3 investigated in this thesis showed no correlation between cardiac expression and circulating levels. Our data again confirm the importance and value of the natriuretic peptides and troponins as cardiac biomarkers. Moreover, we think chances are low that additional protein markers will be found that can provide additional and valuable cardiac specific information, and based on our analyses only the secreted protein BMP-10 might have biomarker potential. Also, it seems that the heart is not a major secreting organ, as was previously proposed. Organs as the lungs, liver, spleen, intestines and kidneys have, inherent to their function, a clear association to the circulating blood on the level of active secretion, reabsorption, clearance, and exchange of fluids and molecules. In the meantime, the endocrine function of the heart is limited. The heart is characterized by the presence of a tight endothelial barrier39, and probably only small proteins can pass this barrier. Troponins are released into the circulation only on active damage of cardiomyocytes4 and thus do not count as secretory proteins. Till date, only clear evidence exists for natriuretic peptides, which are relatively small proteins, being secreted from the myocardium66,67. Together, this suggests that, besides cardiac specificity, myocardial release of markers is a fundamental characteristic of a cardiac specific HF biomarker. In future biomarker studies, we should not simply assume that cardiac secretion is present. Our data show that it is indispensable to prove that the circulating proteins actually originate from the heart and not from peripheral tissues, especially when claiming that circulating levels of a marker correlate with degree of cardiac remodeling.
HF biomarkers: Clinical utility revisited
Previously, criteria for the ideal biomarker were suggested, and finding clinical utility should be the main focus when studying novel biomarkers1,68-71. Naturally, this does not only apply to the field of cardiology but to all biomarker research. In the past decades, many potential novel HF biomarkers were identified. Whilst these markers were associated with HF severity and outcome, based on our results the non-cardiac specific markers amongst them cannot
210
be related to processes of cardiac remodeling. However, HF is a syndrome of the total body and possibly markers of peripheral effects of HF and/or HF comorbidities can be identified. For example markers of processes associated with 1) Lung congestion (Candidates: e.g. TIMP-1, GDF-15), 2) Systemic inflammation (Candidates: e.g. MPO, IL-6, GDF-15), 3) Obesity
Figure 2. Updated hypothesized and schematic depiction of patient stratification based on protein blood plasma biomarkers. Based on the observations in this thesis and observations by others described in the
literature, possible applications of suggested (HF) biomarkers are summarised in this figure. This figure is an update of figure 2 in the introduction. BMI=Body mass index. DKK3=Dickkopf 3. ECG=Electrocardiography. eGFR=Estimated glomerular filtration rate. Gal-3=Galectin-3. GDF-15=Growth differentiation factor 15. HE4=Human epididymis protein 4. HF=Heart failure. H-FABP=Heart-type fatty acid binding protein. HsTN=High sensitive cardiac troponin. IGFBP-7=Insulin growth factor binding protein 7. IL-6=Interleukin 6. MPO=Myeloperoxidase. PENK=Proenkephalin. TIMP-1=Tissue inhibitor of metalloproteinase 1.
TIMP-1 GDF-15 IL-6 IGFBP-7 Metabolites HsTn H-FABP Imaging Imaging ECG ECG MPO DKK3 HE4 PENK eGFR Gal-3 BMI Natriuretic peptides Natriuretic peptides
211 (Candidate: Gal-3), 4) Atrial fibrillation (Candidate: DKK3), 5) Renal failure (Candidates: DKK3, HE4, PENK72) and 6) Metabolic dysfunction (Candidates: Metabolites, IGFBP-7). Based on the current knowledge and data from this thesis, a schematic depiction of patient stratification based on both cardiac specific and non-cardiac specific biomarkers is shown in Figure 2. Additional research is needed to determine the additional clinical value and applicability of these markers. Besides for patient stratification, non-cardiac specific biomarkers could serve as pharmacological treatment targets. For most non-specific markers proof of involvement in cardiac remodeling exists and thus systemic targeting could result in treatment effects at the cardiac level, but due to their non-cardiac specificity also non-cardiac effects can be expected. In this thesis we showed that several HF biomarkers are more abundantly expressed in non-cardiac tissues than in the heart itself and rather associated with HF comorbidities than with cardiac remodeling. Therefore, the non-cardiac effects in pharmacological targeting of markers might be more abundant as compared to the cardiac effects. Possibly, targeting non-cardiac specific biomarkers could have beneficial effects on cardiovascular risk and might eventually have a place in cardiovascular risk management in patients in which comorbidities are difficult to treat by non-pharmacological treatment options. The MPO-inhibition study included in this thesis is an example of such a study. To diminish cardiac remodeling, on the other hand, cardiac specific targeting is likely to be more effective as compared to targeting of a non-cardiac specific factor.
