Rescue High-Frequency Oscillatory Ventilation for Congenital Diaphragmatic Hernia: What about Lung Histopathology and Necropsy Findings?

1 Division of Neonatology, Department of Pediatrics, Hospital de São João, 4200-319 Porto, Portugal 2 Serviço de Neonatologia, Departamento de Pediatria, Hospital de São João–Piso 2, Alameda Prof. Hernâni Monteiro, 4200–319 Porto, Portugal Division of Pediatric Surgery, Department of Pediatrics, Hospital de São João, 4200-319 Porto, Portugal 4 School of Health Sciences, Minho University, 4709-057 Braga, Portugal 5 Faculty of Medicine, Porto University, Porto, Portugal


Introduction
e underlying pathophysiology of Bochdalek congenital diaphragmatic hernia (CDH) is that of pulmonary insufficiency and persistent pulmonary hypertension secondary to pulmonary hypoplasia. e severity of CDH is related mostly to the degree of hypoplasia, which depends on the size of the defect, the presence of the liver in the chest, and how early in gestation the abdominal contents were displaced [1].
A standardized postnatal management of infants with CDH, the CDH EURO Consortium Consensus, has been proposed in 2010 [2]. e different centers treating CDH patients use different mechanical ventilation strategies, and most target the use of gentle ventilation and permissive hypercapnia. e actual recommendations include the use of high-frequency oscillatory ventilation (HFOV) if conventional mechanical ventilation (CMV) fails [3]. Although many centers use HFOV as the primary mode of ventilation, it is not yet a resolved issue whether this approach has bene�t over CMV. e ongoing multicenter randomized controlled trial, VICI trial, will help to clarify which primary respiratory mode, CMV or HFOV, has the most bene�t on CDH patients.
Our center uses HFOV as a rescue ventilation mode for CDH patients. We conducted a study in order to evaluate the success rate of rescue HFOV, and the histological characteristics of the lungs at the autopsy exam of the deceased patients in which rescue HFOV failed.

Material and Methods
Cases of CDH were identi�ed in our database, a tertiary referral centre for neonatal surgery. All neonates diagnosed with CDH from January 1997 to December 2010 were included. Analysed data were retrieved from maternal and infant medical records. Since 2003, the management protocol of CDH has changed, and in 2010 the CDH EURO Consortium Consensus protocol has been adopted. Inhaled nitric oxide (INO, 20 ppm) was routinely used from 2003 aer echocardiography and an oxygenation index (mean airway pressure × fraction of inspired oxygen × 100/partial arterial pressure of oxygen) over 20. Sildena�l has been used in infants with persistent pulmonary hypertension refractory to INO. A policy of delayed surgery aer preoperative stabilization was practiced throughout the study period. Rescue HFOV was done using SensorMedics 3100A (Sensor Medics Corporation, Yorba, Linda, CA, USA), aer a period of unsuccessful conventional mechanical ventilation.
e position of the liver, side of herniation, degree of pulmonary hypertension, settings of CMV before changing to HFOV, settings used on HFOV, effect of rescue HFOV on the respiratory status of the patient evaluated by blood gases samples, rate of success of HFOV and histological exam of the lungs, and other necropsy �ndings of those patients whose parents consented to autopsy were analyzed. Data on the followup of discharged patients who bene�ted from rescue HFOV were also evaluated.
Results are presented in absolute number and percentage of the total sample. Continuous variables are presented in median (minimum and maximum). Mann-Whitney test was used to compare two independent samples and a P value < 0.05 was considered signi�cant.

