Clinical outcomes of the participants according to LVEDP
The 3-year clinical outcomes of patients in the LVEDP >15 mmHg and LVEDP ≤15 mmHg groups are presented in Table 5. There were no significant differences between the two groups. Although the rate of HF hospitalisation tended to be higher in the LVEDP >15 mmHg group, the difference was not statistically significant (15.4% vs 21.1%, p = 0.58). Time-dependent Kaplan–Meier analyses and log-rank tests showed no difference in the rate of HF hospitalisation between the two groups (Fig 1b).
DISSCUSSION
In the present study, we evaluated the relationship of systolic dysfunction as well as the level of LVEDP with the 3-year clinical outcomes in patients with chronic AF. The main results of our study were as follows. First, systolic dysfunction was related with adverse outcomes such as heart failure (HF) hospitalisation and the occurrence of AF with rapid ventricular rhythm during the 3-year observation period. In addition, systolic dysfunction (LVEDP < 50%) was an important independent predictor of HF hospitalisation. Second, elevated LVEDP, which is the key finding in patients with diastolic dysfunction, was not associated with mid-term adverse outcomes. To our knowledge, this is the first study to assess the effect of increased LVEDP on mid-term outcomes in patients with chronic AF. These results highlight the importance of paying more attention to a low EF detected by echocardiography than to diastolic parameters in patients with chronic AF.
AF is one of the most common forms of cardiac arrhythmia requiring treatment, and it is associated with increased rates of mortality and hospitalisation.1 4 Some previous studies of patients with AF also showed that patients with systolic dysfunction more frequently exhibit all-cause mortality than do patients with a preserved EF.13-15. In a previous study, AF was theoretically associated with the development of rapid rhythm, loss of atrial contractility, and left atrial enlargement. These changes can lead to left ventricular wall thickening, neurohormonal changes, and systolic dysfunction, which is an important predictor of adverse cardiac outcomes.5 16 However, in the present study, the mortality rates for patients with REF and those with PEF were comparable; this was probably because of the relatively short observation period, small number of patients in the REF group, and event rates. Although many studies have demonstrated a high mortality rate and increased HF hospitalisation in patients with AF and reduced LVEF, a few studies showed that reduced LVEF did not increase the HF hospitalisation or stroke rate for patients with AF and HF, relative to the rates for patients with AF and diastolic dysfunction.17 18 In the present study, the HF hospitalisation rate was significantly higher in the REF group than in the PEF group, with the rate being as high as 29.2% in the former group. These findings could be carefully explained by the relatively old age of the enrolled patients and the presence of chronic persistent AF in all patients (patients with paroxysmal AF were excluded).
Notably, the prognostic utility of LVEF measurements for patients with AF is well established. However, in a very recent study, the authors compared the long-term prognostic implications of AF and sinus rhythm in patients with acute or chronic HF stratified according to the level of LVEF and found that AF was not associated with worse cardiovascular outcomes in patients with reduced LVEF.19 These varied results could be caused by the complexity of AF. Therefore, prediction of our results was difficult, and we carefully estimated that elevated LVEDP would be related with adverse clinical outcomes. Studies that investigated the relationship between LVEDP and clinical outcomes are very limited because of the invasiveness of measurement methods. Previously, some investigators reported that elevated LVEDP is an important risk factor for adverse clinical outcomes in patients with acute myocardial infarction.20-22 To our knowledge, however, no study has evaluated the association between LVEDP and clinical outcomes in patients with chronic AF. Only left ventricular diastolic dysfunction (LVDD), measured by non-invasive tests such as echocardiography, was proven to be correlated with adverse clinical events such as hospitalisation and stroke.23-25However, in the setting of AF, atrial contractile a synchronicity complicates the assessment of LVDD using classic echocardiographic parameters.26 27 The physiological hallmarks of LVDD are impaired relaxation and reduced diastolic compliance, which increase the LV filling pressure. 28 29 Among patients with normal systolic function, elevated LVEDP could represent early diastolic dysfunction or could be the main physiological consequence of diastolic dysfunction.30 We not only determined the relationship between LVEDP and clinical outcomes in patients with chronic AF and a preserved EF but also clarified the relationship between diastolic dysfunction and adverse clinical outcomes such as mortality, stroke, and HF hospitalisation. We found that echocardiographic parameters that represent diastolic function, such as E/e’ and E/Vp, were worse in the LVEDP> 15 mmHg group. However, during the 3-year observation period, no significant differences were found in terms of adverse clinical outcomes between the LVEDP > 15 mmHg and LVEDP ≤ 15 mmHg groups. We speculate that chronic persistent AF leads to irreversible, chronic structural and neurohormonal changes that could affect both LVEDP and the clinical outcomes.
This study has some limitations, the main one being the retrospective design without randomization. The number of patients was small and the follow-up period relatively short. Because it was a single-center study, it was difficult to collect an adequate number of patients with LV pressure wave data and recruit a relatively homogenous group with chronic persistent AF without valvular heart disease. Long-term clinical follow-up will be necessary after recruitment of more patients, particularly those with reduced LVEF. Even if we performed coronary angiography on the basis of its broad indications, LVEDP measurements were recorded for selective patients with symptoms such as unexplained chest discomfort, unexplained broad-spectrum dyspnoea, and typical and atypical chest pain. Further studies with long-term follow-up periods are necessary after the recruitment of more patients for clarification of the relationship between diastolic dysfunction or elevated LVEDP and clinical outcomes. Furthermore, the number of patients treated with oral anticoagulants was very low in our study, primarily because our hospital is located in a rural area and the insurance system in Korea did not cover treatment with direct oral anticoagulants during the study period. Thus, the level of education and medical compliance were low, and the patients frequently refused warfarin treatment because of the inconvenience of blood sampling. Finally, LVEDP is affected by multiple factors such as dehydration and volume overload (e.g. combination of chronic kidney disease and anaemia), and elevated LVEDP cannot represent diastolic dysfunction, particularly in the setting of chronic AF.
CONCLUSION
Despite the study limitations, the findings are clinically meaningful because we investigated, for the first time, the relationship between elevated LVEDP and mid-term clinical outcomes along with the relationship between systolic dysfunction and clinical outcomes in patients with chronic AF. Also, this study proved that diastolic dysfunction itself does not appear to be a prognostic factor for hospitalization due to heart failure. Clinicians should apply these results during the treatment of patients with chronic AF, with careful evaluation of systolic function and prevention of risk factors that lead to LV systolic dysfunction.
CONTRIBUTORS
JH AHN conceptualized the study. JH AHN and HY YU cleansed the data. JH AHN and HY YU analyzed the data. JH AHN and HY YU reviewed the data analysis. JH AHN and HY YU wrote initial draft. JH AHN and HY YU reviewed the manuscript and interpreted the findings.
COMPETING INTERESTS
None declared