www.elsevier.com/locate/jpedsurg

Journal of Pediatric Surgery (2011) 46, 2047–2053

Original articles

Postoperative regional distribution of pulmonaryventilation and perfusion in infants with congenitaldiaphragmatic herniaKarin C. Björkman a,⁎, Malin Kjellberg b, Sten Erik Bergströmb, Baldvin Jonssonb,Sten Lindahl a, Peter Radell a, Malin Rohdin b,1, Alejandro Sanchez-Crespo c,1

aDepartment of Physiology and Pharmacology, Karolinska Institute, Stockholm, SwedenbDepartment of Woman and Child Health, Karolinska Institute and Astrid Lindgren Children's Hospital, Stockholm, SwedencDepartment of Nuclear Medicine, Karolinska University Hospital, Stockholm, Sweden

Received 23 December 2010; revised 13 May 2011; accepted 26 June 2011

InHfa

0d

Key words:CDH;Pulmonary hypertension;Gas exchange;Lung function

AbstractBackground/Purpose: Advances in management of patients with congenital diaphragmatic hernia(CDH) have improved mortality rates but with a risk of increased pulmonary morbidity. The prognosisfor CDH survivors remains difficult to predict owing to the lack of adequate methods. We used singlephoton emission computed tomography (SPECT) to measure the regional distribution of ventilation andperfusion in CDH infants to quantify the degree of lung function impairment and relate it to neonatalclinical disease severity.Methods: Single photon emission computed tomography was performed in 12 CDH infants at the meanage of six months. Ventilation and perfusion were traced with 5 MBq Technegas and technetium-labelled albumin macro-aggregates, respectively. Neonatal clinical data collected during the patient'sstay in the pediatric intensive care unit was correlated with the SPECT data.Results: Single photon emission computed tomography revealed varying degrees of ventilation-perfusion abnormalities which correlated with the presence of pulmonary artery hypertension, days onventilator and days on extracorporeal membrane oxygenation.Conclusions: The grade of clinical disease severity in infants following CDH repair is closely related tothe ventilation-perfusion abnormality as seen using SPECT. The persistence of pulmonary arteryhypertension into the postoperative neonatal period appears to be an important pathophysiological factorrelated to ventilation-perfusion abnormalities. Single photon emission computed tomography providesvaluable clinical information for patient follow-up.© 2011 Published by Elsevier Inc.

⁎ Corresponding author. Department of Pediatric Anesthesia andtensive Care, Karolinska University Hospital, Astrid Lindgren's Childrenospital, Q9:00, SE 176 34 Stockholm, Sweden. Tel.: +46 8 5177 9909;x: +46 8 517 77265.E-mail address: Karin.Bjorkman@karolinska.se (K.C. Björkman).1 These authors contributed equally to this work.

022-3468/$ – see front matter © 2011 Published by Elsevier Inc.oi:10.1016/j.jpedsurg.2011.06.042

Congenital diaphragmatic hernia (CDH) is a developmentaldefect occurring in one in 2,000 to 4,000 births, and is one ofthe most common life-threatening congenital defects observedin the neonatal period [1]. Modern therapeutic strategies suchas lung protective ventilation with permissive hypercapnia,

2048 K.C. Björkman et al.

extracorporeal membrane oxygenation (ECMO), delayedsurgical repair, pulmonary vasodilators, and maintenance ofright-to-left shunting in cases of severe pulmonary hyperten-sion have contributed to improved survival rates beyond thenewborn period [2-4]. Unfortunately, long-term survival maybe associated with increased pulmonary morbidity andpersistent pulmonary hypertension [5-7]. Disturbances inlung development during early childhood may have importantimplications for pulmonary morbidity throughout life, assignificant alveolarization continues up to 2 to 3 years of age[8]. Although the pathophysiology of this disease is still notfully understood, a correlation between long-term morbidityand degree of lung hypoplasia and structural abnormalities ofthe pulmonary vascular bed (less number of arteries withincreasedmuscular wall thickness) has beenwell-documented,although mainly determined postmortem [9]. It is unclear towhat extent other factors, for example, disturbance in centralregulation, diaphragmatic function, upper airway control, orairway reactivity, may contribute to clinical morbidity.

