- Email: [email protected]
ECMO in CDH: Is there a role?
Author’s Accepted Manuscript ECMO in CDH: Is there a role? David W. Kays
www.elsevier.com/locate/sempedsurg
PII: DOI: Reference:
S1055-8586(17)30045-8 http://dx.doi.org/10.1053/j.sempedsurg.2017.04.006 YSPSU50680
To appear in: Seminars in Pediatric Surgery Cite this article as: David W. Kays, ECMO in CDH: Is there a role?, Seminars in Pediatric Surgery, http://dx.doi.org/10.1053/j.sempedsurg.2017.04.006 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ECMO in CDH: Is there a role?
David W. Kays, MD Professor of Surgery (PAR) Johns Hopkins University School of Medicine Director of Congenital Diaphragmatic Hernia Program Director of Extracorporeal Life Support Program Johns Hopkins All Children’s Hospital St. Petersburg, Florida
601 5th Street South, Suite 306 St. Petersburg, FL 33701 Ph. 727-767-3439 FAX: 727-767-4346 [email protected]
Abstract
Despite its wide use, survival in congenital diaphragmatic hernia (CDH) patients treated with Extracorporeal Membrane Oxygenation (ECMO), as reported by the Extracorporeal Life Support Organization (ELSO), remains unchanged at 50%. High survival rates both with and without utilizing ECMO have been reported, fueling questions about the utility of ECMO support in this difficult population. This review looks at data from the Congenital Diaphragmatic Hernia study group and individual center reports, to evaluate the role of ECMO in CDH, focusing on defining the patients most likely to benefit, and discussing how those benefits can best be achieved. These data show that ECMO improves survival in those CDH patients most severely affected, but potential complications of ECMO delivery outweigh benefit in patients less severely affected. Improved results can be expected by minimizing ECMO complications, and by improving rates of CDH repair in patients that require ECMO.
Keywords: Congenital diaphragmatic hernia ECMO Survival Repair Pulmonary hypoplasia
Introduction
In 1977, German and Bartlett et al. reported the first 4 CDH patients supported with ECMO, of which one survived (1). Four decades later, CDH is now the most common indication for ECMO in neonates (2). Despite its wide availability and use, however, many remain unsure of the utility and benefit of this invasive support in this most difficult population. Multiple centers report excellent survival results using ECMO in 35 – 50% in their CDH population, but in stark contrast, other centers report very good results using ECMO sparingly, if at all. Are these experiences comparable? Several monographs specifically addressing this question of utility of ECMO in CDH have failed to derive clear conclusions (3,4).
The goal of this review is to first address then go beyond the binary question of whether ECMO improves survival in CDH, asking for which CDH patients does ECMO increase survival, how are those results best obtained, and can those results be improved? Is there an outcome difference in veno-venous (VV) versus veno-arterial (VA) ECMO in CDH? What is the optimal timing for CDH repair in the ECMO patient? Unfortunately, reporting of risk stratification data to allow direct comparison of results between series is insufficient in most reports. The task is further complicated by the fact that the ultimate survival success of ECMO in CDH is affected not only by the quality of the ECMO, but also by what comes before and after the ECMO run.
In addressing these questions, we will look to published data from the Congenital Diaphragmatic Hernia Study Group, and from individual center reports focusing on
contemporary series that report high survival rates both with, and without utilizing ECMO. We will also look at how overall treatment strategies affect CDH-ECMO outcomes, specifically the role of patient selection, ventilatory support strategies, type of ECMO, and importantly, how timing of surgical repair affects CDH-ECMO outcomes.
History
ECMO was first used in 1977 and expanded in the 1980s to rescue neonates suffering from lifethreatening hypoxemia and hypercarbia following emergent surgical repair of CDH, their courses also complicated by the ravages of aggressive ventilator support. Langham, Stolar, Newman, and many others demonstrated the life-saving potential of lung rest and the resolution of pulmonary hypertension which often occurred with the use of veno-arterial ECMO post-repair (4-6). In those 3 early reports, 32 of 46 ECMO treated patients (69.6%) survived to discharge, and the future of ECMO support in CDH appeared bright.
Since then, the CDH diagnosis and treatment background has changed. Ventilation strategies have evolved and improved, vastly decreasing the potential for ventilator induced lung injury (5-7). Affected newborns, for better or worse, are now stabilized for days or longer before surgical repair, and as a result the vast majority requiring ECMO arrive unrepaired, rather than post-repair as in the early years (8,9). Pulmonary vasodilators including nitric oxide and others have become ubiquitous in the management of CDH, although their effect on improving, or perhaps worsening CDH outcomes is similarly questioned (10-12). It is possible that
contemporary CDH patients requiring ECMO, as a result of better non-ECMO management, are more severe than in previous eras. It is also possible that, since the majority of CDH patients are now diagnosed prenatally, terminations of more severe CDH fetuses could be truncating the CDH severity spectrum (13). No data currently exist to clarify if the underlying severity of CDH patients has changed over time, but it is clear that the survival of those supported by ECMO, as tracked by the Extra-Corporeal Life Support Organization (ELSO), has not changed over the last decade and remains at just 50% (14). After 30 years of great effort, either the limitations of ECMO benefit in CDH have been reached, or there still remains significant opportunity for improvement.
CDH Spectrum of disease
In 1999, a seminal report from the CDH study group defined the relationship between CDH disease severity, and the survival benefit from ECMO. Compared to conventional therapy alone, ECMO support significantly improved survival for those CDH neonates with a predicted mortality greater than 80%, as defined by a logistic regression equation based on birthweight and 5- minute Apgar scores (15). In contrast, when ECMO was used on patients with less severe disease, ECMO exerted either an equivocal or negative effect on survival, with the negative effect increasing as disease severity decreased (Figure 1). Since support with ECMO adds risks to the CDH patient, such as bleeding from anticoagulation, these risks must be offset with greater benefit before additional CDH survival accrues. Utilizing ECMO in less severe patients increases mortality due to added risk without benefit, and only in the more severe patients is
the added risk offset by increased survival opportunity. Future opportunities for improving survival include minimizing ECMO in those patients on the less severe end of the severity spectrum, and also minimizing risks of ECMO related complications, ie. doing better ECMO, in those on the more severe end of the spectrum that require ECMO.
Understanding and defining CDH disease severity, therefore, is central to understanding the potential benefits of ECMO in CDH. Following the original CDH study group predicted survival equation based on 5-minute Apgar and birthweight, additional work on prenatal anatomic and postnatal physiologic markers of severity have added considerable granularity to our understanding of CDH disease severity (16-19). Of physiologic markers, birth physiology and early blood gases provide more discrimination of underlying disease severity than later blood gases or pre-ECMO physiology (20). This occurs because blood gases tend to improve over the first 12 – 24 hours of life for most CDH patients, and the levels of physiologic derangement immediately before ECMO are similar for the majority of CDH patients. This “homogenization” diminishes the power of those physiologic markers at those time points to differentiate the underlying levels of disease severity. In contrast, anatomic measures of CDH severity including prenatal lung measurements, liver position, and percent liver herniation, along with the postnatal measures of defect size and patch rate, are not affected by treatment factors or clinical course. Anatomic measures are therefore more useful for comparing disease severity between studies where treatment strategies may differ. Of these, thoracic liver position (liverup), especially for left CDH, has proven an important and relatively simple risk stratifier, correlating strongly with increased risk for ECMO and for mortality (19,21,22). A meta-analysis
of liver herniation in CDH showed survival rates of 74% with liver down CDH, which dropped to 45% in liver-up CDH (21). Finally, defect size correlates strongly with survival and risk for ECMO, and “patch rate” functions as a reasonable surrogate for larger defect sizes (23-25). In addition, outborn patients as a group represent a less severe cohort compared to inborn, as the most severe outborn patients do not survive birth and transfer, effectively pruning the severity spectrum. These proven correlations to CDH severity; inborn status, liver position, and patch rate can help inform population severity as we compare historical series.
Trials and Series
Randomized controlled data on the role of ECMO in CDH is limited to two early ECMO studies, and the UK ECMO trial (26-28). In the UK ECMO trial, randomization to conventional ventilation vs ECMO occurred at an oxygenation index of 40. Seventeen of 17 CDH infants randomized to conventional management died, while 4 of 18 in the ECMO arm survived (0% survival vs 22% survival). One of those survivors subsequently died at 2 years of age.
