Document B84Zzq6ZV8bZ56aqq4gXJebdo

Reprinted from " Rritith Journal of Plastic Surgery," Vo/. XV/, No, 1, January 1963 THE USE OF INERT PLASTIC MATERIAL IN RECONSTRUCTIVE SURGERY I. A Biological Test for Tissue Acceptance. II. Tissue Reactions to Commonly Used Materials. By James Calnan, F.R.C.S. Experimental Plastic Surgery Unit, Postgraduate Medical School, University ofLondon. From the earliest days of the specialty, plastic surgeons have been concerned -with inert materials for reconstructive procedures. In spite of half a century of endeavour the use of such material is unsatisfactory. So much so that an authoritative textbook on surgical materials noted: " In spite of all the good results which have been obtained by such means . . . sooner or later the body attempts to rq'ect such foreign material and it therefore can never be considered the ideal substance for the purpose " (Gillis, 1958). Yet there are clearly occasions when the provision of an implant of foreign materials, acceptable to the tissues, helps both patient and surgeon. This paper attempts to evaluate the need for and uses of inert materials, to present a laboratory method for investigating the biological reactions to them, and to examine experimentally such reactions when materials, in present-day use, are implanted in the rat. It thus forms part of a wider investigation which seeks to lay down standards of quality and technique for implant surgery. Surgical opinion is divided on the question of the use of foreign material in reconstructive surgery. There are those who hold that such material should be used only as a temporary measure until the time when reconstruction can be completed by living tissue, while others are prepared to consider it permanent. The division of opinion is not limited to plastic surgeons for the subject is very much alive in orthopaedic, vascular, and cardiac surgery. The main arguments are summarised below:-- The Case Against Foreign Material.--(1) Reaction of the tissues to a foreign body: This may be immediate, resulting in rejection of the material and usually associated with infection, or delayed for weeks, months, or years. Delayed reaction may occur with infection or collections of sterile fluid, but -usually results in eventual extrusion. (2) Movement of a foreign body is well recognised and many well-placed implants have become less satisfactory for this reason. Where bone has been replaced by the implant, fixation presents its own problems: in orthopaedic M surgery where such are used to aid ununited fractures of the long bones, stress ! is a big factor (Scales, 1958). The use of tantalum, much in vogue between the two World Wars, did show that once the material was " on the move," extrusion -was the usual finale. (3) Alteration of physical properties of the implant. Scales et al. (1961) have shown that many of the stainless steels used in orthopaedic surgery corrode and lead to stress fractures, and have demonstrated crazing in acrylics. Polyethylene 1A X THE POSTS VED'IVM. 10*00* BuuAkfc aoiu. uiuUuiL w.u 20554001 BFG03868 2 BRITISH JOURNAL OF PLASTIC SURGERY foam, recommended for breast construction, has altered within a year from a soft to a very hard material (the " marble breasts " mentioned by Moran in 1954). (4) Stimulus to carcinoma: In some experimental animals implant materials do appear to produce malignant disease. This is invariably a sarcoma (Scales, 1958), and never a carcinoma, in the fibrous capsule surrounding the implant. In the rat it takes twelve to eighteen months after implantation for malignancy to appear and this has been found to occur with a wide variety of materials--Polystrene, cellophane, P.V.C. (Northdurft, 1956); Dacron, nylon, Ivalon, Teflon, and silk (Oppenheimer et al., 1955). The very diversity of the nature of these should make one suspect that malignant change is a feature of rat tissue. Scales (1961) has noted that in the guinea-pig malignancy is uncommon. No unequivocal case of malignant change in man has so far been recorded although individual cases where neoplasia was related to metallic implants after thirty to thirty-five years have been published (Siddons and MacArthur, 1952; Penn and Epstein, 1953 ; McDougall, 1956). (5) There is a general feeling of disquiet, difficult to put into words, that a material which is not part of the body and has no blood supply is " at risk." Curiously, the same argument does not apply to homograft cartilage. The Case for Foreign Material.--(1) It makes reconstruction shorter and less traumatic for the patient. (2) There is an endless supply. (3) Some natural tissues when transplanted do not behave well: fat " absorbs," tendons become adherent, fascia becomes vascularised and markedly adherent, bone may absorb in certain situations, and cartilage warps. (4) A special case can be made out for the use of foreign material where the body has an inadequate supply in quantity in children, or in quality such as tendon with paratenon. (5) It is expendable in a growing child, or where uncertainty of success may retard the progress of reconstruction. There is, of course, much to be said for both points of view and indeed protagonists and antagonists are much less consistent in clinical practice than in opinion, although the points for and against the use of foreign material are representative. At least one reason for this ambivalence is the uncertainty associated with the lack of precise knowledge of the tissue reaction to foreign material. Of the thirty-eight papers published op, this subject in the past ten years only ten record histology and a mere four produced experimental evidence. In none was there sufficient detail to allow a reader to make an intelligent assessment of the particular material being recommended for clinical use. In this sense, then, most previous publications should be considered pilot trials. Material and Methods A pilot scheme for the experimental investigation was set up, using Wistar (white Norwegian) rats. The experimental design was of the most simple kind, rats and material, in sterile (autoclaved) containers, being randomly associated. 