- Email: [email protected]
Critical evaluation of prosthetic materials in repair of abdominal wall hernias
Critical Evaluation of Prosthetic Materials in Repair of Abdominal Wall Hernias New Criteria of Tolerance and Resistance
J. P. Arnaud, Strasbourg-Schiltigheim, France R. Eloy, Strasbourg-Hautepierre, France M. Adloff, Strasbourg-Schiltigheim, France J. F. Grenler, Strasbourg-Hautepierre, France
Since the introduction by Witzel [I] and Goepel[2] in 1900 of silver mesh in the repair of large defects of the abdominal wall, the use of an impressive number of foreign materials has been proposed. Up to now, the surgeon has had to choose between many “biologically inert” materials. This selection is generally difficult and despite seventy-five years of endeavor, the use of such material remains unsatisfactory. Moreover, the regular appearance of new prosthetic implants suggests that the definitive material has yet to be decided. The necessary qualities of synthetic materials have been well established by Cumberland [3] and Scales [4]. These are that it should (1) not be physically modified by tissue fluids, (2) be chemically inert, (3) not excite an inflammatory or foreign body reaction, (4) be noncarcinogenic, (5) not produce a state of allergy or hypersensitivity, (6) be capable of resisting mechanical strains, (7) be capable of being fabricated in the form required, and (8) be capable of being sterilized. Many of these requirements are met by some prostheses, but there is little precise knowledge of the tissue reaction around the foreign material. Therefore, several prosthetic materials were compared and evaluated by mechanical resistance, macroscopic findings, systematic bacteriologic investigations, and quantitative histologic examination of the surrounding tissue reaction.
From the Centre Mdic@Xnrgical et ‘Obst&rical. Strasbourg-Schiltigheim. and Unit6 61 INSERM. Surgical Research and Biophysiopathology of the Intestine. Strasbourg-Haute&eve. France. Reprint requests .skuM be x&essad to Rofesseu M. Adloft, Department of Surgery, CMCO Schiltigheim, 67300 StrasbourqSchiltigheim, France.
338
Mat&fat and Methods A scheme for the experimental investigation was set up using Wistar rats weighing between 250 and 500 gm [5]. The prostheses were randomly allocated to the animals. The rats were anesthetized by intraperitoneal injection of Nembutale 1 per cent (0.1 ml/100 gm body weight). All procedures were performed with aseptic technic. A longitudinal skin incision was made and rectangular segments (3 X 4 cm) of the anterior abdominal wall, including both fascia, muscles, and peritoneum, were excised. (Figure 1.) The defect was repaired by stitching a segment of the test material of equal size to the margins of the peritoneum muscles and fascia. Interrupted catgut sutures were placed 5 mm apart and 5 mm from the edge of the defect. The skin was closed over the implant. No antibiotic was used. Six materials in common use were tested. Three were porous weave or mesh materials-nylon woven mesh, Dacron@ knitted mesh, and stainless steel mesh-and three were nonporous weave or cloth materials-silicone cloth, siliconized Dacron cloth, and siliconized Dacron velvet. In each case thirty animals were followed clinically and sacrificed sequentially in groups of five at postoperativedays 3,7,15,30,60, and 90. Before sacrifice and under anesthesia, the abdominal wall resistance was tested using an original method. (Figure 2.) After a skin incision was made to expose the prosthetic implant, the peritoneal cavity was filled with water and methyl blue through a fine needle inserted into the anterior abdominal wall. The intraperitoneal pressure was monitored by a pressure transducer (Statham). When disruption of the wound occurred an immediate decrease in pressure was recorded on a Sefram pressure recorder. The mean (f SEM) value of the bursting pressure at each time was calculated for each foreign material. The tissue reaction to the various prostheses was assessed macroscopically and microscopically. For the former
The American Journal of Surgery
Critical Evaluation of Prosthetic Materials
the peritonization or adherence to intra-abdominal structures and the incorporation, encapsulation, or extrusion of the foreign material were observed. Bacteriologic investigations were regularly performed on the biomaterial itself and also on the serosal fluid which collected. Histologic examination of the whole implant with its surrounding tissue was performed after fixation in 30 per cent formalin. Paraffin-embedded tissue blocks were sectioned and stained with hematoxylin and eosin. Except for the stainless steel prosthesis, the tissue reaction was determined on the basis of four factors. The first factor was the width of the tissue reaction (inflammatory in the early phases, fibroblastic in the later phases) measured in millimeters from the periphery of the implant section. The three other factors concerned the cellular density around the foreign material. The number of inflammatory, giant, and fibroblastic cells in one oil immersion field (X 800) were counted. Finally, the ratio of inflammatory cells to fibroblastic cells at the different postoperative days was calculated. For one tissue section, each of these factors was estimated four times. The average was calculated and the overall averages for all animals at a given period of time were compared using the test of Newman-Keuls.
prosthesis. At postoperative day 7 this “percentage resistance” of the abdominal wall repaired with silicone cloth was only 56 per cent, whereas with the Dacron knitted mesh and the stainless steel mesh it
Results Mechankal Reeletance of the AbdominalWall
Table I summarizes the results of studies in which the bursting strength of wounds in the abdominal wall of rats were determined. In the early postoper-
ative period, the rupture occurred at the junction between the prosthesis and the muscle. After postoperative day 30, the rupture of the abdominal wall occurred in a site other than that of the prosthesis, and these bursting pressures were used as references. Thus, the bursting strength of abdominal wall repairs before postoperative day 30 were also expressed as a percentage of the abdominal pressure needed to
rupture the abdominal wall at a point other than the
Figure 1./mplantat/on of prosthetic material thwgh culature and peritoneum.
