Repaired oesophageal atresia: respiratory morbidity and pulmonary function in adults
- S. Sistonen*#,
- P. Malmberg¶,
- K. Malmström¶,
- T. Haahtela¶,
- S. Sarna+,
- R.J. Rintala* and
- M.P. Pakarinen*
- *Section of Paediatric Surgery, Hospital for Children and Adolescents
- ¶Dept of Allergy, Helsinki University Central Hospital
- +Dept of Public Health, University of Helsinki, Helsinki
- #Dept of Surgery, Päijät-Häme Central Hospital, Lahti, Finland
- S. Sistonen, Section of Paediatric Surgery, Hospital for Children and Adolescents, Helsinki University Central Hospital, P.O. Box 281, FI-00029-HUS, Finland. E-mail: saara.sistonen{at}helsinki.fi.
Abstract
Although after oesophageal atresia (OA) repair in infancy, respiratory problems are common, their natural history remains unclear. We assessed morbidity, pulmonary function (PF), and bronchial hyperresponsiveness (BHR) in adults with repaired OA respiratory.
588 patients who underwent surgery for OA during 1947–1985 were identified and those 262 who were alive and had their native oesophagus were included. Respiratory symptoms and respiratory symptom-related quality of life (RSRQoL) were assessed by questionnaire and interview, and the patients underwent spirometry, a histamine challenge test, and an exhaled nitric oxide test. For the questionnaires, we added 287 carefully matched general population-derived controls.
Among the 101 (58 male) patients, median age 36 yrs (range 22–56 yrs), respiratory morbidity was significantly increased compared to controls. Patients had more respiratory symptoms and infections, as well as asthma and allergies, and more often impaired RSRQoL (p<0.001 for all). PF tests revealed restrictive ventilatory defect in 21 (21%) patients, obstructive ventilatroy defect in 21 (21%) patients, and both in 36 (36%) patients. A total of 41 (41%) had BHR, and in 15 (15%), it was consistent with asthma. The most significant risk factors for restrictive ventilatory defect were thoracotomy-induced rib fusions (OR 3.4, 95% CI 1.3–8.7; p = 0.01) and oesophageal epithelial metaplasia (OR 3.0, 95% CI 1.0–8.9; p = 0.05).
After repair of OA, respiratory-related morbidity, restrictive ventilatory defect and BHR extended into adulthood. Nearly half the patients had BHR and over half had a restrictive ventilatory defect. Thoracotomy-induced rib fusions and gastro-oesophageal reflux-associated oesophageal epithelial metaplasia were the strongest risk factors for restrictive ventilatory defect.
- Atopy
- bronchial hyperresponsiveness
- long-term outcome
- oesophageal atresia
- pulmonary function
- tracheo-oesophageal fistula
The most common congenital anomaly of the oesophagus is oesophageal atresia (OA). It is usually associated with a tracheo-oesophageal fistula (TOF) 1, 2 and affects approximately one in 2,500 live births 3. The embryology of OA is incompletely understood 4, and over half these infants have associated anomalies 3, 5. The malformation itself and surgical trauma result in impaired oesophageal motility, a dysmotility predisposing to gastro-oesophageal reflux (GOR) and its complications 6, 7. A strong connection exists between severity of GOR and persistence of respiratory symptoms among OA survivors 4, but respiratory problems become less frequent with time.
After repair of OA, approximately one-third of these patients suffer respiratory symptoms in adolescence and adulthood 8–12. Prevalence of wheeze ranges 26–43% 10, 13, 14, and that of doctor-diagnosed asthma (DDA) 12–29% 9, 10, 12, 14, 15. This is generally higher than in children (8.8%) 16 and adults (6%) 17. Additionally, pulmonary function (PF) abnormalities, mostly mild restrictive disease 14, 18, 19, occur after repair of OA in childhood 14, adolescence 9, 18, 20, 21 and adulthood 12, 19, 21. Restriction may result from surgical trauma, repeated aspiration or recurrent chest infections, as well as from associated thoracic musculoskeletal defects, such as postoperative rib fusions, scoliosis or other chest deformities.