20554002 BFG03869 4 BRITISH JOURNAL OF PLASTIC SURGERY Table I Weights of Rats at Time of Implants Material Implanted P.T.F.E. film P.T.F.E. felt Polyethylene sponge Polyethylene sheet Orion .... Terylene .... Nylon .... P.V.C........................................ Estane .... Mean Weight g172 156 184 215 l6l 207 155 194 183 Standard Deviation g28 4 26-3 34'8 191 16-9 45'3 297 5i-3 44 7 to the muscle of the anterior abdominal wall, but others may represent the host's reaction to an irritant. The materials implanted, of 2 by 2 cm. area, were as follows:-- 1. Polytetrafluorethylene (P.T.F.E.) sheet, 0-2 mm. thick. 2. P.T.F.E. felt, 3 mm. thick. 3. Polyethylene sponge, 5 mm. thick. 4. Polyethylene sheet, 0-02 mm. thick. 5. Polymethyl methacrylate (Orion) woven fabric, 0-3 mm. thick. 6. Polyethylene terephthalate (Terylene) woven fabric, 0-15 nun. thick. 7. Polyhexamethylene adipamide (nylon) sheet, 0-07 mm. thick. 8. Polyvinylchloride (P.V.C.) sheet, o-i mm. thick. 9. Polyurethrane (Estane) sheet, 1 mm. thick. Results Tissue reaction to the various materials was assessed macroscopically and microscopically. Macroscopic.--Mesentery, playing its part of " policeman " in the abdominal cavity, was frequently found adherent to part or whole of the implant material. Adhesions of bowel, to a marked degree, were commonly found with Orion, Terylene, nylon, P.V.C., and Estane, and to a milder degree with Polythene. No adhesions of bowel were found when the implants were P.T.F.E. The results are summarised in Table II. The Chi square test at the o-i per cent, level of significance confirms that P.T.F.E. behaves much more favourably than any other material. The photographs in Fig. 2 show comparable findings at the time of naked-eye inspection. Microscopic.--Histological notes are given in the legends to the photo micrographs of individual materials (Figs. 3 to 13). The general picture of the findings is described here. The overall microscopic picture of tissue reaction to foreign material concerns round cells, macrophages and giant cells, fibroblasts and fibrous tissue, the vascular supply, and, since these implants were exposed in the peritoneal cavity, the nature of any covering membrane. 20554004 s,rG03870 THE USE OF INERT PLASTIC MATERIAL IN RECONSTRUCTIVE SURGERY J Implant Technique.--Previous experiments in rats, and in man, demon strated that insufficient information concerning the tissue reaction to inert materials is obtained by subcutaneous-implantation only; intraperitoneal implants on the other hand seemed likely to provide more sensitive tests, and offered the advantage of intermittent inspection and biopsy. DIAGRAM OF IMPLANT TECHNIQUE Under ether anaesthesia a midline abdominal incision opened the peritoneal RAT IN POSITION SHOWING LINE OF INCISION PERITONEUM S THIN LAYER OF MUSCLE REMOVED (2*2cm HUr> cavity and an area of 2 by 2 cm. on its anterolateral wall was denuded of peri toneum with a thin layer of underlying muscle. The raw surface so produced was covered by the implant material held in place by silk sutures and the abdomen closed (Fig. 1). A gauze dressing over the wound was secured by a couple of turns of adhesive Elastoplast bandage, to be discarded after one week. Forty to ninety days later the peri toneum was inspected for macroscopic signs of reaction, the animal sacrificed, and the whole implant with its sur rounding tissue removed for histology. Specimens were fixed in formol-saline, Bouin, or Zenker. Paraffin sections cut 5 to 7 microns thick were stained by haematoxylin and eosin, Mayer's add haematoxylin, Harris's haematoxylin, van Gieson, Weigert's elastic, Solochrome cyanin, Aldan blue, Ehrlich's haema? toxylin. At least ten fields were W0UN9 CLOSED BY THROUGH & TMROUGFMWTRESSJ0IRCS examined in each slide. Implant Materials.