I Figure 2. Determinatton of the abdomlnal wall resisthce.
volmm
199, awdl1977
PUMP
INJEClKm
mw-
Arnaud et al
TABLE
I
Mechanical
Resistance
of the Abdominal
Wall
(mm
woven Postoperative Postoperative Postoperative
TABLE
II
day day day
7 15 30
Behavior
Mesh
158 188 > 240
f 1.3 i- 1.6
of the
Incorporation Extrusion Encystment with partial incorporation Encystment without incorporation lntraperitoneal migration
Dacron
Nylon
Foreign
192 197 > 240
Material
Knitted * 2 f 1.4
(after
Stainless Steel Mesh 181 191 > 240
Silicone
* 1.6 f 4.4
postoperative
day
135 138 >240
Siliconized Dacron Cloth
Cloth + 3 r 1.7
154 183 > 240
+ 1.4 * 1.4
Siliconized Dacron Velvet 158 158 > 240
? 1.6 t 3.5
15)
Woven Nylon Mesh
Dacron Knitted Mesh
Stainless Steel Mesh
100% -
100% -
100% -
Silicone -
Cloth
Siliconized Dacron Cloth -
Siliconized Dacron
Velvet -
5% 60%
80%
5% 70%
-
-
-
15%
10%
25%
-
-
-
20%
10%
0%
increased to 75 to 80 per cent. The remaining prostheses gave intermediate values ranging between 64 and 66 per cent. At postoperative day 15, it appears that the resistance of the abdominal wall repaired with silicone cloth (57 per cent) or siliconized Dacron velvet (66 per cent) was not significantly different from that observed at day 7. On the other hand, the resistance of repairs with nylon woven mesh or siliconized Dacron cloth had significantly increased at this time to 78 per cent and 76 per cent, respectively. These values were similar to those observed with Dacron knitted mesh (82 per cent) and stainless steel mesh (79 per cent). MacroscopicFindings Serial clinical examination demonstrated a strong repair of the abdominal wall and no occurrence of hernia except in one case with a silicone cloth prosthesis (postoperative day 7). Fragmentation of the stainless steel wire was observed in three cases but was not associated with impairment of the abdominal wall repair. Peritonization began with the appearance of a glistening membrane which covered the inner surface of the material both at its periphery and around the omental adhesions to the foreign material. Complete peritonization occurred generally with mesh materials by postoperative day 7, whereas 15 days were necessary for complete peritonization in the case of cloth materials. Peritonization, however,
340
Hg * SEMI
was
always incomplete with stainless steel mesh even three months after operation. Omental adhesions were regularly found whatever material was used. They seemed more firmly fixed to mesh material. These omental adhesions were sometimes associated with adhesions of bowel but no intestinal obstruction was observed. After mesh material implantation, fibrous tissue infiltrated between the filaments of the thread, providing a semirigid framework. There was no tissue fluid space in this experimental group. On the other hand, cloth prostheses became loosely invested by thin fibrous tissue and were enclosed in a bursa-like envelope formed by proliferating connective tissue. Sometimes, the cloth materials could be easily removed and serosal collections were found within the capsule. These findings were more marked in the case of Dacron siliconized velvet. As summarized in Table II, the nonincorporation of the foreign material can lead either to its extrusion or to its migration as a foreign body into the peritoneal cavity. Infaction
After implantation of mesh materials, neither wound infections nor serosal collections were observed and this was confirmed by negative bacteriology. On the other hand, serosal subcutaneous collections developed regularly after the use of cloth material. Their evacuation was necessary but this did not always prevent the appearance of subcutaneous wound infections, which occurred in 14.3 per cent
The American Journal of Surgery
Critical Evaluation of Prosthetic Materials
EXTENT WE
OF
CEWUR
FIBROBLASTIC
Figure 3. Extent of the cellular reaction after nyh woven mesh (C), Dacron knltted mesh (M), silicone cloth (S), siNconized Dacron cloth (SR), and slliconlzed Dacron velvet ( R) implantation.
GIANT
15
(siliconized Dacron velvet), 8.5 per cent (siliconized Dacron cloth), and 5.7 per cent (silicone cloth) of the experimental animals. Moreover, a mucopurulent exudate found around the cloth material was noted in 10 per cent of the experimental animals after silicone cloth or silicone Dacron cloth implantation, and in 25 per cent after silicone Dacron velvet implantation. These infections were due to staphylococcus (61 per cent), proteus (22 per cent), and pyocyaneus (16 per cent). MicroscopicFlndings
During the first seven days, the histologic reaction was similar for all tested materials. The following description concerns only the chronic (more than 15 postoperative days) rather than the acute responses of tissues to foreign materials. The overall microscopic picture of tissue reaction concerned the extent of the cellular reaction and the number of inflammatory, giant, and fibroblastic cells. Between postoperative days 15 and 90, the extent of the cellular reaction around the prosthesis varied as follows. (Figure 3.) After siliconized Dacron velvet implantation, it was rectilinear, “a” being not significantly different from zero. The cellular reaction with siliconized Dacron cloth or silicone cloth implantation decreased significantly (p
vohlmo 199, Mwch
mv
30
60
CELLS
CELLS
90 POST
OPERATIVE
MY’S
Ngure 4. Comparative evolution of the different hlstologk parameters under study during SO days after knpbntatkm of the pro&et/c mater/a/s. ( R = Mconked Dacfon velvet; St? = silkzonized Dacron cloth; S = s&one cloth; M = Dacron knitted mesh; and C = nylon woven mesh. )
cellular reaction after Dacron knitted mesh, however, increased between postoperative days 15 and 30 as well as between days 60 and 90 (p
341
Arnaudet al
Figure 5. Emsneration of the inflammatory (/en), fibroblastlc (m/M/e), and giant (right) ceils observedper sectional area. ( R = silkonized Dacron velvet; SR = sMconized Dacron cloth; S = sllkone cloth; M = Dacron knitted mesh; and C = nylon woven mesh.)
