Effect of heart rate control on coagulation status in patients of rheumatic mitral stenosis with atrial fibrillation – A pilot study

Abstract

Background and aim of study

Systemic thromboembolism is a major complication in patients of mitral stenosis (MS) with atrial fibrillation (AF) due to induction of hypercoagulable state. The aim was to assess the relationship, if any, between control of ventricular rate and systemic coagulation factors.

Method

70 patients of moderate to severe MS in AF were studied. 35 patients with average heart rate >100 beats/min over a 24 hour period assessed by Holter monitoring were considered as having a uncontrolled ventricular rate (Group A) and those with average heart rate ≤100 beats/min as controlled ventricular rate (Group B). 30 healthy volunteers acted as controls.

Results

Plasma concentration of prothrombin fragment 1 + 2 (PF1 + 2) 6600 pmol/ml [interquartile range (IQR) 5400.0–9500], thrombin antithrombin III 22.0 ng/ml [IQR 18.6–28.0], and plasminogen activator inhibitor 46.8 ng/ml [IQR 44.0–54.0] were elevated in Group A as compared to Group B (5400 pmol/ml [IQR 3600–7700] p = 0.009, 16.0 ng/ml [IQR 11.0–18.5] p < 0.001, and 25.8 ng/ml [IQR 20.9–34.4] p < 0.001), respectively. A significant correlation was found between heart rate and all three coagulation markers. Multivariate multiple regression analysis showed only heart rate to be an independent predictor of systemic coagulation activation and risk of thrombus formation.

Conclusion

Control of ventricular rate in subjects of MS with AF produces significant reduction in the activation of the coagulation system and may decrease risk of thrombosis.

1. Introduction

Rheumatic heart disease (RHD) is a major health problem in our country. Mitral stenosis (MS) is the most common manifestation, and atrial fibrillation (AF) develops in 40% patients.¹ Though MS per se induces a hypercoagulable state,2, 3, 4, 5 development of AF is an indication for initiating oral anti-coagulant therapy to prevent systemic thromboembolism. However, risk of bleeding often prevents treating physicians of our country from prescribing vitamin K antagonist (VKA) in these patients. This is due to the fact that VKA therapy warrants regular assessment of International Normalized Ratio (INR) to monitor efficacy and safety of drug. However, in our country, RHD often afflicts subjects of poor socioeconomic status, in whom not only lack of facilities to monitor (INR) but also the cost involved makes them reluctant to use VKA. Studies have revealed that rate control strategy is not inferior to rhythm control strategy in preventing cardiovascular morbidity and mortality including risk of thromboembolism in patients of nonvalvular AF.6, 7 Hypothesizing that rate control per se may have a beneficial effect on systemic hypercoagulable state, Atak et al.⁸ in a small interesting study had shown that control of ventricular rate per se with AV nodal blocking drugs produced significant decrease in systemic level of procoagulants in patients of MS with AF. However one the shortcoming of the study was that patients were considered to have controlled or uncontrolled ventricular rate on the basis of a single resting ECG showing ventricular rate ≤100/min and >100/min, respectively. As a result, the level of systemic procoagulants in patients with controlled ventricular rate was reported as similar to those in healthy controls. However, rate control per se is unlikely to decrease level of procoagulants in patients of MS with AF to that in normal subjects. Hence we decided to conduct a pilot study to restudy the systemic coagulation factors in patients with controlled ventricular rate verses uncontrolled ventricular rate (average heart rate ≤100/min or >100/min) assessed over 24 hour period by Holter monitoring. We felt that assessment of the procoagulants in patients with controlled ventricular rate over 24 hours would help to overcome the shortcoming of the previous study.

