- Email: [email protected]
Long term prognosis of atrial fibrillation in ST-elevation myocardial infarction patients undergoing percutaneous coronary intervention
Long Term Prognosis of Atrial Fibrillation in ST-Elevation Myocardial Infarction Patients Undergoing Percutaneous Coronary Intervention Guy Topaz, Nir Flint, Arie Steinvil, Arik Finkelstein, Shmuel Banai, Gad Keren, Yacov Shacham, Lior Yankelson PII: DOI: Reference:
S0167-5273(16)32197-0 doi:10.1016/j.ijcard.2017.03.060 IJCA 24740
To appear in:
International Journal of Cardiology
Received date: Accepted date:
11 September 2016 15 March 2017
Please cite this article as: Topaz Guy, Flint Nir, Steinvil Arie, Finkelstein Arik, Banai Shmuel, Keren Gad, Shacham Yacov, Yankelson Lior, Long Term Prognosis of Atrial Fibrillation in ST-Elevation Myocardial Infarction Patients Undergoing Percutaneous Coronary Intervention, International Journal of Cardiology (2017), doi:10.1016/j.ijcard.2017.03.060
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Long Term Prognosis of Atrial Fibrillation in ST-Elevation Myocardial Infarction Patients Undergoing Percutaneous Coronary Intervention Guy Topaz1, M.D., Nir Flint1 M.D., Arie Steinvil1 M.D., Arik Finkelstein1, M.D., Shmuel Banai1,
PT
M.D., Gad Keren1, M.D., Yacov Shacham1, M.D., and Lior Yankelson1,2, M.D. PHD.
RI
From the (1) Tel-Aviv Sourasky Medical Center and Sackler-School of Medicine, Tel Aviv
NU
SC
University, Israel and (2) NYU Langone Medical Center, New York, USA.
These authors take responsibility for all aspects of the reliability and freedom from bias of the
MA
data presented and their discussed interpretation.
Lior Yankelson
D
Address for correspondence:
TE
Department of Cardiology
AC CE P
Tel-Aviv Sourasky Medical Center 6 Weizmann St. Tel-Aviv, 64239 Israel Phone: +972 3 6973222 Fax: +972 3 6973704 Email: [email protected]
Acknowledgement of grant support: none. Potential conflicts of interest: none. Key Words: Atrial fibrillation (AF); ST elevation MI (STEMI); Percutaneous coronary intervention (PCI); Mortality.
1
ACCEPTED MANUSCRIPT Abstract Background: Atrial fibrillation (AF) is a well-known complication in the setting of ST elevation myocardial infarction (STEMI). Data on the long-term prognostic implications of New-Onset AF
PT
(NOAF) complicating STEMI in the era of complete revascularization remains controversial. Our aim therefore was to evaluate the long-term prognosis of prior AF (pAF) and new-onset AF
RI
(NOAF) in STEMI patients undergoing percutaneous coronary intervention (PCI).
SC
Methods: We studied 1657 consecutive STEMI patients hospitalized in the cardiac intensive care unit during 2008-2014. We reviewed patient records for the occurrence of pAF and NOAF.
MA
followed for a mean period of 3.4±2.1 years.
NU
NOAF was defined as AF occurring within 30 days of the STEMI episode. Patients were
Results: Within our cohort 77 (4.6%) patients had pAF and 47 (2.8%) had NOAF. Patients with any AF were older and had a reduced systolic ejection fraction. Thirty-day mortality and all-
TE
D
cause mortality rates were significantly higher in patients with pAF in comparison to those without AF (9.1% vs. 2.2% p<0.001 and 31.2% vs. 9.4%, p<0.001, respectively). NOAF showed
AC CE P
a trend for increased all-cause mortality (17% vs. 9.1%, p=0.07) and 30-days mortality (6.4% vs. 2.1%. p=0.09). In a multivariate regression model, pAF but not NOAF was a predictor of mortality throughout the follow-up period (HR 2.02, 95% CI 1.2 to 3.1, p=0.005 and HR 1.1, 95% CI 0.56 to 2.2, p=0.75, respectively). Conclusions: Prior AF and not new-onset AF is an independent predictor of both short and long term mortality rates in patients treated with PCI.
Key Words: Atrial fibrillation (AF); ST elevation MI (STEMI); Percutaneous coronary intervention (PCI); Mortality.
2
ACCEPTED MANUSCRIPT Introduction The occurrence of AF, new or old, during the acute phase of a STEMI event is common, with a reported frequency ranging from 2.3% to 21% of patients treated with thrombolytic
PT
therapy or PCI (1,2). The improvement in revascularization techniques has led to a profound
RI
reduction in both STEMI mortality rates and the occurrence of AF (3). Several studies
SC
conducted during both the thrombolytic and the PCI era demonstrated that the occurrence of AF in STEMI patients was associated with worse overall prognosis (4-7), while other studies
NU
contradicted this and demonstrated a better outcome (8). In light of these data, the impact of AF in the setting of STEMI remains controversial. Therefore, we evaluated the short and long-term
MA
prognosis of prior and new onset AF in a contemporary cohort of STEMI patients undergoing PCI.
D
Methods
TE
We analyzed 1657 consecutive cases from our prospective registry of STEMI patients
AC CE P
hospitalized in the Cardiac Intensive Care Unit (CICU) between January 2008 and December 2014. Data was collected as part of the Tel-Aviv Prospective Angiography Study (TAPAS) After obtaining informed consent from each patient as approved by the institutional ethics committee. The TAPAS is a prospective, single-center registry that enrolls all patients undergoing cardiac catheterization at the Tel Aviv Medical Center (9). Diagnosis of STEMI was established by the patient’s history of a typical chest pain, diagnostic electrocardiographic changes, and serial elevation of cardiac biomarkers. Primary PCI was performed on patients with symptoms ≤12 hours in duration as well as on patients with symptoms lasting 12-24 hours in duration if the symptoms continued to persist at the time of admission. Unless contraindicated, a 12-month course of dual antiplatelet therapy (DAPT) with Aspirin and P2Y12 inhibitor was initiated at presentation and additional treatment with anticoagulants was provided. Following primary PCI, left ventricular ejection fraction was assessed in all patients within the first 48 hours of admission. 3
ACCEPTED MANUSCRIPT The diagnosis of pAF was based on patient history, medical records and electrocardiogram at admission. AF was defined according to the AHA guidelines (10). NOAF was defined as new onset AF during a maximum of 30 days of hospitalization. In case of NOAF
PT
occurring during hospitalization, anticoagulation was introduced according to guidelines (10). A
RI
virtual CHADS2VASC2 score was calculated for patients without documented AF (11-13).
SC
All data are displayed as mean (± standard deviation) for continuous variables and as number (percentage) of patients in each group for categorical variables. The p-values were
NU
calculated using the student's t-test or Mann-whitney U test for the continuous variables and the chi-square test or fisher exact test for categorical variables (each when appropriate).