inflammatory cells between days 15 and 30 and days 60 and SO (p
342
knitted prosthesis was associated with an infiltration of fibroblasts which increased in accordance with the formula y = 0.79 t + 280.54 (y = number of fibroblasts, t = postoperative days). At each postoperative time, Dacron knitted mesh, nylon woven mesh, siliconized Dacron velvet, silicone cloth, and siliconized Dacron cloth were prosthetic materials of decreasing potency (Figure 4) in the induction of infiltration by fibroblasts (p <0.05). The number of giant cells (Figure 5) observed was constant throughout the postoperative period for each foreign material except Dacron knitted mesh with which it decreased at postoperative day SO (p <0.05). Plotted together (Figure 4), it appears that the number of giant cells increased progressively when silicone cloth, nylon woven mesh, Dacron knitted mesh, siliconized Dacron cloth, and siliconized Dacron velvet were implanted. The ratio between fibroblasts (Figure 6) and inflammatory cells has been established for each foreign implant at the different postoperative times. This ratio generally increased between postoperative days 15 and SO (p
The American Journal of Surgery
Critical Evaluationof ProstheticMaterials
between each material suggests that the ratio between fibroblasts and inflammatory cells was higher for mesh than for cloth materials. Figure 7 illustrates the different histologic features. Comments
In the present report, mechanical, macroscopic, bacteriologic, and histologic criteria were followed to provide a quantitative critical evaluation of different prosthetic materials which are commonly used. The resistance of the repaired abdominal wall and the tolerance to the foreign material were studied. The bursting strength of abdominal wall repairs with different prosthetic materials suggest that after postoperative day 30 there is no difference whichever material was used. (Figure 7.) Moreover, our findings corroborate previous reports that a healing wound is more resistant than the nonrepaired abdominal wall [6-8]. However, striking differ.ences were observed in the early postoperative period (15 days) in the resistance of the abdominal wall to the different prostheses. During this period, disruption of the wound occurred quite readily and was generally located at the junction between the prosthesis and the muscle, rupture of the prosthesis itself never being seen. This early low strength of the abdominal wall corresponded to the “lag period” of Howes [S] and might be considered to be responsible for the later recurrence of a parietal hernia. Therefore, to produce an early resistant wound it is necessary to choose an adequate prosthetic material. To this end, our results suggested that mesh materials and to a lesser extent the siliconized Dacron cloth enhanced markedly the bursting pressure of the wound. The general appearance of the host tissue reaction around the prosthesis was found to be the most useful macroscopic criteria of tolerance. In all animals repaired with mesh materials there was strong incorporation of the implant within the surrounding tissue. On the other hand implantation of cloth materials was characterized by encystment of the prosthesis with or without its partial incorporation. Similar results have been reported in tests of nonporous weaves [4, 9-111. The nonpermeability of cloth materials might account for the occurrence and persistence of a serosal collection and prevent its drainage towards the peritoneal cavity. As described by Scales [4] and Harrison, Swanson, and Lincoln [12], a fibrous tissue capsule also developed around the nonporous material unlike the microporous mesh material which was invaded by the tissues and incorporated in them [13]. Finally, the cloth materials appeared enclosed but free within a fibrous capsule regularly containing a serosal collection and these
voknn
‘tssl Mwab 1917
92
1
0.1 0
I
15
30
60
90 POST OPEIUTIYE Mw
Figurn 6. Evolution &wing 90 postopefatlve days of ratlo oil&olblsbto~~(R=~Dacron velvet; SR = sWodzed Dacron cbrsl; S = s&-one cloth; M = Dacron knltted mesh; and C = nylon woven mesh. )
conditions may explain the frequency of infection confirmed by bacteriology. Encystment, nonincorporation of the foreign material, and infection lead in a final stage to the extrusion of the implant. These events have never been observed with mesh materials. The tissue reaction to the implanted materials was also examined at the cellular level. The width of the cellular reaction around the prosthesis might be considered as a criterion of tolerance. Excluding the first postoperative month, the extent of this cellular reaction was similar for each material tested except for the Dacron siliconized velvet. In this case even after three months the extent of the reaction was four times greater than that of other prostheses. This histologic assessment corroborated macroscopic findings that the implantation of this kind of material produced the most extensive fibrous capsule. Giant cell infiltration was also used to assess tolerance. Materials which evoked this response were unlikely to be satisfactory. It has been suggested that the presence of those cells reflects instability of the surrounding tissue [IO]. During the postoperative follow-up, this feature remained relatively constant for each material tested. The mean values were compared and except for siliconized Dacron velvet, which exhibited a striking gigantic cell infiltration,
343
Arnaud et al
Figure Mkroscopk 7. plate. Comparative histologk find&s observed 30 days after imp/antatlon of nyIon woven mesh (C), Dacron knitted mesh (M), silicone cloth (S), sillconized Dacron clot% (SR), and siikonized Dacron velvet (R) materials. (Hematoxylin and eosfn st&; magnifkatkn x 25.)
