Downloaded from http://adc.bmj.com/ on March 3, 2015 - Published by group.bmj.com
Original article
Conservative management of antenatally diagnosed cystic lung malformations Christabella Ng, Joanna Stanwell, David M Burge, Michael P Stanton Department of Paediatric Surgery, University Hospital Southampton NHS Foundation Trust, Southampton, UK Correspondence to Dr Michael Stanton, Department of Paediatric Surgery, University Hospital Southampton NHS Foundation Trust, Tremona Road, Southampton SO16 6YD, UK; [email protected] Received 11 March 2013 Revised 30 September 2013 Accepted 14 December 2013 Published Online First 9 January 2014
ABSTRACT Aim To review the outcome of all antenatally diagnosed conservatively managed congenital lung malformations (CLMs) managed at our centre. Methods All patients diagnosed antenatally with cystic lung malformations from 2001 to 2011, at a tertiary referral paediatric surgical centre practising a policy of conservative management of asymptomatic cases, were retrospectively reviewed. Data were collected from medical case notes and radiology reports. Ethical approval was obtained from our institutional research and development department. Results The complete records of 74 fetuses antenatally diagnosed with CLM were reviewed. There were 72 live births, at a median gestation of 39.6 weeks. Emergency lobectomy was performed in one symptomatic neonate. Elective lobectomies were performed at parental request in three asymptomatic infants, one of whom had a family history of synovial sarcoma. Two patients developed pneumonia in the affected lobe during early childhood and proceeded to lobectomy at the age of 3 years. One patient with a bronchopulmonary sequestration required embolisation for cyanotic episodes. The remaining 65 patients have been conservatively managed to date, and none have required hospital admission. Less than a quarter report mild respiratory symptoms such as cough or wheeze. Median follow-up is 5 years. Conclusions This retrospective cohort study of 74 consecutive CLMs diagnosed antenatally over a 10-year period demonstrates that most of these lesions will remain asymptomatic throughout childhood. Although the natural history of CLMs in later years remains to be elucidated, we hope that this report on medium-term outcomes will be useful to clinicians who undertake antenatal counselling and may inform the discussion on how best to manage these children.
INTRODUCTION
To cite: Ng C, Stanwell J, Burge DM, et al. Arch Dis Child 2014;99:432–437. 432
Congenital lung malformations (CLM), comprising congenital cystic adenomatoid malformations (CCAMs), bronchopulmonary sequestrations (BPS), bronchogenic cysts and congenital lobar emphysema (CLE), were first detected on antenatal sonography in 1975 and are now seen with increasing frequency on routine anomaly scans.1–3 Although previous reports estimated the incidence to be between 1 in 10 000 and 1 in 35 000, cystic lung malformations are now identified much more frequently antenatally and may be seen in up to 1 in 3000 pregnancies.4–6 Controversy exists as to whether elective resection should be undertaken in all these infants.6–9 For example, a recent survey of Canadian paediatric surgeons reported that one-third initially plan for conservative
What is already known on this topic Congenital lung malformations are now detected much more frequently on antenatal ultrasound. The majority of published series focus on the outcomes of elective surgery. Seventy per cent of centres worldwide pursue a policy of prophylactic resection, regardless of symptoms or size. There is little known about the natural history if these lesions are left in situ and actively observed.
What this study adds This study demonstrates that antenatally diagnosed congenital lung malformations can be safely managed conservatively if they remain asymptomatic into infancy. Symptoms requiring surgical resection only occur in a small number (5%) of infants after median follow-up of 5 years.
