Systemic to pulmonary collateral blood flow

Downloaded from heart.bmj.com on October 5, 2014 - Published by group.bmj.com

Congenital heart disease

ORIGINAL ARTICLE

Systemic to pulmonary collateral blood flow influences early outcomes following the total cavopulmonary connection Tobias Odenwald,1 Michael A Quail,1 Alessandro Giardini,2 Sachin Khambadkone,2 Marina Hughes,1 Oliver Tann,1 Tain-Yen Hsia,2 Vivek Muthurangu,1 Andrew M Taylor1 1

Centre for Cardiovascular Imaging, UCL Institute of Cardiovascular Science & Great Ormond Street Hospital for Children, London, UK 2 Cardiorespiratory Unit, Great Ormond Street Hospital for Children, London, UK Correspondence to Dr Michael A Quail, Cardiorespiratory Unit, Great Ormond Street Hospital for Children, Level 6 Old Nurses Home, Great Ormond Street, London WC1N 3JH, UK; [email protected] T Odenwald and MA Quail contributed equally to this paper. Accepted 14 February 2012

934

ABSTRACT Background Systemic to pulmonary collaterals (SPCs) represent an additional and unpredictable source of pulmonary blood flow in patients with single ventricle physiology following bidirectional superior cavopulmonary connection (BCPC). Understanding their influence on patient outcomes has been hampered by uncertainty about the optimal method of quantifying SPC flow. Objective To quantify SPC flow by cardiac magnetic resonance (CMR) prior to total cavopulmonary connection (TCPC) in order to identify preoperative risk factors and determine influence on postoperative outcomes. Design Single centre prospective cohort study. Setting Tertiary referral centre. Patients 65 patients with single ventricle physiology undergoing CMR for preoperative assessment of TCPC completion underwent quantification of SPC flow. Clinical outcomes of 41 patients in whom TCPC was completed were obtained. Main outcome measures Early post-TCPC clinical outcomes associated with SPC flow were assessed, including postoperative chest drainage volume, postoperative chest drainage duration and length of intensive care and hospital stays. Additionally preoperative covariates associated with SPC flow were assessed including age at BCPC and CMR, SpO2 at BCPC and CMR, ventricle type, pulmonary artery (PA) crosssectional area and PA pulsatility. Different methods of CMR SPC flow quantification were compared. Results Higher SPC flow was associated with increased postoperative chest drain volume (r¼0.51, p¼0.001), chest drain duration (r¼0.43, p¼0.005), and intensive care unit (r¼0.32, p¼0.04) and log-transformed hospital stays (r¼0.31, p¼0.048). The effect of SPC flow on outcome was independent of fenestration, ventricle type and function. Preoperative covariates associated with SPC flow included age at BCPC (b¼"0.34, p¼0.008), SpO2 at time of CMR (b¼0.34, p¼0.004) and branch PA cross-sectional area (b¼"0.26, p¼0.036), model R2¼0.34. Moreover, patients with pulsatile pulmonary blood flow had lower SPC flow than those without (0.8 vs 1.3 l/min/m2 p¼0.012). SPC flow calculated by the difference between pulmonary venous return and pulmonary artery flow (l/min/m2) showed greatest association with preoperative covariates and strongest correlation with postoperative outcomes compared with other methods of quantification. Conclusions CMR can provide an effective measurement of SPC flow prior to TCPC. Young age at BCPC, high preoperative oxygen saturation and smaller PAs are associated with increased SPC flow, which may

promote increased postoperative pleural drainage and lengthen recovery.