no significant difference was noted between mesh and cloth materials. The importance of what has commonly been called “chronic inflammation” has been investigated by counting the inflammatory cells. Generally, their number decreased with time for each foreign material tested. Again siliconized Dacron velvet showed a marked reaction but mesh materials generally, seemed to induce a slightly greater response than cloth materials. The number of fibroblasts present was taken as a microscopic index of the mechanical
344
resistance of the repaired tissue. In the early postoperative period, these results correlated well with the macroscopic and bursting studies, again showing superiority of mesh materials. Inflammatory cells and fibroblasts represent the main cellular components of a healing wound [14] and they may represent two different evolutionary steps of the cicatrisation procedure. Whereas the former could be considered to reflect the tolerance of the host to foreign material, the second is the histologic feature corresponding to the mechanical re-
The American Journal of
Surgery
Critical Evaluation of Prosthetic Materials
sistance of the repaired wound. Therefore, their ratio appears of particular interest: the larger the ratio, the stronger the wound and the better the tolerance of the host. The order in which the materials are ranked by this ratio, in decreasing magnitude, is mesh material (Dacron knitted and nylon woven mesh) and then cloth materials (silicone and siliconized Dacron cloth) and finally siliconized Dacron velvet. Among the eight properties required classically [3, 41 to characterize foreign prostheses, tolerance and mechanical resistance have been particularly investigated in this study. Macroscopic, mechanical, bacteriologic, and several histologic parameters have been compared in a critical study. It appears that alone none of these parameters is sufficient to fully characterize a foreign material. This combination of macroscopic, mechanical, bacteriologic, and histologic analysis proves that the physical nature of an implant influences directly the biologic reactions which we have considered. Thus, on present evidence it would seem reasonable to recommend mesh rather than cloth materials in the repair of abdominal wall. Among the numerous criteria investigated, it must be outlined that one of them, the ratio of fibroblastic cells to inflammatory cells expresses best the macroscopic, mechanical, and bacteriologic findings and reflects the resistance and tolerance of the foreign materials. This histologic test may be at present a helpful index to evaluate in experimental or clinical trials.
Summary
Six different materials (3 mesh and 3 cloth) commonly used for the repair of abdominal wall hernias were evaluated in 180 rats after implantation through musculature and peritoneum. Serial macroscopic and bacteriologic investigations were done, and bursting strength of the wound was determined in all groups at intervals with an original device. Histologic criteria were used to characterize the resistance of the wound and the tolerance of the host to the foreign material. Statistical analysis of the results demonstrated: (1) after postoperative day 15, the resistance of the wound was similar for each material tested; (2) dur-
vohtmo 133, Much 1977
ing this early period mesh materials exhibited more resistance to bursting pressures than cloth materials; (3) the incorporation of mesh material was constant, whereas encystment or extrusion was always observed after implantation of cloth material; (4) no infection occurred with mesh material, but significant bacteria were found in 18 per cent of cloth material implantations; (5) the extent of the cellular reaction, the enumeration of giant, inflammatory, and fibroblastic cells showed the superiority of mesh materials; and (6) the ratio of fibroblasts to inflammatory cells reflected closely the mechanical resistance and tolerance of the foreign material. References 1. Witzel 0: Ueber die Verschliessung von Bauchwunden und Brustpforten durch Versenkte Silberdrahtnetze. Cent Chir 27: 257, 1900. 2. Goepel R: Uber die Verschliessung Von Bruchpforten Durch Einheilung Geflochtener Fertiger Silberdrahtnetze. Vet-h Msch Ges Chir 29: 4, 1900. 3. Cumberland VH: A preliminary report on the use of prefabricated nylon weave in the repair of ventral hernia. Med J Aust 1: 143, 1952. 4. Scales JT: Discussion on metals and synthetic materials in relation to soft tissues; tissue reaction to synthetic materials. Proc R Sot Med46: 647,1953. 5. Amaud JP, Adloff M, Weill-Bousson M: Prosthetic materials in the repair of the abdominal wall. New criteria of tolerance and resistance. Eur Sot .Exp Surg (10th Congress): 215, 1975. 6. Adamsons RJ, Kahan SA: The rate of healing of incised wound of different tissues in rabbits. Surg C+neco/ Obsfet 130: 237, 1970. 7. Cerise EJ, Busuttil RW, Craighead CC, Dgden WW: The use of Mersiline mesh in repair of abdominal wall hernias. Ann Surg 161: 726, 1975. 6. Howes EL: Healing of fascia defects, p 201. La Cicatrisation (CNRS ed). 1965. 9. Koontz AR, Kimberly RC: Tentalum and marlex mesh: experimental and clinical comparison. Preliminary report. Ann Surg 151: 796, 1960. 10. Calnan JS: The use of inert plastic material in reconstructive surgery. I. A biological test for tissue acceptance. II. Tissue reactions to commonly used materials. Br J Pled Surg 16: 1, 1963. 11. Narat JK, Khedrou LG: Repair of abdominal wall defects with Fort&n Fabric. Ann Surg 136: 272, 1952. 12. Harrison JH. Swanson DS, Lincoln AF: A comparison of tissue reactions to plastic materials. Arch Surg 74: 139, 1957. 13. Usher FC, Gannon JP: Marlex mesh, a new plastic mesh for replacing tissue defects. Experimental studies. Arch Surg 78: 131, 1959. 14. Hartwell SW: Research and clinical observations related to the healing of human wounds, burns and stasis ulcers, p 235. La Cicatrisation (CNRS, ed). 1965.