management of antenatally diagnosed CLM.10 Resection is currently undertaken in approximately two-third paediatric surgical centres worldwide, regardless of size of lesion or presence of symptoms.11 Antenatal and postnatal imaging can define anatomical characteristics12 (microcystic, macrocystic, mixed), and the presence of an aortic feeding vessel may suggest a BPS. However, definitive classification is only possible on histological assessment13 (Stocker types 0–IV), particularly as ‘hybrid’ lesions (with features of both CCAM and BPS) are increasingly described.14 Neither plain radiography nor ultrasonography can reliably define the size or nature of the CLM postnatally, and CT scanning is widely used.15 16 Repeated CT scans during childhood are associated with a small increased risk of later malignancy, which may preclude frequent radiological surveillance as a management strategy.17 The relationship between CLMs and the rare malignant lung tumours of childhood is unclear. Pleuropulmonary blastomas (PPB) and bronchioalveolar carcinomas (BAC) have been reported in patients who were thought preoperatively to have a CLM or even in those in whom lobectomy had been previously been performed for CLM. Some authors believe the lesions may be premalignant.18 Others argue that the risk is negligible or even nonexistent.19 There may be overlap in the histological
Ng C, et al. Arch Dis Child 2014;99:432–437. doi:10.1136/archdischild-2013-304048
Downloaded from http://adc.bmj.com/ on March 3, 2015 - Published by group.bmj.com
Original article definition of type I PPB and type IV CCAM20 and so difficulty arises in accurately comparing cases, but only one report exists in the literature of a PPB identified within an asymptomatic antenatally diagnosed CLM.21 The incidence of pneumonia in infants with CLM who are asymptomatic as neonates has been estimated at 3–5% and to occur at a median age of 7 months.11 It is suggested that surgery may be more difficult, or associated with worse outcomes, if symptoms have developed.11 22 Most authors in favour of elective surgery agree that deferring surgery beyond the neonatal period is beneficial and undertake resection at 3–6 months of age.11 23 A thoracoscopic approach is widely advocated to minimise the musculoskeletal morbidity associated with thoracic surgery.24 25 The question still remains, however, as to whether CLM resection is in the patient’s best interests, given that recognised complications of the procedure include bleeding, infection, pneumothorax and, rarely, death.11 Since 2001 in our centre, we have managed antenatally diagnosed asymptomatic CLMs expectantly,6 reserving surgery for those that develop symptoms. We report on the progress of this 10-year cohort of patients to date.
METHODS Ethical approval was obtained from our institutional research and development department to perform a retrospective review of all cases of CLM detected on antenatal ultrasound, referred to our fetal medicine unit and subsequently managed postnatally by our tertiary level paediatric surgery service 2001–2011. Our management algorithm is summarised in figure 1. Cases were identified from a prospectively maintained database. Antenatal course, postnatal investigations and subsequent outcome were gathered from the hospital records. Radiological investigations of all patients and histology of resected lesions were documented.
RESULTS From 2001 to 2011, 74 fetuses diagnosed with CLMs were identified. Data are expressed as median (range).
Antenatal course Antenatal ultrasound scanning suggested that 63 (85%) of CLMs were CCAM, 8 (11%) were BPS, 2 (3%) were ‘hybrid’ lesions and 1 (1%) was a congenital lobar emphysema (CLE). Four lesions (5%) resolved antenatally by 35 weeks gestation and six (8%) reduced in size. One fetal demise occurred despite the placement of a thoracoamniotic shunt at 23 weeks gestation to treat pleural effusion, and one pregnancy was terminated at 27 weeks gestation for hydrops fetalis. Amniocentesis was performed in one pregnancy due to identification of sandal gap toes (a soft antenatal ultrasound marker describing an increase in the space between the great toe and the rest of the toes). The remainder of patients did not undergo antenatal interventions.
Neonatal course There were 72 live births at a median gestation of 39.6 weeks (25.1–41.9 weeks) and birth weight 3523 g (880–4320 g). During neonatal assessment, respiratory distress was noted in eight (11%) patients. Of these eight patients, five (7%) required stimulation and oxygen, one (1%) required continuous positive airway pressure (CPAP) and two (3%) required intubation and ventilation (one was 25 weeks gestation and one was 34 weeks gestation and underwent lobectomy on day 29). As illustrated in figures 2 and 3, postnatal chest radiographs (CXR) was reported as ‘normal’ in 26 (36%), non-specific changes such as ‘haziness’ Ng C, et al. Arch Dis Child 2014;99:432–437. doi:10.1136/archdischild-2013-304048
Figure 1 Suggested algorithm for management of antenatally diagnosed congenital lung malformations.
