Hernia DOI 10.1007/s10029-014-1220-1
ORIGINAL ARTICLE
Characterisation of the cellular infiltrate in the foreign body granuloma of textile meshes with its impact on collagen deposition U. Klinge • U. Dietz • N. Fet • B. Klosterhalfen
Received: 12 July 2013 / Accepted: 18 January 2014 Ó Springer-Verlag France 2014
Abstract Purpose As part of the foreign body reaction, mesh filaments are surrounded by an infiltrate of inflammatory cells. Though macrophages are considered as being predominant, little is known about the origin of other cells. Methods On 55 meshes explanted from humans, we characterised the cells in the inflammatory infiltrate of the granuloma by immunohistochemistry using 10 cellular markers: CD3? lymphocytes, CD4? T helper cells, CD8? cytotoxic T cells, CD20? B lymphocytes, CD34? stem cells, CD45R0? leucocytes, CD68? macrophages, Mib1 for proliferation, Vimentin for mesenchymal origin, and Desmin for myocytes. Collagen deposits were analysed after staining with Sirius Red. Results More than 80 % of the cells in the infiltrate showed a positive expression of CD68, CD8, CD45R0 and Vimentin. CD4 and Desmin were seen in 30–80 % of the cells, unaffected by material or time. A score summarising the expression of all markers positively correlated significantly with an increased percentage of collagen type III (green) in the mesh wound. The analysis of collagen deposits was only affected to a small degree by size of area for investigation. U. Klinge (&) N. Fet Department of General, Visceral and Transplant Surgery, University Hospital of the RWTH Aachen, Pauwelsstraße 30, 52074 Aachen, Germany e-mail: [email protected] U. Dietz Department of General- Visceral- Vascular- and Paediatric Surgery, University of Wuerzburg, Oberduerrbacher Strasse 6, 97080 Wu¨rzburg, Germany B. Klosterhalfen Institute for Pathology, Du¨ren Hospital, Roonstr 30, 52351 Du¨ren, Germany
Conclusions At the vicinity of the mesh filaments, the accumulated inflammatory cells represent a mixture of cells of various origins. The high expression of at least four markers requires co-expression of different surface markers and thus confirms the existence of multiple transition forms instead of dominance of just macrophages. This offers new options for interventions to attenuate the inflammatory reaction of mesh implants. Keywords Foreign body granuloma Scar collagen Inflammatory cell Hernia mesh
Introduction Any implantation of a foreign body (FB) into tissues is followed by a response of the host’s innate immune system. This is also the case for textile hernia meshes, whose filaments are completely encapsulated by a foreign body granuloma including cells and extracellular matrix. Different types of inflammatory cells migrate to the mesh–host interface, and eventually create a fibrotic capsule in an attempt to separate the mesh from surrounding tissues. Several studies have shown that the inflammatory response of tissues surrounding surgical meshes includes macrophages [1], [2], CD3? lymphocytes [3], [2], CD8? T lymphocytes [4], CD20? B lymphocytes [2], neutrophils [5, 6], mast cells [3], CD15? granulocytes, or foreign body giant cells [2]. Most of these studies, however, focussed on one or only some few cell types demonstrating the influence of mesh material and structure, the time elapsed since implantation, the reason for explantation, or the anatomical location of the mesh, respectively, on the type of cells present [7–9]. In humans, CD68? macrophages and in mice, F4/60? cells were found to be the predominant
123
Hernia
players of the local tissue reaction. However, the fact that many of these cells in the infiltrate of the FB granuloma do not show a positive expression of these markers and thus may not be considered as macrophages, suggests the involvement of other cells in the process of local wound healing, as well. The cell signalling in the vicinity of surgical meshes obviously influences the extracellular matrix and specifically the collagen deposition. Correspondingly, there is a strong relationship between the inflammatory process and the extent of fibrosis around implanted meshes [8]. The predominant extracellular component of scar is collagen, either the mature and cross-linked collagen type I or the immature collagen type III. A ratio of collagen type I/III above 3 is assumed to be an indicator of good quality of the scar tissue (fibrosis). Klosterhalfen et al. demonstrated by means of immunohistochemical staining of 623 explanted meshes, that the content of collagen type I and type III of the fibrosis in the granuloma showed an impaired wound healing with a reduced ratio in half of the sections. Complementary results were found by means of polarisation microscopy after Sirius red staining [10]. Junge et al. [11] described a decreased collagen ratio of 2.1 ± 1.4 on 87 explanted meshes. With Sirius red staining, collagen type I appears as red, whereas collagen type III as green; thus an automatic image analysis process was used, which analysed the distribution of red or green pixels, respectively, in small selected fields in the scar around the mesh filaments (0.1 mm2, 400 magnification). However, as the analysis was limited to a circumscriptive area adjacent to the mesh filaments, it does not reflect the collagen deposition within the pores in between the mesh filaments. The aim of this study was first, to characterise the cell types of the infiltrate beneath the filaments of explanted hernia meshes; and second, to describe the collagen ratio in the mesh wound including a representative scar area of about 8 mm2. Both approaches have not been previously described.
As reference for collagen in physiologic fascia, we took samples of eight patients, of which the tissue had been excised during laparotomy, with no scar tissue (Ethics Committee approval for fascia tissue biopsy, University of Wuerzburg, Protocol 060/10 from the 16th of November, 2010).
Materials and methods
Table 1 Antibodies and dilution used for identification of inflammatory cells
Fifty-five samples of hernia meshes explanted in the years 2001–2012, and previously fixed in paraffin blocks were analysed. The mesh explants were taken from 22 female and 33 male patients with a mean age of 57 years (range 29–89). The mean time elapsed between implantation and explantation was 3.1 years. The recovered meshes were reclassified according to Klinge and Klosterhalfen [9] into one of five classes. The samples included 23 meshes of class I (large pore), 14 of class II (small pore), 7 of class III (intra-abdominal position), and 11 of class V (plug). The reasons for explantation were pain in 10 patients, infection in 13 and hernia recurrence in 37 patients.
Antigene
Antibodies
Dilution
CD3
Menarini Klon PSI
1:200
CD20
Menarini Klon L26 B cell
1:50
CD34
Menarini Klon QBEnd/10
1:50
CD45RO
Menarini Klon UCHL1
1:200
CD68
Menarini Klon 514H12
1:40
CD4
DAKO Klon 4B12
1:200
CD8
DAKO Klon C8/144B
1:100
Desmin
DAKO Klon D33
1:50
123
Immunohistochemistry Mesh samples were stained for HE and for expression of the following cell markers:
Marker
Cell type
CD3
Lymphocytes
CD4
T helper cells
CD8
Cytotoxic T cells
CD20
B lymphocytes
CD34
Stem cells
CD45R0
Leucocytes
CD68
Macrophages
Mib1 Vimentin Desmin
Function
Proliferation Mesenchymal origin Myocytes
Lymphoid tissue served as positive control for CD3, CD4, CD8, CD20, CD34, CD45RO, CD68, Mib1, Vimentin, whereas myometrium for Desmin. Pretreatment with Citrate (pH 6) 1:100 was done for CD3, CD20, CD34, CD45RO, Mib1, Vimentin, 20 min at 96 °C. Pretreatment with Tris–EDTA (pH 9) 1:100 was done for 20 min at 96 °C for CD4, CD8, CD68, Desmin (see Table 1 for details of antibodies and dilution). The protocol of autostainer includes 30 min primary antibody 300 ll, 10 min H2O2 300 ll, 10 min secondary
Ki67 (Mib1)
DAKO Klon Mib1
1:100
Vimentin
Zytomed Klon V9
1:200
Hernia
antibody 300 ll, 15 min Poly-HRP-Goat anti Mouse/Rabbit/Rat IgG 300 ll, 10 min DAB 300 ll. Washing with TRIS-Buffer pH 7 with non-ionic detergent Tween 20. Counterstaining was done with haematoxylin (30 s, Sigma). Fixation was done in Xylol. Analysis was performed only within the cellular infiltrate of the scar tissue beneath the mesh filaments, coding the intensity of expression for positive cells according to the following score: I (\5 %), II (5–30 %), III (30–80 %) and IV ([80 %).