INTRODUCTION Patients with single ventricle physiology are prone to develop systemic to pulmonary collaterals (SPCs). SPCs can contribute a significant amount of flow to the pulmonary circulation. In the sequence of single ventricle palliation, they are more prominent after bidirectional cavopulmonary connection (BCPC) than after Fontan completion (total cavopulmonary connection (TCPC))1e3 and increase with time following BCPC.2 4 5 The underlying pathophysiological mechanisms for SPC formation are not fully understood. A range of factors that might promote the formation of SPCs have been proposed, including hypoxaemia, decreased volume and velocity of pulmonary blood flow, absent pulsatility, high transpulmonary gradients and lack of hepatic venous effluent.3e5 Interestingly, if patients proceed to TCPC, the SPCs seem to regress.1 The clinical consequences of SPC flow are uncertain because their development is unpredictable and they remain difficult to assess. The additional pulmonary blood flow provided can relieve systemic hypoxaemia prior to TCPC.5 6 However, SPC flow may result in important volume overloading of the single ventricle, compete with and limit blood flow via the pulmonary arteries (PAs), and increase pulmonary arterial pressure and vascular resistance. As a result, there is no consensus as to the appropriate management of SPCs. Some centres pursue a strategy of aggressive interventional treatment of SPCs,7 8 while others have recommended that routine coil embolisation is not warranted.9 10 Moreover, the majority of SPCs do not consist of well-defined vessels but a network of small, microscopic connections between the systemic and pulmonary arteries. Angiographic assessment of SPCs is difficult and there is no validated grading system.1 9 However, SPC flow can now be quantified non-invasively using cardiac magnetic resonance (CMR) phasecontrast imaging to measure great vessel flow.2 3 SPC flow can be calculated either as the difference between aortic flow volume and systemic venous return (SVC+IVC flow) or as the difference between pulmonary venous return and PA flow. Heart 2012;98:934e940. doi:10.1136/heartjnl-2011-301599


Congenital heart disease Accordingly, we studied the clinical and haemodynamic profiles of single ventricle patients during the period following BCPC but prior to TCPC to determine the association of SPC flow volume with early clinical outcomes following TCPC completion. In addition, we aimed to identify preoperative variables that are associated with SPC development. A comparison of the different methods of SPC flow quantification was performed to characterise the correlation with early postoperative outcomes.

METHODS Patient population Between January 2009 and June 2011, 66 consecutive patients underwent CMR as part of their routine clinical management for possible TCPC completion (all patients had undergone BCPC). One patient was excluded because pulmonary venous flow could not be measured. Consent for use of anonymised MR and clinical data for research purposes was obtained from all patients before the investigation. The study received institutional research approval. CMR reports were reviewed and DICOM data were re-analysed where necessary. Forty-one patients later proceeded to TCPC, 24 had not had their TCPC yet or were not suitable for Fontan completion (figure 1). The following clinical data was retrieved from hospital records: patient demographics, diagnosis, date and type of initial surgical or catheter interventions, oxygen saturations, haemoglobin and haematocrit before BCPC, date of BCPC, saturations, haemoglobin and haematocrit shortly after BCPC, and saturations 1 year after BCPC and at the time of CMR. Of the 41 patients who proceeded to TCPC, the following data were collected: date and type of TCPC, intraoperative transpulmonary gradient, operation time, cardiopulmonary bypass (CPB) time, arterial flow during CPB, red blood cells given during TCPC and within 24 h, duration and volume of chest drainage, days in the intensive care unit (ICU), days in hospital, complications and saturations at discharge.

CMR technique MR was performed using a 1.5 T MR scanner (Avanto; Siemens Medical Systems, Erlangen, Germany).

MR anatomy 3D MR images covering the entire heart were obtained in a sagittal orientation by using a magnetisation-prepared 3D balanced, steady-state free precession sequence with navigator respiratory gating (repetition time (TR) 4.6 ms, echo time (TE) 2.3 ms, flip angle 908, number of lines per segment acquired per cardiac cycle 30e40, sensitivity-encoding factor 2.0, bandwidth per pixel 590 Hz, field of view 28032803120 mm, acquisition matrix 2563152380, reconstruction matrix 5123304396 and reconstructed iso-volumetric voxel size 1.531.531.5 mm).

MR angiography Gadolinium-enhanced MR angiography was performed as previously described11 with a coronal 3D fast-field-echo sequence. Gadolinium (Dotarem, Guerbet) was injected into a peripheral vein and tracked into the heart with a dynamic coronal 2D fast-field-echo sequence. The gadolinium dose was 0.2 mmol/kg. The MR angiographic sequence was started when contrast reached the ventricle. Two consecutive angiograms were acquired in a single 15e20-second period of apnoea.

MR ventricular function Retrospectively gated steady-state free precession cine CMR of the heart was acquired in the vertical long-axis, four-chamber and the short-axis covering the entirety of both ventricles (9e12 slices). Image parameters were TR 2.4 ms, TE 1.2 ms, flip angle 688, slice thickness 6e8 mm, matrix 2003240, field of view 300e380 mm and temporal resolution 25 phases acquired during a single breath-hold. Assessment of single ventricle volumes was performed by manual segmentation of short-axis cine images at end diastole and end systole (OsiriX; OsiriX Foundation,

Figure 1 Flow diagram of included study patients. BCPC, bidirectional cavopulmonary connection; PA, pulmonary artery; TCPC, total cavopulmonary connection.

