Primary paediatric chest wall tumours necessitating surgical management

Abstract

Introduction

Limited literature exists on oncological chest wall reconstruction in the paediatric population, with the field still largely undecided on the best surgical reconstructive techniques to employ. The use of biological grafts/meshes is gaining popularity in certain adult surgical procedures but their use in paediatric procedures is rarely reported in the literature. We present the outcomes of our institution’s multidisciplinary approach to managing paediatric chest wall tumours as well as our experience with the use of biological grafts for chest wall reconstruction following oncological resections.

Methods

Data were analysed retrospectively from eight paediatric patients who were treated for primary chest wall tumours between 2010 and 2018.

Results

The tumours comprised two lipoblastomas, three Ewing’s sarcomas, an undifferentiated sarcoma with osteosarcomatous differentiation, a high grade undifferentiated sarcoma and a myofibroma. Seven of the eight patients underwent chest wall reconstruction with a biological graft. There were no postoperative mortalities and no evidence of recurrence in any of the patients in the series. No further chest wall operations were required and there were no postoperative infection related complications.

Conclusions

We support the use of biological grafts for chest wall reconstruction after oncological resections and maintain that a multidisciplinary approach is essential for the management of paediatric chest wall tumours.

Introduction

Chest wall tumours are rare in the paediatric population. Successful management is optimised by early detection and prompt treatment by a multidisciplinary team. These patients pose a significant challenge to clinicians across the globe since they can present with a wide variety of tumours, which may potentially require large chest wall resections along with reconstructive procedures. Most will also warrant neoadjuvant chemotherapy and/or radiotherapy.

Chest wall reconstruction after tumour resection has been well described in adults but there is a paucity of information in the literature when it comes to the paediatric population. Although multiple techniques and the use of various prosthetics/grafts have been reported,^1–5 there is no clear evidence suggesting the most appropriate technique for reconstruction.⁶

We present the multidisciplinary management strategies employed as well as the outcomes attained by our tertiary centre over an eight-year period. All patients were managed by integrated paediatric surgical and oncological teams. Although only five of the eight patients received adjuvant therapy, all underwent surgical resection. A biological graft, Biodesign^® (Cook Medical, Bloomington, IN, US), was used in all patients requiring chest wall reconstruction, which resulted in favourable outcomes.

Methods

Data from all paediatric patients who were treated for isolated primary chest wall tumours (both benign and malignant) between 2010 and 2018 were analysed retrospectively using electronic patient records. Only patients who underwent surgical intervention and combined oncological follow-up review at our centre were included in the study. Data collected included age at diagnosis, tumour type, details of neoadjuvant and adjuvant therapy, age at surgery, number of ribs resected and treatment related complications as well as details of follow-up review. Institutional ethical approval was granted by the hospital trust for the retrospective data analysis.

Results

Eight patients were identified as having undergone oncological resection for chest wall tumours. The tumours comprised two lipoblastomas, three Ewing’s sarcomas, an undifferentiated sarcoma with osteosarcomatous differentiation, a high grade undifferentiated sarcoma and a myofibroma. The age of the children included in the review ranged from one month to fourteen years. The mean age at diagnosis was 49 months.

Seven (88%) of the eight children presented with a left-sided chest wall tumour. Six of the patients presented primarily with the development of a chest wall mass while two were investigated after presenting with a persistent cough and subsequently had abnormal imaging. Both of these children were later diagnosed with Ewing’s sarcoma. Five of the eight masses were malignant in nature. All malignant cases received neoadjuvant chemotherapy but only one (high grade undifferentiated sarcoma) had preoperative radiotherapy. Table 1 summarises the preoperative data.

Table 1.

Preoperative data

Patient sex	Age at diagnosis	Tumour type	Presentation	Side of lesion	Location in relation to rib	Preoperative treatment
M	9 months	Lipoblastoma	Mass	Left	Middle third	Nil
M	10 months	Lipoblastoma	Mass	Left	Posterior	Nil
F	95 months	Ewing’s sarcoma	Mass	Left	Posterior	Chemotherapy
F	34 months	Ewing’s sarcoma	Cough	Left	Posterior	Chemotherapy
F	69 months	Ewing’s sarcoma	Cough	Left	Anterior	Chemotherapy
F	7 months	Undifferentiated sarcoma	Mass	Right	Anterior	Chemotherapy
F	171 months	High grade undifferentiated sarcoma	Mass	Left	Posterior	Chemotherapy + radiotherapy
F	1 month	Myofibroma	Mass	Left	Middle third	Nil

The mean age at the time of surgery was 53 months. Seven of the eight patients required rib resections as well as chest wall reconstructions. The myofibroma was the only tumour that was resected without any surrounding tissue. On average, two ribs (ranging from one to three) required partial resection in each case. The same biological graft, Biodesign^®, was used for all chest wall reconstructions. Following resection of the tumour with appropriate margins, the Biodesign^® patch was secured to the surrounding ribs and fascia with a non-absorbable monofilament suture. All five of the patients with malignant tumours received postoperative chemotherapy and three underwent postoperative radiotherapy. Table 2 summarises the postoperative data.

