1188
JACC Vol. 17, No.5 April 1991:1188-95
SEMINAR ON BALLOON AORTIC VALVULOPLASTY-V William W. O'Neill, MD, FACC, Guest Editor
Follow-Up Recatheterization After Balloon Aortic Valvuloplasty THOMAS M. BASHORE, MD, FACC, CHARLES J. DAVIDSON, MD, FACC AND THE MANSFIELD SCIENTIFIC AORTIC VALVULOPLASTY REGISTRY INVESTIGATORS Durham, North Carolina
The results of recatheterization were assessed in a select group of 95 patients enrolled in the Mansfield Scientific Aortic Valvuloplasty Registry to determine whether any procedural or patientrelated variables at baseline predicted either initial immediate or follow-up (6.2 ± 3.3 months) results. At the follow-up catheterization, 39 (41 %) of the patients were in improved condition and 56 patients (59%) had recurrence of symptoms, allowing for analysis of the effect of the procedure in two symptomatic patient subsets. In the total group the aortic valve area increased initially from 0.56 ± 0.16 to 0.87 ± 0.27 cm2 but partial return to the baseline valve area was evident at follow-up (0.63 ± 0.25 cm2). Similarly, the mean aortic gradient initially decreased from 72 ± 30 to 35 ± 16 mm Hg but then increased to 55 ± 25 mm Hg at follow-up. Neither the initial nor the late hemodynamic results appeared affected by any definable procedural variable at the time of valvuloplasty, including the maximal diameter of balloons, number of balloons simultaneously used, mean inflation time or total number of inflations.
Such technical concerns also did not seem to affect short- or long-term outcome. Similarly, no baseline hemodynamic variable clearly separated those who became increasingly symptomatic from those whose condition was improved at the 6 month interval. At recatheterization, a reduction in the aortic valve area toward baseline was observed in 24 (62 %) of the 39 improved patients and in 45 (80%) of the 56 who were symptomatic. At follow-up the less symptomatic group did have a greater aortic valve area (0.70 ± 0.32 versus 0.58 ± 0.18 cm 2; p = 0.02) and a slightly higher ejection fraction (59 ± 19% versus 48 ± 20%; P = 0.07) than the symptomatic group. Thus, in this select group of patients, differences in procedural techniques at the time of aortic valvuloplasty did not appear to influence the return toward the baseline aortic valve gradient at follow-up. Fewer symptoms at follow-up were observed in those patients achieving the greatest aortic valve area after the procedure and in those with maintenance of left ventricular contractile performance. (J Am Coli CardioI1991;17:1188-95)
Percutaneous balloon aortic valvuloplasty has emerged as a therapeutic option in selected patients with severe aortic stenosis who are poor candidates for surgery (1-5). However, clinical studies (2,6,7) have revealed that evidence for restenosis, with or without return of clinical symptoms, occurs frequently. The acute relief of aortic obstruction observed using balloon catheter techniques appears to result from fracture of valvular calcium, tearing of commissural scar and perhaps stretching of the cusps or the aortic anulus (8-13). In the early days after the procedure, partial return of a pressure gradient occurs because of both remodeling of the aortic valve and an improvement in stroke volume (14,15). Restenosis may be an inevitable consequence at some point,
after the procedure, although the mechanism of restenosis has yet to be well defined and its definition is controversial especially in light of the vagaries in the measurement of aortic valve area. It remains unclear, for example, which factors, such as valve architecture (12), underlying ventricular function (16-18) or variations in the procedure itself (19-23), might be most important in delaying restenosis or the return of symptoms. Preliminary studies (3) have suggested that the final aortic valve area probably best predicts early restenosis, and it has been advocated that a final aortic valve area> 1.0 cm2 after valvuloplasty should always be sought. To date, follow-up data in patients after aortic valvuloplasty have primarily used noninvasive methods, such as Doppler echocardiography, to estimate the severity of the valvular gradient or to measure aortic valve area. Although Doppler gradients correlate reasonably well with invasive measures, substantial standard errors for the measurement of the mean aortic gradient and of the aortic valve area using the continuity equation have been described (24). For these
Manuscript received September 10, 1990; revised manuscript received October 6, 1990, accepted October 24, 1990. Parts I through IV of this Seminar appeared in the January, February, March 1 and March 15, 1991 issues of the Journal (Vol 17, pages 187-98.485-91. 828-33 and 909-13). Address for reprints: Thomas M. Bashore. MD, Cardiovascular Laboratory, Box 3012, Duke Medical Center. Durham. North Carolina 27710. ©1991 by the American College of Cardiology
0735-1097/91/$3.50
BASHORE ET AL. RECATHETERIZATION AFTER AORTIC VALVULOPLASTY
JACC Vol. 17, No, 5 April 1991:1188-95
reasons, follow-up invasive data are important to allow for comparison of serial hemodynamic changes with use of similar techniques. Analysis of data in the Mansfield Scientific Aortic Valvuloplasty Registry from December I, 1986 to September I, 1988 reveals that 95 patients underwent repeat catheterization after aortic valvuloplasty. This time interval is greater than that reported elsewhere in this seminar. Both patient characteristics and details of the technique at the time of the initial procedure were available for review. This presentation is an assessment of the effect of these variables on the subsequent follow-up invasive hemodynamic outcome In these patients.