Heart 2012;98:934e940. doi:10.1136/heartjnl-2011-301599

935


Congenital heart disease Switzerland). End diastolic and end systolic volumes were calculated by use of Simpson’s rule, and from these volumes, stroke volume (SV) and ejection fraction (EF) were calculated. Cardiac output was obtained by multiplying SV by heart rate.

MR flow quantification Superior vena cava (SVC), inferior vena cava (IVC), pulmonary trunk (if present), branch PAs, pulmonary veins and ascending aorta data were acquired using a flow-sensitive gradient-echo sequence (TR 29.9 ms, TE 2.18 ms, flip angle 308, slice thickness 5 mm and matrix 1923256) during free breathing. Throughplane flow data (30 phases per cardiac cycle) were acquired by use of retrospective cardiac gating. Arterial and venous blood flow was calculated from phase contrast images by use of a semiautomatic vessel edge-detection algorithm (OsiriX; OsiriX Foundation, Switzerland) with operator correction. PA pulsatility was determined by the presence of phasic forward flow in the main PA. All volume and flow measurements were indexed for body surface area.

SPC flow calculation SPC flow was quantified by four different methods. Method A calculated SPC flow from the pulmonary circulation as: SPC volume ðpulmonaryÞ ¼ Pulmonary venous ðPVÞ flow " PA flow: Method B used the systemic circulation, where SPC volume ðaorticÞ ¼ Aortic flow " systemic venous flow ðSVC þ IVCÞ:

In addition, SPC flow obtained from Methods A and B were separately expressed as a fraction/percentage of pulmonary venous flow, Method C : SPC ðpulmonaryÞ percentage ¼ SPC ðpulmonaryÞ=PV flow: Method D : SPC ðaorticÞ percentage ¼ SPC ðaorticÞ=PV flow: . Statistical analysis

Statistical analysis was performed using SPSS V.19.0.0.1 for Mac. All parameters are expressed as mean695% CI (1.963SE of mean) or median and IQR. Proportions are expressed as percentages. The four methods of SPC measurement were compared by BlandAltman analysis, with the limits of agreement expressed as mean difference 61.963 the SD of the difference. Multiple linear regression analysis was used to determine covariates associated with clinical outcomes and SPC flow. Area under the receiver operating characteristics curve was used to assess the diagnostic accuracy of SPC flow for the prediction of 5 l/m2 postoperative pleural effusion. Pearson’s correlation coefficient was used to analyse simple linear relationships between measures of SPC flow and acute post-TCPC clinical outcomes. Where possible, nonnormally distributed data were transformed to their log-normal forms for parametric analysis; otherwise, non-parametric testing was performed (Kendall’s Tau-b). Independent samples t tests and ManneWhitney U tests were used to compare groups with normal and non-normal distribution, respectively.

RESULTS Demographics Sixty-five patients underwent CMR assessment after BCPC prior to consideration of TCPC completion, with a median age of 3.8 years (IQR 3.1e5.2). The morphology of the single ventricle in the study patients was left in 41.5%, right in 53.9% and indeterminate in 4.6%. Coil embolisation of SPCs had been 936

performed in 10.8% of patients prior to CMR. SPC flow in these patients was higher than in those who were not coiled (1.8 vs 1.1 l/min/m2, p¼0.005). However, in the majority of patients, collateral vessels visible on CMR angiography were small, multiple and originated from variable sources; only two patients had large, dominant vessels. Forty-one patients had TCPC completion, median age was 3.9 years (IQR 3.4e4.5). All patients underwent TCPC with an extracardiac conduit on CPB, 10 patients had fenestration. Two patients died during follow-up (table 1).

SPC measurement methods The four methods used to quantify SPC flow were assessed by Bland-Altman analysis (figure 2). SPC flow obtained from Methods A and B showed strong correlation within 2SD limits of agreement (r¼0.83, p5 l/m2) with either high sensitivity or specificity. Moreover, preoperative factors such as older age at BCPC, higher oxygen saturations, reduced branch PA size and the absence of PA pulsatility are associated with higher SPC flow. Patients included in this study were found to have multiple small vessels or networks of SPC on CMR angiography, which were not amenable to preoperative interventional coiling. Interestingly, in patients who had undergone coil embolisation before TCPC completion, the SPC flow quantified by the subsequent CMR remained significantly higher than the average value (1.8 vs 1.1 l/min/m2) of the entire cohort. This suggests larger collateral vessels only contribute partially to the total SPC volume, and that both minor and major collaterals develop concurrently. However, because preintervention quantitative assessment of SPC flow was not performed, the benefit of emoblising the large collaterals to achieve important reduction in collateral flow cannot be assessed. 938