Table 2.

Postoperative data

Tumour type	Number of ribs resected	Graft used for reconstruction	Postoperative treatment	Complications
Lipoblastoma	2	Biodesign®	Nil
Lipoblastoma	2	Biodesign®	Nil	Temporary abdominal wall paresis
Ewing’s sarcoma	3	Biodesign®	Chemotherapy + radiotherapy
Ewing’s sarcoma	2	Biodesign®	Chemotherapy + radiotherapy	Scoliosis, hyposplenism
Ewing’s sarcoma	3	Biodesign®	Chemotherapy + radiotherapy
Undifferentiated sarcoma	2	Biodesign®	Chemotherapy
High grade undifferentiated sarcoma	1	Biodesign®	Chemotherapy + radiotherapy	Hearing loss
Myofibroma	0	Biodesign®	Nil

The time from diagnosis to completion of treatment ranged from one to two months in the benign cases (only surgical resection and reconstruction necessary) and from eight to fourteen months (mean: 10.8 months) in the malignant cases. Table 3 shows the neoadjuvant and adjuvant treatment regimens based on tumour histology.

Table 3.

Malignant tumour classification, neoadjuvant regimens and adjuvant therapies

Tumour type	Histological features / biomarkers	Neoadjuvant chemotherapy	Protocol followed	Preoperative radiotherapy	Adjuvant chemotherapy	Postoperative radiotherapy
Ewing’s sarcoma	CD99 positive, EWSR1 translocation	VDC/IE (9 cycles)	Euro Ewing 2012 (experimental)	Nil	VC/IE (5 cycles)	PBT 45 CGE + 9 CGE boost
Ewing’s sarcoma	CD99 positive, EWSR1 translocation	VIDE (6 cycles)	Euro Ewing 99	Nil	IVA (8 cycles)	PBT 42 CGE
Ewing’s sarcoma	CD99 positive, EWSR1 translocation	VIDE (6 cycles)	Euro Ewing 99	Nil	IVA (8 cycles)	PBT 45 CGE + 9 CGE boost
Undifferentiated sarcoma	Desmin positive, osteosarcomatous differentiation	IVA (3 cycles)	EpSSG RMS 2005	Nil	Cis/Dox (4 cycles)	Nil
High grade undifferentiated sarcoma	No biomarkers	Ifos/Dox (5 cycles)	EpSSG NRSTS 2005	PBT 45 CGE + 4.5 CGE boost	VAC (8 cycles)	Nil

CD99 = cluster of differentiation 99; CGE = cobalt gray equivalent; Cis/Dox = cisplatin, doxorubicin; EpSSG = European Paediatric Soft Tissue Sarcoma Study Group; EWSR1 = Ewing’s sarcoma breakpoint region 1; Ifos/Dox = ifosfamide, doxorubicin; IVA = ifosfamide, vincristine, dactinomycin; NRSTS = non-rhabdomyosarcoma soft tissue sarcoma; PBT = proton beam therapy; RMS = rhabdomyosarcoma; VAC = vincristine, dactinomycin, cyclophosphamide; VC/IE = vincristine, cyclophosphamide/ifosfamide, etoposide; VDC/IE = vincristine, doxorubicin, cyclophosphamide/ifosfamide, etoposide; VIDE = vincristine, ifosfamide, doxorubicin, etoposide

Complications

None of the patients developed postoperative wound or graft infections. One patient developed a transient paresis of the upper left abdominal wall following chest wall resection and reconstruction. This presented as a bulge in the left upper abdomen. Complete recovery was noted two years after surgery. The remaining complications were predominantly related to the administration of chemotherapy/radiotherapy. One patient developed scoliosis. It is uncertain whether this was a result of local radiotherapy (which had caused decreased vertebral height) or a direct result of the chest wall resection. The same patient also developed hyposplenism secondary to splenic irradiation. One patient developed high frequency hearing loss following chemotherapy (Table 2).