Methods The Mansfield Scientific Aortic Valvuloplasty Registry. The Registry is a corporate-sponsored data base established through Mansfield Scientific and conducted under Food and Drug Administration-approved investigation device exemption. Clinical sites were required to complete a questionnaire regarding a series of procedural events, hemodynamic results and complications in all patients undergoing aortic valvuloplasty using balloon valvuloplasty catheters designed by the parent company. In addition to clinical follow-up, data from repeat catheterization were also sought. During this time, 785 patients were enrolled in the Registry. At the time of this report, 525 patients were both alive and eligible for a 6 month follow-up catheterization. Follow-up catheterization was not a routine requirement; therefore, many patients underwent recatheterization because of recurrence of symptoms. In some centers recatheterization was incorporated into an institutionally approved protocol for the evaluation of the aortic valvuloplasty procedure. Therefore, the Registry included a mixed group of patients who had undergone repeat catheterization. Of the 525 eligible patients, 95 (18%) underwent recatheterization and these patients form the basis of this report. An aortic valvuloplasty data base form allowed recording of information regarding the valvuloplasty procedure, as well as pertinent hemodynamic data and clinical symptoms. The procedural data base included the number of balloons used, the maximal diameter of each balloon, the balloon length, the number of inflations, the duration of each inflation and information on whether balloon rupture occurred and whether any complications ensued. A baseline data base was also derived regarding the symptomatic status of each patient. This included documentation of fatigue, shortness of breath, angina, syncope, congestive heart failure class (New York Heart Association) and angina class (Canadian Cardiovascular Society). The overall clinical status was also recorded as stable, improving or worsening. Recatheterization data were tabulated in a manner similar to that recorded at baseline. Recatheterization was performed at an average of 6.2 ± 3.3 months after the
1189
baseline procedure. At recatheterization 42 patients (44%) underwent repeat aortic valvuloplasty. Another five patients had repeat valvuloplasty during the same readmission. Thirteen patients (14%) underwent aortic valve replacement during the follow-up hospital admission and seven patients (7%) died during that hospital stay. Patient characteristics. Table 1 outlines pertinent baseline clinical descriptors before valvuloplasty in the 95 patients who underwent recatheterization and contrasts them with descriptors in the subgroup of patients who did not undergo repeat catheterization. The two groups had similar baseline characteristics except for slightly more coronary artery disease and slightly less congestive heart failure in the study cohort. At 6 months more symptoms were present in the recatheterization group. Because 30% of patients died before they were eligible for recatheterization, the significance of this observation is unclear except that it confirms the view that many patients were restudied because symptoms were present. Because all descriptors were not recorded for all data items, Table 1 includes the actual number of patients from whom the data were derived for each variable. Overall, the patient group is elderly with mild to moderate coronary artery disease, moderately well preserved ventricular function and a modest degree of aortic regurgitation. Table 2 outlines the symptomatic status of the 95 patients at baseline and at the time of the repeat study. Symptomatic improvement at recatheterization was evident in regard to heart failure classification, syncope and angina. However, the overall number of patients with any symptom of fatigue or dyspnea was unchanged. At the time of recatheterization, the condition of 39 (41 %) was considered improved, whereas 56 (59%) had either similar or worse symptoms compared with baseline. Statistical analysis. Data were expressed as the mean values ± SD or as mean values and the 95% confidence limits for the mean. Statistical analysis between groups was performed using either a paired or unpaired t test for continuous data and Fisher's exact test for dichotomous variables. A p value < 0.05 was considered significant.