Reformulated criteria for the ideal biomarker
In this thesis, we focused on circulating proteins as HF biomarkers, but also non-proteins like for example microRNAs and metabolites are possible HF biomarkers and these have been studied extensively. As is true for protein biomarkers, it is also indispensable to investigate tissue origin of microRNAs and metabolites to fully understand their biology and clinical value. For example, several microRNAs that seemed promising based on clinical studies appeared not to be related to cardiac function in animal models of HF73. Apparently, microRNA levels are also affected by comorbidities and extra-cardiac production. Together, tissue origin, tissue specificity and tissue release appear to be important qualities of HF biomarkers. Probably, this applies to all types of markers (including proteins, microRNAs and others) and thus this rule can be applied not only in HF biomarker research but in the entire biomarker field. Based on the results and data of this thesis, in Figure 3 we reformulated the previously suggested criteria for the ideal biomarker1,68-71, now also with more emphasis on tissue origin, tissue specificity and tissue release. We think that future biomarker studies should use these three pillars, namely organ specificity, analytical quality and clinical utility, as a central backbone. This will improve the quality of biomarker research and, hopefully, result in the discovery of novel markers which can be applied in the clinic.
The value of reusing databases
In this thesis, a bioinformatics approach was used. Herein, we aimed to use publically available datasets to answer our research questions, instead of regaining the data ourselves.
212
We used these datasets to determine the cardiac specificity of the total human genome and the secreted protein encoding genome. Possibly, this dataset can also be used to answer other research questions. For example, tissue specific genes could be selected that can serve as specific pharmacological treatment targets. Moreover, similarly as we did for HF biomarkers, the same approach could be applied to investigate (novel) biomarkers for other organs/diseases. Databases composed and material gathered to answer a certain research question could be reused to answer alternative research questions, can serve as a guide to improve research questions prior to performing the actual experiment, or can be used as pilot data. This can save time and money, and can reduce the use of animal and patient material, thereby improving the efficiency of research.
Figure 3. Reformulated criteria for the ideal biomarker. Overview of criteria that characterize a high quality
biomarker. These criteria can be applied to cardiac biomarkers, but are also applicable for biomarkers for other organs and tissues, and in non-cardiac diseases. Moreover, they also apply to HF biomarkers of peripheral HF effects. Based on previous suggestions1,68-71 and this thesis.
Organ specificit
• High tissue specificity
• Tissue release towards circulation • Proven relation between disease
process and tissue origin
Anal tical qualit
• Rapid and accurate detection • Long half-life after withdrawal • Low patient burden
• Low costs for measuring
Clinical utilit
• High disease sensitivity and specificity • Sensitive to disease severity alterations • Early disease detection
• Improves clinical outcome
213
CONCLUSION
In this thesis, three different approaches to study biomarkers (bioinformatics, animal models and clinical cohorts) were combined to improve our knowledge on HF biomarker biology, clinical utility, and HF biomarkers as treatment targets. The results of our unique and elaborate approach shed new light on the biomarker field as a whole. We showed that many novel HF biomarkers lack cardiac specificity and are produced by non-cardiac organs and tissues. Relating the levels of these markers to processes of cardiac remodeling seems therewith impossible. Nevertheless, non-specific markers might still have clinical value, for example as markers of peripheral effects of HF or as pharmacological treatment targets, as we showed by inhibiting the inflammatory marker (MPO) which resulted in the improvement of cardiovascular comorbidities. Finally, this thesis showed that for future biomarker studies it is indispensable to investigate tissue origin of markers in order to improve their specificity and therewith their clinical utility.