in 36 (50%), ‘cystic changes’ in 7 (10%) and ‘mediastinal shift’ in 2 (3%). An emergency lobectomy was performed in one (1%) symptomatic premature neonate with a right upper lobe microcystic lesion, compressing the middle and lower lobes leading to significant mediastinal shift. This patient had been diagnosed with CLM at 20 weeks gestation and delivered by elective lower segment Caesarean section at 34 weeks gestation due to intrauterine growth restriction (birth weight 1485 g). CPAP was required for 3 days, following which the infant was selfventilating in ambient oxygen. An increasing oxygen requirement followed, leading to intubation and ventilation on day 26. Due to prematurity and anticipated complex surgery, he was transferred to a supra-regional centre (day 27) and underwent lobectomy on day 29. A postoperative pneumothorax required intercostal drainage for 10 days. No other cases underwent surgical intervention in the neonatal period as resolution of any initial symptoms occurred. This included those that had visible cystic change or mediastinal shift on plain radiographs. The two patients with initial mediastinal shift went on to have elective 433
Downloaded from http://adc.bmj.com/ on March 3, 2015 - Published by group.bmj.com
Original article
Figure 2 Plain chest radiograph taken on day 1 of life of an infant with an antenatally diagnosed congenital lung malformation, showing an increased density in the right lower zone (white arrows). This infant underwent CT scan (figure 4) at 3 months of age. CT scan at 1 month and 3 months of age, respectively. Both have CT imaging consistent with CCAM. The second patient has had respiratory syncytial virus-positive bronchiolitis, neither has undergone lobectomy.
Infant course CT images of the CLMs were obtained at 3 months of age in 68 of the 71 conservatively managed infants. As illustrated in figures 4 and 5, these demonstrated: 5 (7%) normal with no lung lesion identified, CCAM in 36 (53%), BPS in 21 (29%), ‘hybrid’ lesions in 1 (1%) and CLE in 5 (7%). The CLE patients had features reported as hyperinflation, hyperexpansion or focal air trapping, but without cystic or solid mass identifiable. Median CLM dimensions were 1.7×2.5×2.6 cm (0.4–4.5 cm) for CCAMs and 1.6×2×2.4 cm (0.4–3 cm) for BPSs. Three patients (4%) did not undergo CT scan at our centre. One infant was one of triplets who were all asymptomatic and all had normal postnatal CXR. As initial conservative
Figure 3 Contrast CT of the infant shown in figure 2 at 3 months of age, showing multiple small cysts in the right lower lobe (likely congenital cystic adenomatoid malformation). This infant remained asymptomatic and has not undergone resection (now aged 3 years). 434
Figure 4 Plain chest radiograph of a neonate with an antenatally diagnosed congenital lung malformation taken on day 1 of life, showing a left lower zone abnormality (cystic change). This patient was asymptomatic and underwent CT scan aged 2 months (figure 5). management was planned and it was unknown which infant had a CLM, a decision was made not to scan all three infants. One patient was managed in Oman at parental request. The third patient had a very small lesion detected antenatally and was asymptomatic. After discussion with the family, no CT was performed, and clinical follow-up was undertaken. Elective surgery was performed at parental request in three (4%) asymptomatic infants: one age of 10 months at another UK centre performing thoracoscopic surgery (microcystic CCAM with no evidence of malignancy), one age 8 months at our centre due to a family history of synovial sarcoma (microcystic CCAM with no evidence of malignancy) and one age 1 year in Oman (histology unavailable).
Childhood course Sixty infants with a persisting CLM ( postnatal CT scan in 58 patients, and two patients with CLM on antenatal scan but postnatal CT not performed) did not undergo surgery and were followed-up into childhood. During this time, 3/60 (5%) patients developed symptoms associated with the CLM. One
Figure 5 Contrast CT scan of the same infant shown in figure 3 at 2 months of age, revealing a parenchymal abnormality in the left lower zone. A feeding vessel arising from the abdominal aorta and venous drainage into the pulmonary venous system was present. This infant remained asymptomatic and has not undergone resection (now aged 4 years). Ng C, et al. Arch Dis Child 2014;99:432–437. doi:10.1136/archdischild-2013-304048
Downloaded from http://adc.bmj.com/ on March 3, 2015 - Published by group.bmj.com
Original article patient (2%) with a BPS experienced episodes of cyanosis at the age of 5 years, having previously not had symptoms attributable to shunting. He proceeded to uncomplicated radiological embolisation. Two patients (3%) were admitted with pneumonia requiring intravenous antibiotics. These two subsequently underwent lobectomy both at the age of 3 years. One was complicated postoperatively by a pneumothorax that required intercostal drainage for 5 days. Histology demonstrated type 1 CCAM in one and a localised area of right lower lobe bronchial atresia with fibrosis and scarring (consistent with congenital segmental atresia) in the other. The remaining 57 patients have been conservatively managed to date. Of these 57, 7 (12%) reported upper respiratory tract infections, 4 (7%) reported episodes of wheeziness and 46 (81%) remained symptom-free. None have required hospital admission, with corresponding radiographic change, for these symptoms to date. Median follow-up is 5 years (1–10 years). Nine asymptomatic patients in this study period had further CT scans as part of ongoing conservative management. These showed reduction in size or resolution in 8/9 (89%). During the study period, two further patients presented with symptomatic CLMs that were not detected on antenatal scans. These were a term neonate and a 3-year-old boy who presented with respiratory distress and right upper lobe pneumonia, respectively. Uncomplicated lobectomy was performed, with histology demonstrating microcystic CCAM in both. Outcomes to date are summarised in figure 6.
