Keywords
chest wall - sarcoma - wound healing - sternum - tumor
Introduction
Chest wall resections due to malignancies require extended surgical procedures to
ensure wide safety margins. Reconstruction of larger defects, and in particular complete
sternal resections, presents a major challenge.[1] Sufficient chest wall stability, absence of respiratory impairments, adequate tissue
coverage, and satisfactory cosmetics results must all be achieved.[2]
Synthetic materials combined with soft-tissue transfers, such as pectoralis or latissimus
muscle flaps, have been employed in chest wall reconstructions for decades.[2]
[3] This strategy has been associated with significant morbidity and mortality in 46
to 69% of patients.[4]
[5] Respiratory impairments may pose major problems such as paradox chest wall movements
and rigidity due to excessive encapsulation.[6] Seroma formation and wound healing impairment are paving the way for infection of
synthetic materials which often jeopardizes long-term success and result in significant
failure rates.[7]
In the past, the use of a wide variety of synthetic materials such as polypropylene,
polytetrafluoroethylene (PTFE), solid methylmethacrylate sandwich constructions, as
well as various composite absorbable meshes has been explored. PTFE materials have
been most frequently used.[8] However, due to the inconclusiveness of available studies, the type of allomaterial
and the technique of reconstruction still depend largely on the expertise and preference
of the individual surgeon.
Recently, collagen matrixes harvested from human or animal sources and processed for
medical use have emerged as an alternative. The implanted matrix undergoes a remodeling
process with cellular infiltration, neovascularization, and exchange of extracellular
matrix. Some of these materials, such as the product used in the present study, have
been chemically cross-linked to increase strength and durability.[9]
[10]
[11]
[12]
[13]
The implantation of biological collagen matrixes has been described in abdominal hernia
repair and various other fields of surgery.[11]
[12]
[14] However, in thoracic surgery, experience is limited to single cases and small series.[15]
[16]
[17] Unlike permanently fixated biological and synthetic materials, acellular collagen
matrixes allow for integration and remodeling. The long-term integrity, strength,
and stability of a collagen matrix are unknown. In this pilot study, we evaluated
the feasibility and long-term results in the usage of a porcine dermal collagen matrix
for thoracic wall reconstruction.
Materials and Methods
Study Setting and Design
Beginning in December 2009, six consecutive sarcoma patients scheduled for major thoracic
wall or sternum resection were enrolled in the study ([Table 1]). Written informed consent was obtained from all patients. Data were analyzed retrospectively
and included patient demographics, diagnoses, operative data, and both pre- and postoperative
laboratory and technical examinations. All patients underwent preoperative and postoperative
computed tomography (CT) scans required for tumor evaluation and oncological follow-up.
Table 1
Patient characteristics
|
Patient
|
Sex (M/F)
|
Age (y)
|
Diagnosis
|
Location
|
Defect size (cm2)
|
Pretreatment
|
|
1
|
M
|
22
|
Osteosarcoma (G3)
|
Ninth rib left
|
9 × 18 cm (162)
|
Preoperative chemotherapy
|
|
2
|
F
|
48
|
High-grade sarcoma (G3)
|
Dorsal chest wall left (ribs 7–9)
|
8 × 10 cm (80)
|
None
|
|
3
|
M
|
44
|
Chondrosarcoma (G2)
|
Sternum
|
7 × 19 cm (133)
|
None
|
|
4
|
F
|
50
|
Ewing sarcoma (G4)
|
Seventh rib right
|
14 × 10 cm (140)
|
Preoperative chemotherapy
|
|
5
|
M
|
66
|
Metastasis from synovial sarcoma (G3) of the right femur
|
Chest wall right (ribs 4–6)
|
14 × 18 cm (252)
|
Preoperative chemotherapy
|
|
6
|
F
|
46
|
Chondrosarcoma (G2)
|
Sternum
|
8 × 16 cm (128)
|
None
|
Characteristics of Biological Implant
Thoracic wall reconstructions were performed using collagen matrix patches (Permacol;
Covidien, Mansfield, Massachusetts, United States; 1.5 mm thickness in sizes of up
to 15 × 20 cm) ([Figs. 1] and [2]). This collagen matrix is derived from porcine dermis in which cells, cell debris,
DNA, and RNA were removed. The resulting acellular matrix with its constituent collagen