The prognosis for children with severe pulmonary mor-bidity remains difficult to predict owing to the lack ofsensitive investigative methods. Chest radiographs common-ly used for surveillance and follow-up of CDH patientsprovide only limited information about morphologicalchanges [10]. Computed tomographic scan and magneticresonance imaging visualize detailed internal structures butoffer only limited information about pulmonary function.Although infant spirometry and multiple-breath washoutprovide more functionally oriented data, they only relate tooverall pulmonary function [11,12]. In order to obtaininformation about the distribution of lung ventilation (V) andperfusion (Q), planar scintigraphic methods can be employed[13-15]. Furthermore, single photon emission computedtomography (SPECT) provides improved contrast resolutionand more complete 3-dimensional spatial information of theV and Q regional distribution and their matching and, hence,offers a better functional analysis. To perform a SPECTinvestigation, the radiotracers used to mark V and Q mustremain fixed in the alveolar walls and in the lung capillarybed, respectively. In this study, ventilation and lung bloodflow distribution were imaged using Technegas aerosoland macroaggregates of human albumin marked with techne-tium 99m.

Routine application of SPECT in infants has previouslybeen problematic owing to high radiation doses, lack ofsuitable inhalation agents for the ventilation scan, anddifficulties with patient cooperation. In this study weapplied an innovative SPECT method specifically designedfor infants [16] with the aims of determining (a) the lung Vand Q distribution following CDH repair; (b) the relation-ship between the degree of V and Q abnormalities andclinical disease severity; and (c) the role of various riskfactors as predictors of subsequent lung function abnormal-ities. We hypothesized that V and Q regional distributionabnormalities are important elements in the pathophysiol-ogy of CDH.

1. Materials and methods

1.1. Patients

Twelve infants born with CDH at Astrid Lindgren'sChildren Hospital, Stockholm, Sweden (average gestationalage 37 weeks, (range, 35-41 weeks) were consecutively re-cruited between December 2006 and March 2008. All patientswere admitted to the pediatric intensive care unit andunderwent surgical repair of CDH in the newborn period. Atthe time of the SPECT examination, the patients were, onaverage, 6 months of age (range, 3-12 months). They were allclinically stable and discharged from the pediatric intensivecare unit, where the average length of stay was 25 days (range,3-64 days).

The hernia was left sided in 10 and right sided in 2 of theinfants. Nine of the patients had patch repair and 8 hadlarge defects. None of the patients in this series hadcongenital heart defects or chromosomal anomalies. Sixinfants required ECMO, and 8 infants had pulmonary arteryhypertension (PHTN) postoperatively. Two patients re-quired supplemental oxygen at the time of the SPECTinvestigation. Clinical parameters from the postnatal periodincluding days on ventilator, days on ECMO, days withsupplemental oxygen, diagnosed PHTN, and Apgar score at1 and 5 minutes were recorded for each patient. Thepresence of PHTN was retrieved retrospectively frompatient charts and diagnosed by ultrasound examinationsperformed by experienced pediatric cardiologists. Thedegree of PHTN was examined serially and generallydecreased over time, with variation according to clinicalcondition, and no child filled diagnostic criteria for PHTNat the time of the SPECT study. We did not assign thediagnosis at a particular point in time postoperatively butrather included all children who had significant PHTN atsome point postoperatively. Likewise, no single measure-ment could be used, pulmonary artery pressure could not bedetermined on all occasions even when other signsof PHTN (eg, short acceleration times, septal deviation,M-formed deceleration pattern, etc) clearly were present.Thus, the diagnosis was applied after comprehensive con-sultant evaluation, using both quantitative (eg, tricuspidinsufficiency, estimated pressure gradients) and qualitative(eg, flow characteristics) criteria.

The parents received written information about thestudy, and informed consent was obtained. The regionalethical research board and the radiation protection committeeof the Karolinska Hospital, Stockholm, Sweden, approvedthe study.