MA
Multivariate logistic regression analysis was used to identify independent predictors for the development of new-onset atrial fibrillation following STEMI. The variables entered in the
D
multivariate model were age, gender, hypertension, heart failure, tobacco use, history of prior MI
TE
and Creatinine level at admission. Multivariable Cox proportional hazard model was applied to estimate hazard ratios for all-cause mortality. The variables accounted for in the multivariate cox
AC CE P
proportional hazard model were age, gender, hypertension, heart failure, diabetes mellitus and Creatinine level at admission. Prior AF and new-onset AF were referenced to no-occurrence of AF and adjusted to other potential predictors of mortality following STEMI. The propensity scores were evaluated for each gender by fitting the logistic model. The data of each gender were stratified by quartiles of the propensity scores. The observed and expected rates of mortality among pAF and that among patients with no pAF were evaluated for each stratum. A two-tailed p value of < 0.05 was considered significant for all analyses. All analyses were performed with the SPSS software (SPSS Inc., Chicago, IL). Results Baseline characteristics of 1657 patients are presented in Table 1. Mean follow-up time was 40±25 months. The overall rate of AF was 7.4%, with 4.6% of patients diagnosed with pAF and 2.8% presenting with NOAF. 4
ACCEPTED MANUSCRIPT Patients with prior AF. Patients with pAF had more often paroxysmal AF (71%) and not persistent AF (data was available for 87% of patients). Compared to patients without pAF, patients with pAF were older (74.5±11 vs. 60.8±12, p<0.01), had increased rate of hypertension
PT
(62.3% vs. 41.9%, p<0.01) and history of MI (18.2% vs. 10.3%, p=0.03). Interestingly, patients with pAF had significantly lower rate of tobacco use (26% vs. 51.4%, p<0.01). Complication
RI
rates were higher in patients with pAF, with greater need for an intra-aortic balloon pump (IABP)
SC
(9.1% vs. 3.5%, p=0.02), mechanical ventilation (15.6% vs. 3.9%, p<0.01) and lower EF (44.4%±9.1 vs. 47.7%±7.9, p<0.01). Patients with prior AF also had higher rates of VT/VF (13%
NU
vs. 5.8%, p=0.02) (Table 1).
MA
Patients with NOAF. Most of NOAF episodes occurred after the PCI (61%). Compared to patients without AF, patients with NOAF were older (69.9±13 vs. 60.5±12, p<0.01), had higher
D
rates of heart failure (21.3% vs. 7.5%, p<0.01), hypertension (61.7% vs. 41.3%, p<0.01), and a
TE
history of MI (23.4% vs. 9.9%, p<0.01). As with pAF, among patients with NOAF there was a reduced rate of tobacco use (29.8% vs. 52.1%, p<0.01). Complication rates for NOAF patients
AC CE P
were higher, with statistically significant need for mechanical ventilation (17% vs. 3.5%, p<0.01) and lower EF (45.2%±7.6 vs. 47.6%±8, p=0.03). In a multivariate analysis adjusted for age, gender, hypertension, heart failure, tobacco use, prior MI and creatinine level the significant risk factors for the occurrence of NOAF were older age and prior MI (Table 2). Mortality. Thirty-day and long-term mortality rates were significantly higher for patients with pAF in comparison to those without AF (9.1% vs. 2.2%, p<0.01and 31.2% vs. 9.4%, p<0.01, respectively). As can be seen in figures 1 and 2, the same trend was noted when comparing patients with and without NOAF, with thirty-day and long-term mortality rates higher in the NOAF group (6.4% vs. 2.1%, p=0.08 and 17% vs. 9.1%, p=0.07, respectively). However, in a multivariate adjusted Cox proportional hazard model for mortality, pAF but not NOAF was a predictor of mortality throughout the follow-up period (HR 2.02, 95% CI 1.2 to 3.1, p<0.01and HR 1.1, 95% CI 0.56 to 2.2, p=0.75, respectively), as depicted in table 3 and figure 3. 5
ACCEPTED MANUSCRIPT Propensity scores among pAF patients. The numbers of pAF in each of the first 2 propensity scores quartiles were too small (1.5% in men and 0.05% in women) for an adequate analysis and were therefore omitted from further analyses. However, pAF patients in the 3rd
PT
and 4th propensity score quartiles from both genders, had higher mortality rates, which were statistically significant among women and not significant among men (65% vs. 23.2%, P=0.004
SC
RI
and 26.1% vs. 14.7%, p=0.24, respectively).
AF events in pAF patients. Of the 77 pAF patients, 16 (21%) had AF during the index
NU
hospitalization. All the pAF patients with documented AF at any point of the hospitalization survived the first 30-days after STEMI, in comparison to 7 (11.5%) mortality events among
MA
patients with pAF but without documentation of the arrhythmia during the index hospitalization (p=0.33). Long-term all-cause mortality rates, however, were higher in the group with AF during
D
the index hospitalization in compare to those without (43.8% vs. 27.9%, p=0.22). The prognostic
TE
value of pAF remained significant after subtracting those who had AF during the index hospitalization. Short and long-term mortality rate remained higher among the pAF vs. patients
AC CE P
without AF (11.5% vs. 2.2%, <0.01 and 27.9% vs. 9.4%, p<0.01, respectively). Mortality according to CHADS2VASC2 score. Patients were analyzed according to real or “virtual” CHADS2VASC2 score (11-13) for patients with and without AF, and stratified into lower (0-2) and higher (3+) risk groups. Within the cohort, 1245 (75%) patients had a lower CHADS2VASC2 score. Prior AF, compared to baseline sinus rhythm was associated with higher 30-day and long-term mortality rates, regardless to CHADS2VASC2 group (as depicted in figure 4). Patients with NOAF in the higher CHADS2VASC2 group had short-term and long-term mortality rates similar to patients without NOAF (Figure 5). Interestingly, among patients in the lower CHADS2VASC2 group, NOAF was associated with a trend toward increased mortality, both at 30-days (4.8% vs. 0.7%, p=0.14) and long term (9.5% vs. 4.2%, p=0.22).
6
ACCEPTED MANUSCRIPT Discussion In this study of AF in STEMI patients receiving contemporary care, our main findings are: 1. Incidence of AF was 4.6% and 2.8% for prior AF and NOAF, respectively.
PT
2. Prior AF was associated with a 5 –fold increase in short and long term mortality.
RI
3. The occurrence of an AF episode during the index hospitalization, when evaluated in
SC
patients with pAF, did not affect the short-term outcome but was associated with positive trend for worse long-term outcome.
NU
4. Among patients with pAF, there was no interaction between mortality and CHADS2VASC2 score, with both high and low risk groups demonstrating increased
MA
mortality compared to patient without AF.
5. Patients with NOAF had a trend for increased short and long term mortality rate.
D
However, in contrast to pAF, this was not statistically significant.
TE
6. When looking at CHADS2VASC2 risk regardless of AF, patients with NOAF had a positive interaction between CHADS2VASC2 score and mortality with a trend for higher
AC CE P
mortality risk surprisingly occurring in the lower CHADS2VASC2 group. Albeit a number of studies had previously looked at the impact of AF on STEMI outcome, results thus far have been contradicting. Aligned with the findings in the OACIS study (14) patients with AF in our cohort were older, with higher co-morbidity rate. It can be speculated that this may explain why patients with prior AF have an alarmingly worse prognosis. Yet, we found pAF to be an independent robust risk modifier in a multivariate analysis, inferring of a particular contribution of AF to mortality, perhaps by auxiliary mechanisms that are yet unknown. For example, AF causes impairment in ventricular filling and contractility and therefore reduces cardiac performance (15,16). In the OACIS study patients with AF were more frequently hypotensive with faster heart rate and higher Killip class, they experienced hemodynamic instability and significantly lower post-procedural TIMI flow rates, corresponding to the higher
7
ACCEPTED MANUSCRIPT complication rates evident in our study. It is interesting to consider implementing AF in to risk stratification models such as the TIMI and GRACE scores. Studies from the PCI era have found a variety of outcomes: some demonstrated that the
PT
presence of AF, regardless whether pAF or NOAF, was an independent marker of increased
RI
risk for short and long-term mortality (5-6). In the HORIZONS-AMI sub-study Rene et al. found
SC
higher 3-year mortality amongst patients with NOAF (17). Another study found an association between NOAF and short term mortality but did not find an association between pAF and NOAF
NU
with long-term mortality (18). In contrast, Beukema et al. found that AF after but not before PCI was associated with long term mortality (19). Podolecki et al (7) found that only permanent AF,
MA
diagnosed before the AMI was an independent predictor of mortality (and not prehospital paroxysmal AF or NOAF). Jons et al. found in a CARISMA sub-study (20) that indeed new-
D
onset AF in post-MI patients increased major cardiovascular risk but only those events lasting
TE
more than 30 seconds.
