Most patients with congenital heart defects involving the right ventricular outflow tract (RVOT) undergo surgical repair early in life. This usually consists of reconstructing the RVOT using a patch, conduit, and/or a bioprosthetic valve. However, RVOT dysfunction is a common occurrence with resultant stenosis and/or regurgitation. In recent years, many patients have had the option to undergo transcatheter pulmonary valve replacement (TCPVR) with balloon-expandable valves instead of traditional cardiac surgery. A dangerous complication of TCPVR is compression of the coronary arteries, which can cause ischemia and infarction. Coronary compression testing with balloon dilation of the RVOT with concurrent coronary angiography or aortography is considered standard of care prior to TCPVR. If coronary compression occurs with balloon inflation then TCPVR is not performed. Non-invasive cross-sectional imaging with CT angiography (CTA) prior to TCPRV may help identify those at risk for coronary artery compression. Our objective was to measure the distance between coronary arteries and the RVOT on CTA to determine if a certain cut-off distance could reliably predict coronary artery compression.
All patients that underwent coronary artery compression testing with the intention of undergoing TCPVR at UCLA between October 2010 and May 2018 were assessed. Those that underwent CT angiography were included in the analysis. For each patient, a variety of parameters were collected, including the distance from the nearest coronary artery branch to the target landing zone for TCPVR. X-ray images and videos from the actual catheterization procedure were used to locate the exact landing zone of the valve on CTA. Then this area was examined in a multiplanar reconstruction to find the closest and therefore the most “at risk” coronary artery and measure its proximity to the landing zone.
A total of 327 patients underwent coronary artery compression testing. Of these, 50 patients underwent CT angiography and were included in this analysis. The mean age at catheterization was 36.5 ± 16.9 years, 26 patients were female, 28% had a native RVOT, 18% had a bioprosthetic valve, and 54% had a conduit. A coronary artery to landing zone distance of ≤ 3 mm was highly predictive of coronary artery compression. Out of 11 patients with a distance ≤ 3 mm, 7 (~64%) experienced coronary compression upon catheterization. Out of the 39 patients with a distance > 3 mm, only 1 (~3%) experienced coronary compression. The sensitivity of the ≤ 3 mm cut-off for predicting compression is 88%, specificity is 90%. None of the 24 patients with a coronary artery distance ≥ 8mm had evidence of coronary artery compression. Seven of the 8 patients with compression had a conduit, 1 had a native RVOT and none had a bioprosthetic valve.
An easily performed assessment of chest CTA data can predict the likelihood of coronary artery compression in patients being considered for TCPVR. A coronary artery to landing zone distance of ≤ 3 mm was highly predictive of coronary artery compression upon balloon sizing of the RVOT. Patients with conduits are at higher risk of compression than those with a native RVOT or bioprosthetic valve.
Distance between coronary arteries and the RVOT is not the only factor that contributes to coronary compression. Other factors such as anomalous coronary arteries, reimplanted coronary arteries, and size and location of other structures in the heart may also contribute to the likelihood of compression. Further analysis is needed to explore the relationship between these factors and coronary compression.