1.2. Single photon emission computed tomography

Lung perfusion was traced with human albumin macroaggregates (Mallinckrodt Medical, The Netherlands) labelledwith technetium 99m (99mTc) and administered through a

2049Postoperative regional distribution of pulmonary ventilation

peripheral venous catheter. The total amount of administeredactivity and macro aggregate particles were 5 MBq and lessthan one seventieth of the total prepared batch volume,respectively. Lung ventilation was traced using 5 MBq ofTechnegas aerosol (Tetley Manufacturing Ltd, Sydney,Australia) and administered during normal tidal breathing.The SPECT technique used in this work has previously beendescribed in detail [16]. In brief, a heart ultrasound wasperformed to exclude cardiac right-to-left shunt. Thirtyminutes before the SPECT examination, all patients werelightly sedated with 50 mg/kg oral chloral hydrate andimmobilized in a vacuum bag on the gamma camera bench.Chloral hydrate is thought to affect respiratory drive andmechanics minimally [17]. Hemoglobin oxygen saturationand heart rate were continuously monitored with pulse oxi-metry during the examination. SPECT imaging was per-formed in a three-headed gamma camera (TRIAD XLT,Trionix Research Laboratory, Twinsburg, Ohio), with 90projections and a total acquisition time of 15 minutes perscan. Two consecutive SPECT scans were performed, thefirst one corresponding to lung perfusion and the second tothe combined perfusion and ventilation distribution. Purelung ventilation images were obtained by pixel-wise sub-traction between first and second scan. Filtered back-projection was used for image reconstruction. The patientsreceived an estimated total dose of 1.5 mSv [18].

1.3. Image quantification and analysis

The SPECT analysis was performed blinded to the clini-cal history. From the set of SPECT images, the ipsilateral,same side as the hernia (i), and contralateral (c) ventilationand perfusion were defined as the total amount of imagevoxels with intensity values above noise. The relative distri-bution of the total V and Q between sides was scored andpair-wise compared at group level (n = 12). Significancetesting was performed using a non-parametric Wilcoxonmatched-pairs signed-ranks test.

From the three-dimensional SPECT data, the ratio ofthe relative distribution of ventilation and perfusion betweenthe ipsilateral and contralateral sides Vi/Vc and Qi/Qc

respectively, were quantified. In addition, we quantifiedfunctional lung volume as the fraction of the total lungvolume containing V/Q values within [0.6 and 1.4] for theipsilateral and contralateral sides, Voli, and Volc, respec-tively. With this definition, two different group stratifica-tions were tested:

1. According to the ipsilateral to contralateral functionalvolume ratio Voli/Volc; in this case, owing to the smallnumber of observations (n = 12), the patients weregrouped only in 2 subgroups, Voli/Volc higher and loweror equal than 0.5. The distribution of all consideredclinical parameters within these 2 groups were comparedusing a 2-tailed, 2-sample Student t test.

2. According to the newborn clinical status using ECMOas a surrogate measure of disease severity; the averageof the Vi/Vc, Qi/Qc, and Voli/Volc at group level werecompared between patients treated with and withoutECMO. Statistical significance was tested using a2-tailed, 2-sample Student t test.

In addition, single linear regression analysis was used toinvestigate separately the correlation between each of thedifferent SPECT derived lung function parameters and eachof the described clinical variables in each patient at grouplevel (n = 12). As test statistics, we used t score (slope oflinear regression/standard error of the slope) to determinewhether the slop of the regression differs significantly fromzero. Finally, multiple linear regression analysis was per-formed between each of the SPECT parameters and thoseclinical variables for which a single linear correlation waspreviously obtained. Using a t score, we tested the nullhypothesis that none of the clinical parameters, in a multiplevariable model, could explain any of the variations observedin the SPECT results beyond the variations obtained by anyof them separately. Statistical analyses were performed withSTATA 7.0 (Stata Corp LP, College Station, Tex). In thisformulation, the obtained regression coefficients representsthe independent contributions from each considered clinicalvariable to the prediction of the SPECT parameter.