In 2006, Morini et al published a systematic review of ECMO in CDH, identifying 658 publications of which 21 (2043 patients) met entry criteria (29). Looking at the most rigorous of the non-randomized studies, they concluded ECMO use was associated with a reduction in CDH mortality. Zalla et al recently reviewed a single center CDH experience dividing 16 years of treatment into 4 eras, the latter 2 with ECMO availability. Post-hoc analysis suggested a 73%
reduction in risk of death in the ECMO eras compared to the pre-ECMO eras despite increases in CDH disease severity (p<0.001) (30).
Table 1 shows a compilation of single center publications since 2000 that show high survival, from centers reporting either low or higher ECMO usage, that were assembled for this publication. These reports were chosen because they document very good results, appear to have minimal selection bias, and provide some anatomic data for risk stratification. While all the results are excellent and the authors are to be congratulated, it appears notable that the highest survival rates and the highest severity, as reflected in the higher patch rates, were reported in the centers with higher ECMO use. While hardly definitive, these data support the concept that although the majority of CDH patients can be successfully supported without ECMO, those of higher to highest severity will have a survival advantage with well performed ECMO support.
Highest Severity Patients
The most compelling argument that ECMO can increase survival in CDH comes from the patients most severely affected, as defined in three contemporary reports. In 2012, Yoder et al looked at the highly severe subset of patients reported to the CDH Study Group patients that failed to achieve pre-ductal saturations of 85% in the first 24 hours, or before going onto ECMO(31). Twenty-three percent survived, and of those over 80% were treated with ECMO. In 2012 Stoffan reported seventeen patients with severe pulmonary hypoplasia as defined by
predicted lung volumes of less than 15%. Seven of 17 survived (41%) and all required treatment with ECMO (32). And in 2015, Kays et al used multi-variate modelling to define the worst 10% of 172 consecutive inborn, prenatally diagnosed CDH patients (33). Of the 19 worst, all of whom were aggressively resuscitated at birth and showed an average pH at 1 hour of life of 6.83 with a PCO2 great than 100, 10 of 16 ECMO-eligible patients survived to discharge (63%), while all 3 ECMO-ineligible prematures died. Fifteen of those 16 survivors were rescued with ECMO. While the reported use of ECMO never proves its necessity, these data are compelling evidence of the potential for ECMO to rescue patients at the highly severe end of the CDH spectrum, a capability not currently well documented for non-ECMO therapies.
Improving survival on ECMO
Considerable evidence suggests that CDH-ECMO survival results can be improved. Price and Stolar showed as early as 1991 that CDH patients treated with ECMO often died of potentially preventable treatment complications, not pulmonary hypoplasia (34). In a more recent report from 2013, 24 of 96 CDH –ECMO patients died (75% survival) but only 6 of the 24 deaths were attributed to pulmonary hypoplasia. The other 18 were ascribed to potentially preventable complications such as technical ECMO issues, intracranial bleeding, surgical bleeding, compartment syndrome, sepsis, and failure to achieve stability (19). Since end-stage pulmonary hypoplasia appears to account for a minority of deaths in CDH-ECMO patients, even in a series with 75% CDH-ECMO survival, the development and dissemination of best ECMO practices in this highly difficult population could lead to substantial survival improvements.
As an example, many authors have demonstrated the negative effect of ventilator induced lung injury on CDH survival. In 2011, Antonoff et al reported improved CDH-ECMO survival from 20% to 82% after instituting a treatment protocol centered on consistent lung protective ventilation. The institution of “gentle” ventilation strategies have increased ECMO survival and total survival in multiple series, and is a current mainstay of CDH management (5,6,35).
Veno-venous vs. Veno-arterial ECMO The choice of veno-venous (VV) versus veno-arterial (VA) ECMO in CDH has so far not been demonstrated to affect survival, but the current available data is poorly controlled for underlying disease severity. Dimmitt in 2001, and Guner et al in 2009 reviewed the cumulative ELSO experience on the topic, the latter covering 15 years from 1991 – 2006 (36,37). Guner et al found that VA ECMO was used 82% vs 18% of the time for VV. When controlling for various physiologic measures at the time of cannulation neither report demonstrated a difference in survival by ECMO delivery type in CDH, but Guner reported that 18% of VV ECMO patients required conversion to VA, with survival dropping from 54% (VV) to 44% (VV to VA) in those patients, compared to 50% when ECMO started as VA. Unfortunately, neither birth nor anatomic risk stratification was included. Kugelman et al reported no difference in survival with VV vs. VA (69% vs 68%) but the VV patients were larger (3.44 kg vs 2.77 kg, p<0.05) and less inborn (31 vs 58%, p=ns), both markers of less severity in the VV group (38). Neither liver position nor prenatal lung measurements were included in any of these reports. That VV patients were larger is not a surprise, as VV cannulas are larger requiring larger veins for safe
insertion. This issue is compounded in CDH patients, as they appear to have smaller jugular veins than their non-CDH neonatal ECMO counterparts (39).
Timing of CDH repair relative to ECMO
CDH patients supported with ECMO but never repaired do not survive. Repair confers a survival advantage, especially in the most severe. In the highly severe CDH patients reported by Yoder et al from the CDH study group that failed to attain a pre-ductal saturation of 85% in the first 24 hours of life or before ECMO, survival rose from 23% in the total to 44% in the subset where CDH repair was performed (31). However since ECMO adds significant risk and difficulty to repair, timing of repair relative to ECMO has become an important topic. Options include repair before ECMO, repair “early” during ECMO, repair “late” during ECMO, or repair after ECMO. Each strategy has its own risk-benefit profile, and its own proponents.
A 2009 study using CDH Study Group data showed that compared to repair on ECMO (44% survival), repair after ECMO showed a survival advantage (77% survival) (40). Others have provided supportive data to this important paper, and the strategy of repair after ECMO is currently widely practiced (40,41). The authors controlled for disease severity based on anatomic and physiologic measures and did not find statistically significant differences in the patients repaired on or off of ECMO. The authors did not, however, discuss the significant potential for selection bias inherent in comparing unrepaired CDH patients that had successfully weaned from ECMO, a major milestone, to patients that had not yet and might not
reach that critical level of recovery. Although the patient groups were not statistically different, those repaired on ECMO had lower 5 minute Apgar scores, more frequent need for patch repair, and a higher rate of “D” size defects (diaphragm agenesis), all suggestive of worse disease compared to those who successfully weaned from ECMO unrepaired. Additionally, the mean length of the ECMO run for those repaired after ECMO was 8.4 days, while the mean age of repair of those repaired while still on ECMO was 8.9 days, clarifying that as a group, those repaired on ECMO had not achieved the same degree of physiologic stability at the same time point as those discontinued from ECMO. While this study showed clearly that repair after ECMO works well for those CDH patients that successfully weaned from ECMO in an acceptable time frame, it did not address the fate of similar appearing patients that failed to wean from ECMO.
Unrepaired patients that fail to wean from ECMO either undergo repair late on ECMO, or never get repaired. Golden reported 66 CDH-ECMO patients of whom 33 were successfully weaned from ECMO unrepaired at a mean of 11.8 days of ECMO (41). However, twenty-two failed to wean and were repaired at a mean of 22 days on ECMO, and 11 died unrepaired after a mean of 16 days of ECMO. Eighty-five percent of those repaired after ECMO survived, but only 36% of those repaired late on ECMO survived, and 17% died on ECMO unrepaired. Overall CDH-ECMO survival in this study was 54%. Two additional single center studies defined an 11% and 15% mortality attributed to non-repair on ECMO (19,41). Together these studies highlight the potential fate of those unrepaired CDH patients that fail to wean from ECMO in a timely
fashion, and whose outcomes must be considered in a comprehensive evaluation of the strategy of CDH repair after ECMO.
To avoid deaths attributable to late or non-repair on ECMO, Dassinger et al performed “early” repair on ECMO, repairing 34 CDH on ECMO at an average of 55 hours after ECMO initiation. Only nine percent suffered bleeding complications requiring intervention, and a total of 22 (71%) survived. Fallon et al reported 46 CDH-ECMO patients. Survival was 73% for those repaired early on ECMO, 50% for those repaired late, and 64% for those repaired after ECMO. These data provide increasing evidence both for the benefit of repair, and of the potential advantage of early versus late repair in the ECMO run. To minimize the risk of operative bleeding on ECMO, Aminocaproic acid (Amicar) is used by many but not all (42,43) (44,45).