DISCUSSION This series represents the largest cohort of patients with antenatally diagnosed asymptomatic CLMs reported to date and confirms that the majority of these remain asymptomatic during infancy. The frequency of postnatal diagnoses we observed is similar to other series16 with regard to CCAM and BPS, but the proportion of CLE (7%) is possibly higher. No patients with presumed CLE have become symptomatic or undergone surgery. Our series largely relies on radiological interpretation of postnatal imaging to define CLM. It is possible that some of the patients we report as having CLE would histologically have evidence of bronchial atresia and be diagnosed as congenital segmental atresia, an entity now considered to lie within the CCAM/BPS spectrum.26 27 The second of the two patients with antenatally diagnosed lesions who underwent surgery in infancy had features consistent with congenital segmental atresia (bronchial atresia and fibrosis). The 5% incidence of symptom development in our series is consistent with other reports in the literature. Two of the three of our patients who became symptomatic during infancy had CCAM histologically and one had a sequestration treated with radiological emobolisation and not excision. A recent systematic review and meta-analysis11 of published reports 1996–2009 estimated that 3% of patients who had not developed symptoms in the neonatal period went on to become symptomatic as infants. A proportion of these cases were part of series where surgery was planned at some stage for all cases. The authors noted that in the majority of cases reported (73%), postnatal surgery was undertaken. Whether a child has symptoms that are secondary to a known CLM or are within the range of universal childhood respiratory illnesses such as upper respiratory tract infections is difficult to define. We used an objective definition of hospital admission with pneumonia and radiographic changes in the area of the known CLM, or requirement for surgical excision. There are few other series describing symptom development in conservatively management of antenatal CLMs. Chetcuti and Crabbe (in Ng C, et al. Arch Dis Child 2014;99:432–437. doi:10.1136/archdischild-2013-304048
Figure 6 Outcomes of all fetuses with antenatally diagnosed cystic lung malformations from 2001 to 2011. a letter) described a cohort of >100 patients who were conservatively managed and in whom approximately 5% became symptomatic.19 Details of length of follow-up and surgical outcome were not described. In another series of expectant management, Hammond et al28 reported that 13/21 (62%) infants, from a cohort of 26 antenatal CLMs, were asymptomatic after nonoperative treatment. Four (19%) underwent (early) surgery and four (19%) died in the early neonatal period without surgery. Other series describe small numbers of patients who were initially asymptomatic and managed conservatively but later required surgery. For example, Aziz et al reported 3/29 (10%) of asymptomatic patients undergoing surgery22 and Hsieh et al29 reported bronchopneumonia in 1/7 (14%). Pneumothorax in infancy has been described in association with CLM22 28 but did not occur in any of our patients. Most series reporting the outcome for antenatally diagnosed CLMs beyond the neonatal period describe mainly elective operative results. Overall, complication rates of 5% have been described in a recent systematic review,11 including air leak, infection and residual disease. One death following elective surgery in infancy has been reported.30 Thoracoscopic resection may prove to be advantageous compared with open thoracotomy in terms of regional anaesthesia use, hospital stay and cosmesis,31 32 although high conversion rates (>20%) are reported initially33 and other surgical complication rates are unlikely to be significantly lower.34 35 Segmentectomy of macroscopically involved tissue, rather than anatomical lobectomy, has been reported to result in a higher rate of disease recurrence.11 435
Downloaded from http://adc.bmj.com/ on March 3, 2015 - Published by group.bmj.com
