Abstract
In this issue of the World Journal for Pediatric and Congenital Heart Surgery, the article by Rodriguez et al 1 from Children’s Hospital of Eastern Ontario captures what we know and don’t know about a procedure that has been now utilized for over two decades in the treatment of aortic root enlargement for a variety of diagnostic categories, ranging from connective tissue disorders to the broader group of conotruncal abnormalities following repair. Although the rationale for the operation is relatively straightforward (to replace all tissue at risk of rupture or dissection between the aortic annulus and the distal normal caliber aorta), the nuances of the procedure can often make the difference in intermediate-term outcomes. Among these is the management of the aortic valve during valve-sparing aortic root replacement (VSARR). Aortic valve regurgitation can be just as much of an indication for surgical intervention as the dilation of the proximal aorta, especially in patients with congenital heart disease.
The authors embarked on a formal systematic review of the literature over a 20-year period (1999-2019). The search yielded 689 citations, but in the end, only five observational studies met the criteria set forth in the study methods (among which was intentional limitation to cohorts with at least ten patients operated on at 18 years of age or less for aortic root dilatation). Ultimately, based on the authors’ criteria, the review eventually focuses on 81 patients in five series, with a mean age range of 9.9 to 13.9 years.
As mentioned, although VSARR is currently utilized worldwide, this systematic review highlights what we currently know about the procedure. To start with, there is a fact that the literature lacks large pediatric case series with enough granularity to guide decision-making regarding the size of the aorta as an indication for surgery and details of management of aortic regurgitation. And, it is known that the procedure can be performed with very low perioperative mortality and attrition over time and that reintervention rates tend to increase in the long run. Of the total of 16 reinterventions (17.8%) required during the follow-up period of the studies reviewed, 7 (41.2% of all reinterventions) were secondary to aortic valve insufficiency, which seems to be more prevalent in patients undergoing root remodeling. Lastly, the mean follow-up range of only 1.2 to 4.4 years highlights the fact that there is insufficient long-term data to guide us on how best to manage enlarged aortas and regurgitant valves in pediatric patients.
Given the difficulty in finding data to guide evidence-based decision-making, I was kindly asked by our editor to provide a comment about what I think we know about this procedure as it applies to this rather heterogeneous group of young patients and, specifically, to expand on the topic of aortic valve management during VSARR.
Two articles recently published might be of interest to the readership, summarizing the lessons that we have learned with now well over 100 consecutive cases of VSARRs performed in pediatric patients over two decades. 2,3 Our case series has a large prevalence of patients with structurally normal hearts in the setting of connective tissue disorders but also includes patients with conotruncal anomalies and failed Ross procedures. As congenital heart surgeons, we are now also seeing a modest increase in the number of patients who have undergone three-stage palliation of hypoplastic left heart syndrome and present with competent or mildly regurgitant pulmonary (neoaortic) valves and aneurysmal enlargement of the proximal aortopulmonary amalgamation. 4
At which diameter should we intervene? As far as maximal root diameter is concerned, the indications we currently utilize for surgical intervention have been reported elsewhere. 2 In pediatric patients with Marfan syndrome (MFS), we currently use criteria that have been put forward for adult patients, given the rarity of rupture before teen-age years even at sizeable diameters and very high Z-scores. Criteria for intervention are more elusive for patients with Loeys-Dietz syndrome (LDS). Intervention on aortas with diameter between 3.5 and 4 cm is indicated for particularly aggressive genotypes (LDS type I and II) and cases with associated severe craniofacial dysmorphic features and high vertebral artery tortuosity index. 5 In patients with LDS subtypes associated with less aggressive vascular phenotype (LDS IV), we have intervened at diameters above 4.5 cm. In patients with connective tissue disorder, we also advocate prophylactic surgery at lower diameters (eg, not quite exceeding 5 cm in MFS) in the setting of severe family history of aortic catastrophe or worsening degree of aortic regurgitation. In patients with Turner syndrome, we modify the threshold for intervention to when the diameter exceeds 2.5 cm/m2. 6 As far as nonsyndromic patients (those without genetic aortopathies; largely including patients with congenital heart disease and mostly patients with conotruncal anomalies), indication for surgical intervention is often that of severe aortic valve regurgitation, frequently (at least in our experience) secondary to a malformed aortic or neoaortic valve with foreshortened, dysplastic, or severely prolapsing leaflets in the setting of a bicuspid or trileaflet valve. Rupture at diameters less than 5.5 to 6.0 cm is furthermore very rare in this group of patients, 7 and given the high likelihood of valve replacement, decision for intervention should be tailored to the individual patient.