345
J. P. Arnaud, Strasbourg-Schiltigheim, France R. Eloy, Strasbourg-Hautepierre, France M. Adloff, Strasbourg-Schiltigheim, France J. F. Grenler, Strasbourg-Hautepierre, France
Since the introduction by Witzel [I] and Goepel[2] in 1900 of silver mesh in the repair of large defects of the abdominal wall, the use of an impressive number of foreign materials has been proposed. Up to now, the surgeon has had to choose between many “biologically inert” materials. This selection is generally difficult and despite seventy-five years of endeavor, the use of such material remains unsatisfactory. Moreover, the regular appearance of new prosthetic implants suggests that the definitive material has yet to be decided. The necessary qualities of synthetic materials have been well established by Cumberland [3] and Scales [4]. These are that it should (1) not be physically modified by tissue fluids, (2) be chemically inert, (3) not excite an inflammatory or foreign body reaction, (4) be noncarcinogenic, (5) not produce a state of allergy or hypersensitivity, (6) be capable of resisting mechanical strains, (7) be capable of being fabricated in the form required, and (8) be capable of being sterilized. Many of these requirements are met by some prostheses, but there is little precise knowledge of the tissue reaction around the foreign material. Therefore, several prosthetic materials were compared and evaluated by mechanical resistance, macroscopic findings, systematic bacteriologic investigations, and quantitative histologic examination of the surrounding tissue reaction.
From the Centre Mdic@Xnrgical et ‘Obst&rical. Strasbourg-Schiltigheim. and Unit6 61 INSERM. Surgical Research and Biophysiopathology of the Intestine. Strasbourg-Haute&eve. France. Reprint requests .skuM be x&essad to Rofesseu M. Adloft, Department of Surgery, CMCO Schiltigheim, 67300 StrasbourqSchiltigheim, France.
338
Mat&fat and Methods A scheme for the experimental investigation was set up using Wistar rats weighing between 250 and 500 gm [5]. The prostheses were randomly allocated to the animals. The rats were anesthetized by intraperitoneal injection of Nembutale 1 per cent (0.1 ml/100 gm body weight). All procedures were performed with aseptic technic. A longitudinal skin incision was made and rectangular segments (3 X 4 cm) of the anterior abdominal wall, including both fascia, muscles, and peritoneum, were excised. (Figure 1.) The defect was repaired by stitching a segment of the test material of equal size to the margins of the peritoneum muscles and fascia. Interrupted catgut sutures were placed 5 mm apart and 5 mm from the edge of the defect. The skin was closed over the implant. No antibiotic was used. Six materials in common use were tested. Three were porous weave or mesh materials-nylon woven mesh, Dacron@ knitted mesh, and stainless steel mesh-and three were nonporous weave or cloth materials-silicone cloth, siliconized Dacron cloth, and siliconized Dacron velvet. In each case thirty animals were followed clinically and sacrificed sequentially in groups of five at postoperativedays 3,7,15,30,60, and 90. Before sacrifice and under anesthesia, the abdominal wall resistance was tested using an original method. (Figure 2.) After a skin incision was made to expose the prosthetic implant, the peritoneal cavity was filled with water and methyl blue through a fine needle inserted into the anterior abdominal wall. The intraperitoneal pressure was monitored by a pressure transducer (Statham). When disruption of the wound occurred an immediate decrease in pressure was recorded on a Sefram pressure recorder. The mean (f SEM) value of the bursting pressure at each time was calculated for each foreign material. The tissue reaction to the various prostheses was assessed macroscopically and microscopically. For the former
The American Journal of Surgery
Critical Evaluation of Prosthetic Materials
the peritonization or adherence to intra-abdominal structures and the incorporation, encapsulation, or extrusion of the foreign material were observed. Bacteriologic investigations were regularly performed on the biomaterial itself and also on the serosal fluid which collected. Histologic examination of the whole implant with its surrounding tissue was performed after fixation in 30 per cent formalin. Paraffin-embedded tissue blocks were sectioned and stained with hematoxylin and eosin. Except for the stainless steel prosthesis, the tissue reaction was determined on the basis of four factors. The first factor was the width of the tissue reaction (inflammatory in the early phases, fibroblastic in the later phases) measured in millimeters from the periphery of the implant section. The three other factors concerned the cellular density around the foreign material. The number of inflammatory, giant, and fibroblastic cells in one oil immersion field (X 800) were counted. Finally, the ratio of inflammatory cells to fibroblastic cells at the different postoperative days was calculated. For one tissue section, each of these factors was estimated four times. The average was calculated and the overall averages for all animals at a given period of time were compared using the test of Newman-Keuls.
prosthesis. At postoperative day 7 this “percentage resistance” of the abdominal wall repaired with silicone cloth was only 56 per cent, whereas with the Dacron knitted mesh and the stainless steel mesh it
Results Mechankal Reeletance of the AbdominalWall
Table I summarizes the results of studies in which the bursting strength of wounds in the abdominal wall of rats were determined. In the early postoper-
ative period, the rupture occurred at the junction between the prosthesis and the muscle. After postoperative day 30, the rupture of the abdominal wall occurred in a site other than that of the prosthesis, and these bursting pressures were used as references. Thus, the bursting strength of abdominal wall repairs before postoperative day 30 were also expressed as a percentage of the abdominal pressure needed to
rupture the abdominal wall at a point other than the
Figure 1./mplantat/on of prosthetic material thwgh culature and peritoneum.