Sirius red staining Sections of 5 lm were stained for 1 h in Picrosirius solution (0.1 % solution of Sirius Red F3BA in saturated aqueous picric acid, pH 2) according to Junqueira et al. [10]. Sections were washed in 0.01 N HCl for 2 min, dehydrated, cleared, and mounted in synthetic resin. Visualisation by crossed Polaroid filters allowed estimation of collagen type I and type III at 409 magnification. A photograph was taken with 640 9 480 pixels (=307200 pixels), 16 million colours, and saved as jpg. The image reflects an area of 3.3–2.5 mm = 8.1 mm2, which was completely analysed for the proportion of red and green colours by use of ‘‘Image J 1.46r’’ (http://imagej.nih. gov/ij/). Furthermore, we performed an analysis of collagen at a magnification of 4009 reflecting an area of interest of 0.1 mm2 close to the filaments (mean of three measurements per specimen). The colour histogram of the analysed function provided the separate percentage of green, red and blue pixels in an image. The percentage of red as indicator of collagen I is calculated as percentage of red in relation to red and green pixels; the percentage of green as indicator of collagen III resulted from subtraction of red from 100. Statistical calculations were done with IBM SPSS statistics 20. In general, a p \ 0.05 was considered as significant. Tests were done as indicated in the text.
Results The 10 meshes explanted because of pain were recovered after a mean time of 4.3 years; in this group, the number of mesh class V (plugs) was significantly increased (Pearson Chi Square p = 0.005) (Table 2). The meshes explanted because of infection were recovered after a mean time of 2.4 years, without influence of the mesh type. Also recurrence was not related to the type of explanted meshes in 32 specimens, explanted after a mean of 3.1 years.
Table 2 Number of explanted meshes (n = 55) with reason for explantation (pain, infection or recurrence) and mesh class. More than one reason for explantation may be present Mesh class I
II
Total III
V
Pain No
22
12
6
5
45
Yes
1
2
1
6*
10
No
16
10
5
11
42
Yes
7
4
2
0
13
No
4
5
4
5
18
Yes
19
9
3
6
37
23
14
7
11
55
Infection
Recurrence
Total number of explants
* Increase for meshes of class V reaches significance (Pearson Chi Square p = 0.004)
Characterisation of the cellularity in the vicinity of the mesh Most of the markers were expressed on cells in the inflammatory infiltrate around the mesh filaments. We observed an expression of CD8, CD45R0, CD68, and of Vimentin on more than 80 % cells. Whereas CD34-positive cells were rarely seen within the granuloma infiltrate, a variable distribution of 5–80 % of the cells showed a positive expression of CD3, CD4, CD20, Desmin, or Ki67 (Table 3). The extent of the cellular infiltrate (less, equal or more than scar tissue) was significantly positively correlated with the number of CD3-positive lymphocytes in the infiltrate (r = ?0.353). The number of CD68 ? macrophages was significantly correlated with the number of CD4 ? T helper cells (r =?0.316). The cellular expression of Desmin was influenced by the mesh class with highest values for class III and lowest values for class I meshes (3.8 ± 0.2 versus 3.3 ± 0.5; p = 0.06). The expression of CD45R0 is inversely related with the risk for pain (Fisher’s exact test mean score of 4 ± 0 without pain versus 3.9 ± 0.1 with pain; p = 0.04), mainly because two of the ten meshes explanted due to pain showed a slightly lower expression of CD45R0 with a score of \4 (3.7). When summing up all expression scores of the 10 cellular markers, we got a mean value for the 10 scores of 29.7 ± 1.5 with a range 24–33 (possible maximum score = 40): Highest values were seen for class V meshes, whereas class I meshes showed the lowest values (ANOVA with Bonferroni adjustment for Post Hoc tests 29.0 ± 1.4 versus 30.6 ± 1.6, p = 0.025) (Table 4).
123
Hernia Table 3 Scores of expression for 10 analysed cell markers at the vicinity of the mesh filaments as minimum, maximum, mean and standard deviation Marker
Na
Cell type/indicator
Score Minimum
Maximum
Mean
Std. dev.
CD3
Lymphocyte
50
1.7
3.0
2.2
0.4
CD4
T helper cell
50
2.0
4.0
3.0
0.5
CD8 CD20
Cytotoxic T cell Activated B cell
50 50
4.0 1.0
4.0 3.0
4.0 2.0
0.0 0.5
CD34
Hematopoietic stem cell
50
1.0
1.0
1.0
0.0
CD45R0
Leucocytes
50
3.7
4.0
4.0
0.1
CD68
Macrophage
49
3.0
4
4.0
0.1
Desmin
Myocyte, myofibroblast
50
2.5
4.0
3.5
0.4
Ki-67
Proliferation
50
1.0
3.0
2.1
0.3
Vimentin
Mesenchymal lineage
50
4.0
4.0
4.0
0.0
a
Due to inability to identify mesh structures in some sections, the total number is \55
Table 4 Summed up expression levels for 10 different cell markers in relation to mesh class (mean, standard deviation, confidence interval) Mesh class
Na
Mean
Std. deviation
95 % confidence interval for mean Lower bound
Upper bound
Class I
18
29.0
1.4
28.3
29.7
Class II
9
29.6
1.2
28.8
30.3
Class III
6
30.3
1.0
29.4
31.2
Class V
10
30.6
1.6
29.5
31.7
Total
49
29.7
1.5
29.3
30.1
a
Due to inability to identify mesh structures in some sections, the total number is \55
A positive expression of four markers CD8, CD68, CD45R0 and Vimentin each in more than 80 % of cells implies a simultaneous expression of these proteins at least on some cells. Obviously, these four markers constantly take part in the tissue response to foreign body materials like meshes (Fig. 1). Characterisation of collagen around the mesh filaments The proportion of red-coloured collagen, reflecting mature collagen type I, was about 80 % in physiologic fascia (n = 8) and 57–60 % in the tissue at the vicinity of mesh filaments. In physiologic fascia tissue, the percentage of green, which reflects immature collagen type III, was about 20 %, correspondingly. In contrast, in the area around the mesh filaments, the proportion of coloured green collagen was markedly increased and usually reached values of 40–45 %. In general, meshes of class II, III and V showed higher mean values than meshes of class I without reaching
123
statistical significance due to considerable variations between the samples (Fig. 2). Similarly, the ratio of collagen I/III (red/green) was 1.3–1.5 in the vicinity of mesh filaments and at a mean of 3.8 in physiologic fascia. The quality of collagen was not influenced by the elapsed time since mesh implantation or by the reason for mesh explantation. Considering the reason for explantation, only pain showed a non-significant relation with a decreased amount of collagen type I (58.9 % of 38 samples without pain versus 56.6 % of 10 samples with an explantation because of pain, ANOVA p = 0.068), whereas recurrence and infection did not induce any considerable changes of the distribution for red and green or collagen type I and III. However, there were significant correlations between the number of CD20? and Desmin? cells in the granuloma with an increase of type III collagen (green)(correlation coefficient of r = 0.291, p = 0.043, and r = 0.328, p = 0.022, respectively). When summing up all 10 histological scores from the cellular markers, a highly positive value was closely related with an increased deposition of collagen type III as well (r = 0.335, p = 0.021) (Fig. 3). Analysis of collagen deposits at 409 in a large area of interest of 8 mm2 showed close correlations to corresponding measurements at a magnification of 400 in a small area of interest of 0.1 mm2 (Pearson correlation coefficient for green collagen III r = 0.688, red collagen type I = 0.703, ratio collagen I/III r = 0.704; Table 5).