fibers is cross-linked with hexamethylene diisocyanate for additional stability and
reduction of collagenase degradation. The biomechanical characteristics of several
collagen materials have been tested in experimental settings. The native collagen
patch used in this series sustained a maximum load of 317.2 ± 23.6 N during uniaxial
tensile testing. The tensile strength per unit width was 105.7 ± 7.9 N/cm and stiffness
58.3 ± 4.0 N/mm (mean ± standard error of the mean). One month after abdominal hernia
repair in a porcine animal model, the maximum tolerated force was 35.6 ± 3.5 N compared
with the native abdominal wall with 16.8 ± 1.8 N, similarly the tensile strength was
11.9 ± 1.2 N/cm compared with 5.6 ± 0.6 N/cm and stiffness values were 5.8 ± 0.9 N/mm
compared with 1.1 ± 0.2 N/mm.[10]
[11]
[12]
[14]
[18]
[19]
Fig. 1 Complete sternum resection in a patient with chondrosarcoma (left panel). Subsequent
sternum reconstruction with a 15 × 20 cm porcine dermal derived patch (right panel).
Fig. 2 Extended thoracic wall resection in a patient with osteosarcoma of the ninth left
rib (upper panels). Partial diaphragm resection and pulmonary wedge resections were
necessary. The chest wall and diaphragm defects were reconstructed using a 15 × 20
cm Permacol (Covidien, Mansfield, Massachusetts, United States) patches (lower panel).
A pedicled latissimus dorsi transfer flap was used for soft-tissue coverage.
Patients and Surgical Techniques
Indications for surgery were resection of primary chest wall sarcomas (n = 5) or sarcoma metastasis (n = 1). Patient data are listed in [Table 1]. Lateral chest wall resections were performed via thoracotomies to ensure wide safety
margins (n = 4). Complete sternal resections were performed through a midline incision ([Figs. 1] and [2]).
Before treatment, all patients had undergone biopsy to determine definite histopathological
diagnosis. All tumor surgeries were en bloc resections, which included the biopsy
sites.
Resections were performed with a minimum margin of 4 cm for ribs, sternum, and soft
tissue. The underlying pleura was always resected, as were adherent parts of the lung,
diaphragm, and mediastinum ([Table 1]).
Thoracic cavities were drained with 28 Charrière chest tubes. Patches were shaped
with scissors to match the resected defect with an overlap of approximately 1 cm on
all margins to ensure complete coverage. Under tension the patch was anchored with
multiple single-tied coated nonabsorbable polyethylene terephthalate sutures (Ethibond;
Ethicon, Norderstedt, Germany) to neighboring ribs and transosseously drilled holes
in the adjoining rib stumps. The same sutures were additionally used as running stiches
([Fig. 3]) to increase tension and to approximate soft tissues to the patch.
Fig. 3 Computed tomography (CT) scan of patient with osteosarcoma of the ninth rib showing
the extent of the tumor mass (upper left panel). Postoperative follow-up CT scan after
thoracic wall resection and reconstruction after 6 months (upper right panel) and
after 12 months (lower panel).
Redon drainages were placed on top, and left in place for at least 7 days. Adjacent
muscles, subcutaneous tissue, and skin were mobilized, and approximated with polyglactin
(Vicryl, Ethicon, Norderstedt, Germany) sutures. In two cases, soft-tissue coverage
was augmented by transposition of lifted muscle flaps.
Before incision, a single shot of cefuroxime was administered for perioperative antibiotic
prophylaxis. All patients received continuous postoperative analgesia via a peridural
catheter. Patients were mobilized without restrictions from postoperative day 1. Clinical
wound assessment, laboratory tests, and radiological imaging were performed according
to routine surgical procedures.
Follow-Up
Chest wall reconstructions were evaluated during postoperative oncological follow-up
visits at 3, 6, 12, and 24 months. The subjective impressions or complaints from the
patients were documented. Wound healing and soft-tissue coverage were examined. Chest
wall stability was tested using deep breath maneuvers and coughing tests. The stability
was classified as excellent, good, medium, or poor according to the surgical consultant's
impression. Routine oncological CT scans were used to evaluate the patches for structural
changes and integrity.