A different approach to avoid deaths due to non-repair on extracorporeal support is to identify and repair patients at high risk for needing ECMO, when feasible, before they decompensate enough to require ECMO, providing subsequent ECMO support after repair only when needed. Kays et al developed multivariate modeling employing prenatal lung measurements, liver position, birth physiology, and one hour blood gas data to predict subsequent risk of ECMO at one hour of life. They showed 95% survival in those left liver-up patients that were repaired early before ECMO (mean time to repair = 21 hours), compared to 65% survival achieved in an equivalent group of left liver-up patients that arrived to ECMO unrepaired (46). In the patients that arrived to ECMO unrepaired, 4 were repaired late when they failed to improve on ECMO and did not survive, while 3 developed organ failure on ECMO and died unrepaired. The authors
attributed their increased survival on a strategy that both assured repair in high-risk infants, and which also minimized bleeding risk by performing repair before the anticoagulation associated with ECMO.
Duration of ECMO
Although poor CDH survival beyond 2 weeks of ECMO has been reported (17), recent data suggests longer ECMO runs are needed for patients with more severe disease as defined by a variety of physiologic and anatomic stratifiers (19). In that series, survival of 43% of patients still on ECMO at 4 weeks was demonstrated. Based on these data, the authors recommended against arbitrarily defined durations of ECMO support.
ECMO Delivery systems
Rarely discussed but potentially important, especially in CDH-ECMO because of the longer length of ECMO runs, are the outcomes differences for infants supported with roller head ECMO systems versus centrifugal pump systems. In 2012, Barrett et al reviewed 1592 neonates that received ECMO, 163 by centrifugal pump and 1492 by roller head pump, and showed that those supported with centrifugal pumps had a nearly 8 fold risk of hemolysis, 21 fold risk of hyperbilirubinemia, 3 fold risk of hypertension, and greater than a 2 fold risk of renal failure compared to patients supported with roller head pumps (47). Although survival to discharge
wasn’t different between pump types, these effects could be magnified in the long runs of CDH infants, and this topic deserves monitoring.
Summary
There is wide center variation in CDH treatment and results, shown by Reickert et al. in 1998, and this remains the situation today (48). While many centers have shown very good survival employing minimal to no ECMO in CDH, the highest overall CDH survival rates reported in the literature and likely including the sickest patients, are reported by centers that utilize ECMO. Three of four centers shown in Table 1 with high overall survival rates also report ECMO survival rates in excess of 70%, substantially different from the ELSO reported survival of 50%. It is likely that these unsatisfying survival rates contribute to the ambivalence regarding benefit from ECMO in CDH. But since higher survival rates are possible, and ECMO has the unique potential to rescue CDH patients with profound pulmonary hypoplasia, the role of ECMO in severe CDH seems secure.
Optimal patient selection for ECMO in CDH requires refinement of non-ECMO support techniques so that this higher risk but higher potential reward modality is focused primarily on those patients with more severe CDH as defined by smaller lungs, worse birth physiology, liverup and with larger defects. Ventilatory support that strictly limits lung injury can improve ECMO outcomes, and the importance of operative repair in achieving survival of patients with CDH, especially of those on the more severe end of the spectrum cannot be overstated. While repair
after ECMO works well for those that successfully wean from ECMO, a strategy of early repair on ECMO, or possibly repair before ECMO, appears to provide a higher potential for overall survival, provided the repair can be done with acceptably low bleeding rates on support.
At this point 4 decades later, it seems unlikely that a prospective randomized study of ECMO support specific to CDH will be performed, highlighting the importance of the collective data provided by the CDH Study Group and that of well-reported single center experiences. Ideally future clinical CDH reports will include a full cadre of both early physiologic and anatomic markers, especially liver position, liver-up size, patch rate, and defect size, to support better risk assessment and comparison of outcomes between studies.
Finally, based on the data presented above and significant professional experience with CDH care, I feel confident that ECMO improves survival potential in more severe CDH over and above that currently available by non-ECMO techniques. Significant upside still exists for improving CDH outcomes, and since the majority of deaths occur in those more severely affected, improved ECMO outcomes will likely be central to realizing it.
Table 1
ECMO Reference
Year
Patients
Isol Surv
% ECMO
2004 & 2006
65
83%
(51)
2016
59
(52)
2001
Surv
% Inborn
% Patch
0%
nr
nr
73%
3%
100%
29%
81
80%
0%
49%
37%
2002
120
80%
13%
63%
56%
7%
(54)
2003
39
92%
36%
86%
72%
38%
(55)
2012
49
86%
29%
50%
nr
49%
(44)
2013
268
88%
40%
75%
72%
65%
Little or No ECMO
(49,50)
Low ECMO
(53)
Higher ECMO Use
nr = not reported Isol = Isolated Surv= survival
Table 1:
This table highlights single center high survival CDH reports, published since 2000, from centers utilizing either minimal or no ECMO, a low rate of ECMO, or higher (average) rates of ECMO.
References
1.
German JC, Gazzaniga AB, Amlie R, et al. Management of pulmonary insufficiency in diaphragmatic hernia using extracorporeal circulation with a membrane oxygenator (ECMO). J Pediatr Surg. 1977;12(6):905–12.
2.
Rais-Bahrami K, Van Meurs KP. Venoarterial versus venovenous ECMO for neonatal respiratory failure. Semin Perinatol. 2014;38(2):71–7.
3.
Harrington KP, Goldman AP. The role of extracorporeal membrane oxygenation in congenital diaphragmatic hernia. Semin Pediatr Surg. 2005;14(1):72–6.
4.
Morini F, Goldman A, Pierro A. Extracorporeal membrane oxygenation in infants with congenital diaphragmatic hernia: a systematic review of the evidence. Eur J Pediatr Surg. 2006;16(6):385–91.
5.
Wilson JM, Lund DP, Lillehei CW, et al. Congenital diaphragmatic hernia--a tale of two cities: the Boston experience. J Pediatr Surg. 1997;32(3):401–5.
6.
Kays DW, Langham MR Jr, Ledbetter DJ, et al. Detrimental effects of standard medical therapy in congenital diaphragmatic hernia. Ann Surg. 1999; 230(3):34051.
7.
Wung JT, James LS, Kilchevsky E, et al. Management of infants with severe respiratory failure and persistence of the fetal circulation, without hyperventilation. Pediatrics. 1985;76(4):488–94.
8.
Tsao K, Lally KP. Surgical management of the newborn with congenital diaphragmatic hernia. Fetal Diagn Ther. 2011;29(1):46–54.
9.
Moyer V, Moya F, Tibboel R, et al. Late versus early surgical correction for congenital diaphragmatic hernia in newborn infants. Moyer VA, editor. Cochrane Database Syst Rev; 2002;(3):CD001695.
10.
Malowitz JR, Hornik CP, Laughon MM, et al. Management Practice and Mortality for Infants with Congenital Diaphragmatic Hernia. Am J Perinatol. 2015 Jul;32(9):887–94.
11.
Hagadorn JI, Brownell EA, Herbst KW. Trends in treatment and in-hospital mortality for neonates with congenital diaphragmatic hernia. J Perinatol; 2015 Sep;35(9):748–54.
12.
Putnam LR, Tsao K, Morini F, et al. Evaluation of Variability in Inhaled Nitric Oxide Use and Pulmonary Hypertension in Patients With Congenital Diaphragmatic Hernia. JAMA Pediatr. 2016 1;170(12):1188-1194.
13.
Burgos CM, Frenckner B. Addressing the hidden mortality in CDH: A populationbased study. J Pediatr Surg. 2016 . In Press (epub ahead of print)
14.
ELSO International Summary. Extracorporeal Life Support Organization, January, 2017
15.
The Congenital Diaphragmatic Hernia Study Group. Does extracorporeal membrane oxygenation improve survival in neonates with congenital diaphragmatic hernia? J Pediatr Surg. 1999;34(5):720–4–discussion724–5.
16.
The Congenital Diaphragmatic Hernia Study Group. Estimating disease severity of congenital diaphragmatic hernia in the first 5 mi nutes of life. J Pediatr Surg 2001;36(1):141–5.
17.
Tiruvoipati R, Vinogradova Y, Faulkner G, et al. Predictors of outcome in patients with congenital diaphragmatic hernia requiring extracorporeal membrane oxygenation. J Pediatr Surg. 2007 Aug;42(8):1345–50.
18.
Seetharamaiah R, Younger JG, Bartlett RH, et al. Factors associated with survival in infants with congenital diaphragmatic hernia requiring extracorporeal membrane oxygenation: a report from the Congenital Diaphragmatic Hernia Study Group. J Pediatr Surg. 2009;44(7):1315–21.
19.
Kays DW, Islam S, Richards DS, et al. Extracorporeal life support in patients with congenital diaphragmatic hernia: how long should we treat? J Am Coll Surg. 2014;218(4):808–17.
