Document wrERB0meNjzx628a3ZZevBdOE

BIOMAT., MED. DEV., ART. ORG., 4(344), 235-261 (1976) #jMU- QUANTITATIVE CELL CULTURE BIOCOMPATIBILITY TESTING OF MEDICAL DEVICES AND CORRELATION TO ANIMAL TESTS R. E. Wilsnack, D.V.M. Huntingdon Research Center Division of Becton, Dickinson and Company Post Office Box 527 Brooklandville, Maryland 21022 ABSTRACT The biocompatibillty of a wide variety of blomaterials was quantitatively assessed, in a physiologically normal environment, as to cytotoxicity induced in WI-38 cells by cell culture medium extracts. Materials tested included PVC plastic, rubber, sili cone rubber, polyethylene, polypropylene, acetal, polyurethane, Teflonr, nylon, epoxy, and polystyrene. Cell culture test results were correlated to U.S.P. animal tests. Potential test artifacts, lead, barium, cadmium, and endotoxin were tested for cytotoxicity in WI-38 cells.' r^Cell culture methods yielded more positive tests, particu larly rubber, PVC plastic and silicone rubber compounds, than observed in U.S.P. animal tests'^ Positivity in animal tests did not correlate quantitatively to cytotoxic titers in cell culture. Discrepancies between cell culture tests and animal tests, specifically rubber compounds, were attributable, in some instances, to differentials In elution efficiency between saline, cottonseed oil, and complete MEM cell culture medium. In other instances, particularly PVC plastics, differences between cell culture and animal test results were due to an inherent differ ence in the two Indicator systems to respond to specific toxic moieties. I. INTRODUCTION An array of in vivo and in vitro biocompatibillty tests (1-25), for the appraisal of medical devices, has been evolved with the presumptive goal of predicting and preventing adverse 235 Copyright 1977 by Marcel Dekker, Inc. All Rights Reserved. Neither this work nor any part may be reproduced or transmitted in any form or by any meafts, electronic or mechanical, including photocopying, microfilming, and recording, or by any information storage and retrieval system, without permission in writing from the publisher. Tooeeosg BFG60783 236 WILSNACK reactions during clinical application in humans. Previous test results have been difficult to correlate and interpret because of the nonquantitative endpoint parameters Inherent to many tests, the lack of statistically adequate numbers and generic variety of test specimens, wide biologic variations in indicator systems, and diverse physical/biochemical ambience employed in each test system. To this end, we have attempted to quantify the biocompati bility of specific formulations of biomaterial specimens in a human (WI-38) cell culture system (26) and have attempted to plot the correlation to U.S.P. animal tests (27). The cell cul ture test system was designed to encompass physiologically normal biochemical and physical environments in order that insight might be gained Into the short-term bioavailability of leachable toxic moieties which could be pertinent to clinical biomedical device usage. II. MATERIALS AND METHODS A. Cell Cultures WI-38 cells (28) were received from American Type Culture Collection and were propagated in 32 oz prescription bottles with growth medium consisting of Minimum Essential Medium (Eagle) with Earle's salts (MEM), 10% fetal bovine serum (FBS) (BBL, Division of BioQuest) (Table 1), 2 mM glutamine, penicil lin (100 units/ml), streptomycin (lOOmcg/ml), Fungizone (2.5 mcg/ml), and Gentamicin (5 mcg/ml) (the total mixture referred to as complete MEM medium). -All cell cultures were maintained in a humidified atmosphere (90% relative humidity) with 5-7% carbon dioxide at a temperature of 37C. All suspen sions were prepared with the aid of 0.25% trypsin solutions and were counted in a haemocytometer. Cells for cytotoxicity scoring were placed in 60 ran polystyrene cell culture dishes (Falcon Plastics) at 250,000 cells per dish and usually reached monolayer BFG60784 BIOCOMPATIBILITY OF BIOMATERIALS TABLE 1 Chemical Characterization of Fetal Bovine Serum Specific gravity Total protein A1bumin Globulin Blood urea nitrogen Total lipids Total bilirubin Hemoglobin 1.018 4.1 g% 