For all statistical analysis, P b .05 was consideredsignificant.

2. Results

2.1. The ipsilateral and contralateral lungventilation and perfusion distributions

The SPECT results differed significantly among patients.Some of the patients had almost no ventilation and perfusionon the ipsilateral side, resulting in only one functional lung,while others had almost normal V and Q distribution in bothlungs. Five patients had significantly compromised ipsilat-eral lung capacity, defined as less than 20% of both the totalair and blood flow to the ipsilateral lung. Seven patients hadgood relative ventilation and perfusion distribution betweenipsilateral and contralateral sides. For the group as a whole(n = 12) there is a strong imbalance in the distribution ofV and Q between ipsilateral and contralateral side (P b .001).The ipsilateral lung contributed 30 % of ventilation and 28%of perfusion while the contralateral lung contributed 70%of ventilation and 72% of perfusion. Interestingly, there isno significant difference between the relative distribution ofV and Q within each lung side (P = .266 for both lung sides),suggesting that V and Q are equally affected.

As an example, Fig. 1 shows the results of the V, Q, andV/Q distribution for three patients with good, moderate, andpoor ipsilateral lung function, respectively.

Fig. 1 The images to the left of the panels depict the SPECT results for three different left-sided diaphragmatic hernia patients withrespectively good (A), moderate (B), and poor (C) relative ventilation (V) and perfusion (Q) distribution between the ipsilateral andcontralateral lung sides. The plots to the right of the panels represent the corresponding lung volume histograms of V/Q ratio values in eachlung side. The colored region on the plot represents the fraction of the lung volume within V/Q [0.6, 1.4].

2050 K.C. Björkman et al.

Table 1 The average of the patient's baseline clinical parameters stratified by Voli/Volc, the ipsilateral to contralateral lung volume ratiowith V/Q values within [0.6,1.4]

Voli/Volc N 0.5 (n = 7) Voli/Volc b 0.5 (n = 5) P

Group average Voli/Volc 0.80 ± 0.17 0.22 ± 0.08 .0002Days on ventilator 10.7 ± 10.7 35.4 ± 16.2 .011Days on ECMO 2.6 ± 4.5 18.4 ± 11.3 .008Days with supplement oxygen 18.9 ± 21.5 187.4 ± 184.1 .036Apgar (1 min) 6.3 ± 3.2 4.4 ± 2.9 .325Apgar (5 min) 7.6 ± 3.5 6.8 ± 2.9 .692No. with PHTN 3 5 .042

Values are mean ± SD.

2051Postoperative regional distribution of pulmonary ventilation

2.2. Ipsilateral to contralateral functional lungvolume and correlation to clinical parameters

Group stratification by the ratio of functional volume,Voli/Volc, is presented in Table 1. As this table demonstrates,patients with ipsilateral to contralateral functional lungvolume ratio larger than 0.5 (average 0.8 at group level)showed also the least severe clinical parameters, indicatingthat the degree of clinical disease severity correlates wellwith overall V/Q matching.

2.3. Relation of ECMO to the ipsilateralto contralateral V, Q, and functional lungvolume ratios

The ipsilateral to contralateral ratio of ventilation,perfusion and functional volume in patients treated withand without ECMO are shown as box plots in Fig. 2. As thisfigure shows, patients not requiring ECMO showed onaverage ratios closer to 1 than patients requiring ECMO.Median values for the ventilation, perfusion, and functionalvolume were 0.9, 0.7, and 0.8, respectively. The correspond-ing values for the patient group that required ECMO in thenewborn period were 0.2, 0.2, and 0.3, respectively. Whencomparing the two groups, significant differences werefound in the ipsilateral to contralateral relative ventilationdistribution (P = .036) as well as the functional lung volume(P = .037), whereas a borderline difference was found in

Table 2 The linear regression coefficients between each of the baselin

Vi/Vc

Coef P

Days on ventilator −0.016 .006Days on ECMO −0.001 .007Days with supplement oxygen −0.001 .051PHTN presence −0.641 .001Apgar (1 min) 0.045 .240Apgar (5 min) 0.013 .360

Coef indicates regression coefficient.

ipsilateral to contralateral relative perfusion distribution(P = .054).

