Evaluation of Superficial Femoral Artery Compromise and Limb Growth Retardation After Transfemoral Artery Balloon Dilatations

Inclusion/Exclusion Criteria

Children who underwent balloon dilatation procedures via their femoral arteries and who were being evaluated for their primary cardiac problems in the outpatient clinic were studied. All patients seen during the 12-month period ending August 1994 were included. Patients who had balloon dilatation procedures <12 months before evaluation were excluded from the study. Also excluded were patients whose femoral arteries were used for cannulation during cardiac surgery or repeat balloon dilatation in the interval between the balloon procedure and present study. However, diagnostic retrograde arterial catheterization via the same femoral artery as used for balloon dilatation or via the contralateral artery was not a criterion for exclusion.

Groups

Forty-three children, 18 months to 20 years old at the time of study, met the criteria set forth in the above section and formed group 1. They were 1 day to 15.5 years old (median, 36 months) at the time of balloon dilatation, and 19 of 43 (44%) were <2 years old. Group 2 consisted of 35 children (age range, 26 months to 16 years) who had previously undergone diagnostic femoral arterial catheterization and were 1 day to 12 years old (mean, 37 months) at the time of catheterization. Sixteen of 35 (46%) were <2 years old. The final group, group 3, was composed of 47 children, 8 months to 15.6 years old, who had had neither an interventional procedure nor an arterial catheterization and who served as controls.

Technique of Balloon Dilatation/Catheterization

After administration of a sedative mixture and local Xylocaine anesthesia, percutaneous femoral venous and arterial diagnostic catheterization and cineangiography were performed as appropriate; Desilets and Hoffman's²² modification of Seldinger's²³ technique using short sheaths was used in all balloon cases (group 1) during the diagnostic component of the study. After diagnostic catheterization, an Amplatz extra-stiff exchange guidewire (Cook) was positioned at the desired location through a multipurpose catheter inserted via the arterial sheath. The sheath and the catheter were removed, leaving the guidewire in place. Then the balloon dilatation catheter (without preinflation) was inserted directly over the guidewire, a method similar to Seldinger's technique,²³ and positioned across the stenotic lesion, and balloon angioplasty/valvuloplasty was performed. After completion of balloon dilatation, the balloon catheter was exchanged over a guidewire for a short sheath one French size smaller than the size of the shaft of the balloon dilatation catheter. Postprocedural data were recorded with a catheter introduced via the arterial sheath. In group 2 patients, percutaneous retrograde arterial catheterization was performed via a sheath inserted into the femoral artery by the Desilets-Hoffman²² method. Percutaneous transvenous catheterization via the ipsilateral femoral vein was also performed in 34 of 35 patients (97%).

A bolus of heparan sulfate, 100 U/kg, was administered immediately before introduction of the arterial catheter in both groups 1 and 2. If the procedure time exceeded 60 minutes, an additional 50 U/kg heparin was given. At the conclusion of the procedure, the effect of heparin was not reversed, nor was the heparin continued.

Pedal pulses, Doppler pressures, and capillary filling after catheterization were monitored in group 1 patients. If there was decreased pulse or perfusion in the catheterized extremity, warming of the contralateral extremity was undertaken immediately. If the perfusion deficiency continued to persist beyond 3 to 4 hours, low-molecular-weight dextran or heparan sulfate as a 50-U/kg bolus was given, a heparin drip 20 to 25 U·kg⁻¹·h⁻¹ was started, and partial thromboplastin time was maintained at 1.5 to 2 times the control value. If the pulses and perfusion were not normalized by the next morning, 16 to 20 hours after catheterization, treatment with thrombolytic agents (urokinase or streptokinase) was instituted.