2.0 g% 2.1 q% 14.0 mg% 109.0 mg% 0.4 mg% 12.2 mg% adata provided by BBL, Division of BioQuest 237 in 5-7 days. Alternatively, cells for cytotoxicity titrations were plated in Microtest II tissue culture plates (6 mm wells) (Falcon Plastics) at 20,000 cells per well and normally achieved monolayer in 5 days. WI-38 cells were not used beyond the 35th passage. The cell cultures were free of mycoplasma, as determined by aerobic and anaerobic culture (29), at the beginning and at the termination of the study. B. WI-38 Cell Culture MEM Elution Test Test materials were cut Into pieces not to exceed 20 mm in any dimension, or specimens made up of particles less than 20 ran in cross-sectional dimension were tested as received. None of the test specimens were sterile at the time of test nor were specimens aseptically handled. Materials were placed into (150 ml) borosilicate glass bottles containing MEM + 10% fetal bovine serum with antibiotics (as described under cell cultures) (hereafter referred to as complete MEM medium) at a ratio of 1 g test material to 5 ml of complete MEM medium. Bottles containing test specimens and complete MEM medium were incubated (stationary) at ^ Q W W i 238 WILSNACK 37C for 24 hours. After the 24-hour elution period, the mix tures were centrifuged at 500 x g and supernates were decanted for testing. Two modes of testing were employed: cytotoxicity scoring and cytotoxicity titration. To determine the cytotoxicity score, 5 ml of undiluted test eluate was placed into each of two 60 mm cell culture dishes containing monolayers of WI-38 cells from which complete MEM medium had been aspirated. The plates con taining test eluates were held at 37C in a humidified (90% relative humidity) atmosphere containing 5-7% carbon dioxide for 24 hours, after which time all fluids were decanted and cells were fixed in methanol and stained with Papanicolaou hema toxylin stain (Fisher Scientific Co.). Culture dishes were microscopically (100X) scored as to the degree of morphologi cally discernable cytotoxicity. Cytotoxicity was subjectively scored from 0 to 4, 0 indicating a negative test and 4 indicat ing virtually complete cell destruction. Cytotoxicity titers were determined by preparing two-fold dilutions (2-128) of test eluates in complete MEM medium with subsequent application of 0.25 ml of each dilution of test eluate to each of two wells in a Microtest II plate which con tained monolayers of WI-38 cells from which all medium had pre viously been aspirated. Microtest II plates, containing monolayers of WI-38 cells and two-fold dilutions of test eluates were Incubated, fixed, stained, and morphologically scored as described under procedure for cytotoxicity score. The cytotoxi city titer was defined as the reciprocal of the highest dilution of test eluate that induced morphologically discernable altera tion of WI-38 cells as determined by microscopic (100 X) obser vation of stained cell preparations. C. Systemic Injection Tests of Saline and Cottonseed Oil Eluates in Mice Methods used in tha performance of this test have been delineated in U.S.P. XIX (27). Test materials were eluted with BFG60785 BIOCOMPATIBILITV OF BIOMATERIALS sodium chloride injection, U.S.P. (Cutter Laboratories) (here after referred to as saline) and refined cottonseed oil (VWR Scientific Co.) Random-bred female mice, ranging in weight between 17 and 23 g, were used as indicators of toxicity. Saline eluates were injected intravenously at a dose of 50 ml/kg body weight. In a similar manner, cottonseed oil eluates were Injected intraperitoneally at a dose rate of 50 ml/kg body weight. All mice were observed for abnormal clinical signs and death, at 4 hours post inoculation and at 24-hour intervals thereafter for a total of 72 hours. D. Intracutaneous Tests of Saline, Complete MEM Medium, and Cottonseed Oil Eluates in Rabbits Test protocol.was based on U.S.P. XIX (27) and utilized female New Zealand white rabbits, weighing 2.5-3.5 kg. Test eluates of saline (50 or 70C), complete MEM medium (37C), and cottonseed oil (50 or 70C) were Injected (0.2 ml) intracutane- ously at ten linearly arranged sites, parallel to the spinal