Discussion The present study of 55 meshes explanted from humans clearly showed that the inflammatory infiltrate of the
Hernia
A
A
HE 40x
HE 400x
Sirius red 40x
Sirius red 400x
B
B
CD68
CD20
CD3
CD45R0
CD4
CD8
Desmin
Vimentin
Mib1
CD34
C
Fig. 1 Mesh explanted after 2 years from the groin of a male patient (48 years) because of pain. a Staining with HE (top) and Sirius red (bottom) showing the inflammatory infiltrate in the vicinity of the mesh filaments; red collagen type I, green collagen type III, magnification of 409 (left) and of 4009 (right). b Immunohistochemistry (at magnification of 4009) for CD68, CD20, CD3, CD45R0, CD4, CD8, Desmin, Vimentin, and Mib1
foreign body granuloma is not only purely an accumulation of CD68? macrophages though most of the cells express this surface marker. We could demonstrate that this cluster of cells express other markers in a similar intensity, e.g. CD8, CD45R0 and Vimentin. In additional, a positive expression of CD4 and of Desmin was abundant. The intensity of positive stain strongly indicates that many cells
Fig. 2 Deposition of collagen around the mesh structure according to the mesh classes and in physiologic fascia tissue. Collagen type I is indicated by the percentage of red pixels in an image of a Sirius red staining (409 magnification, cross Polaroid filter) (a), collagen type III is indicated by the percentage of green pixels (b), ratio of collagen type I/III is depicted in (c)
123
Hernia 34
sum of expression scores
33 32 31 30 29 28 27 25
30
35
40
45
50
collagen type III (green)
Fig. 3 Correlation of the presence of collagen III in the vicinity of mesh filaments and the summed up mean expression level of the 10 evaluated immunohistochemical cell markers (correlation coefficient r = ?0.335)
Table 5 Analysis of red collagen type I, green collagen type III and ratio collagen I/III at magnification of 409 and of 4009 (mean ± standard deviation) magnification
Collagen type I
Collagen type III
Ratio collagen I/III
409
61.4 ± 6.0
38.6 ± 6.0
1.7 ± 0.4
4009
58.4 ± 3.6
41.6 ± 3.6
1.4 ± 0.2
should express simultaneously at least two of the markers. These findings question the assumption that the infiltrate consists of distinct groups of macrophages, leucocytes or lymphocytes, but instead favours the conception of a dynamic collaboration and interchange of the cells of the innate immune system. Obviously, it is not only a crosstalk between macrophages and fibroblasts, as it was suggested by Jansen et al. [12], but the influx of immunologically competent cells is considerably more complex. Macrophages are known to present a high degree of heterogeneity, which reflects a specialisation of function [13]. The strong positive staining for CD45R0 supports the idea that some of the cells represent fibrocytes. These fibrocytes should not be mixed up with the fibroblasts that are the differentiated cells for the formation of connective tissue. In 1994, Richard Bucala described and named for the first time circulating fibrocytes as a new subpopulation of leucocytes [14]. Fibrocytes derived from CD14? precursor cells, immigrate in tissues and particularly in wounds, and can differentiate in macrophages, myofibroblasts or fibroblasts. It was shown that surface markers can change in relation to their functional status [15], e.g. with increasing differentiation, the expression of CD45
123
decreases [16, 17]. Recently, Reich et al. [18] could demonstrate in an animal model of a kidney injury that depletion of monocytes did not affect the accumulation of fibrocytes in the kidney, whereas depletion of CD11b? pr Ly-6G? cells did. However, the variable and changing expression of surface markers on cells hinders any clear assigning to a specific lineage. Thus, it is still under discussion whether local resident (stem) cells are able to transform into fibrocytes, or whether its origin mainly is the bone marrow. Whereas Bucala already used wound chambers to attract fibrocytes [14], it was Baker and Tang who pointed out the importance of fibrocytes for the foreign body reaction [19, 20]. Thevenot et al. could demonstrate that an inhibition of mast cells reduces the local transformation into fibrocytes [19, 20], which supports the finding of Orenstein et al. [21] who found an improved biocompatibility of synthetic meshes after systemic stabilisation of mast cells by chromoglycin acid. Further studies have to consider the variable patterns of CD markers on cells of the infiltrate, not the least to find the best combinations, which are able to represent the inflammatory activity. The major reasons for explantation of our meshes were infection, pain or recurrence. Explantations were done in the mean 3 years after implantation, which is in line with the findings of Klosterhalfen in his series of 1,000 explanted polypropylene meshes [9]. However, in our cohort, we included 7 explantations that have been done within the first 3 month due to infections or technical problems. This may explain that we found a higher rate of infections (40 %) as reason for the revision, as in the previous report (20 %). Since infection, and an early one in particular, is always accompanied by an intense inflammation, any effect of material or time on the deposition of collagens or the accumulation of specific cells may be hidden. Any quantitative and qualitative measurement of collagens that is deposited around the mesh has its limitations, as it widely depends on the amount of adjacent tissue that is considered and which contributes to the measurement. Since the first report of Junqueira in 1978, a qualitative distinction in the mature collagen I and the immature collagen III was done by Sirius red staining with use of a polarisation filter [10]. Previous studies could demonstrate that the ratio collagen I/collagen III is quite low at the interface of a mesh [11]. However, these measurements were done with 1009 magnification at small areas (0.3 mm2) of the fibrotic capsule. In this study, we made the analysis from pictures that were taken with a 409 magnification. Thus, we could analyse an area of about 8 mm2, which usually includes the entire mesh width and the complete scar that surrounds the mesh. Interestingly, the ratio of collagen I/III was quite similar to the
Hernia
measurement of the fibrotic capsule of the foreign body granuloma, if the measurements were done close to the filaments at a 4009 magnification. Due to the accumulation of collagen type III, again the ratio was found quite low. Measurements at fascia revealed a ratio of 3.8, which reflects the predominance of mature collagen I in physiological tissue. In contrast, the entire scar around a mesh showed a significantly increased amount of immature collagen type III, reflecting the permanent inflammatory process at the interface of the foreign body. This is in accordance with the findings of Vaz et al. [22], who found in rats after 30 days a type I/type III ratio of 1.2, analysing microscopic fields at a 200 magnification. However, in our small cohort, we could not confirm an enhanced amount of collagen III in the subgroup of patients with a recurrence [8]. This may be due to the fact that the area for analysis was considerably larger, or that the high number of explantations for infection with its similar increase of collagen III lessens any difference. It has to be considered that the amount of collagen III in relation to collagen I is affected both by indication of explantation, time since implantation, and configuration of the material and last but not the least, patients’ immunological defence capabilities. As these factors may superpose each other, any interpretation of a causal relationship is a challenge. Recently, Turner et al. [23] published an alternative staining with Herovici’s polychrome for the assessment of collagen and proposed a magnification of 10 to analyse an area of 1 mm2. They described a collagen I/III ratio of 4.0 for normal skin, and of 3.7 for hypertrophic scar, which is close to the ratio we found in our control fascia.