AC CE P
The use of CHADS2 and CHA2DS2-VASc scores for predicting adverse cardiovascular outcome in patients with AF has been extensively studied (21,22). Recently, a number of studies evaluated the application of CHADS2 and CHA2DS2-VASc scores on patients without AF (“virtual” score), demonstrating the feasibility of these scores as a reliable method of riskstratifying non-AF patients (11,12). In our work, we chose to use this method to compare patients with and without AF on the basis of CHA2DS2-VASc score. Indeed, we found that the occurrence of NAOF was a significant adverse modifier in the lower strata (0-2) rather than in the higher strata (3 and above). It is possible that in the higher risk group, mortality is high regardless of AF so that even without the arrhythmia, patients are compromised. This is in contrast to the lower risk group, where patients have relatively lower adverse event rate so that the occurrence of AF either represents an additional ‘concealed” risk mechanism or directly increases the likelihood of an adverse event.
8
ACCEPTED MANUSCRIPT In our study, pAF had a major impact on mortality, greater than NOAF. This finding is aligned with other studies (7). The modest impact of NOAF in our study may be explained by a relatively low event rate occurring in the NOAF group (47 cases). Also, the mere occurrence of
PT
an AF episode during the index hospitalization in patients with prior AF showed a trend for long term worse prognosis. It seems the occurrence of AF during the STEMI hospitalization, whether
RI
detected for the first time and considered to be NOAF or among pAF patients, is associated with
SC
worse outcome.
NU
In light of the major impact on prognosis, it is critical to consider the method of screening and detection of AF. Jons et al found that when using an implantable loop recorder in AMI
MA
patients, nearly 25% of the patients had AF during 2-years follow-up, an incidence 4 fold higher than previously believed (20). In this regard, it is possible that the incidence of AF in our cohort
D
is underestimated and so that the impact of AF on mortality is even higher, as some of the death
TE
cases occurring in patients presumably in sinus rhythm in fact occurred in undiagnosed AF patients. Currently, there is no recommendation for distinct AF detection or monitoring strategies
AC CE P
in patients who experience AMI. The impact of AF on prognosis, as well as the high rates of sub-clinical AF in this population group may call for implementing a more aggressive and proactive approach of AF detection in this patient population. Limitations. This was a retrospective analysis of a prospective consecutive cohort rather than a prospectively designed trial. This limitation is negated by the all-comer approach, including all patients admitted to our facility. We did not stratify patients by the degree of coronary involvement or the location of the culprit lesion. We were not able to analyze the exact duration of detected AF episodes. To ensure adequate power, we did not separate pAF patients according to the duration of the arrhythmia (paroxysmal versus persistent). Finally, only 63 AF episodes were detected during the index hospitalization (47 new onset AF and 16 AF events among the pAF patients) limiting the ability to achieve statistically significant results regarding the impact of in-hospital AF on the outcome. 9
ACCEPTED MANUSCRIPT Conclusion. Atrial fibrillation is associated with poor outcome in STEMI patients, but only prior AF and not new-onset AF is an independent significant proven predictor of long-term mortality. It may be reasonable to intensively seek for AF among STEMI patients.
AC CE P
TE
D
MA
NU
SC
RI
PT
Acknowledgments: none.
10
ACCEPTED MANUSCRIPT References: 1. Schmitt J, Duray G, Gersh BJ, Hohnloser SH. Atrial fibrillation in acute myocardial infarction: a systematic review of the incidence, clinical features and prognostic
PT
implications. Eur Heart J. 2009 May;30(9):1038-45. 2. Jabre P, Roger VL, Murad MH et al. Mortality associated with atrial fibrillation in patients
RI
with myocardial infarction: a systematic review and meta-analysis. Circulation. 2011 Apr
SC
19;123(15):1587-93.
3. McManus DD, Huang W, Domakonda KV et al. Trends in atrial fibrillation in patients
NU
hospitalized with an acute coronary syndrome. Am J Med. 2012 Nov;125(11):1076-84.
MA
4. Crenshaw BS, Ward SR, Granger CB, Stebbins AL, Topol EJ, Califf RM. Atrial fibrillation in the setting of acute myocardial infarction: the GUSTO-I experience. Global Utilization of Streptokinase and TPA for Occluded Coronary Arteries. J Am Coll Cardiol. 1997
TE
5.
D
Aug;30(2):406-13.
Pizzetti F, Turazza FM, Franzosi MG et al. Tognoni G; GISSI-3 Investigators. Incidence
AC CE P
and prognostic significance of atrial fibrillation in acute myocardial infarction: the GISSI-3 data. Heart. 2001 Nov;86(5):527-32. 6. Lehto M, Snapinn S, Dickstein K, Swedberg K, Nieminen MS; OPTIMAAL investigators. Prognostic risk of atrial fibrillation in acute myocardial infarction complicated by left ventricular dysfunction: the OPTIMAAL experience. Eur Heart J. 2005;26:350–6. 7. Podolecki T, Lenarczyk R, Kowalczyk J et al. Effect of Type of Atrial Fibrillation on Prognosis in Acute Myocardial Infarction Treated Invasively. Am J Cardiol. 2012 Jun 15;109(12):1689-93. 8. Eldar M, Canetti M, Rotstein Z et al. Significance of paroxysmal atrial fibrillation complicating acute myocardial infarction in the thrombolytic era. SPRINT and Thrombolytic Survey Groups. Circulation. 1998 Mar 17;97(10):965-70. 9. Steinvil A, Sadeh B, Arbel Y et al. Prevalence and predictors of concomitant carotid and coronary artery atherosclerotic disease. J Am Coll Cardiol. 2011 Feb 15;57(7):779-83. 11
ACCEPTED MANUSCRIPT 10. January CT, Wann LS, Alpert JS et al. ACC/AHA Task Force Members. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: executive summary: a report of the American College of Cardiology/American Heart Association
PT
Task Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014 Dec 2;130(23):2071-104.
RI
11. Biancari F, Asim Mahar MA, Kangasniemi OP. CHADS2 and CHA2DS2-VASc scores for
SC
prediction of immediate and late stroke after coronary artery bypass graft surgery. J Stroke Cerebrovasc Dis. 2013 Nov;22(8):1304-11.
NU
12. Mitchell LB, Southern DA, Galbraith D, et al. Prediction of stroke or TIA in patients
MA
without atrial fibrillation using CHADS2 and CHA2DS2-VASc scores. Heart. 2014 Oct;100(19):1524-30.
13. Liu FD, Shen XL, Zhao R et al. Predictive role of CHADS2 and CHA2DS2-VASc scores
TE
D
on stroke and thromboembolism in patients without atrial fibrillation: a meta-analysis. Ann Med. 2016 May 6:1-9.
AC CE P
14. Kinjo K, Sato H, Sato H et al. Osaka Acute Coronary Insufficiency Study (OACIS) Group. Prognostic significance of atrial fibrillation/atrial flutter in patients with acute myocardial infarction treated with percutaneous coronary intervention. Am J Cardiol. 2003 Nov 15;92(10):1150-4.
15. Gupta DK, Giugliano RP, Ruff CT et al. Effective Anticoagulation with Factor Xa Next Generation in AF–Thrombolysis in Myocardial Infarction 48 (ENGAGE AF–IMI 48) Echocardiographic Study Investigators. The Prognostic Significance of Cardiac Structure and Function in Atrial Fibrillation: The ENGAGE AF-TIMI 48 Echocardiographic Substudy. J Am Soc Echocardiogr. 2016 Apr 20. 16. Krezowski JT, Wilson BD, McGann CJ, Marrouche NF, Akoum N. Changes in left ventricular filling parameters following catheter ablation of atrial fibrillation. J Interv Card Electrophysiol. 2016 Apr 13.