20.
Khmour AY, Konduri GG, Sato TT, et al. Role of admission gas exchange measurement in predicting congenital diaphragmatic hernia survival in the era of gentle ventilation. J Pediatr Surg. 2014;49(8):1197–201.
21.
Mullassery D, Ba'ath ME, Jesudason EC,et al. Value of liver herniation in prediction of outcome in fetal congenital diaphragmatic hernia: a systematic review and meta-analysis. Ultrasound Obstet Gynecol. 2010 May;35(5):609–14.
22.
Lazar DA, Ruano R, Cass DL, et al. Defining “liver-up”: does the volume of liver herniation predict outcome for fetuses with isolated left-sided congenital diaphragmatic hernia? J Pediatr Surg. 2012;47(6):1058–62.
23.
Congenital Diaphragmatic Hernia Study Group, Morini F, Valfrè L, Capolupo I, Lally KP, Lally PA, et al. Congenital diaphragmatic hernia: defect size correlates with developmental defect. J Pediatr Surg. 2013 Jun;48(6):1177–82.
24.
Lally KP, Lasky RE, Lally PA, et al. Standardized reporting for congenital diaphragmatic hernia - An international consensus. J Pediatr Surg.; 2013 Dec;48(12):2408–15.
25.
Migliazza L, Bellan C, Alberti D, et al. Retrospective study of 111 cases of congenital diaphragmatic hernia treated with early high-frequency oscillatory ventilation and presurgical stabilization. J Pediatr Surg. 2007;42(9):1526–32.
26.
Bartlett RH, Gazzaniga AB, Toomasian J, et al. Extracorporeal membrane oxygenation (ECMO) in neonatal respiratory failure. 100 cases. Ann Surg. 1986;204(3):236–45.
27.
Schumacher RE, Roloff DW, Chapman R, et al. Extracorporeal membrane oxygenation in term newborns. A prospective cost-benefit analysis. ASAIO J. 1993;39(4):873–9.
28.
The UK Collaborative ECMO Trial Group. UK collaborative randomised trial of neonatal extracorporeal membrane oxygenation. The Lancet. 1996 ;348(9020):75–82.
29.
Morini F, Goldman A, Pierro A. Extracorporeal membrane oxygenation in infants with congenital diaphragmatic hernia: a systematic review of the evidence. Eur J Pediatr Surg. 2006;16(6):385–91.
30.
Zalla JM, Stoddard GJ, Yoder BA. Improved mortality rate for congenital diaphragmatic hernia in the modern era of management: 15 year experience in a single institution. J Pediatr Surg. 2015;50(4):524–7.
31.
Yoder BA, Lally PA, Congenital Diaphragmatic Hernia Study Group. Does a highest pre-ductal O(2) saturation <85% predict non-survival for congenital diaphragmatic hernia? J Perinatol. 2012;32(12):947–52.
32.
Stoffan AP, Wilson JM, Jennings RW, et al. Does the ex utero intrapartum treatment to extracorporeal membrane oxygenation procedure change outcomes for high-risk patients with congenital diaphragmatic hernia? J Pediatr Surg. 2012;47(6):1053–7.
33.
Kays DW, Islam S, Perkins JM, et al. Outcomes in the physiologically most severe congenital diaphragmatic hernia (CDH) patients: Whom should we treat? J
Pediatr Surg. 2015;14;50(6):893–7. 34.
Price MR, Galantowicz ME, Stolar CJ. Congenital diaphragmatic hernia, extracorporeal membrane oxygenation, and death: a spectrum of etiologies. J Pediatr Surg. 1991;26(9):1023–6–discussion1026–7.
35.
Antonoff MB, Hustead VA, Groth SS, et al. Protocolized management of infants with congenital diaphragmatic hernia: effect on survival. J Pediatr Surg. 2011;46(1):39–46.
36.
Dimmitt RA, Moss RL, Rhine WD, et al. Venoarterial versus venovenous extracorporeal membrane oxygenation in congenital diaphragmatic hernia: the Extracorporeal Life Support Organization Registry, 1990-1999. J Pediatr Surg. 2001;36(8):1199–204.
37.
Guner YS, Khemani RG, Qureshi FG, et al. Outcome analysis of neonates with congenital diaphragmatic hernia treated with venovenous vs venoarterial extracorporeal membrane oxygenation. J Pediatr Surg. 2009;44(9):1691–701.
38.
Kugelman A, Gangitano E, Pincros J, et al. Venovenous versus venoarterial extracorporeal membrane oxygenation in congenital diaphragmatic hernia. J Pediatr Surg. 2003;38(8):1131–6.
39.
Frenckner B, Palmér K, Lindén V. Neonates with congenital diaphragmatic hernia have smaller neck veins than other neonates-An alternative route for ECMO cannulation. J Pediatr Surg. 2002;37(6):906–8.
40.
Congenital Diaphragmatic Hernia Study Group, Bryner BS, West BT, et al. Congenital diaphragmatic hernia requiring extracorporeal membrane oxygenation: does timing of repair matter? J Pediatr Surg. 2009;44(6):1165–71– discussion1171–2.
41.
Golden J, Jones N, Zagory J, et al. Outcomes of congenital diaphragmatic hernia repair on extracorporeal life support. Pediatr Surg Int.; 2016 Nov 11;:1–7.
42.
Wilson JM, Wilson JM, Lund DP, et al. Congenital diaphragmatic hernia—A tale of two cities: The Boston experience. J Pediatr Surg. 1997;32(3):401–5.
43.
Keijzer R, Wilschut DE, Houmes RJ, et al. Congenital diaphragmatic hernia: to repair on or off extracorporeal membrane oxygenation? J Pediatr Surg. 2012;47(4):631–6.
44.
Kays DW, Islam S, Larson SD, et al. Long-term maturation of congenital diaphragmatic hernia treatment results: toward development of a severity-specific treatment algorithm. Ann Surg. 2013;258(4):638–44–discussion644–5.
45.
Prabhu S, Mattke AC, Anderson B, et al. Repair of congenital diaphragmatic hernia during extracorporeal life support: experience with six neonates. ANZ J Surg. 2016 Sep;86(9):711–6.
46.
Kays DW, Talbert JL, Islam S, et al. Improved Survival in Left Liver-Up Congenital Diaphragmatic Hernia by Early Repair Before Extracorporeal Membrane Oxygenation: Optimization of Patient Selection by Multivariate Risk Modeling. J Am Coll Surg. 2016;222(4):459–70.
47.
Barrett CS, Jaggers JJ, Cook EF, et al. Outcomes of neonates undergoing extracorporeal membrane oxygenation support using centrifugal versus roller blood pumps. Ann Thorac Surg. 2012;94(5):1635–41.
48.
Reickert CA, Hirschl RB, Atkinson JB, et al. Congenital diaphragmatic hernia survival and use of extracorporeal life support at selected level III nurseries with multimodality support. Surgery. 1998;123(3):305–10.
49.
Bagolan P, Casaccia G, Crescenzi F, et al. Impact of a current treatment protocol on outcome of high-risk congenital diaphragmatic hernia. J Pediatr Surg. 2004;39(3):313–8–discussion313–8.
50.
Casaccia G, Crescenzi F, Dotta A, et al. Birth weight and McGoon Index predict mortality in newborn infants with congenital diaphragmatic hernia. J Pediatr Surg. 2006;41(1):25–8–discussion25–8.
51.
Straňák Z, Krofta L, Haak LA, et al. Antenatal assessment of liver position, rather than lung-to-head ratio (LHR) or observed/expected LHR, is predictive of outcome in fetuses with isolated left-sided congenital diaphragmatic hernia. J Matern Fetal Neonatal Med; 2016 Mar 24:1–5.
52.
Al-Shanafey S, Giacomantonio M, Henteleff H. Congenital diaphragmatic hernia: experience without extracoporeal membrane oxygenation. Pediatr Surg Int.; 2002;18(1):28–31.
53.
Boloker J, Bateman DA, Wung JT, et al. Congenital diaphragmatic hernia in 120 infants treated consecutively with permissive hypercapnea/spontaneous
respiration/elective repair. J Pediatr Surg. 2002. 37(3):357-66. 54.
Downard CD, Jaksic T, Garza JJ, et al. Analysis of an improved survival rate for congenital diaphragmatic hernia. J Pediatr Surg. 2003;38(5):729–32.
55.
Guidry CA, Hranjec T, Rodgers BM, et al. Permissive hypercapnia in the management of congenital diaphragmatic hernia: our institutional experience. J Am Coll Surg. 2012;214(4):640–645–647.e1–discussion646–7.