Although after repair of OA, respiratory symptoms are common, the long-term outcome of PF and respiratory morbidity remain unclear. As growing numbers of OA survivors are reaching adulthood, interest in their long-term outcome is increasing. To this end, we performed a population-based, long-term follow-up study in adults with repaired OA, assessing respiratory morbidity, respiratory symptom-related quality of life (RSRQoL), PF, bronchial responsiveness and atopy. The predictive factors for restrictive ventilation in adulthood were also studied.
PATIENTS AND METHODS
The hospital records of all 588 patients treated for OA at the Hospital for Children and Adolescents, Helsinki University Central Hospital (Helsinki, Finland) between 1947–1985 were reviewed. Vital status and postal addresses came from the database of the Population Register Centre of Finland based on a personal identification code given to all residents of Finland. Of the 588 infants who underwent surgery for OA, 296 were alive at the beginning of the study in November 2005. The 34 who had undergone oesophageal replacement were excluded. 16 patients had emigrated, and the postal addresses of 11 were unknown. The remaining 235 (90%) eligible survivors with a native oesophagus were contacted by mail. Letters described the study protocol and purpose. Those returning signed consent forms received invitations to enter the study. Data drawn from the hospital records included survival, type of OA, associated anomalies, surgical treatment and complications.
During their outpatient visit, each participant filled in a symptom questionnaire and underwent an interview, clinical assessment, flow–volume spirometry for PF, histamine challenge test (HCT) for bronchial responsiveness, exhaled nitric oxide fraction (FeNO) for airway inflammation, skin prick test (SPT) for common allergens, and measurement of serum level of immunoglobulin (Ig)E.
The present study is part of a larger evaluation of the long-term outcomes of OA. Findings concerning oesophageal morbidity and function 22, incidence of cancer 23, and natural history of spinal anomalies and scoliosis 24 among adults with repaired OA have already been published. GOR symptoms occurred in 34 of the study participants, and 85 had symptoms of dysphagia due to oesophageal dysmotility after repair of OA; these occurrences were significantly more frequent than among the general population-derived controls 22. A total of 21 adults with repaired OA had oesophageal epithelial metaplasia in their oesophageal biopsy samples 22. In addition, chest wall deformities were noted in 15 study participants 24.
Interview and questionnaires
All participants were interviewed by the same investigator with a structured form, providing information on medical history, respiratory symptoms and diseases, symptoms of allergy, need for asthma medication, and smoking habits. Symptoms of respiration included wheeze, attacks of shortness of breath, recurrent cough and respiratory infections. Symptoms of allergy included allergic rhinitis, conjunctivitis, rash and atopic eczema. The questionnaire was a modification of validated questionnaires on asthma and allergy symptoms 25 as well as on RSRQoL 26. The identical symptom questionnaire was sent to a group of 915 controls randomly chosen from the Population Register Centre of Finland, matched for age, sex and municipality of residence.
Pulmonary function, histamine provocation and exhaled nitric oxide
Flow–volume spirometry was performed according to the guidelines of the European Respiratory Society (ERS) 27, and bronchial hyperresponsiveness (BHR) was estimated using a validated dosimetric HCT 28. Forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1) were recorded, as well as their reference value percentages. We used national reference values for spirometry published previously 29. Ventilatory function was defined as restrictive when FVC was <80%, which corresponds to a z-score of -2.0, and obstructive when FEV1/FVC was <87% predicted, which also corresponds to a z-score of -2.0. Defects were classified as mild (z-score -2.0– -3.5), moderate (z-score -3.5– -5.5), or severe (z-score -5.5– -7.5). Through use of the dose–response curve, the provocative dose of inhaled histamine producing a decrease of 15% in FEV1 (PD15) could be determined, and, based on bronchial hyperreactivity, was graded as mild (0.41–1.60 mg), moderate (0.11–0.40 mg) or severe (≤0.1 mg) 30. Patients with PD15 ≤1.6 mg were considered to have increased bronchial responsiveness. Previous study has shown that moderate to severe BHR (PD15 ≤0.40 mg) has a high specificity for asthma 28.