12
ACCEPTED MANUSCRIPT 17. Rene AG, Généreux P, Ezekowitz M et al. Impact of atrial fibrillation in patients with STelevation myocardial infarction treated with percutaneous coronary intervention (from the HORIZONS-AMI [Harmonizing Outcomes With Revascularization and Stents in Acute
PT
Myocardial Infarction] trial). Am J Cardiol. 2014 Jan 15;113(2):236-42. 18. Consuegra-Sánchez L, Melgarejo-Moreno A, Galcerá-Tomás J et al. Short- and long-
RI
term prognosis of previous and new-onset atrial fibrillation in ST-segment elevation
SC
acute myocardial infarction. Rev Esp Cardiol (Engl Ed). 2015 Jan;68(1):31-8. 19. Beukema RJ, Elvan A, Ottervanger JP et al. Atrial fibrillation after but not before primary
MA
Neth Heart J. 2012 Apr;20(4):155-60.
NU
angioplasty for ST-segment elevation myocardial infarction of prognostic importance.
20. Jons C, Jacobsen UG, Joergensen RM et al. Cardiac Arrhythmias and Risk Stratification after Acute Myocardial Infarction (CARISMA) Study Group. The incidence and
TE
D
prognostic significance of new-onset atrial fibrillation in patients with acute myocardial infarction and left ventricular systolic dysfunction: a CARISMA sub study. Heart Rhythm.
AC CE P
2011 Mar;8(3):342-8.
21. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA. 2001 Jun 13;285(22):2864-70. 22. Olesen JB, Lip GY, Hansen ML, Validation of risk stratification schemes for predicting stroke and thromboembolism in patients with atrial fibrillation: nationwide cohort study. BMJ. 2011 Jan 31;342:d124.
13
ACCEPTED MANUSCRIPT Table 1: Baseline patient characteristics.
p Value
NOAF (47)
NO-NOAF (1533)
61.4±13
69.9±13
60.5±12
Male Gender, n (%)
1331 (80.3)
32 (68.1)
1247 (81.3)
Heart Failure, n (%)
134 (8.1)
10 (21.3)
Dyslipidemia, n (%)
777 (46.9)
Hypertension, n (%)
Prior AF (77)
No-Prior AF (1580) <0.01
0.03
52 (67.5)
1279 (80.9)
<0.01
115 (7.5)
<0.01
9 (11.7)
125 (7.9)
0.28
22 (46.8)
722 (47.1)
0.97
33 (42.9)
744 (47.1)
0.47
710 (42.8)
29 (61.7)
633 (41.3)
<0.01
48 (62.3)
662 (41.9)
<0.01
Past MI, n (%)
177 (10.7)
11 (23.4)
152 (9.9)
<0.01
14 (18.2)
163 (10.3)
0.03
Diabetes Mellitus, n (%)
359 (21.7)
14 (29.8)
328 (21.4)
0.2
17 (22.1)
342 (21.6)
0.88
Tabaco Use, n (%)
832 (50.2)
14 (29.8)
798 (52.1)
<0.01
20 (26.0)
812 (51.4)
<0.01
Family Hx.
290 (17.5)
4 (8.5%)
279 (18.2)
0.12
7 (9.1)
283 (17.9)
0.04
454±631
362±565
0.28
353±545
365±568
0.86
3 (6.4)
52 (3.4)
0.22
7 (9.1)
55 (3.5)
0.02
33 (2.0)
0 (0%)
31 (2%)
1.0
2 (2.6)
31 (2.0)
0.66
74 (4.5)
8 (17.0)
54 (3.5)
<0.01
12 (15.6)
62 (3.9)
<0.01
102 (6.2)
5 (10.6)
87 (5.7)
0.19
10 (13.0)
92 (5.8)
0.02
51 (3.1)
6 (12.8)
43 (2.8)
<0.01
2 (2.6)
49 (3.1)
1
1.15±0.26
1.17±0.28
1.14±0.26
0.42
1.2±0.25
1.14±0.26
0.05
1.2±0.5
1.41±0.66
1.18±0.47
0.01
1.47±0.8
1.19±0.48
<0.01
1246±1378
1008±1004
1260±13878
0.22
1121±1391
1252±1378
0.42
47.6±8
45.2±7.6
47.8±7.9
0.03
44.4±9.1
47.7±7.9
<0.01
CHA2DS2, mean±SD
0.95±1.0
1.66±1.3
0.89±0.9
<0.01
1.79±1.4
0.91±1.0
<0.01
CHA2DS2VASC2, mean±SD
1.55±1.5
2.77±1.8
1.44±1.4
<0.01
3.0±1.8
1.48±1.4
<0.01
IABC, n (%)
62 (3.7)
CABG, n (%) Mechanical Ventilation, n (%) VT/VF, n (%) Bradycardia
Creatinine admission, mean±SD Creatinine peak, mean±SD CPK, mean±SD EF, mean±SD
NU
MA
D
AC CE P
365±566
SC
60.8±12
Time to ER, n (%)
<0.01
p Value
74.5±11
TE
Age, mean ± SD
Prior AF vs. no prior AF (n=1657)
PT
NOAF vs. no NOAF (n=1580)
RI
Total (n=1657)
Coefficients
Abbreviations: MI, Myocardial infarction; ER, Emergency Room; IABC, Intra-Aortic Balloon Counterpulsation; CABG, Coronary artery bypass graft; VT/VF, Ventricular Tachycardia/Ventricular Fibrillation, CPK, Creatine PhosphoKinase; EF, Ejection Fraction.
14
ACCEPTED MANUSCRIPT Table 2: Multivariate analysis for predictors of new onset atrial fibrillation following STEMI adjusted odds ratios and 95% CI.
1.04 1.14 1.36 1.94 0.58 2.62 0.7
Lower
Upper
1.01 0.56 0.71 0.89 0.29 1.27 0.24
1.07 2.33 2.59 4.24 1.14 5.39 2.1
AC CE P
TE
D
MA
Abbreviations: MI, Myocardial infarction.
<0.01 0.71 0.34 0.96 0.11 <0.01 0.53
PT
p Value
SC
Age Male Gender Hypertension Heart Failure Tobacco Use Prior MI Creatinine at Admission
95% CI
RI
OR
NU
Coefficients
15
ACCEPTED MANUSCRIPT Table 3.
95% CI Lower
Upper
<0.01 0.75 0.085 <0.01 <0.01 <0.01 0.77 0.43
MA
NU
3.19 2.21 1.04 1.08 2.98 5.04 1.48 1.63
D
No atrial fibrillation Reference Prior atrial fibrillation 2.02 1.28 New Onset atrial fibrillation 1.11 0.56 Male Gender 0.74 0. 52 Age 1.07 1.05 Creatinine admission 2.074 1.44 Heart Failure 3.55 2.49 Diabetes Mellitus 1.05 0.75 Hypertension 1.15 0.81
p Value
RI
HR
SC
Coefficients
PT
Cox proportional hazard model for mortality.
AC CE P
TE
CI = confidence interval; HR = hazard ratio.
16
ACCEPTED MANUSCRIPT Figure legends Figure 1. cumulative rates of mortality in patients with and without prior AF. AF = atrial fibrillation.
PT
Figure 2. cumulative rates of mortality in patients with and without New-onset AF. AF = atrial fibrillation.
RI
Figure 3. cox proportional hazards survival curves. AF = atrial fibrillation. CI = confidence
SC
interval; HR = hazard ratio.
CHADS2VASC2 score. AF = atrial fibrillation.
NU
Figure 4. cumulative rates of mortality in patients with and without Prior AF according to
Figure 5. cumulative rates of mortality in patients with and without New onset AF according to
AC CE P
TE
D
MA
CHADS2VASC2 score. AF = atrial fibrillation.