Histopathology and Immunohistochemistry
Formalin fixated and paraffin embedded sections of the explanted patch material were
stained for hematoxylin and eosin and periodic acid Schiff reaction. Immunohistochemistry
(IHC) was used to stain for CD31 positive capillary endothelial cells indicating neoangiogenesis.
IHC was performed by means of a standardized avidin-biotin complex method.
Results
Resection of chest wall sarcomas, including complete removal of the sternum in two
cases, was successfully performed in all patients. Details of patient data and surgical
procedures are listed in [Tables 1] and [2]. A minimal area of three ribs with adjacent soft tissues was resected. Mean size
of the defects was 149 cm2 (range: 80–252 cm2). Complete resection (R0) and wide safety margins were confirmed by histopathological
examination in all patients.
Table 2
Surgical procedures and follow up
|
Patient
|
Resection
|
Additional resections
|
Reconstruction
|
Stability
|
Complications
|
Follow-up
|
|
Allomaterial
|
Soft tissue coverage
|
|
1
|
Chest wall tumor en bloc resection including skin spindle, musculature, partial resection
of the 8th–10th ribs
|
Partial diaphragm resection, pulmonary wedge resection left lower lobe
|
15 × 20 cm × 1.5 mm Permacol and 10 × 10 cm × 1.5 mm Permacol for diaphragm
|
Latissimus dorsi flap graft
|
Excellent
|
None
|
Reoperation for local tumor recurrence at 14 mo, death at 32 mo due to distant metastasis
|
|
2
|
Chest wall tumor en bloc resection including the seventh—ninth ribs
|
None
|
10 × 10 cm × 1.5 mm Permacol
|
Direct (mobilized subcutaneous tissue and skin)
|
Good
|
None
|
Tumor relapse, patient alive under chemotherapy 42 mo
|
|
3
|
Complete sternum resection including skin
|
Partial resection of pleura and pericardium
|
15 × 20 cm × 1.5 mm Permacol
|
Pectoralis muscle
|
Good
|
None
|
Lost to regular follow-up, no CT scans, patient alive, no tumor relapse after 37 mo
|
|
4
|
Chest wall tumor en bloc resection including the complete seventh rib, partial resection
of the sixth and eighth ribs
|
None
|
15 × 20 cm × 1.5 mm Permacol
|
Direct (mobilized subcutaneous tissue and skin)
|
Good
|
None
|
No local tumor recurrence, death at 17 mo due to cerebral metastasis
|
|
5
|
Chest wall tumor en bloc resection including partial resection of the fourth–sixth
ribs
|
Pulmonary wedge resections right middle and lower lobes
|
15 × 20 cm × 1.5 mm Permacol
|
Direct (mobilized subcutaneous tissue and skin)
|
Moderate
|
None
|
No local tumor recurrence, death at 8 mo due to pelvic tumor relapse
|
|
6
|
Complete sternum resection
|
Partial resection of right pleura and pericardium
|
15 × 20 cm × 1.5 mm Permacol
|
Direct (mobilized subcutaneous tissue and skin)
|
Excellent
|
None
|
No local tumor recurrence, patient alive after 30 mo
|
The collagen patch material proved to be implantable under tension. The material was
pliable but not stretchable, and resulted in immediate stability of the thoracic wall.
Mobilized subcutaneous tissue and skin were sufficient to cover the patches in four
patients. Pediculate muscle flaps were necessary for two patients (Musculus latissimus
dorsi and bilateral M. pectoralis major flap). Primary skin closure was achieved in
all patients. Mean duration of surgery was 178 (range: 110–305) minutes.