2.4. The significance of each single clinicalparameter on the SPECT findings

Table 2 shows that days on ventilator, days on ECMO andpresence of PHTN in the postoperative period were linearlycorrelated with the SPECT parameters (P b .018 for allcombinations). Days with supplemental oxygen only showeda borderline significant correlation with Vi/Vc (P = .051), andno correlation with the rest of SPECT derived parameters(all combinations P N .06). Apgar at 1 and 5 minutes did notshow any significant correlation with SPECT findings (allcombinations P N .2). Hence, only days on ECMO, days onventilator and diagnosis for PHTN were carried on to themultiple linear regression analysis, shown in Table 3. Theresults presented in Table 3 support the assertion that onlypresence of PHTN significantly correlates with the degreeof V and Q distribution abnormality (P b .05).

3 Discussion

The main aims of this study were to apply a novel SPECTmethod to the study of ventilation and perfusion conditionsin infants following CDH repair, and to determine if V-Qabnormalities correlated with clinical disease severity. We

e clinical factors and the SPECT derived lung function parameters

Qi/Qc Voli /Volc

Coef P Coef P

−0.011 .020 −0.013 .009−0.0007 .014 −0.0009 .007−0.001 .068 −0.001 .058−0.500 .002 −0.542 .0010.039 .209 0.035 .2820.012 .717 .010 .760

Fig. 2 Box plots representing the ipsilateral to contralateralventilation (Vi/Vc) and perfusion (Qi/Qc) distribution ratio and theratio of the ipsilateral and contralateral lung volumes with V/Qratios within [0.6, 1.4] (Voli/Volc) stratified for diaphragmatichernia patients treated with (n = 6) and without (n = 6) ECMO. Thebox plots represent the median and the upper and lower quartiles.The bars represent the maximum and minimum value of the group.The circular marker represents an outlier.

2052 K.C. Björkman et al.

also examined several clinical parameters to see if they werepredictive of later V-Q abnormality.

The method proved applicable and well tolerated in thispatient group. Unique functional information provided bySPECT is important in understanding the degree and severityof impairment in lung function and is generally not avail-able in traditional examinations. For example, in the presentstudy, the ventilation and perfusion characteristics variedconsiderably in the study group. This variation in lungfunction was not evident from other investigations, forexample, growth, chest radiographs, or clinical evaluation ofrespiratory status at follow-up. Well-appearing children withclinically normal respiratory parameters (respiratory rate,oxygen saturation, work of breathing) could have relativelysevere abnormalities in ipsilateral ventilation and perfusionand depend almost completely on gas exchange from thecontralateral lung. Such children may have limited functionalreserve and be very susceptible to infection or other com-plications. Thus, SPECT seemed to provide valuable infor-mation useful for the clinical management of these patients.

Clinical disease severity in infants born with congenitaldiaphragmatic hernia correlates well with abnormality inventilation, perfusion, and V/Q matching as provided bySPECT. This method provided a quantitative assessment ofregional V, Q, and V/Q matching in infants. SPECT may be

Table 3 Multiple linear regression coefficients between eachof the SPECT based lung function parameters and the combineeffect of days on ventilator, days on ECMO and presenceof PHTN

Ventilator (d) ECMO (d) PHTN

Coef P Coef P-value Coef P

Vi/Vc −0.003 .762 −0.009 .585 −0.450 .034Qi/Qc 0.002 .818 −0.011 .424 −0.389 .039Voli /Volc 0.0008 .921 −0.0004 .380 −0.3961 .029

unique in its ability to evaluate local prerequisites for gasexchange (ie, ventilation and perfusion). Interestingly, thesickest infants at birth, that is, those requiring ECMO treat-ment, continued to have the most marked V-Q abnormalitiesat the time of the SPECT examination months later, con-firming our suspicion that ventilation and perfusion abnor-malities are important elements in the pathophysiology ofthis disorder.