17
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
Figure 1
18
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
Figure 2
19
Figure 3
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
20
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
Figure 4
21
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
Figure 5
22
S0167-5273(16)32197-0 doi:10.1016/j.ijcard.2017.03.060 IJCA 24740
To appear in:
International Journal of Cardiology
Received date: Accepted date:
11 September 2016 15 March 2017
Please cite this article as: Topaz Guy, Flint Nir, Steinvil Arie, Finkelstein Arik, Banai Shmuel, Keren Gad, Shacham Yacov, Yankelson Lior, Long Term Prognosis of Atrial Fibrillation in ST-Elevation Myocardial Infarction Patients Undergoing Percutaneous Coronary Intervention, International Journal of Cardiology (2017), doi:10.1016/j.ijcard.2017.03.060
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Long Term Prognosis of Atrial Fibrillation in ST-Elevation Myocardial Infarction Patients Undergoing Percutaneous Coronary Intervention Guy Topaz1, M.D., Nir Flint1 M.D., Arie Steinvil1 M.D., Arik Finkelstein1, M.D., Shmuel Banai1,
PT
M.D., Gad Keren1, M.D., Yacov Shacham1, M.D., and Lior Yankelson1,2, M.D. PHD.
RI
From the (1) Tel-Aviv Sourasky Medical Center and Sackler-School of Medicine, Tel Aviv
NU
SC
University, Israel and (2) NYU Langone Medical Center, New York, USA.
These authors take responsibility for all aspects of the reliability and freedom from bias of the
MA
data presented and their discussed interpretation.
Lior Yankelson
D
Address for correspondence:
TE
Department of Cardiology
AC CE P
Tel-Aviv Sourasky Medical Center 6 Weizmann St. Tel-Aviv, 64239 Israel Phone: +972 3 6973222 Fax: +972 3 6973704 Email: [email protected]
Acknowledgement of grant support: none. Potential conflicts of interest: none. Key Words: Atrial fibrillation (AF); ST elevation MI (STEMI); Percutaneous coronary intervention (PCI); Mortality.
1
ACCEPTED MANUSCRIPT Abstract Background: Atrial fibrillation (AF) is a well-known complication in the setting of ST elevation myocardial infarction (STEMI). Data on the long-term prognostic implications of New-Onset AF
PT
(NOAF) complicating STEMI in the era of complete revascularization remains controversial. Our aim therefore was to evaluate the long-term prognosis of prior AF (pAF) and new-onset AF
RI
(NOAF) in STEMI patients undergoing percutaneous coronary intervention (PCI).
SC
Methods: We studied 1657 consecutive STEMI patients hospitalized in the cardiac intensive care unit during 2008-2014. We reviewed patient records for the occurrence of pAF and NOAF.
MA
followed for a mean period of 3.4±2.1 years.
NU
NOAF was defined as AF occurring within 30 days of the STEMI episode. Patients were
Results: Within our cohort 77 (4.6%) patients had pAF and 47 (2.8%) had NOAF. Patients with any AF were older and had a reduced systolic ejection fraction. Thirty-day mortality and all-
TE
D
cause mortality rates were significantly higher in patients with pAF in comparison to those without AF (9.1% vs. 2.2% p<0.001 and 31.2% vs. 9.4%, p<0.001, respectively). NOAF showed
AC CE P
a trend for increased all-cause mortality (17% vs. 9.1%, p=0.07) and 30-days mortality (6.4% vs. 2.1%. p=0.09). In a multivariate regression model, pAF but not NOAF was a predictor of mortality throughout the follow-up period (HR 2.02, 95% CI 1.2 to 3.1, p=0.005 and HR 1.1, 95% CI 0.56 to 2.2, p=0.75, respectively). Conclusions: Prior AF and not new-onset AF is an independent predictor of both short and long term mortality rates in patients treated with PCI.
Key Words: Atrial fibrillation (AF); ST elevation MI (STEMI); Percutaneous coronary intervention (PCI); Mortality.
2
ACCEPTED MANUSCRIPT Introduction The occurrence of AF, new or old, during the acute phase of a STEMI event is common, with a reported frequency ranging from 2.3% to 21% of patients treated with thrombolytic
PT
therapy or PCI (1,2). The improvement in revascularization techniques has led to a profound
RI
reduction in both STEMI mortality rates and the occurrence of AF (3). Several studies
SC
conducted during both the thrombolytic and the PCI era demonstrated that the occurrence of AF in STEMI patients was associated with worse overall prognosis (4-7), while other studies
NU
contradicted this and demonstrated a better outcome (8). In light of these data, the impact of AF in the setting of STEMI remains controversial. Therefore, we evaluated the short and long-term
MA
prognosis of prior and new onset AF in a contemporary cohort of STEMI patients undergoing PCI.
D
Methods
TE
We analyzed 1657 consecutive cases from our prospective registry of STEMI patients
AC CE P
hospitalized in the Cardiac Intensive Care Unit (CICU) between January 2008 and December 2014. Data was collected as part of the Tel-Aviv Prospective Angiography Study (TAPAS) After obtaining informed consent from each patient as approved by the institutional ethics committee. The TAPAS is a prospective, single-center registry that enrolls all patients undergoing cardiac catheterization at the Tel Aviv Medical Center (9). Diagnosis of STEMI was established by the patient’s history of a typical chest pain, diagnostic electrocardiographic changes, and serial elevation of cardiac biomarkers. Primary PCI was performed on patients with symptoms ≤12 hours in duration as well as on patients with symptoms lasting 12-24 hours in duration if the symptoms continued to persist at the time of admission. Unless contraindicated, a 12-month course of dual antiplatelet therapy (DAPT) with Aspirin and P2Y12 inhibitor was initiated at presentation and additional treatment with anticoagulants was provided. Following primary PCI, left ventricular ejection fraction was assessed in all patients within the first 48 hours of admission. 3
ACCEPTED MANUSCRIPT The diagnosis of pAF was based on patient history, medical records and electrocardiogram at admission. AF was defined according to the AHA guidelines (10). NOAF was defined as new onset AF during a maximum of 30 days of hospitalization. In case of NOAF
PT
occurring during hospitalization, anticoagulation was introduced according to guidelines (10). A
RI
virtual CHADS2VASC2 score was calculated for patients without documented AF (11-13).
SC
All data are displayed as mean (± standard deviation) for continuous variables and as number (percentage) of patients in each group for categorical variables. The p-values were
NU
calculated using the student's t-test or Mann-whitney U test for the continuous variables and the chi-square test or fisher exact test for categorical variables (each when appropriate).
MA
Multivariate logistic regression analysis was used to identify independent predictors for the development of new-onset atrial fibrillation following STEMI. The variables entered in the
D
multivariate model were age, gender, hypertension, heart failure, tobacco use, history of prior MI
TE
and Creatinine level at admission. Multivariable Cox proportional hazard model was applied to estimate hazard ratios for all-cause mortality. The variables accounted for in the multivariate cox
AC CE P
proportional hazard model were age, gender, hypertension, heart failure, diabetes mellitus and Creatinine level at admission. Prior AF and new-onset AF were referenced to no-occurrence of AF and adjusted to other potential predictors of mortality following STEMI. The propensity scores were evaluated for each gender by fitting the logistic model. The data of each gender were stratified by quartiles of the propensity scores. The observed and expected rates of mortality among pAF and that among patients with no pAF were evaluated for each stratum. A two-tailed p value of < 0.05 was considered significant for all analyses. All analyses were performed with the SPSS software (SPSS Inc., Chicago, IL). Results Baseline characteristics of 1657 patients are presented in Table 1. Mean follow-up time was 40±25 months. The overall rate of AF was 7.4%, with 4.6% of patients diagnosed with pAF and 2.8% presenting with NOAF. 4
ACCEPTED MANUSCRIPT Patients with prior AF. Patients with pAF had more often paroxysmal AF (71%) and not persistent AF (data was available for 87% of patients). Compared to patients without pAF, patients with pAF were older (74.5±11 vs. 60.8±12, p<0.01), had increased rate of hypertension
PT
(62.3% vs. 41.9%, p<0.01) and history of MI (18.2% vs. 10.3%, p=0.03). Interestingly, patients with pAF had significantly lower rate of tobacco use (26% vs. 51.4%, p<0.01). Complication
RI
rates were higher in patients with pAF, with greater need for an intra-aortic balloon pump (IABP)
SC
(9.1% vs. 3.5%, p=0.02), mechanical ventilation (15.6% vs. 3.9%, p<0.01) and lower EF (44.4%±9.1 vs. 47.7%±7.9, p<0.01). Patients with prior AF also had higher rates of VT/VF (13%
NU
vs. 5.8%, p=0.02) (Table 1).
