Predicting surgical outcome by echocardiography in complete atrioventricular canal defect repair

Anatomic repair of complete atrioventricular canal defects (CAVC) is nowadays a routine procedure, which, despite its relative frequency, remains surgically challenging. Freedom from re-operation/re-intervention is reported to be 85% after 10−20 years of follow-up with the most common cause of re-operation/re-intervention being left atrioventricular valve regurgitation (LAVVR).1, 2 Therefore, the non-invasive assessment of surgical outcome by means of echocardiography and the establishment of validated follow-up protocols, which specifically address (I) the timing of the examination, (II) the mode of an investigation by either transthoracic or transesophageal echocardiography (TEE/TTE), and (III) the relevant echocardiographic variables that should be assessed, is of pivotal importance in CAVC repair patients. However, despite the need for such protocols, data to support validated approaches are still sparse. Recently, Gellis et al. have shown that after primary CAVC surgery re-intervention rates remained high for left atrioventricular valve (LAVV) repairs, particularly for patients with LAVV stenosis gradients ≥ 5 mmHg and mild or greater LAVV regurgitation on postoperative TEE.2 In their study, an age at LAVV repair of less than 72 months, partial AVC anatomy, left ventricular dysfunction, mean LAVV stenosis gradient ≥ 5 mmHg, and multiple jets of regurgitation on postoperative LAVV repair TEE were associated with LAVV re-intervention. However, the grading of LAVVR on postoperative TEE was not found to be an independent risk factor.2 In a very thoughtful accompanying commentary article, Dr. Bacha states that patients with residual or recurrent LAVV lesions after CACV repair represent a surgically challenging group of patients. Specifically, these patients exhibit a rate of re-intervention of one-third at a fairly short median follow-up time of 18 months. Thus, he concludes, that we should all aim for as little regurgitation on the LAVV as possible, and that any more than mild regurgitation should be regarded as a risk factor.3 However, from the aforementioned study it remains unclear whether this holds also true for primary CAVC repair and not only for those patients who underwent LAVV re-intervention after CAVC repair. Furthermore, the optimal timing of echocardiographic evaluation remains unclear (immediate postoperative TEE vs. follow-up TTE). In this issue of Echocardiography, Freeman et al.4 addressed these issues by investigating pre-operative and post-operative LAVVR at different time points. The authors aimed to identify echocardiographic characteristics that predicted post-operative LAVVR at discharge and at 1-year follow-up. Echocardiograms were performed pre-operatively and post-operatively (TTE/TEE), on post-operative day 1 (TTE), at discharge (TTE), and after 1-year (TTE). In their analyses, the authors included a total of 52 CAVC patients (92% with known trisomy 21), out of which 71% were classified as Rastelli type A and 29% as Rastelli type C (29%) defects. Only two patients had moderate or greater LAVVR pre-operatively. Freeman et al.4 found that out of the patients with less than moderate LAVVR on post-operative TEE, 20% showed a worsening to more than moderate LAVVR at discharge, but, importantly, only 9% remained that way after 1 year. Of those with more than moderate LAVVR on post-operative TEE, 40% improved to less than moderate LAVVR after 1 year follow-up. Two patients worsened at 1 year, which was likely secondary to cleft suture dehiscence. These data indicate that intraoperative/post-operative TEE assessment of LAVVR does not correlate well with the situation after 1 year. In fact, Freeman et al.4 highlight that pre-operative LAVVR was found to be the only significant predictor of LAVVR severity at 1-year follow-up. Given that CAVC is a rare condition, this retrospective study on 52 pediatric patients can be considered "large", especially in comparison to other published pediatric studies. A major strength of this study is its design, which simplifies a very complex congenital anatomy and its peri-operative and post-operative assessment in a clear and easy understandable manner. More importantly, the results by Freeman et al. have several practical implications: First, as stated above, LAVVR is likely to change over time and, thus, albeit the need for close monitoring and medical treatment (if indicated), decisions to re-intervene in the immediate post-operative period should be made with caution (Freeman et al. saw the highest percentage of more than moderate LAVVR at discharge!).4 Second, LAVVR assessment on post-operative day 1 (POD) was a poor predictor of later AVV function, which led the authors to question its overall value and to suggest deciding on the necessity of POD 1 echocardiography on a case to case basis.4 While we agree with the authors that the value of POD 1 echocardiography may be limited in regard to its ability to predict future LAVVR, we believe that several other important information can be gained from POD 1 echocardiography in the setting of CAVC repair (i.e., development of ventricular function and pulmonary hypertension, pericardial effusion, etc.). Third, in their study, the only significant predictor of LAVVR severity at 1 year was the pre-operative severity of LAVVR.4 Considering the relatively homogenous cohort studied by Freeman et al., which excluded partial atrioventricular canal defects as well as those with other congenital cardiac defects, future investigations on these even more complex patients are required. In addition, it would be interesting to compare the findings in patients with versus those without trisomy 21, which was not possible in the current study due to the high number of trisomy 21 patients.5 Furthermore, the non-invasive prognostication of AVV function by simple echocardiographic measures is especially relevant in unbalanced AV canal patients undergoing univentricular palliation, since AVV competency is of pivotal importance for a well-functioning Fontan circulation.6 In addition, several echocardiographic scores, which incorporate AVV parameters, have been developed to predict successful biventricular versus univentricular repair in unbalanced CAVC.7 Another issue raised by Freeman et al.4 regards the role of intraoperative TEE. While TEE results are certainly important for acute intraoperative decisions (i.e., to go off-pump), their role in the prediction of long-term surgical results or outcome is limited. In fact, one previous study showed considerable discrepancies between the examinations of residual AVVR and potential AVV stenosis: most of the residual regurgitations were underestimated by intraoperative TEE.8 Importantly, validated intraoperative TEE guidelines for children with congenital heart disease and specifically for those with CAVC have still to be developed,8-10 in contrast to clear guidelines for the use of TEE in the operating room for the adult population with acquired heart disease.11 One has to be aware, that intraoperative TEE and post-operative TTE in an ICU setting acquire data on very artificial, non-physiologic conditions. Specifically, hemodynamics may be altered by factors such as mechanical ventilation, open chest, volume status, and inotropic support, and so forth. While these obstacles cannot be overcome by a study such as the one provided by Freeman et al., the authors report essential information on our understanding of specific aspects of the peri-operative echocardiographic evaluation of CAVC patients. We believe that the results by Freeman et al. provide a valuable framework for further prospective studies on the role of peri-operative echocardiography in CAVC. Drs. Koestenberger and Sallmon indicate no conflicts of interest related to the content of this article.