Abstract
The electrical remodelling is considered to play a role in promoting arrhythmogenic substrate of atrial fibrillation (AF), and intracellular calcium overload may play a key role, especially in its early phase. The effect of oral verapamil on repetitive paroxysmal AF (PAF) was evaluated in clinical cases.
Thirty-five patients with repetitive PAF (total PAF duration >2/24 h) were divided into two groups with and without verapamil administration (240 mg/day) and they were followed-up for 12 months. Before and after the follow-up period, 24 h Holter ECG was recorded. In each Holter recording, total PAF duration and the longest PAF duration was evaluated and spectral analysis was performed for fibrillation waves in PAF episodes to evaluate the fibrillatory frequency. Total PAF duration was prolonged by 45 ± 79 min in the control group ( n = 18) whereas shortened by 25 ± 55 min in the verapamil group ( n = 17, P = 0.005). The fibrillatory frequency was increased from 5.66 ± 1.05 to 6.73 ± 1.02 Hz in the control group and was unchanged in the verapamil group. There was inverse relationship between Δtotal PAF duration and Δfibrillatory frequency ( P = 0.0002).
Verapamil prevented the increase in fibrillatory frequency in PAF patients in relatively long-term observation. Verapamil might be effective for prevention of the electrophysiological change and increase in PAF episodes at least in specific type of PAF cases.
Introduction
Electrical remodelling is considered as one of the important mechanisms of changing a paroxysmal atrial fibrillation (PAF) to a chronic AF. 1–4 In experimental studies, although the effect of L-type calcium channel blocker, verapamil on the electrical remodelling is still controversial, 5 , 6 several reports documented negative results for prevention of electrical remodelling, especially in relatively long-term observation. 7 , 8 Although, there are a few reports documenting the attenuation of the increase in atrial fibrillatory frequency by verapamil administration in clinical cases with long-lasting atrial fibrillation, 9 , 10 a preventive effect of verapamil on clinical PAF is unclear. However, L-type calcium channel blocker may effective for prevention of the progression of electrical remodelling in cases with repetitive PAF because early phase of electrical remodelling, i.e. the initial phase of each episode of PAF, is frequently repeated in those cases. 11 In the present study, we hypothesized that long-term administration of verapamil would be effective for preventing the increase in PAF episodes in clinical cases with repetitive PAF, and chronic effect of oral verapamil on the PAF episodes and atrial fibrillatory frequency was evaluated in cases with repetitive PAF in a prospective manner.
Methods
Patients
The subjects consisted of 35 cases with PAF, 15 were female, 20 were male, and the mean age was 60 ± 9 years old. As the criteria for patient selection, total PAF duration should be longer than 2 h during 48 h (twice 24 h) monitoring as a result of repetitive PAF. Cases with atrial enlargement (left atrial dimension >45 mm) or obvious structural heart disease were excluded to avoid the influence of concomitant cardioactive medications. The cases taking other antiarrhythmic drugs, including digitalis, were excluded because those medications would affect on PAF episodes and atrial electrophysiological properties. 12 The cases who suffer severe symptom during PAF were also excluded because they had to be assigned to active rhythm or rate control therapies. The 35 cases were randomly divided into two groups with and without verapamil administration at a dose of 240 mg per day, i.e. verapamil and control groups, and they were followed-up for 12 months. All patients were prescribed optimal dose of warfarin or aspirin to prevent cardiogenic embolism. All studies were performed under the permission of the Clinical Studies and Ethics Committee of Kitasato University Hospital.
Holter monitoring and evaluation of PAF episodes
Holter monitoring was performed in each patient before and after the follow-up period (Model 563 StrataScan, Del Mar Avionics Co. Ltd., Boston, MA, USA). In the Holter recording, total duration of PAF (total PAF duration) and the longest duration of PAF (longest PAF duration) were evaluated as previously described. 13 For the analysis of the data, the PAF episodes with the duration >5 s were detected as PAF in the trend data in the Holter system. The appearance of the PAF episode was confirmed in the minimized ECG trace, and the total and the longest durations of PAF could be measured in the trend graph. 13 The changes in total PAF duration and longest PAF duration was calculated as follows: [Δtotal PAF duration] = [total PAF duration after the follow-up] − [total PAF duration before the follow-up], [Δlongest PAF duration] = [longest PAF duration after the follow-up] − [longest PAF duration before the follow-up]; therefore, positive and negative number indicate increase and decrease in PAF durations, respectively.
