Orthopaedics & Traumatology: Surgery & Research (2013) 99, 959—964
Available online at
ScienceDirectwww.sciencedirect.com
ORIGINAL ARTICLE
Histological characteristics of inducedmembranes in subcutaneous, intramuscularsites and bone defect
H. Liua,1, G. Hub,1, P. Shangc, Y. Shena, P. Niea, L. Penga,b,d,∗,H. Xua
a Department of Orthopaedic Surgery, Second Affiliated Hospital of Wenzhou Medical University, 109,Xueyuanxi Road, Wenzhou 325000, Chinab Trauma Center, Affiliated Hospital of Hainan Medical College, 31, Long Hua Road, Haikou 571100, Chinac Department of Rehabilitation, Second Affiliated Hospital of Wenzhou Medical University, 109, XueyuanxiRoad, Wenzhou 325000, Chinad Department of Orthopaedic Surgery, Hainan Provincial People’s Hospital, 19, Hua Xiu Road, Haikou570311, China
SummaryBackground: The induced membrane technique was proposed as a treatment of large seg-mental bone defects. The influence of the surrounding tissues on its characteristics remainsunknown. It is therefore not known which kind of plastic surgery procedure (muscular or facio-cutaneous flap) would optimize bone osteointegration within a bone defect reconstructed usingthe induced-membrane technique.Hypothesis: We hypothesized that membrane characteristics could be influenced by the soft-tissue environment either subcutaneous or muscular.Objective: To evaluate the histological characteristics of poly-methylmethacrylate (PMMA)induced membranes in intramuscular, subcutaneous and bony environment (radius defects) at2 steps: spacer implantation; secondary bone graft and its subsequent osteintegration afterspacer removal.
Methods: PMMA-induced membranes were obtained in the three sites of 15 rabbits. Subse-quent new bone formation was studied in the same environments in 24 other rabbits. Sixweeks after the initial implantation, PMMA spacers were replaced with iliac autografts. Animalswere euthanized at 2, 4, and 8 weeks postoperatively. Tissue samples were harvested and
∗ Corresponding author. Department of Orthopaedic Surgery, Second Affiliated Hospital of Wenzhou Medical College, 109, Xueyuanxi Road,Wenzhou 325000, China. Tel.: +86-577-88829799; fax: +86-577-88816191.
stained with hematoxylin and eosin. The histological characteristics of the membrane (thicknessand microvessel density) and the newly-formed bone (cortical thickness) were quantitativelyanalyzed.Results: The membranes in the subcutaneous sites developed quicker, were thicker and hadthe lowest microvessel density (P < 0.01). The membranes in the intramuscular sites developedlater and were thinner (P < 0.01). The membranes in the osseous defects had the greatestmicrovessel density (P < 0.01). After bone grafting, induced membranes became thinner andtheir microvessel density decreased substantially, but maintained better in osseous site. Thenewly-formed bone that developed in the radius defects, had the thickest cortices (P < 0.01).Conclusions: The evolution of membranes induced in the intramuscular and subcutaneous envi-ronments was close to that of the bone defect model, although bone formation appearedweaker.Level of evidence: Basic science study III.
he induced membrane technique described by Masquelett al. has emerged as a well-established technique of recons-ructing large bone defects of the tibia, femur, humerus,and, wrist, ulna and mandibular [1—5]. In contrast to theemur and tibia that are covered by muscles, the mandibularnd wrist bones are subcutaneous. These anatomic dif-erences may be of importance, because surrounding softissues are hypothesized to influence the histological char-cteristics of membranes [6]. In particular, plastic surgery isften needed in association with this large bone reconstruct-ons [6,7] and it is therefore not known which kind of plasticurgery (muscular or facio-cutaneous flap) would optimizeone osteointegration within a bone defect reconstructedsing the induced-membrane technique.