2. Clinical materials and methods

2.1. Patients

Patients were recruited for the study from the Cardiology Outpatient Department of our hospital, a tertiary care Centre between September 2013 and August 2014. Patients with clinical evidence of significant MS in AF not on any VKA therapy (either never prescribed by treating physician before referral or had stopped therapy for at least 3 weeks prior to presentation in the OPD) were considered for the study. A screening 2D Echo and resting 12 lead ECG was done to confirm the presence of significant MS (moderate to severe; MVA <1.5 cm²) and AF without associated significant mitral regurgitation (MR) (moderate to severe) or aortic valve disease (moderate to severe aortic stenosis and/or aortic regurgitation). All patients underwent detailed Trans Thoracic Echocardiography (TTE) in order to evaluate valvular involvement in the study group and to exclude cardiac disease in control subjects using Philips iE33 fitted with a commercially available 5 MHz transducer. The mitral valve area (MVA) was calculated using both planimetry and pressure half-time methods. Mean transmitral diastolic gradients and pulmonary artery systolic pressure (PASP) were also measured by Doppler studies. All echocardiographic examinations were performed 5–10 times, as patients were in AF⁹; the mean values of the replicate measurements were calculated. Patients with moderate to severe MS (MVA <1.5 cm²) with concomitant mild and/or absent MR were included in the study. Trans Esophageal Echocardiography was done in all cases to look for LA/LAA thrombus and left atrial spontaneous echo contrast (LASEC). Patients with LA/LAA thrombus and LASEC were excluded from study, as like LA thrombus,¹⁰ presence of LASEC has also been shown to increase systemic levels of peptide byproducts of coagulation cascade like PF1 + 2.¹¹ Other exclusion criteria were coexistent left ventricular (LV) dysfunction (ejection fraction <45%), pregnancy, prolonged activated partial prothrombin time (aPTT) or INR, antiplatelet or anticoagulation therapy, diabetes mellitus, renal and hepatic dysfunction, overt malignancy, chronic inflammatory disease, history of systemic and/or pulmonary embolism, history of deep venous thrombosis, and NYHA class IV patients.

After preliminary evaluation, eligible patients were subjected to 24 hour Holter monitoring. We defined controlled ventricular rate as patients having average heart rate ≤100/min along with no individual rate >110% of maximum predicted heart rate for the age of the patient over a 24 hour period as was used in AFFIRM trial.⁷ Based on the reports of the Holter, patients were accordingly stratified as either having controlled (≤100/min) or uncontrolled ventricular rate (>100/min).

We decided to study 100 subjects [70 cases and 30 controls] in this preliminary cross sectional pilot study to assess effect of heart rate on systemic coagulation factors (prothrombin fragment 1 + 2 (PF1 + 2), thrombin antithrombin III (TAT)) and plasminogen activator inhibitor (PAI-1) by ELISA. Accordingly, we recruited 35 patients, each with either controlled or uncontrolled ventricular rate (as defined in our study) and 30 healthy volunteers as control group for our study. The controls had no echocardiographic evidence of structural heart disease, were in sinus rhythm and had all other exclusion criteria identical to study patients. The study protocol was approved by the ethical committee of our institution, and informed consent was obtained from each subject included in the study.

2.2. Hematological investigation

Baseline laboratory tests [(hemogram, erythrocyte segmentation rate, renal function tests, liver function tests, PT (prothrombin time), aPTT, fibrinogen and platelet count)] were done in all subjects of our study (cases and controls). Peripheral venous samples of the study subjects for measuring levels of plasma coagulation parameters, including PAI-1, TAT and PF1 + 2 were drawn in the morning between 8 and 10 AM. Samples were taken using 21-G vacuum tube phlebotomy needles into 3.8% 1:9 trisodium citrate-containing tubes, without venous stasis. The plasma was immediately separated by centrifugation of blood (3000 × g for 15 min), and then stored in several aliquots at −70 °C until used for assay.

Plasma PF1 + 2 (enzyme linked immunosorbent assay; Sincerebio, China; normal human plasma values are 31.2–2000 pmol/ml; intra- and inter-assay coefficient of variation of <8% and <10%, respectively) and plasma TAT (enzyme linked immunosorbent assay; Assaypro, USA; normal human plasma values are 0.5–10 ng/ml; intra- and inter-assay coefficients of variation were <5% and <8%) concentrations as markers of in vivo thrombin generation were measured using solid-phase sandwich enzyme linked immunosorbent assay method, respectively. Measurement of PAI-1 employed a quantitative sandwich enzyme immunoassay (Assaypro, USA; normal plasma values are 5–40 ng/ml; intra- and inter-assay of coefficients of variation were <5% and <8%) and was a reflection of endogenous fibrinolytic activity. Analysis of the plasma samples was done in Department of Hematology, AIIMS, New Delhi.