Follow-up review

Both patients with lipoblastomas will soon be completing their five-year follow-up period. They were seen on an annual basis with repeat chest x-rays. There has been no evidence of recurrence.

Two of the patients diagnosed with Ewing’s sarcomas are in remission. They are continuing to be followed up (at seven and four years after treatment respectively). Both are seen on a six-monthly basis with repeat chest x-rays and annual magnetic resonance imaging (MRI). For the patient receiving prolonged follow-up (seven years), this is largely a result of treatment related to scoliosis. She is under regular review with a spinal team and is being managed non-operatively with a spinal brace. The third patient with an Ewing’s sarcoma is currently two years post-treatment and is being seen every four months with repeated MRI. MRI interpretation in this patient has proved to be difficult owing to the radiological appearances of the biological reaction of the Biodesign^® graft. She is therefore under close surveillance.

The patient diagnosed with an undifferentiated sarcoma with osteosarcomatous differentiation is in remission and continues to undergo six-monthly whole body MRI. She is close to reaching the five-year follow-up mark. The initial biopsy was consistent with an undifferentiated/anaplastic sarcoma. At the time of formal resection, there was evidence of osteosarcomatous differentiation. Adjuvant therapy was adjusted appropriately (Table 3).

The patient with a high grade undifferentiated sarcoma only completed treatment in early 2018 and is consequently under regular four-monthly follow-up review with repeated imaging. The infant diagnosed with the myofibroma was followed up for six months after surgery and was subsequently discharged with no evidence of recurrence.

None of the eight patients required additional surgery to the chest wall. All showed a good response to the combination of surgery and adjuvant therapy. No tumour recurrence has been reported to date.

Discussion

Paediatric chest wall tumours are rare and varied in histotype. They range from benign to malignant and slow growing to aggressive.

Lipoblastomas are benign tumours arising from abnormal proliferation of immature white fat cells.⁷ They often grow aggressively and invasively, and therefore require complete surgical excision with the involved rib segments. They are seen predominantly in childhood and tend to occur in boys under three years of age.^8,9 Both patients with a lipoblastoma in our series fell into this category. They were both followed up for five years after treatment. The literature varies in terms of the recommended follow-up duration, with minimums of two–three years being suggested,^9,10 but up to five years being described as reasonable in patients with no recurrence.⁸

Ewing’s sarcomas behave aggressively with a high degree of metastatic spread.¹¹ All three patients with Ewing’s sarcoma diagnoses received preoperative chemotherapy, which has been proved to achieve greater local control and result in less extensive surgery with more accurate margin resection.¹² All patients received postoperative radiotherapy in keeping with the current guidance.¹³

Myofibromas are one of the most common tumours of infancy. They are benign in nature, are more likely to present as solitary lesions and have an extremely good prognosis. They generally do not warrant any treatment beyond surgical excision and it has been shown that positive surgical margins do not affect outcomes.¹⁴ The patient recovered well postoperatively and was discharged from our care with no further follow-up review required.

The tumour with osteosarcomatous differentiation proved somewhat difficult to manage based on changing histological diagnoses before and after surgery (as described above). Despite being the most common bone tumours in childhood, osteosarcomas are incredibly rare to develop primarily in the chest wall.¹³ Very little literature exists about the recommended management strategies for paediatric chest wall osteosarcomas, except the need for resection with clear bony and soft tissue margins. The patient in our series has so far had positive outcomes.

Surgical management of chest wall tumours involves two key components: gross total resection and reconstruction. The aim of resection is to achieve adequate tumour clearance from soft tissue and bony margins whereas the aim of reconstruction is to enable effective ventilation, avoid deformity and allow for adequate protection of internal organs.⁶ There are arguments to suggest that using a rigid material to maintain the chest structure is beneficial¹⁵ although there is also evidence to suggest that less rigid biological meshes/bioabsorbable plates provide better results in young, growing children.^3,4,16 The use of a rigid prosthetic is thought to maintain good chest wall contour and provide better protection of internal structures.¹⁵ They have, however, also been associated with a risk of dislocation/dislodgement, chronic periprosthetic infection and chronic pain.^5,17

The largest resection in our series involved the two children requiring resection of a portion of three ribs. The resultant defect was within the limits of using the Biodesign^® patch without need for additional reconstructive methods. Sana et al found that surgeons generally support the idea that defects of <5cm or resections involving fewer than four ribs do not require complex reconstructions.¹⁶

