| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Clinical Radiology, University Hospital Mannheim, University of
Heidelberg, Theodor Kutzer Ufer 1-3, Mannheim 68167, Germany.
2 Department of Pediatrics, University Hospital Mannheim, University of
Heidelberg, Mannheim, Germany.
3 Department of Pediatric Surgery, University Hospital Mannheim, University of
Heidelberg, Mannheim, Germany.
Received October 5, 2006;
accepted after revision December 29, 2006.
Address correspondence to K. A. Buesing
(karen.buesing{at}rad.ma.uni-heidelberg.de).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. A total of 30 infants (17 boys, 13 girls; 2 ± 0.26 years) were included. Of these, 18 (60%) infants received arteriovenous ECMO therapy with subsequent reconstruction of the right common carotid artery (RCCA). Two years postoperatively, the children were examined with cerebral MRI, including 3D time-of-flight and contrast-enhanced 3D MR angiography of the intra- and extracranial brain-supplying arteries. The pathologic findings were analyzed for the ability to predict impaired neurologic development.
RESULTS. The RCCA was occluded or highly stenotic in 13 (72%) of 18 children. All infants showed intra- and extracranial collaterals and a patent internal carotid artery. The average duration of ECMO was not longer than in cases of successful reconstruction (p =1). The ECMO group showed a significantly greater incidence of cerebral injuries (p = 0.007) but no relevant impairment in neurologic development compared with controls (p = 0.26). Unsuccessful RCCA repair had no predictive value for a poor neurologic outcome (p =1).
CONCLUSION. The outcome of RCCA repair after ECMO is possibly poorer than expected, with vascular occlusion or high-grade stenosis occurring in almost three quarters of patients. Although reocclusion of the RCCA does not increase the risk for cerebral lesions or an impaired neurologic development during the first 2 years postoperatively, the overall benefit of RCCA repair remains doubtful, and the potential long-term risk arising from these plaques has yet to be assessed.
Keywords: carotid artery • congenital diaphragmatic hernia • extracorporeal membrane oxygenation therapy • MR angiography
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Using Bartlett's classic venoarterial perfusion technique, the right common carotid artery (RCCA) and the internal jugular vein are ligated. Thus, despite the life-saving character of this method, ligation of the RCCA has often been the primary objection to using this therapy. Although the increase in severe neurologic disabilities such as infantile cerebral palsy is insignificant [2], up to 75% of surviving CDH patients will display neurodevelopmental problems [6, 7].
To possibly prevent further cerebral complications, revascularization surgery of the RCCA was performed after arteriovenous cannulation. So far, several authors have shown a good short-term vessel patency [8–12], but there is still no evidence of any improvement in neurologic development after RCCA repair as compared with ligation [1, 8, 9, 13]. Indeed, results may worsen with increasing follow-up intervals [14, 15], and the structural and neurodevelopmental impact of reocclusion of the RCCA on the patient is still unknown.
In all investigations to date, the patient populations were inhomogeneous, with a variety of underlying diagnoses. However, except for CDH, most causes of neonatal respiratory failure are self-limited, and ECMO allows time for the lung to recover from the underlying disease process. Thus, among all neonates undergoing ECMO therapy, CDH patients have the worst prognosis, with cumulative survival statistics of up to 58% compared with 94–100% for meconium aspiration [9, 16], and a significantly higher morbidity of neurologic origin [1, 6, 7, 9]. These statistics emphasize that a separate analysis of CDH children is warranted to assess the therapeutic benefits from RCCA repair.
The purpose of this study was to evaluate carotid artery repair 2 years after neonatal ECMO therapy in a homogeneous group of CDH patients, focusing in particular on the structural and neurologic impact of vascular reocclusion. Using MRI and, to our knowledge, for the first time contrast-enhanced MR angiography, we prospectively assessed the condition of the reconstructed carotid artery and the incidence of cerebral injuries in a single comprehensive scan. The neurologic development was correlated with the neuroimaging results and compared with the outcome in children who did not receive ECMO therapy for CDH.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Informed parental consent was obtained for all infants enrolled in the study, and separate consents were obtained for the ECMO therapy and for carotid artery reconstruction surgery. The study was approved by the local research ethics committee.