All patients were extubated in the operating room and transferred to the intensive
care unit (n = 3) or surgical ward (n = 3). The postoperative course was uncomplicated in all cases. In particular, no
early or late respiratory impairment was observed. No postoperative bleeding, no infections,
or wound healing problems were encountered. Inflammatory markers peaked in average
(C-reactive protein [CRP] 29.2 mg/dL; white blood cell count [WBC] 12,950/µL) on postoperative
day 2. Mean CRP value at 3 weeks postoperatively was 1.2 (range: 0.5–4) mg/dL and
mean WBC was 7,270 (range: 4,310–11,500)/µL. Chest drains were removed after a mean
of 5 (range: 2–8) days. Mean hospital stay was 10 (range: 8–15) days. The 30-day mortality
was 0%.
Mean duration of follow-up was 27.6 (range 8–42) months. None of the patients suffered
from subjective respiratory limitations or complaints. No fluid accumulation inside
or outside the thorax, no excessive fibrous tissue formation, or signs of herniation
were observed ([Table 2]).
Following lateral chest wall resections, provoking tests caused only minimal movements
of the patch area. At 3-month follow-up, the stability was rated as good (n = 2) and excellent (n = 2).
Initially, both cases of complete sternal reconstruction presented with noteworthy
visible respiratory movements of the anterior chest. Neither of the patients suffered
from respiratory impairment. In the course of follow-up, the stability improved significantly.
At 3-month follow-up, the stability was rated good (n = 1) and moderate (n = 1) by the surgeon under deep breath and coughing tests ([Table 2]). In all six patients, no change in stability was observed at follow-up visits beyond
3 months. On routine follow-up CT scans, all implanted patches were identified as
intact without bulging, herniation, rupture, or loss of structural integrity. The
appearance of the patch material in the CT scans did not change during follow-up.
No excessive encapsulation or seroma formations were evident. A representative series
of follow-up CT scans demonstrating the intact patch material is shown in [Fig. 3].
Follow-up and oncological outcome are shown in [Table 2]. Despite wide safety margins, patient 1 developed a local tumor relapse and underwent
reoperation 14 months after his first tumor resection. The primary implanted patch
was removed en bloc with the recurrent sarcoma. During gross and histological examination,
the patch proved to be intact and well integrated into the surrounding tissues ([Fig. 4]). IHC staining for CD31 showed remarkable neovascularization of the collagen matrix
within the surrounding tissue. This observation is a substantial indication for biological
integration of the collagen matrix patch.
Fig. 4 Hematoxylin and eosin stain with immunohistochemistry of CD31 for neovascularization
in explanted collagen matrix patch 14 months after initial implantation. Intimal cells
are stained red. Tenfold magnification on left panel and 40-fold on right panel.
Upon returning to his mid-eastern country of residence 4 months postoperatively, patient
3 was lost to personal follow-up. According to e-mail contact, the patient is healthy,
without local impairments and is free of tumor recurrence.
Discussion
Following its introduction nearly 30 years ago, PTFE has emerged as the most frequently
applied material for chest wall reconstructions. It allows for implantation under
tension and results in a durable replacement of bony and soft tissues. Some surgeons
even used methylmethacrylate sandwich techniques or large metal prostheses to create
sufficient stability in large defects, especially in cases of sternal resections.[2]
[3]
[4]
[5]
[6]
[17] Nevertheless, reconstructions with synthetic materials are associated with serious
complications. Synthetic materials are known to cause seroma formation and excessive
encapsulation by dense fibrous tissue. The latter may not only result in patient discomfort
and impaired respiration but also in serious deformities such as scoliosis following
chest wall replacement in children.[15]
[16]
More importantly, PTFE materials do not overcome the main risk of all synthetic durable
implants: infection. Although low complication rates have been described in some series,
infection is a common problem, particularly in large reconstructions and sternal replacements
ranging from 5 to 12%.[2]
[3]
[4]
[5]
[6]
[8]
[9] Therefore, effective soft-tissue coverage with muscle flaps is essential, when applying
synthetic allomaterials. A considerable number of synthetic patches had to be revised
or explanted for these reasons. Mansour et al report 7% of flap complications or loss,
and another 5% of surgical site infection.[4] In 2006, Weyant et al showed that within 30 days complication rates were 38% for
rigid methylmethacrylate sandwich techniques and a 4.5% 90-day prosthesis removal
rate was observed. For PTFE or polypropylene mesh, the 30-day mortality was 27% and
the 90-day prosthesis removal rate was 4.1%.[6] Supposedly, absorbable or composite meshes have a reduced infection risk. However,
these implants may result in significant mid- and long-term instabilities, paradoxical
respiratory movements, and finally herniation in the area of the reconstructed chest
wall.[1]
[3]
[4]
[8]
The rationale for employing collagen matrix patches in thoracic wall reconstruction
is to combine the rigidity and durability of nonabsorbable synthetic materials with
a reduced risk of infection.[15]
[19]
A growing variety of biological collagen matrix meshes is commercially available.