--Nine materials in common use were available from the hospital operating theaftp. Each rat received a single implant and the experiment was repeated sn times, i.e., a total of fifty-four rats, this being considered the minimal number required for statistical assessment. The mean weights and standard, deviations of rats used for the various implants are presented in Table I. In each group of six there were an equal number of males and females. One aninli` died (Estane implant) and eight implants became free in the peritoneum (two nylon sheet, three P.V.C. sheet, three Estane sheet) and could not therefore be assessed histologically. Five became completely enveloped by mesentery (three polyethylene sheet, one nylon sheet, one Estane sheet) and provided little microscopic information of tissue reaction to them. Examination of the remaining forty specimens form the basis for the assessment of the cellular response. One disadvantage of the technique of peritoneal implantation presented here is the loss of information that can occur when materials fail to remain in situ. Some of these were probably due to inadequate suturing of the material fin p- 20554003 BFG03871 20554005 L809d8 h* uoubuuoj sip sSbjttosus 01 psuisss `uuoj ,, ureoj ,, usdo Xjsa sip ui ApBissdss `snsipi{oj '(^j sfquj,) oojXu puB *nopo ` D'A J sjsass oj `jssqs pqos (zi Sy) sirejsg pire H-Tl'd qJiM punnim uiojj ItnXjBA sjOB|dun jp punoJB usss sbm uopoBSJ snsBjqojqg jo ssjgsp uibjjsd y--'siso^qtj puv sjsviqouqij (HI siqBx) JOEsgraSis AiqSiq si qsiqM `uiopssij jo ssiSsp suo joj fz-zi = zX `sjqBi -aim z Xq z b sb psjusssjj pimoj JSASU SJSM Sips JUBlS jssqs `H'dX'd qjTM pUB HOpSBSl pSS-]OBT sssi XipsqiBui psqoAS uuoj jssqs ipoouis JO SJBUSJBJY `01 PUB `i `9 ` -sglj ui usss sq obs sisqjo spqM `[busjbui jnBidim sip jo sjusio2bjj Sutuibjuod sssip jo snios ssjajjsuounp n -Sij -(sSuods snsqjXjoj `susjXjsx `nopo `JRJ 'H'd l'j) nononxTsnos qssta nsdo jo usaom jo sibusjbui ssoip jp mqjiM puB panojB panoj sjsai `mioj sjJBZiq jo usyo `sqss tobtq--`s]pj juvtQ puv sadvqiojovyq 'z -amsuoiusd ibuuou Xq psJSAOD [eusins ptre suoisaqpE on (-sXsp aAg-XiusAas ie j jbh) -jssqs -g j'X'J 'in uotntwsqo pranswii jo tssiSsp Jourai qM pMoq jo uotssqpv s; sisqj qioq uj (-sXvp SAg-Xuns jf 'zz irjj) qiop bsaom susiAjsj. -ii XqBspspBjs jnBsqraSis jsnf Xpo sbm ssusisgrp sqj jnq c(oi pire `6 `8 sSuj) spiJsiBui jsipo UBip uopBuigui ipo-panoi sssj sshbo oj psnunuos SABq oi psjBsddB (Z oj 'sSy) susjXjsj, pus `uopo *'3 d'l'J : ,, sjsass siEjspoui ,, jo ,, jqSqs ,, sb Surapuisi uopsBSj sqj spBjS oj sjqissod sbm it sqssM xts jsjjv aonwnBiduii jsjjb uoos psaiurexs spusjBui pe punojB punoj sjsm `uoijdbsj XjojBuiurepui ire jo usd sb `spso panoj sSjbi puB pBuis--'SU3D puno# -i S AJiaoBas 3AiisnHisNosa ni iviHaivw diisvtj xaaNi ao asn 3Hi 6 BRITISH JOURNAL OF PLASTIC SURGERY Table II Gross Appearances of Implants : the Presence of Adhesions of Mesentery, Bowel Material P.T.F.E. . All others . Totals 1 No Adhesions 1 j9 i3 j 12 Adhesions 3 37 40 x*=20-5 d = i P<o-ooi Totals 12 40 52 Table III Influence of Physical Form of Implants. The Presence of Giant Cells Implant Sheet .... Woven or open mesh Totals Giantjjglls Present Absent 6 16 18 0 24 l6 Totals 22 18 40 Xs = 12-2 d.f. = 1 P<0-001 Table IV Reaction by Fibroblasts and Fibrosis Material Reaction Slight Moderate or Severe Totals P.T.F.E. All others 9 3 12 4 24 28 Totals 13 27 Xs =9 d.f. =i P<o-oo5>o-ooi 40 2055400& BFG03813 20554007 Fig. 3 Polytetrafluorethylene (P.T.F.E.) felt: biological modification of its physical structure i. Specimen of P.T.F.E. felt before implantation ( x 67). Note its fibrous nature and spiked ends. It does not pick up histology stains. 11. P.T.F.E. felt four days after implantation (Rat 58). The material is still spiky and' giant cells (e) can be seen at the free edge of the material. (Mayer's acid hzmatoxylin stain, x 67.) ill. A later stage at seventy-six days (Rat 3). There are fewer giant cells (e) but an increase of interstitial fibrosis (b). (Mayer's H. & E. x 67.) iv. At ninety-four days the P.T.F.E. has become matted and smoothed off. Giant cells are now scanty (E). (Rat 46.) (Ehrlich's H. & E. < 67.) BFG03874 8 BRITISH JOURNAL OF PLASTIC SURGERY of fibrous tissue within its interstices to such an extent that the original shape macroscopically was always deformed. Fibroblasts, fibrous plaques, and giant cells were always seen in every field of every specimen of Polythene foam examined (Fig. io). This perhaps is not surprising when one remembers that Dunphy (i960) has recommended the use of this material as an implant to harvest young fibrous tissue. From the table it will be noted that P.T.F.E. produces less fibrous tissue reaction than other materials, significant at the 0-5 per cent, level of confidence. It is notable that with woven materials (Orion, Terylene) fibrous penetration of the material was unusual (Figs. 5 and 6). I. P.T.F.E. felt to show lack of reaction and covering of mesentery (m). The interstitial reaction is still cellular and the lack of bundles of fibrous tissue is notable (compare Polythene, Fig. 10). (Rat 58 at four days.) (Ehrlich's H. & E. X67.) II. P.T.F.E. felt at seventy days to show localised abscess (p) within the material. The clear spaces are artefacts occurring when this tough material is cut on the microtome and the soft tissues are dragged away. 4. Vascular Connections.--In the majority of materials it was not possible to show any abnormality of the vessels proceeding to and from the area of the implant, and the fibrous capsules were relatively avascular. With solid materials this was expected except for the unusual phenomenon of " incorporation " of P.T.F.E. sheet which will be described later. With woven materials, as already noted, tissue reaction was confined to the surface and, although there appeared to be an increased number of small vessels at the edges of these implants, the vessels themselves did not appear abnormal. By contrast, in those materials of more open texture (Polythene foam, P.T.F.E. felt) primitive vascular spaces were commonly seen within the implant. These spaces, varying in size from 10 to 100 microns and even larger (Fig. 10) did not appear to be lined by endothelium 20554008 BFG03875 THE USE OF INERT PLASTIC MATERIAL IN RECONSTRUCTIVE SURGERY 9 in every case. Serial sections confirmed that they were not areas of haemorrhage within the implant, as at first thought. Although it was not possible to demonstrate their connection with afferent or efferent vessels, their pattern of direction was consistent with such an assumption. Vascular spaces such as these would appear to be unsatisfactory and possibly could produce a haemolytic anaemia of mechanical origin. 