column, on each of two rabbits. The procedure was modified from that described in U.S.P. XIX in that 7 test eluates and 1 control were injected (10 injection sites per eluate) on each rabbit. Injection lines were demarcated by outlines prepared on the shaved skin with picric acid. Injection sites were examined at 24 , 48, and 72 hours postinoculation for evidence of erythema, edema, and necrosis and were scored according to the following numerical scale: 0 = no discernable response 1 = trace response 2 = slight response 3 = mild response 4 = moderate response E. Intramuscular Implantation Test in Rabbits Female New Zealand white rabbits, weighing 2.5-3.5 kg, were used as test subjects. Test procedure was basically as described ?0 tn o os CO o C3 U) 240 WILSNACK in U.S.P. XIX (27). Test materials and U.S.P. negative controls were prepared in strips 1 x 1 x 10 mm and Inserted into the para vertebral muscles, of each of 2 rabbits, through a 15 ga needle, with 4 test strips and 4 control strips per rabbit. After 3-7 days, rabbits were euthanized, muscle surrounding the sample was excised, and the tissues circumscribing the implants were macroscopically examined for evidence of hemorrhage, inflammation, and encapsulation necrosis. F. Preparation and Quantitation of Heavy Metal Solutions Solutions of BaCl (9 g/25 ml), 3 CdS04*8H20 (28 g/25 ml) and Pb (N03)2 (9g/25 ml) were prepared in complete MEM medium and were allowed to stabilize for 24 hours at 37C. Solutions were then adjusted to pH 7.2 with ION NaOH and precipitates were removed by centrifugation. Control media were prepared for each metal solution with a comparable volume of ION NaOH added with subsequent adjustment of all control solutions to pH 7.2 with ION HC1. Final concentration of metal ions, in each respective solution, was determined by atomic absorption spectroscopy on an atomic absorption spectrophotometer (Perkin-Elmer Model 303). Five-fold dilutions of each metal solution, as well as accompanying control solutions, were prepared in complete MEM medium. Each appropriate dilution (0.25 ml) was placed into each of 4 wells of a Microtest II plate containing a monolayer of WI-38 cells from which growth medium had been previously aspir ated. Plates containing monolayers of WI-38 cells and dilutions of metal solutions were held for 24 hours at 37C in a humidified atmosphere (90* relative humidity) containing 5-7* carbon dioxide. Plates were then fixed, stained, and scored for cytotoxicity as previously described. G. Preparation of Endotoxin Solutions Endotoxin (E. coli) was purchased (Mallinckrodt) In powder form, and two-fold dilutions were prepared in complete MEM BFG60786 BIOCOMPATIBH.ITY OF BIOMATERIALS 241 medium. Only those dilutions without visible precipitate v/ere tested for cytotoxic effect. Cytotoxicity titer was determined utilizing procedures described for heavy metals. H. Test Specimens Materials for test were selected at random from a total group submitted for various forms of biocompatlbillty testing. Test specimens included finished products and raw materials; however, the majority of the materials tested were unfabricated raw material. Raw materials were tested In a variety of config urations Including sheets, tubing, pellets, granules, blocks, cylinders, and screens. III. RESULTS A. Cytotoxicity of Barium. Lead, and Cadmium in WI-38 Cell Culture Cadmium Induced morphologic alteration of exposed cells at concentrations as low as 25 ppm, barium was cytotoxic at a higher level of 500 ppm, and cytotoxicity was not observed with lead concentrations up to 1000 ppm (Table 2). Higher concentrations of lead could not be maintained In solution in complete MEM medium without formation of a precipitate. Barium, lead, and cadmium cytotoxicity was quantified since these heavy metals have historically been present in eluates of some materials used in the fabrication of medical devices and could be a factor in the ingredient-specific Interpretation of cytotoxicity. B. Cytotoxicity of Endotoxin in WI-38 Cell Culture U1 o to Cytotoxicity was noted in those test preparations containing 167 pg/ml or more of endotoxin (.. coll) (Table 3). Endotoxin 