FeNO was measured using computerised equipment with a chemiluminescence analyser (Niox; Aerocrine AB, Sweden), according to American Thoracic Society (ATS) recommendations 30. The patients were seated, wore no nose clip, and were asked to fill their lungs completely with NO-free air, and thereafter to exhale with a mean±10% and instantaneous flow of 50±5 mL·s−1. An exhalation time of 10 s served as the default. The system was calibrated with a certified NO calibration gas mixture according to manufacturer's instructions. Exhalations not meeting ATS requirements were rejected by the system, and the patient was asked to perform new exhalation manoeuvres, until providing three reproducible FeNO values within 10% of each other. The mean FeNO of three acceptable end-expiratory plateau measurements was calculated and recorded 31, 32. Previously specified reference values for adult lifelong never-smokers, according to the distribution of FeNO by similar technique, and exhalation rate depending on age and height, were used 33.
Atopy
Atopy was defined by a positive SPT to common allergens. To evaluate sensitisation to common allergens, SPTs were carried out for the following common allergens: birch, grass, mugwort, cat, dog, cow, horse, Cladosporium herbarum, and Dermatophagoides pteronyssinus (Soluprick SQ; ALK, Horsholm, Denmark). The patient was regarded as atopic if one or more reactions were positive, determined as a wheal diameter of ≥3 mm in the presence of an expected response to control solutions; the positive control was histamine dihydrochloride (10 mg·mL−1) and as the negative control was saline. The level of serum IgE was determined based on the standard values of our laboratory, 0–110 kU·L−1 for adults.
Statistical analysis
Data was analysed using SPSS 16.0 (SPSS Inc., Chicago, IL, USA). Statistical analysis used the Chi-squared and Kruskal–Wallis tests. Data are given as frequencies or as mean and range. The multivariate logistic regression model for occurrence of restrictive ventilation defect, included age, any other associated anomaly, rethoracotomies, thoracotomy-induced rib fusions, oesophageal anastomotic leakage, recurrent TOF, primary pulmonary complications, oesophageal epithelial metaplasia and smoking. A study of the urban adult population of Helsinki served as a comparison group for SPTs 34. Level of significance was set at p<0.05.
Ethics
The ethics committee of the Hospital for Children and Adolescents, Helsinki University Central Hospital approved this study. Written informed consent was obtained from each participant.
RESULTS
Subjects
A total of 72% of the 235 subjects contacted replied, and the first 101 (43%) patients who returned the signed informed consent form were invited to enter the study (fig. 1). Most of those participating (91%) had OA with distal TOF. Because patients who had undergone oesophageal replacement were excluded, each patient had a native oesophagus. Of the 101 participating patients, 95 had undergone primary oesophageal anastomosis, two had undergone a staged anastomosis, and three had undergone a closure of an isolated H-type fistula. Myotomy of the proximal oesophageal pouch to enable primary anastomosis was employed in five subjects. Of the 101 participants, 58 were male, and the mean follow-up time was 36 yrs (range 21–57 yrs) (table 1). Participants and nonparticipants were comparable in terms of age, sex, type of OA, associated anomalies and surgical complications (p>0.05). A total of 287 (31%) of the controls responded to the mail questionnaires (table 1).
Tracheopulmonary morbidity in infancy
Of the patients, 34 had suffered pneumonia as a primary pulmonary complication and 32 needed no postoperative ventilation, with the mean postoperative ventilation time being 2.6 days (range 0–7 days) (table 2). During childhood, tracheomalacia was diagnosed in 15 of the patients, but only one had undergone aortopexy. A total of 10 patients had recurrent TOF.
Questionnaires
Participants reported increased incidence of allergy, pneumonia, bronchitis, repeated childhood infections, persistent cough and DDA in relation to controls (p<0.001 for all). Among the patients, RSRQoL was significantly decreased (p<0.001). Occurrence of tracheomalacia in infancy was associated with neither current respiratory symptoms nor with tendency to respiratory infections (p>0.05) (table 3).
Pulmonary function, histamine provocation and exhaled nitric oxide
In flow–volume spirometry, 21 patients had restriction only, 21 had only obstruction only and 36 had both. Accordingly, 57% of the patients had restrictive ventilatory defect. The PF test (PFT) results were similar in males and females (p>0.05) (table 4). A total of 41 had BHR. BHR was mild in 26 patients, moderate in 11 and severe in four; 15 patients had moderate to severe BHR compatible with asthma. BHR was associated with atopy (p<0.007) and with current respiratory symptoms (p<0.03). FeNO was clearly elevated only in four patients, and slightly elevated in seven of the patients, but was associated neither with BHR, atopy, current respiratory symptoms, nor PF abnormalities.
Atopy
37 patients had at least one positive reaction to common allergens in SPT, and sensitisation to multiple allergens was evident in 22. Positive SPT reactions to the most common allergens were dog (28%), cat (22%), birch (23%), mugwort (22%), and timothy grass (21%). Elevated levels of serum IgE >110 kU·L−1 occurred in 20 patients (median 285 kU·L−1, range 126–2,436 kU·L−1). Elevated IgE was associated with multiple sensitisations (p<0.001), current respiratory symptoms (p = 0.015) and BHR (p<0.007), whereas multiple sensitisations were associated with current respiratory symptoms (p = 0.008).
Predictors of restrictive ventilation
Restrictive ventilation defect was only associated with the occurrence of oesophageal epithelial metaplasia in adulthood (p = 0.038), but not with tracheomalacia in infancy, history of primary pneumonia, age, airway inflammation in adulthood, or current respiratory symptoms (p>0.05 for each). In a multivariate logistic regression analysis, thoracotomy-induced rib fusions (OR 3.4, 95% CI 1.3–8.7; p = 0.01) and GOR-associated oesophageal epithelial metaplasia in adulthood (OR 3.0, 95% CI 1.0–8.9; p = 0.05) were the most significant risk factors for restrictive ventilation defect. Neither increasing age, sex, other associated anomalies, primary lung complications, primary surgical complications, nor number of thoracotomies, predicted the incidence of restriction.
DISCUSSION
To the best of our knowledge, this is the largest and only population-based study on respiratory morbidity and PF in adults with repaired OA. We used validated questionnaires with controls derived from the general population that were carefully matched for age, sex, and municipalities of residency. A drop-out analysis between participants and nonparticipants suggested that the two groups were comparable with respect to primary disease characteristics, minimising the risk of selection bias. We found an increased incidence of daily respiratory symptoms and asthma, as well as decreased RSRQoL. PF abnormalities were detectable in 78% of the patients: restriction in 57%, obstruction in 55% and both these ventilation defects in 36%. Occurrence of restrictive PF after repair of OA was associated with thoracotomy-induced rib fusion, which was present in one-third, and GOR-associated oesophageal epithelial metaplasia in one-fifth of the patients. However, the restrictive ventilation defect showed no association with current respiratory symptoms.
Environmental factors, such as cold climate, and inhaled allergens and irritants, have a substantial effect on respiratory symptoms among the general population 35. In Finland, prevalence of DDA among children is 8.8% 16 and is 6% among adults 17, whereas in patients with OA it is even higher 9, 10, 12, 14, 15. In the present study, 16% of the adult patients with repaired OA and 6% of the general population-derived controls have DDA. These figures are in line with other findings 9, 10, 12, 14, 15. Self-reported wheeze occurred in 37% of the patients and in 30% of the controls. Self-reported wheeze usually occurs in ∼37% of the survivors of OA with no tendency toward improvement with age 4, 10, 11, 13, 14.