ECMO Therapy
Children received ECMO therapy according to the following postnatal
management schedule: All neonates were intubated immediately after birth and
given gentle conventional ventilation. If the postductal
PO2 was < 50 mm Hg (or saturation < 85%) at 1
hour, inhaled nitric oxide was added to the ventilation therapy. Infants in
whom PaCO2 exceeded 75 mm Hg were switched from
conventional ventilation to high-frequency oscillatory ventilation (mean
airway pressure, 20 cm H2O; pressure difference, < 40 cm
H2O; frequency, 8–12 Hz). Patients were placed on ECMO
therapy if the postductal PaO2 failed to rise above 40
mm Hg and preductal saturation stayed at < 90% for more than 2 hours or if
the postductal PaO2 failed to rise above 50 mm Hg and
preductal saturation remained at < 95% for more than 4 hours. ECMO therapy
was also initiated when the difference between pre- and postductal saturation
was greater than 15% for more than 12 hours.
The procedure was performed arteriovenously via cannulation of the RCCA and the internal jugular vein. During ECMO, heparin (Liquemin N 5000, Roche) was given IV at a daily dose of 400 IU/kg.
Arterial Reconstruction
The arterial reconstruction procedure was performed by six pediatric
surgeons, each with at least 7 years of clinical experience, including
pediatric vascular surgery. The number of surgeons involved in the study was
not restricted to minimize an operator-related risk.
At the time of decannulation, the jugular vein was ligated. The arterial catheter was removed and the proximal and distal vessels were secured with vascular clamps. A segment of artery was excised, including the region of the ligatures, and an end-to-end anastomosis was constructed with interrupted 6-0 Prolene sutures (Fig. 1). Postoperatively, IV anticoagulation with heparin at 200 IU/kg was continued for the next 3 days and at 100 IU/kg daily for another 4 weeks.
|
The patients were examined using 1.5-T MR units (Magnetom Sonata or Avanto, Siemens Medical Solutions), a circular polarized head coil, and a body-phased-array coil. Before contrast media injection, cranial MR images were acquired using a spin-echo T1-weighted sequence (TR/TE, 450/11; 5-mm section thickness), a turbo spin-echo T2-weighted sequence (5,000/101; 5-mm section thickness), a T2*-weighted sequence (565/35, 5-mm section thickness), a FLAIR sequence (8,000/120; 5-mm section thickness), and an isotropic diffusion-weighted (DW) single-shot echo-planar sequence (6,000/100; 5-mm section thickness; b values, 0, 500, and 1,000 s/mm2). This was followed by 3D time-of-flight (TOF) angiography of the circle of Willis in all patients. The postoperative state of the thoracocervical vasculature was subsequently assessed by applying fast low-angle shot (FLASH) 3D MR angiographic sequences with bolus administration of a single dose of gadopentetate dimeglumine (Magnevist, Schering).
Imaging Assessment
All MR images were independently assessed by two radiologists with 12 and 6
years of experience, respectively. They were blinded to previous ECMO therapy
and the neurologic status of the children.
Neuroimages were evaluated for cerebral infarction, hemorrhage, or their residua (including porencephaly) and for lateralized or diffuse cerebral atrophy as indicated by unilateral or bilateral ventriculomegaly with enlarged sulci. The 3D TOF MR angiography images were analyzed with particular regard to the right internal carotid artery and an intracranial collateralization of the circle of Willis.
The patency of the reconstructed RCCAs was assessed using contrast-enhanced FLASH 3D MR angiography. A stenosis was classified according to generally accepted grading standards (narrowing < 30% = low, 30–70% = medium, > 70–99% = high grade).
2-Year Follow-Up Neurologic Assessment
Neurologic examinations were performed during hospitalization; before
discharge; and at 3, 6, 12, and 24 months old. Bayley Scales of Infant
Development were administered at each visit by a pediatrician trained to
perform the test.