Conclusion In the present study on 55 meshes that were explanted from humans, we could clearly show that the inflammatory infiltrate of the foreign body granuloma is not only an accumulation of purely CD68? macrophages but that this cluster of cells express other markers in a similar intensity, e.g. CD8, CD45R0 and Vimentin. In addition, a positive expression of CD4 and of Desmin was abundant. Furthermore, we could demonstrate that collagen deposits around meshes to a great extent were of collagen type III. This can be analysed at a magnification of 40 reflecting an area of 8 mm2, which is less biased by the selection of the area of interest as it is at a magnification of 4009. The detection of a huge variety of immunological cells around meshes with a simple procedure to quantify collagen deposits offers enhanced options to monitor and improve biocompatibility of meshes in future studies.
Acknowledgments The project is co-funded by the European Union (European Regional Development Fund-Investing in your future) and the German federal state North Rhine-Westphalia (NRW).
References 1. Bellon JM, Bujan J, Contreras L, Hernando A, Jurado F (1994) Macrophage response to experimental implantation of polypropylene prostheses. Eur Surg Res 26(1):46–53 2. Klosterhalfen B, Junge K, Hermanns B, Klinge U (2002) Influence of implantation interval on the long-term biocompatibility of surgical mesh. Br J Surg 89(8):1043–1048. doi:10.1046/j.13652168.2002.02149.x 3. Rosch R, Junge K, Schachtrupp A, Klinge U, Klosterhalfen B, Schumpelick V (2003) Mesh implants in hernia repair. Inflammatory cell response in a rat model. Eur Surg Res 35(3):161–166 7004570045 4. Lee J, Lee SH, Chung WH, Kim DH, Chung DJ, Do SH, Kim HY (2013) The evaluation of canine small intestinal submucosal xenograft and polypropylene mesh as bioscaffolds on an abdominal full-thickness resection model in growing rats. J Vet Sci. Pii: jvs.2013.021 5. Pans A, Pierard GE (1992) A comparison of intraperitoneal prostheses for the repair of abdominal muscular wall defects in rats. Eur Surg Res 24(1):54–60 6. Zogbi L, Portella AO, Trindade MR, Trindade EN (2010) Retraction and fibroplasia in a polypropylene prosthesis: experimental study in rats. Hernia 14(3):291–298. doi:10.1007/s10029009-0607-x 7. Binnebosel M, Klink CD, Otto J, Conze J, Jansen PL, Anurov M, Schumpelick V, Junge K (2010) Impact of mesh positioning on foreign body reaction and collagenous ingrowth in a rabbit model of open incisional hernia repair. Hernia 14(1):71–77. doi:10. 1007/s10029-009-0580-4 8. Klosterhalfen B, Klinge U (2013) Retrieval study at 623 human mesh explants made of polypropylene—impact of mesh class and indication for mesh removal on tissue reaction. J Biomed Mater Res B Appl Biomater. doi:10.1002/jbmb.32958 9. Klinge U, Klosterhalfen B (2012) Modified classification of surgical meshes for hernia repair based on the analyses of 1,000 explanted meshes. Hernia 16(3):251–258. doi:10.1007/s10029012-0913-6 10. Junqueira LC, Cossermelli W, Brentani R (1978) Differential staining of collagens type I, II and III by Sirius Red and polarization microscopy. Arch Histol Jpn 41(3):267–274 11. Junge K, Klinge U, Rosch R, Mertens PR, Kirch J, Klosterhalfen B, Lynen P, Schumpelick V (2004) Decreased collagen type I/III ratio in patients with recurring hernia after implantation of alloplastic prostheses. Langenbecks Arch Surg 389(1):17–22 12. Jansen PL, Kever M, Rosch R, Krott E, Jansen M, Alfonso-Jaume A, Dooley S, Klinge U, Lovett DH, Mertens PR (2007) Polymeric meshes induce zonal regulation of matrix metalloproteinase-2 gene expression by macrophages and fibroblasts. FASEB J 21(4):1047–1057. doi:10.1096/fj.06-6755com 13. Gordon S, Taylor PR (2005) Monocyte and macrophage heterogeneity. Nat Rev Immunol 5(12):953–964 nri173310.1038/ nri1733 14. Bucala R, Spiegel LA, Chesney J, Hogan M, Cerami A (1994) Circulating fibrocytes define a new leukocyte subpopulation that mediates tissue repair. Mol Med 1(1):71–81 15. Pilling D, Fan T, Huang D, Kaul B, Gomer RH (2009) Identification of markers that distinguish monocyte-derived fibrocytes from monocytes, macrophages, and fibroblasts. PLoS One 4(10):e7475. doi:10.1371/journal.pone.0007475
123
Hernia 16. Grieb G, Steffens G, Pallua N, Bernhagen J, Bucala R (2011) Circulating fibrocytes-biology and mechanisms in wound healing and scar formation. Int Rev Cell Mol Biol 291:1–19. doi:10.1016/ B978-0-12-386035-4.00001-X 17. Mori L, Bellini A, Stacey MA, Schmidt M, Mattoli S (2005) Fibrocytes contribute to the myofibroblast population in wounded skin and originate from the bone marrow. Exp Cell Res 304(1):81–90. doi:10.1016/j.yexcr.2004.11.011 18. Reich B, Schmidbauer K, Rodriguez Gomez M, Johannes Hermann F, Gobel N, Bruhl H, Ketelsen I, Talke Y, Mack M (2013) Fibrocytes develop outside the kidney but contribute to renal fibrosis in a mouse model. Kidney Int 84(1):78–89. doi:10.1038/ ki.2013.84ki201384 19. Baker DW, Liu X, Weng H, Luo C, Tang L (2011) Fibroblast/ fibrocyte: surface interaction dictates tissue reactions to micropillar implants. Biomacromolecules 12(4):997–1005. doi:10. 1021/bm1013487
123
20. Thevenot PT, Baker DW, Weng H, Sun MW, Tang L (2011) The pivotal role of fibrocytes and mast cells in mediating fibrotic reactions to biomaterials. Biomaterials 32(33):8394–8403. doi:10.1016/j.biomaterials.2011.07.084 21. Orenstein SB, Saberski ER, Klueh U, Kreutzer DL, Novitsky YW (2010) Effects of mast cell modulation on early host response to implanted synthetic meshes. Hernia 14(5):511–516. doi:10.1007/ s10029-010-0680-1 22. Vaz M, Krebs RK, Trindade EN, Trindade MR (2009) Fibroplasia after polypropylene mesh implantation for abdominal wall hernia repair in rats. Acta Cir Bras 24(1):19–25 S010286502009000100005 23. Turner NJ, Pezzone MA, Brown BN, Badylak SF (2013) Quantitative multispectral imaging of Herovici’s polychrome for the assessment of collagen content and tissue remodelling. J Tissue Eng Regen Med 7(2):139–148. doi:10.1002/term.508
ORIGINAL ARTICLE
Characterisation of the cellular infiltrate in the foreign body granuloma of textile meshes with its impact on collagen deposition U. Klinge • U. Dietz • N. Fet • B. Klosterhalfen
Received: 12 July 2013 / Accepted: 18 January 2014 Ó Springer-Verlag France 2014
Abstract Purpose As part of the foreign body reaction, mesh filaments are surrounded by an infiltrate of inflammatory cells. Though macrophages are considered as being predominant, little is known about the origin of other cells. Methods On 55 meshes explanted from humans, we characterised the cells in the inflammatory infiltrate of the granuloma by immunohistochemistry using 10 cellular markers: CD3? lymphocytes, CD4? T helper cells, CD8? cytotoxic T cells, CD20? B lymphocytes, CD34? stem cells, CD45R0? leucocytes, CD68? macrophages, Mib1 for proliferation, Vimentin for mesenchymal origin, and Desmin for myocytes. Collagen deposits were analysed after staining with Sirius Red. Results More than 80 % of the cells in the infiltrate showed a positive expression of CD68, CD8, CD45R0 and Vimentin. CD4 and Desmin were seen in 30–80 % of the cells, unaffected by material or time. A score summarising the expression of all markers positively correlated significantly with an increased percentage of collagen type III (green) in the mesh wound. The analysis of collagen deposits was only affected to a small degree by size of area for investigation. U. Klinge (&) N. Fet Department of General, Visceral and Transplant Surgery, University Hospital of the RWTH Aachen, Pauwelsstraße 30, 52074 Aachen, Germany e-mail: [email protected] U. Dietz Department of General- Visceral- Vascular- and Paediatric Surgery, University of Wuerzburg, Oberduerrbacher Strasse 6, 97080 Wu¨rzburg, Germany B. Klosterhalfen Institute for Pathology, Du¨ren Hospital, Roonstr 30, 52351 Du¨ren, Germany