Because of the difficulties in measuring SPC flow through conventional modalities, CMR provides a unique non-invasive method to allow quantitative assessment of SPC. Theoretically, the calculation of SPC flow using either pulmonary or systemic flow volumes is equivalent. However, technical difficulties can introduce both systematic and random measurement errors, resulting in significant differences between techniques in the same patient. Such errors result in deterioration of the method’s capacity to predict outcomes and reduce the variability accounted for by baseline covariates. In this study, absolute measures, rather than percentages, more robustly predicted outcome and had greater agreement. The difference between pulmonary venous return and PA flow was more strongly associated with clinical outcomes than SPC quantified from the difference between ascending aortic flow and systemic venous return. Furthermore, 34% of the variability of SPC volume based on pulmonary circulation was accounted for by baseline covariates, in contrast to 19% in the systemic circulation-based SPC calculation. On this basis, we believe that SPC quantification based on the pulmonary circulation is superior, which is in agreement with previous studies.2 3 However, unlike in the study by Grosse-Wortmann et al,3 where the descending aorta flow was used as a surrogate for IVC flow with the potential to miss more distal SPC, we measured the IVC flow directly. In our study, in addition to confirming that SPC flow corresponds with postoperative pleural effusion and hospital stay, the results also highlight some of the preoperative factors associated with SPC formation. While speculative, small-branch PAs may engender SPC development by reducing pulmonary blood flow and later age at BCPC may prolong the first stage physiology where pulmonary blood flow is purposefully limited. Similarly, better branch PA growth in patients with a pulsatile BCPC has been reported, which may blunt SPC formation by allowing higher pulmonary blood flow. Higher oxygen saturation is probably the result of having important collateral flow. In combination with the observation that decreased flow in the left PA is associated with increased left-sided SPC flow, evidence appears to suggest that pulmonary blood flow plays an important role in SPC pathogenesis. These results support data presented by Ichikawa et al4 who found an association between increased cardiac return on bypass (indicating increased SPC flow) and small PAs, increased oxygen saturations and the age of Fontan. The confirmation of findings obtained intraoperatively using non-invasive methods attests to the accuracy of CMR flow quantification. Heart 2012;98:934e940. doi:10.1136/heartjnl-2011-301599


Congenital heart disease

Figure 3 Correlation between pulmonary SPC volume (Method A) and (A) indexed chest drain volume following TCPC and (B) chest drain duration presented in descending order. SPC, systemic to pulmonary collateral; TCPC, total cavopulmonary connection.

Limitations The observed association of collateral flow and early postoperative clinical outcomes does not describe a mechanistic relationship. Similarly, preoperative factors identified to correspond with high SPC flow cannot definitively be shown to cause collateral formation. However, these relationships shed light on some of the risk factors for SPC formation and its subsequent effect on early post-TCPC outcome. Our study identified an association of fenestration with increased ICU stay. Fenestration is not routinely performed in our institution; however, it may be performed at the time of operation if deemed necessary by the operating surgeon; intraoperative factors considered may include Heart 2012;98:934e940. doi:10.1136/heartjnl-2011-301599

high transpulmonary gradients and low cardiac output following cessation of CPB. Therefore, the association of fenestration with longer ICU stay may reflect a more difficult perioperative course. This study also cannot evaluate the efficacy of interventions to embolise SPC prior to TCPC completion. We are in the process of undertaking a systematic study of catheterbased intervention of SPC in single ventricle patients to answer this question. Furthermore, this study did not investigate the long-term effect of SPC flow, or whether the relief of cyanosis with completion of the Fontan circulation would lead to SPC regression. Lastly, our methodology assumes that increased pulmonary venous flow results from SPC flow; however, the 939


Congenital heart disease Contributors All authors have contributed to (1) the conception and design, acquisition of data or analysis and interpretation of data; (2) drafting of the article or revising it critically for important intellectual content; and (3) final approval of the version to be published. TO and MAQ contributed equally to this work. Funding MAQ is funded by a British Heart Foundation Clinical Research Training Fellowship. TO is funded by a Marie Curie fellowship. VM is funded by the British Heart Foundation as a BHF Intermediate Fellow. AG, SK and T-YH are funded by the Fondation Leducq. AMT is funded by the National Institute of Health Research as an NIHR Senior Research Fellow and by the Fondation Leducq. Competing interests None. Ethics approval This study was conducted with the approval of the Institute of Child Health Research Ethics Committee. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2.