2.3. Statistical analysis

The data were analyzed by statistical software SPSS version 16 (IL, Chicago). Normality was tested by skewness and kurtosis test and also visually by Box–Whisker plot. One-way analysis of variance was followed by multiple comparison Tukey's test for normally distributed variables and having homogenous variances across the groups. Welch test and Dunnett T3 test were applied, when variances were heterogeneous across the groups. Non-normal variables were compared by performing Kruskal–Wallis test, and pair-wise comparison was done by Mann–Whitney test with Bonferroni adjustment. The Spearman correlation test was done to assess the correlation of coagulation factors with heart rate. To find the predictors of systemic hypercoagulable state in subjects of MS with AF, five potential variables considered in the previous study were included.⁸ The multivariate multiple regression was applied considering the three coagulation markers of our study as dependent and the five variables (heart rate, mean diastolic gradient, PASP, MVA, and LA diameter) as independent variables. The univariate and multivariate results were determined. The normality of residuals was tested for the regression assumption.

3. Results

The clinical, echocardiographic and hematological variables in patients with uncontrolled ventricular rate (Group A) and controlled ventricular rate (Group B) are presented in Table 1, Table 2. There was no significant difference in age or sex ratio between these groups. However patients with uncontrolled ventricular rate were more symptomatic with significantly higher percentage of patients in NYHA functional class III [80% vs 57%, p ≤ 0.05]. There was no significant difference in LA diameter, MVA or PASP between the two groups, but the mean transmitral gradient was significantly higher in Group A compared to Group B (p < 0.01) (Table 2). The platelet count, INR and aPTT were also similar between the two groups, while serum fibrinogen was significantly higher in Group A than Group B (Table 2). The mean heart rate in Group A was 125.6 ± 12.5 beats/min compared to 75.68 ± 11.8 beats/min in Group B patients, and the difference between two groups was statistically significant. Though there was no significant difference in the percentage of patients on beta blockers (β blockers) or non-dihydropyridine calcium channel blocker (CCB), higher percentage of patients in Group B were on digoxin or on combination of digoxin and β blockers or CCB. Further in Group A, majority of patients were on subtherapeutic doses of AV nodal blocking drugs. The mean plasma concentration of PF1 + 2, TAT and PAI-1 was significantly higher in Group A patients compared to Group B and control subject (Table 1). Similarly the plasma concentration of all three coagulation factors was significantly elevated in Group B as compared to control subjects.

Table 1.

Baseline demographics and levels of coagulation factors in patients with uncontrolled (Group A) and controlled heart rate (Group B) and in controls (Group C).

Variable	Mitral stenosis group		Control group C (n = 30)	Over all significance	Multiple comparison (p value)
	Group A (n = 35)	Group B (n = 35)			Group A vs B	Group B vs C	Group A vs C
Age (years)	31.86 (5.24)	28.97 (5.48)	31.14 (3.35)	0.068	NS	NS	NS
Sex (male), n (%)	16 (45.7%)	13 (37.1%)	14 (40.0%)	0.759	NS	NS	NS
PF1 + 2 (pmol/ml)	6600 [5400.0–9500]	5400 [3600–7700]	1110 [360–1750]	<0.001	0.009	<0.001	<0.001
TAT (ng/ml)	22.0 [18.6–28.0]	16.0 [11.0–18.5]	9.30 (6.4–11.30)	<0.001	<0.001	<0.001	<0.001
PAI-1 (ng/ml)	46.8 [44.0–54.0]	25.8 [20.9–34.4]	12.4 (7.5–21.8)	<0.001	<0.001	<0.001	<0.001

Mean (SD), n (%), median [IQR].

NS, non-significant; PAI-1, plasminogen activator inhibitor; TAT, thrombin antithrombin; PF, prothrombin fragment 1 + 2.

Table 2.

Clinical, hematological, and echocardiographic characteristics of patients with uncontrolled and controlled heart rate.