5. The Covering.--As already mentioned under macroscopic findings, mesentery was the commonest covering found on the peritoneal aspect of the Fio. 5 P.T.F.E. sheet. 1. General view of encapsulated implant. (Rat 25 at sixty-four days.) Note thin capsule of very cellular fibrous tissue (a) surrounding the implant (B), the relative absence of round cell infiltration or giant cells. (E. H. E. x 67.) 11. Junction of implant (b) and capsule (a) where a cellular element with vascular spaces is notable. (Rat 40 at sixty-five days.) (Mayer's H. & E. x 67.) (Set also Fig. 13 of incorporated P.T.F.E. sheen) implants. From the clinical point of view the most satisfactory implant is one which has become completely- covered by a fibrous tissue envelope and is thus virtually extracorporeal. It is known that this of itself carries the risk that any blood-borne infection which settles within the fibrous envelope will also exclude the natural body mechanisms for its defence. Hence abscess formation requiring removal of an implant many years later has become well recognised. The fibrous tissue envelope or capsule does, however, ensure that at the cellular level all is quiet. While sheet (solid) materials were usually encapsulated in this way, open mesh materials never were completely. At one or more areas giant cells and round cells were always found, indicating that biological activity was still continuing (Table V, p. 17). Such histological findings must indicate their unsatisfactory nature for use in reconstructive surgery. BFG03876 20554009 10 BRITISH JOURNAL OF PLASTIC SURGERY Fig. 6 Orion woven doth. I. General view. (A) Thick fibrous covering of implant. (B) Woven cloth with no tissue permeation, (c) Individual loose fibres surrounded by fibrous tissue, containing giant cells, (d) Deep portion of fibrous capsule and normal muscle of abdominal wall. (Rat 88 at forty-nine days.) (H. & E. (transverse section), x 67.) n. Longitudinal section to show fibrous capsule (A), lack of tissue permeating the woven cloth (B) and multiple giant cells (e). Beyond the fibrous capsule there is little cellular reaction. (Rat 88 at forty-nine days.) (Mayer's H. & E. x 67.) ill. Local encapsulation of frayed ends of implant (a) within the main fibrous capsule. Note giant cells (e). (Rat 24 at sixty-four days.) (Mayer's H. & E. x 67.) iv. Bizarre-looking giant cells (e) ingesting Orion fibres. (Rat 90 at forty-four days.) (Mayer's haematoxylin-van Gieson. x 250.) 20554010 BFG03877 THE USE OF INERT PLASTIC MATERIAL IN RECONSTRUCTIVE SURGERY II Fig. 7 Terylene woven cloth. i. General view. (Rat 23 at forty days.) As for Orion (a), (b), (c). Note giant cells (s) at cut edge of implant. (Ehrlich's H. & E. * 67.) 11. To show fibrous capsule (a) dissected free from surrounding tissue and subsidiary fibrous tissue enveloping individual terylene fibres at cut end of implant (G). (Rat 22 at sixty-five days.) (Mayer's hsematoxylin-van Gieson. x 67.) in. Giant cells (e) enveloping free ends of woven cloth. Oblique section. (Rat 4 at seventy-six days.) (Ehrlich's H. & E. x 67.) Where mesentery or bowel did not cover the implant, a layer of condensed fibrous tissue with occasional fibroblasts was found. Over only two implants of solid P.T.F.E. sheet were endothelial cells found microscopically. 6. Incorporation.--It has been stated by several writers (Yeager and Cowley, 1948; Grindlay and Waugh, 1951) that normal tissue has been found permeating polyethylene sponge and the impression given that this state of affairs is ideal. While concurring with the latter, it must be pointed out that the findings from 20554011 BFG03878 12 BRITISH JOURNAL OF PLASTIC SURGERY Fig. 8 Nylon sheet. Rat 36 at sixty-five days, (a) Fibrous capsule, (b) Fragmented portions of nylon associated with giant cells. Note the general disorganisation of the cellular elements. (Mayer's haemataxylin-van Gieson. x 67.) Fig. 9 P.V.C. sheet. Rat 38 at fifty-nine days. Note the fragmented material (b), giant cells (e), and highly cellular reaction (m) composed of round cells and histiocytes. (E. H. E. * 67.) 20554012 BFG03879 THE USE OF INERT PLASTIC MATERIAL IN RECONSTRUCTIVE SURGERY 13 Fig. IO Polythene sponge. I. General view. (Rat 18 at seventy-seven days.) Relatively avascular fibrous tissue (a) permeating the interstices of the sponge (B). Note frequent giant cells (e). (Mayer's H. & E. x 67.) II. Giant cells (e) invading the Polythene (B). Note accompanying histiocytes and fibroblasts. (Rat 80 at fifty days.) (Mayer's H. & E. x 250.) ill. Small vascular spaces (l) within the Polythene sponge--some lined by endothelium, the majority unlined. (Rat 81 at fifty days.) (H. & E. x 250.) iv. Large vascular spaces (L). As ill above. ( x 100.) 