We found that the relative regional distribution of V, Q,and V/Q matching in postoperative CDH patients can varyfrom an almost normal distribution in both lungs to severelycompromised function on the ipsilateral side. The high pro-portion of patients with abnormal ipsilateral V-Q distributionmay indicate a lack of normal postnatal lung growth duringthe first year.

Two different tendencies among the infants with severelylimited ipsilateral lung function were observed; those withequally impaired V and Q and those with a tendency to amore impaired perfusion distribution. This finding mightsupport the claim that vascular growth in some cases doesnot match alveolar growth in CDH patients as previouslysuggested [13,14,19]. Previous studies have also shown thatthe pulmonary vascular bed in CDH patients is not onlyreduced in size but responds abnormally to vasodilators andthe muscular arteries are hypertrophied and extend moreperipherally than normal [5,6].

Among the various tested clinical variables used to repre-sent the clinical severity of disease, the presence of pulmo-nary hypertension in the post-operative period is moststrongly correlated to late V-Q distribution abnormalityand may be an important indicator for the need of closefollow-up. While the small sample size limits our ability toconclude a causal relationship, the association betweenpulmonary hypertension and prolonged V-Q abnormalitiesmerits further study.

3.1. Methodological issues

The ventilation technique used in this work is based on amodification of a commercial Technegas generator [16]. Thesemodifications allow for a safe and controlled aerosoladministration, which results in a greater aerosol alveolardeposition and less impaction in central airways. This tech-nique overcomes some of the limitations of commercialventilation systems by making patient compliance unneces-sary, avoiding difficult breathing maneuvers and by mini-mizing both facemask dead space and inspiratory-expiratoryresistance. Hence, this technique fits the requirements forroutine applications in neonates and even older patients. Limi-tations hampering the application of this SPECT technique inclinical routine include patient movements between SPECTscans, limited system spatial resolution (7 mm at the center ofthe field of view), and the long image acquisition time.

In addition, a general problem for this group of patientsis the lack of established reference values for the regionaldistribution of V, Q, and V/Q from healthy infants. In this

2053Postoperative regional distribution of pulmonary ventilation

study, we compared ipsilateral to contralateral lung functionfor every patient, assuming that the contralateral lung hadrelatively “normal” function, an assumption that we know islikely not completely accurate [9]. Nonetheless, the com-parison between ipsilateral and contralateral lung functionhas been used previously for lung scintigraphy in CDHpatients [14] and ethical considerations concerning radiationexposure precluded the option of recruiting healthy infants toserve as controls. The present definition of functional lungvolume as the fraction of the total volume with V/Q values inthe interval [0.6, 1.4] was based on previous SPECT scansfrom our group in healthy adults [20,21] in which about65% of the entire healthy lung has V/Q ratios in the interval[0.8, 1.2] and more than 85% in the interval [0.6, 1.4].

These results indicate that regional maldistribution oflung ventilation and perfusion are important determinants ofpulmonary morbidity and compromised lung capacity ininfants with CDH. The SPECT technique used in this studyprovides us with valuable information in infants with CDHnot obtainable by other means. These data correlate wellwith several clinical variables representing disease severity.Among these variables, pulmonary hypertension appears tobe an important pathophysiological factor related to V-Qabnormality. This SPECT technique is a sensitive investiga-tive tool, which may help predict on-going disease severityin this vulnerable patient group and may be a valuable aid inthe management and follow-up of these infants. For example,early measurement of significant V-Q abnormalities mayhelp identify children who need close follow-up, nutritionalsupport, prolonged therapy for infections, obstructive diseaseor elevated pulmonary resistance, or even the need for furthersurgery owing to recurrent infection and/or non-function.

Acknowledgments

The authors would like to express their gratitude to thefamilies who have participated in this study. This work wassupported with grants from the Swedish Society for MedicalResearch, Stiftelsen Samariten Stockholm and SwedishHeart and Lung Foundation.

References

[1] Pober BR. Overview of epidemiology, genetics, birth defects, andchromosome abnormalities associated with CDH. Am J Med GenetPart C Semin Med Genet 2007;145C(2):158-71.