Variable	Mitral stenosis group		p value
	Group A (n = 35)	Group B (n = 35)
Heart rate (beats/min)	125.6 (12.5)	75.68 (11.8)	<0.001
NYHA functional class
NYHA II	7 (20%)	15 (42.9%)	NS
NYHA III	28 (80%)	20 (57.1%)	<0.01
Palpitation	30 (85%)	28 (80%)	NS
Medication for rate control
Digitalis	21 (60%)	31 (89%)	<0.001
Beta blocker	28 (80%)	31 (89%)	NS
Calcium channel blocker	4 (11%)	7 (20.0%)	NS
Digitalis + beta blocker	14 (40%)	26 (74%)	NS
Digitalis + calcium channel blocker	4 (11%)	7 (20%)	NS
Hematological variables
INR	1.04 (0.10)	1.04 (0.07)	NS
APTT	31.8 (5.8)	32.2 (6.0)	NS
Platelets (*109/l)	260 (47)	256 (35)	NS
Fibrinogen (mg/dl)	483 (141.9)	388.7 (132.4)	<0.01
Echo cardio graphic parameters
Mitral valve area (2D, cm2)	0.9 (0.14)	0.87 (0.15)	NS
Left atrium diameter (mm)	47.54 (6.59)	45.26 (5.42)	NS
Mean trans mitral diastolic gradient (mm Hg)	18.57 (5.60)	14.37 (4.29)	<0.01
PASP (mm Hg)	48.59 (12.62)	NS

Mean (SD), n (%).

NS, non-significant; NYHA, New York Heart Association; INR, International Normalized Ratio, APTT, activated partial thromboplastin time; PASP, pulmonary artery systolic pressure.

On univariate multiple linear regression analysis, when the effect of heart rate along with four other variables was regressed on individual coagulation markers (at alpha of 0.05), heart rate was found to have significant effect on the level of all three coagulation markers, while MVA and PASP did not have significant effect on level of any of the coagulation markers (Table 3). Mean diastolic gradient was found to have significant effect on levels of PAI-1 (p = 0.02) and TAT (p = 0.02) but not on PF1 + 2 while LA diameter has significant effect on TAT only (p = 0.017).

Table 3.

Univariate and multivariate multiple regression analyses of various factors predictive of hypercoagulable state in subjects of MS with AF.

Coagulation markers	p value						R2
	Heart rate (beats/min)	MV area (cm2)	LA diameter (mm)	Mean diastolic gradient (mm Hg)	PASP (mm Hg)	All five together
PAI-1 (ng/ml)	<0.001	0.139	0.469	0.020	0.967	<0.001	0.7810
TAT (ng/ml)	<0.001	0.260	0.017	0.018	0.961	<0.001	0.4774
PF1 + 2 (pmol/ml)	<0.001	0.674	0.828	0.124	0.822	<0.001	0.3266
All coagulation markers together	<0.001	0.474	0.069	0.073	0.996	<0.001

PAI-1, plasminogen activator inhibitor; TAT, thrombin antithrombin; PF, prothrombin fragment 1 + 2.

On multivariate multiple linear regression, when all the three coagulation markers were analyzed together, ventricular rate was found to have significant independent effect on the coagulation markers (p < 0.001). LA diameter (p = 0.06) and mean transmitral gradient (p = 0.07) failed to show significant independent effect on the coagulation activity at α = 0.05, but if type I error is relaxed to 8%, then these two variables were also independent predictors of increased systemic coagulation. However MVA and PASP were found to have no significant independent effect on coagulation markers in our study (Table 3).

A significant correlation was found between heart rate and plasma levels of all the coagulation factors assessed by Spearman correlation test [PAI-1: r = 0.711, p < 0.001; TAT: r = 0.545, p = 0.001; PF1 + 2: r = 0.645, p < 0.001].

4. Discussion

Our present study has shown that systemic coagulation activity is significantly increased in patients of MS in AF with fast ventricular rate (average heart rate >100/min) compared to those with controlled heart rate (average heart rate ≤100/min). In addition, mean diastolic gradient was significantly higher in patients with fast ventricular rate and could also be a predictor of systemic coagulation activity. Our study results are in agreement with the study by Atak et al.⁸ To the best of our knowledge, no other study data are available in world literature evaluating the effect of heart rate control on systemic coagulation factors in patients of nonvalvular AF. Unlike the study by Atak et al.,⁸ where there was no statistically significant difference in level of procoagulants in patients with controlled ventricular rate (resting heart rate 79.8 ± 12.1/min) compared to healthy controls, in our study the level of procoagulants was significantly higher in patients with controlled ventricular rate (mean heart rate 75.86 ± 11.8/min) compared to healthy controls (in sinus rhythm with resting heart rate 68.9 ± 9.7/min). So we feel, based on our pilot study results, adequately powered future studies should be planned to assess the effect of strict heart rate control (resting heart rate between 60 and 80 beats/min) compared to moderate rate control (resting heart rate <100/min) in patients with valvular AF.