20554015 BFG03880 14 BRITISH JOURNAL OF PLASTIC SURGERY Fig. 12 Fig. II. -- Giant cells containing material. (Arrows indicate giant cellscontaining portion of material within cytoplasm.) I. Nylon sheet. (Rat 36 at sixty-five days.) (H. & E. x 250.) II. Orion woven doth. (Rat 88 at forty- nine days.) (Ehrlich's H.&E. x 250.) in. Polythene sponge. (Rat 40 at sixty- one days.) (Ehrlich's H.&E. x 250.) Fig. 12.--Estane sheet. Complete en capsulation of material (B) which does not accept histological stains. Note (a) fine fibrous tissue capsule, (e) two marginal giant cells. (Rat 13 at fifty days.) ' (Ehrlich's H. & E. x 67.) Fig. 11 our experiment indicate that scar tissue infiltrates polyethylene, and can hardly' be called " normal tissue." Two specimens out of six, of P.T.F.E. sheet 0 3 mm. thick, were found to be covered by normal peritoneal endothelium (Fig. 13). Microscopically the endothelial cells stained normally with basophilic dyes and clearly received 20554014 BFG03881 meooja STOfeSSOZ SBq 3]OtJ3)JE J38jE[ 3qj_ ('SSz x ) -JB03 JEipaUI PUE IBUmUI p33|JBUI-I13.Vi '3'J'l'd J 133q$ 3q3 UIipiAV S|3SS3A 8uil|OOHEUUON 'A1 (SSz x ) 'snssn ui spsssA psiuEMun jo Jsqumu sip Supnpsj joj poipsui lEoiSojoiq jBnsn sqi si siq: puE psiojejut Xiqsqojd si SIJ33 ipiM psjpus psssASip : uoiE3i anssu jo 5)3E[ sip 3io(q -(o) spsnui pus (a) 'a'd'X'd uonaunf -hi (SZ X ) -(a) 3UE]dun 8UU3A03 (v) $1133 |E3UO)U3d JBUUOfq 'II pEM JEUIUIOpqE JO spsnoi |BtUJO{q (D) SJ133 (EU0ISB330 pUB S|3SS3A Xq p3)B3UU3d 133qS 'H'Jld (8) `HIBldun p3SOdX3 XjSIlOIASjd JO 8UU3A03 JB3U01U3J (v) (`001 X `a 3J `fj) `D0n33S 3(OI|^ `I -Xixis }B Xsdoiq `z a*a) (`sXsp 33iqi (`3pnp unu z-o `3qs '3'J'l'd) l `Old lUBidtui jo uoiiEJodioouj Si Aaaoans 3AiionnxsNODaH ni ivmaxvw oixsvid xaaNi ao asn am i6 BRITISH JOURNAL OF PLASTIC SURGERY Fig. 14 Adenocarcinoma of large bowel: no relation to implants. I. Clinically obvious ring carcinoma of descending colon (arrow) with subsequent spread. (Rat 56, male, 255 g. weight.) (a = site of P.T.F.E. sheet implant.) II. General view of section of bowel, showing well-differentiated adenocarcinoma (r) invading muscle layer (s). (H. & E. x 25.) ill. General features of adenocarcinoma. This rat was one of two in which carcinoma of bowel was found out of 100 in which the viscera were examined at laparotomy. In this rat a peritoneal implant of P.T.F.E. sheet had been placed on the left anterior abdominal wall one month ago. A diagnosis of bowel neoplasm was made on clinical examination: cachexia and a palpable abdominal mass were present. (H. & E. x 250.) 20554016- BFG03883 THE USE OF INERT PLASTIC MATERIAL IN RECONSTRUCTIVE SURGERY 17 nourishment through the implant material. Many vessels of normal appearance were seen within the sheet of P.T.F.E. and only occasional round cells. At the junction of P.T.F.E. and underlying abdominal muscle the cellular reaction is minimal and in some areas muscle and P.T.F.E. abut without any intervening layer. The P.T.F.E. sheet, which before implantation refused to accept any of the common histology dyes, was readily stained an even bluish-grey with Table V Influence of Physical Nature of Implant on the Total Biological Reaction Material ; l Incorporated or i completely i Encysted Partially Encysted and Giant Cells, etc. Totals Sheet . . . 10 8 18 Others 0 24 24 (woven, felt, sponge) | Totals . I 10 32 4* x=952 d.f. = 1 P<0'0O5 >0-001 hxmatoxylin. This indicated that some biological transformation had occurred in the material which microscopically now measured o-6 mm. thick. Increased water retention (normally o-i per cent.) due to separation of the fibrous elements of the sheet would explain the post-implantation stainability and apparent increase in thickness of the sheet. The P.T.F.E. used was made from granular polymer and no explanation can be offered for the elongated fibres seen in the histological sections. They do bear some resemblance to collagen fibres and it is intriguing to believe that they have been accepted biologically as such. The term " incorporation " is suggested for this particular tissue reaction (Calnan, 1961). Discussion There are several facets of this work worthy of further exploration. 1. On the Technique Presented.--Barr (1953) called attention to the need for laboratory testing of implant materials used in orthopaedic surgery. The requirement is no less urgent in plastic or any other branch of surgery. Barr went further and condemned the present order whereby materials are used clinically and investigated only when things go wrong (cf. Gibson and Davis, 1953). " Scientific testing methods are available which are much more accurate than clinical trial, and the human guinea-pig technique is to a large extent outmoded and indefensible." While no biological study can give a complete guarantee that failures will not occur, the risk will be so reduced as to justify the cost and time of the research. The negative attitude, whereby laboratory ,B 20554017 Bj?G03884 l8 BRITISH JOURNAL OF PLASTIC SURGERY investigations disclose which materials are unsuitable for use in man, is perhaps of even greater benefit. The technique of the testing in vivo of materials in the rat differs from that offered by others (LeVeen and Barberio, 1949; Grindlay and Waugh, 1951; Moore and Brown, 1952; Brown et al., 1954; Harrison et al., 1957). By placing the test material in the peritoneum a macroscopic assessment which does not