18 years of the Fontan operation at a single institution

In this study by Dr. Lindsay S. Rogers et al. (2012, Journal of the American College of Cardiology 60, 1018-1025), the authors report their experiences of performing the Fontan operation (palliation) on 771 patients from 1992-2009 at the Children’s Hospital of Philadelphia. The authors recorded a variety of variables about the patient (demographic and anatomical) and the actual surgical procedure (e.g, time on cardiopulmonary bypass, amount of drainage from the pleural – chest – tubes, length of stay, readmission to the hospital following the procedure) and report their findings here.

The interesting part of this study is that they split their analysis into 3 different ‘eras’. Era 1 were Fontan operations performed from 1992-1997 (6 years), Era 2 was 1998-2002 (5 years), and Era 3 was 2003-2009 (7 years). This is important because how patients with a congenital heart defect born in 1992-1993 that had the Fontan procedure may have been treated much differently than those born in 2007-2008 that exhibited the same defect and had the same Fontan procedure. Obviously we hope that science and medical research in general should advance how we treat human diseases and how we perform operations and so it is predicted that outcomes for those born in Era 3 (2003-2009) that had the Fontan procedure may have higher survival rates than those born in Era 1 (1992-1997). This study highlights the shift in treating children with congenital heart defects among these three eras. For example, as the authors indicate, Era 1 (1992-1997) represents a time period when most children were treated with the lateral tunnel type of Fontan (see #1 below for an introduction to this point) , during Era 2 (1998-2002), there was a shift towards using the extra-cardiac conduit method (but still relatively equal number of both method) and routine use of the “modified ultrafiltration” method of cardiopulmonary bypass (see link below) started during this Era 2 (in 1996), and in Era 3 most Fontan operations were performed using the extra-cardiac conduit method (though note that this differs from other hospitals such as C.S. Mott Children’s Hospital – see below). Again, this study highlights the importance of allowing researchers to use such data gathered from children having the Fontan procedure as it allows these types of analyses.

Here are the major findings or those that I find interesting:

1) As Fig. 1 indicates, the number of Fontan procedures performed per year at this hospital ranges from 25-70, not terribly high and somewhat surprising to me. It also shows how the type of Fontan performed has changed dramatically across these years. In Era 1 (1992-1997), most Fontan procedures had the lateral tunnel method whereas in the modern era (2003-2009), most Fontan procedures used the extra-cardiac conduit method. From what I understand, the lateral tunnel method is older (well, introduced in 1987) than the extra-cardiac conduit method (introduced in 1990). The lateral tunnel (LT) procedure involves placing a ‘baffle’ (piece of Gore-tex) inside the atrium. One predicted risk of the LT procedure is an increased chance of developing heart rhythm problems, which might not be surprising given that you are sewing something inside of the atrium. The LT procedure is still used by many major hospitals (e.g., at C.S. Mott Children’s Hospital at the University of Michigan, 92% of Fontan procedures performed from 1992-2007 used the LT method: Hirsch et al. 2008, Annals of Surgery 248, 402-410). The extra-cardiac conduit (literally outside the heart tunnel…) or ECC method basically does not involve sewing this baffle into the atrium and is theoretically associated with decreased postoperative complications. This topic (lateral tunnel vs. ECC) should clearly be a focus of a future blog post as I have a major interest in this area. Moving on…

2) From Era 1 (1992-1997), the median age at Fontan was 2.3 years, whereas it was 2.8 years in Era 3 (2003-2009). This is somewhat surprising given our personal experiences that the age of Fontan has been steadily decreasing over the years. Also, as these authors show, the age at stage 2 surgery (e.g., “hemi-Fontan”) decreased from Era 1 (6.4 months) to Era 3 (5.9 months). It would be interesting to conduct a more fine-grained analysis where we look at how differences in age at Fontan affect other parameters (e.g., do children that have Fontan at 18 months have a different outcome than those at 36 months?). As far as I can tell, most studies that have looked at this break the data up into larger chunks (e.g., Fontan performed >3 years or ❤ years as in Shiraishi et al. (2009, Annals Thoracic Surgery 87, 555-561). In this study by Shiraishi et al. (2009), they found that patients with dominant left ventricle (e.g., having Tricuspid atresia or a very small/reduced right ventricle) that had the Fontan procedure performed before 3 years of age had a higher cardiac index (basically heart performance corrected for variation in body size) at 5 and 10 years after operation and higher peak oxygen consumption (might view this as exercise capacity).

On the other hand, this increase in age at Fontan from Era 1 to Era 3 is likely due to the observation that body weight at the time of the Fontan procedure is a predictor of short- and long-term outcomes. Indeed, weight at Fontan from Era 1 (12 pounds) has increased to Era 3 (12.9 pounds).