Unlike permanently fixated biological materials, this novel generation of materials
allows for a remodeling process with neovascularization and complete integration.
This process starts with the migration of host cells and the formation of new blood
vessels. In this respect, our histological findings ([Fig. 4]) in a patch explanted after 14 months support the data from other reports.[7]
[19]
Our positive experiences with respect to strength, durability, and resistance to infection
with this type of biological implant for thoracic wall reconstruction are in accordance
with experimental and clinical findings of others.[15]
[16]
[17]
[20] Most of the available data stem from the field of abdominal hernia repair.[7]
[10]
[14]
[19] Only very few reports describe the application of collagen matrix patches in thoracic
surgery.[15]
[16]
[17]
[20] In abdominal hernia surgery, biological materials proved to be superior even to
synthetic meshes with an overall success rate of 90%.[21] However, to our knowledge, no experimental study exists which compares the biomechanical
characteristics of collagen matrixes and synthetic materials directly. In an animal
study of hernia repair contaminated by staphylococci, all cases with synthetic implants
failed, whereas 67% of the animals with the biological patch type used by us experienced
uncomplicated wound healing.[22] Clinical studies in patients with contamination and overt infection of the surgical
field resulted in similar outcomes.[23]
[24]
However, for reconstruction of the thoracic wall characteristics of ideal materials
and techniques of implantation are different. To minimize paradox respiratory movements,
a patch must be implanted under tension and should not be stretchable. The ideal characteristics
and strength of the biological material should resemble that of 1- to 2-mm-thick PTFE
membranes. Compared with other collagen matrixes, the product used in this study provides
very good strength and durability owing to a thickness of 1.5 mm.
At this point, we would like to stress that origin and processing of a collagen matrix
determines remodeling, stability, and durability. Whereas reduced susceptibility to
infection is a common attribute of all these implants, strength and durability are
not.
Tissues derived from dermis, intestinal submucosa, or pericardium of porcine, bovine,
and human origin have been explored. The process of chemical cross-linking restricts
the extent of degradation by collagenase digestion and the turnover of collagen fibers
following implantation. On the contrary, cellular infiltration, neovascularization,
and integration are delayed as well.[7]
[17] Compared with a collagen matrix derived from porcine dermis, significantly higher
failure rates were reported for non–cross-linked human acellular dermis in a systematic
review of the published series of hernia repair. Non–cross-linked collagen mesh from
human dermis, which undergoes extensive remodeling and is known to stretch with time,
seems to be less suitable for use in thoracic surgery.[25]
Application of collagen matrixes may also be beneficial for pediatric patients. Deformities
of the spine are a well-known complication following thoracic wall reconstruction
in children.[15] Lin et al demonstrated that chest wall reconstructions with collagen matrix patches
in pediatric patients did not lead to the development of clinically relevant scoliosis.
Handling and tailoring of the thick collagen patch was easy, it proved to be blood
tight and a safe abutment for sutures. No seroma formation, wound healing problems,
or infections were observed in this pilot study. The clinical, radiological, and histological
findings in this study support long-term stability and effective integration of this
type of collagen matrix implant. Our experience demonstrates that a porcine dermal–derived
acellular collagen matrix is a viable alternative to PTFE for thoracic wall reconstructions.
However, it is evident that this pilot study only allows limited conclusions. Also,
the advantages of collagen matrix materials must be weighed against the considerably
higher expense. Nevertheless, in complicated settings such as large defects, total
sternal replacement, difficult soft-tissue coverage, contaminated surgical field,
or pediatric surgeries, implantation of this biological patch appears to be justified.