Analysis of fibrillation waves
During the PAF episode (>5 min) recorded in the Holter monitoring, spectral analysis of fibrillation waves was performed. 14 , 15 The PAF episodes with shorter duration were eliminated from the analysis to avoid the influence of change in fibrillatory frequency at the time of termination. To eliminate the influence of early change in fibrillatory frequency as the result of short-term electrical remodelling, the analysis was performed in the record within 2 min from the initiation of PAF episode. When plural episodes of AF were recorded in one Holter recording, up to five representative episodes were randomly selected and the mean fibrillatory frequency of those episodes was employed as the data. The data from the surface ECG lead of CM5, which was digitally stored on-line on a microcomputer at sampling rate of 1 kHz were used and the QRS–T complexes were subtracted using a template matching algorithm. 11 , 12 Frequency analysis was performed off-line on a microcomputer (Bimutas II, Kissei Komtec Co. Ltd., Tokyo, Japan). Frequency analysis of the subtracted ECG involved three steps, including (i) bandpass filtering, (ii) application of a Hamming window, and (iii) 4096-point fast Fourier transformation. A 50% overlap of adjacent spectral analyses allowed the use of an average of 20 epochs of analyses within a single 44 s data set. 15 Power spectra were quantified by measuring the peak frequency signal with the maximum magnitude derived from each epoch. The most dominant frequency of the spectrum in the 3–12 Hz range was defined as fibrillatory frequency in each epoch, and was averaged from 20 epochs. As the index for the change in fibrillatory frequency during the follow-up period, Δfibrillatory frequency was calculated by subtracting baseline fibrillatory frequency from fibrillatory frequency after the follow-up period. 14
Statistics
All values are expressed as means ± standard deviation. Statistical analysis was performed with one-way ANOVA test. A P -value of <0.05 was considered significant.
Results
There was no significant difference in clinical characteristics between the two groups with and without verapamil before the observation ( Table 1 ). All patients were followed-up for 12 months without adding any cardioactive medications. Although the total PAF duration was increased in some cases, no patients suffered from persistent or permanent AF during the observation period.
Comparison of the data between the control and verapamil groups
| Verapamil group | Control group | P -value | |
|---|---|---|---|
| Basic clinical characteristics | |||
| n (cases) | 17 | 18 | |
| Age (years old) | 60 ± 9 (42–75) | 61 ± 8 (46–76) | 0.828 |
| Gender | 7 female, 10 male | 8 female, 10 male | 0.845 |
| LAD (mm) | 35 ± 4 (29–42) | 36 ± 4 (29–42) | 0.329 |
| Basic diseases | Mild HT 3 | Mild HT 4 | 0.898 |
| Concomitant medications | Warfarin 17 | Warfarin 18 | 0.998 |
| Total VR under observation (b.p.m.) | 72 ± 11 (54–90) | 75 ± 6 (64–91) | 0.291 |
| VR during PAF under observation (b.p.m.) | 87 ± 14 (64–112) | 96 ± 14 (68–122) | 0.086 |
| Parameters for PAF | |||
| Before observation | |||
| Total PAF duration (min) | 174 ± 53 (126–336) | 173 ± 45 (120–276) | 0.955 |
| Longest PAF duration (min) | 53 ± 41 (8–167) | 46 ± 33 (12–110) | 0.608 |
| Fibrillatory frequency during PAF (Hz) | 5.67 ± 1.00 (3.98–7.58) | 5.66 ± 1.05 (3.91–8.33) | 0.985 |
| After observation | |||