Conclusions At the vicinity of the mesh filaments, the accumulated inflammatory cells represent a mixture of cells of various origins. The high expression of at least four markers requires co-expression of different surface markers and thus confirms the existence of multiple transition forms instead of dominance of just macrophages. This offers new options for interventions to attenuate the inflammatory reaction of mesh implants. Keywords Foreign body granuloma Scar collagen Inflammatory cell Hernia mesh
Introduction Any implantation of a foreign body (FB) into tissues is followed by a response of the host’s innate immune system. This is also the case for textile hernia meshes, whose filaments are completely encapsulated by a foreign body granuloma including cells and extracellular matrix. Different types of inflammatory cells migrate to the mesh–host interface, and eventually create a fibrotic capsule in an attempt to separate the mesh from surrounding tissues. Several studies have shown that the inflammatory response of tissues surrounding surgical meshes includes macrophages [1], [2], CD3? lymphocytes [3], [2], CD8? T lymphocytes [4], CD20? B lymphocytes [2], neutrophils [5, 6], mast cells [3], CD15? granulocytes, or foreign body giant cells [2]. Most of these studies, however, focussed on one or only some few cell types demonstrating the influence of mesh material and structure, the time elapsed since implantation, the reason for explantation, or the anatomical location of the mesh, respectively, on the type of cells present [7–9]. In humans, CD68? macrophages and in mice, F4/60? cells were found to be the predominant
123
Hernia
players of the local tissue reaction. However, the fact that many of these cells in the infiltrate of the FB granuloma do not show a positive expression of these markers and thus may not be considered as macrophages, suggests the involvement of other cells in the process of local wound healing, as well. The cell signalling in the vicinity of surgical meshes obviously influences the extracellular matrix and specifically the collagen deposition. Correspondingly, there is a strong relationship between the inflammatory process and the extent of fibrosis around implanted meshes [8]. The predominant extracellular component of scar is collagen, either the mature and cross-linked collagen type I or the immature collagen type III. A ratio of collagen type I/III above 3 is assumed to be an indicator of good quality of the scar tissue (fibrosis). Klosterhalfen et al. demonstrated by means of immunohistochemical staining of 623 explanted meshes, that the content of collagen type I and type III of the fibrosis in the granuloma showed an impaired wound healing with a reduced ratio in half of the sections. Complementary results were found by means of polarisation microscopy after Sirius red staining [10]. Junge et al. [11] described a decreased collagen ratio of 2.1 ± 1.4 on 87 explanted meshes. With Sirius red staining, collagen type I appears as red, whereas collagen type III as green; thus an automatic image analysis process was used, which analysed the distribution of red or green pixels, respectively, in small selected fields in the scar around the mesh filaments (0.1 mm2, 400 magnification). However, as the analysis was limited to a circumscriptive area adjacent to the mesh filaments, it does not reflect the collagen deposition within the pores in between the mesh filaments. The aim of this study was first, to characterise the cell types of the infiltrate beneath the filaments of explanted hernia meshes; and second, to describe the collagen ratio in the mesh wound including a representative scar area of about 8 mm2. Both approaches have not been previously described.
As reference for collagen in physiologic fascia, we took samples of eight patients, of which the tissue had been excised during laparotomy, with no scar tissue (Ethics Committee approval for fascia tissue biopsy, University of Wuerzburg, Protocol 060/10 from the 16th of November, 2010).
Materials and methods
Table 1 Antibodies and dilution used for identification of inflammatory cells
Fifty-five samples of hernia meshes explanted in the years 2001–2012, and previously fixed in paraffin blocks were analysed. The mesh explants were taken from 22 female and 33 male patients with a mean age of 57 years (range 29–89). The mean time elapsed between implantation and explantation was 3.1 years. The recovered meshes were reclassified according to Klinge and Klosterhalfen [9] into one of five classes. The samples included 23 meshes of class I (large pore), 14 of class II (small pore), 7 of class III (intra-abdominal position), and 11 of class V (plug). The reasons for explantation were pain in 10 patients, infection in 13 and hernia recurrence in 37 patients.
Antigene
Antibodies
Dilution
CD3
Menarini Klon PSI
1:200
CD20
Menarini Klon L26 B cell
1:50
CD34
Menarini Klon QBEnd/10
1:50
CD45RO
Menarini Klon UCHL1
1:200
CD68
Menarini Klon 514H12
1:40
CD4
DAKO Klon 4B12
1:200
CD8
DAKO Klon C8/144B
1:100
Desmin
DAKO Klon D33
1:50
123
Immunohistochemistry Mesh samples were stained for HE and for expression of the following cell markers:
Marker
Cell type
CD3
Lymphocytes
CD4
T helper cells
CD8
Cytotoxic T cells
CD20
B lymphocytes
CD34
Stem cells
CD45R0
Leucocytes
CD68
Macrophages
Mib1 Vimentin Desmin
Function
Proliferation Mesenchymal origin Myocytes
Lymphoid tissue served as positive control for CD3, CD4, CD8, CD20, CD34, CD45RO, CD68, Mib1, Vimentin, whereas myometrium for Desmin. Pretreatment with Citrate (pH 6) 1:100 was done for CD3, CD20, CD34, CD45RO, Mib1, Vimentin, 20 min at 96 °C. Pretreatment with Tris–EDTA (pH 9) 1:100 was done for 20 min at 96 °C for CD4, CD8, CD68, Desmin (see Table 1 for details of antibodies and dilution). The protocol of autostainer includes 30 min primary antibody 300 ll, 10 min H2O2 300 ll, 10 min secondary
Ki67 (Mib1)
DAKO Klon Mib1
1:100
Vimentin
Zytomed Klon V9
1:200
Hernia
antibody 300 ll, 15 min Poly-HRP-Goat anti Mouse/Rabbit/Rat IgG 300 ll, 10 min DAB 300 ll. Washing with TRIS-Buffer pH 7 with non-ionic detergent Tween 20. Counterstaining was done with haematoxylin (30 s, Sigma). Fixation was done in Xylol. Analysis was performed only within the cellular infiltrate of the scar tissue beneath the mesh filaments, coding the intensity of expression for positive cells according to the following score: I (\5 %), II (5–30 %), III (30–80 %) and IV ([80 %).