214
REFERENCES
1. 1. Hlatky MA, Greenland P, Arnett DK, et al. Criteria for evaluation of novel markers of cardiovascular risk: A scientific statement from the american heart association. Circulation. 2009;119(17):2408-2416.
2. Ponikowski P, Voors AA, Anker SD, et al. 2016 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure: The task force for the diagnosis and treatment of acute and chronic heart failure of the european society of cardiology (ESC)developed with the special contribution of the heart failure association (HFA) of the ESC. Eur Heart J. 2016;37(27):2129-2200.
3. Yancy CW, Jessup M, Bozkurt B, et al. 2017 ACC/AHA/HFSA focused update of the 2013 ACCF/AHA
guideline for the management of heart failure: A report of the american college of cardiology/american heart association task force on clinical practice guidelines and the heart failure society of america. J Card Fail. 2017;23(8):628-651.
4. Omland T, Rosjo H, Giannitsis E, Agewall S. Troponins in heart failure. Clin Chim Acta. 2015;443:78-84. 5. Giannitsis E, Steen H, Kurz K, et al. Cardiac magnetic resonance imaging study for quantification of infarct
size comparing directly serial versus single time-point measurements of cardiac troponin T. J Am Coll Cardiol. 2008;51(3):307-314.
6. Steen H, Giannitsis E, Futterer S, Merten C, Juenger C, Katus HA. Cardiac troponin T at 96 hours after acute myocardial infarction correlates with infarct size and cardiac function. J Am Coll Cardiol.
2006;48(11):2192-2194.
7. Licka M, Zimmermann R, Zehelein J, Dengler TJ, Katus HA, Kubler W. Troponin T concentrations 72 hours
after myocardial infarction as a serological estimate of infarct size. Heart. 2002;87(6):520-524.
8. Kakkar R, Lee RT. The IL-33/ST2 pathway: Therapeutic target and novel biomarker. Nat Rev Drug Discov.
2008;7(10):827-840.
9. Pascual-Figal DA, Januzzi JL. The biology of ST2: The international ST2 consensus panel. Am J Cardiol. 2015;115(7 Suppl):3B-7B.
10. Pascual-Figal DA, Ordonez-Llanos J, Tornel PL, et al. Soluble ST2 for predicting sudden cardiac death in patients with chronic heart failure and left ventricular systolic dysfunction. J Am Coll Cardiol.
2009;54(23):2174-2179.
11. Yu L, Ruifrok WP, Meissner M, et al. Genetic and pharmacological inhibition of galectin-3 prevents cardiac remodeling by interfering with myocardial fibrogenesis. Circ Heart Fail. 2013;6(1):107-117.
12. de Boer RA, Daniels LB, Maisel AS, Januzzi JL,Jr. State of the art: Newer biomarkers in heart failure. Eur J Heart Fail. 2015;17(6):559-569.
13. Kirchhoff C, Habben I, Ivell R, Krull N. A major human epididymis-specific cDNA encodes a protein with sequence homology to extracellular proteinase inhibitors. Biol Reprod. 1991;45(2):350-357.
14. Bingle CD, Vyakarnam A. Novel innate immune functions of the whey acidic protein family. Trends Immunol. 2008;29(9):444-453.
15. LeBleu VS, Teng Y, O'Connell JT, et al. Identification of human epididymis protein-4 as a fibroblast-derived mediator of fibrosis. Nat Med. 2013;19(2):227-231.
16. de boer R, Cao Q, Postmus D, et al. The WAP four-disulfide core domain protein HE4: A novel biomarker for heart failure. JACC. 2013;1(2):164-169.
17. Hertlein L, Stieber P, Kirschenhofer A, et al. Human epididymis protein 4 (HE4) in benign and malignant diseases. Clin Chem Lab Med. 2012;50(12):2181-2188.