To the best of our knowledge, no study has yet investi-ated the difference of induced-membrane characteristicsn these different environments (intramuscular or subcuta-eous). Furthermore, it is not known how much time inducedembranes would maintain their histological characteristics
nd effect on bone healing following bone grafting.We hypothesized that membrane characteristics would
e influenced by the soft-tissue environment subcutaneousr muscular. To test this hypothesis, the histological charac-eristics of induced-membrane and of the subsequent boneormation in subcutaneous, intramuscular and osseous sitesere compared. The protocol was built so as to respond to
questions:
is the model of paravertebral implantation (subcutaneousand intramuscular) realistic, compared to a real bonedefect model of the distal radius?
are the membrane characteristics in the 2 sites compara-ble (subcutaneous and intramuscular)?
what are the degradation characteristics of induced mem-brane in the 3 sites after completion of the bone graft?
ethod
nimal
hirty-nine New Zealand white rabbits from the experi-ental animal center of Wenzhou Medical University were
atfiw
rights reserved.
perated on (avg wt: 3.0 kg) under general anesthesia using solution of Chloral Hydrate (Solarbio, China) in semi-sterileonditions.
urgical procedure
n 15 rabbits, PMMA cylinders (4.5 mm in diameter, 15 mm inength) were moulded ex vivo and placed into bilateral sub-utaneous, intramuscular sites and radial defects. Two PMMAylinders were respectively placed in paravertebral musclesnd in the subcutaneous tissues of the lumbar area. In theadius defect model serving as a control, a 3-cm longitudi-al incision was made along the lower third of the forearm.
15-mm bone resection was performed in each distaladius, before being filled using a PMMA cylinder. Penicillin0,000 IU/kg was administered intramuscularly immediatelyreoperatively and 24, 48 hours postoperatively. The oper-ted limbs were immobilized for 2—8 weeks using splints.ive animals were euthanized and the ten specimens of eachite served for the histological examination at 2, 4 and 8eeks postoperatively, respectively.
Twenty-four other rabbits served for studying the evo-ution of the membrane after bone graft completion. Thenimals underwent unilateral implantation of PMMA cylin-ers in the three sites. The cylinders were then removedt 6 weeks and the voids inside the induced membraneere filled in with morcelized bone autografts harvested
rom the iliac crests. Finally the membranes incisions wereutured using 4-0 prolene. Postoperative anti-inflammatoryrugs (Penicillin 20,000 IU/kg) were ordered and a splintmmobilization was used. Eight animals were euthanized andhe eight specimens of each site served for the histologicalxamination at 2, 4 and 8 weeks postoperatively, respec-ively.
istological examination
issue samples were fixed in 10% formalin. Samples con-aining bone tissue were additionally demineralized in a0% formic acid solution. Cross-sections of the membrane
nd longitudinal sections through the mid-sagittal plane ofhe bone samples were obtained and embedded in paraf-n, stained with hematoxylin and eosin, and examinedith a microscope (Olympus, Japan) and photographed
Histological characteristics of induced membranes 961
Figure 1 Photomicrographs of representative sections show the histological changes of membrane around a PMMA cylinder inintramuscular tissue (a), subcutaneous tissue (b) and radial defect (c) of rabbits in the phase of membrane formation. 1 = 2 weeks;
(Sony, Tokyo, Japan). There were 10 specimens in thefirst series and 8 specimens in the second series in eachsite at each time point. Three measurements were carriedout in 3 separate fields of each specimen and averaged.Computed-assisted measurements of the histological param-eters (membrane thickness, microvessel density and corticalbone thickness) were obtained using an automated imageanalysis system (Image-Pro Plus, Media Cybernetics, SilverSpring, MD, USA). Membrane thickness was defined as thedistance separating the cement from the adjacent connec-tive tissues. Microvessel density was defined as the numberof microvessels per membrane area in microscopic fields(×400). Cortical bone thickness was defined as the dis-tance between the marrow cavity and the membrane. Thedifference of membrane thickness and microvessel densitybetween the end and the beginning of each time intervalwere calculated, so as to determined the speed of mem-brane growth and microvessel proliferation per week.