From J Ped Surg 34: 720 -725, page 722
www.elsevier.com/locate/sempedsurg
PII: DOI: Reference:
S1055-8586(17)30045-8 http://dx.doi.org/10.1053/j.sempedsurg.2017.04.006 YSPSU50680
To appear in: Seminars in Pediatric Surgery Cite this article as: David W. Kays, ECMO in CDH: Is there a role?, Seminars in Pediatric Surgery, http://dx.doi.org/10.1053/j.sempedsurg.2017.04.006 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ECMO in CDH: Is there a role?
David W. Kays, MD Professor of Surgery (PAR) Johns Hopkins University School of Medicine Director of Congenital Diaphragmatic Hernia Program Director of Extracorporeal Life Support Program Johns Hopkins All Children’s Hospital St. Petersburg, Florida
601 5th Street South, Suite 306 St. Petersburg, FL 33701 Ph. 727-767-3439 FAX: 727-767-4346 [email protected]
Abstract
Despite its wide use, survival in congenital diaphragmatic hernia (CDH) patients treated with Extracorporeal Membrane Oxygenation (ECMO), as reported by the Extracorporeal Life Support Organization (ELSO), remains unchanged at 50%. High survival rates both with and without utilizing ECMO have been reported, fueling questions about the utility of ECMO support in this difficult population. This review looks at data from the Congenital Diaphragmatic Hernia study group and individual center reports, to evaluate the role of ECMO in CDH, focusing on defining the patients most likely to benefit, and discussing how those benefits can best be achieved. These data show that ECMO improves survival in those CDH patients most severely affected, but potential complications of ECMO delivery outweigh benefit in patients less severely affected. Improved results can be expected by minimizing ECMO complications, and by improving rates of CDH repair in patients that require ECMO.
Keywords: Congenital diaphragmatic hernia ECMO Survival Repair Pulmonary hypoplasia
Introduction
In 1977, German and Bartlett et al. reported the first 4 CDH patients supported with ECMO, of which one survived (1). Four decades later, CDH is now the most common indication for ECMO in neonates (2). Despite its wide availability and use, however, many remain unsure of the utility and benefit of this invasive support in this most difficult population. Multiple centers report excellent survival results using ECMO in 35 – 50% in their CDH population, but in stark contrast, other centers report very good results using ECMO sparingly, if at all. Are these experiences comparable? Several monographs specifically addressing this question of utility of ECMO in CDH have failed to derive clear conclusions (3,4).
The goal of this review is to first address then go beyond the binary question of whether ECMO improves survival in CDH, asking for which CDH patients does ECMO increase survival, how are those results best obtained, and can those results be improved? Is there an outcome difference in veno-venous (VV) versus veno-arterial (VA) ECMO in CDH? What is the optimal timing for CDH repair in the ECMO patient? Unfortunately, reporting of risk stratification data to allow direct comparison of results between series is insufficient in most reports. The task is further complicated by the fact that the ultimate survival success of ECMO in CDH is affected not only by the quality of the ECMO, but also by what comes before and after the ECMO run.
In addressing these questions, we will look to published data from the Congenital Diaphragmatic Hernia Study Group, and from individual center reports focusing on
contemporary series that report high survival rates both with, and without utilizing ECMO. We will also look at how overall treatment strategies affect CDH-ECMO outcomes, specifically the role of patient selection, ventilatory support strategies, type of ECMO, and importantly, how timing of surgical repair affects CDH-ECMO outcomes.
History
ECMO was first used in 1977 and expanded in the 1980s to rescue neonates suffering from lifethreatening hypoxemia and hypercarbia following emergent surgical repair of CDH, their courses also complicated by the ravages of aggressive ventilator support. Langham, Stolar, Newman, and many others demonstrated the life-saving potential of lung rest and the resolution of pulmonary hypertension which often occurred with the use of veno-arterial ECMO post-repair (4-6). In those 3 early reports, 32 of 46 ECMO treated patients (69.6%) survived to discharge, and the future of ECMO support in CDH appeared bright.
Since then, the CDH diagnosis and treatment background has changed. Ventilation strategies have evolved and improved, vastly decreasing the potential for ventilator induced lung injury (5-7). Affected newborns, for better or worse, are now stabilized for days or longer before surgical repair, and as a result the vast majority requiring ECMO arrive unrepaired, rather than post-repair as in the early years (8,9). Pulmonary vasodilators including nitric oxide and others have become ubiquitous in the management of CDH, although their effect on improving, or perhaps worsening CDH outcomes is similarly questioned (10-12). It is possible that
contemporary CDH patients requiring ECMO, as a result of better non-ECMO management, are more severe than in previous eras. It is also possible that, since the majority of CDH patients are now diagnosed prenatally, terminations of more severe CDH fetuses could be truncating the CDH severity spectrum (13). No data currently exist to clarify if the underlying severity of CDH patients has changed over time, but it is clear that the survival of those supported by ECMO, as tracked by the Extra-Corporeal Life Support Organization (ELSO), has not changed over the last decade and remains at just 50% (14). After 30 years of great effort, either the limitations of ECMO benefit in CDH have been reached, or there still remains significant opportunity for improvement.
CDH Spectrum of disease
In 1999, a seminal report from the CDH study group defined the relationship between CDH disease severity, and the survival benefit from ECMO. Compared to conventional therapy alone, ECMO support significantly improved survival for those CDH neonates with a predicted mortality greater than 80%, as defined by a logistic regression equation based on birthweight and 5- minute Apgar scores (15). In contrast, when ECMO was used on patients with less severe disease, ECMO exerted either an equivocal or negative effect on survival, with the negative effect increasing as disease severity decreased (Figure 1). Since support with ECMO adds risks to the CDH patient, such as bleeding from anticoagulation, these risks must be offset with greater benefit before additional CDH survival accrues. Utilizing ECMO in less severe patients increases mortality due to added risk without benefit, and only in the more severe patients is
the added risk offset by increased survival opportunity. Future opportunities for improving survival include minimizing ECMO in those patients on the less severe end of the severity spectrum, and also minimizing risks of ECMO related complications, ie. doing better ECMO, in those on the more severe end of the spectrum that require ECMO.
Understanding and defining CDH disease severity, therefore, is central to understanding the potential benefits of ECMO in CDH. Following the original CDH study group predicted survival equation based on 5-minute Apgar and birthweight, additional work on prenatal anatomic and postnatal physiologic markers of severity have added considerable granularity to our understanding of CDH disease severity (16-19). Of physiologic markers, birth physiology and early blood gases provide more discrimination of underlying disease severity than later blood gases or pre-ECMO physiology (20). This occurs because blood gases tend to improve over the first 12 – 24 hours of life for most CDH patients, and the levels of physiologic derangement immediately before ECMO are similar for the majority of CDH patients. This “homogenization” diminishes the power of those physiologic markers at those time points to differentiate the underlying levels of disease severity. In contrast, anatomic measures of CDH severity including prenatal lung measurements, liver position, and percent liver herniation, along with the postnatal measures of defect size and patch rate, are not affected by treatment factors or clinical course. Anatomic measures are therefore more useful for comparing disease severity between studies where treatment strategies may differ. Of these, thoracic liver position (liverup), especially for left CDH, has proven an important and relatively simple risk stratifier, correlating strongly with increased risk for ECMO and for mortality (19,21,22). A meta-analysis
of liver herniation in CDH showed survival rates of 74% with liver down CDH, which dropped to 45% in liver-up CDH (21). Finally, defect size correlates strongly with survival and risk for ECMO, and “patch rate” functions as a reasonable surrogate for larger defect sizes (23-25). In addition, outborn patients as a group represent a less severe cohort compared to inborn, as the most severe outborn patients do not survive birth and transfer, effectively pruning the severity spectrum. These proven correlations to CDH severity; inborn status, liver position, and patch rate can help inform population severity as we compare historical series.
Trials and Series
Randomized controlled data on the role of ECMO in CDH is limited to two early ECMO studies, and the UK ECMO trial (26-28). In the UK ECMO trial, randomization to conventional ventilation vs ECMO occurred at an oxygenation index of 40. Seventeen of 17 CDH infants randomized to conventional management died, while 4 of 18 in the ECMO arm survived (0% survival vs 22% survival). One of those survivors subsequently died at 2 years of age.