What procedure should we utilize? The risk against which all technical modifications of the procedure are measured is that of early or late need for aortic valve replacement (AVR). In our experience, the reimplantation technique has yielded far superior results than the remodeling approach. The theoretical advantage of stabilizing the root below the nadir of the sinuses in the reimplantation procedure has seemingly paid off in our pediatric patients, with freedom from AVR of 97% at five years. This might be secondary to the large proportion of patients in our practice with connective tissue disorders and the inherent propensity for late enlargement of the ventriculo-aortic junction if not contained by the replacement graft. The report by Myers et al 8 referenced in Rodriguez’ review article implies that the reimplantation technique might be associated with a higher risk of reintervention on the aortic valve. Undersizing of the graft and a higher prevalence of patients with congenital heart disease in the Boston Children’s experience might have accounted for a relative higher need for AVR following VSARR in that series, with a perhaps looser association between the techniques utilized (reimplantation vs remodeling) and the need for AVR. We believe, however, that there is still a substantial role for root remodeling in children and young adults undergoing VSARR. In patients without connective tissue disorders, the annulus is not prone to late enlargement, and root remodeling is most certainly an option. Roots with unfavorable coronary patterns (juxta-commissural location, coronary anomalies, or anterior coronaries after the arterial switch operation, for example) are perhaps best treated with the remodeling technique. A large proportion of patients with repaired congenital heart disease will have dense scarring in the area between the aorta and pulmonary artery as well as below the coronaries, rendering the extensive dissection required for the reimplantation procedure more hazardous and less desirable. Also, in the rare young patient with exceedingly large sinuses and a small (15-17 mm) annulus, root remodeling will allow for growth, delaying annular stabilization either to the time when dilatation occurs or to when the patient outgrows the smaller graft utilized to replace the root and match the distal ascending aorta.
How do we manage the aortic valve? Most patients with enlarged aortas and connective tissue disorders will have central (usually mild or moderate) or mildly eccentric aortic valve regurgitation. Although the sinuses are often asymmetrically enlarged, high reimplantation of the commissural posts in an appropriately sized graft usually restores central coaptation. In these cases, it is imperative to avoid symmetric prolapse with lack of height in the zone of coaptation, as late recurrence of aortic regurgitation is likely. Central plication of the leaflets is usually successful in increasing the effective height of the area of coaptation. 3 Truly asymmetric (noncentral) aortic regurgitation is more difficult to manage and is typically associated with aortic root intervention in the setting of congenital (nonsyndromic) aortic root replacement. In these cases, the root is often asymmetrically enlarged and/or the leaflets might be thickened, retracted, and disclose fusion or fenestrations. We have gone as far as excluding a sinus and bicuspidalizing the neoaortic valve in a child with a completely retracted anterolateral leaflet late after undergoing an arterial switch operation. Aortic root enlargement was corrected in this case with a remodeling technique.
How about the bicuspid aortic valve? In our experience, 12% of patients undergoing root intervention had a bicuspid aortic valve. Most patients with bicuspid aortic valves who have not required open or balloon aortic valvuloplasty earlier in infancy will typically have minimal aortic regurgitation at the time of aortic root intervention. In these cases, the critical step in valve preservation at the time of VSARR is the faithful preservation of the geometry of the valve when reimplanting the commissural posts. Aside from the pure (Sievers type 0) bicuspid morphology or the near 180° opening (Sievers type 2), we find that the reimplantation technique seems to be better than root remodeling in accomplishing appropriate positioning of the posts, at least in our hands. Other groups 9 have however been very successful in repairing and preserving even calcified bicuspid aortic valves with substantial stenosis and/or insufficiency in adult patients undergoing VSARR. For pediatric patients, “repair at all cost” needs to be balanced with the likely need to reintervene in the growing patient, the durability of prostheses, and other issues, such as compliance with medications (anticoagulants) and the desire and possibility of carrying out a successful pregnancy later in life.
When should we not utilize the procedure? As mentioned before, most patients will have a sizeable aortic annulus when VSARR is contemplated. Whereas syndromic children usually have an adult-size annulus even during early childhood, older patients with repaired congenital heart disease are likely to have a normal annulus when an aneurysm is present with associated severe valvular regurgitation and less often stenosis. Therefore, a Bentall-type operation should be considered as a safe and valid alternative. In addition to severe intrinsic valvular anomalies unlikely to sustain repair in the long run, ventricularization of the sinuses, extensive fenestrations, concomitant need for mitral valve replacement, and acute/emergent presentation are also reasons in our opinion to hesitate about repair. Congenitally malformed hearts can also disclose an abnormal relationship of the great vessels (eg, in transposition of the great arteries and double-outlet right ventricle). In these cases, carrying out a deep dissection around the root, exposing, and then recreating a symmetric, competent aortic valve within a graft can be very difficult to achieve. A second period of aortic cross-clamping to replace the aortic valve after a lengthy attempt at repair might be poorly tolerated by the patient; accepting a suboptimal result following repair can on the other hand lead to early need for valve replacement. The high rate of reintervention in patients with aortic enlargement and substantially abnormal valves have led us to often follow the teaching that “a good prosthesis is always better than a bad repair.”