What range of extra-cardiac conduit flow velocity is detectable intraoperatively following the completion of a total cavo-pulmonary connection?

Eight patients who underwent TCPC completion from October 2012 to October 2013 were enrolled. This retrospective study was approved by the institutional ethics committee with a waiver of the requirement to obtain written informed consent from the parents or the guardians of the patients. Medical charts, anesthetic records, surgical records, and TEE records were retrospectively reviewed. Demographic data and ECC flow profiles observed via intraoperative TEE were collected for every patient.

The ECC flow profile was obtained at the site just proximal to the IVC-ECC anastomosis. We set the sample volume at a site near the center between the ECC walls, where the color signal was most intense. The m-Vs were measured during inspiration and expiration and averaged based on an inspiration-to-expiration ratio of 1:2.

The difference between the measured m-V and predicted m-V was analyzed using a paired t test. Correlations of the measured m-V with the predicted m-V, BSA, and age were analyzed by Spearman’s rank correlation test. A p value of less than 0.05 was considered statistically significant.

To our knowledge, the present study is the first to elucidate an acceptable ECC flow velocity immediately after TCPC completion. The measured m-V was significantly faster than the predicted value in our study, both collectively and in every patient individually.

Two findings have the potential to invalidate the third assumption described above in calculating the predicted values and could account for the faster velocities measured in our cases: first, the discrepancy in size between the IVC and graft; second, the compression of the ECC by the patient’s heart. The former finding must have made the graft shrink at the site of the anastomosis, and the latter must have squeezed the cross-section of the ECC into a narrower form than the pure circle. Both findings are closely related to the size mismatch between the small physique of a child and the relatively large-bore graft selection to accommodate the child’s growth into an adult.

Mosbahi et al. studied the spatial distribution of the flow velocity from the right ventricular outflow tract (RVOT) to both of the branch pulmonary arteries (PAs) at three different flow rates of constant flow (low [75 %], normal [100 %], and high [125 %] versus normal baseline measurements taken in vivo) using a computational fluid dynamics model [3]. The constant flow in this model mimics Fontan circulation well in the absence of a pumping chamber driving the pulsatile pressure for pulmonary circulation. The distribution of the flow velocity in their study turned out to be quite homogenous in a straight trunk running from the RVOT to the main PA. Most of the points on the cross-sectional plane in the straight trunk had the same velocities, but a number of very thin circumferential points along the vessel wall had notably low velocities. This is quite consistent with the flat profile flow of the ECC we assumed. Mosbahi’s group also observed higher flow velocities around the corner and after the change in vessel diameters from the main PA to branch PAs and much higher velocities reaching about 2-fold the velocity in the main PA at the “high” flow rate. This condition could be easily reproduced in the presence of ECC compression and a hyperdynamic state concomitantly in our cases. Shrinkage was also unavoidable, because a 16- or 18-mm graft was selected for the ECCs for the 2- to 5-year-old children. The ECC flow velocity could have thus been closer to the IVC flow velocity and necessarily faster than the predicted velocities at the site of the anastomosis with the IVC. Given the IVC diameter range of 5.9 to 14.3 mm (median of 9.7 mm) in our patients, the IVC flow velocities could have been roughly estimated to be 3.2-fold the predicted velocities. The calculated IVC flow velocity was highest (68.8 cm/s) in our smallest patient, and a corresponding ECC velocity (56.7 cm/s) was actually observed. The ECC must also have been compressed, at least to some extent. With only negligible compression, the ECC flow velocity slows to a level close to the theoretical value due to the relative release of the shrinking. While this compression can be significant, it further enhances the elevated ECC flow velocity. These findings are likely to account for the detection of the fastest m-V in the smallest patient, the greater velocity in the present study versus the previous study, and the 2-3-fold higher m-V versus the predicted value in the remaining cases.

We extrapolate the CO in healthy subjects to our TCPC patients at the time of ECC flow measurement, that is, at about 1–2 h after separation from CPB. The CO in patients with Fontan circulation typically decreases to about 70 % of that in normal subjects at rest [4]. In other observations after the Fontan operation, Nakazawa and colleagues measured a CO of 2.45 ± 0.48 L/min/m² in 10 patients with tricuspid atresia (TA) and a CO of 2.75 ± 0.72 L/min/m² in another 10 patients with single ventricle physiology excluding TA [5]. Williams et al., meanwhile, found that dopamine infusion at a rate of 7.5 μg/kg/min just after the classical Fontan operation increased the cardiac index by 40 % (from 1.98 ± 0.86 to 2.75 ± 1.05 L/min/m²) in 9 patients [6]. Yet in both of the foregoing studies, the CO values were roughly estimated at only 50 % of the value in the healthy children investigated by Salim et al. (5.2 ± 1.4 L/min/m²). The hyperdynamic states in our patients must have counterbalanced the low CO typical of Fontan cases to an extent similar to that observed in the report by Williams et al. but not to an extent sufficient to permit the extrapolation of the CO of healthy children to our cases. By assuming a CO lower than the CO we actually applied in predicting the ECC m-V in the present study, the predicted m-V must be smaller. This may imply that an ECC flow of a low velocity, even a velocity lower than that previously reported, is very likely to be obtained when the shrinkage and compression of the ECC is negligible.

Fenestration was created in 3 patients in a side-to-side anastomosis fashion between the ECC and atrial wall at the mid-portion of the conduit. We measured the ECC flow at a site distal from the fenestration, so the fenestration had no effect whatsoever on the measured flow.

Our study is subject to several limitations. First, the study was retrospective. Ideally, the site, point in time, and respiratory setting for ECC flow measurement should be strictly determined. Second, we worked with only a small number of cases with widely variable backgrounds in age, diagnosis, and PA morphology. The preferred policy for intervention in recent years has been to complete the TCPC at an earlier age in life, ideally in patients younger than 3 years of age [7, 8]. A prospective investigation of a larger number of subjects with a peak age in this younger age range is therefore awaited.

In conclusion, the range of ECC flow velocities widely varied in our series. Whereas a low velocity of around 15–20 cm/s was detected in most of our cases, a m-V ranging up to 60 cm/s was observed immediately after the completion of the TCPC even in the absence of significant obstruction of the ECC pathway. The observed range of m-V at the ECC, about 15–60 cm/s, may be acceptable for the establishment of TCPC circulation.

The present study was approved by the ethics committee of Tokyo Women’s Medical University with a waiver of the requirement to obtain written informed consent from the parents or the guardians of the patients (Reference No. 3283).