Statistical Analysis
The Fisher's exact test, the Mann-Whitney U test, or the
chi-square test was applied to evaluate the differences between the two
groups. Data were analyzed using SAS release 8.02 (SAS Institute,
1999–2000) for Windows (Microsoft). A p value < 0.05 was
considered to be significant.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
2-Year Follow-Up
Carotid artery reconstruction and neuroimaging—MR
angiography revealed successful RCCA repair in five (28%) of 18 children
(Fig. 2). The reconstructed
RCCA was found to be occluded in 10 (56%) of 18 children (Fig.
3A,
3B) and another three patients
(17%) presented with a high-grade stenosis
(Fig. 4). Thus, intracranial
blood supply from the reconstructed RCCA was lacking or highly impaired in 13
(72%) of 18 patients. A poor postoperative result was not related to a
significant extension of ECMO time as compared with children with successful
RCCA repair (8.1 ± 2.9 days vs 8 ± 3.1 days, p =1).
|
|
|
|
All infants in whom outcome of the RCCA reconstruction was poor presented with an intra- and extracranial collateral blood supply, with cervical collaterals mainly arising via the external carotid artery. The corresponding internal carotid artery was patent in all patients but reduced in diameter in most cases (Figs. 3A, 3B, 4, 5).
|
Brain imaging was without identifiable abnormalities in 10 (83%) of 12 children in the non-ECMO group. Two boys presented with a mild ventricular dilatation, indicating minor brain atrophy. After neonatal ECMO therapy, however, MRI revealed focal cerebral lesions as a result of previous ischemia or hemorrhage in four (22%) of 18 children (p = 0.13). Of all lesions, only one small infarction could be allocated to the perfusion area of the right internal cerebral artery. The remaining defects were related to the perfusion area of the right posterior circulation (one patient) or the left middle cerebral artery (two patients) (Fig. 6). Furthermore, diffusely enlarged CSF spaces occasionally associated with periventricular white matter changes could be found in 67% (12/18) of all patients who had received ECMO therapy (Fig. 7A, 7B). Compared with controls, this finding was statistically significant (p = 0.007). All focal cerebral lesions were associated with enlarged CSF spaces.
|
|
|
No relevant increase in risk for focal cerebral injuries or distended CSF spaces was seen in patients showing occlusion or stenosis of the reconstructed RCCA as compared with those in whom repair was successful (p = 1). All children of the non-ECMO group showed a normal cerebral vascular supply. The results are summarized in Table 2 and illustrated in Figure 8.
|
|
Assessment of neurologic development— Neurologic development was normal for their age in all children of the non-ECMO group. Among the 18 patients who received ECMO therapy, three children (17%) developed neurologic deficits: one patient suffered from infantile cerebral palsy and two children presented with mental retardation and a limp leg, respectively. This difference did not reach statistical significance, however (p = 0.26) (Table 2 and Fig. 8).
An assessment of focal brain injures and accentuated CSF spaces on MRI revealed an adversely affected neurologic development in two (50%) of four and two (14%) of 14 children, respectively, which was significantly or mildly increased as compared with one (4%) of 26 and one (6%) of 16 infants without corresponding cerebral lesions (p = 0.04 and p = 0.59). However, occlusion or a relevant stenosis after RCCA repair was of no consequence for the neurologic development (p = 1) (Table 3).
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Ours is the first study systematically applying contrast-enhanced 3D FLASH MR angiography for postoperative follow-up of RCCA repair after ECMO therapy. Contrast-enhanced MR angiography is much faster than 3D TOF MR angiography and therefore less susceptible to motion artifacts. The brain parenchyma and the intra- and extracranial brain-supplying arteries can be comprehensively evaluated when MRI, including 3D TOF MR angiography of the circle of Willis with subsequent contrast-enhanced 3D FLASH MR angiography of the thoracocervical vasculature, is used. Hence, a supplementary Doppler study, potentially with additional sedation, can be avoided.