Results Initial and follow-up hemodynamic changes observed. Table 3 outlines the initial hemodynamic changes observed in the patients at the time of their original aortic valvuloplasty procedure and contrasts the data with data observed at follow-up catheterization. Ejection fraction values obtained immediately after the procedure were not recorded on the data sheets and were not available for review. Figure 1 displays the immediate effects of aortic valvuloplasty on the aortic valve area and contrasts these initial measurements with values observed at follow-up catheterization. It is obvious that a return toward the baseline gradient and aortic valve area was common. When patients with fewer symptoms were compared with those whose
1190
BASHORE ET AL. RECA THETERIZA TION AFTER AORTIC VALVULOPLASTY
JACC Vol. 17, No.5 April 1991: 1188-95
Table 1. Comparison of Patients Undergoing Recatheterization With Patients Who Had No Recatheterization Performed Recatheterization
Baseline characteristic Male(%) Age (yr) CHF NYHA I or II NYHA III or IV Fatigue Dyspnea Angina Syncope Aortic regurgitation oto mild Moderate to severe CAD None or IVD 2VD or 3VD Prevalvuloplasty Hemodynamics AVA (cm 2) Peak to peak aortic gradient (mm Hg) LVEDP (mm Hg) LVEF(%) Postvalvuloplasty Hemodynamics AVA (cm 2) Mean aortic pressure gradient (mm Hg) Follow-up symptoms Mean FlU time (mo.) CHF NYHA I or II NYHA III or IV Fatigue Dyspnea Angina Syncope Improved (%)
No Recatheterization
No.(%)
n
No(%)
n
p Value
47 (49%) 75 ± 10 51 (54%) 23 (24%) 68 (71%) 56 (59%) 83 (87%) 53 (56%) 22 (23%)
95 95 95 91 91 95 95 95 95
310 (45%) 78 ± 5 296 (68%) 84 (19%) 360 (81%) 317 (71%) 431 (91%) 239 (51%) 113 (25%)
690 690 435 444 444 445 472 469 455
NS <0.001 <0.01 NS NS <0.05 NS NS NS
57 (85%) 10 (15%)
67 67
380 (88%) 53 (12%)
433 433
NS NS
70 (74%) 19 (26%)
89 89
424 (76%) 136 (24%)
560 560
NS NS
± 0.16 ± 30 ± 11 ± 20
95 95 95 65
0.50 60 19 49
± 0.18 ± 23 ± 14 ± 19
684 690 670 316
<0.01 <0.001 NS NS
0.87 ± 0.27 35 ± 16
95 95
0.82 ± OJI 30 ± 13
673 678
NS <0.001
6.2 47 (49%) 35 (37%) 54 (57%) 58 (61%) 68 (71%) 31 (33%) II (12%) 39 (41%)
95 95 95 95 95 95 95 95 95
6.7 102 (34%) 241 (62%) 145 (38%) 147 (49%) 185 (60%) 63 (21%) 20 (7%) 21\ (65%)
443 298 386 386 303 307 303 293 325
NS <0.01 <0.0005 <0.005 <0.05 <0.05 <0.025 NS <0.0005
0.56 72 19 49
The baseline characteristics, hemodynamic values before and immediately after valvuloplasty and the symptoms at the follow-up (FlU) interval are displayed for those who had, compared with those who had not, undergone recatheterization. There are no differences noted in any baseline characteristic except for a slight increase in fatigue and congestive heart failure (CHF) in the subsets not undergoing recatheterization. No hemodynamic measurement differed between the two groups. At follow-up the recatheterization group of patients had more symptoms in all categories and were generally less improved. This suggests that in the majority of patients recatheterization was performed because of symptoms. AVA = aortic valve area; CAD = coronary artery disease; LVEDP = left ventricular end-diastolic pressure; LVEF = left ventricular ejection fraction; NYHA = New York Heart Association; VD = vessel disease.
symptoms had returned to baseline severity, there was no difference in any measurable baseline hemodynamic variables. Similarly, when the immediate results of the valvuloplasty were assessed, there was no acute postprocedural hemodynamic variable seen more commonly in the patients with improved status than in those with recurrent symptoms (Table 4). At follow-up catheterization, many patients with improvement demonstrated evidence for a return of the valve gradient. The follow-up aortic valve area was larger in patients who were clinically improved at 6 months. These patients also had a higher ejection fraction and had experienced a greater change in the ejection fraction from baseline (from 50 ± 21% to 59 ± 19%) compared with that in patients whose symptoms had worsened or returned to baseline
Table 2. Symptomatic Status of Patients Undergoing Recatheterization
Congestive heart failure Fatigue Dyspnea NYHA I or II NYHA III or IV Angina Syncope
At Baseline
At 6 Month Recatheterization
p Value
51 (54%) 56 (59%) 83 (87%) 23 (24%) 68 (71%) 53 (53%) 22 (23%)
47 (49%) 58 (61%) 68 (71%) 35 (37%) 54 (57%) 31 (33%) II (12%)
NS NS NS <0.05 <0.05 <0.05 <0.05
The change in symptomatic status in patients undergoing recatheterization is redisplayed from Table I. At recatheterization a reduction in anginal and syncopal symptoms was noted but there was no change in overall symptoms of heart failure in this select subset. More patients reported functional class I or II symptoms at follow-up than at baseline.