Figure 4 Receiver operating characteristic curve of SPC flow (Method A) for prediction of post-TCPC chest drain volume >5000 ml/m2. SPC, systemic to pulmonary collateral; TCPC, total cavopulmonary connection.

3. 4.

contribution of systemic venous to pulmonary venous collateral flow cannot be excluded.

5. 6.

CONCLUSION CMR quantification of SPC flow in single ventricle patients can be reliably obtained by calculating the difference between pulmonary venous return and PA flow. Using this method, increased SPC flow was found to be strongly associated with higher volume and prolonged postoperative pleural effusion drainage and ICU and hospital stay independent of conventional risk factors. Moreover, higher preoperative saturation, smallerbranch PAs and later age at BCPC correlated with increased SPC flow. As most patients in this study had multiple small collateral vessels with few obvious targets for intervention, the role of aggressive preoperative SPC embolisation remains unclear.

7. 8. 9. 10. 11. 12.

Triedman JK, Bridges ND, Mayer JE Jr, et al. Prevalence and risk factors for aortopulmonary collateral vessels after Fontan and bidirectional Glenn procedures. J Am Coll Cardiol 1993;22:207e15. Whitehead KK, Gillespie MJ, Harris MA, et al. Noninvasive quantification of systemic-to-pulmonary collateral flow: a major source of inefficiency in patients with superior cavopulmonary connections. Circ Cardiovasc imaging 2009;2:405e11. Grosse-Wortmann L, Al-Otay A, Yoo SJ. Aortopulmonary collaterals after bidirectional cavopulmonary connection or Fontan completion: quantification with MRI. Circ Cardiovasc Imaging 2009;2:219e25. Ichikawa H, Yagihara T, Kishimoto H, et al. Extent of aortopulmonary collateral blood flow as a risk factor for Fontan operations. Ann Thorac Surg 1995;59:433e7. McElhinney DB, Reddy VM, Tworetzky W, et al. Incidence and implications of systemic to pulmonary collaterals after bidirectional cavopulmonary anastomosis. Ann Thorac Surg 2000;69:1222e8. Powell AJ. Aortopulmonary collaterals in single-ventricle congenital heart disease: how much do they count? Circ Cardiovasc Imaging 2009;2:171e3. Stern HJ. The argument for aggressive coiling of aortopulmonary collaterals in single ventricle patients. Catheter Cardiovasc Interv 2009;74:897e900. Stern HJ. Aggressive coiling of aortopulmonary collaterals in single-ventricle patients is warranted. Pediatr Cardiol 2010;31:449e53. Bradley SM, McCall MM, Sistino JJ, et al. Aortopulmonary collateral flow in the Fontan patient: does it matter? Ann Thorac Surg 2001;72:408e15. Bradley SM. Management of aortopulmonary collateral arteries in Fontan patients: routine occlusion is not warranted. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu 2002;5:55e67. Muthurangu V, Taylor AM, Hegde SR, et al. Cardiac magnetic resonance imaging after stage I Norwood operation for hypoplastic left heart syndrome. Circulation 2005;112:3256e63. Spicer RL, Uzark KC, Moore JW, et al. Aortopulmonary collateral vessels and prolonged pleural effusions after modified Fontan procedures. Am Heart J 1996;131:1164e8.

PAGE fraction trail=6.75

940

Heart 2012;98:934e940. doi:10.1136/heartjnl-2011-301599


Systemic to pulmonary collateral blood flow influences early outcomes following the total cavopulmonary connection Tobias Odenwald, Michael A Quail, Alessandro Giardini, et al. Heart 2012 98: 934-940

doi: 10.1136/heartjnl-2011-301599

Updated information and services can be found at: http://heart.bmj.com/content/98/12/934.full.html

These include:

References

This article cites 12 articles, 4 of which can be accessed free at: http://heart.bmj.com/content/98/12/934.full.html#ref-list-1

Article cited in: http://heart.bmj.com/content/98/12/934.full.html#related-urls

Email alerting service

Topic Collections

Receive free email alerts when new articles cite this article. Sign up in the box at the top right corner of the online article.

Articles on similar topics can be found in the following collections Epidemiology (3249 articles)

Notes

To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to: http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to: http://group.bmj.com/subscribe/