I Figure 2. Determinatton of the abdomlnal wall resisthce.
volmm
199, awdl1977
PUMP
INJEClKm
mw-
Arnaud et al
TABLE
I
Mechanical
Resistance
of the Abdominal
Wall
(mm
woven Postoperative Postoperative Postoperative
TABLE
II
day day day
7 15 30
Behavior
Mesh
158 188 > 240
f 1.3 i- 1.6
of the
Incorporation Extrusion Encystment with partial incorporation Encystment without incorporation lntraperitoneal migration
Dacron
Nylon
Foreign
192 197 > 240
Material
Knitted * 2 f 1.4
(after
Stainless Steel Mesh 181 191 > 240
Silicone
* 1.6 f 4.4
postoperative
day
135 138 >240
Siliconized Dacron Cloth
Cloth + 3 r 1.7
154 183 > 240
+ 1.4 * 1.4
Siliconized Dacron Velvet 158 158 > 240
? 1.6 t 3.5
15)
Woven Nylon Mesh
Dacron Knitted Mesh
Stainless Steel Mesh
100% -
100% -
100% -
Silicone -
Cloth
Siliconized Dacron Cloth -
Siliconized Dacron
Velvet -
5% 60%
80%
5% 70%
-
-
-
15%
10%
25%
-
-
-
20%
10%
0%
increased to 75 to 80 per cent. The remaining prostheses gave intermediate values ranging between 64 and 66 per cent. At postoperative day 15, it appears that the resistance of the abdominal wall repaired with silicone cloth (57 per cent) or siliconized Dacron velvet (66 per cent) was not significantly different from that observed at day 7. On the other hand, the resistance of repairs with nylon woven mesh or siliconized Dacron cloth had significantly increased at this time to 78 per cent and 76 per cent, respectively. These values were similar to those observed with Dacron knitted mesh (82 per cent) and stainless steel mesh (79 per cent). MacroscopicFindings Serial clinical examination demonstrated a strong repair of the abdominal wall and no occurrence of hernia except in one case with a silicone cloth prosthesis (postoperative day 7). Fragmentation of the stainless steel wire was observed in three cases but was not associated with impairment of the abdominal wall repair. Peritonization began with the appearance of a glistening membrane which covered the inner surface of the material both at its periphery and around the omental adhesions to the foreign material. Complete peritonization occurred generally with mesh materials by postoperative day 7, whereas 15 days were necessary for complete peritonization in the case of cloth materials. Peritonization, however,
340
Hg * SEMI
was
always incomplete with stainless steel mesh even three months after operation. Omental adhesions were regularly found whatever material was used. They seemed more firmly fixed to mesh material. These omental adhesions were sometimes associated with adhesions of bowel but no intestinal obstruction was observed. After mesh material implantation, fibrous tissue infiltrated between the filaments of the thread, providing a semirigid framework. There was no tissue fluid space in this experimental group. On the other hand, cloth prostheses became loosely invested by thin fibrous tissue and were enclosed in a bursa-like envelope formed by proliferating connective tissue. Sometimes, the cloth materials could be easily removed and serosal collections were found within the capsule. These findings were more marked in the case of Dacron siliconized velvet. As summarized in Table II, the nonincorporation of the foreign material can lead either to its extrusion or to its migration as a foreign body into the peritoneal cavity. Infaction
After implantation of mesh materials, neither wound infections nor serosal collections were observed and this was confirmed by negative bacteriology. On the other hand, serosal subcutaneous collections developed regularly after the use of cloth material. Their evacuation was necessary but this did not always prevent the appearance of subcutaneous wound infections, which occurred in 14.3 per cent
The American Journal of Surgery
Critical Evaluation of Prosthetic Materials
EXTENT WE
OF
CEWUR
FIBROBLASTIC
Figure 3. Extent of the cellular reaction after nyh woven mesh (C), Dacron knltted mesh (M), silicone cloth (S), siNconized Dacron cloth (SR), and slliconlzed Dacron velvet ( R) implantation.
GIANT
15
(siliconized Dacron velvet), 8.5 per cent (siliconized Dacron cloth), and 5.7 per cent (silicone cloth) of the experimental animals. Moreover, a mucopurulent exudate found around the cloth material was noted in 10 per cent of the experimental animals after silicone cloth or silicone Dacron cloth implantation, and in 25 per cent after silicone Dacron velvet implantation. These infections were due to staphylococcus (61 per cent), proteus (22 per cent), and pyocyaneus (16 per cent). MicroscopicFlndings
During the first seven days, the histologic reaction was similar for all tested materials. The following description concerns only the chronic (more than 15 postoperative days) rather than the acute responses of tissues to foreign materials. The overall microscopic picture of tissue reaction concerned the extent of the cellular reaction and the number of inflammatory, giant, and fibroblastic cells. Between postoperative days 15 and 90, the extent of the cellular reaction around the prosthesis varied as follows. (Figure 3.) After siliconized Dacron velvet implantation, it was rectilinear, “a” being not significantly different from zero. The cellular reaction with siliconized Dacron cloth or silicone cloth implantation decreased significantly (p
vohlmo 199, Mwch
mv
30
60
CELLS
CELLS
90 POST
OPERATIVE
MY’S
Ngure 4. Comparative evolution of the different hlstologk parameters under study during SO days after knpbntatkm of the pro&et/c mater/a/s. ( R = Mconked Dacfon velvet; St? = silkzonized Dacron cloth; S = s&one cloth; M = Dacron knitted mesh; and C = nylon woven mesh. )
cellular reaction after Dacron knitted mesh, however, increased between postoperative days 15 and 30 as well as between days 60 and 90 (p
341
Arnaudet al
Figure 5. Emsneration of the inflammatory (/en), fibroblastlc (m/M/e), and giant (right) ceils observedper sectional area. ( R = silkonized Dacron velvet; SR = sMconized Dacron cloth; S = sllkone cloth; M = Dacron knitted mesh; and C = nylon woven mesh.)