MA
Patients with NOAF. Most of NOAF episodes occurred after the PCI (61%). Compared to patients without AF, patients with NOAF were older (69.9±13 vs. 60.5±12, p<0.01), had higher
D
rates of heart failure (21.3% vs. 7.5%, p<0.01), hypertension (61.7% vs. 41.3%, p<0.01), and a
TE
history of MI (23.4% vs. 9.9%, p<0.01). As with pAF, among patients with NOAF there was a reduced rate of tobacco use (29.8% vs. 52.1%, p<0.01). Complication rates for NOAF patients
AC CE P
were higher, with statistically significant need for mechanical ventilation (17% vs. 3.5%, p<0.01) and lower EF (45.2%±7.6 vs. 47.6%±8, p=0.03). In a multivariate analysis adjusted for age, gender, hypertension, heart failure, tobacco use, prior MI and creatinine level the significant risk factors for the occurrence of NOAF were older age and prior MI (Table 2). Mortality. Thirty-day and long-term mortality rates were significantly higher for patients with pAF in comparison to those without AF (9.1% vs. 2.2%, p<0.01and 31.2% vs. 9.4%, p<0.01, respectively). As can be seen in figures 1 and 2, the same trend was noted when comparing patients with and without NOAF, with thirty-day and long-term mortality rates higher in the NOAF group (6.4% vs. 2.1%, p=0.08 and 17% vs. 9.1%, p=0.07, respectively). However, in a multivariate adjusted Cox proportional hazard model for mortality, pAF but not NOAF was a predictor of mortality throughout the follow-up period (HR 2.02, 95% CI 1.2 to 3.1, p<0.01and HR 1.1, 95% CI 0.56 to 2.2, p=0.75, respectively), as depicted in table 3 and figure 3. 5
ACCEPTED MANUSCRIPT Propensity scores among pAF patients. The numbers of pAF in each of the first 2 propensity scores quartiles were too small (1.5% in men and 0.05% in women) for an adequate analysis and were therefore omitted from further analyses. However, pAF patients in the 3rd
PT
and 4th propensity score quartiles from both genders, had higher mortality rates, which were statistically significant among women and not significant among men (65% vs. 23.2%, P=0.004
SC
RI
and 26.1% vs. 14.7%, p=0.24, respectively).
AF events in pAF patients. Of the 77 pAF patients, 16 (21%) had AF during the index
NU
hospitalization. All the pAF patients with documented AF at any point of the hospitalization survived the first 30-days after STEMI, in comparison to 7 (11.5%) mortality events among
MA
patients with pAF but without documentation of the arrhythmia during the index hospitalization (p=0.33). Long-term all-cause mortality rates, however, were higher in the group with AF during
D
the index hospitalization in compare to those without (43.8% vs. 27.9%, p=0.22). The prognostic
TE
value of pAF remained significant after subtracting those who had AF during the index hospitalization. Short and long-term mortality rate remained higher among the pAF vs. patients
AC CE P
without AF (11.5% vs. 2.2%, <0.01 and 27.9% vs. 9.4%, p<0.01, respectively). Mortality according to CHADS2VASC2 score. Patients were analyzed according to real or “virtual” CHADS2VASC2 score (11-13) for patients with and without AF, and stratified into lower (0-2) and higher (3+) risk groups. Within the cohort, 1245 (75%) patients had a lower CHADS2VASC2 score. Prior AF, compared to baseline sinus rhythm was associated with higher 30-day and long-term mortality rates, regardless to CHADS2VASC2 group (as depicted in figure 4). Patients with NOAF in the higher CHADS2VASC2 group had short-term and long-term mortality rates similar to patients without NOAF (Figure 5). Interestingly, among patients in the lower CHADS2VASC2 group, NOAF was associated with a trend toward increased mortality, both at 30-days (4.8% vs. 0.7%, p=0.14) and long term (9.5% vs. 4.2%, p=0.22).
6
ACCEPTED MANUSCRIPT Discussion In this study of AF in STEMI patients receiving contemporary care, our main findings are: 1. Incidence of AF was 4.6% and 2.8% for prior AF and NOAF, respectively.
PT
2. Prior AF was associated with a 5 –fold increase in short and long term mortality.
RI
3. The occurrence of an AF episode during the index hospitalization, when evaluated in
SC
patients with pAF, did not affect the short-term outcome but was associated with positive trend for worse long-term outcome.
NU
4. Among patients with pAF, there was no interaction between mortality and CHADS2VASC2 score, with both high and low risk groups demonstrating increased
MA
mortality compared to patient without AF.
5. Patients with NOAF had a trend for increased short and long term mortality rate.
D
However, in contrast to pAF, this was not statistically significant.
TE
6. When looking at CHADS2VASC2 risk regardless of AF, patients with NOAF had a positive interaction between CHADS2VASC2 score and mortality with a trend for higher
AC CE P
mortality risk surprisingly occurring in the lower CHADS2VASC2 group. Albeit a number of studies had previously looked at the impact of AF on STEMI outcome, results thus far have been contradicting. Aligned with the findings in the OACIS study (14) patients with AF in our cohort were older, with higher co-morbidity rate. It can be speculated that this may explain why patients with prior AF have an alarmingly worse prognosis. Yet, we found pAF to be an independent robust risk modifier in a multivariate analysis, inferring of a particular contribution of AF to mortality, perhaps by auxiliary mechanisms that are yet unknown. For example, AF causes impairment in ventricular filling and contractility and therefore reduces cardiac performance (15,16). In the OACIS study patients with AF were more frequently hypotensive with faster heart rate and higher Killip class, they experienced hemodynamic instability and significantly lower post-procedural TIMI flow rates, corresponding to the higher
7
ACCEPTED MANUSCRIPT complication rates evident in our study. It is interesting to consider implementing AF in to risk stratification models such as the TIMI and GRACE scores. Studies from the PCI era have found a variety of outcomes: some demonstrated that the
PT
presence of AF, regardless whether pAF or NOAF, was an independent marker of increased
RI
risk for short and long-term mortality (5-6). In the HORIZONS-AMI sub-study Rene et al. found
SC
higher 3-year mortality amongst patients with NOAF (17). Another study found an association between NOAF and short term mortality but did not find an association between pAF and NOAF
NU
with long-term mortality (18). In contrast, Beukema et al. found that AF after but not before PCI was associated with long term mortality (19). Podolecki et al (7) found that only permanent AF,
MA
diagnosed before the AMI was an independent predictor of mortality (and not prehospital paroxysmal AF or NOAF). Jons et al. found in a CARISMA sub-study (20) that indeed new-
D
onset AF in post-MI patients increased major cardiovascular risk but only those events lasting
TE
more than 30 seconds.
