Cardiopulmonary Morbidity in Adults Born With Congenital Diaphragmatic Hernia

This prospective, nationwide, observational cohort study included adults born with CDH between 1989 and 2001, who had been treated in either of the 2 ECMO centers in the Netherlands (Erasmus MC Rotterdam, Radboudumc Nijmegen) with expertise in CDH since the 1990s. Recruitment occurred between October 2019 and May 2021. Standardized assessments were performed in Erasmus MC Rotterdam (Supplemental Information). Perinatal and demographic characteristics were retrieved from medical records and from participants themselves. This study was performed in compliance with the Code of Ethics of the World Medical Association (Declaration of Helsinki) and was approved by our institutional medical ethics review board (NL167096.078.18; study title: CDH-FU; approved: January 11, 2019). All participants signed informed consent.

Details on baseline data, equipment, and methodology are provided in Supplemental Information.

Height (centimeters) and weight (kilograms) were measured, and the BMI calculated. Body composition was assessed by air displacement plethysmography using whole-body densitometry (BOD POD).

Spirometry was performed before and after 400 µg inhaled salbutamol. Airway patency (forced expiratory volume in 1 second [FEV₁]), forced vital capacity (FVC), FEV_1/FVC, and maximal midexpiratory flow rate were measured with an electronic spirometer; total lung capacity and inspiratory vital capacity (VCIN) by whole body plethysmography; and diffusion capacity using the single-breath method. Spirometry data are expressed as SD scores (SDS) on sex-, age-, and height-related references values.¹¹ 

The participant’s anonymized end-inspiratory and -expiratory chest CT scans were reviewed by a thoracic radiologist (P.C.) unaware of medical history other than CDH. The following findings were recorded: (1) skeletal deformities, (2) architectural lung distortion on end-inspiratory images, (3) low attenuation areas on the end-expiratory images, (4) fibrotic rest abnormalities, and (5) recurrence of diaphragmatic defect and its contents (herniating organs).

CPET was performed on an electronically braked cycle. We recorded for analysis: Peak workload and peak oxygen (O₂) consumption (VO_2peak) per kilogram calculated as percentage of predicted, peak workload per bodyweight, absolute VO_2peak and VO_2peak per body weight, breathing reserve, ventilatory anaerobic threshold, and O₂ pulse.¹²  CPET results were included of participants with a peak respiratory exchange ratio (RER) >1.1 and peak heart rate >80% of predicted.

On echocardiography, pulmonary hypertension (PH) was defined as tricuspid regurgitation (TR)jet >2.8 m per second using the updated classification of PH or, in the absence of TR, as a shift in the interventricular septum position.¹³  The right ventricular systolic pressure (RVSP) was estimated using the Bernoulli formula (TRjet 2 x 4). On the basis of the position of the interventricular septum, right ventricular dilation, or RVSP, PH was classified as absent, mild, or severe.

Continuous data are expressed as median (range) or mean (SD), and categorical data are expressed as counts and percentages. Group comparisons (participants versus nonparticipants, Rotterdam versus Nijmegen, and ECMO treatment versus non-ECMO treatment) of baseline characteristics were evaluated using Mann-Whitney U tests for continuous variables and χ² tests for categorical variables. To compare lung function and CPET parameters with reference data (SDS = 0 of 100%), we used a 1 sample t test, and between ECMO treatment and non-ECMO treatment, we used an independent samples t test. A χ² test served to evaluate differences in pulmonary imaging and in cardiac evaluation between groups. To determine whether particular baseline characteristics had a significant influence on SDS FVC after bronchodilation, SDS FEV₁ after bronchodilation, SDS VCIN-single-breath method, SDS transfer factor for carbon monoxide_, and VO_2peak per kilogram as percentage of predicted, we performed multivariable linear regression analyses with the following characteristics: Type of operation (primary or patch), use of ECMO, birth weight (grams) and time on ventilator (days). Analyses were performed using SPSS 28.0 (IBM, Chicago, IL), using a 2-sided significance level of 0.05. No correction for multiple testing was performed.