inflammatory cells between days 15 and 30 and days 60 and SO (p
342
knitted prosthesis was associated with an infiltration of fibroblasts which increased in accordance with the formula y = 0.79 t + 280.54 (y = number of fibroblasts, t = postoperative days). At each postoperative time, Dacron knitted mesh, nylon woven mesh, siliconized Dacron velvet, silicone cloth, and siliconized Dacron cloth were prosthetic materials of decreasing potency (Figure 4) in the induction of infiltration by fibroblasts (p <0.05). The number of giant cells (Figure 5) observed was constant throughout the postoperative period for each foreign material except Dacron knitted mesh with which it decreased at postoperative day SO (p <0.05). Plotted together (Figure 4), it appears that the number of giant cells increased progressively when silicone cloth, nylon woven mesh, Dacron knitted mesh, siliconized Dacron cloth, and siliconized Dacron velvet were implanted. The ratio between fibroblasts (Figure 6) and inflammatory cells has been established for each foreign implant at the different postoperative times. This ratio generally increased between postoperative days 15 and SO (p
The American Journal of Surgery
Critical Evaluationof ProstheticMaterials
between each material suggests that the ratio between fibroblasts and inflammatory cells was higher for mesh than for cloth materials. Figure 7 illustrates the different histologic features. Comments
In the present report, mechanical, macroscopic, bacteriologic, and histologic criteria were followed to provide a quantitative critical evaluation of different prosthetic materials which are commonly used. The resistance of the repaired abdominal wall and the tolerance to the foreign material were studied. The bursting strength of abdominal wall repairs with different prosthetic materials suggest that after postoperative day 30 there is no difference whichever material was used. (Figure 7.) Moreover, our findings corroborate previous reports that a healing wound is more resistant than the nonrepaired abdominal wall [6-8]. However, striking differ.ences were observed in the early postoperative period (15 days) in the resistance of the abdominal wall to the different prostheses. During this period, disruption of the wound occurred quite readily and was generally located at the junction between the prosthesis and the muscle, rupture of the prosthesis itself never being seen. This early low strength of the abdominal wall corresponded to the “lag period” of Howes [S] and might be considered to be responsible for the later recurrence of a parietal hernia. Therefore, to produce an early resistant wound it is necessary to choose an adequate prosthetic material. To this end, our results suggested that mesh materials and to a lesser extent the siliconized Dacron cloth enhanced markedly the bursting pressure of the wound. The general appearance of the host tissue reaction around the prosthesis was found to be the most useful macroscopic criteria of tolerance. In all animals repaired with mesh materials there was strong incorporation of the implant within the surrounding tissue. On the other hand implantation of cloth materials was characterized by encystment of the prosthesis with or without its partial incorporation. Similar results have been reported in tests of nonporous weaves [4, 9-111. The nonpermeability of cloth materials might account for the occurrence and persistence of a serosal collection and prevent its drainage towards the peritoneal cavity. As described by Scales [4] and Harrison, Swanson, and Lincoln [12], a fibrous tissue capsule also developed around the nonporous material unlike the microporous mesh material which was invaded by the tissues and incorporated in them [13]. Finally, the cloth materials appeared enclosed but free within a fibrous capsule regularly containing a serosal collection and these
voknn
‘tssl Mwab 1917
92
1
0.1 0
I
15
30
60
90 POST OPEIUTIYE Mw
Figurn 6. Evolution &wing 90 postopefatlve days of ratlo oil&olblsbto~~(R=~Dacron velvet; SR = sWodzed Dacron cbrsl; S = s&-one cloth; M = Dacron knltted mesh; and C = nylon woven mesh. )
conditions may explain the frequency of infection confirmed by bacteriology. Encystment, nonincorporation of the foreign material, and infection lead in a final stage to the extrusion of the implant. These events have never been observed with mesh materials. The tissue reaction to the implanted materials was also examined at the cellular level. The width of the cellular reaction around the prosthesis might be considered as a criterion of tolerance. Excluding the first postoperative month, the extent of this cellular reaction was similar for each material tested except for the Dacron siliconized velvet. In this case even after three months the extent of the reaction was four times greater than that of other prostheses. This histologic assessment corroborated macroscopic findings that the implantation of this kind of material produced the most extensive fibrous capsule. Giant cell infiltration was also used to assess tolerance. Materials which evoked this response were unlikely to be satisfactory. It has been suggested that the presence of those cells reflects instability of the surrounding tissue [IO]. During the postoperative follow-up, this feature remained relatively constant for each material tested. The mean values were compared and except for siliconized Dacron velvet, which exhibited a striking gigantic cell infiltration,
343
Arnaud et al
Figure Mkroscopk 7. plate. Comparative histologk find&s observed 30 days after imp/antatlon of nyIon woven mesh (C), Dacron knitted mesh (M), silicone cloth (S), sillconized Dacron clot% (SR), and siikonized Dacron velvet (R) materials. (Hematoxylin and eosfn st&; magnifkatkn x 25.)