The mechanism proposed to explain prothrombotic state in MS is obstruction to flow across the mitral valve leading to increase in the pressure in the LA and blood stasis. This phenomenon in turn leads to activation of platelets, the coagulation system and predisposition to thrombus formation.¹² When the heart rate increases, the diastolic filling time decreases more than that of systole and thus results in an increase of the trans mitral gradient, LA pressure and blood stasis in left atrium. Accordingly, there is higher degree of activation of the coagulation system with higher rates compared to lower rates in background of similar degree of stenosis of mitral valve as shown in our study. Similarly AF per se is a risk factor for thrombus formation and systemic thromboembolism due to induction of hypercoagulable state.13, 14, 15, 16 During thrombin generation, the amino terminus of the prothrombin molecule is released as the inactive F 1 + 2 fragment. The generated thrombin is inhibited by the endogenous heparin sulfate-antithrombin III mechanism to form thrombin-antithrombin III (TAT). Thus PF 1 + 2 and TAT have been used as markers of in vivo thrombin generation, and the presence of LA thrombus in MS has been reported to be associated with increased systemic levels of by-products of coagulation cascade like PF1 + 2, TAT, and D-dimer.10, 17 In patients of MS without LA thrombus studies have also reported about increased peripheral venous level of PF1 + 2, TAT and fibrinopeptide A implying that systemic hypercoagulable state precedes and perhaps predisposes to thrombus formation.8, 18 Our study results are also in agreement with these studies. It has been suggested that coagulation activity may be increased locally within the left atrium compared to the peripheral systemic bed in patients of MS with conflicting reports. While Li-Saw-Hee et al.¹⁹ in their study could not find significant difference in level of pro coagulants in LA and peripheral veins of patients with MS and chronic AF, ATAK et al.¹⁰ and Pevrill et al.²⁰ reported increased level of procoagulants in LA compared to systemic venous system in patients of MS either in sinus rhythm or in AF. However, in both of these studies, it was shown that the presence of LASEC in LA was responsible for the enhanced activation of local coagulation system, resulting in higher local level of procoagulants compared to that in the systemic venous system.10, 20 The third factor, PAI-1, we studied is regarded as a major determinant of fibrinolytic activity and high levels denote decreased fibrinolytic activity and a prothrombotic state.

In patients with concomitant MS and AF, the risk of stroke increases 17 fold compared to 5 fold in patients with nonvalvular AF.²¹ Hence these patients are candidates of oral anticoagulant therapy. In patients with AF secondary to mitral valve disease, only VKA are recommended for treatment, in spite of their myriad drug–drug and drug–food interaction, mandating regular assessment of its efficacy and safety by monitoring INR.²² But in a developing country like ours where RHD is a major health problem primarily affecting poor section of the society, prescribing and follow-up of patients on VKA are a big challenge for treating physicians due to reluctance on part of patients to periodically monitor INR. Though the newer anticoagulants evaluated in patients with nonvalvular AF have been found to be safer and more efficacious than VKA,²³ they have not been adequately evaluated in patients with valvular AF and are presently not recommended in this subset of patients.²⁴

Our study has shown that control of ventricular rate per se produces significant decrease in the hypercoagulable state and thus decreases the risk of clot formation and thromboembolism. As rate control decreases the hypercoagulable state, we can infer that it would also decrease the dose of oral anticoagulants required to achieve therapeutic INR in patients with controlled ventricular rate compared to uncontrolled ventricular rate. In this context, we feel that the results of our study should be encouraging for the treating physicians. Aggressive control of ventricular rate in patients with MS with AF will not only decrease the risk of systemic thromboembolism per se but also make therapy with VKA safer by decreasing the dose requirement.

5. Limitation of study

Due to small sample size, we were also unable to compare the coagulation factor levels in subjects with average heart rate ≤80/min compared to those with ≤100/min, which should be assessed in future studies.

6. Conclusion

The control of ventricular rate in subjects of MS with AF produces significant reduction in activation of the coagulation system and may decrease risk of thromboembolism.

Conflicts of interest

The authors have none to declare.