In 2006, Morini et al published a systematic review of ECMO in CDH, identifying 658 publications of which 21 (2043 patients) met entry criteria (29). Looking at the most rigorous of the non-randomized studies, they concluded ECMO use was associated with a reduction in CDH mortality. Zalla et al recently reviewed a single center CDH experience dividing 16 years of treatment into 4 eras, the latter 2 with ECMO availability. Post-hoc analysis suggested a 73%
reduction in risk of death in the ECMO eras compared to the pre-ECMO eras despite increases in CDH disease severity (p<0.001) (30).
Table 1 shows a compilation of single center publications since 2000 that show high survival, from centers reporting either low or higher ECMO usage, that were assembled for this publication. These reports were chosen because they document very good results, appear to have minimal selection bias, and provide some anatomic data for risk stratification. While all the results are excellent and the authors are to be congratulated, it appears notable that the highest survival rates and the highest severity, as reflected in the higher patch rates, were reported in the centers with higher ECMO use. While hardly definitive, these data support the concept that although the majority of CDH patients can be successfully supported without ECMO, those of higher to highest severity will have a survival advantage with well performed ECMO support.
Highest Severity Patients
The most compelling argument that ECMO can increase survival in CDH comes from the patients most severely affected, as defined in three contemporary reports. In 2012, Yoder et al looked at the highly severe subset of patients reported to the CDH Study Group patients that failed to achieve pre-ductal saturations of 85% in the first 24 hours, or before going onto ECMO(31). Twenty-three percent survived, and of those over 80% were treated with ECMO. In 2012 Stoffan reported seventeen patients with severe pulmonary hypoplasia as defined by
predicted lung volumes of less than 15%. Seven of 17 survived (41%) and all required treatment with ECMO (32). And in 2015, Kays et al used multi-variate modelling to define the worst 10% of 172 consecutive inborn, prenatally diagnosed CDH patients (33). Of the 19 worst, all of whom were aggressively resuscitated at birth and showed an average pH at 1 hour of life of 6.83 with a PCO2 great than 100, 10 of 16 ECMO-eligible patients survived to discharge (63%), while all 3 ECMO-ineligible prematures died. Fifteen of those 16 survivors were rescued with ECMO. While the reported use of ECMO never proves its necessity, these data are compelling evidence of the potential for ECMO to rescue patients at the highly severe end of the CDH spectrum, a capability not currently well documented for non-ECMO therapies.
Improving survival on ECMO
Considerable evidence suggests that CDH-ECMO survival results can be improved. Price and Stolar showed as early as 1991 that CDH patients treated with ECMO often died of potentially preventable treatment complications, not pulmonary hypoplasia (34). In a more recent report from 2013, 24 of 96 CDH –ECMO patients died (75% survival) but only 6 of the 24 deaths were attributed to pulmonary hypoplasia. The other 18 were ascribed to potentially preventable complications such as technical ECMO issues, intracranial bleeding, surgical bleeding, compartment syndrome, sepsis, and failure to achieve stability (19). Since end-stage pulmonary hypoplasia appears to account for a minority of deaths in CDH-ECMO patients, even in a series with 75% CDH-ECMO survival, the development and dissemination of best ECMO practices in this highly difficult population could lead to substantial survival improvements.
As an example, many authors have demonstrated the negative effect of ventilator induced lung injury on CDH survival. In 2011, Antonoff et al reported improved CDH-ECMO survival from 20% to 82% after instituting a treatment protocol centered on consistent lung protective ventilation. The institution of “gentle” ventilation strategies have increased ECMO survival and total survival in multiple series, and is a current mainstay of CDH management (5,6,35).
Veno-venous vs. Veno-arterial ECMO The choice of veno-venous (VV) versus veno-arterial (VA) ECMO in CDH has so far not been demonstrated to affect survival, but the current available data is poorly controlled for underlying disease severity. Dimmitt in 2001, and Guner et al in 2009 reviewed the cumulative ELSO experience on the topic, the latter covering 15 years from 1991 – 2006 (36,37). Guner et al found that VA ECMO was used 82% vs 18% of the time for VV. When controlling for various physiologic measures at the time of cannulation neither report demonstrated a difference in survival by ECMO delivery type in CDH, but Guner reported that 18% of VV ECMO patients required conversion to VA, with survival dropping from 54% (VV) to 44% (VV to VA) in those patients, compared to 50% when ECMO started as VA. Unfortunately, neither birth nor anatomic risk stratification was included. Kugelman et al reported no difference in survival with VV vs. VA (69% vs 68%) but the VV patients were larger (3.44 kg vs 2.77 kg, p<0.05) and less inborn (31 vs 58%, p=ns), both markers of less severity in the VV group (38). Neither liver position nor prenatal lung measurements were included in any of these reports. That VV patients were larger is not a surprise, as VV cannulas are larger requiring larger veins for safe
insertion. This issue is compounded in CDH patients, as they appear to have smaller jugular veins than their non-CDH neonatal ECMO counterparts (39).
Timing of CDH repair relative to ECMO
CDH patients supported with ECMO but never repaired do not survive. Repair confers a survival advantage, especially in the most severe. In the highly severe CDH patients reported by Yoder et al from the CDH study group that failed to attain a pre-ductal saturation of 85% in the first 24 hours of life or before ECMO, survival rose from 23% in the total to 44% in the subset where CDH repair was performed (31). However since ECMO adds significant risk and difficulty to repair, timing of repair relative to ECMO has become an important topic. Options include repair before ECMO, repair “early” during ECMO, repair “late” during ECMO, or repair after ECMO. Each strategy has its own risk-benefit profile, and its own proponents.
A 2009 study using CDH Study Group data showed that compared to repair on ECMO (44% survival), repair after ECMO showed a survival advantage (77% survival) (40). Others have provided supportive data to this important paper, and the strategy of repair after ECMO is currently widely practiced (40,41). The authors controlled for disease severity based on anatomic and physiologic measures and did not find statistically significant differences in the patients repaired on or off of ECMO. The authors did not, however, discuss the significant potential for selection bias inherent in comparing unrepaired CDH patients that had successfully weaned from ECMO, a major milestone, to patients that had not yet and might not
reach that critical level of recovery. Although the patient groups were not statistically different, those repaired on ECMO had lower 5 minute Apgar scores, more frequent need for patch repair, and a higher rate of “D” size defects (diaphragm agenesis), all suggestive of worse disease compared to those who successfully weaned from ECMO unrepaired. Additionally, the mean length of the ECMO run for those repaired after ECMO was 8.4 days, while the mean age of repair of those repaired while still on ECMO was 8.9 days, clarifying that as a group, those repaired on ECMO had not achieved the same degree of physiologic stability at the same time point as those discontinued from ECMO. While this study showed clearly that repair after ECMO works well for those CDH patients that successfully weaned from ECMO in an acceptable time frame, it did not address the fate of similar appearing patients that failed to wean from ECMO.
Unrepaired patients that fail to wean from ECMO either undergo repair late on ECMO, or never get repaired. Golden reported 66 CDH-ECMO patients of whom 33 were successfully weaned from ECMO unrepaired at a mean of 11.8 days of ECMO (41). However, twenty-two failed to wean and were repaired at a mean of 22 days on ECMO, and 11 died unrepaired after a mean of 16 days of ECMO. Eighty-five percent of those repaired after ECMO survived, but only 36% of those repaired late on ECMO survived, and 17% died on ECMO unrepaired. Overall CDH-ECMO survival in this study was 54%. Two additional single center studies defined an 11% and 15% mortality attributed to non-repair on ECMO (19,41). Together these studies highlight the potential fate of those unrepaired CDH patients that fail to wean from ECMO in a timely
fashion, and whose outcomes must be considered in a comprehensive evaluation of the strategy of CDH repair after ECMO.
To avoid deaths attributable to late or non-repair on ECMO, Dassinger et al performed “early” repair on ECMO, repairing 34 CDH on ECMO at an average of 55 hours after ECMO initiation. Only nine percent suffered bleeding complications requiring intervention, and a total of 22 (71%) survived. Fallon et al reported 46 CDH-ECMO patients. Survival was 73% for those repaired early on ECMO, 50% for those repaired late, and 64% for those repaired after ECMO. These data provide increasing evidence both for the benefit of repair, and of the potential advantage of early versus late repair in the ECMO run. To minimize the risk of operative bleeding on ECMO, Aminocaproic acid (Amicar) is used by many but not all (42,43) (44,45).