BASHORE ET AL. RECATHETERIZATION AFTER AORTIC VALVULOPLASTY
JACC Vol. 17, No.5 April 1991:1188-95
1191
Table 3. Hemodynamic Data
(cm 2)
Aortic valve area Mean aortic gradient (mm Hg) Peak to peak aortic gradient (mm Hg) LVEF(%) LVEDP (mm Hg)
Baseline Before BAV
n
0.56 :!: 72 :!: 59:!: 49 :!: 19:!:
95 95 95 65 95
0.18 30 21 20 11
Immediately After BAV
n
0.87:!: 0.27 35 :!: 16 32:!: 12
95 95 95
At Follow-Up Catheterization 0.63 :!: 55:!: 51 :!: 53 :!: 19:!:
NA 16:!: 9
94
0.25 25 21 21 \0
n
92 94 94 48 95
Baseline, immediate postvalvuloplasty and follow-up catheterization hemodynamic data in 95 patients. The initially improved hemodynamic values observed after balloon aortic valvuloplasty (BAV) have returned toward baseline in many patients. Abbreviations as in Table 1.
levels. This finding may have clinical relevance because it appears symptoms may not recur despite evidence for a return toward baseline of the aortic valve area and gradient in some patients. Analysis of the procedural events and their impact on outcome. Several valvuloplasty technique variables were assessed to determine their relation to either the immediate or at 6 month follow-up catheterization outcome. Figure 2 displays the effect of maximal balloon size (diameter) on the change in aorta valve area observed, the aortic valve area after the valvuloplasty procedure and the aortic valve area at recatheterization. Although there is a slight trend toward a larger resultant aortic valve area at recatheterization when larger balloons were used initially, in general the use of a larger balloon size did not seem to correlate with a better valve area at restudy. The number of patients for whom each balloon size Figure 1. Aortic valve area data in 95 patients. Left panel, the effect of balloon aortic valvuloplasty on the aortic valve area (A VA) immediately after the procedure. Note the upward shift in data points. Right panel, at recatheterization the aortic valve area has returned to baseline in many patients, although there are still more patients above the line of identity than below it. Individual dots often represent more than one patient. 2.0
2.0
1.8
1.8
1.6
1.6
was used is recorded at the bottom of the figure and it is important to note that the overwhelming majority of patients had either a single 20 mm balloon (63%) or a combination of balloons with a total summed diameter of 30 mm 01%). Because of the small number of patients with balloons of other sizes, it is difficult to discern the true effect of balloon size, other than that of the two most commonly used. The effect of the total number of balloon inflations on aortic valve area either immediately after the procedure or at recatheterization is shown in Figure 3. Again, the number of inflations appears to be unrelated to outcome; from two to six inflations were performed in most instances. On the assumption that the duration of each inflation may playa role, the hemodynamic status immediately after the procedure was assessed relative to inflation time. However, Figure 4 reveals no obvious relation between inflation time and aortic valve area. Similar findings were obtained for all hemodynamic variables. The majority of inflations lasted :0:;30 s. When these same technical variables were reassessed in the symptomatic subsets at 6 month follow-up study, no relation between any technical event and symptomatic outcome was found. For instance, the largest diameter balloon or balloons in the patients with symptomatic improvement
•
N
E U ......
1.4
< > < fen
1.2
0
N
E
...... :c I< ()
1.0
1.2 1.0
I W 0.8
0.8
c: < > <
0W f- 0.6
::::l
()
<
1.4
U
0.6 0.4
0.4
••• • •• •
n = 95 0.4 0.6
0.8 1.0 1.2
1.4 1.6 1.8 2.0
AVA BASELINE (cm 2 )
•
0.2
00
0.2
n = 95 0.4 0.6 0.8 1.0
1.2
1.4
1.6 1.8 2.0
AVA BASELINE (cm 2 )
1192
BASHORE ET AL. RECATHETERIZA TION AFTER AORTIC VALVULOPLASTY
lACC Vol. 17, No.5
April 1991:1188-95
Table 4. Comparative Analysis of 39 Patients Symptomatically Improved Versus Those 56 With Similar or Worse Symptoms at the 6 Month Follow-up Catheterization
AVA (cm 2) Baseline Acute post Follow-up Mean aortic gradient (mm Hg) Baseline Acute post Follow-up Ejection fraction (%) Baseline Follow-up LVEDP (mm Hg) Baseline Acute post Follow-up
Improved Compared With Baseline (n = 39)
Similar or Worse Compared With Baseline (n = 56)
P
0.57 ± 0.19 0.91 ± 0.29 0.70 ± 0.32
0.56 ± 0.18 0.85 ± 0.24 0.58 ± 0.18
0.71 0.32 0.02
62 ± 21 32 ± II 50 ± 20
57 ± 20 31 ± 18 52 ± 22
0.20 0.47 0.72
50 ± 21 59 ± 19
47 ± 20 48 ± 20
0.58 0.07
21 ± 14 16 ± 9 20 ± 11
19 ± 12 17 ± 8 18 ± 9
0.42 0.76 0.44
At follow-up catheterization, the aortic valve area (A VA) was greater and the ejection fraction marginally higher in the group of patients with clinical improvement. No other identifiable variable was found that was more common in the improved subset. Abbreviations as in Table I.