Results
A total of 80 newborns with CDH were treated at our institution during the considered period. e demographics of patient population are reported in Table 1, and the clinical characteristics in Table 2. Rescue HFOV was used in ten patients, Table 3. HFOV was started between two hours and four days of life. e reason for rescue HFOV from CMV was hypercapnia (PaCO 2 > 60 mmHg) in �ve patients (50%) and hypercapnia plus hypoxia (PaCO 2 > 60 mmHg plus preductal saturation < 80%) in �ve patients (50%). Ventilation settings and indexes before and aer HFOV are reported in Table  4. Aer starting on rescue HFOV there was no bene�t over oxygenation in any patient, and there was a signi�cant decrease of PaCO 2 values, although not to normal values (45 and 60 mmHg) in three patients, and a decrease of PaCO 2 values to normal values in �ve patients. Two (20%) patients survived and were discharged. e rate of success of rescue HFOV in this subgroup of ten patients was 20% (2/10).
e histological �ndings of the six patients whose parents consented to necropsy are reported in Table 5.
Clinical neonatal characteristics and followup of the two survivors of CDH aer rescue HFOV are reported on Table  6.

Discussion
Our center's overall survival of CDH was 49% until 2003. Since 2003, the overall survival has raised to 67% with the implementation of new protocols of treatment. HFOV is used as rescue ventilation and according to the results of our study, two patients out of ten (20%) bene�ted from this respiratory mode.
Many studies in the literature reported that HFOV for preoperative stabilization and for intra-and postoperative respiratory treatment of CDH has been shown to be effective and associated with a superior survival rate when compared 4 ISRN Critical Care T 5: Histological �ndings of the six patients whose parents consented to necropsy. to CMV [4][5][6][7][8][9][10]. On the other hand, some other wellconducted studies did not show improvement in survival of CDH patients when HFOV was used as rescue or �rst intension respiratory mode [11][12][13]. Although some of these studies were performed years ago, in an era where therapies for pulmonary hypertension and CDH treatment protocols were limited, the comparison to CMV did favour HFOV in some studies, and did not in others. It is still an unde�ned issue whether HFOV has supremacy over CMV for CDH patients. In our study, two patients bene�ted from rescue HFOV, and the �ve years followup does not show neurodevelopmental or other signi�cant disabilities. �e other eight patients to whom HFOV was offered as rescue respiratory mode did not survive. Looking at the necropsy �ndings of the six patients whose parents consented to necropsy we can understand why ventilation was unsuccessful. One patient with a coarctation of aorta not diagnosed prenatally, and other �ndings such as pneumonia, meconium aspiration, hyaline membranes, severe muscular hypertrophy of medium and small diameter lung arteries, severe lung hypoplasia, pleural effusions, and haemorrhagic diatesis, may explain the CMV and HFOV failure. It seems that the question is how are the lungs of the CDH patients? Are they hypoplastic without other histological �ndings that may worsen the outcome, or are there other �ndings that may worsen the diagnosis? In our small series, it seems that the other �ndings on the lungs, and in one case a coarctation of aorta, could be the explanation to the fatal outcome.
Two out of ten patients who were ventilated with HFOV had a good outcome, suggesting that these lungs probably did not present the ominous �ndings of the deceased patients. In these two patients rescue, HFOV proved to be superior to CMV.
Although limited by the small number of patients and the signi�cant period of time considered (1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010) in which treatment strategies and protocols for CDH have changed, the results of this study support the idea that there is a role for HFOV in CDH, when used as rescue ventilation for selected patients or as �rst respiratory mode. Also, the results of this study show that we cannot judge the efficacy of one strategy of ventilation, if we do not know the histopathological characteristics of the lung. e reason why in some studies HFOV was, or was not, superior to CMV, could be better explained if a necropsy study and histological assessment of the lung were done.
Other studies, related to other pathologies, have shown that the necropsy �ndings and histological assessment of the lungs may change the diagnosis in a signi�cant number of cases [14,15].
More studies are needed to establish which respiratory mode, CMV or HFOV, is better for �rst respiratory treatment of CDH patients. We hope that the ongoing VICI trial may clarify this issue. e results of this study support the idea that HFOV may increase survival of CDH patients, when conventional ventilation fails.
We suggest that all deceased CDH patients should have a necropsy study in order to identify other congenital malformations that may pass undetected on prenatal and postnatal evaluation, and the lung histological �ndings that may explain the failure of the respiratory mode. In randomized studies comparing CMV and HFOV, the necropsy study is of great importance and should be part of the study, when comparing the failure of each respiratory mode.