Although chest wall dynamics and protection of the intrathoracic viscera are the obvious concerns when faced with a chest wall defect, attention needs to also be placed on movement of the shoulder girdle and hindrance of rotation of the scapula. Large posterior chest wall defects can be protected by the shoulder girdle but those defects in line with the rotation of the tip of the scapula may provide an area of entrapment, thereby necessitating a more definitive reconstruction. The paper by Sana et al demonstrates the extensive armamentarium (including biological and non-biological grafts, local/pedicle/free muscle flaps, methylmethacrylate moulds and titanium plating) available to surgeons for chest wall defects large enough to require a more sturdy reconstruction to ensure chest wall dynamics are not compromised.¹⁶

Surgisis^® (Cook Medical) grafts have been shown to be very useful in paediatric chest wall reconstruction.^17,18 A refined, enhanced version of the Surgisis^® product line was developed and renamed Biodesign^®.¹⁹ Both Surgisis^® and Biodesign^® are developed from porcine small intestinal submucosa. Biodesign^®, however, is a purer biomaterial with a lower lipid content than Surgisis^®. It has also demonstrated improved neovascularisation. In order to produce the Biodesign^® mesh/graft, cells are removed from porcine small intestinal submucosa, leaving behind a three-dimensional extracellular matrix composed predominantly of protein. The graft provides a favourable scaffold for the patient’s own cells to grow on, which is responsible for tissue remodelling. When healing is complete, the graft is undetectable as it is completely resorbed and replaced with new collagen.²⁰

The use of biological grafts has been shown to be favourable in a variety of respects: they are easily cut and moulded intraoperatively to allow for precise, patient specific reconstruction,⁴ they provide adequate stability and allow for growth capacity of the developing child (as the graft is incorporated and remodelled with continued growth),¹⁷ and they are thought to have less of a foreign body reaction and a lower risk of infection.²⁰ The use of a biological graft may therefore prevent the need for further chest wall surgical intervention.¹⁷ None of the patients in our series required additional procedures to the thoracic wall. Although one patient developed scoliosis (a well recognised complication of chest wall resection and reconstruction),^6,21 it is possible that this was also related to the direct administration of radiotherapy to the posterolateral vertebral region.

A significant finding from our series, which we deem to be particularly important in the paediatric oncology population, is the absence of postoperative wound infection related complications. This has been noted to be especially problematic with rigid prosthetics and can cause significant morbidity.¹⁷ The overall outcomes from the use of a biological graft in our centre, however, have proved to be favourable.

Nevertheless, from our experience with the use of biological grafts, we do recommend that postoperative radiological findings are interpreted with caution. The Biodesign^® graft elicits a short-term inflammatory tissue reaction, which can be difficult to distinguish from tumour recurrence on MRI (particularly on T2 weighted imaging).¹⁹ Regular repeated imaging as well as assessment by an experienced paediatric radiologist with oncology imaging expertise is recommended. The appearance of the inflammatory reaction does lessen with time and MRI can still be used as the primary imaging modality. Positron emission tomography may be performed for further clarity if there are areas of concern.

Other available biological grafts described in the literature include porcine dermal grafts and human dermal allografts (Table 4).²⁰ The use of biological grafts has been well documented in adult hernia repairs²² and pelvic floor reconstructions as well as for the repair of abdominal wall defects and ventral hernias.^23–25 Owing to a lack of high level evidence, no clear guidelines exist for the use and role of biological meshes in adult surgical procedures. There is also little information available regarding long-term outcomes. There is an incredibly limited amount of literature on the use of biological meshes in paediatric surgical procedures and as such, more research (including randomised controlled trials) will be needed in order to create guidelines/recommendations for their routine use.

Table 4.

Comparison of commonly used biological grafts

	Porcine		Human
	Small intestinal submucosa	Dermal	Dermal (cadaveric)
Trade names	Biodesign®, Surgisis®	Permacol™	AlloDerm®
Manufacturer	Cook Medical	Covidien	LifeCell
Cross-linked	No	Yes	No
Rehydration required	Yes	No	No
Tissue remodelling	Rapid	Slower	Rapid
Duration of graft*	Shorter term	Longer term	Shorter term

^*There is no current literature providing data regarding the specific lengths of time taken for the grafts to be resorbed. Meshes that are not cross-linked are speculated to last months while cross-linked meshes may last years.

Conclusions

A multidisciplinary approach has proved to yield successful outcomes in the management of both benign and malignant paediatric chest wall tumours at our institution. We support the use of Biodesign^® biological grafts in paediatric chest wall reconstruction after oncological resections. Favourable outcomes can be achieved with the combined use of evidence-based optimal adjuvant oncological therapies.