was 23 ± 6 mm versus 23 ± 6 mm in those with recurrent symptoms. Likewise, the average inflation time in the improved group was 32 ± 19 s versus 29 ± 27 s in the group with recurrent symptoms (p = 0.56). Seventy patients (74%) had a single balloon procedure; the double balloon technique was used in 25 (26%). After the use of two balloons, the postvalvuloplasty area was 0.85 ± 0.09 cm2 and the mean gradient was 38.5 ± 4.7 mm Hg. With use of a single balloon, the postprocedure aortic valve area was 0.88 ± 0.06 cm2 and the mean gradient was 29.6 ± 2.7 mm Hg. However, this study was not designed to determine if the final aortic valve area is improved by the use of two balloons rather than one. Most patients underwent double balloon valvuloplasty when a single balloon procedure was deemed inadequate.
Figure 2. Effect of maximal balloon sizes. Maximal balloon diameter was determined by using maximal balloon size derived from the maximal diameter of a single inflated balloon or the simple summation of the maximal diameter of two balloons. The relation between this diameter and the acute change (t;) in the aortic valve area (top panel), the acute post valvuloplasty aortic valve area (middle panel) and the aortic valve area at recatheterization (RECATH) (bottom panel) is shown. The number of patients at each maximal balloon diameter size is shown at the bottom of the figure. No clear pattern is evident, although most patients had either a 20 or 30 mm maximal diameter used. Data are expressed as mean ± 95% confidence interval. 0.83
95 PERCENT CONFIDENCE INTERVALS FOR FACTOR MEANS
-0. 17L-:,7 5 "",78-;;2';;-0-=2';;-2-=2';;-3-=2=-7-:!30;;--;!3-::-2-::33;:--;:'35;--.!:38;-'
Discussion
MAX. BALLOON DIAMETER(s)
Aortic valvuloplasty results in initial improvement in the measured outflow gradient in most patients. However, the final aortic valve area that results is frequently less than that observed after aortic valve replacement (24). A return of the aortic valve area toward baseline is probably a certainty at some time after the procedure.
0~L-L-L-L-L-~~~~~~
,5
'8 20 22 23
27 30 32 33
35 38
MAX. BALLOON DIAMETER(s)
1.3
Mechanisms of Restenosis Why restenosis occurs has not been completely elucidated. Our study suggests that there is no measurable technique-related variable that predicts a rapid return of the aortic gradient by 6 months. It is more likely that restenosis is an inherent property of the architecture of the stenotic aortic valve itself. It may also be a function of the status of
:I:
~
o
1.0
1IItt tIl 11I
~~ 0.7 ~ E <~
~ <
0.4 0.1
-0.2
15
18
20 22
23 27
30 32
33
36 38
MAX. BALLOON DIAMETER(s) n
=
3
2 60 1
5
1 10 3
3
3
4
BASHORE ET AL. RECATHETERIZATION AFTER AORTIC VALVULOPLASTY
JACC Vol. 17, No.5 April 1991:1188-95
..... 1.6 N
.3e 1.2 f-
~ 0.8 11.
0.4
0(
>
n-
0(
I IIIIIII I I !f I I I
I
3
1
5 7 9 11 13 NO. OF INFLATIONS
15
1 11 12 8 1413 7 7
2
O~-L~~~~-L~~~~~
3
17
..... 1.4
e
lN
!~U,~,~ ,~ I, I!I!jII I : f
f
::: O. 1 L....J..-'-~-'---'--L-L.-'-.l-.L-'---'--''-'--'---' -.
3 5 7 9 11 13 15 TOTAL NO. OF INFLATIONS
17
Figure 3. Effect of number of inflations on aortic valve area. In the top panel the total number of inflations is displayed relative to the final aortic valve area (A VA) observed after the procedure; in the bottom panel the number of inflations is shown relative to the aortic valve area at recatheterization (RECATH). The number of patients represented is shown within each panel. No obvious relation is observed between the aortic valve area and the frequency of balloon inflation at the time of the valvuloplasty procedure. Data are expressed as mean values ± the 95% confidence interval. underlying left ventricular systolic performance. Patients with a low ejection fraction and a low aortic gradient can safely undergo aortic valvuloplasty (25,26), but baseline ventricular performance appears to profoundly affect eventual outcome (16,17). Role of symptomatic status. Fifty-nine percent of this patient group had recurrent symptoms at recatheterization and 41% were considered clinically improved from their baseline state. Despite symptomatic improvement, 69 patients (72%) had some reduction in the aortic valve area at recatheterization, including 62% of those with symptomatic Figure 4. Relation between average inflation time and the acute aortic valve area (AVA) achieved after the procedure. No relation is demonstrable between the two variables. 2.or--------------------. 1.8 1.6
Ne
1.4
()
-
1.2
«
•••
~'.o
:: •••••
t- 0.8
--
(f)
00.6 CL
_
••
•
•••
._.