Sirius red staining Sections of 5 lm were stained for 1 h in Picrosirius solution (0.1 % solution of Sirius Red F3BA in saturated aqueous picric acid, pH 2) according to Junqueira et al. [10]. Sections were washed in 0.01 N HCl for 2 min, dehydrated, cleared, and mounted in synthetic resin. Visualisation by crossed Polaroid filters allowed estimation of collagen type I and type III at 409 magnification. A photograph was taken with 640 9 480 pixels (=307200 pixels), 16 million colours, and saved as jpg. The image reflects an area of 3.3–2.5 mm = 8.1 mm2, which was completely analysed for the proportion of red and green colours by use of ‘‘Image J 1.46r’’ (http://imagej.nih. gov/ij/). Furthermore, we performed an analysis of collagen at a magnification of 4009 reflecting an area of interest of 0.1 mm2 close to the filaments (mean of three measurements per specimen). The colour histogram of the analysed function provided the separate percentage of green, red and blue pixels in an image. The percentage of red as indicator of collagen I is calculated as percentage of red in relation to red and green pixels; the percentage of green as indicator of collagen III resulted from subtraction of red from 100. Statistical calculations were done with IBM SPSS statistics 20. In general, a p \ 0.05 was considered as significant. Tests were done as indicated in the text.
Results The 10 meshes explanted because of pain were recovered after a mean time of 4.3 years; in this group, the number of mesh class V (plugs) was significantly increased (Pearson Chi Square p = 0.005) (Table 2). The meshes explanted because of infection were recovered after a mean time of 2.4 years, without influence of the mesh type. Also recurrence was not related to the type of explanted meshes in 32 specimens, explanted after a mean of 3.1 years.
Table 2 Number of explanted meshes (n = 55) with reason for explantation (pain, infection or recurrence) and mesh class. More than one reason for explantation may be present Mesh class I
II
Total III
V
Pain No
22
12
6
5
45
Yes
1
2
1
6*
10
No
16
10
5
11
42
Yes
7
4
2
0
13
No
4
5
4
5
18
Yes
19
9
3
6
37
23
14
7
11
55
Infection
Recurrence
Total number of explants
* Increase for meshes of class V reaches significance (Pearson Chi Square p = 0.004)
Characterisation of the cellularity in the vicinity of the mesh Most of the markers were expressed on cells in the inflammatory infiltrate around the mesh filaments. We observed an expression of CD8, CD45R0, CD68, and of Vimentin on more than 80 % cells. Whereas CD34-positive cells were rarely seen within the granuloma infiltrate, a variable distribution of 5–80 % of the cells showed a positive expression of CD3, CD4, CD20, Desmin, or Ki67 (Table 3). The extent of the cellular infiltrate (less, equal or more than scar tissue) was significantly positively correlated with the number of CD3-positive lymphocytes in the infiltrate (r = ?0.353). The number of CD68 ? macrophages was significantly correlated with the number of CD4 ? T helper cells (r =?0.316). The cellular expression of Desmin was influenced by the mesh class with highest values for class III and lowest values for class I meshes (3.8 ± 0.2 versus 3.3 ± 0.5; p = 0.06). The expression of CD45R0 is inversely related with the risk for pain (Fisher’s exact test mean score of 4 ± 0 without pain versus 3.9 ± 0.1 with pain; p = 0.04), mainly because two of the ten meshes explanted due to pain showed a slightly lower expression of CD45R0 with a score of \4 (3.7). When summing up all expression scores of the 10 cellular markers, we got a mean value for the 10 scores of 29.7 ± 1.5 with a range 24–33 (possible maximum score = 40): Highest values were seen for class V meshes, whereas class I meshes showed the lowest values (ANOVA with Bonferroni adjustment for Post Hoc tests 29.0 ± 1.4 versus 30.6 ± 1.6, p = 0.025) (Table 4).
123
Hernia Table 3 Scores of expression for 10 analysed cell markers at the vicinity of the mesh filaments as minimum, maximum, mean and standard deviation Marker
Na
Cell type/indicator
Score Minimum
Maximum
Mean
Std. dev.
CD3
Lymphocyte
50
1.7
3.0
2.2
0.4
CD4
T helper cell
50
2.0
4.0
3.0
0.5
CD8 CD20
Cytotoxic T cell Activated B cell
50 50
4.0 1.0
4.0 3.0
4.0 2.0
0.0 0.5
CD34
Hematopoietic stem cell
50
1.0
1.0
1.0
0.0
CD45R0
Leucocytes
50
3.7
4.0
4.0
0.1
CD68
Macrophage
49
3.0
4
4.0
0.1
Desmin
Myocyte, myofibroblast
50
2.5
4.0
3.5
0.4
Ki-67
Proliferation
50
1.0
3.0
2.1
0.3
Vimentin
Mesenchymal lineage
50
4.0
4.0
4.0
0.0
a
Due to inability to identify mesh structures in some sections, the total number is \55
Table 4 Summed up expression levels for 10 different cell markers in relation to mesh class (mean, standard deviation, confidence interval) Mesh class
Na
Mean
Std. deviation
95 % confidence interval for mean Lower bound
Upper bound
Class I
18
29.0
1.4
28.3
29.7
Class II
9
29.6
1.2
28.8
30.3
Class III
6
30.3
1.0
29.4
31.2
Class V
10
30.6
1.6
29.5
31.7
Total
49
29.7
1.5
29.3
30.1
a
Due to inability to identify mesh structures in some sections, the total number is \55
A positive expression of four markers CD8, CD68, CD45R0 and Vimentin each in more than 80 % of cells implies a simultaneous expression of these proteins at least on some cells. Obviously, these four markers constantly take part in the tissue response to foreign body materials like meshes (Fig. 1). Characterisation of collagen around the mesh filaments The proportion of red-coloured collagen, reflecting mature collagen type I, was about 80 % in physiologic fascia (n = 8) and 57–60 % in the tissue at the vicinity of mesh filaments. In physiologic fascia tissue, the percentage of green, which reflects immature collagen type III, was about 20 %, correspondingly. In contrast, in the area around the mesh filaments, the proportion of coloured green collagen was markedly increased and usually reached values of 40–45 %. In general, meshes of class II, III and V showed higher mean values than meshes of class I without reaching
123
statistical significance due to considerable variations between the samples (Fig. 2). Similarly, the ratio of collagen I/III (red/green) was 1.3–1.5 in the vicinity of mesh filaments and at a mean of 3.8 in physiologic fascia. The quality of collagen was not influenced by the elapsed time since mesh implantation or by the reason for mesh explantation. Considering the reason for explantation, only pain showed a non-significant relation with a decreased amount of collagen type I (58.9 % of 38 samples without pain versus 56.6 % of 10 samples with an explantation because of pain, ANOVA p = 0.068), whereas recurrence and infection did not induce any considerable changes of the distribution for red and green or collagen type I and III. However, there were significant correlations between the number of CD20? and Desmin? cells in the granuloma with an increase of type III collagen (green)(correlation coefficient of r = 0.291, p = 0.043, and r = 0.328, p = 0.022, respectively). When summing up all 10 histological scores from the cellular markers, a highly positive value was closely related with an increased deposition of collagen type III as well (r = 0.335, p = 0.021) (Fig. 3). Analysis of collagen deposits at 409 in a large area of interest of 8 mm2 showed close correlations to corresponding measurements at a magnification of 400 in a small area of interest of 0.1 mm2 (Pearson correlation coefficient for green collagen III r = 0.688, red collagen type I = 0.703, ratio collagen I/III r = 0.704; Table 5).