18. Escudero JM, Auge JM, Filella X, Torne A, Pahisa J, Molina R. Comparison of serum human epididymis protein 4 with cancer antigen 125 as a tumor marker in patients with malignant and nonmalignant diseases. Clin Chem. 2011;57(11):1534-1544.
19. Kappelmayer J, Antal-Szalmas P, Nagy B,Jr. Human epididymis protein 4 (HE4) in laboratory medicine and an algorithm in renal disorders. Clin Chim Acta. 2015;438:35-42.
20. Bolstad N, Oijordsbakken M, Nustad K, Bjerner J. Human epididymis protein 4 reference limits and natural variation in a nordic reference population. Tumour Biol. 2012;33(1):141-148.
215
21. Wan J, Wang Y, Cai G, et al. Elevated serum concentrations of HE4 as a novel biomarker of disease severity and renal fibrosis in kidney disease. Oncotarget. 2016;7(42):67748-67759.
22. Hellstrom I, Raycraft J, Hayden-Ledbetter M, et al. The HE4 (WFDC2) protein is a biomarker for ovarian carcinoma. Cancer Res. 2003;63(13):3695-3700.
23. Galgano MT, Hampton GM, Frierson HF,Jr. Comprehensive analysis of HE4 expression in normal and malignant human tissues. Mod Pathol. 2006;19(6):847-853.
24. Koene RJ, Prizment AE, Blaes A, Konety SH. Shared risk factors in cardiovascular disease and cancer. Circulation. 2016;133(11):1104-1114.
25. Meijers WC, Maglione M, Bakker SJL, et al. Heart failure stimulates tumor growth by circulating factors. Circulation. 2018;138(7):678-691.
26. Jungbauer CG, Riedlinger J, Block D, et al. Panel of emerging cardiac biomarkers contributes for prognosis rather than diagnosis in chronic heart failure. Biomark Med. 2014;8(6):777-789.
27. de Boer RA, Lok DJ, Jaarsma T, et al. Predictive value of plasma galectin-3 levels in heart failure with reduced and preserved ejection fraction. Ann Med. 2011;43(1):60-68.
28. Kempf T, von Haehling S, Peter T, et al. Prognostic utility of growth differentiation factor-15 in patients with chronic heart failure. J Am Coll Cardiol. 2007;50(11):1054-1060.
29. Kempf T, Zarbock A, Widera C, et al. GDF-15 is an inhibitor of leukocyte integrin activation required for survival after myocardial infarction in mice. Nat Med. 2011;17(5):581-588.
30. Moore L, Fan D, Basu R, Kandalam V, Kassiri Z. Tissue inhibitor of metalloproteinases (TIMPs) in heart failure. Heart Fail Rev. 2012;17(4-5):693-706.
31. Grupper A, Nativi-Nicolau J, Maleszewski JJ, et al. Circulating galectin-3 levels are persistently elevated after heart transplantation and are associated with renal dysfunction. JACC Heart Fail. 2016;4(11):847-856.
32. Ahmad T, Wang T, O'Brien EC, et al. Effects of left ventricular assist device support on biomarkers of cardiovascular stress, fibrosis, fluid homeostasis, inflammation, and renal injury. JACC Heart Fail. 2015;3(1):30-39.
33. Hollander Z, Lazarova M, Lam KK, et al. Proteomic biomarkers of recovered heart function. Eur J Heart Fail. 2014;16(5):551-559.
34. Levine B, Kalman J, Mayer L, Fillit HM, Packer M. Elevated circulating levels of tumor necrosis factor in severe chronic heart failure. N Engl J Med. 1990;323(4):236-241.
35. Ricchiuti V, Voss EM, Ney A, Odland M, Anderson PA, Apple FS. Cardiac troponin T isoforms expressed in renal diseased skeletal muscle will not cause false-positive results by the second generation cardiac troponin T assay by boehringer mannheim. Clin Chem. 1998;44(9):1919-1924.
36. Dufour DR, Lott JA, Nolte FS, Gretch DR, Koff RS, Seeff LB. Diagnosis and monitoring of hepatic injury. I. performance characteristics of laboratory tests. Clin Chem. 2000;46(12):2027-2049.