[2] Clark RH, Hardin Jr WD, et al. Current surgical management ofcongenital diaphragmatic hernia: a report from the CongenitalDiaphragmatic Hernia Study Group. J Pediatr Surg 1998;33(7):1004-9.

[3] Okuyama H, Kubota A, et al. Inhaled nitric oxide with early surgeryimproves the outcome of antenatally diagnosed congenital diaphrag-matic hernia. J Pediatr Surg 2002;37(8):1188-90.

[4] Chiu PPL, Sauera C, et al. The price of success in the management ofcongenital diaphragmatic hernia: is improved survival accompanied byan increase in long-term morbidity? J Ped Surg 2006;41(5):888-92.

[5] Mohseni-Bod H, Bohn D. Pulmonary hypertension in congenitaldiaphragmatic hernia. Semin Pediatr Surg 2007;16(2):126-33.

[6] Hayward MJ, Kharasch V, et al. Predicting inadequate long-term lungdevelopment in children with congenital diaphragmatic hernia: ananalysis of longitudinal changes in ventilation and perfusion. J PediatrSurg 2007;42(1):112-6.

[7] Khan AM, Lally KP. The role of extracorporeal membraneoxygenation in the management of infants with congenital diaphrag-matic hernia. Semin Perinatol 2005;29(2):118-22.

[8] Hislop AA. Airway and blood vessel interaction during lung deve-lopment. J Anat 2002;201(4):325-34.

[9] Beals DA, Schloo BL, et al. Pulmonary growth and remodeling ininfants with high- risk congenital diaphragmatic hernia. J Pediatr Surg1992;27(8):997-1001 [discussion 1001-2].

[10] Holt PD, Arkovitz MS, et al. Newborns with diaphragmatic hernia:initial chest radiography does not have a role in predicting clinicaloutcome. Pediatr Radiol 2004;34(6):462-4.

[11] Arena F, Baldari S, et al. Mid- and long-term effects on pulmonaryperfusion, anatomy and diaphragmatic motility in survivors of conge-nital diaphragmatic hernia. Ped Surg Int 2005;21:954-9.

[12] Proquitté H, Friberger H, et al. The effect of surgery on lung volumeand conventional monitoring parameters in ventilated newborn infants.Eur Respir J 2010;35(5):1072-8.

[13] Jeandot R, Lambert B, et al. Lung ventilation and perfusionscintigraphy in the follow up of repaired congenital diaphragmatichernia. Eur J Nucl Med 1989;15(9):591-6.

[14] Okuyama H, Kubota A, et al. Correlation between lung scintigraphyand long-term outcome in survivors of congenital diaphragmatichernia. Pediatr Pulmonol 2006;41(9):882-6.

[15] Falconer AR, Brown RA, et al. Pulmonary sequelae in survivors ofcongenital diaphragmatic hernia. Thorax 1990;45(2):126-9.

[16] Sánchez-Crespo A, Rohdin M, et al. A technique for lung ventilation-perfusion SPECT in neonates and infants. Nucl Med Commun 2008;29(2):173-7.

[17] D'Agostino J, Terndrup TE. Chloral hydrate versus midazolam forsedation of children for neuroimaging: a randomized clinical trial.Pediatr Emerg Care 2000;16:1-4.

[18] ICRP: Radiation doses to patients from radiopharmaceuticals. Annualof International Commission on Radiological Protection. Publication1998;80(28):31-33, 69.

[19] O'Toole SJ, Irish MS, et al. Pulmonary vascular abnormalities incongenital diaphragmatic hernia. Clin Perinatol 1996;23(4):781-94.

[20] Sánchez-Crespo A, Petersson J, et al. A novel quantitative dual-isotopemethod for simultaneous ventilation and perfusion lung SPET. Eur JNucl Med Mol Imaging 2002 Jul;29(7):863-75.

[21] Petersson J, Sánchez-Crespo A, et al. Physiological evaluation of anew quantitative SPECT method measuring regional ventilation andperfusion. J Appl Physiol 2004;96(3):1127-36.