Original article Some series have attempted to describe differences in the degree of difficulty at surgery, postoperative complications, length of stay and postresection histology for patients who have become symptomatic with chest infections prior to surgery.22 33 A meta-analysis of published data estimated that the ‘risk’ of adverse events after surgery for symptomatic cases (including both neonatal and infant cases) may be up to 2–3 times greater than for elective cases.11 No difference in outcome for cases undertaken beyond the neonatal period was demonstrated. The purported risk of malignancy in patients with CLMs is commonly used to justify pre-emptive surgical resection. BAC has been reported in association with type I CLM in approximately 20 cases (8 were
Original article
Conservative management of antenatally diagnosed cystic lung malformations Christabella Ng, Joanna Stanwell, David M Burge, Michael P Stanton Department of Paediatric Surgery, University Hospital Southampton NHS Foundation Trust, Southampton, UK Correspondence to Dr Michael Stanton, Department of Paediatric Surgery, University Hospital Southampton NHS Foundation Trust, Tremona Road, Southampton SO16 6YD, UK; [email protected] Received 11 March 2013 Revised 30 September 2013 Accepted 14 December 2013 Published Online First 9 January 2014
ABSTRACT Aim To review the outcome of all antenatally diagnosed conservatively managed congenital lung malformations (CLMs) managed at our centre. Methods All patients diagnosed antenatally with cystic lung malformations from 2001 to 2011, at a tertiary referral paediatric surgical centre practising a policy of conservative management of asymptomatic cases, were retrospectively reviewed. Data were collected from medical case notes and radiology reports. Ethical approval was obtained from our institutional research and development department. Results The complete records of 74 fetuses antenatally diagnosed with CLM were reviewed. There were 72 live births, at a median gestation of 39.6 weeks. Emergency lobectomy was performed in one symptomatic neonate. Elective lobectomies were performed at parental request in three asymptomatic infants, one of whom had a family history of synovial sarcoma. Two patients developed pneumonia in the affected lobe during early childhood and proceeded to lobectomy at the age of 3 years. One patient with a bronchopulmonary sequestration required embolisation for cyanotic episodes. The remaining 65 patients have been conservatively managed to date, and none have required hospital admission. Less than a quarter report mild respiratory symptoms such as cough or wheeze. Median follow-up is 5 years. Conclusions This retrospective cohort study of 74 consecutive CLMs diagnosed antenatally over a 10-year period demonstrates that most of these lesions will remain asymptomatic throughout childhood. Although the natural history of CLMs in later years remains to be elucidated, we hope that this report on medium-term outcomes will be useful to clinicians who undertake antenatal counselling and may inform the discussion on how best to manage these children.
INTRODUCTION
To cite: Ng C, Stanwell J, Burge DM, et al. Arch Dis Child 2014;99:432–437. 432
Congenital lung malformations (CLM), comprising congenital cystic adenomatoid malformations (CCAMs), bronchopulmonary sequestrations (BPS), bronchogenic cysts and congenital lobar emphysema (CLE), were first detected on antenatal sonography in 1975 and are now seen with increasing frequency on routine anomaly scans.1–3 Although previous reports estimated the incidence to be between 1 in 10 000 and 1 in 35 000, cystic lung malformations are now identified much more frequently antenatally and may be seen in up to 1 in 3000 pregnancies.4–6 Controversy exists as to whether elective resection should be undertaken in all these infants.6–9 For example, a recent survey of Canadian paediatric surgeons reported that one-third initially plan for conservative
What is already known on this topic Congenital lung malformations are now detected much more frequently on antenatal ultrasound. The majority of published series focus on the outcomes of elective surgery. Seventy per cent of centres worldwide pursue a policy of prophylactic resection, regardless of symptoms or size. There is little known about the natural history if these lesions are left in situ and actively observed.
What this study adds This study demonstrates that antenatally diagnosed congenital lung malformations can be safely managed conservatively if they remain asymptomatic into infancy. Symptoms requiring surgical resection only occur in a small number (5%) of infants after median follow-up of 5 years.