Statistical analysis
Statistical analysis was performed with SPSS16 software
(SPSS Inc., Chicago, IL). The non-parametric Kruskal-Wallistest was used to examine differences in membrane thick-ness and microvessel density among different sites at eachtime point. Cortical bone thickness was compared among
t7
2
lary; ST, subcutaneous tissue. Hematoxylin and eosin staining;
ifferent sites using one-way analysis of variance. Theeast-significant difference (LSD) test was used for pairwiseomparison. Differences were considered as statistically sig-ificant with a P-value of less than 0.05.
esult
nduced membranes in the distal radius were thinnerut displayed a greater microvessel density. Substantialifferences appeared between the 2 subcutaneous models,he former in the paravertebral region, the latter in theistal radius. Microvessel density was lower and progressedlower in the non-osseous site. Mean membrane thicknessas 256.31 ± 92.52 �m, and 178.44 ± 50.94 �m in the
ubcutaneous and the distal radius, respectively (P < 0.01,igs. 1 and 2). The mean speed of thickness increaseithin the first 4 weeks was 50.61 ± 21.85 �m/week, and4.63 ± 6.45 �m/week respectively (P < 0.01). Mem-ranes showed a maximum microvessel density of.03 ± 2.09 × 104/�m2, and 32.71 ± 8.10 × 104/�m2 inhe subcutaneous, and the distal radius respectively. Theean speed of microvessel density proliferation within
he first 4 weeks was 1.76 ± 0.52 × 104/�m2 week and.05 ± 1.35 × 104/�m2 week respectively (P < 0.01).
There were also substantial differences between the non-osseous sites. Induced membranes were thicker in
962
Figure 2 Change of membrane thickness and microvesseldensity in the phases of membrane formation and degradation.The figure shows the evolution of membrane formation and ofmembrane degradation over time in the 3 sites of implantation.Tt
td5ms5Mosmt3
iiont4n
r5cAaiC
D
TmaItisbiipsgm
ttmoaAafimi
rmtmenttsnwRiaf
dbdtt
he y-axis indicates the mean value in micrometers (±1SD) ofhe induced membrane thickness, and microvessel density.
he subcutaneous site and displayed a lower microvesselensity. Mean membrane thickness was 256.31 ± 92.52 �m,5.43 ± 10.41 �m in the subcutaneous, and in the intra-uscular site respectively (P < 0.01, Fig. 1). The mean
peed of thickness increase within the first 4 weeks was0.61 ± 21.85 �m/week, 13.85 ± 2.60 �m/week (P < 0.01).embranes showed a maximum microvessel densityf 7.03 ± 2.09 × 104/�m2, 15.32 ± 3.2 × 104/�m2 in theubcutaneous and intramuscular site respectively. Theean speed of microvessel density proliferation within
he first 4 weeks was 1.76 ± 0.52 × 104/�m2 week and.83 ± 0.80 × 104/�m2 week respectively (P < 0.01).
In the rabbits that received a bone graft within theirnduced membrane, a significant and constant decreasen membrane thickness and their microvessel density wasbserved (Figs. 2 and 3). It was quicker in the subcuta-
eous site and slower in the intramuscular site than inhe osseous site. Four-weeks membrane thicknesses were.72%, 59.34% and 17.04% of the original membrane thick-ess in the subcutaneous, intramuscular and osseous site,
plpi
H. Liu et al.
espectively. The 2-week microvessel densities were 14.90%,8.10% and 14.29% of their original values in the sub-utaneous, intramuscular and osseous site, respectively.lthough their membranes underwent the slower degener-tion, bone cortices reabsorbed (Figs. 3 and 4) significantlyn the intramuscular site only (P < 0.01). Finally (8 th week),ortical thickness was best preserved in the distal radius.