AC CE P
The use of CHADS2 and CHA2DS2-VASc scores for predicting adverse cardiovascular outcome in patients with AF has been extensively studied (21,22). Recently, a number of studies evaluated the application of CHADS2 and CHA2DS2-VASc scores on patients without AF (“virtual” score), demonstrating the feasibility of these scores as a reliable method of riskstratifying non-AF patients (11,12). In our work, we chose to use this method to compare patients with and without AF on the basis of CHA2DS2-VASc score. Indeed, we found that the occurrence of NAOF was a significant adverse modifier in the lower strata (0-2) rather than in the higher strata (3 and above). It is possible that in the higher risk group, mortality is high regardless of AF so that even without the arrhythmia, patients are compromised. This is in contrast to the lower risk group, where patients have relatively lower adverse event rate so that the occurrence of AF either represents an additional ‘concealed” risk mechanism or directly increases the likelihood of an adverse event.
8
ACCEPTED MANUSCRIPT In our study, pAF had a major impact on mortality, greater than NOAF. This finding is aligned with other studies (7). The modest impact of NOAF in our study may be explained by a relatively low event rate occurring in the NOAF group (47 cases). Also, the mere occurrence of
PT
an AF episode during the index hospitalization in patients with prior AF showed a trend for long term worse prognosis. It seems the occurrence of AF during the STEMI hospitalization, whether
RI
detected for the first time and considered to be NOAF or among pAF patients, is associated with
SC
worse outcome.
NU
In light of the major impact on prognosis, it is critical to consider the method of screening and detection of AF. Jons et al found that when using an implantable loop recorder in AMI
MA
patients, nearly 25% of the patients had AF during 2-years follow-up, an incidence 4 fold higher than previously believed (20). In this regard, it is possible that the incidence of AF in our cohort
D
is underestimated and so that the impact of AF on mortality is even higher, as some of the death
TE
cases occurring in patients presumably in sinus rhythm in fact occurred in undiagnosed AF patients. Currently, there is no recommendation for distinct AF detection or monitoring strategies
AC CE P
in patients who experience AMI. The impact of AF on prognosis, as well as the high rates of sub-clinical AF in this population group may call for implementing a more aggressive and proactive approach of AF detection in this patient population. Limitations. This was a retrospective analysis of a prospective consecutive cohort rather than a prospectively designed trial. This limitation is negated by the all-comer approach, including all patients admitted to our facility. We did not stratify patients by the degree of coronary involvement or the location of the culprit lesion. We were not able to analyze the exact duration of detected AF episodes. To ensure adequate power, we did not separate pAF patients according to the duration of the arrhythmia (paroxysmal versus persistent). Finally, only 63 AF episodes were detected during the index hospitalization (47 new onset AF and 16 AF events among the pAF patients) limiting the ability to achieve statistically significant results regarding the impact of in-hospital AF on the outcome. 9
ACCEPTED MANUSCRIPT Conclusion. Atrial fibrillation is associated with poor outcome in STEMI patients, but only prior AF and not new-onset AF is an independent significant proven predictor of long-term mortality. It may be reasonable to intensively seek for AF among STEMI patients.
AC CE P
TE
D
MA
NU
SC
RI
PT
Acknowledgments: none.
10
ACCEPTED MANUSCRIPT References: 1. Schmitt J, Duray G, Gersh BJ, Hohnloser SH. Atrial fibrillation in acute myocardial infarction: a systematic review of the incidence, clinical features and prognostic
PT
implications. Eur Heart J. 2009 May;30(9):1038-45. 2. Jabre P, Roger VL, Murad MH et al. Mortality associated with atrial fibrillation in patients
RI
with myocardial infarction: a systematic review and meta-analysis. Circulation. 2011 Apr
SC
19;123(15):1587-93.
3. McManus DD, Huang W, Domakonda KV et al. Trends in atrial fibrillation in patients
NU
hospitalized with an acute coronary syndrome. Am J Med. 2012 Nov;125(11):1076-84.
MA
4. Crenshaw BS, Ward SR, Granger CB, Stebbins AL, Topol EJ, Califf RM. Atrial fibrillation in the setting of acute myocardial infarction: the GUSTO-I experience. Global Utilization of Streptokinase and TPA for Occluded Coronary Arteries. J Am Coll Cardiol. 1997
TE
5.
D
Aug;30(2):406-13.
Pizzetti F, Turazza FM, Franzosi MG et al. Tognoni G; GISSI-3 Investigators. Incidence
AC CE P
and prognostic significance of atrial fibrillation in acute myocardial infarction: the GISSI-3 data. Heart. 2001 Nov;86(5):527-32. 6. Lehto M, Snapinn S, Dickstein K, Swedberg K, Nieminen MS; OPTIMAAL investigators. Prognostic risk of atrial fibrillation in acute myocardial infarction complicated by left ventricular dysfunction: the OPTIMAAL experience. Eur Heart J. 2005;26:350–6. 7. Podolecki T, Lenarczyk R, Kowalczyk J et al. Effect of Type of Atrial Fibrillation on Prognosis in Acute Myocardial Infarction Treated Invasively. Am J Cardiol. 2012 Jun 15;109(12):1689-93. 8. Eldar M, Canetti M, Rotstein Z et al. Significance of paroxysmal atrial fibrillation complicating acute myocardial infarction in the thrombolytic era. SPRINT and Thrombolytic Survey Groups. Circulation. 1998 Mar 17;97(10):965-70. 9. Steinvil A, Sadeh B, Arbel Y et al. Prevalence and predictors of concomitant carotid and coronary artery atherosclerotic disease. J Am Coll Cardiol. 2011 Feb 15;57(7):779-83. 11
ACCEPTED MANUSCRIPT 10. January CT, Wann LS, Alpert JS et al. ACC/AHA Task Force Members. 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: executive summary: a report of the American College of Cardiology/American Heart Association
PT
Task Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014 Dec 2;130(23):2071-104.
RI
11. Biancari F, Asim Mahar MA, Kangasniemi OP. CHADS2 and CHA2DS2-VASc scores for
SC
prediction of immediate and late stroke after coronary artery bypass graft surgery. J Stroke Cerebrovasc Dis. 2013 Nov;22(8):1304-11.
NU
12. Mitchell LB, Southern DA, Galbraith D, et al. Prediction of stroke or TIA in patients
MA
without atrial fibrillation using CHADS2 and CHA2DS2-VASc scores. Heart. 2014 Oct;100(19):1524-30.
13. Liu FD, Shen XL, Zhao R et al. Predictive role of CHADS2 and CHA2DS2-VASc scores
TE
D
on stroke and thromboembolism in patients without atrial fibrillation: a meta-analysis. Ann Med. 2016 May 6:1-9.
AC CE P
14. Kinjo K, Sato H, Sato H et al. Osaka Acute Coronary Insufficiency Study (OACIS) Group. Prognostic significance of atrial fibrillation/atrial flutter in patients with acute myocardial infarction treated with percutaneous coronary intervention. Am J Cardiol. 2003 Nov 15;92(10):1150-4.
15. Gupta DK, Giugliano RP, Ruff CT et al. Effective Anticoagulation with Factor Xa Next Generation in AF–Thrombolysis in Myocardial Infarction 48 (ENGAGE AF–IMI 48) Echocardiographic Study Investigators. The Prognostic Significance of Cardiac Structure and Function in Atrial Fibrillation: The ENGAGE AF-TIMI 48 Echocardiographic Substudy. J Am Soc Echocardiogr. 2016 Apr 20. 16. Krezowski JT, Wilson BD, McGann CJ, Marrouche NF, Akoum N. Changes in left ventricular filling parameters following catheter ablation of atrial fibrillation. J Interv Card Electrophysiol. 2016 Apr 13.