management of antenatally diagnosed CLM.10 Resection is currently undertaken in approximately two-third paediatric surgical centres worldwide, regardless of size of lesion or presence of symptoms.11 Antenatal and postnatal imaging can define anatomical characteristics12 (microcystic, macrocystic, mixed), and the presence of an aortic feeding vessel may suggest a BPS. However, definitive classification is only possible on histological assessment13 (Stocker types 0–IV), particularly as ‘hybrid’ lesions (with features of both CCAM and BPS) are increasingly described.14 Neither plain radiography nor ultrasonography can reliably define the size or nature of the CLM postnatally, and CT scanning is widely used.15 16 Repeated CT scans during childhood are associated with a small increased risk of later malignancy, which may preclude frequent radiological surveillance as a management strategy.17 The relationship between CLMs and the rare malignant lung tumours of childhood is unclear. Pleuropulmonary blastomas (PPB) and bronchioalveolar carcinomas (BAC) have been reported in patients who were thought preoperatively to have a CLM or even in those in whom lobectomy had been previously been performed for CLM. Some authors believe the lesions may be premalignant.18 Others argue that the risk is negligible or even nonexistent.19 There may be overlap in the histological
Ng C, et al. Arch Dis Child 2014;99:432–437. doi:10.1136/archdischild-2013-304048
Downloaded from http://adc.bmj.com/ on March 3, 2015 - Published by group.bmj.com
Original article definition of type I PPB and type IV CCAM20 and so difficulty arises in accurately comparing cases, but only one report exists in the literature of a PPB identified within an asymptomatic antenatally diagnosed CLM.21 The incidence of pneumonia in infants with CLM who are asymptomatic as neonates has been estimated at 3–5% and to occur at a median age of 7 months.11 It is suggested that surgery may be more difficult, or associated with worse outcomes, if symptoms have developed.11 22 Most authors in favour of elective surgery agree that deferring surgery beyond the neonatal period is beneficial and undertake resection at 3–6 months of age.11 23 A thoracoscopic approach is widely advocated to minimise the musculoskeletal morbidity associated with thoracic surgery.24 25 The question still remains, however, as to whether CLM resection is in the patient’s best interests, given that recognised complications of the procedure include bleeding, infection, pneumothorax and, rarely, death.11 Since 2001 in our centre, we have managed antenatally diagnosed asymptomatic CLMs expectantly,6 reserving surgery for those that develop symptoms. We report on the progress of this 10-year cohort of patients to date.
METHODS Ethical approval was obtained from our institutional research and development department to perform a retrospective review of all cases of CLM detected on antenatal ultrasound, referred to our fetal medicine unit and subsequently managed postnatally by our tertiary level paediatric surgery service 2001–2011. Our management algorithm is summarised in figure 1. Cases were identified from a prospectively maintained database. Antenatal course, postnatal investigations and subsequent outcome were gathered from the hospital records. Radiological investigations of all patients and histology of resected lesions were documented.
RESULTS From 2001 to 2011, 74 fetuses diagnosed with CLMs were identified. Data are expressed as median (range).
Antenatal course Antenatal ultrasound scanning suggested that 63 (85%) of CLMs were CCAM, 8 (11%) were BPS, 2 (3%) were ‘hybrid’ lesions and 1 (1%) was a congenital lobar emphysema (CLE). Four lesions (5%) resolved antenatally by 35 weeks gestation and six (8%) reduced in size. One fetal demise occurred despite the placement of a thoracoamniotic shunt at 23 weeks gestation to treat pleural effusion, and one pregnancy was terminated at 27 weeks gestation for hydrops fetalis. Amniocentesis was performed in one pregnancy due to identification of sandal gap toes (a soft antenatal ultrasound marker describing an increase in the space between the great toe and the rest of the toes). The remainder of patients did not undergo antenatal interventions.
Neonatal course There were 72 live births at a median gestation of 39.6 weeks (25.1–41.9 weeks) and birth weight 3523 g (880–4320 g). During neonatal assessment, respiratory distress was noted in eight (11%) patients. Of these eight patients, five (7%) required stimulation and oxygen, one (1%) required continuous positive airway pressure (CPAP) and two (3%) required intubation and ventilation (one was 25 weeks gestation and one was 34 weeks gestation and underwent lobectomy on day 29). As illustrated in figures 2 and 3, postnatal chest radiographs (CXR) was reported as ‘normal’ in 26 (36%), non-specific changes such as ‘haziness’ Ng C, et al. Arch Dis Child 2014;99:432–437. doi:10.1136/archdischild-2013-304048
Figure 1 Suggested algorithm for management of antenatally diagnosed congenital lung malformations.