Angiographic Evaluation of Femoral Artery Used for Balloon Dilatation

In patients who underwent balloon dilatation of aortic coarctation or aortic stenosis, evaluation of the femoral artery was undertaken at the time of restudy 1 year after balloon dilatation according to the protocol for follow-up.^{24 25} Successful retrograde catheterization through the femoral artery used for balloon dilatation is considered to indicate patency of that vessel. If the catheterization was performed via the contralateral femoral artery, descending aortography to visualize the femoral artery used in the initial balloon dilatation procedure was performed.

Procedure and Methods of Recording Arterial Pulse and Limb Growth

A cardiology nurse and/or the senior author (P.S.R.) identified patients to be selected for study from the clinic patient list. One of the two investigators (H.Y.L. or S.C.B.R.), without prior knowledge of the patient's status (group 1, 2, or 3), obtained a history pertaining to ischemic symptoms, palpation of femoral pulses, Doppler blood pressures, and measurements of lower extremities as detailed below. The protocol of this study was approved by the Institutional Review Board. Informed consent was obtained in each case.

(1) The history was scrutinized for symptoms of lower-extremity vascular insufficiency, including pain at rest, paresthesia, coolness, color change, intermittent claudication, and ischemic ulceration. (2) Femoral artery pulses in both groins were palpated, and the pulse volume was arbitrarily graded as previously described³ : 0, absent; 1, diminished; 2, normal; 3, slightly hyperactive; and 4, bounding. (3) Doppler-derived blood pressure measurements from both ankles (posterior tibial artery) and both right and left brachial arteries were measured as previously described.²⁶ Two measurements 5 minutes apart were made for each limb. If there was a difference, a third measurement was made 5 minutes later and all three values were averaged. (4) Lower limb length³ was measured with a tape from the anterosuperior iliac spine to the lower end of the medial malleolus with the patient in the supine position. The length of the thigh was measured from the anterosuperior iliac spine to the lower end of the medial femoral condyle. The circumference of the thigh was measured at the gluteal fold, perpendicular to the long axis of the lower extremity. Calf circumference was measured at the maximum calf muscle mass as observed from a lateral profile of the leg. Interobserver variability was studied in 10 randomly selected subjects and was 5%, 7%, and 3% for brachial and ankle Doppler blood pressures and leg length, respectively.

Calculations

The ratio of Doppler-derived ankle blood pressure to brachial pressure (ABI) may indicate the degree of SFA compromise²⁶ and was calculated by indexing ankle to brachial pressure for the balloon or catheterized leg and for the control leg in groups 1 and 2. For group 3, the ratios of right and left ankle pressure, respectively, to right brachial pressure were used for ABIs. In a substantial proportion of our study subjects (group 1), aortic coarctations were treated by surgery or balloon angioplasty, and residual aortic coarctation, if any, may variably affect ABI. Therefore, the ratio of Doppler pressures of interventional ankle/control ankle (AAI) was also calculated. It was thought that this index would remove the effect of coarctation and provide an accurate assessment of SFA compromise in the interventional leg.

Indices of lower limb measurements were also calculated for groups 1 and 2: LLI, TLI, TCI, and CCI as interventional/control measurements. LLD was calculated by subtracting the control leg length from the interventional leg length and was expressed in millimeters. For the control group, right/left extremity measurements were calculated to derive AAI, LLI, TLI, TCI, CCI, and LLD. It was thought that control subjects without balloon intervention or femoral arterial catheterization were necessary because of documented lower limb length inequality (3 to 20 mm; mean, 7 mm) in normal subjects.^{27 28 29}

Statistical Methods

The data are expressed as mean±SD for continuous, normally distributed variables. Medians and ranges are given if the data were not normally distributed. One-way ANOVAs were used for between-group comparisons of normally distributed variables. For variables that were categorical, ordinal, or not normally distributed, between-group comparisons were made with Fisher's exact, Kruskal-Wallis, McNemars, or χ² tests, as appropriate. Simple and multiple linear regression was used to assess the relationship between independent variables such as age and catheter size and various indices of arterial pulse and limb growth. Contingency tables were prepared at arbitrary cutoff points to discern differences in arterial insufficiency and limb growth, and between-group comparisons were made by χ² or Fisher's exact tests. The level of statistical significance was set at P<.05 and Bonferroni-adjusted in the case of multiple comparisons.