| Total PAF duration (min) | 148 ± 48 (44–226) | 218 ± 85 (52–442) | 0.006* |
| Longest PAF duration (min) | 33 ± 25 (10–102) | 64 ± 45 (12–162) | 0.019* |
| Fibrillatory frequency during PAF (Hz) | 5.70 ± 1.13 (4.07–7.75) | 6.73 ± 1.02 (4.93–8.93) | 0.004* |
| Changes after the observation | |||
| Δtotal PAF duration (min) | −26 ± 55 (−130 to +56) | +45 ± 79 (−100 to +216) | 0.005* |
| Δlongest PAF duration (min) | −19 ± 38 (−131 to +35) | +17 ± 26 (−24 to +100) | 0.002* |
| Δfibrillatory frequency during PAF (Hz) | −0.07 ± 0.89 (−1.62 to +1.46) | +1.07 ± 0.90 (−0.32 to +3.08) | 0.001* |
| Verapamil group | Control group | P -value | |
|---|---|---|---|
| Basic clinical characteristics | |||
| n (cases) | 17 | 18 | |
| Age (years old) | 60 ± 9 (42–75) | 61 ± 8 (46–76) | 0.828 |
| Gender | 7 female, 10 male | 8 female, 10 male | 0.845 |
| LAD (mm) | 35 ± 4 (29–42) | 36 ± 4 (29–42) | 0.329 |
| Basic diseases | Mild HT 3 | Mild HT 4 | 0.898 |
| Concomitant medications | Warfarin 17 | Warfarin 18 | 0.998 |
| Total VR under observation (b.p.m.) | 72 ± 11 (54–90) | 75 ± 6 (64–91) | 0.291 |
| VR during PAF under observation (b.p.m.) | 87 ± 14 (64–112) | 96 ± 14 (68–122) | 0.086 |
| Parameters for PAF | |||
| Before observation | |||
| Total PAF duration (min) | 174 ± 53 (126–336) | 173 ± 45 (120–276) | 0.955 |
| Longest PAF duration (min) | 53 ± 41 (8–167) | 46 ± 33 (12–110) | 0.608 |
| Fibrillatory frequency during PAF (Hz) | 5.67 ± 1.00 (3.98–7.58) | 5.66 ± 1.05 (3.91–8.33) | 0.985 |
| After observation | |||
| Total PAF duration (min) | 148 ± 48 (44–226) | 218 ± 85 (52–442) | 0.006* |
| Longest PAF duration (min) | 33 ± 25 (10–102) | 64 ± 45 (12–162) | 0.019* |
| Fibrillatory frequency during PAF (Hz) | 5.70 ± 1.13 (4.07–7.75) | 6.73 ± 1.02 (4.93–8.93) | 0.004* |
| Changes after the observation | |||
| Δtotal PAF duration (min) | −26 ± 55 (−130 to +56) | +45 ± 79 (−100 to +216) | 0.005* |
| Δlongest PAF duration (min) | −19 ± 38 (−131 to +35) | +17 ± 26 (−24 to +100) | 0.002* |
| Δfibrillatory frequency during PAF (Hz) | −0.07 ± 0.89 (−1.62 to +1.46) | +1.07 ± 0.90 (−0.32 to +3.08) | 0.001* |
LAD, left atrial dimension; HT, hypertension; PAF, paroxysmal atrial fibrillation; VR, ventricular rate.
Numbers in parentheses indicate the ranges of data.
*Indicates statistical significance.
Comparison of the data between the control and verapamil groups
| Verapamil group | Control group | P -value | |
|---|---|---|---|
| Basic clinical characteristics | |||
| n (cases) | 17 | 18 | |
| Age (years old) | 60 ± 9 (42–75) | 61 ± 8 (46–76) | 0.828 |
| Gender | 7 female, 10 male | 8 female, 10 male | 0.845 |
| LAD (mm) | 35 ± 4 (29–42) | 36 ± 4 (29–42) | 0.329 |
| Basic diseases | Mild HT 3 | Mild HT 4 | 0.898 |
| Concomitant medications | Warfarin 17 | Warfarin 18 | 0.998 |
| Total VR under observation (b.p.m.) | 72 ± 11 (54–90) | 75 ± 6 (64–91) | 0.291 |
| VR during PAF under observation (b.p.m.) | 87 ± 14 (64–112) | 96 ± 14 (68–122) | 0.086 |
| Parameters for PAF | |||
| Before observation | |||
| Total PAF duration (min) | 174 ± 53 (126–336) | 173 ± 45 (120–276) | 0.955 |
| Longest PAF duration (min) | 53 ± 41 (8–167) | 46 ± 33 (12–110) | 0.608 |
| Fibrillatory frequency during PAF (Hz) | 5.67 ± 1.00 (3.98–7.58) | 5.66 ± 1.05 (3.91–8.33) | 0.985 |
| After observation | |||
| Total PAF duration (min) | 148 ± 48 (44–226) | 218 ± 85 (52–442) | 0.006* |
| Longest PAF duration (min) | 33 ± 25 (10–102) | 64 ± 45 (12–162) | 0.019* |