iscussion
he main findings of the present work are that the inducedembrane formed quicker in the non-osseous site, but they
ppeared less active, with a lower microvessel density.t suggests that the subcutaneous paravertebral implanta-ion may underestimate the real osteogenic properties ofnduced membranes developing in a real bone defect. Theame difference was observed between the 2 paraverte-ral implantations, respectively in the subcutaneous and thentramuscular site. Thicker induced membrane was observedn subcutaneous sites, which may play a mechanical role inreventing soft tissue protrusion [2], subsequently trying tolow down bone resorption. In fact bone resorption was thereatest in the intramuscular site, which had the thinnerembrane.As the membrane in different sites showed different
hickness in our study, a possible explanation was warran-ed: the tissue origin of fibroblasts is a crucial factor forembrane thickness. Fibroblasts are the major components
f the induced membrane [8] and induced membrane is product of soft-tissue reaction to the PMMA spacer [9].
significant cellular response is activated when implantsre placed in the subcutaneous site [10], leading to thickerbrous membrane. Whereas, in intramuscular environ-ents, only early-appearing myogenic cells differentiate
nto myofibroblasts [11], leading to a relatively thinner one.Second, although the paravertebral implantation rep-
esents an imperfect model of segmental bone defects,embrane characteristics follow the same evolution over
ime. These results suggest that a subcutaneous environ-ent would produce less bone resorption than a muscular
nvironment. It shows that the induced membrane tech-ique suits to subcutaneous defects, and looks to justifyhe use of facio-cutaneous flap associated with the skele-on reconstruction in the treatment of large post-traumaticegmental defects. Previous studies [12—14] reported thato difference in the rate of complications and failuresas found between muscular and facio-cutaneous flaps.egarding to their osteogenic properties, further clinicalnvestigations should be conducted to examine the char-cteristic of induced membranes covered by muscular oracio-cutaneous flaps.
Third, the membranes in each site exhibited significantegradation after bone graft implantation inside the mem-rane cavity. Although the membranes in the osseous siteisplayed a moderate thickness, they had a slow degrada-ion speed, and induced the stronger bone formation. Inhe contrary, the membrane in non-osseous site did not
revent graft resorption in particular in the intramuscu-ar environment. It is possible that membrane degradationrovoked direct bone graft contact with the surround-ng soft tissues, thus favoring bone resorption. Clinical
Histological characteristics of induced membranes 963
Figure 3 Photomicrographs of representative sections show the histological changes of membrane degradation and new boneformation in the phase of membrane degradation in intramuscular tissue (a), subcutaneous tissue (b) and radial defects (c) ofrabbits. 1 = 2weeks; 2 = 4weeks; 3 = 8weeks (ST, subcutaneous tissuegraft; C, capillary; EO, endochondral ossification. Hematoxylin and e
Figure 4 Evolution of cortical bone thickness following bonegrafting inside the induced membrane in the 3 sites of implan-tation. The y-axis indicates the thickness of new cortical bonein micrometers (±1SD). The control is represented by the meanthickness of non-operated distal radius in the rabbit.
eglamavnt
tidtiacsfrt
D
Tc
; IM, induced membrane; M, muscle; NB, new bone; BG, boneosin staining; original magnification 200×).
xperience also suggested that morsellized autologous boneraft reabsorbed if placed within a well-vascularised muscu-ar environment [15,16]. The microvessel density decreasedfter bone graft, raising the question of the capacity of theseembranes to promote durably bone vascularization. Some
uthors suggested that induced membranes could facilitateascularization of bone graft [5,8], because they observedo resorption of the within-membrane grafts [6], in the con-rary to other studies [4,17].
Based on the current study, we showed that the soft-issue environment influenced the characteristics of thenduced-membrane, as well in the formation than in theegradation phase. The subcutaneous paravertebral implan-ations may not represent an ideal model of the bone defectsn the osseous sites. Although their membrane were thicker,nd their degradation speed in the subcutaneous site waslose to that of the distal radius, they had a lower microves-el density, and were finally not able to optimize boneormation as well as the induced membrane of the distaladius. In addition, the phenomenon of membane degrada-ion cannot be ignored and its maintenance is challenging.
isclosure of interest
he authors declare that they have no conflicts of interestoncerning this article.