disturb the implant may be added to the microscopic examination and so increase the precision of observation. By choosing a suitable experimental design and statistical test of significance, it becomes possible to compare accurately and validly the biological reactions between different materials. The combination of macroscopic, microscopic, and statistical analysis ensures that a sufficient number of animals will be used in the test and that the results will be applicable to the animal population concerned. While the exceptional response will be noted it will not become confused with the usual and faith can be distinguished from fact. While it is not claimed that the technique presented cannot be improved upon, the information so derived from it docs allow a valid comparison to be made between the biological reactions to the various implant materials. 2. Macroscopic Findings.--Gross observation of the peritoneum and its contents at varying periods revealed evidence of the intensity of the inflammatory reaction to the implanted material (LeVeen and Barberio, 1949). With P.V.C. and nylon, multiple adhesions of bowel, producing kinking and minor degrees of intestinal obstruction, were invariably found. Rough-surfaced materials (Polythene foam, woven Orion, and Terylene) usually had a thick covering of mesentery but P.T.F.E. felt implants had only a thin film of mesentery covering them. This suggested that there were two factors involved--the physical nature of an open mesh and a chemical difference in this biological situation. When the gross appearances of P.T.F.E. in both sheet and felt form are compared with those of other materials (Table VI) it is clear that there is a highly significant Table VI Cellular Reaction to Physical Form of Implant Materials Materials Cellularicy Slight Moderate or Gross Totals Smooth sheet . (P.T.F.E. and Polythene) Rough .... (P.T.F.E. and Polythene) 5 2 49 10 12 Totals 7 14 X- = i-84 d.f. = 1 P< 0-2 >o-i 21 20554018 BFG03885 THE USE OF INERT PLASTIC MATERIAL IN RECONSTRUCTIVE SURGERY 19 difference in the gross appearances, indicating that of the two factors the chemical is probably more important to the tissue reaction than the physical. The influence of physical form should not be neglected, however. From Table III it is clear that smooth-surfaced materials are less commonly associated with giant cells microscopically. 3. Microscopic Findings.--Clearly the microscopic findings are of the greatest interest because it is possible only in this way to appreciate the reaction to implant material at the cellular level. Yet new materials are recommended for use in reconstructive surgery in man either without any microscopic observations (Blaine, 1946; Farmer, 1947; Rubin et al., 1948; Jeremiah, 1950; Aldunate, 1957 ; Edwards and Lyons, 1958; Gordon, 1958 ; Neuman, 1958) or with inadequate observation and description of the histological changes that have occurred around such materials (Frantz, 1943 > Brown et al., 1948, 1954, i960). Others draw conclusions more favourable towards the new material under examination than could possibly be accepted by others not equally biased. This is an understandable attitude in clinical practice where the urgent need is to find materials to help our patients. Unfortunately, history does not support the good intentions. A pertinent example is the use of polyethylene sponge. In 1951 Grindlay and Waugh reported favourably on its use following implantation for eighteen months in twenty-eight dogs. Much support followed in the literature (Moore and Brown, 1952; Johnson and Grindlay, 1954; Struthers, 1955) which ceased when clinical failures obtruded themselves, to be followed by a new wave of enthusiasm when polyurethane sponge became available. It is only very recently that Gilmer and his colleagues (1961) have pointed out that die " living tissue " which grows into the sponges (in dogs) is " in situ fibrogenesis developing an avascular connective tissue which cannot be likened to granulation tissue." Nor can it be fairly called " normal tissue " for, although the individual elements are not abnormal, the total composition of the tissue is abnormal and unwanted. Sponge has been advocated as a framework for bone formation (Struthers, 1955), and again Gilmer et al. (1961) point out that any bone found within die sponge is formed by a process of metaplasia within the collagen at the periphery: they also noted that while the sponge might form a useful framework for bone, it did in fact cause a distinct delay in bone formation in dogs which prevented healing for as long as one year. From our own observations on polyethylene sponge implanted in rats, giant cells were always seen, and usually in every microscopic field examined, while the fibrosis within its interstices was considerable. These two findings alone should be sufficient to deter anyone from using the material in man except under experimental conditions. It is difficult to understand why they were not noted by early investigators for they would certainly have prevented the wholesale use of moulded implants in breast reconstruction and the subsequent misery of the recipients. It would seem worth while to consider four features of the histological reaction--the intensity of the round cell infiltration, the presence of giant cells, the amount of fibrosis, and the condition of