The debate over whether Ashton's deterioration was driven primarily by a failing mitral valve or by an infection appeared to have been firmly settled in the autopsy report.

The autopsy was performed on November 13, 1994, by Dr. Joseph de Nanassy, and was restricted to an examination of the heart and lungs. While a preliminary report was issued on December 13, 1994, the final report was not completed until February 20, 1995. In the preliminary report, there was no mention of a finding of evidence of a viral infection in the lungs.

In the final report, de Nanassy determined there was extensive recent hemorrhage in both lungs and that there was pulmonary necrosis related to cytomegalovirus infection (CMV). He also noted marked bilateral myocardial hypertrophy. He concluded the most likely cause of death was the pulmonary hemorrhage. The implication from this was that the CMV attacked tissue in Ashton's lungs, leading to pulmonary necrosis. The weakening generated by this necrosis in turn could have led to the hemorrhaging and to Ashton's death.

As noted above, CMV is a very serious viral threat to children in Ashton's condition. While Embree had raised it as a possibility, no one involved in Ashton's treatment had thought it likely that he suffered from such an infection. Giddins testified that he was surprised at the results of the autopsy.

Consulting witnesses to this Inquest have come to a different conclusion about the cause of Ashton's death. Soder, Duncan, Cornel and Taylor each concluded that the problem lay with the dysfunctional mitral valve and not with CMV. Taylor testified, in fact, that he did not believe that Ashton had a CMV infection.

When Cornel examined the heart, he said that the degree of disruption of the AV valve was very severe.

The leaflet was almost completely detached. So that the degree of competence of the valve would have been minimal. It is an extremely severe lesion. (Evidence, pages 44,891-44,892)

When asked how this could happen, Cornel testified that the suture line bears the entire stress of left ventricular contraction and pressure.

If there is a tiny defect anywhere in that suture line, there is going to be added stress at that point and it may start to tear. Once a tear has begun, it becomes a progressive weakness, like tearing a stamp along the perforation. So the tear may very well be progressive. It is one of the reasons I don't do this type of repair anymore. There is a higher incidence of leaflet dehiscence in the one patch repair than some other types of repair. (Evidence, page 44,892)

Dehiscence means the coming apart of the surgical incision or suture line. It was Cornel's view that the leaflet tissue remaining on the mitral valve that had been sutured to the patch had come away from the suture line. When asked about his positive assessment of the surgery and the mitral valve regurgitation, Duncan testified:

It doesn't matter which surgeon where does an AV septal defect repair, the left-sided inlet valve will always leak after surgery, some more, some less. It depends how much tissue you have to work with.

This valve was behaving satisfactorily at the completion of the surgery, but I presume either the sutures pulled out or maybe became infected and pulled through, I don't know. But for whatever reason, the valve began to perform much worse in a relatively short period of time and just became progressively more so. (Evidence, pages 41,452-41,453)

De Nanassy disagreed that there was valve dehiscence, because, he said, there was no valve tissue to dehisce in the first place. Taylor, however, said that if that had been the case, Ashton would have been in trouble immediately after surgery.

So in my opinion the leaflet valve tissue was attached to that patch. The suture line was there, although part of those sutures are for the other side, the tricuspid valve. And the leaflet pulled away from the patch, apart from the suture line, dehisced over the course of the clinical events leading up to the child's death. (Evidence, page 43,260)

Odim said that he was surprised to read Taylor's conclusion about dehiscence.

Certainly when I saw this I was a little bit surprised because certainly when Ashton was alive we were looking for evidence of this on his studies, echos and cath. So I was a little surprised by that finding.

I was also surprised because I think I recall looking at the specimen and I certainly don't remember picking this up and we may have looked at it as a group, I don't remember whether this was done in the setting of the M & M round. (Evidence, pages 25,771-25,772)

Taylor said there were three potential explanations for the suture line coming apart. They can be summarized as follows:

• There could have been an infection that undermined the repair. If the wound had become infected, the suture line would have 'softened', allowing the sutures to loosen and pull away.
• Technique: the sutures needed to be placed deep enough to have a sufficient grasp on the tissue being sutured. If not, they could have easily pulled out.
• Ashton could have had an inherent problem in the mitral valve tissue that prevented it from holding the suture.

As was mentioned, Taylor disagreed with the conclusion that the child had a CMV infection that caused the necrosis. He noted that the child had been checked for infection numerous times before death and no one had noted any infection. He said that CMV grows fairly easily and therefore would have been easily detectable. Embree confirmed this. It should be noted that de Nanassy did not actually test for the presence of CMV, but reached his conclusion based on observation.

Tests conducted by Taylor on sections of the lung were negative for the presence of CMV. He did feel that Ashton had bronchitis and pneumonia (both bacterial, rather than viral, infections) and that these bacterial infections led to the necrosis. He pointed to the fact that samples of endotracheal secretions that were taken from Ashton on November 7 tested positive for bacterial infection. It is not known when the results of these laboratory tests of November 7 were available to the team caring for Ashton.

Taylor thought that the mitral valve regurgitation could have led to a backup of fluid in the lungs that would have led to an infection in the lungs. This would then have led to the pulmonary hemorrhaging that eventually led to Ashton's death.

Taylor also thought that the infection could have moved along the suture line, causing the valve to dehisce. This would also have led to regurgitation of the mitral valve and the resultant backup of fluid in the lungs, leading to pulmonary hemorrhaging. As a result, Taylor concluded that the two key factors in Ashton's death were the mitral valve regurgitation and the necrotizing pneumonia.

In his report, Taylor wrote:

The clinical history of the post-operative period indicates progressive development of mitral regurgitation. Initial echocardiogram showed no regurgitation while the last echocardiograms showed moderate to severe regurgitation. On the last day of this child's life it was clinically appreciated that he needed mitral valve replacement. My examination of the post-mortem heart specimen showed the anteromedial leaflet of the mitral valve to be separated from the suture line to which it was suspended from the atrial ventricular septal defect patch. This was also reported on the initial autopsy examination of the heart. Progressive dehiscence of this leaflet from the patch would cause the clinical findings recorded in the postoperative days. The finding of massive pulmonary haemorrhage is in keeping with fulminant mitral valve regurgitation. (Exhibit 336, page 10.1)