Fifty-four of 122 eligible patients declined participation. Thus, 68 participants (56% of eligible) were included (Fig 1). Because of missing data in Nijmegen, baseline data of participants and nonparticipants could be compared only for the Rotterdam group; this did not reveal significant differences (data not shown). The participants’ baseline data did not differ between both centers (Supplemental Information, Supplemental Table 5). Altogether, 26 participants (38%) had been treated with ECMO; these had significantly more often a right-sided defect (P = .026), patch closure (P < .001), and longer duration of ventilation (P < .001) than the others. Background characteristics, current treatment (bronchodilators, inhaled steroids), self-reported gastroesophageal reflux, and smoking habits at follow-up did not significantly differ between the 2 groups, but a history of surgical correction of thoracic deformities (pectus excavatum or carinatum) did (P = .046). Moreover, ECMO-treated men had lower body weight and height than the non-ECMO–treated men (P = .028 and .003, respectively; Table 1).

Lung function testing, pulmonary imaging, and CPET were performed in all participants. For organizational reasons or quality issues, these data were missing: BOD POD n = 7, echocardiography n = 5.

The height of 14 of 30 women and 21 of 38 men was shorter than the average population height.¹⁴  ECMO-treated men had significantly lower height and weight than those without ECMO (Table 1). BMI was 20 to 25 in 36 participants (47% of men and 60% of women, respectively); BMI was <20 in 8 men and 7 women (21% and 23%, respectively) and >25 in 12 men and 5 women (32% and 17%, respectively). Median (range) fat mass in women was 29% (16–47), and in men was 19% (6–45).

SDS FVC, FEV₁, FEV₁/FVC, maximal and midexpiratory flow rate, total lung capacity, VCIN, transfer factor for carbon monoxide, and diffusion capacity corrected for alveolar volume were significantly below normal (Supplemental Information, Supplemental Table 6). All lung function parameters in the ECMO-treated group were significantly lower than those in the non-ECMO–treated group (Table 2). In 8 participants (12%), airflow obstruction was considered reversible, 4 of whom underwent ECMO.

All chest CT scans had good to excellent image quality. ECMO-treated participants all had architectural distortion on end-inspiration, expiratory low attenuation, and linear fibrotic rest abnormalities (Table 3; Supplemental Fig 3). A recurrence of the diaphragmatic defect was seen in 38 participants (56%), of whom 20 had undergone ECMO (77%) and 18 had not (43%). In 27 of 38 participants with a recurrence (71%), a patch was used for initial CDH repair. The herniated content varied from fatty tissue to solid and hollow organs (Fig 2). During the study period, 2 participants with CDH recurrence underwent redo surgery for incarceration; in both cases, a patch was applied. The remaining 36 participants remained asymptomatic.

Four participants were excluded because they did not reach maximal exercise capacity: 2 had saturation <88% before or during CPET, on account of which CPET was stopped, and the other 2 did not reach an RER >1.1 by subjective maximal exercising. Two nonexcluded participants had saturation of 85% at the end of the CPET.

Parameters of exercise capacity were analyzed separately for men and women, except for parameters calculated as percentage of predicted. The non-ECMO–treated group showed workload as percentage of predicted within the normal range, whereas these were below normal in the ECMO-treatment group. VO_2peak as percentage of predicted was below normal in both groups, and lower in the ECMO-treated group. The absolute VO_2peak and VO_2peak per body weight was lower in the ECMO-treated group compared with the non-ECMO–treated group. Breathing reserve was below normal range only in men treated with ECMO. O₂ pulse was below normal in men treated with ECMO and in women in both groups. Table 4 shows exercise parameters for both groups.

None of the participants had signs of PH. TR was present in 30 participants and occurred irrespective of ECMO treatment (48% and 45%). RVSP was significantly higher in the ECMO-treated group (25 mmHg versus 18 mmHg, P < .001, Supplemental Table 7).