Subjects

The ages and weights at evaluation of arterial pulse and limb growth in three groups, namely group 1 children (n=43) who had prior transfemoral artery balloon dilatations, group 2 children (n=35) who had prior transfemoral artery diagnostic catheterization, and group 3 children (n=47) who had neither catheterization nor balloon dilatation (Table 1⇓) were similar (P>.1), as were the ages and weights at intervention and duration of follow-up for groups 1 and 2 (Table 1⇓). The sex distribution was also similar (P>.1).

In group 1, balloon dilatations were performed for relief of obstruction of aortic coarctation in 20 children (native, 16; postsurgical recoarctation, 4), aortic stenosis in 17 (valvar, 10; subvalvar, 7), Blalock-Taussig shunt in 5, and pulmonary venous baffle after Mustard repair for transposition of the great arteries in 1. In group 2, retrograde arterial catheterization was undertaken to evaluate preoperative or postoperative status of cyanotic heart defects in 16 patients, left-to-right shunts in 10, obstructive (acyanotic) lesions in 7, and cardiomyopathy in 2. The diagnoses in group 3 children were left-to-right shunts in 18 (ventricular septal defect, 12; atrial septal defect, 4; patent ductus arteriosus, 2), obstructive lesions in 10 (aortic stenosis, 4; pulmonary stenosis, 6), functional murmur in 11, and miscellaneous cardiac abnormalities in 8 (Kawasaki disease, 4; cardiomyopathy, 2; mitral insufficiency, 2).

In group 1, the catheter shaft size carrying the balloon used for dilatation varied between 5F and 9F, with a median of 5F. The diameter of the balloon was 4 to 20 mm, with a median of 10 mm. The median catheter size used for retrograde arterial catheterization in group 2 was 5F, with a range of 4F to 7F.

Vascular Compromise

None of the 125 patients from any of the three groups had a history or symptoms indicating ischemia. There was a higher (P=.04) incidence of diminished (1+) pulses in the femoral artery used for balloon dilatation (11 [26%] of 43) than in the contralateral femoral artery (4 [9%] of 43). The prevalence of diminished pulses was also greater (P=.018) in the group 1 balloon dilatation extremity (11 [26%] of 43) than in the arterial catheterization (3 [9%] of 35) and right lower extremity (3 [6%] of 47) in groups 2 and 3.

The ABI in the interventional leg in group 1 was less (P=.001) than that in groups 2 and 3 (Table 2⇓) but was similar (P>.1) to that in the control extremity. Twenty (47%) of the group 1 patients had aortic coarctations, and their ABI (0.9±0.14) was lower (P<.001) than that of the noncoarctation group (1.07±0.14); therefore, it was difficult to assess the significance of ABI. For this reason, the AAI was calculated, which may circumvent the coarctation problem. The AAI (Table 2⇓) in the balloon group (0.95±0.13) was lower (P=.023) than that in the arterial catheterization (1.0±0.1) and control (1.01±0.09) groups. A Doppler index ≤0.9 is considered to reflect arterial stenosis or occlusion^{14 30} ; therefore, we used this index cutoff point to determine whether there was a higher prevalence of patients with an AAI ≤0.9 (Fig 1⇓) in group 1 than in groups 2 and 3. As can be seen in Fig 1⇓, the prevalence of subjects with AAI ≤0.9 in group 1 (16 [37%] of 43) was higher (P=.003) than that in groups 2 (6 [17%] of 35) and 3 (4 [8.5%] of 47). Even after group 3 patients with AAI >1.1 were combined with AAI ≤0.9 (8 [17%] of 47), a significant difference (P=.043) between groups remains.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Distributions of ankle/ankle Doppler blood pressure index (AAI) in all three groups. •, AAI ≤0.9; ○, AAI >0.9. The number of subjects in group 1 (16 [37%] of 43) with AAI ≤0.9 is larger (P=.003) than that in groups 2 (6 [17%] of 35) and 3 (4 [8.5%] of 47).