Discussion The present study of 55 meshes explanted from humans clearly showed that the inflammatory infiltrate of the
Hernia
A
A
HE 40x
HE 400x
Sirius red 40x
Sirius red 400x
B
B
CD68
CD20
CD3
CD45R0
CD4
CD8
Desmin
Vimentin
Mib1
CD34
C
Fig. 1 Mesh explanted after 2 years from the groin of a male patient (48 years) because of pain. a Staining with HE (top) and Sirius red (bottom) showing the inflammatory infiltrate in the vicinity of the mesh filaments; red collagen type I, green collagen type III, magnification of 409 (left) and of 4009 (right). b Immunohistochemistry (at magnification of 4009) for CD68, CD20, CD3, CD45R0, CD4, CD8, Desmin, Vimentin, and Mib1
foreign body granuloma is not only purely an accumulation of CD68? macrophages though most of the cells express this surface marker. We could demonstrate that this cluster of cells express other markers in a similar intensity, e.g. CD8, CD45R0 and Vimentin. In additional, a positive expression of CD4 and of Desmin was abundant. The intensity of positive stain strongly indicates that many cells
Fig. 2 Deposition of collagen around the mesh structure according to the mesh classes and in physiologic fascia tissue. Collagen type I is indicated by the percentage of red pixels in an image of a Sirius red staining (409 magnification, cross Polaroid filter) (a), collagen type III is indicated by the percentage of green pixels (b), ratio of collagen type I/III is depicted in (c)
123
Hernia 34
sum of expression scores
33 32 31 30 29 28 27 25
30
35
40
45
50
collagen type III (green)
Fig. 3 Correlation of the presence of collagen III in the vicinity of mesh filaments and the summed up mean expression level of the 10 evaluated immunohistochemical cell markers (correlation coefficient r = ?0.335)
Table 5 Analysis of red collagen type I, green collagen type III and ratio collagen I/III at magnification of 409 and of 4009 (mean ± standard deviation) magnification
Collagen type I
Collagen type III
Ratio collagen I/III
409
61.4 ± 6.0
38.6 ± 6.0
1.7 ± 0.4
4009
58.4 ± 3.6
41.6 ± 3.6
1.4 ± 0.2
should express simultaneously at least two of the markers. These findings question the assumption that the infiltrate consists of distinct groups of macrophages, leucocytes or lymphocytes, but instead favours the conception of a dynamic collaboration and interchange of the cells of the innate immune system. Obviously, it is not only a crosstalk between macrophages and fibroblasts, as it was suggested by Jansen et al. [12], but the influx of immunologically competent cells is considerably more complex. Macrophages are known to present a high degree of heterogeneity, which reflects a specialisation of function [13]. The strong positive staining for CD45R0 supports the idea that some of the cells represent fibrocytes. These fibrocytes should not be mixed up with the fibroblasts that are the differentiated cells for the formation of connective tissue. In 1994, Richard Bucala described and named for the first time circulating fibrocytes as a new subpopulation of leucocytes [14]. Fibrocytes derived from CD14? precursor cells, immigrate in tissues and particularly in wounds, and can differentiate in macrophages, myofibroblasts or fibroblasts. It was shown that surface markers can change in relation to their functional status [15], e.g. with increasing differentiation, the expression of CD45
123
decreases [16, 17]. Recently, Reich et al. [18] could demonstrate in an animal model of a kidney injury that depletion of monocytes did not affect the accumulation of fibrocytes in the kidney, whereas depletion of CD11b? pr Ly-6G? cells did. However, the variable and changing expression of surface markers on cells hinders any clear assigning to a specific lineage. Thus, it is still under discussion whether local resident (stem) cells are able to transform into fibrocytes, or whether its origin mainly is the bone marrow. Whereas Bucala already used wound chambers to attract fibrocytes [14], it was Baker and Tang who pointed out the importance of fibrocytes for the foreign body reaction [19, 20]. Thevenot et al. could demonstrate that an inhibition of mast cells reduces the local transformation into fibrocytes [19, 20], which supports the finding of Orenstein et al. [21] who found an improved biocompatibility of synthetic meshes after systemic stabilisation of mast cells by chromoglycin acid. Further studies have to consider the variable patterns of CD markers on cells of the infiltrate, not the least to find the best combinations, which are able to represent the inflammatory activity. The major reasons for explantation of our meshes were infection, pain or recurrence. Explantations were done in the mean 3 years after implantation, which is in line with the findings of Klosterhalfen in his series of 1,000 explanted polypropylene meshes [9]. However, in our cohort, we included 7 explantations that have been done within the first 3 month due to infections or technical problems. This may explain that we found a higher rate of infections (40 %) as reason for the revision, as in the previous report (20 %). Since infection, and an early one in particular, is always accompanied by an intense inflammation, any effect of material or time on the deposition of collagens or the accumulation of specific cells may be hidden. Any quantitative and qualitative measurement of collagens that is deposited around the mesh has its limitations, as it widely depends on the amount of adjacent tissue that is considered and which contributes to the measurement. Since the first report of Junqueira in 1978, a qualitative distinction in the mature collagen I and the immature collagen III was done by Sirius red staining with use of a polarisation filter [10]. Previous studies could demonstrate that the ratio collagen I/collagen III is quite low at the interface of a mesh [11]. However, these measurements were done with 1009 magnification at small areas (0.3 mm2) of the fibrotic capsule. In this study, we made the analysis from pictures that were taken with a 409 magnification. Thus, we could analyse an area of about 8 mm2, which usually includes the entire mesh width and the complete scar that surrounds the mesh. Interestingly, the ratio of collagen I/III was quite similar to the
Hernia
measurement of the fibrotic capsule of the foreign body granuloma, if the measurements were done close to the filaments at a 4009 magnification. Due to the accumulation of collagen type III, again the ratio was found quite low. Measurements at fascia revealed a ratio of 3.8, which reflects the predominance of mature collagen I in physiological tissue. In contrast, the entire scar around a mesh showed a significantly increased amount of immature collagen type III, reflecting the permanent inflammatory process at the interface of the foreign body. This is in accordance with the findings of Vaz et al. [22], who found in rats after 30 days a type I/type III ratio of 1.2, analysing microscopic fields at a 200 magnification. However, in our small cohort, we could not confirm an enhanced amount of collagen III in the subgroup of patients with a recurrence [8]. This may be due to the fact that the area for analysis was considerably larger, or that the high number of explantations for infection with its similar increase of collagen III lessens any difference. It has to be considered that the amount of collagen III in relation to collagen I is affected both by indication of explantation, time since implantation, and configuration of the material and last but not the least, patients’ immunological defence capabilities. As these factors may superpose each other, any interpretation of a causal relationship is a challenge. Recently, Turner et al. [23] published an alternative staining with Herovici’s polychrome for the assessment of collagen and proposed a magnification of 10 to analyse an area of 1 mm2. They described a collagen I/III ratio of 4.0 for normal skin, and of 3.7 for hypertrophic scar, which is close to the ratio we found in our control fascia.
