Fetal cardiac intervention: Why, who & when?

Director and Senior Lecturer in Perinatal Cardiology, Imperial College,
Queen Charlotte's and the Royal Brompton Hospitals,
London, UK
Helena M. Gardiner

Worldwide experience in fetal valvuloplasty is limited and previously published outcomes have been poor. Recent reports of success in opening aortic and pulmonary valves and the inter-atrial septum have led to renewed interest in these procedures.
Essential steps in introducing and new procedure include an evaluation of WHY valvuloplasty may be of benefit, WHO may benefit and WHEN an intervention should best be undertaken to produce maximum effect. The answers to these questions will necessitate a true understanding of the natural history of the disease process including the in-utero and postnatal course, and surgical outcomes. Ideally data from unselected clearly defined demographic populations should be examined to reduce bias. For pulmonary atresia/intact septum such data exist. These studies confirm outcome of unselected cases detected over the 5 year period 1991-1996 in the UK & Eire. Nine year follow up of the 240 cases identified (which included 86 fetal diagnoses) has shown that one third of the 183 liveborn infants died and of those surviving only 49% had a final biventricular circulation. Small single centre series from other countries support these results.
The surgical outcomes for critical aortic stenosis are predominantly single centre reports and there appears to have been considerable improvement over the last 15 years. However, there is likely to be some selection bias in these reports and one of the most useful demographic series is from the Czech Republic examining the outcome following aortic valvuloplasty in a large single centre series. This 5 year follow-up reports on 269 patients and includes outcomes on 80 neonates. Compared to the overall group, the neonates had higher mortality (30% vs 10.4%), a lower surgery-free survival post valvuloplasty (29% vs 50%) and re-intervention-free survival (26% vs 39%).
However, outcome data from neonatal series underestimate the extent and range of the problems seen by the perinatal cardiologist. Additional adverse factors include closure of the oval foramen with hydrops, initiation of severe pulmonary disease, progression of the disease process towards hypoplastic left heart syndrome and considerable secondary damage to the ventricle in response to volume and pressure loading. All these factors reduce the number of neonates with critical valvular disease who actually reach a cardiac centre and thus bias the data towards a more favourable outcome. An understanding of the true natural history of these diseases and the relatively few that achieve a biventricular circulation may encourage a re-evaluation of fetal valvuloplasty.
Valvuloplasty has been performed at Imperial College on both the aortic and pulmonary valves of selected fetuses since 2000 from 21 weeks' gestation. The technical issues are relatively simple to overcome and we perform the procedure as a day case under maternal local anaesthesia with percutaneous access using 16-17.5 gauge needles & coronary balloons from 2.5 to 4.5mm. Other centres have used general anaesthesia and some have performed a mini-laparotomy to gain access. In our experience the procedure has reversed hydrops and prolonged pregnancy with the ultimate aim of reducing secondary ventricular damage and normalising fetal flow patterns.
All of our cases have been technically successful but not without complications and these will be discussed. In the report by Reich et al, a unifoliate, unicommissural aortic valve was seen in 34% of neonates compared with 12% in the infant group. Fetuses presenting with critical aortic or pulmonary stenosis are likely to have a high prevalence of rudimentary unifoliate, unicommissural semilunar valves compared with those that require valvuloplasty later in the first year of life and this may explain why we have repeated the valvuloplasty in one fetus because of closure of the valve six weeks later and observed closure in others.
The effect that an intervention may have on ventricular and valvar growth is difficult to predict, even in the child following postnatal valvuloplasty. All of our cases have required early postnatal procedures but repeat postnatal aortic valvuloplasty may not provide sufficient relief of obstruction and aggressive early postnatal surgery in the form of the Ross-Konno operation may be the procedure of choice in some cases of aortic stenosis.
However, there are still insufficient data to inform a consensus on selection criteria and enable a risk-benefit analysis of fetal intervention. The outcomes of large, perhaps randomised or “intention to treat” series will be required to answer the question whether an intra-uterine intervention will alter the outcome for a specific individual with critical aortic or pulmonary stenosis.