12
ACCEPTED MANUSCRIPT 17. Rene AG, Généreux P, Ezekowitz M et al. Impact of atrial fibrillation in patients with STelevation myocardial infarction treated with percutaneous coronary intervention (from the HORIZONS-AMI [Harmonizing Outcomes With Revascularization and Stents in Acute
PT
Myocardial Infarction] trial). Am J Cardiol. 2014 Jan 15;113(2):236-42. 18. Consuegra-Sánchez L, Melgarejo-Moreno A, Galcerá-Tomás J et al. Short- and long-
RI
term prognosis of previous and new-onset atrial fibrillation in ST-segment elevation
SC
acute myocardial infarction. Rev Esp Cardiol (Engl Ed). 2015 Jan;68(1):31-8. 19. Beukema RJ, Elvan A, Ottervanger JP et al. Atrial fibrillation after but not before primary
MA
Neth Heart J. 2012 Apr;20(4):155-60.
NU
angioplasty for ST-segment elevation myocardial infarction of prognostic importance.
20. Jons C, Jacobsen UG, Joergensen RM et al. Cardiac Arrhythmias and Risk Stratification after Acute Myocardial Infarction (CARISMA) Study Group. The incidence and
TE
D
prognostic significance of new-onset atrial fibrillation in patients with acute myocardial infarction and left ventricular systolic dysfunction: a CARISMA sub study. Heart Rhythm.
AC CE P
2011 Mar;8(3):342-8.
21. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA. 2001 Jun 13;285(22):2864-70. 22. Olesen JB, Lip GY, Hansen ML, Validation of risk stratification schemes for predicting stroke and thromboembolism in patients with atrial fibrillation: nationwide cohort study. BMJ. 2011 Jan 31;342:d124.
13
ACCEPTED MANUSCRIPT Table 1: Baseline patient characteristics.
p Value
NOAF (47)
NO-NOAF (1533)
61.4±13
69.9±13
60.5±12
Male Gender, n (%)
1331 (80.3)
32 (68.1)
1247 (81.3)
Heart Failure, n (%)
134 (8.1)
10 (21.3)
Dyslipidemia, n (%)
777 (46.9)
Hypertension, n (%)
Prior AF (77)
No-Prior AF (1580) <0.01
0.03
52 (67.5)
1279 (80.9)
<0.01
115 (7.5)
<0.01
9 (11.7)
125 (7.9)
0.28
22 (46.8)
722 (47.1)
0.97
33 (42.9)
744 (47.1)
0.47
710 (42.8)
29 (61.7)
633 (41.3)
<0.01
48 (62.3)
662 (41.9)
<0.01
Past MI, n (%)
177 (10.7)
11 (23.4)
152 (9.9)
<0.01
14 (18.2)
163 (10.3)
0.03
Diabetes Mellitus, n (%)
359 (21.7)
14 (29.8)
328 (21.4)
0.2
17 (22.1)
342 (21.6)
0.88
Tabaco Use, n (%)
832 (50.2)
14 (29.8)
798 (52.1)
<0.01
20 (26.0)
812 (51.4)
<0.01
Family Hx.
290 (17.5)
4 (8.5%)
279 (18.2)
0.12
7 (9.1)
283 (17.9)
0.04
454±631
362±565
0.28
353±545
365±568
0.86
3 (6.4)
52 (3.4)
0.22
7 (9.1)
55 (3.5)
0.02
33 (2.0)
0 (0%)
31 (2%)
1.0
2 (2.6)
31 (2.0)
0.66
74 (4.5)
8 (17.0)
54 (3.5)
<0.01
12 (15.6)
62 (3.9)
<0.01
102 (6.2)
5 (10.6)
87 (5.7)
0.19
10 (13.0)
92 (5.8)
0.02
51 (3.1)
6 (12.8)
43 (2.8)
<0.01
2 (2.6)
49 (3.1)
1
1.15±0.26
1.17±0.28
1.14±0.26
0.42
1.2±0.25
1.14±0.26
0.05
1.2±0.5
1.41±0.66
1.18±0.47
0.01
1.47±0.8
1.19±0.48
<0.01
1246±1378
1008±1004
1260±13878
0.22
1121±1391
1252±1378
0.42
47.6±8
45.2±7.6
47.8±7.9
0.03
44.4±9.1
47.7±7.9
<0.01
CHA2DS2, mean±SD
0.95±1.0
1.66±1.3
0.89±0.9
<0.01
1.79±1.4
0.91±1.0
<0.01
CHA2DS2VASC2, mean±SD
1.55±1.5
2.77±1.8
1.44±1.4
<0.01
3.0±1.8
1.48±1.4
<0.01
IABC, n (%)
62 (3.7)
CABG, n (%) Mechanical Ventilation, n (%) VT/VF, n (%) Bradycardia
Creatinine admission, mean±SD Creatinine peak, mean±SD CPK, mean±SD EF, mean±SD
NU
MA
D
AC CE P
365±566
SC
60.8±12
Time to ER, n (%)
<0.01
p Value
74.5±11
TE
Age, mean ± SD
Prior AF vs. no prior AF (n=1657)
PT
NOAF vs. no NOAF (n=1580)
RI
Total (n=1657)
Coefficients
Abbreviations: MI, Myocardial infarction; ER, Emergency Room; IABC, Intra-Aortic Balloon Counterpulsation; CABG, Coronary artery bypass graft; VT/VF, Ventricular Tachycardia/Ventricular Fibrillation, CPK, Creatine PhosphoKinase; EF, Ejection Fraction.
14
ACCEPTED MANUSCRIPT Table 2: Multivariate analysis for predictors of new onset atrial fibrillation following STEMI adjusted odds ratios and 95% CI.
1.04 1.14 1.36 1.94 0.58 2.62 0.7
Lower
Upper
1.01 0.56 0.71 0.89 0.29 1.27 0.24
1.07 2.33 2.59 4.24 1.14 5.39 2.1
AC CE P
TE
D
MA
Abbreviations: MI, Myocardial infarction.
<0.01 0.71 0.34 0.96 0.11 <0.01 0.53
PT
p Value
SC
Age Male Gender Hypertension Heart Failure Tobacco Use Prior MI Creatinine at Admission
95% CI
RI
OR
NU
Coefficients
15
ACCEPTED MANUSCRIPT Table 3.
95% CI Lower
Upper
<0.01 0.75 0.085 <0.01 <0.01 <0.01 0.77 0.43
MA
NU
3.19 2.21 1.04 1.08 2.98 5.04 1.48 1.63
D
No atrial fibrillation Reference Prior atrial fibrillation 2.02 1.28 New Onset atrial fibrillation 1.11 0.56 Male Gender 0.74 0. 52 Age 1.07 1.05 Creatinine admission 2.074 1.44 Heart Failure 3.55 2.49 Diabetes Mellitus 1.05 0.75 Hypertension 1.15 0.81
p Value
RI
HR
SC
Coefficients
PT
Cox proportional hazard model for mortality.
AC CE P
TE
CI = confidence interval; HR = hazard ratio.
16
ACCEPTED MANUSCRIPT Figure legends Figure 1. cumulative rates of mortality in patients with and without prior AF. AF = atrial fibrillation.
PT
Figure 2. cumulative rates of mortality in patients with and without New-onset AF. AF = atrial fibrillation.
RI
Figure 3. cox proportional hazards survival curves. AF = atrial fibrillation. CI = confidence
SC
interval; HR = hazard ratio.
CHADS2VASC2 score. AF = atrial fibrillation.
NU
Figure 4. cumulative rates of mortality in patients with and without Prior AF according to
Figure 5. cumulative rates of mortality in patients with and without New onset AF according to
AC CE P
TE
D
MA
CHADS2VASC2 score. AF = atrial fibrillation.
17
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
NU
Figure 1
18
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
MA
Figure 2
19
Figure 3
AC CE P
TE
D
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
20
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
Figure 4
21
MA
NU
SC
RI
PT
ACCEPTED MANUSCRIPT
AC CE P
TE
D
Figure 5
22