9
A
TOCvN
R[
[
[
[
[
[
[
64
cknowledgments
he authors thank all the staff in the Laboratory ofrthopaedic Research Institute and Scientific Researchenter of Second Affiliated Hospital of Wenzhou Medical Uni-ersity. This study was supported by grants from the Nationalatural Science Foundation of China (31060135/C100302).
eferences
[1] Masquelet AC, Fitoussi F, Begue T, Muller GP. Reconstruction ofthe long bones by the induced membrane and spongy autograft.Ann Chir Plast Esthet 2000;45:346—53.
[2] Masquelet AC, Begue T. The concept of induced membranefor reconstruction of long bone defects. Orthop Clin North Am2010;41:27—37.
[3] Masquelet AC, Obert L. Induced membrane technique for bonedefects in the hand and wrist. Chir Main 2010;29:221—4.
[4] Viateau V, Guillemin G, Bousson V, Oudina K, Hannouche D,Sedel L, Logeart-Avramoglou D, Petite H. Long-bone critical-size defects treated with tissue-engineered grafts: a study onsheep. J Orthop Res 2007;25:741—9.
[5] Zwetyenga N, Catros S, Emparanza A, Deminiere C, SiberchicotF, Fricain JC. Mandibular reconstruction using induced mem-branes with autologous cancellous bone graft and HA-betaTCP:animal model study and preliminary results in patients. Int JOral Maxillofac Surg 2009;38:1289—97.
[6] Masquelet AC. Muscle reconstruction in reconstructive surgery:soft tissue repair and long bone reconstruction. Langenbecks
Arch Surg 2003;388:344—6.
[7] Uzel AP, Lemonne F, Casoli V. Tibial segmental bone defectreconstruction by Ilizarov type bone transport in an inducedmembrane. Orthop Traumatol Surg Res 2010;96:194—8.
[
H. Liu et al.
[8] Pelissier P, Masquelet AC, Bareille R, Pelissier SM, Amedee J.Induced membranes secrete growth factors including vascularand osteoinductive factors and could stimulate bone regener-ation. J Orthop Res 2004;22:73—9.
[9] Ryhanen J, Kallioinen M, Tuukkanen J. In vivo biocompatibilityevaluation of nickel-titanium shape memory metal alloy: mus-cle and perineural tissue responses and encapsule membranethickness. J Biomed Mater Res 1999;41:481—8.
10] Hadfield G. The tissue of origin of the fibroblasts of granulationtissue. Br J Surg 1963;50:870—81.
11] Li Y, Huard J. Differentiation of muscle-derived cells intomyofibroblasts in injured skeletal muscle. Am J Pathol2002;161:895—907.
12] Yazar S, Lin CH, Lin YT, Ulusal AE, Wei FC. Outcome compar-ison between free muscle and free fasciocutaneous flaps forreconstruction of distal third and ankle traumatic open tibialfractures. Plast Reconstr Surg 2006;117:2468—75.
13] Wettstein R, Schurch R, Banic A, Erni D, Harder Y. Review of 197consecutive free flap reconstructions in the lower extremity. JPlast Reconstr Aesthet Surg 2008;61:772—6.
14] Danino AM, Gras M, Coeugniet E, Jebrane A, Harris PG.Is muscle the best coverage for leg Gustilo IIIb fractures?A retrospective comparative study. Ann Chir Plast Esthet2008;53:473—9.
15] Ermis I, Poole M. The effects of soft tissue coverage on bonegraft resorption in the craniofacial region. Br J Plast Surg1992;45:26—9.
16] Romana MC, Masquelet AC. Vascularized periosteum associ-ated with cancellous bone graft: an experimental study. PlastReconstr Surg 1990;85:587—92.
17] Catros S, Zwetyenga N, Bareille R, Brouillaud B, Renard M,Amedee J, Fricain JC. Subcutaneous-induced membranes haveno osteoinductive effect on macroporous HA-TCP in vivo. JOrthop Res 2009;27:155—61.