In a cohort of 18 2-year-olds, we found 72% of all reconstructed carotid arteries to be occluded or highly stenotic.
Because the surgical procedure is largely standardized and does not differ from what has previously been described by other investigators [8, 12, 14, 15], the unsatisfactory results probably cannot be explained exclusively by procedural or technical problems. A simple lack of operating experience or an operator-related risk seems fairly unlikely because we deliberately included different surgeons, each with several years of experience in pediatric vascular surgery.
Surgical results depend on the state of the vessel at the time of decannulation [11, 12, 15] and are therefore related to ECMO duration [8, 13, 17]. Because of the underlying respiratory problem, the duration of ECMO is usually longer in children with CDH than in those with other primary diagnoses. However, in our cohort there was no relevant difference in ECMO duration for children in whom RCCA repair was successful as compared with those showing reocclusion (p = 1), and the mean ECMO time is within the scope of recent reports of other study groups [7, 8]. Other risk factors related to the underlying diagnosis are uncertain but seem unlikely. Although the reason for the poor postoperative outcome remains debatable, we assume that the outcome of RCCA repair in neonates after ECMO therapy after the first postoperative year may be poorer than previously expected.
For all CDH patients in whom the state of the RCCA was unfavorable, MR angiography revealed intra- and extracranial collateral vessels that maintained bilateral cerebral perfusion and prevented territorial brain damage. As a result, we found the risk for cerebral injury to be the same in children with reocclusion of the RCCA as in infants with successfully reconstructed vessels. However, because RCCA repair was successful in only a small number of patients in our cohort (28%), this finding may lack statistical power.
Neuroimaging
In view of the fact that ours is the first study focusing on the structural
and clinical impact of RCCA reocclusion, we cannot directly compare our
observation with previously published data. Nevertheless, our findings can be
indirectly assessed and compared with earlier investigations of children with
either successful repair or with permanent ligation of the RCCA. Most previous
studies revealed a lower incidence of brain lesions after RCCA repair than
after RCCA ligation
[10–15],
but these data still need to be evaluated critically. In some studies
assessing patients with RCCA repair, the mean duration of ECMO was relatively
short at 76–135 hours [8,
13]. Consequently, the risk
for cerebral injury is reduced. Furthermore, neuroimaging studies were usually
performed just before discharge. At that time, differences in cerebral
disorders between the study groups are unlikely to be due to a beneficial
effect of the RCCA repair. According to clinical and experimental studies,
early cerebral injury in ECMO patients can mainly be attributed to several
events, including both pre-ECMO- and ECMO-related phenomena, or their combined
effects—for example, exposure to profound hypoxia or asphyxia, systemic
heparinization, alteration of pulsatile flow, and microthrombi from the ECMO
circuit [18].
In addition, the overall incidence of cerebral abnormalities varies noticeably among the studies, which may partially be due to the chosen imaging technique. After RCCA repair, Baumgart et al. [13], Levy et al. [12], and Cheung et al. [15] independently observed cerebral abnormalities in up to 22% of the patients when using CT. On the other hand, based on MRI, Taylor et al. [10] detected various cerebral disorders in five (38%) of eight children after successful reconstructive vascular surgery, whereas Sarioglu et al. [8] revealed cerebral abnormalities in up to 45% of the children. In contrast, Spector et al. [11] evaluated 18 neonates after carotid artery reconstruction but did not find any indication of cerebral disorders on MRI.
Analyzing data from studies in patients with a permanently ligated RCCA, Lago et al. [19] identified focal cerebral lesions in 23% and enlarged CSF spaces in 51% of a cohort of 31 newborns. A few years later, Ahmad et al. [20] investigated 51 children with CDH and found cerebral hemorrhage or infarction in 14% and enlarged CSF spaces in 42%. Also, McGahren et al. [1] identified cerebral abnormalities in 50% (6/12) of patients undergoing ECMO therapy with permanent ligation of the RCCA.