| Fibrillatory frequency during PAF (Hz) | 5.70 ± 1.13 (4.07–7.75) | 6.73 ± 1.02 (4.93–8.93) | 0.004* |
| Changes after the observation | |||
| Δtotal PAF duration (min) | −26 ± 55 (−130 to +56) | +45 ± 79 (−100 to +216) | 0.005* |
| Δlongest PAF duration (min) | −19 ± 38 (−131 to +35) | +17 ± 26 (−24 to +100) | 0.002* |
| Δfibrillatory frequency during PAF (Hz) | −0.07 ± 0.89 (−1.62 to +1.46) | +1.07 ± 0.90 (−0.32 to +3.08) | 0.001* |
| Verapamil group | Control group | P -value | |
|---|---|---|---|
| Basic clinical characteristics | |||
| n (cases) | 17 | 18 | |
| Age (years old) | 60 ± 9 (42–75) | 61 ± 8 (46–76) | 0.828 |
| Gender | 7 female, 10 male | 8 female, 10 male | 0.845 |
| LAD (mm) | 35 ± 4 (29–42) | 36 ± 4 (29–42) | 0.329 |
| Basic diseases | Mild HT 3 | Mild HT 4 | 0.898 |
| Concomitant medications | Warfarin 17 | Warfarin 18 | 0.998 |
| Total VR under observation (b.p.m.) | 72 ± 11 (54–90) | 75 ± 6 (64–91) | 0.291 |
| VR during PAF under observation (b.p.m.) | 87 ± 14 (64–112) | 96 ± 14 (68–122) | 0.086 |
| Parameters for PAF | |||
| Before observation | |||
| Total PAF duration (min) | 174 ± 53 (126–336) | 173 ± 45 (120–276) | 0.955 |
| Longest PAF duration (min) | 53 ± 41 (8–167) | 46 ± 33 (12–110) | 0.608 |
| Fibrillatory frequency during PAF (Hz) | 5.67 ± 1.00 (3.98–7.58) | 5.66 ± 1.05 (3.91–8.33) | 0.985 |
| After observation | |||
| Total PAF duration (min) | 148 ± 48 (44–226) | 218 ± 85 (52–442) | 0.006* |
| Longest PAF duration (min) | 33 ± 25 (10–102) | 64 ± 45 (12–162) | 0.019* |
| Fibrillatory frequency during PAF (Hz) | 5.70 ± 1.13 (4.07–7.75) | 6.73 ± 1.02 (4.93–8.93) | 0.004* |
| Changes after the observation | |||
| Δtotal PAF duration (min) | −26 ± 55 (−130 to +56) | +45 ± 79 (−100 to +216) | 0.005* |
| Δlongest PAF duration (min) | −19 ± 38 (−131 to +35) | +17 ± 26 (−24 to +100) | 0.002* |
| Δfibrillatory frequency during PAF (Hz) | −0.07 ± 0.89 (−1.62 to +1.46) | +1.07 ± 0.90 (−0.32 to +3.08) | 0.001* |
LAD, left atrial dimension; HT, hypertension; PAF, paroxysmal atrial fibrillation; VR, ventricular rate.
Numbers in parentheses indicate the ranges of data.
*Indicates statistical significance.
PAF duration in Holter monitoring
Table 1 shows the whole data obtained from the two groups before and after the observation period. There were no significant differences between the two groups before the observation, but total PAF duration and longest PAF duration became longer in the control group than the verapamil group after the observation period. Figure 1 shows the changes in total PAF duration and longest PAF duration after the observation period in the two groups with and without verapamil. The Δtotal PAF duration and the Δlongest PAF duration in the verapamil group showed negative numbers indicating the decrease in duration of PAF in the Holter monitoring. In contrast, the control group showed positive numbers in these parameters indicating the increase in PAF duration. There were significant differences between the two groups in these parameters.
ΔTotal paroxysmal AF (PAF) duration and Δlongest PAF duration in the two groups. The left and right panels show Δtotal PAF duration and Δlongest PAF duration, respectively. Closed circles show the individual data, and thick bars indicate the mean and the range of standard deviation. The control group showed an increase in PAF frequency and duration whereas verapamil group showed a decrease in PAF frequency. See text for details.