in 36 (50%), ‘cystic changes’ in 7 (10%) and ‘mediastinal shift’ in 2 (3%). An emergency lobectomy was performed in one (1%) symptomatic premature neonate with a right upper lobe microcystic lesion, compressing the middle and lower lobes leading to significant mediastinal shift. This patient had been diagnosed with CLM at 20 weeks gestation and delivered by elective lower segment Caesarean section at 34 weeks gestation due to intrauterine growth restriction (birth weight 1485 g). CPAP was required for 3 days, following which the infant was selfventilating in ambient oxygen. An increasing oxygen requirement followed, leading to intubation and ventilation on day 26. Due to prematurity and anticipated complex surgery, he was transferred to a supra-regional centre (day 27) and underwent lobectomy on day 29. A postoperative pneumothorax required intercostal drainage for 10 days. No other cases underwent surgical intervention in the neonatal period as resolution of any initial symptoms occurred. This included those that had visible cystic change or mediastinal shift on plain radiographs. The two patients with initial mediastinal shift went on to have elective 433
Downloaded from http://adc.bmj.com/ on March 3, 2015 - Published by group.bmj.com
Original article
Figure 2 Plain chest radiograph taken on day 1 of life of an infant with an antenatally diagnosed congenital lung malformation, showing an increased density in the right lower zone (white arrows). This infant underwent CT scan (figure 4) at 3 months of age. CT scan at 1 month and 3 months of age, respectively. Both have CT imaging consistent with CCAM. The second patient has had respiratory syncytial virus-positive bronchiolitis, neither has undergone lobectomy.
Infant course CT images of the CLMs were obtained at 3 months of age in 68 of the 71 conservatively managed infants. As illustrated in figures 4 and 5, these demonstrated: 5 (7%) normal with no lung lesion identified, CCAM in 36 (53%), BPS in 21 (29%), ‘hybrid’ lesions in 1 (1%) and CLE in 5 (7%). The CLE patients had features reported as hyperinflation, hyperexpansion or focal air trapping, but without cystic or solid mass identifiable. Median CLM dimensions were 1.7×2.5×2.6 cm (0.4–4.5 cm) for CCAMs and 1.6×2×2.4 cm (0.4–3 cm) for BPSs. Three patients (4%) did not undergo CT scan at our centre. One infant was one of triplets who were all asymptomatic and all had normal postnatal CXR. As initial conservative
Figure 3 Contrast CT of the infant shown in figure 2 at 3 months of age, showing multiple small cysts in the right lower lobe (likely congenital cystic adenomatoid malformation). This infant remained asymptomatic and has not undergone resection (now aged 3 years). 434
Figure 4 Plain chest radiograph of a neonate with an antenatally diagnosed congenital lung malformation taken on day 1 of life, showing a left lower zone abnormality (cystic change). This patient was asymptomatic and underwent CT scan aged 2 months (figure 5). management was planned and it was unknown which infant had a CLM, a decision was made not to scan all three infants. One patient was managed in Oman at parental request. The third patient had a very small lesion detected antenatally and was asymptomatic. After discussion with the family, no CT was performed, and clinical follow-up was undertaken. Elective surgery was performed at parental request in three (4%) asymptomatic infants: one age of 10 months at another UK centre performing thoracoscopic surgery (microcystic CCAM with no evidence of malignancy), one age 8 months at our centre due to a family history of synovial sarcoma (microcystic CCAM with no evidence of malignancy) and one age 1 year in Oman (histology unavailable).
Childhood course Sixty infants with a persisting CLM ( postnatal CT scan in 58 patients, and two patients with CLM on antenatal scan but postnatal CT not performed) did not undergo surgery and were followed-up into childhood. During this time, 3/60 (5%) patients developed symptoms associated with the CLM. One
Figure 5 Contrast CT scan of the same infant shown in figure 3 at 2 months of age, revealing a parenchymal abnormality in the left lower zone. A feeding vessel arising from the abdominal aorta and venous drainage into the pulmonary venous system was present. This infant remained asymptomatic and has not undergone resection (now aged 4 years). Ng C, et al. Arch Dis Child 2014;99:432–437. doi:10.1136/archdischild-2013-304048
Downloaded from http://adc.bmj.com/ on March 3, 2015 - Published by group.bmj.com
Original article patient (2%) with a BPS experienced episodes of cyanosis at the age of 5 years, having previously not had symptoms attributable to shunting. He proceeded to uncomplicated radiological embolisation. Two patients (3%) were admitted with pneumonia requiring intravenous antibiotics. These two subsequently underwent lobectomy both at the age of 3 years. One was complicated postoperatively by a pneumothorax that required intercostal drainage for 5 days. Histology demonstrated type 1 CCAM in one and a localised area of right lower lobe bronchial atresia with fibrosis and scarring (consistent with congenital segmental atresia) in the other. The remaining 57 patients have been conservatively managed to date. Of these 57, 7 (12%) reported upper respiratory tract infections, 4 (7%) reported episodes of wheeziness and 46 (81%) remained symptom-free. None have required hospital admission, with corresponding radiographic change, for these symptoms to date. Median follow-up is 5 years (1–10 years). Nine asymptomatic patients in this study period had further CT scans as part of ongoing conservative management. These showed reduction in size or resolution in 8/9 (89%). During the study period, two further patients presented with symptomatic CLMs that were not detected on antenatal scans. These were a term neonate and a 3-year-old boy who presented with respiratory distress and right upper lobe pneumonia, respectively. Uncomplicated lobectomy was performed, with histology demonstrating microcystic CCAM in both. Outcomes to date are summarised in figure 6.