Limb Growth Retardation

The length of the lower extremity and thigh and the circumference of the leg in the balloon extremity (group 1) normalized (indexed) to the control extremity were not significantly different (P>.1) from those of the arterial catheterization (group 2) and right (group 3) lower extremities (Table 3⇓). However, the TCI was lower (P=.001) in group 1 than in groups 2 and 3 (Table 3⇓); the magnitude of this difference, however, is small and may not be clinically significant.

The mean LLD between groups was not significantly different either (Table 3⇑). The numbers of subjects with longer or shorter extremities were not statistically different (P=.89) between the three groups (Table 4⇓). Detailed review of LLD data suggest that although the incidence of significant LLD, defined as a difference ≥5 mm in the balloon group (4 [9.3%] of 43) ascribable to the procedure, is of some concern, a similar (P>.1) magnitude of LLDs (4 [8.5%] of 47) seemed to occur in the control population as well.

Further Characterization of Study Subjects in Group 1

Decreased femoral pulses with a cool extremity on the side used for balloon catheter insertion occurred immediately after balloon dilatation in 4 (9%) of 43 children. Warming of the contralateral extremity alone (in 1) and additional administration of low-molecular-weight dextran (in 1) or heparin (in 2) was undertaken, with improvement of circulation within hours of the procedure. At follow-up, two of these children were found to have blocked (1 complete, 1 partial) femoral arteries, and the other two children had patent femoral arteries. None of the children in this group required thrombolytic therapy.

Twenty-seven of 43 children (63%) were evaluated for patency of the femoral artery 14±10 months after balloon dilatation. Follow-up retrograde arterial catheterization was undertaken in 10 patients via the same femoral artery through which the initial balloon dilatation had been performed; this was considered evidence for patency of the femoral artery. The contralateral femoral artery was used at the follow-up catheterization in 17 patients; descending aortography with angiographic visualization of the femoral artery initially used for balloon dilatation (Fig 2⇓) was performed in these patients. Complete blockage (Fig 2B⇓) was observed in 6 and partial blockage (Fig 2C⇓) in 2. Thus, 8 of 27 patients (30%) had blockage of the femoral arteries, but there was good collateral circulation (Fig 2B⇓). None of these children were symptomatic either at the time of follow-up catheterization or at the time of enrollment into the present study. There is statistical concordance (P=.027) between AAI and femoral artery patency. The AAI in children with blocked femoral arteries (but with “good” collateral circulation) was lower (P=.011) than in children with patent femoral arteries (0.83±0.12 versus 0.96±0.11). In 2 children with stenosed (partially blocked) femoral arteries, the respective AAIs were 0.786 and 1.048. All 6 children with complete blockage of the femoral artery had AAIs ≤0.9. However, an AAI ≤0.9 did not always indicate a blocked femoral artery, in that 8 other children in the study group with AAI ≤0.9 had patent femoral arteries, as demonstrated by angiography.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Selected frames from descending aortograms during filming of femoral arteries. Balloon dilatation had been performed via right (A and B) or left (C) femoral artery 1 year before this study. Catheters (c) were introduced via left (A and B) or right (C) femoral artery, and tips of catheters were positioned in lower part of abdominal descending aorta (not shown). Note good opacification of right iliac and femoral arteries (A) in a patient without blockage of femoral artery. In another child (B), note complete blockage of artery (open arrows). Also note good collateral circulation (arrowheads) opacifying distal femoral artery. In third child (C), there is partial blockage of left femoral artery, which had been used for balloon dilatation 1 year before this study. Ao indicates aorta.