The comparisons of the data along the time course were summarized in Table 2 , and Figure 2 shows the change in total PAF duration during the observation period in the two groups. Although total PAF duration tended to be prolonged in the control and rather shortened in verapamil groups, the differences were not significant.
The change in total paroxysmal AF (PAF) duration along the time course. Each thin line indicates the individual measurements of one patient. Left and right sides of each panel show the data before and after the follow-up period, respectively. Although total PAF duration tended to be prolonged in the control and rather shortened in verapamil groups, the differences were not significant. See text for details.
Changes in parameters after the observation
| Before | After | P -value | |
|---|---|---|---|
| Control group ( n = 18) | |||
| Total PAF duration (min) | 173 ± 45 (120–276) | 218 ± 85 (52–442) | 0.058 |
| Longest PAF duration (min) | 46 ± 33 (12–110) | 64 ± 45 (12–162) | 0.191 |
| Fibrillatory frequency (Hz) | 5.66 ± 1.05 (3.91–8.33) | 6.73 ± 1.02 (4.93–8.93) | 0.004* |
| Verapamil group ( n = 17) | |||
| Total PAF duration (min) | 174 ± 53 (126–336) | 148 ± 48 (44–226) | 0.139 |
| Longest PAF duration (min) | 53 ± 41 (8–167) | 33 ± 25 (10–102) | 0.107 |
| Fibrillatory frequency (Hz) | 5.67 ± 1.00 (3.98–7.58) | 5.70 ± 1.13 (4.07–7.75) | 0.856 |
| Before | After | P -value | |
|---|---|---|---|
| Control group ( n = 18) | |||
| Total PAF duration (min) | 173 ± 45 (120–276) | 218 ± 85 (52–442) | 0.058 |
| Longest PAF duration (min) | 46 ± 33 (12–110) | 64 ± 45 (12–162) | 0.191 |
| Fibrillatory frequency (Hz) | 5.66 ± 1.05 (3.91–8.33) | 6.73 ± 1.02 (4.93–8.93) | 0.004* |
| Verapamil group ( n = 17) | |||
| Total PAF duration (min) | 174 ± 53 (126–336) | 148 ± 48 (44–226) | 0.139 |
| Longest PAF duration (min) | 53 ± 41 (8–167) | 33 ± 25 (10–102) | 0.107 |
| Fibrillatory frequency (Hz) | 5.67 ± 1.00 (3.98–7.58) | 5.70 ± 1.13 (4.07–7.75) | 0.856 |
PAF, paroxysmal atrial fibrillation.
Numbers in parentheses indicate the ranges of data.
*Indicates statistical significance.
Changes in parameters after the observation
| Before | After | P -value | |
|---|---|---|---|
| Control group ( n = 18) | |||
| Total PAF duration (min) | 173 ± 45 (120–276) | 218 ± 85 (52–442) | 0.058 |
| Longest PAF duration (min) | 46 ± 33 (12–110) | 64 ± 45 (12–162) | 0.191 |
| Fibrillatory frequency (Hz) | 5.66 ± 1.05 (3.91–8.33) | 6.73 ± 1.02 (4.93–8.93) | 0.004* |
| Verapamil group ( n = 17) | |||
| Total PAF duration (min) | 174 ± 53 (126–336) | 148 ± 48 (44–226) | 0.139 |
| Longest PAF duration (min) | 53 ± 41 (8–167) | 33 ± 25 (10–102) | 0.107 |
| Fibrillatory frequency (Hz) | 5.67 ± 1.00 (3.98–7.58) | 5.70 ± 1.13 (4.07–7.75) | 0.856 |
| Before | After | P -value | |
|---|---|---|---|
| Control group ( n = 18) | |||
| Total PAF duration (min) | 173 ± 45 (120–276) | 218 ± 85 (52–442) | 0.058 |
| Longest PAF duration (min) | 46 ± 33 (12–110) | 64 ± 45 (12–162) | 0.191 |
| Fibrillatory frequency (Hz) | 5.66 ± 1.05 (3.91–8.33) | 6.73 ± 1.02 (4.93–8.93) | 0.004* |
| Verapamil group ( n = 17) | |||
| Total PAF duration (min) | 174 ± 53 (126–336) | 148 ± 48 (44–226) | 0.139 |
| Longest PAF duration (min) | 53 ± 41 (8–167) | 33 ± 25 (10–102) | 0.107 |
| Fibrillatory frequency (Hz) | 5.67 ± 1.00 (3.98–7.58) | 5.70 ± 1.13 (4.07–7.75) | 0.856 |
PAF, paroxysmal atrial fibrillation.