..
••••
0.4
..
0.2
n=95
°0~-~370--~6~0-~970-~12~0--~15~0-~18~0~
INFLATION TIME (sec)
1193
improvement and 80% with recurrent symptoms. The follow-up ejection fraction in the improved subgroup was higher than those with recurrent symptoms (59 ± 19% versus 48 ± 20%) despite a similar value at follow-up for the mean aortic gradient (50 ± 20 versus 52 ± 22 mm Hg) and only a slightly better aortic valve area (0.7 ± OJ versus 0.6 ± 0.2 cm2). This finding is consistent with the view that symptomatic status may be more a function of ventricular performance than of the measured aortic valve area or gradient (27) . These data also reemphasize that valvular stenosis often precedes the onset of symptoms in patients with aortic stenosis. Patients undergoing aortic valvuloplasty should, therefore, be closely monitored for both return of the transvalvular gradient and further deterioration of left ventricular performance. The lack of recurrent symptoms did not indicate that the aortic valve gradient had not increased. Conversely, 20% of patients with recurrent symptoms had little evidence for a return of the aortic gradient. Because symptoms alone did not consistently correlate with the status of the aortic outflow obstruction, the implication is that many patients suffer from the combined effects of impaired myocardial function together with variable degrees of outflow obstruction. Symptomatic status appears particularly dependent on diastolic ventricular performance characteristics after aortic valvuloplasty (27) and this may explain the apparent discrepancy observed in this study. Role of procedural factors. In addition, this study reveals that follow-up status in these patients is probably not procedure related. Despite the logical concept that aggressive efforts should be made to achieve the largest aortic valve area possible, the initial aortic valve area after valvuloplasty predicted neither symptomatic status nor aortic valve area at follow-up. Technical variables such as maximal balloon size, mean inflation time and the number of inflations seemed to bear no relation to either the initial or the follow-up hemodynamic outcome. These data reemphasize that patientrelated variables are probably more important than procedure-related variables in determining outcome. Valve architecture. The intrinsic valve architecture probably plays a key role in the success of the procedure. For example, it has been observed that a poorer outcome and more frequent complications may result when valvuloplasty is performed on bicuspid aortic valves (12,28,29) and better results may be observed when aortic valve stenosis is due to commissural fusion, as in rheumatic aortic stenosis (30). Most of the patients in this study probably had calcific aortic stenosis with a tricuspid leaflet and variable degrees of commissural fusion. Better characterization of the inherent valve structure in aortic stenosis might allow for better selection of patients who would benefit from aortic valvuloplasty. This has certainly been the case for percutaneous mitral valvotomy (31,32) and may be equally true for aortic valvuloplasty. A similar observation has been made regarding the effect of valvuloplasty on congenital aortic valve
1194
BASHORE ET AL. RECATHETERIZATION AFTER AORTIC VALVULOPLASTY
stenosis, where valve morphology is also a major issue in determining efficacy of the procedure (33). Role of ventricular function. The supporting role that underlying ventricular function plays in preventing return of the valve gradient and in lessening symptoms is of considerable interest. In this study the symptomatic patients had little change in ejection fraction at 6 months, whereas the patients with clinical improvement had an increase from a mean of 50% to 59%. Ejection phase indexes, such as the ejection fraction, are obviously very sensitive to altered afterload (34) and hypertrophy alone leads to some degree of intrinsic myocardial dysfunction (35). Significant improvement in ejection fraction has been observed in a few patients after aortic valvuloplasty (2) and dramatic improvement has occasionally been observed in patients after aortic valve replacement (36,37). Whether the improvement in left ventricular systolic performance is related to inherent augmentation in contractile function or simply to improved ejection dynamics is unclear (38,39). Limitations. This study has several limitations. This Registry was designed simply to tabulate data regarding the aortic valvuloplasty procedure in a systematic manner. For that reason, though prospectively collected, not all baseline and follow-up data were available for all variables, complicating analysis of these data. Whenever possible, we have completed the actual number of data points from which the results were compiled. This limitation is common to registry studies such as this and the data should be interpreted in this light. An additional major limitation is that this is purely an observational study and most patients who underwent recatheterization did so for clinical reasons. In fact, this study represents only 18% of all potentially eligible patients who might have been included. This makes generalizations regarding the population as a whole tenuous. Fortunately, less symptomatic patients were also included because some institutions did attempt to recatheterize patients as part of a routine protocol. This does allow for some comparative observations. Though not the case at the 6 month follow-up, the baseline variables in those restudied were similar to those patients not recatheterized (Table 1). It is important to appreciate that 30% of the patients expired before the 6 month follow-up anniversary. Obviously, the data are biased in favor of patients who were both alive and willing to have a repeat catheterization procedure performed. Symptoms at follow-up were more common in the recatheterization group compared with those who were alive and not undergoing recatheterization. We do not know the role that procedural and other related factors might have in those who are not able or willing to return for repeat catheterization. Conclusions. Because noninvasive measures for assessing the results of aortic valvuloplasty may differ from invasive measures, recatheterization data are important to characterize the results of aortic valvuloplasty in a systemic manner. This study of 95 patients who underwent recathe-
lACC Vol. 17, No.5
April 1991: 1188-95
terization suggest that, despite an early reduction in the aortic outflow gradient, a return to the baseline gradient occurs in a high percentage of patients even as early as 6 months. Restenosis appears unrelated to any identifiable procedural factor. Because symptoms did not always correlate with a return toward the baseline gradient, it is important to follow up all patients closely after the procedure. Improved definition of the structural characteristics of the stenotic aortic valve may result in a better understanding of the long-term results from aortic valvuloplasty than will modifications in the technique of aortic valvuloplasty itself.