3) Perhaps the most important part of this paper is the “outcomes” section. The good news is that only 3.5% of the patients that underwent the Fontan procedure died from 1992-2009 (27/771) and the probability of death of the individual <30 days after the Fontan procedure has declined significantly from Era 1 (9.3% died) to Era 3 (1.2%), though there hasn’t been any improvement in increasing survival <30 days after the Fontan from Era 2 (1.0%) to Era 3 (1.2%).

4) More good news is that duration in the ICU, total time in the hospital, and the frequency of lengthy (>14 days)  pleural effusions (drainage from the chest tubes) has declined from Era 1 to Era 3. Though the average i) ICU duration (2-3 days across Era 1 to Era 3) has not changed from 1992-2009, the variation has changed such that were fewer lengthy stays in the ICU in Era 3 (range of stay was 1-45 days) compared to Era 1 (0-181 days). Similarly, the duration of chest tube drainage in Era 1 (mean was 3 days) was similar to Era 3 (4 days) but again the frequency of chest tube drainage that lasted >14 dyas has declined from Era 1 (28.6%) to Era (17.5%). Moreover, the length of hospital stay has declined from Era 1 (12 days) to Era 3 (8 days) and the frequency of lengthy hospital stays (>14 days) after the Fontan has declined from Era 1 (46.7% of Fontan procedures performed here involved patients staying >14 days after procedure) compared to Era 3 (19.5% patients stayed >14 days after Fontan).

5) The final and important part of this paper is indicating the risk factors that predicted whether patients that had the Fontan procedure died, had lengthy hospital stays or chest tube drainage after the Fontan. For death or Fontan takedown within 30 days of the Fontan procedure, those patients that had longer times on deep hypothermic circulatory arrest had an increased probability of dying or Fontan takedown. However, the use of modified ultrafiltration during cardiopulmonary bypass has decreased the risk of death or Fontan takedown, which again is good news and evidence that progress in surgical techniques has benefited patients having the Fontan procedure performed in the modern era. This also highlights the importance of asking your surgeon or their support team how long your child was on cardiopulmonary bypass as it provides some potentially useful information about the future or risks for the future. Also of note, whether patients had the i) lateral tunnel or extra-cardiac conduit method of the Fontan and ii) whether fenestration was or was not used did not affect the risk of death or Fontan takedown within 30 days of the Fontan, though this is again only 30 days after the Fontan and we need to know more about the long-term outcomes of these different procedures.

6) Because the authors found that those patients with longer support times (basically longer cardiopulmonary bypass time) had longer hospital stays and longer periods of chest tube drainage after the Fontan, they also investigated what factors of the patient affected total support time. They found that patients that were larger at the Fontan had longer total support times and that those patients that had the extra-cardiac conduit method also had longer total support times. These are interesting results when comparing the benefits and costs of the lateral tunnel vs. extra-cardiac conduit method as well as the age (and weight) at which to perform the Fontan procedure. However, the longer total support time and the longer time on deep hypothermic circulatory arrest (see above) likely just reflect that the surgery was more complicated because of a complex heart defect. The authors do indicate that their results suggest that the extra-cardiac conduit method is associated with greater short-term complications (longer chest tube drainage and hospital stay duration because of longer total support time) but the preferential use of this procedure over the lateral tunnel method is because it is thought to lower the risk of long-term complications (heart rhythm problems associated with lateral tunnel methods). Yet, we need to see those data showing a reduction in heart rhythm problems in order to justify the conclusion that the extra-cardiac conduit method has long-term benefits compared to the lateral tunnel method!

7) The final and interesting point of this paper is found in Table 9, which summarizes the post-operative outcomes from several similar studies using data collected from several hospitals that commonly perform these procedures (e.g., Children’s Hospital of Philadelphia, C.S. Mott Children’s Hospital at the University of Michigan, Children’s Mercy Hospitals and Clinics in Kansas City, Children’s Hospital of Wisconsin, Children’s Hospital in Boston,  etc.). What is interesting here is that you might make some comparisons among the different studies (and really hospitals) for mortality rates, hospital stay times, bypass times, etc. Though I will resist making comparisons here because really it isn’t good science to make comparisons when the hospitals and surgeons take in different numbers of patients (some do more than others) and take in patients with varying degrees of difficulty (who may have higher mortality rates). For example, at the Children’s Mercy Hospitals and Clinics in Kansas City, surgeons performed 145 Fontan procedures (all nonfenestrated, extra-cardiac conduit method) from 1997-2008 and 5.5% of those patients died and 2.8% of those patients had Fontan takedown. In contrast, at Children’s Hospital of Wisconsin in Milwaukee, surgeons performed 256 Fontan procedures (fenestration used selectively) from 1994-2007, 2.0% of those patients died and 0.8% of those patients had Fontan takedown. These differences may or may not be statistically different from one another and we do not know if the degree of complexity of the defects treated at the two hospitals differ from one another.

Link to this paper:


Link to information about Modified Ultrafiltration during cardiopulmonary bypass: