The effect of morphologic subtype on outcomes following the Sano–Norwood procedure

Abstract

OBJECTIVES

Controversy exists concerning outcomes of patients with different morphologic subtypes of hypoplastic left heart syndrome undergoing the Norwood procedure, in particular, aortic atresia-mitral stenosis (AA-MS) patients receiving a systemic–pulmonary modified Blalock-Taussig (mBT) shunt. We sought to determine the influence of known risk factors and morphology on early survival in our cohort of Sano–Norwood patients with right ventricle–pulmonary artery (RV–PA) conduits as the source of pulmonary blood flow.

METHODS

We studied early survival in patients with Sano-modified Norwood procedures from 2002 to 2010 and included both typical and atypical (including unbalanced atrioventricular canal defect) morphologic variants. We included a comparison group composed of classical Norwood patients with mBT shunts.

RESULTS

Of 264 Sano–Norwood patients, 49 had AA-MS and 30 had atypical morphologies. Weight ≤2.5 kg was associated with a poorer 1-year survival (P = 0.0007), though ascending aorta (AscAo) size <2.0 mm was not. We did not observe a difference in 30-day or 1-year Kaplan–Meier (K–M) survival across typical morphologic variants for either a Sano or mBTS group. Atypical variants exhibited a trend towards lower 30-day and 1-year survival in both Sano and mBTS groups though this was not significant. Direct comparison of 30-day K–M survival for Sano versus mBTS in AA-MS patients showed similar outcomes (P = NS).

CONCLUSIONS

Use of the RV–PA conduit results in good early survival, even in those with a small AscAo size. Atypical morphologic variants seem to do worse irrespective of the Sano or mBTS group. Further studies will be required to determine conclusively whether the RV–PA shunt confers protective effects in the AA-MS subgroup compared with the mBTS.

INTRODUCTION

The Norwood Stage I palliation has improved survival for patients with hypoplastic left heart syndrome (HLHS) since its introduction in 1983 [1]. With continued improvements in perioperative management, in particular those aimed at improving the pulmonary:systemic blood flow ratio (Qp:Qs) in favour of Qs, as well as surgical technique, many centres have reported 30-day survival following the Norwood procedure to be >75% [2–5]. The majority of deaths within the early follow-up period occur in the immediate perioperative period and up until the time of Stage II palliation at an age of 3–6 months [2]. Classically, the Norwood procedure has utilized a systemic–pulmonary artery (modified Blalock-Taussig, mBT) shunt as the source of pulmonary blood flow (PBF). Using this technique, some centres have reported poorer early survival in the aortic atresia-mitral stenosis (AA-MS) morphogic variant of HLHS [6–8]. The Sano modification of the Norwood procedure uses a right ventricle–pulmonary artery (RV–PA) shunt as the source of PBF and it has been suggested that this may benefit patients following Stage I palliation, in particular those with AA-MS due to improved coronary perfusion and more stable perioperative haemodynamics [6, 9]. The main objective of this study was therefore to examine the influence of previously described risk factors including weight and ascending aorta (AscAo) size on early survival following the Sano-modified Norwood procedure and to determine the influence of morphologic subtype, in particular the AA-MS variant. In addition, we compared the outcome by morphologic subtype in the main Sano–Norwood patient cohort with a control cohort who had undergone the classical Norwood procedure using an mBT shunt.

MATERIALS AND METHODS

Patients

We conducted a retrospective, single-centre analysis of patients undergoing the Sano-modified Norwood procedure between May 2002 and January 2010. Patients with both typical and atypical HLHS morphologic variants were included. Typical variants included patients with aortic stenosis-mitral stenosis (AS-MS), AA-MS and aortic atresia-mitral atresia (AA-MA). Atypical variants included patients with unbalanced complete atrioventricular septal defects, those with partial or total anomalous pulmonary venous drainage (P/TAPVD), patients with left or right atrial isomerism (LAI, RAI) and those with double inlet/double outlet right ventricle (DIRV/DORV). All patients had a transthoracic echocardiogram (TTE) prior to surgery to confirm diagnosis and assign each patient into one of these four categories according to morphology. All scan images were independently assessed by two cardiologists and the concordant morphologic category assigned. The distinction between valvar atresia and stenosis was made by assessment of flow using colour flow Doppler imaging on the two-dimensional images and was assessed independently by two investigators (OS; DW). Prior to Stage I, documentation of moderate to severe impairment of right ventricular (RV) function was made on TTE; prior to Stage II, RV functional assessment was made on preoperative cardiac catheter in all patients using a semi-quantitative scale: (1) good; (2) moderate; (3) poor.

Patients who had undergone classical Norwood procedures using an mBT shunt were excluded from the main study group; these patients comprised the control group. We also excluded all patients who had a left ventricle-dependent circulation. In addition, during the study period, patients who had undergone initial bilateral PA banding with maintenance of the arterial duct (prostaglandin E1 or ductal stenting) were also excluded. The study was a retrospective review that had been registered with the institutional Research and Development Board and ethical approval had been waived by the Central Office for Research Ethics Committees due to the use of retrospective, anonymized data. Follow-up data for all patients were collected through patient records and data which had been stored either in written documents or electronically (Heartsuite Database; Systeria Inc., Glasgow, UK). All patients were followed up either at the Birmingham Children's Hospital or at their local institution within the UK.

Surgical technique

All patients had been stabilized preoperatively with prostaglandin E1, either on the ward or in the paediatric intensive care unit (ICU). The conduct of the surgery involved deep hypothermic circulatory arrest with systemic cooling to a nasopharyngeal temperature of 18–22° Celsius with arterial perfusion via a 3.0 mm GORE-TEX^® shunt (WL Gore & Associates UK Ltd, Livingstone, UK) into the innominate artery as previously described [10]. Cold crystalloid cardioplegia solution was employed as a single dose for myocardial protection and aortic arch reconstruction performed using pulmonary homograft patch. Selective antegrade cerebral perfusion was used during arch reconstruction. The RV–PA connection was constructed using a GORE-TEX^® graft, which was 5.0 mm for patients ≥2.5 kg and 4.0 mm for patients <2.5 kg. The RV–PA shunt was placed wither to the right or left of the Damus connection. For the Classical Norwood patients ≥2.5 kg, a 3.5 mm GORE-TEX^® shunt was constructed between the origin of the subclavian artery and the main PA on the same side, either right or left. A 3.0 mm shunt was used in patients <2.5 kg.

Patients were weaned from cardiopulmonary bypass (CPB) on a standard inotrope regimen of 0.2–0.5 mcg/kg/min milrinone and 0.01–0.15 mcg/kg/min epinephrine. All patients underwent epicardial echocardiographic assessment of the heart prior and immediately after separation from CPB. Delayed sternal closure was routinely adopted for all patients in the study. Upon arrival in the ICU, target arterial oxygen saturations were 70–80% with mixed venous oxygen saturations of 40–50% on fractional inspired oxygen concentration of 0.5. All patients were maintained on aspirin until the time of Stage II palliation.

Statistical analysis

All data were analysed using the statistical software package ‘R’ version 2.11.1 (R Foundation, Vienna, Austria). Continuous variables are presented as means with the standard deviation or as medians with the stated range. Comparative univariable analyses were performed using the t-test or the Wilcoxon signed-rank test. Binomial or ordinal data are expressed as percentages, and comparative univariable analyses performed using the two-sided Fisher's exact test. Kaplan–Meier (K–M) analysis was used for survival with the Log-rank test used to determine significant differences. A probability value of <0.05 was taken to represent a statistically significant difference between groups. Cox Multivariate Logistic regression was further used for both cohorts for analysing prognostic factors.

RESULTS

Patient characteristics and intra-operative data

Two hundred and sixty-four Sano–Norwood patients were included in the main study (163 males; 101 females). Antenatal diagnosis had been made in 179 patients (67.8%). The median age of the patients at the time of surgery was 4.0 days (range 0–309 days), while the median weight of patients was 3.1 kg (range 1.7–5.3 kg). Of the 264 patients, the assignment to morphologic categories was as follows: n = 88 AS-MS (33.3%); n = 49 AA-MS (18.6%); n = 97 AA-MA (36.7%). Thirty patients (11.4%) had atypical morphology and these are listed in Table 1. The median size of the AscAo was 3.0 mm (range 0.5–8.0 mm), with 84 patients having an AscAo ≤2.0 mm. Thirty-seven patients had weight ≤2.5 kg. The smallest AscAo size was associated with the AA-MA morphologic subtype. Of the 264 patients, 74 (28.0%) were assessed to have at least moderate impairment of RV function prior to surgery. The mean CPB time was 116.9 ± 32.5 min; mean circulatory arrest time was 17.6 ± 14.5 min; mean cross-clamp time was 55.2 ± 17.3 min. We did not find that longer CPB, circulatory arrest or cross-clamp times were significantly associated with higher perioperative mortality. In our series, 30 patients had undergone RV–PA conduit placement to the left of the Damus connection and 234 patients to the right of the Damus anastamosis.

In the comparison group, 84 patients underwent a classical Norwood procedure using an mBT shunt. For these patients, the median weight was 3.2 kg (range 1.9–4.7 kg) and the median AscAo size was 4.0 mm (range 1.5–6.5 mm). This cohort included 10 patients (11.9%) with atypical morphologies as shown in Table 1.

Early mortality in the Sano cohort: weight and AscAo size

The overall 30-day mortality in the Sano cohort was 15.2% (40/264 patients). This compared with an overall 30-day mortality of 25.0% (21/84) in the mBT shunt control cohort. Patient weight >2.5 kg was significantly associated with higher 1-year survival compared with survival for patients ≤2.5 kg (Fig. 1A; P = 0.0007). Further, on Cox multivariate analysis, weight emerged as a significant prognostic factor in the Sano cohort (hazard ratio, HR = 1.58; confidence interval, CI: 0.43–0.93; P = 0.02). Although there was a small difference in survival for patients with AscAo ≤ versus > 2.0 mm beyond 50 days (Fig. 1B), this was not statistically significant (P = 0.355). Interestingly, however, on multivariate analysis, AscAo size did emerge as a significant prognostic factor for survival in the mBT shunt cohort, though not in the Sano cohort (for mBTS cohort: HR = 1.45; CI: 0.50–0.96; P = 0.026). Follow-up was 100% complete over a median period of 17.1 months (range 0.0–98.8 months). Of the 264 patients, 219 patients were discharged from hospital and 197 patients survived to Stage II. One patient was still awaiting second-stage palliation at the time of the last follow-up. Overall 30-day mortality was 40/264 patients (15.2%), with overall 1-year mortality of 70/264 (26.5%). The interstage mortality was defined as mortality after the first 30 days and up until the time of Stage II palliation and was 26/224 patients (11.6%). At the time of cardiac catheterization pre-Stage II, ventricular function was documented as good in 156 patients (76.9%) and poor in 8 patients (3.9%). The median age at the time of Stage II was 5.1 months (range 1.3–17.4 months).

Early mortality: influence of morphology

In the main Sano patient cohort, we initially examined the early survival to 1-year among patients with typical morphologic subtypes. The overall 1-year mortality for all subtypes in the Sano cohort was 29.6% (78/264 patients). There was no significant difference in 30-day, interstage or 1-year survival among the three morphologic groups, in particular, the AA-MS patients did not exhibit a worse outcome (P = 0.751; Fig. 2A). We further compared survival in these patients with typical morphologies as a single group with the group of patients comprising atypical morphologic variants. This showed a trend towards reduced survival in patients with atypical variant morphologies, though this was not statistically significant (P = 0.0865; Fig. 2B). The 30-day survival was 86 versus 77% for the typical versus atypical variants, respectively; 1-year survival was 72 versus 57% for these two groups, respectively.

For the mBT shunt cohort, the overall 1-year mortality was 32.1% (27/84 patients). In examining the control group of patients with mBT shunts, we found a similar trend towards reduced 1-year survival in patients with atypical variant morphology (Fig. 3A; P = 0.50), with divergence between the two K–M plots occurring within the first 30 days, as had been observed in patients who had undergone the Sano–Norwood (comparison with Fig. 2B). The 30-day survival was 77% for the typical morphologies versus 60% for the atypical morphologies. The 1-year survival was 69 versus 60% for these two groups, respectively.

In the mBT shunts patients, when analysing the typical variants by individual morphologic group, a lower survival in the AA-MS patients was seen compared with the AA-MA/AS-MS patients within the first 30 days, with the AA-MS curve diverging and following that for the atypical morphologic variants, though this was not statistically significant (Fig. 3B; P = 0.899). By 100 days, however, survival for the AA-MS patients was not dissimilar to that for AA-MA/AS-MS patients, though the ‘other’ variants continued to show a poorer survival (Fig. 3B). We further performed a direct comparison of 30-day survival for patients with the AA-MS morphology who had undergone the Sano–Norwood compared with AA-MS patients who had received an mBT shunt. Although the K–M survival was lower beyond 5 days in the mBT shunt cohort compared with that in the Sano cohort, the numbers of patients in both groups were small and the difference was not statistically significant (Fig. 4; P = 0.204). There were no significant differences in the median weights or AscAo sizes for either the Sano or mBT shunt cohorts when comparing typical with atypical morphologic groups (Table 1). Pre-Stage II RV function was similar between the overall group of patients with RV–PA conduits versus mBT shunts (Table 1; P = 0.809).

DISCUSSION

We have reported early outcomes for 264 patients undergoing the Sano-modified Norwood procedure, with an overall 30-day and 1-year mortality of 15.2 and 29.5%, respectively. Of the patient-related factors we examined, only weight ≤2.5 kg was associated with a poorer outcome. Recent large series, in accordance with our data, have identified weight ≤2.5 kg as an important predictor of survival [4, 11], though this finding has not been universal [3]. The interstage mortality of 11.6% for the Sano cohort is similar to that reported in other series, and is important in that it has been shown that this is potentially greatly modifiable by way of home surveillance programmes [3]. The main findings of our study were: firstly, the poorer survival in patients with atypical variant morphologies, with only a 57% 1-year survival in this group; secondly, the lack of difference in outcome for the AA-MS subtype in the Sano–Norwood cohort. Finally, in examining outcomes in a comparison group of 84 patients receiving mBT shunts, the trend towards lower survival in atypical variant morphologies persisted, though notably, the AA-MS patients in this cohort did worse within the first 30 days compared with other typical morphologies. A further direct comparison between 30-day survival for the AA-MS patients receiving either the RV–PA conduit or the mBT shunt suggested higher mortality in the latter group, though was also not statistically significant.

Survival in atypical morphologic variants: Sano cohort

HLHS represents a spectrum of disorders characterized by underdevelopment of left-sided heart structures with various morphologies as described. The distribution of patients across typical morphologic groups in our series is comparable with those reported by others [6]. It is of note, however, that some investigators have excluded patients with atypical morphology [5]. Importantly, the group with the highest mortality in our series was that with atypical variant morphologies. While the report from Milwaukee excluded these variants, Sano et al. found that these patients had a higher perioperative mortality compared with typical HLHS morphologies in accordance with our own findings [4, 6]. This latter report used exclusively RV–PA conduits and further found that there was no difference in outcomes for patients with different typical HLHS morphologies.

Survival in the AA-MS morphologic subtype: Sano cohort

Several studies have reported that the AA-MS morphologic subtype is a risk factor for the Norwood procedure [6–8]. Recent data has also suggested that aortic atresia in itself in patients with HLHS following the Norwood procedure is associated with decreased systemic, splanchnic and cerebral perfusion [12]. Most recent published series of Norwood procedures, however, have included patients with predominantly classical Norwood procedures using the mBT shunt [3, 6, 11]. In contrast, we have not found AA-MS patients to have a worse outcome in our patients receiving the RV–PA conduit. As hypothesized by Tweddell and colleagues [6], this may be related to reduced diastolic run-off using the RV–PA conduit rather than a classical systemic–pulmonary artery shunt. Pathological studies reported by several groups have found various coronary abnormalities in the AA-MS subtype of patients including thicker-walled coronary arteries with increased tortuosity and the presence of ventricular–epicardial coronary artery fistulae [13, 14]. Further, hearts with AA-MS morphology may have greater degrees of endocardial fibro-elastosis [8]. One report has described a 53% incidence of fistulae in AA-MS patients, with the presence of fistulae in itself being a predictor of poorer early survival [8]. It has therefore been proposed that this HLHS subtype might be more vulnerable to the effects of systemic diastolic flow run-off due to a fragile coronary circulation, particularly if undergoing a Norwood procedure using an mBT shunt [6]. Glatz et al. [6] reported worse outcomes for AA-MS patients in their series of 72 patients, all of whom had received mBT shunts. Similarly, Furck et al. [3] found AA-MS to be an important risk factor in their series of 157 patients, with only 2 patients receiving RV–PA conduits. Vida et al. [8] included a greater proportion of patients with RV–PA conduits in their series (almost 25%), and found a 29% hospital mortality/need for transplantation in the AA-MS subgroup versus 7.9% in other subgroups. It is a strong possibility, therefore, that the differences we and Sano et al. have observed may be at least in part attributable to improved coronary and systemic perfusion in patients receiving RV–PA conduits. To explore this issue further, we compared outcomes by morphologic subtype in our main patient cohort of RV–PA conduits with a smaller group who had received mBT shunts.

Survival by morphologic subtype: mBT shunt versus RV–PA conduit

We found a persistent trend towards lower 1-year survival in patients with atypical variants of HLHS in the comparison cohort of patients who had received mBT shunts. As with the RV–PA conduit cohort, the K–M survival plots for typical versus atypical variants diverged within the first 30 days, after which the difference persisted. Stasik et al. included 100 patients in their study with 94% receiving mBT shunts and included 20 atypical variants who had a tendency to poorer survival (significant on univariable analysis; [11]). The report by Glatz et al. [6] who had used only mBT shunts excluded atypical morphologic variants. Interestingly, we found that among the typical morphologic variants who had received mBT shunts, patients with AA-MS had lower 30-day survival compared with AA-MA and AS-MS patients. Beyond 100 days, however, this difference became less apparent. These findings are therefore consistent with data reported by Glatz et al. and Vida et al. [6, 8]. Further, direct comparison for AA-MS patients between the RV–PA conduit and the mBT shunt suggested poorer survival in the latter, with curves divergent around 5 days, though this was not significant statistically. As with the main Sano cohort of patients, it should be noted that there were no significant differences in weight or AscAo size for the typical versus atypical morphologic groups which might account for the trends in survival that we observed.

Only one recent important study has reported a comparison between patients receiving the mBT shunt and those receiving the RV–PA conduit. This multi-centre trial randomized 555 patients to either the RV–PA conduit or the mBT shunt at the time of Norwood procedure [15]. The main finding was a better early transplant-free survival up to 12 months in patients receiving the RV–PA conduit, though no detailed survival data by morphologic subgroup has yet been reported. In our series, we have found that the overall 30-day and 1-year survival were lower in patients undergoing mBT shunts compared with the RV–PA conduit irrespective of morphology, though this was not statistically significant. Of note, a recent, large, single-centre study of patients undergoing the Norwood procedure using the RV–PA shunt similarly found no difference in mid-term survival among typical or atypical HLHS anatomical variants [16].

mBT shunt versus RV–PA conduit: other issues

It is of interest that AscAo size emerged as a significant prognostic factor in the mBT shunt cohort on multivariate analysis, though not in the Sano cohort. This may be partially attributable to the lesser diastolic run-off and improved coronary perfusion with RV–PA shunts, though this would need to be substantiated by further studies. Although we have observed good outcomes in our group of patients, the use of RV–PA conduits is not without problems. One is the potential direct backflow and volume loading on the RV. Another is the long-term effect of the right ventriculotomy in itself. With respect to the first point, this has led others to consider the use of valved homograft conduits to provide a limitation or prevent backflow into the RV [17]. With respect to the second point, we did not quantitatively measure RV function long term in the present group of patients, though we have recently reported in a smaller series of patients with the Sano modification that RV function was well preserved in the majority of patients [2]. The results from the recent multi-centre trial comparing the RV–PA shunt versus the mBT shunt for the Norwood procedure similarly did not find significant differences in RV function at 12 months between the two groups [15].

Study limitations and conclusions

The present study was retrospective in its design. We did not make a detailed assessment of the causes of interstage mortality, and analysis of RV function prior to Stage II palliation was only made semi-quantitatively. Further, in comparing the RV–PA patient cohort with the control group who had received an mBT shunt, the inferences that may be made are limited by the historical nature of the control group as well as differences in the distribution across morphologic subtypes. Our findings confirm the importance of patient-related factors such as weight as a risk factor for mortality following the Norwood I procedure, and lend some support for the survival benefits of the RV–PA shunt in Norwood patients in those with both atypical morphologic types as well as in the AA-MS subgroup, though these findings will need to be corroborated by further data.

ACKNOWLEDGEMENTS

We thank the Cardiology Department (J. De Giovanni, P. Miller, A. Chickermane, T. Desai) and the ICU department for the care of all the patients included in this study. We also thank Joe Eurell for assistance in data collection.

Conflict of interest: none declared.

REFERENCES