A different approach to avoid deaths due to non-repair on extracorporeal support is to identify and repair patients at high risk for needing ECMO, when feasible, before they decompensate enough to require ECMO, providing subsequent ECMO support after repair only when needed. Kays et al developed multivariate modeling employing prenatal lung measurements, liver position, birth physiology, and one hour blood gas data to predict subsequent risk of ECMO at one hour of life. They showed 95% survival in those left liver-up patients that were repaired early before ECMO (mean time to repair = 21 hours), compared to 65% survival achieved in an equivalent group of left liver-up patients that arrived to ECMO unrepaired (46). In the patients that arrived to ECMO unrepaired, 4 were repaired late when they failed to improve on ECMO and did not survive, while 3 developed organ failure on ECMO and died unrepaired. The authors
attributed their increased survival on a strategy that both assured repair in high-risk infants, and which also minimized bleeding risk by performing repair before the anticoagulation associated with ECMO.
Duration of ECMO
Although poor CDH survival beyond 2 weeks of ECMO has been reported (17), recent data suggests longer ECMO runs are needed for patients with more severe disease as defined by a variety of physiologic and anatomic stratifiers (19). In that series, survival of 43% of patients still on ECMO at 4 weeks was demonstrated. Based on these data, the authors recommended against arbitrarily defined durations of ECMO support.
ECMO Delivery systems
Rarely discussed but potentially important, especially in CDH-ECMO because of the longer length of ECMO runs, are the outcomes differences for infants supported with roller head ECMO systems versus centrifugal pump systems. In 2012, Barrett et al reviewed 1592 neonates that received ECMO, 163 by centrifugal pump and 1492 by roller head pump, and showed that those supported with centrifugal pumps had a nearly 8 fold risk of hemolysis, 21 fold risk of hyperbilirubinemia, 3 fold risk of hypertension, and greater than a 2 fold risk of renal failure compared to patients supported with roller head pumps (47). Although survival to discharge
wasn’t different between pump types, these effects could be magnified in the long runs of CDH infants, and this topic deserves monitoring.
Summary
There is wide center variation in CDH treatment and results, shown by Reickert et al. in 1998, and this remains the situation today (48). While many centers have shown very good survival employing minimal to no ECMO in CDH, the highest overall CDH survival rates reported in the literature and likely including the sickest patients, are reported by centers that utilize ECMO. Three of four centers shown in Table 1 with high overall survival rates also report ECMO survival rates in excess of 70%, substantially different from the ELSO reported survival of 50%. It is likely that these unsatisfying survival rates contribute to the ambivalence regarding benefit from ECMO in CDH. But since higher survival rates are possible, and ECMO has the unique potential to rescue CDH patients with profound pulmonary hypoplasia, the role of ECMO in severe CDH seems secure.
Optimal patient selection for ECMO in CDH requires refinement of non-ECMO support techniques so that this higher risk but higher potential reward modality is focused primarily on those patients with more severe CDH as defined by smaller lungs, worse birth physiology, liverup and with larger defects. Ventilatory support that strictly limits lung injury can improve ECMO outcomes, and the importance of operative repair in achieving survival of patients with CDH, especially of those on the more severe end of the spectrum cannot be overstated. While repair
after ECMO works well for those that successfully wean from ECMO, a strategy of early repair on ECMO, or possibly repair before ECMO, appears to provide a higher potential for overall survival, provided the repair can be done with acceptably low bleeding rates on support.
At this point 4 decades later, it seems unlikely that a prospective randomized study of ECMO support specific to CDH will be performed, highlighting the importance of the collective data provided by the CDH Study Group and that of well-reported single center experiences. Ideally future clinical CDH reports will include a full cadre of both early physiologic and anatomic markers, especially liver position, liver-up size, patch rate, and defect size, to support better risk assessment and comparison of outcomes between studies.
Finally, based on the data presented above and significant professional experience with CDH care, I feel confident that ECMO improves survival potential in more severe CDH over and above that currently available by non-ECMO techniques. Significant upside still exists for improving CDH outcomes, and since the majority of deaths occur in those more severely affected, improved ECMO outcomes will likely be central to realizing it.
Table 1
ECMO Reference
Year
Patients
Isol Surv
% ECMO
2004 & 2006
65
83%
(51)
2016
59
(52)
2001
Surv
% Inborn
% Patch
0%
nr
nr
73%
3%
100%
29%
81
80%
0%
49%
37%
2002
120
80%
13%
63%
56%
7%
(54)
2003
39
92%
36%
86%
72%
38%
(55)
2012
49
86%
29%
50%
nr
49%
(44)
2013
268
88%
40%
75%
72%
65%
Little or No ECMO
(49,50)
Low ECMO
(53)
Higher ECMO Use
nr = not reported Isol = Isolated Surv= survival
Table 1:
This table highlights single center high survival CDH reports, published since 2000, from centers utilizing either minimal or no ECMO, a low rate of ECMO, or higher (average) rates of ECMO.
References
1.
German JC, Gazzaniga AB, Amlie R, et al. Management of pulmonary insufficiency in diaphragmatic hernia using extracorporeal circulation with a membrane oxygenator (ECMO). J Pediatr Surg. 1977;12(6):905–12.
2.
Rais-Bahrami K, Van Meurs KP. Venoarterial versus venovenous ECMO for neonatal respiratory failure. Semin Perinatol. 2014;38(2):71–7.
3.
Harrington KP, Goldman AP. The role of extracorporeal membrane oxygenation in congenital diaphragmatic hernia. Semin Pediatr Surg. 2005;14(1):72–6.
4.
Morini F, Goldman A, Pierro A. Extracorporeal membrane oxygenation in infants with congenital diaphragmatic hernia: a systematic review of the evidence. Eur J Pediatr Surg. 2006;16(6):385–91.
5.
Wilson JM, Lund DP, Lillehei CW, et al. Congenital diaphragmatic hernia--a tale of two cities: the Boston experience. J Pediatr Surg. 1997;32(3):401–5.
6.
Kays DW, Langham MR Jr, Ledbetter DJ, et al. Detrimental effects of standard medical therapy in congenital diaphragmatic hernia. Ann Surg. 1999; 230(3):34051.
7.
Wung JT, James LS, Kilchevsky E, et al. Management of infants with severe respiratory failure and persistence of the fetal circulation, without hyperventilation. Pediatrics. 1985;76(4):488–94.
8.
Tsao K, Lally KP. Surgical management of the newborn with congenital diaphragmatic hernia. Fetal Diagn Ther. 2011;29(1):46–54.
9.
Moyer V, Moya F, Tibboel R, et al. Late versus early surgical correction for congenital diaphragmatic hernia in newborn infants. Moyer VA, editor. Cochrane Database Syst Rev; 2002;(3):CD001695.
10.
Malowitz JR, Hornik CP, Laughon MM, et al. Management Practice and Mortality for Infants with Congenital Diaphragmatic Hernia. Am J Perinatol. 2015 Jul;32(9):887–94.
11.
Hagadorn JI, Brownell EA, Herbst KW. Trends in treatment and in-hospital mortality for neonates with congenital diaphragmatic hernia. J Perinatol; 2015 Sep;35(9):748–54.
12.
Putnam LR, Tsao K, Morini F, et al. Evaluation of Variability in Inhaled Nitric Oxide Use and Pulmonary Hypertension in Patients With Congenital Diaphragmatic Hernia. JAMA Pediatr. 2016 1;170(12):1188-1194.
13.
Burgos CM, Frenckner B. Addressing the hidden mortality in CDH: A populationbased study. J Pediatr Surg. 2016 . In Press (epub ahead of print)
14.
ELSO International Summary. Extracorporeal Life Support Organization, January, 2017
15.
The Congenital Diaphragmatic Hernia Study Group. Does extracorporeal membrane oxygenation improve survival in neonates with congenital diaphragmatic hernia? J Pediatr Surg. 1999;34(5):720–4–discussion724–5.
16.
The Congenital Diaphragmatic Hernia Study Group. Estimating disease severity of congenital diaphragmatic hernia in the first 5 mi nutes of life. J Pediatr Surg 2001;36(1):141–5.
17.
Tiruvoipati R, Vinogradova Y, Faulkner G, et al. Predictors of outcome in patients with congenital diaphragmatic hernia requiring extracorporeal membrane oxygenation. J Pediatr Surg. 2007 Aug;42(8):1345–50.
18.
Seetharamaiah R, Younger JG, Bartlett RH, et al. Factors associated with survival in infants with congenital diaphragmatic hernia requiring extracorporeal membrane oxygenation: a report from the Congenital Diaphragmatic Hernia Study Group. J Pediatr Surg. 2009;44(7):1315–21.
19.
Kays DW, Islam S, Richards DS, et al. Extracorporeal life support in patients with congenital diaphragmatic hernia: how long should we treat? J Am Coll Surg. 2014;218(4):808–17.
20.