References I. Cribier A, Savin T, Berland J, et al. Percutaneous transluminal baUoon valvuloplasty of adult aortic stenosis: report of 92 cases. J Am Coli CardioI1987;9:381-6. 2. Safian RD, Berman AD, Diver DJ, et al. Balloon aortic valvuloplasty in 170 consecutive patients. N Engl J Med 1988;319:125--30. 3. Letac B, Cribier A, Koning R, Bellefleur JP. Results of percutaneous transluminal valvuloplasty in 218 adults with valvular aortic stenosis. Am J Cardiol 1988;62:598-605. 4. Brady ST, Davis CA, Kussmaul WG, Laskey WK, Hirshfield JW, Herrmann HC. Percutaneous aortic balloon valvuloplasty in octogenarians: morbidity and mortality. Ann Intern Med 1989;110:761-6. 5. Dodek A, Hooper RO, Kiess M. Percutaneous balloon valvuloplasty for aortic stenosis: improved quality of life for elderly patients. Can J Cardiol 1988;4:223-7. 6. Block PC, Palacios IF. Clinical and hemodynamic follow-up after percutaneous aortic valvuloplasty in the elderly. Am J CardioI1988;62:76O-3. 7. Nishimura RA, Holmes DR Ir, Reeder GS, et al. Doppler evaluation of results of percutaneous aortic balloon valvuloplasty in calcific aortic stenosis. Circulation 1988;78:791-9. 8. BerdotfRL, Strain J, Crandall C, Ghali V, Golman B. Pathology of aortic valvuloplasty: findings after postmortem successful and failed dilations. Am Heart J 1989;117:688-9. 9. Safian RD, Mandell VS, Thurer RE, et al. Postmortem and intraoperative balloon valvuloplasty of calcific aortic stenosis in elderly patients: mechanisms of successful dilation. J Am Coll Cardiol 1987;9:655-60. 10. McKay RG, Safian RD, Lock JE, et al. Balloon dilatation of calcific aortic stenosis in elderly patients: postmortem, intraoperative, and percutaneous valvuloplasty studies. Circulation 1986;74: 119-25. II. Isner JM, Samuels DA, Siovenkai GA, et al. Mechanism of aortic balloon valVUloplasty: fracture of valvular calcific deposits. Ann Intern Med 1988;108:377-80. 12. Kennedy KD, Hauck AJ, Edwards WD, Holmes DR Jr, Reeder GS, Nishimura RA. Mechanism of reduction of aortic valvular stenosis by percutaneous transluminal balloon valvuloplasty: report of five cases and review of literature. Mayo Clin Proc 1988;63:769-76. 13. McKay CR, Kawanishi DT, Chatetief S, Reid CL, Rahimtoola SH. Stenotic aortic and mitral valves treated with catheter balloon valvuloplasty in a patient with a small valve annulus. Ann Intern Med 1988; 108: 568-70. 14. Davidson CJ, Harpole DA, Kisslo K, et al. Analysis of the early rise in aortic trans valvular gradient after aortic valvuloplasty. Am Heart J 1989;117:411-7. 15. Grollier G, Commeau P, Sesboue B, Huret B, Potier JC, Foucault IP. Short-term clinical and haemodynamic assessment of balloon aortic valvuloplasty in 30 elderly patients: discrepancy between immediate and eight-day haemodynamic values. Eur Heart J 1988;9(suppl E): 155-62. 16. Davidson CJ, Harrison JK, Leithe M, Kisslo KB, Bashore TM. Failure of balloon aortic valvuloplasty to result in sustained clinical improvement in patients with depressed left ventricular function. Am J Cardiol 1990;65: 72-7. 17. Sherman W, Hershman R, Lazzam C, Cohen M, Ambrose J, Gorlin R. Balloon valvuloplasty in adult aortic stenosis: determinants of clinical outcome. Ann Intern Med 1989;110:421-5.