no significant difference was noted between mesh and cloth materials. The importance of what has commonly been called “chronic inflammation” has been investigated by counting the inflammatory cells. Generally, their number decreased with time for each foreign material tested. Again siliconized Dacron velvet showed a marked reaction but mesh materials generally, seemed to induce a slightly greater response than cloth materials. The number of fibroblasts present was taken as a microscopic index of the mechanical
344
resistance of the repaired tissue. In the early postoperative period, these results correlated well with the macroscopic and bursting studies, again showing superiority of mesh materials. Inflammatory cells and fibroblasts represent the main cellular components of a healing wound [14] and they may represent two different evolutionary steps of the cicatrisation procedure. Whereas the former could be considered to reflect the tolerance of the host to foreign material, the second is the histologic feature corresponding to the mechanical re-
The American Journal of
Surgery
Critical Evaluation of Prosthetic Materials
sistance of the repaired wound. Therefore, their ratio appears of particular interest: the larger the ratio, the stronger the wound and the better the tolerance of the host. The order in which the materials are ranked by this ratio, in decreasing magnitude, is mesh material (Dacron knitted and nylon woven mesh) and then cloth materials (silicone and siliconized Dacron cloth) and finally siliconized Dacron velvet. Among the eight properties required classically [3, 41 to characterize foreign prostheses, tolerance and mechanical resistance have been particularly investigated in this study. Macroscopic, mechanical, bacteriologic, and several histologic parameters have been compared in a critical study. It appears that alone none of these parameters is sufficient to fully characterize a foreign material. This combination of macroscopic, mechanical, bacteriologic, and histologic analysis proves that the physical nature of an implant influences directly the biologic reactions which we have considered. Thus, on present evidence it would seem reasonable to recommend mesh rather than cloth materials in the repair of abdominal wall. Among the numerous criteria investigated, it must be outlined that one of them, the ratio of fibroblastic cells to inflammatory cells expresses best the macroscopic, mechanical, and bacteriologic findings and reflects the resistance and tolerance of the foreign materials. This histologic test may be at present a helpful index to evaluate in experimental or clinical trials.
Summary
Six different materials (3 mesh and 3 cloth) commonly used for the repair of abdominal wall hernias were evaluated in 180 rats after implantation through musculature and peritoneum. Serial macroscopic and bacteriologic investigations were done, and bursting strength of the wound was determined in all groups at intervals with an original device. Histologic criteria were used to characterize the resistance of the wound and the tolerance of the host to the foreign material. Statistical analysis of the results demonstrated: (1) after postoperative day 15, the resistance of the wound was similar for each material tested; (2) dur-
vohtmo 133, Much 1977
ing this early period mesh materials exhibited more resistance to bursting pressures than cloth materials; (3) the incorporation of mesh material was constant, whereas encystment or extrusion was always observed after implantation of cloth material; (4) no infection occurred with mesh material, but significant bacteria were found in 18 per cent of cloth material implantations; (5) the extent of the cellular reaction, the enumeration of giant, inflammatory, and fibroblastic cells showed the superiority of mesh materials; and (6) the ratio of fibroblasts to inflammatory cells reflected closely the mechanical resistance and tolerance of the foreign material. References 1. Witzel 0: Ueber die Verschliessung von Bauchwunden und Brustpforten durch Versenkte Silberdrahtnetze. Cent Chir 27: 257, 1900. 2. Goepel R: Uber die Verschliessung Von Bruchpforten Durch Einheilung Geflochtener Fertiger Silberdrahtnetze. Vet-h Msch Ges Chir 29: 4, 1900. 3. Cumberland VH: A preliminary report on the use of prefabricated nylon weave in the repair of ventral hernia. Med J Aust 1: 143, 1952. 4. Scales JT: Discussion on metals and synthetic materials in relation to soft tissues; tissue reaction to synthetic materials. Proc R Sot Med46: 647,1953. 5. Amaud JP, Adloff M, Weill-Bousson M: Prosthetic materials in the repair of the abdominal wall. New criteria of tolerance and resistance. Eur Sot .Exp Surg (10th Congress): 215, 1975. 6. Adamsons RJ, Kahan SA: The rate of healing of incised wound of different tissues in rabbits. Surg C+neco/ Obsfet 130: 237, 1970. 7. Cerise EJ, Busuttil RW, Craighead CC, Dgden WW: The use of Mersiline mesh in repair of abdominal wall hernias. Ann Surg 161: 726, 1975. 6. Howes EL: Healing of fascia defects, p 201. La Cicatrisation (CNRS ed). 1965. 9. Koontz AR, Kimberly RC: Tentalum and marlex mesh: experimental and clinical comparison. Preliminary report. Ann Surg 151: 796, 1960. 10. Calnan JS: The use of inert plastic material in reconstructive surgery. I. A biological test for tissue acceptance. II. Tissue reactions to commonly used materials. Br J Pled Surg 16: 1, 1963. 11. Narat JK, Khedrou LG: Repair of abdominal wall defects with Fort&n Fabric. Ann Surg 136: 272, 1952. 12. Harrison JH. Swanson DS, Lincoln AF: A comparison of tissue reactions to plastic materials. Arch Surg 74: 139, 1957. 13. Usher FC, Gannon JP: Marlex mesh, a new plastic mesh for replacing tissue defects. Experimental studies. Arch Surg 78: 131, 1959. 14. Hartwell SW: Research and clinical observations related to the healing of human wounds, burns and stasis ulcers, p 235. La Cicatrisation (CNRS, ed). 1965.
345