DISCUSSION This series represents the largest cohort of patients with antenatally diagnosed asymptomatic CLMs reported to date and confirms that the majority of these remain asymptomatic during infancy. The frequency of postnatal diagnoses we observed is similar to other series16 with regard to CCAM and BPS, but the proportion of CLE (7%) is possibly higher. No patients with presumed CLE have become symptomatic or undergone surgery. Our series largely relies on radiological interpretation of postnatal imaging to define CLM. It is possible that some of the patients we report as having CLE would histologically have evidence of bronchial atresia and be diagnosed as congenital segmental atresia, an entity now considered to lie within the CCAM/BPS spectrum.26 27 The second of the two patients with antenatally diagnosed lesions who underwent surgery in infancy had features consistent with congenital segmental atresia (bronchial atresia and fibrosis). The 5% incidence of symptom development in our series is consistent with other reports in the literature. Two of the three of our patients who became symptomatic during infancy had CCAM histologically and one had a sequestration treated with radiological emobolisation and not excision. A recent systematic review and meta-analysis11 of published reports 1996–2009 estimated that 3% of patients who had not developed symptoms in the neonatal period went on to become symptomatic as infants. A proportion of these cases were part of series where surgery was planned at some stage for all cases. The authors noted that in the majority of cases reported (73%), postnatal surgery was undertaken. Whether a child has symptoms that are secondary to a known CLM or are within the range of universal childhood respiratory illnesses such as upper respiratory tract infections is difficult to define. We used an objective definition of hospital admission with pneumonia and radiographic changes in the area of the known CLM, or requirement for surgical excision. There are few other series describing symptom development in conservatively management of antenatal CLMs. Chetcuti and Crabbe (in Ng C, et al. Arch Dis Child 2014;99:432–437. doi:10.1136/archdischild-2013-304048
Figure 6 Outcomes of all fetuses with antenatally diagnosed cystic lung malformations from 2001 to 2011. a letter) described a cohort of >100 patients who were conservatively managed and in whom approximately 5% became symptomatic.19 Details of length of follow-up and surgical outcome were not described. In another series of expectant management, Hammond et al28 reported that 13/21 (62%) infants, from a cohort of 26 antenatal CLMs, were asymptomatic after nonoperative treatment. Four (19%) underwent (early) surgery and four (19%) died in the early neonatal period without surgery. Other series describe small numbers of patients who were initially asymptomatic and managed conservatively but later required surgery. For example, Aziz et al reported 3/29 (10%) of asymptomatic patients undergoing surgery22 and Hsieh et al29 reported bronchopneumonia in 1/7 (14%). Pneumothorax in infancy has been described in association with CLM22 28 but did not occur in any of our patients. Most series reporting the outcome for antenatally diagnosed CLMs beyond the neonatal period describe mainly elective operative results. Overall, complication rates of 5% have been described in a recent systematic review,11 including air leak, infection and residual disease. One death following elective surgery in infancy has been reported.30 Thoracoscopic resection may prove to be advantageous compared with open thoracotomy in terms of regional anaesthesia use, hospital stay and cosmesis,31 32 although high conversion rates (>20%) are reported initially33 and other surgical complication rates are unlikely to be significantly lower.34 35 Segmentectomy of macroscopically involved tissue, rather than anatomical lobectomy, has been reported to result in a higher rate of disease recurrence.11 435
Downloaded from http://adc.bmj.com/ on March 3, 2015 - Published by group.bmj.com
Original article Some series have attempted to describe differences in the degree of difficulty at surgery, postoperative complications, length of stay and postresection histology for patients who have become symptomatic with chest infections prior to surgery.22 33 A meta-analysis of published data estimated that the ‘risk’ of adverse events after surgery for symptomatic cases (including both neonatal and infant cases) may be up to 2–3 times greater than for elective cases.11 No difference in outcome for cases undertaken beyond the neonatal period was demonstrated. The purported risk of malignancy in patients with CLMs is commonly used to justify pre-emptive surgical resection. BAC has been reported in association with type I CLM in approximately 20 cases (8 were