vascular spaces. The technique advocated here allows one to observe whether living tissue will grow over the implant either to seal it off or become a part of it. The main differences between the histological features associated with the various implants have already been 20554019 BFG03886 20 BRITISH JOURNAL OF PLASTIC SURGERY pointed out. The presence of clusters of round cells, commonly called chronic inflammation, near foreign material at any time later than say three weeks after implantation, must always be viewed with suspicion. Clinically, one knows that such foci can erupt into an acute inflammation with little warning. Giant cells always indicate instability of tissues at the cellular level and materials which evoke this response are unlikely to give satisfaction. In our own investigation, a giant cell response seemed to be associated more with the physical form of an implant than with its chemical nature, and such has been noted with silica in man. Silica in the lung produces no giant cell reaction and is by necessity of small particle size, whereas silica under the skin is invariably associated with giant cells. It would be naive to believe that no reaction occurs around any one material, but the superiority of P.T.F.E. in this respect should be noted (LeVeen and Barberio, 1949), fork is only in comparative experiment* that~a greater or lesser biological reaction to foreign material may be demonstrated with certainty, and this has been the error of many previous investigations. Harrison et al. (1957), investigating the tissue reactions to Dacron, Ivalon sponge, nylon, Qflon, and Teflon (P.T.F.E.) implanted int& the subcutaneous, tissue of the anterior abdominal wall in dogs, reached the same conclusions concerning P.T.F.E. Indeed, the photomicrographs of the tissue reaction in their paper bear striking resemblances to those found in our rats. The order in which they ranked materials by the tissue response, in decreasing magnitude, were nylon. Dacron, polyvinyl sponge, Orion, Teflon--a ranking which differ* little from that reported here. The paper by Hvrison et al. (1957) offers strong support thsrt the biological reaction to implants is similar in dog and rat. While comparative studies in man have not yet been published, individual reports (Teplitsky and Rubin, 1949; Rubin, 1949, 1951) would seem to imply that extrapolation of these findings to man are likely to Be accurate. 4. Implications for Cluaical Practice.--It is clear from what has already been written, both., here agjPin other publications, that implants of foreign material may initiate six (Efferent biological reactions :-- 1. Immediate inflammation, either chemical or bacterial, with immediate rejection of the implant. 2. Delayed rejection of the implant, of the order of weeks, months, or years, and probably due to an " uneasy " acceptance of the material by the host; round cell infiltration, giant cells, and unorganised vascular patterns are seen histologically many weeks after implantation. 3. Encystment without reaction--i.e., a fibrous tissue envelope. 4. Encystment, with an incomplete fibrous capsule where cellular reaction is continuing. 5. Slow absorption, where giant cells dominate the histological picture. 6. Incorporation or complete acceptance of the material. At the clinical level one has to define what is meant by an " inert " material. Oifmiral inertness is evidently not sufficient. Physical inertness would appear to be of equal importance, for we have shown that the biological reaction to smooth, solid implants differs materially from that to porous, woven, or spongy implants. It is hoped to publish more detailed experimental evidence concerning the influence of physical form in the near future. On present evidence it would seem reasonable to recommend that materials 20554020 BFG03887 THE USE OF INERT PLASTIC MATERIAL IN RECONSTRUCTIVE SURGERY 21 other than P.T.F.E. should be abandoned and that, where possible, smooth surfaced implants are preferable to those of open structure. 5. The Risk of Neoplasia.--Neoplasms, and invariably sarcomas, have been noted arising from the fibrous capsule surrounding implants in rats (Oppenheimer et a/., 1948, 1955 ; Druckrey and Schmahl, 1952 ; Laskin et al., 1954), but the great variety of materials which may cause them should make one suspect that this may largely be a species reaction (Scales, 1958, 1961). Northdurft (1956) suggests that neoplasia is related to the physical form of the implant and not to its chemical nature, and he believes that malignant change occurs in the cells of the avascular capsule. Leighton et al. (1956) support this and suggest " anaerobiosis could have contributed to the formation of a morphologically malignant variant." It is not generally known that carcinoma can occur spontaneously in rats. In the execution of the present and other experiments, two well differentiated carcinomas of the colon were found during laparotomy of 100 rats. One had received an implant of P.T.F.E. sheet thirty days earlier, the other not. It would seem unlikely that P.T.F.E. in this instance had contributed to the carcinoma of the bowel. Neoplasms do not appear to develop in less than one year after implantation (Turner, 1942; Oppenheimer et al., 1948) and hence none in our series was expected. In man, other than in association with very obvious chronic irritation, malignant changes have not been reported. The risk that it could occur would seem to be slight although Scales (1958) suggests that one may have to wait thirty to forty years before reaching a firm decision. The story of the treatment of thyrotoxicosis by irradiation and the subsequent development of skin carcinomas would support Scales' guarded prognosis. CONCLUSIONS 1. Arguments for and against the use of inert foreign materials in recon structive surgery have been considered. It is unrealistic to believe that their use should, or can be, abandoned entirely. 