Khmour AY, Konduri GG, Sato TT, et al. Role of admission gas exchange measurement in predicting congenital diaphragmatic hernia survival in the era of gentle ventilation. J Pediatr Surg. 2014;49(8):1197–201.
21.
Mullassery D, Ba'ath ME, Jesudason EC,et al. Value of liver herniation in prediction of outcome in fetal congenital diaphragmatic hernia: a systematic review and meta-analysis. Ultrasound Obstet Gynecol. 2010 May;35(5):609–14.
22.
Lazar DA, Ruano R, Cass DL, et al. Defining “liver-up”: does the volume of liver herniation predict outcome for fetuses with isolated left-sided congenital diaphragmatic hernia? J Pediatr Surg. 2012;47(6):1058–62.
23.
Congenital Diaphragmatic Hernia Study Group, Morini F, Valfrè L, Capolupo I, Lally KP, Lally PA, et al. Congenital diaphragmatic hernia: defect size correlates with developmental defect. J Pediatr Surg. 2013 Jun;48(6):1177–82.
24.
Lally KP, Lasky RE, Lally PA, et al. Standardized reporting for congenital diaphragmatic hernia - An international consensus. J Pediatr Surg.; 2013 Dec;48(12):2408–15.
25.
Migliazza L, Bellan C, Alberti D, et al. Retrospective study of 111 cases of congenital diaphragmatic hernia treated with early high-frequency oscillatory ventilation and presurgical stabilization. J Pediatr Surg. 2007;42(9):1526–32.
26.
Bartlett RH, Gazzaniga AB, Toomasian J, et al. Extracorporeal membrane oxygenation (ECMO) in neonatal respiratory failure. 100 cases. Ann Surg. 1986;204(3):236–45.
27.
Schumacher RE, Roloff DW, Chapman R, et al. Extracorporeal membrane oxygenation in term newborns. A prospective cost-benefit analysis. ASAIO J. 1993;39(4):873–9.
28.
The UK Collaborative ECMO Trial Group. UK collaborative randomised trial of neonatal extracorporeal membrane oxygenation. The Lancet. 1996 ;348(9020):75–82.
29.
Morini F, Goldman A, Pierro A. Extracorporeal membrane oxygenation in infants with congenital diaphragmatic hernia: a systematic review of the evidence. Eur J Pediatr Surg. 2006;16(6):385–91.
30.
Zalla JM, Stoddard GJ, Yoder BA. Improved mortality rate for congenital diaphragmatic hernia in the modern era of management: 15 year experience in a single institution. J Pediatr Surg. 2015;50(4):524–7.
31.
Yoder BA, Lally PA, Congenital Diaphragmatic Hernia Study Group. Does a highest pre-ductal O(2) saturation <85% predict non-survival for congenital diaphragmatic hernia? J Perinatol. 2012;32(12):947–52.
32.
Stoffan AP, Wilson JM, Jennings RW, et al. Does the ex utero intrapartum treatment to extracorporeal membrane oxygenation procedure change outcomes for high-risk patients with congenital diaphragmatic hernia? J Pediatr Surg. 2012;47(6):1053–7.
33.
Kays DW, Islam S, Perkins JM, et al. Outcomes in the physiologically most severe congenital diaphragmatic hernia (CDH) patients: Whom should we treat? J
Pediatr Surg. 2015;14;50(6):893–7. 34.
Price MR, Galantowicz ME, Stolar CJ. Congenital diaphragmatic hernia, extracorporeal membrane oxygenation, and death: a spectrum of etiologies. J Pediatr Surg. 1991;26(9):1023–6–discussion1026–7.
35.
Antonoff MB, Hustead VA, Groth SS, et al. Protocolized management of infants with congenital diaphragmatic hernia: effect on survival. J Pediatr Surg. 2011;46(1):39–46.
36.
Dimmitt RA, Moss RL, Rhine WD, et al. Venoarterial versus venovenous extracorporeal membrane oxygenation in congenital diaphragmatic hernia: the Extracorporeal Life Support Organization Registry, 1990-1999. J Pediatr Surg. 2001;36(8):1199–204.
37.
Guner YS, Khemani RG, Qureshi FG, et al. Outcome analysis of neonates with congenital diaphragmatic hernia treated with venovenous vs venoarterial extracorporeal membrane oxygenation. J Pediatr Surg. 2009;44(9):1691–701.
38.
Kugelman A, Gangitano E, Pincros J, et al. Venovenous versus venoarterial extracorporeal membrane oxygenation in congenital diaphragmatic hernia. J Pediatr Surg. 2003;38(8):1131–6.
39.
Frenckner B, Palmér K, Lindén V. Neonates with congenital diaphragmatic hernia have smaller neck veins than other neonates-An alternative route for ECMO cannulation. J Pediatr Surg. 2002;37(6):906–8.
40.
Congenital Diaphragmatic Hernia Study Group, Bryner BS, West BT, et al. Congenital diaphragmatic hernia requiring extracorporeal membrane oxygenation: does timing of repair matter? J Pediatr Surg. 2009;44(6):1165–71– discussion1171–2.
41.
Golden J, Jones N, Zagory J, et al. Outcomes of congenital diaphragmatic hernia repair on extracorporeal life support. Pediatr Surg Int.; 2016 Nov 11;:1–7.
42.
Wilson JM, Wilson JM, Lund DP, et al. Congenital diaphragmatic hernia—A tale of two cities: The Boston experience. J Pediatr Surg. 1997;32(3):401–5.
43.
Keijzer R, Wilschut DE, Houmes RJ, et al. Congenital diaphragmatic hernia: to repair on or off extracorporeal membrane oxygenation? J Pediatr Surg. 2012;47(4):631–6.
44.
Kays DW, Islam S, Larson SD, et al. Long-term maturation of congenital diaphragmatic hernia treatment results: toward development of a severity-specific treatment algorithm. Ann Surg. 2013;258(4):638–44–discussion644–5.
45.
Prabhu S, Mattke AC, Anderson B, et al. Repair of congenital diaphragmatic hernia during extracorporeal life support: experience with six neonates. ANZ J Surg. 2016 Sep;86(9):711–6.
46.
Kays DW, Talbert JL, Islam S, et al. Improved Survival in Left Liver-Up Congenital Diaphragmatic Hernia by Early Repair Before Extracorporeal Membrane Oxygenation: Optimization of Patient Selection by Multivariate Risk Modeling. J Am Coll Surg. 2016;222(4):459–70.
47.
Barrett CS, Jaggers JJ, Cook EF, et al. Outcomes of neonates undergoing extracorporeal membrane oxygenation support using centrifugal versus roller blood pumps. Ann Thorac Surg. 2012;94(5):1635–41.
48.
Reickert CA, Hirschl RB, Atkinson JB, et al. Congenital diaphragmatic hernia survival and use of extracorporeal life support at selected level III nurseries with multimodality support. Surgery. 1998;123(3):305–10.
49.
Bagolan P, Casaccia G, Crescenzi F, et al. Impact of a current treatment protocol on outcome of high-risk congenital diaphragmatic hernia. J Pediatr Surg. 2004;39(3):313–8–discussion313–8.
50.
Casaccia G, Crescenzi F, Dotta A, et al. Birth weight and McGoon Index predict mortality in newborn infants with congenital diaphragmatic hernia. J Pediatr Surg. 2006;41(1):25–8–discussion25–8.
51.
Straňák Z, Krofta L, Haak LA, et al. Antenatal assessment of liver position, rather than lung-to-head ratio (LHR) or observed/expected LHR, is predictive of outcome in fetuses with isolated left-sided congenital diaphragmatic hernia. J Matern Fetal Neonatal Med; 2016 Mar 24:1–5.
52.
Al-Shanafey S, Giacomantonio M, Henteleff H. Congenital diaphragmatic hernia: experience without extracoporeal membrane oxygenation. Pediatr Surg Int.; 2002;18(1):28–31.
53.
Boloker J, Bateman DA, Wung JT, et al. Congenital diaphragmatic hernia in 120 infants treated consecutively with permissive hypercapnea/spontaneous
respiration/elective repair. J Pediatr Surg. 2002. 37(3):357-66. 54.
Downard CD, Jaksic T, Garza JJ, et al. Analysis of an improved survival rate for congenital diaphragmatic hernia. J Pediatr Surg. 2003;38(5):729–32.
55.
Guidry CA, Hranjec T, Rodgers BM, et al. Permissive hypercapnia in the management of congenital diaphragmatic hernia: our institutional experience. J Am Coll Surg. 2012;214(4):640–645–647.e1–discussion646–7.
From J Ped Surg 34: 720 -725, page 722