In the present study, occurrence of respiratory symptoms was clearly lower among the adult study population than in children and adolescents with repaired OA 9, 10, 12, suggesting that prevalence of respiratory problems decreases with age. Nevertheless, decreased RSRQoL and daily respiratory symptoms were still significantly more common among patients than among controls. Incidences of asthma and BHR were in line with previous results, whereas the PF abnormalities were more common among our adults than in children and adolescents 9, 14, 18, 20, 21, 36, in accordance with other findings among adults 19.
The proportion of children and adolescents with repaired OA having a restrictive PF defect has ranged 21–40%, and 16–54% with obstructive ventilation defect 9, 18, 20, 21, 36. Our respective figures were 57% and 55%. PF abnormalities have not shown a correlation with respiratory or oesophageal symptoms 9, 21, 36. Although a clear link exists between severity of GOR and persistence of respiratory symptoms among OA survivors 4, the reason for PF abnormalities is unclear. They have been suggested to result from recurrent aspiration 21, from poor tracheal clearance leading to recurrent episodes of bronchitis and pneumonia causing lung damage 15, 36, or from poor lung growth during infancy 18. In the present study, thoracotomy-induced rib fusion and GOR-associated epithelial metaplasia of the oesophagus were significant predictors of restrictive ventilation defect. Thus, epithelial metaplasia served as a surrogate marker for significant GOR, which may result in repeated aspiration and lung damage. However, rib fusion leads to an immobile and thus restrictive thoracic cage.
Prevalence of BHR in the general Finnish population with no previous diagnosis of asthma or chronic bronchitis and with normal lung volumes is 17% 37. In our study population, BHR occurred markedly more frequently in 41%, which is in accordance with the results of other studies 9, 18, 20, 36. It has been postulated that increased bronchial reactivity in these patients would merely reflect sequalae of chronic lung disease from damaged epithelium in the airways caused by recurrent aspiration of acidic gastric contents 18. Severe or moderate BHR has been associated with a more restrictive ventilatory defect among adolescents 9, but no correlation has emerged between increased BHR and history of asthma or atopic eczema 9, 36. In the present study, BHR was associated with current respiratory symptoms and atopy. Atopy prevalence was 37%, and 22% had multiple sensitisations, in accordance with incidences of 34–47% and 16–42% among the general population, respectively 34.
Increased FeNO is associated with the steroid-responsive eosinophilic airway inflammation, higher particularly in patients with atopic asthma 32. Elevated FeNO was detectable in 23% of the adolescents with repaired OA 9. Elevated values were detectable only in patients with atopy and increased BR, but with no relation to respiratory symptoms 9. Among the present study population, elevated FeNO occurred in 11% of subjects. FeNO was associated with neither current respiratory symptoms, atopy, nor pulmonary abnormalities. In short, in adult patients with repaired OA, airway inflammation was uncommon in the absence of atopy.
In conclusion, respiratory-related morbidity, restrictive ventilatory defect and BHR extend into adulthood after repair of OA in a significant number of patients, whereas obstructive ventilatory defect is not the main pulmonary problem. Nearly half the patients had BHR and over half had restrictive ventilatory defect. Thoracotomy-induced postoperative rib fusions and GOR-associated oesophageal epithelial metaplasia in adulthood were the strongest risk factors for restrictive ventilatory defect.
Footnotes
Support Statement
This study was supported by the Foundation for Paediatric Research, Foundation for Gastroenterologic Research, Helsinki University Central Hospital, and Central Hospital of Pä ijät-Häme (all Finland).
Statement of interest
None declared.
- Received September 28, 2009.
- Accepted February 21, 2010.
- ©ERS 2010