2. A technique for the biological testing of materials in rats is presented and particular features discussed. With a laboratory test of some precision available, the " clinical trial " of new materials in man without prior investigation in animals is indefensible and outmoded. 3. The macroscopic and microscopic tissue reactions to materials commonly used in reconstructive surgery (nylon, Polythene, Orion, Terylene, P.V.C., Estane, and P.T.F.E.) have been examined by this laboratory technique and the findings assessed statistically. 4. Tissue reaction to P.T.F.E. is less than with any other material and for this reason it is recommended that the use of all less satisfactory materials should be discontinued. 5. The influence on the biological reaction of the physical form of an implant has been demonstrated and this will be amplified in a further publication. 6. Although consideration of the biological tissue reaction to implant materials in man is beyond the scope of the present paper, certain resemblances indicate that experimental results may be extrapolated with fair accuracy. It is hoped to publish at a later date experimental and clinical support for this. f `-|i 1 20554021 22 BRITISH JOURNAL OF PLASTIC SURGERY 7. Incorporation of an implant as opposed to encapsulation is a new phenomenon which would appear to be the ideal. It is a pleasure to acknowledge the technical assistance and help provided by Miss L. Greenaway, who was responsible for the histological sections and the care of the animals; by Mr W. H. Brackenburyfor the photomicrography; by Mr G. Williamsfor the photographs of Fig. 2 ; and by Mr D. Banks, Medical Artist, for Fig. 1, all of the Department of Medical Illustration, Postgraduate Medical School. The sheet P.T.F.E. was kindly supplied gratis by the Crane Packing Company. The expenses of this experiment were defrayed by a generous grant from the Medical Research Council, and this financial assistance is gratefully acknowledged. REFERENCES Aldunate, E. (1951). Plast. reconstr. Surg., 7, 147. Barr, J. S. (1953). J. Bonejt Surg., 34A, 249. Blaine, G. (1946). Lancet, 2, 52$. Brown, J. B., Fryer, M. P., and Lu, M. (1954). Arch. Surg., 68, 744. 'v Brown, J. B., Ohlwiler, D. A., and Fryer, M. P. (i960). Ann. Surg., 152, 534. Brown, M. H., Grindlay, J. H., and Craig, W. K. (1948). Surg. Gynec. Obstet., 86,663. - Calnan, J. (1961). Nature, tax, 1017. ^ Dunphy, J. E. (i960). Ann. R. Coll. Surg. Engl., 26, 69. Edwards, W. S., and Lyons, C. (1958). Surg. Gynec. Obstet., 107, 62. Farmer, A. W. (1947). Plast. reconstr. Surg., a, 207. Frantz, V. K. (1943). Arm. Surg., 118, 116. Gibson, T. V., and Davies, W. B. (1953). Brit. J. plast. Surg., 6, 4. ------ (1955)- Brit. J. plast. Surg., 8, 85. Gillis, L. (1958). Modem Trends in Surgical Materials." London: Butterworth. Gilmer, W. S., Tooms, R. E., and Salvatore, J. E. (1961). Surg. Gynec. Obstet., 1x3, 143GORDON, S. D. (1958). Amer. J. Surg., 95, 263. Grindlay, J. H., and Waugh, J. M. (1951). Arch. Surg., 63,288. Harrison, J. H., Swanson, D. S., and Lincoln, A. F. (1937)- Arch. Surg., 74, 139. Jeremiah, B. S. (1950). Plast. reconstr. Surg., 5, 232. Johnson, H. A., and Grindlay, J. H. (1934)- Plast. reconstr. Surg+ 14, 293. Lasjun, D. M., Robertson, I. B., and Weinmann, J. P. (1954X Preo.Soe. exp. Biol., N.y.,87,329. Leighton, J., Kline, 1., and Ork, H. C (1936). Science, 123, 512. LeVeen, H. H., and Barberio, J. R. (1949)- Ann. Surg., X29, 74. McDoUGALL, J. (1956). 3- Bone 3t Surg., 38B, 709. Moore, A. M., and Brown, J. B. (1932). Plast. reconstr. Surg., so, 453. Moran, P. (1934). Personal communication. Neuman, Z. (1957). Brit. 3 plast. Surg., 9, 195. . ., Northduret, H. (1956). Strahlentherapie, 100, 192. Oppenheimer, B. S., Oppenheimer, E. T., and Stout, A. P. (1932). Proc. Soc. exp. Biol., N.Y., 70, 366. Oppenheimer, B. S., Oppenheimer, E. T., Damshelfsxy, I., and Eirich, F. R. (1933). Cancer Res., 15, 333. Penn, J., and Epstein, E. (1953). Plast. reconstr. Surg., 12, 148. Rubin, L. R. (1949). Plast. reconstr. Surg., 4, 269. ------ (1931)* Plast. reconstr. Surg., 7, 131. Rubin, L. R., Robertson, G. W., and Shapiro, R. N. (1948). Plast. reconstr. Surg., 3, 586. Scales, T. (1938). " Biological and Mechanical Factors in Prosthetic Surgery "; in " Modem Trends in Surgical Materials." Ed. Leon Gillis. London: Butterworth. ------ (1961). Personal communication. Scales, J. T., Winter, G. D., and Shirley, H. T. (1961). Brit. med. 3., a, 478. Smdons, A. H. M., and MacArthur, A. M. (1952). Brit. j. Stag., 39, 542. Struthers, A. M. (1955). Plast. reconstr. Surg., 15, 274. Teplitsky, D., and Rubin, R. (1949). Plast. reconstr. Surg., 4, 274. Turner, F. C. (1942). 3 "at. Cancer Inst., 2, 81. Yeager, G. A., and Cowley, P. M. (1948). Ann. Surg., 128, 509. 2t)8640Z2 BFG03889