So far, only a few studies have directly compared children undergoing RCCA repair with infants with ligation of the carotid artery. Unfortunately, the results of these studies are contradictory: In cohorts of 140 and 61 patients, respectively, Baumgart et al. [13] and Sarioglu et al. [8] did not observe a relevant difference in either focal or generalized cerebral pathology, whereas Desai et al. [14] found a significantly reduced number of cerebral abnormalities in the carotid repair cohort (26%) compared with a historical control group (52%) in a total patient population of 69.
Thus, with respect to the literature, it is uncertain as to whether reocclusion of the RCCA increases the risk for cerebral lesions compared with children in whom repair was successful. However, in our cohort of 18 ECMO patients, no differences could be detected during the first 2 years postoperatively. Our findings also show that the risk for cerebral lesions in cases of RCCA reocclusion is not higher than in patients with a ligated RCCA. This is almost certainly due to the fact that there are sufficient intra- and extracranial collateral vessels. This theory is supported by data from Hunter et al. [21] and Matsumoto et al. [22] that show only a temporary decrease in right hemispheric cerebral blood flow after carotid artery ligation. In addition, Baumgart et al. [13] did not reveal any differences in neuroimaging and neurocognitive outcome between children with RCCA ligation and those with carotid artery repair despite significantly reduced cerebral blood flow velocities in the right hemispheric cerebral arteries. Furthermore, corresponding to our findings, previous data have shown that the right hemisphere is not predominantly affected by cerebral lesions in patients undergoing ECMO therapy [4, 13, 14, 19, 23].
Our data further imply that children with CDH do not have an increased risk for cerebral lesions compared with other underlying diagnoses of severe respiratory failure.
Carotid Artery Repair and Neurodevelopment
Neurodevelopment has been documented in ECMO survivors, and the overall
results are encouraging [8,
20,
24]. Both short-term and
long-term neurologic complications are mostly related to the occurrence of
severe intracranial hemorrhage or cerebral infarction. Griffin et al.
[25] reported that the absence
of intracranial hemorrhage and cerebral infarction before, during, and after
ECMO has been associated with near to normal short-term outcome. Using
logistic regression, Glass et al.
[26] showed that the presence
of neurologic abnormalities on CT is the strongest predictor of long-term
outcome. Our results confirm this finding, with focal cerebral lesions being
the only significant indicator for an impaired neurologic development
(p = 0.04) as compared with enlarged CSF spaces (p = 0.59)
and ECMO therapy itself (p = 0.26). However, other investigators
could not verify this association
[1].
Furthermore, our data now show that reocclusion of the RCCA after reconstructive surgery is not a predictive indicator for an impaired short-term neurologic development (p = 1). Although the postoperative outcome of the RCCA repair in our cohort was disappointing, approximately 95% of these children present with a normal neurologic development or suffer from only slight neurologic handicaps (12%) up to 2 years postoperatively. This overall short-term outcome is similar to that in patients with successful reconstruction [8, 13, 14]. Sarioglu et al. [8] evaluated neurocognitive outcome in 18 children with RCCA repair compared with 16 controls with carotid artery ligation and found that 90% of infants who had reconstructive surgery developed normally compared with 56% of the ligation group. Also, Baumgart et al. [13] and Desai et al. [14] independently revealed a normal neurodevelopment in 85% and 75%, respectively, of the children with a reconstructed RCCA as compared with 70% and 68% in the ligation group.
In conclusion, according to the data presently available and including our results, the clinical benefit from RCCA repair after ECMO is still a matter of debate. Although the perioperative risk of vascular reconstruction has been proven to be low [13] and reocclusion of the RCCA does not increase the risk for cerebral lesions or an impaired neurologic development during the first 2 years postoperatively, differences in long-term neurologic morbidity need to be further evaluated. Dodge et al. [27] have described problems in neurologic development not apparent before the age of 2 years. Furthermore, potential long-term risk arising from the postoperative plaques in the RCCA has yet to be assessed.
MRI, including 3D TOF MR angiography and contrast-enhanced 3D MR angiography, is a suitable method for comprehensive evaluation of the cerebrum and intra- and extra-cranial vessels after ECMO therapy.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