Numbers in parentheses indicate the ranges of data.
*Indicates statistical significance.
Analysis of atrial fibrillatory frequency
Figure 3 shows the representative examples of fibrillation wave analysis in the two groups. Figure 3 A exhibits the recordings from a case in the control group. The fibrillatory frequency was 5.62 Hz before the observation period of this study and was increased to 8.24 Hz after the observation period. Figure 3 B shows the recordings from a case in the verapamil group. The fibrillatory frequency before the observation period was 5.59 Hz and that after the observation period was 5.52 Hz, which was almost unchanged during this observation.
Representative examples of atrial fibrillation wave analysis. A and B exhibit the examples of fast Fourier transformation analysis in the control and verapamil groups, respectively. The fibrillatory frequency was increased from 5.62 to 8.24 Hz in a case in the control group whereas it was almost unchanged in a case in the verapamil group. See text for details.
Figure 4 shows the data of the fibrillatory frequency before and after the observation period in the two groups. The fibrillatory frequency was significantly increased in the control group from 5.56 ± 1.05 to 6.73 ± 1.02 Hzl; however, it was unchanged in the verapamil group. Figure 5 shows the relationship between the Δtotal PAF duration and the Δfibrillatory frequency. There was significant direct relationship between them, indicating that the increase in total PAF duration was correlated with the increase in atrial fibrillatory frequency during PAF episode.
Change in fibrillatory frequency during the follow-up period. The left and right sides of each group indicate the data before and after the follow-up period, respectively. Each thin line indicates the individual data of each case. The fibrillatory frequency was increased in the control group; however, it was unchanged in the verapamil group. See text for details.
Relationship between Δtotal paroxysmal AF (PAF) duration and Δfibrillatory frequency. There was significant direct relationship between Δtotal PAF duration and Δfibrillatory frequency in observation of total patients. See text for details.
Discussion
Atrial electrical remodelling and progression of AF
Although the importance of premature atrial contraction as a trigger for AF has been emphasized recently, 16 changes in atrial electrophysiological properties, i.e. shortening of atrial refractoriness or decrease in conduction velocity, are still important to understand the promotion of arrhythmogenic substrate for re-entrant mechanism as a driver for AF. 3 , 4 Atrial electrical remodelling has been proven to be observed even in clinical cases and now widely understood as one of the mechanisms of progression of AF in various situations. 1–4 , 17–19 Our previous report has documented the shortening of f-f interval of fibrillation waves in surface ECG recording along with the clinical time course of AF; 13 therefore, progressive shortening of atrial refractoriness would be corresponded with the increase in frequency and duration of AF in clinical patients with PAF. In the present study, we utilized spectral analysis of the fibrillation waves in surface ECG of Holter monitoring, which is more objective method to analyse electrophysiological properties of fibrillated atrium. 7 , 8 , 20 The result was that increase in fibrillatory frequency, i.e. Δfibrillatory frequency, was corresponded with the increase in total PAF duration, i.e. positive Δtotal PAF duration, so that the progression of shortening of atrial refractoriness was considered to be corresponded with the progression of the atrial electrical remodelling and the increase in PAF frequency.
Possibility of L-type calcium channel blocker for prevention of electrical remodelling
It has been documented that intracellular calcium overload plays an important role in promotion of electrical remodelling, especially in its early phase within 1–3 days. 17 , 18 There are several reports about suppressive effect of verapamil on the shortening of atrial refractoriness within 24–48 h, 2–5 but the effect might be weakened later because L-type calcium channel would be down-regulated at least in mRNA or functional levels during long-term observation as a result of the progression of electrical remodelling. 7 , 17 Our previous experimental study documented that the effect of verapamil appeared differently at least between the right and left atria and it was effective for suppression of shortening of atrial refractoriness in the left atrium even in later phase of the evaluation, 5 so that the effect of verapamil on suppression of the atrial electrical remodelling would be still controversial.