37. Ismail OZ, Bhayana V. Lipase or amylase for the diagnosis of acute pancreatitis? Clin Biochem. 2017;50(18):1275-1280.
38. De Angelis G, Rittenhouse HG, Mikolajczyk SD, Blair Shamel L, Semjonow A. Twenty years of PSA: From prostate antigen to tumor marker. Rev Urol. 2007;9(3):113-123.
39. Aird WC. Phenotypic heterogeneity of the endothelium: I. structure, function, and mechanisms. Circ Res. 2007;100(2):158-173.
40. Chen H, Shi S, Acosta L, et al. BMP10 is essential for maintaining cardiac growth during murine cardiogenesis. Development. 2004;131(9):2219-2231.
41. Somi S, Buffing AA, Moorman AF, Van Den Hoff MJ. Expression of bone morphogenetic protein-10 mRNA during chicken heart development. Anat Rec A Discov Mol Cell Evol Biol. 2004;279(1):579-582.
42. Nakano N, Hori H, Abe M, et al. Interaction of BMP10 with tcap may modulate the course of hypertensive cardiac hypertrophy. Am J Physiol Heart Circ Physiol. 2007;293(6):H3396-403.
43. Qu X, Liu Y, Cao D, et al. BMP10 preserves cardiac function through its dual activation of SMAD-mediated and STAT3-mediated pathways. J Biol Chem. 2019;294(52):19877-19888.
216
44. Tillet E, Ouarne M, Desroches-Castan A, et al. A heterodimer formed by bone morphogenetic protein 9 (BMP9) and BMP10 provides most BMP biological activity in plasma. J Biol Chem. 2018;293(28):10963-10974.
45. Diercks GF, van Boven AJ, Hillege HL, et al. Microalbuminuria is independently associated with ischaemic electrocardiographic abnormalities in a large non-diabetic population. the PREVEND (prevention of REnal and vascular ENdstage disease) study. Eur Heart J. 2000;21(23):1922-1927.
46. Yu B, Kiechl S, Qi D, et al. A cytokine-like protein dickkopf-related protein 3 is atheroprotective. Circulation. 2017;136(11):1022-1036.
47. Zenzmaier C, Sklepos L, Berger P. Increase of dkk-3 blood plasma levels in the elderly. Exp Gerontol. 2008;43(9):867-870.
48. Schunk SJ, Zarbock A, Meersch M, et al. Association between urinary dickkopf-3, acute kidney injury, and subsequent loss of kidney function in patients undergoing cardiac surgery: An observational cohort study. Lancet. 2019.
49. Zewinger S, Rauen T, Rudnicki M, et al. Dickkopf-3 (DKK3) in urine identifies patients with short-term risk of eGFR loss. J Am Soc Nephrol. 2018;29(11):2722-2733.
50. Schunk SJ, Speer T, Petrakis I, Fliser D. Dickkopf 3-a novel biomarker of the 'kidney injury continuum'. Nephrol Dial Transplant. 2020.
51. Federico G, Meister M, Mathow D, et al. Tubular dickkopf-3 promotes the development of renal atrophy and fibrosis. JCI Insight. 2016;1(1):e84916.
52. Paulus WJ, Tschope C. A novel paradigm for heart failure with preserved ejection fraction: Comorbidities drive myocardial dysfunction and remodeling through coronary microvascular endothelial inflammation. J Am Coll Cardiol. 2013;62(4):263-271.
53. Fu X, Kassim SY, Parks WC, Heinecke JW. Hypochlorous acid generated by myeloperoxidase modifies adjacent tryptophan and glycine residues in the catalytic domain of matrix metalloproteinase-7
(matrilysin): An oxidative mechanism for restraining proteolytic activity during inflammation. J Biol Chem. 2003;278(31):28403-28409.
54. Fu X, Kassim SY, Parks WC, Heinecke JW. Hypochlorous acid oxygenates the cysteine switch domain of pro-matrilysin (MMP-7). A mechanism for matrix metalloproteinase activation and atherosclerotic plaque rupture by myeloperoxidase. J Biol Chem. 2001;276(44):41279-41287.