To the best of our knowledge, we are the first to report results of cardiopulmonary assessment in a large cohort of adult CDH patients born after introduction of ECMO. Previously published data on cardiopulmonary morbidities in CDH mainly related to children or to adults born before ECMO became available. Overall, we found airflow obstruction, reduced lung volumes, reduced diffusion capacity, and reduced exercise capacity, with worse outcomes in the ECMO-treated population. Many of the ECMO-treated and non-ECMO–treated participants had structural abnormalities on chest CT, including usculoskeletal deformities, low attenuation regions (either air trapping or emphysema), and recurrence of the diaphragmatic defect. Interestingly, evaluation at rest revealed no signs of PH. After the assessments, we referred participants to a dedicated pulmonary outpatient clinic. So far, 21 participants have attended this clinic; 13 were successfully treated with inhalers.

In childhood, ECMO-treated CDH patients are likely to have more growth problems,¹⁵  persistent airflow obstruction,¹⁶  reduced exercise tolerance,¹⁰  and motor function problems¹⁷  than those who did not need ECMO. Although in the current study the men in the ECMO-treated group were significantly shorter and thinner, our data are inconclusive on persistency of growth retardation after ECMO-treatment.

In a pre-ECMO cohort, Spoel and coworkers found a mild deterioration of airflow obstruction and diffusion capacity from childhood into adulthood⁷ ; their spirometry and diffusion capacity results are in line with ours in the non-ECMO–treated group. Toussaint-Duyster and coworkers found that ECMO-treated children born with CDH had more severe airflow obstruction than those without ECMO, although the slope of deterioration between 8 and 12 years was similar in groups.¹⁶  Dao and coworkers found a continuous decrease in FEV₁, FVC, and FEV₁/FVC from ages 5 to 20 years in 7 participants who had reached that age.¹⁸  Indicators of CDH severity, such as defect size, were associated with the baseline lung function results but had little effect on the decline, suggesting that, even in the less-severe cases, lung function declined.¹⁸  Because ECMO treatment has been associated with longer need for ventilatory support, these combined factors may have resulted in a tendency to more severe respiratory morbidity. In line with Dao and coworkers,¹⁸  we showed irreversible airflow obstruction in many participants. Bronchopulmonary dysplasia (BPD) in prematurely born children has been likened to the impaired lung development in CDH,¹⁹  but so far has not been proven to be associated with lung function decline over time.^20,21  Ongoing changes in the lung morphology and physiology might lead to a further decline of lung function in CDH, which therefore should be considered an entity of its own.

Most of the centers that treat CDH patients routinely perform chest x-rays to assess recurrence and lung morphology.²²  Data on the use of CT scans in childhood are sparse and absent in adults, although CT scans are superior to chest x-rays when it comes to visualizing lung abnormalities and diaphragm changes.^23–25  Recurrence of the diaphragmatic defect has been described in 9% to 32% of children between the ages of 1 and 12 years.^24,25  Our study is the first reporting recurrence in adults assessed with CT. Even when considering herniation of fatty tissue being a local relaxation of the native diaphragm and therefore not clinically significant, the prevalence of “true” recurrence was up to 37% (25 of 68). We only performed chest CT scans, but we assume that the recurrence would have been missed on x-ray because of the mainly posterior location. Only 2 participants with a radiologic recurrence (5%) were symptomatic. The remaining 36 participants with a recurrence were advised to seek medical counseling in cases of early warning symptoms such as unexplained abdominal or thoracic pain. It is debatable whether an extensive elective reoperation should be recommended to avoid potential incarceration or to enhance pulmonary capacity. The clinical consequences of recurrence later in life (eg, in pregnant women) are currently unknown. Pregnancies had not yet occurred in our cohort.

Our data do not allow speculating whether defect recurrence affects lung function. All ECMO-treated, and the majority of non-ECMO–treated, participants showed structural lung abnormalities (architectural distortion and low attenuation areas) comparable to those in BPD.^26,27  In our cohort, low attenuation areas, including low-density parenchyma both from air-trapping to emphysema, were mainly classified as mild to moderate, whereas fibrotic rest abnormalities were less pronounced and classified as mild. The latter observation seemed surprising because postinflammatory fibrosis could be anticipated after artificial ventilation and exogenous O₂ therapy. Our data do not allow for detailed comparison of the parenchymal changes in CDH with previously performed CT scans in BPD. To further elucidate the morphologic changes in chest CT scans, we are currently performing more sophisticated digital analyses.