Although the short-term preventive effect of verapamil on the progression of electrical remodelling has been suggested, 1–5 the long-term effects of verapamil on the atrial electrophysiological properties are still unresolved. Theoretically, verapamil may shorten the action potential duration by suppressing calcium influx, but the degree of the effect seems to be low. It has been reported that acute administration of verapamil can cause a shortening of the atrial fibrillatory cycle length. 21 But, it is probably because of acute haemodynamic alterations or changes in the autonomic tones, 9 , 22 and does not match with the result in our study. In contrast, Meurling et al . 9 and Bollmann et al . 10 have reported the attenuation of atrial fibrillatory frequency caused by verapamil administration in patients with long-lasting AF. Although the mechanism of this phenomenon was unclear, the result was compatible to the data in our present study. Wijffels et al . 23 have suggested that the decrease in conduction velocity would be the main factor to increase the fibrillatory cycle length by antiarrhythmic agents in persistent AF, but it is considered that verapamil would not affect on the conduction velocity even though d -isomer of verapamil may have some effect on sodium channels in the ventricle. 24 Decreased ventricular rate might be involved as the mechanism because the lower ventricular rate might result in better haemodynamic situation and lower atrial wall stress during AF. 9 , 10 In our observation, the ventricular rate during PAF tended to be lower in the verapamil group than in the control, but the difference was not significant ( Table 1 ), so that it seems not to be a strong factor in this observation. However, because PAF repeated frequently with short duration in the patients in our observation, it is speculated that early phase of electrical remodelling, i.e. the phase in which calcium overload plays an important role, is also repeated frequently. Therefore, suppression of calcium influx by L-type calcium channel blocker might be effective at least this type of patients by suppressing intracellular calcium overload, i.e. the early phase of the electrical remodelling, in atrial muscle.
Clinical implications
Because the technique for catheter ablation for triggering focus of AF has been developed, especially in younger patients with repetitive PAF, 16 the clinical impact of pharmacological prevention of AF might have been decreased at least in some population. However, there should be large population of ablation refractory patients or patients suitable for pharmacological therapy, so that pharmacological rhythm management is still an important issue for managing clinical patients with AF. Although the potassium or sodium channel blockers may play a main role in rhythm management, the rate control therapy utilizing digitalis, calcium channel blocker, or beta-blocker plays another important role in the management of AF patients. 25 Verapamil is commonly used medication for rate-control in patients with AF and it has been documented that the rate control with verapamil is superior to the other rate control medications, including beta-blockers or digitalis at least in the point of view of quality of life. 26 The usefulness of verapamil in patients with PAF should be reconsidered at least in specific type of patients with PAF like the population in this study.
Limitations
This study includes several limitations. First, the number of the patients and the follow-up period for this observation were limited. Secondly, the evaluating method for PAF duration was depending on 1 or 2 times Holter ECG recording, so that the reproducibility of the data was limited. To minimize this error in the reproducibility, we chose patients with repetitive PAF in this study because these patients used to show relatively reproducible data in total PAF duration. 13 Thirdly, the fibrillatory frequency evaluated in this study may be reflecting the atrial refractoriness of the right atrium but not the left atrium because of the data obtained from the surface ECG. 14 , 20 Finally, because we employ only the single most dominant frequency as the fibrillatory frequency in this study, the changes in frequency distribution were not reflected in the parameters. The usage of the other parameters such as coefficient of variation should be considered. These limitations should be solved in future studies with larger number of patients and the electrophysiological study.
Conclusion
Oral L-type calcium channel blocker, verapamil, was administered for 12 months in patients with frequent type of PAF with no structural heart disease. The Δtotal PAF duration and Δlongest PAF duration were significantly smaller in the verapamil group in comparison with the control group. The fibrillatory frequency during PAF was unchanged in the verapamil group, whereas it was increased in the control group. Verapamil may prevent the increase in PAF frequency possibly through preventing the progression of the atrial electrical remodelling, i.e. the shortening of atrial refractoriness at least in patients with repetitive PAF.
Conflict of interest : none declared.
Funding
This study was partly supported by a grant for scientific research from the Ministry of Education Science and Culture of Japan (No. 20070423).