55. Tromp J, Khan MA, Klip IT, et al. Biomarker profiles in heart failure patients with preserved and reduced ejection fraction. J Am Heart Assoc. 2017;6(4):10.1161/JAHA.116.003989.
56. Nauseef WM. Myeloperoxidase in human neutrophil host defence. Cell Microbiol. 2014;16(8):1146-1155. 57. Van der Zwan LP, Scheffer PG, Dekker JM, Stehouwer CD, Heine RJ, Teerlink T. Hyperglycemia and
oxidative stress strengthen the association between myeloperoxidase and blood pressure. Hypertension. 2010;55(6):1366-1372.
58. Gomez Garcia A, Rivera Rodriguez M, Gomez Alonso C, Rodriguez Ochoa DY, Alvarez Aguilar C.
Myeloperoxidase is associated with insulin resistance and inflammation in overweight subjects with first-degree relatives with type 2 diabetes mellitus. Diabetes Metab J. 2015;39(1):59-65.
59. Tang WH, Brennan ML, Philip K, et al. Plasma myeloperoxidase levels in patients with chronic heart failure. Am J Cardiol. 2006;98(6):796-799.
60. Wang Q, Xie Z, Zhang W, et al. Myeloperoxidase deletion prevents high-fat diet-induced obesity and insulin resistance. Diabetes. 2014;63(12):4172-4185.
61. Pulli B, Ali M, Iwamoto Y, et al. Myeloperoxidase-hepatocyte-stellate cell cross talk promotes hepatocyte injury and fibrosis in experimental nonalcoholic steatohepatitis. Antioxid Redox Signal. 2015;23(16):1255-1269.
62. Cheng D, Talib J, Stanley CP, et al. Inhibition of MPO (myeloperoxidase) attenuates endothelial
dysfunction in mouse models of vascular inflammation and atherosclerosis. Arterioscler Thromb Vasc Biol. 2019;39(7):1448-1457.
63. Rudolph V, Andrie RP, Rudolph TK, et al. Myeloperoxidase acts as a profibrotic mediator of atrial fibrillation. Nat Med. 2010;16(4):470-474.
217
64. Mollenhauer M, Friedrichs K, Lange M, et al. Myeloperoxidase mediates postischemic arrhythmogenic ventricular remodeling. Circ Res. 2017;121(1):56-70.
65. Levin ER, Gardner DG, Samson WK. Natriuretic peptides. N Engl J Med. 1998;339(5):321-328.
66. Kinnunen P, Vuolteenaho O, Ruskoaho H. Mechanisms of atrial and brain natriuretic peptide release from rat ventricular myocardium: Effect of stretching. Endocrinology. 1993;132(5):1961-1970.
67. Liang F, Wu J, Garami M, Gardner DG. Mechanical strain increases expression of the brain natriuretic peptide gene in rat cardiac myocytes. J Biol Chem. 1997;272(44):28050-28056.
68. Tijsen AJ, Pinto YM, Creemers EE. Circulating microRNAs as diagnostic biomarkers for cardiovascular diseases. Am J Physiol Heart Circ Physiol. 2012;303(9):H1085-95.
69. Weber JA, Baxter DH, Zhang S, et al. The microRNA spectrum in 12 body fluids. Clin Chem. 2010;56(11):1733-1741.
70. Nadar SK, Shaikh MM. Biomarkers in routine heart failure clinical care. Card Fail Rev. 2019;5(1):50-56. 71. Ahmad T, Fiuzat M, Pencina MJ, et al. Charting a roadmap for heart failure biomarker studies. JACC Heart
Fail. 2014;2(5):477-488.
72. Emmens JE, Ter Maaten JM, Damman K, et al. Proenkephalin, an opioid system surrogate, as a novel comprehensive renal marker in heart failure. Circ Heart Fail. 2019;12(5):e005544.
73. Vegter EL, Ovchinnikova ES, Sillje HHW, et al. Rodent heart failure models do not reflect the human circulating microRNA signature in heart failure. PLoS One. 2017;12(5):e0177242.
