Let the intramural segment choose: unroofing, reimplantation, or a neo-ostium

In intramural AAORCA a length of the artery does not run free on the epicardium but sits within the aortic wall, sharing a layer of it with the aorta. Unroofing divides that shared wall along the segment so the intramural portion opens directly into the aortic lumen.

That accomplishes two things at once: it relieves the lateral compression of the intramural segment, and it converts a slit-like ostium into a larger new one. For a sufficiently long intramural segment, those two correspond precisely to the mechanisms currently thought to matter most.

What it does not do is relocate the artery to a normal right sinus. Comparative postoperative CTA shows this directly: after unroofing an acute take-off persisted in 100% and the interarterial course fell only from 43/45 to 35/45. A course still running between the great arteries on postoperative imaging is therefore not a failed operation; that course is simply not what this technique changes.

Its provenance and limits come first. It is not a threshold set by any guideline but a direction read off two datasets: postoperative CTA showed that a preoperative intramural length under 5 mm was associated with a residual thickened intercoronary pillar after unroofing in right-sided lesions; and in a direct comparison of 61 patients, intramural length was in fact similar between those reimplanted and those unroofed (median 5 mm versus 6 mm), showing that length is one input rather than a switch.

The same 61-patient study states the indication more concretely: transection and reimplantation is a useful alternative when the course lies below the commissure, when unroofing does not relocate the ostium to the appropriate sinus, or when unroofing would leave compression by the intercoronary pillar. None of the three is “length below a number”; all three are “unroofing cannot achieve a result.”

A 17-child surgical series from China raises the same concern: with a short intramural segment, unroofing may not effectively relocate the coronary ostium. The series is small with limited follow-up, but it shows the concern is not one centre's idiosyncrasy.

The aortic leaflets meet and are supported at the commissures, and an intramural coronary often passes near exactly that region. If the segment runs at or below a commissure, opening it completely may require taking the commissure down first — which turns a coronary operation into one that also involves the aortic valve.

Resuspending the commissure afterwards is an established answer. In 26 consecutive unroofing operations, the first 9 had unroofing alone and the subsequent 17 routinely had commissural resuspension; on follow-up no patient in the resuspension group had aortic regurgitation, while 6 of 9 (67%) in the earlier group had mild-or-greater regurgitation. The follow-up durations differed, and this is a before-and-after comparison within one centre rather than a randomised one.

The other answer is to avoid the commissure altogether. In an early series of 9 patients, every intramural segment lay at or below the commissure, and 2 had a new ostium created at an appropriate site without extensive unroofing, specifically to avoid disturbing the intercoronary commissure; the authors suggested this might spare the aortic valve. The same series reports one patient requiring a Ross procedure for severe aortic insufficiency 44 months later, which shows the concern is real.

Multicentre figures give the background magnitude: among the 395 CHSS operations, of the 358 with paired evaluations about 8% (27/358) developed new mild-or-greater aortic insufficiency and about 2% (7/358) moderate-or-greater. The probability is low, and it is not zero.

In principle, reimplantation can eliminate the anomalous sinus of origin, the slit-like ostium, the intramural course, the acute take-off, and the intercoronary pillar at once. Comparative postoperative CTA supports this: after reimplantation those features were essentially gone, with an acute take-off persisting in only 2 of 17 (12%).

The cost is an anastomosis that did not previously exist and must remain unnarrowed, unkinked, untwisted, tension-free, and adequate as the child grows. In the same study, 2 of 17 (12%) in the reimplantation group developed severe coronary stenosis requiring urgent revision, a category of problem the unroofing group did not have.

An adult series points to the same place. Of 16 AAORCA reimplantations, 15 (94%) succeeded with no operative mortality; one was judged unsatisfactory intraoperatively with difficulty weaning from bypass and was converted to a saphenous vein graft. Those completed were symptom-free with no stenosis, kinking, or pulmonary artery compression on follow-up imaging. The principal risk of reimplantation declares itself in the operating room rather than slowly afterwards.

In one single-centre series of 230 AAOCA operations (median age at surgery 17 years), 86 had unroofing, 123 coronary reimplantation, 13 repair of an intraconal left main, and 8 other procedures. Over a median follow-up of 4 years there were no early or late deaths; 6 patients (2.6%) were reoperated for recurrent symptoms or ischemia, 3 for anatomic coronary narrowing and 3 for a previously undiagnosed myocardial bridge. There were no reoperations in the reimplantation group and 5 (5.8%) in the unroofing group. On that basis the centre states a current preference for reimplantation.

At another centre, 148 protocolized unroofing operations (median age 44.4 years, 130 right-sided) showed 10- and 15-year survival of 94.5% over a median follow-up of 9.5 years, with no early or late deaths in the pediatric subgroup at a median follow-up of 10.9 years. It is one of the longest-followed unroofing experiences available.

They are not directly comparable: age structure, anatomic mix, technique allocation, and follow-up length all differ, and in both centres the technique was chosen by the surgeon according to anatomy. What holds is that both techniques achieve very low mortality at experienced centres. What does not hold is that either is superior in the long run.

One further limit applies to all of it: follow-up after AAOCA repair is currently measured in years, while a school-age child's horizon is decades. No existing dataset answers what these repairs look like thirty or forty years on.