Toussaint-Duyster and coworkers reported a decline in maximal exercise endurance time, irrespective of ECMO treatment, in children between the ages of 5 and 12 years.¹⁶  Van der Cammen-van Zijp and coworkers studied the exercise capacity of young adults born with CDH between 1975 and 1986.²⁸  That cohort seemed to have had a milder clinical course than our cohort with limited ventilatory support; data of VO_2peak in the non-ECMO group in our cohort were comparable to the data of Van der Cammen-van Zijp.²⁸  Especially, ECMO-treated CDH patients with more severe cardiopulmonary morbidity may benefit from tailormade lifestyle and exercise programs.²⁹ 

Only 1 participant in our study had a cardiac malformation. We therefore focused on evaluation of PH, which usually resolves within the first weeks of life and has not been seen in further follow-up of children born with CDH.³⁰  PH with poor overall outcome is more prevalent in adults born preterm and in those suffering from chronic obstructive lung disease.³¹  PH in chronic obstructive lung disease may further increase in sleep and during exercise.³²  In our cohort, PH was absent during noninvasive measurements at rest. Considering the impaired lung function and reduced exercise tolerance in our cohort, we wondered whether they would develop PH during exercise and states of hypoxia. O₂ pulse in CPET is thought to indirectly reflect stroke volume.¹²  In our cohort, O₂ pulse was lower than normal in men treated with ECMO and in women in both groups, which raises the question whether they developed PH during CPET. Furthermore, CDH patients might experience states of hypoxia while traveling by air, as it has been described in young CDH survivors.³³  Therefore, a fit-to-fly test and, if necessary, additional O₂ should be considered in more severely affected CDH patients.

Because baseline data from nonparticipants initially treated in Nijmegen and the rate of nonparticipants was 50%, we cannot fully exclude bias. Still, its probability seems to be negligible because baseline data from the participants from both centers were comparable. All participants were born before the currently used standardized classification for defect sizes was implemented.³⁴  CPET data were subdivided according to sex, resulting in relatively small numbers of participants per subgroup.

In current clinical practice, most adults born with CDH will not receive structural follow-up. The results of this study indicate that cardiopulmonary morbidity and CDH recurrence can be expected in all young adult CDH patients, and that outcomes may be worse in the most severely affected individuals (eg, reflected by the need for ECMO treatment). To prevent further deterioration of cardiopulmonary morbidity, active counseling on healthy lifestyle, frequent exercise, and a strict nonsmoking policy should be offered. Medical treatment of airflow obstruction might be beneficial in daily life activities, during exacerbations, or during physical exercise. All CDH patients should be counseled on the possibility and risks of recurrence. A low-dose chest CT scan could be performed to screen for significant CDH recurrence and as baseline for monitoring of low attenuation regions at age 18 years. A specific preconception counseling policy in all female CDH survivors could prevent unexpected complications in those with recurrence and/or severely impaired lung function. Cardiac evaluation might be restricted to patients with severely impaired lung function and reduced exercise tolerance. The latter group may need a fit-to-fly test before long-distance flying.

We thank Joke Dunk, Ellen van Deutekom, Arno van Heijst, Sanne Botden, and all members of the lung function and cardiac laboratory for their contributions. Ko Hagoort provided editorial advice.

BPD

bronchopulmonary dysplasia

CDH

congenital diaphragmatic hernia

CPET

cardiopulmonary exercise testing

CT

computed tomography

ECMO

extracorporeal membrane oxygenation

FEV₁

forced expiratory volume in 1 second

FVC

forced vital capacity

O₂

oxygen

PH

pulmonary hypertension

RER

respiratory exchange ratio

RVSP

right ventricular systolic pressure

SDS

SD scores

TR

tricuspid regurgitation

VCIN

inspiratory vital capacity

VO_2peak

peak oxygen consumption