Conclusion In the present study on 55 meshes that were explanted from humans, we could clearly show that the inflammatory infiltrate of the foreign body granuloma is not only an accumulation of purely CD68? macrophages but that this cluster of cells express other markers in a similar intensity, e.g. CD8, CD45R0 and Vimentin. In addition, a positive expression of CD4 and of Desmin was abundant. Furthermore, we could demonstrate that collagen deposits around meshes to a great extent were of collagen type III. This can be analysed at a magnification of 40 reflecting an area of 8 mm2, which is less biased by the selection of the area of interest as it is at a magnification of 4009. The detection of a huge variety of immunological cells around meshes with a simple procedure to quantify collagen deposits offers enhanced options to monitor and improve biocompatibility of meshes in future studies.
Acknowledgments The project is co-funded by the European Union (European Regional Development Fund-Investing in your future) and the German federal state North Rhine-Westphalia (NRW).
References 1. Bellon JM, Bujan J, Contreras L, Hernando A, Jurado F (1994) Macrophage response to experimental implantation of polypropylene prostheses. Eur Surg Res 26(1):46–53 2. Klosterhalfen B, Junge K, Hermanns B, Klinge U (2002) Influence of implantation interval on the long-term biocompatibility of surgical mesh. Br J Surg 89(8):1043–1048. doi:10.1046/j.13652168.2002.02149.x 3. Rosch R, Junge K, Schachtrupp A, Klinge U, Klosterhalfen B, Schumpelick V (2003) Mesh implants in hernia repair. Inflammatory cell response in a rat model. Eur Surg Res 35(3):161–166 7004570045 4. Lee J, Lee SH, Chung WH, Kim DH, Chung DJ, Do SH, Kim HY (2013) The evaluation of canine small intestinal submucosal xenograft and polypropylene mesh as bioscaffolds on an abdominal full-thickness resection model in growing rats. J Vet Sci. Pii: jvs.2013.021 5. Pans A, Pierard GE (1992) A comparison of intraperitoneal prostheses for the repair of abdominal muscular wall defects in rats. Eur Surg Res 24(1):54–60 6. Zogbi L, Portella AO, Trindade MR, Trindade EN (2010) Retraction and fibroplasia in a polypropylene prosthesis: experimental study in rats. Hernia 14(3):291–298. doi:10.1007/s10029009-0607-x 7. Binnebosel M, Klink CD, Otto J, Conze J, Jansen PL, Anurov M, Schumpelick V, Junge K (2010) Impact of mesh positioning on foreign body reaction and collagenous ingrowth in a rabbit model of open incisional hernia repair. Hernia 14(1):71–77. doi:10. 1007/s10029-009-0580-4 8. Klosterhalfen B, Klinge U (2013) Retrieval study at 623 human mesh explants made of polypropylene—impact of mesh class and indication for mesh removal on tissue reaction. J Biomed Mater Res B Appl Biomater. doi:10.1002/jbmb.32958 9. Klinge U, Klosterhalfen B (2012) Modified classification of surgical meshes for hernia repair based on the analyses of 1,000 explanted meshes. Hernia 16(3):251–258. doi:10.1007/s10029012-0913-6 10. Junqueira LC, Cossermelli W, Brentani R (1978) Differential staining of collagens type I, II and III by Sirius Red and polarization microscopy. Arch Histol Jpn 41(3):267–274 11. Junge K, Klinge U, Rosch R, Mertens PR, Kirch J, Klosterhalfen B, Lynen P, Schumpelick V (2004) Decreased collagen type I/III ratio in patients with recurring hernia after implantation of alloplastic prostheses. Langenbecks Arch Surg 389(1):17–22 12. Jansen PL, Kever M, Rosch R, Krott E, Jansen M, Alfonso-Jaume A, Dooley S, Klinge U, Lovett DH, Mertens PR (2007) Polymeric meshes induce zonal regulation of matrix metalloproteinase-2 gene expression by macrophages and fibroblasts. FASEB J 21(4):1047–1057. doi:10.1096/fj.06-6755com 13. Gordon S, Taylor PR (2005) Monocyte and macrophage heterogeneity. Nat Rev Immunol 5(12):953–964 nri173310.1038/ nri1733 14. Bucala R, Spiegel LA, Chesney J, Hogan M, Cerami A (1994) Circulating fibrocytes define a new leukocyte subpopulation that mediates tissue repair. Mol Med 1(1):71–81 15. Pilling D, Fan T, Huang D, Kaul B, Gomer RH (2009) Identification of markers that distinguish monocyte-derived fibrocytes from monocytes, macrophages, and fibroblasts. PLoS One 4(10):e7475. doi:10.1371/journal.pone.0007475
123
Hernia 16. Grieb G, Steffens G, Pallua N, Bernhagen J, Bucala R (2011) Circulating fibrocytes-biology and mechanisms in wound healing and scar formation. Int Rev Cell Mol Biol 291:1–19. doi:10.1016/ B978-0-12-386035-4.00001-X 17. Mori L, Bellini A, Stacey MA, Schmidt M, Mattoli S (2005) Fibrocytes contribute to the myofibroblast population in wounded skin and originate from the bone marrow. Exp Cell Res 304(1):81–90. doi:10.1016/j.yexcr.2004.11.011 18. Reich B, Schmidbauer K, Rodriguez Gomez M, Johannes Hermann F, Gobel N, Bruhl H, Ketelsen I, Talke Y, Mack M (2013) Fibrocytes develop outside the kidney but contribute to renal fibrosis in a mouse model. Kidney Int 84(1):78–89. doi:10.1038/ ki.2013.84ki201384 19. Baker DW, Liu X, Weng H, Luo C, Tang L (2011) Fibroblast/ fibrocyte: surface interaction dictates tissue reactions to micropillar implants. Biomacromolecules 12(4):997–1005. doi:10. 1021/bm1013487
123
20. Thevenot PT, Baker DW, Weng H, Sun MW, Tang L (2011) The pivotal role of fibrocytes and mast cells in mediating fibrotic reactions to biomaterials. Biomaterials 32(33):8394–8403. doi:10.1016/j.biomaterials.2011.07.084 21. Orenstein SB, Saberski ER, Klueh U, Kreutzer DL, Novitsky YW (2010) Effects of mast cell modulation on early host response to implanted synthetic meshes. Hernia 14(5):511–516. doi:10.1007/ s10029-010-0680-1 22. Vaz M, Krebs RK, Trindade EN, Trindade MR (2009) Fibroplasia after polypropylene mesh implantation for abdominal wall hernia repair in rats. Acta Cir Bras 24(1):19–25 S010286502009000100005 23. Turner NJ, Pezzone MA, Brown BN, Badylak SF (2013) Quantitative multispectral imaging of Herovici’s polychrome for the assessment of collagen content and tissue remodelling. J Tissue Eng Regen Med 7(2):139–148. doi:10.1002/term.508