SFA compromise as assessed by AAI was not significantly associated with age and weight at study, age and weight at intervention, duration of follow-up, or the sizes of the shaft of the balloon catheter and of the balloon diameter used for balloon dilatation (Table 5⇓). Similarly, no relationship was observed between LLI and LLD on the one hand and the other independent variables on the other hand (Table 5⇓). Examination by stepwise multiple regression analysis (not shown) did not identify any predictive factors. The AAI was also not significantly different (P>.1) between the subgroups of patients with aortic coarctation (0.927±0.144; n=20), aortic stenosis (0.947±0.125; n=17), and Blalock-Taussig shunt (0.993±0.065; n=5). However, an AAI ≤0.9 was present in 10 (50%) of 20 aortic coarctation, 6 (35%) of 17 aortic stenosis, and 0 (0%) of 5 Blalock-Taussig patients. This may be related to the use of small-diameter balloons (4 to 6 mm) in the subgroup of Blalock-Taussig shunt patients. The LLIs and LLDs were also not different (P>.1) between these subgroups. All 6 children with completely blocked femoral arteries had normal LLIs (1.0) and no LLDs (0). Two children with partial blockage had LLIs of 0.991 and 1.0, respectively, and their corresponding LLDs were −5 and 0.

Simple linear regression of LLI and LLD as a function of AAI (Fig 3⇓) suggested no correlation between AAI and either outcome. In addition, multiple regression analysis showed that AAI was not a predictor of either leg growth retardation parameter even after the covariate effects of age and weight at intervention were controlled for.

• Download figure
• Open in new tab
• Download powerpoint

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Relationship of AAI with LLI (A) and LLD (B); note that there is no correlation, suggesting that leg growth parameters cannot be predicted by AAIs. Controlling for covariate effects of age and weight at intervention did not improve correlation (not shown).

Study Limitations

Although the study subjects did not include all transarterial balloon dilatations performed at our institution, there was no selection bias, in that all patients seen in the outpatient clinic for evaluation of their cardiac defect during the year preceding August 1994 were included. The high incidence of limb growth discrepancy in the control subjects is another issue of concern, but a similar prevalence of LLD has been observed in other normal populations.^{27 28 29} Inaccuracy of tape measurement of limb length may be another source of concern, but interobserver variation was small (3%), and such measurements have been used effectively by other workers.³ Although radiographic bone length (orthoroentgenography or scanography^{1 27 40 41} ) may be more accurate, tape measurements of the lower limb as detailed above were thought to be adequate for this study because of the simplicity, avoidance of radiographic exposure, and lower (or no) expense. The ABIs and AAIs may not reflect arterial occlusion. Although this is an indirect method, its utility has been well demonstrated in other studies.^{14 16 26 30 42} Furthermore, in our own study, there was concordance (P=.027) between angiographically determined femoral arterial patency and AAIs. Finally, the duration of follow-up may be short. As mentioned earlier in the discussion, the AAIs, LLIs, and LLDs are similar (P>.1) in children with follow-up of <3.5 years to those in the group of children with follow-up ≥3.5 years. A longer follow-up duration may be necessary to confirm our results. Retrograde arterial catheterization after balloon dilatation may have influenced the results. The AAI in the group of children without catheterization (1.04±0.12; n=16) is not significantly different (P>.05) from that (0.98±0.09; n=19) in the group who had catheterization but had demonstrated patent femoral arteries. These data suggest that follow-up catheterization did not significantly influence the results. Thus, although a number of potential study limitations are identified, none of them seem to have adversely affected the results of this study.

On the basis of these data, we conclude that although SFA compromise was present in the balloon group, it did not seem to result in limb linear growth retardation but may affect thigh mass growth. This may be related to development of good collateral circulation, which may have blunted the indices of limb growth retardation. It is suggested that (1) study of a larger number of patients at a longer follow-up interval may be warranted and (2) assessment of SFA compromise and measurement of limb growth parameters of patients who have undergone transarterial balloon dilatation procedures are undertaken at the time of evaluation of their primary cardiac condition.