JACC Vo\. 17, No.5 April 1991:1188-95
18. Drobinski G, Lechat P, Metzger JP, Lepailleur C, Vacheron A, Grosgogeat Y. Results of percutaneous catheter valvuloplasty for calcified aortic stenosis in the elderly. Eur Heart J 1987;8:322-8. 19. Rocha P, Baron B, Lacome P, et a1. Aortic percutaneous transluminal valvuloplasty in elderly patients by balloon larger than aortic anulus. Cathet Cardiovasc Diagn 1988;15:81-8. 20. Voudris V, Drobinski G, L'epine Y, Sotirov I, Moussallem N, Canny M. Results of percutaneous valvuloplasty for calcific aortic stenosis with different balloon catheters. Cath Cardiovasc Diagn 1989;17:80-3. 21. Dorros G, Lewin RF, King JF, Janke LM. Percutaneous transluminal valvuloplasty in calcific aortic stenosis: the double balloon technique. Cathet Cardiovasc Diagn 1987;13: 151-6. 22. Meier B, Friedli B, Oberhansli 1. Trefoil balloon for aortic valvuloplasty. Br Heart J 1986;56:292-3. 23. Orme EC, Wray RB, Barry WH, Krueger SK, Mason JW. Comparison of three techniques for percutaneous balloon aortic valvuloplasty of aortic stenosis in adults. Am Heart J 1989;117:11-7. 24. Rahimtoola SH. Perspective on valvular heart disease: an update. J Am Coli CardioI1989;14:1-23. 25. Berland J, Cribier A, Savin T, Lefebvre E, Koning R, Letac B. Percutaneous balloon valvuloplasty in patients with severe aortic stenosis and low ejection fraction. Circulation 1989;79:1189-96. 26. Safian RD, Warren SE, Berman AO, et al. Improvement in symptoms and left ventricular performance after balloon aortic valvuloplasty in patients with aortic stenosis and depressed left ventricular ejection fraction. Circulation 1988;78: 1181-91. 27. Sheikh KH, Davidson CJ, Honan MB, et al. Changes in left ventricular diastolic performance after aortic balloon valvuloplasty: acute and late effects. J Am Coli Cardiol 1990;16:795-803. 28. Davidson CJ, Skelton TN, Kisslo KB, et al. The risk for systemic embolization associated with percutaneous balloon valvuloplasty in adults: a prospective comprehensive evaluation. Ann Intern Med 1988; 108:557-60.
BASHORE ET AL. RECATHETERIZATJON AFfER AORTIC VALVULOPLASTY
1195
29. Isner JM, Salem DN, Desnoyners MR, et al. Treatment of calcific aortic stenosis by balloon valvuloplasty. Am J CardioI1987;59:313-7. 30. Medina A, Bethencourt A, Coello I, et al. Combined percutaneous mitral and aortic balloon valvuloplasty. Am J CardioI1989;64:620-4. 31. Abascal VM, Wilkins GT, Choong CY, et a1. Echocardiographic evaluation of mitral valve structure and function in patients followed for at least 6 months after percutaneous balloon mitral valvuloplasty. J Am Coli CardioI1988;12:606-15. 32. Reid CL, McKay CR, Chandraratna PAN, Kawanishi DT, Rahimtoola SH. Mechanisms of increase in mitral valve area and influence of anatomic features in double balloon, catheter valvuloplasty in adults with rheumatic mitral stenosis: an echocardiographic Doppler study. Circulation 1987;76:628-36. 33. Sholler GF, Keane JF, Perry SB, Sanders SP, Lock JE. Balloon dilatation of congenital aortic valve stenosis: results and influence of technical and morphological features on outcome. Circulation 1988;78:351-60. 34. Gunther S, Grossman W. Determinants of ventricular function in pressure-overload hypertrophy in man. Circulation 1981;64:126-34. 35. Huber D, Grimm J, Koch R, Krayenbuehl HP. Determinants of ejection performance in aortic stenosis. Circulation 1981;64:126-34. 36. Croke RP, Pifarre R, Sullivan H, Gunnar R, Loeb H. Reversal of advanced left ventricular dysfunction following aortic valve replacement for aortic stenosis. Ann Thorac Surg 1977;24:38-43. 37. O'Toole JD, Geiser EA, Reddy PS, Curtiss EI, Landfair RM. Effect of preoperative ejection fraction on survival and hemodynamic improvement following aortic valve replacement. Circulation 1978;58:1175-84. 38. Dorn GW, Donner R, Assey ME, Spann JF, Wiles HB, Carabello BA. Alterations in left ventricular geometry, wall stress and ejection performance after correction of congenital aortic stenosis. Circulation 1988;78: 1358-64. 39. Harpole DH, Davison CJ, Skelton TN, Kisslo KB, Jones RH, Bashore TM. Early and late changes in left ventricular systolic performance after percutaneous aortic balloon valvuloplasty. Am J Cardioll990;66:327-32.