242 WILSNACK TABLE 2 Detection Limits of Barium, Lead, and Cadmium in WI-38 Cell Cultures Cytotoxicity of Compounds Tested Concentration, ppma Bari urn Leadc Cadmium 1,000 500 100 50 25 10 4b 0 40 00 00 00 00 4 4 4 4 2 0 Compounds diluted in MEM + 10% FBS bCytotoxicity score cHigher concentrations (>1,000 ppm) could not be maintained in solution without precipitation cytotoxicity was titrated to provide a point of reference in test environments involving contamination of a biomaterial with Gramnegative bacteria. TABLE 3 $ Detection Limits of Endotoxin (E. coll) in WI-38 Cell Cultures Concentration of Endotoxin, ug/ml a Cytotoxicity Score 167 3 83 0 41 0 aEndotoxin diluted in MEM + 10% FBS BFG60787 RIOCOKPATIBILITY OF BIOMATERIALS 243 C. Overall Correlation of WI-38 Cell Culture to In Vivo Tests Cell culture-positive specimens were identified in all classes of test material except polypropylene and nylon (Table 4). Polyvinyl chloride (PVC) plastics, rubber, and silicone rubbers evoked the highest proportion of positive responses in cell culture tests. A total of 1013 specimens were tested yield ing one class of 115 materials positive in cell culture and one or more in vivo tests and a second reactivity class consisting of specimens positive in cell culture and negative in in vivo tests. Cytotoxicity titers, an approximation of the bioavailability of the toxic moiety, were clearly highest among the rubber com pounds. Six hundred thirteen of the specimens were negative by all tests. With the exception of one specimen, none of the test samples negative In the cell culture test were positive by an in vivo test (Table 4). This specimen was a Teflon polymer, and inadequate quantity was available to confirm a borderline reaction of a cottonseed oil eluate in the intracutaneous rabbit test. As a measure of the reproducibility of the cell culture test, 50 specimens were selected by random number table to be retested at the conclusion of the study. On the basis of cyto toxicity score, none of the retests deviated more than 2 units from the original score (Table 5). 0. Polyvinyl Chloride (PVC) Plastics One hundred fifty-two PVC plastic specimens, representing a wide range of formulations, were tested yielding 74 (495!) samples cell culture-positive. Sixty-two (84%) of these speci mens were negative in all animal tests (Table 6). Systemic Injection tests on cell culture-positive materials were negative with all eluates; intracutaneous tests were negative with saline eluates and positive in 9 of 69 specimens with cottonseed oil Surmary o f Specimen* Tested f o r C y to to x ic ity in WI-38 C e ll C ultures No. o f specimens negative In c e ll culture & untested In animals BFG60788 BIOCOMPATIBILITY OF BIOMATERIALS TABLE 5 Reproducibility of Cytotoxicity (MEM Eluate) in WI-38 Cell Cultures 245 Number of samples yielding same cytotoxicity score on repeat test Number of samples yielding 1 unit deviation in cyto toxicity score on repeat test Number of samples yielding 2 unit deviation in cyto toxicity score on repeat test Total Number of Samples Original Cytotoxicity Score T) 1 2 3 V 31 0 0 1 13 00Z00 0101 1 50 aSamples selected for repeat test by random number table eluates (Table 6). Intramuscular Implant tests were positive in 1 of 22 tests on cell culture-positive PVC plastics (Table 6). The proportion of cell culture-positive materials yielding posi tive results on an in vivo testing was not a function of the cytotoxic titer, i.e. the two PVC plastic specimens with the highest cytotoxic titer (1:8) were negative in vivo (Tables 6 and 7). None of the cell culture-negative specimens were posi tive in. vivo. Patterns and titers of cytotox(c1ty were formulationspecific; and differences In response, as measured by cell cul ture and in vivo tests, were attributable to the elution and/or the target phase of the cell culture test. The elution phase of the test was assessed by performing intracutaneous tests with complete MEM eluates and negative results were observed in most instances (Table 8). Clearly, the target phase of the cell to CJ1 o to ! 246 NO LU COc wn O OoN 00 VO m u 4-> in ro Q. CO O> r-- f Q. 03 4P out. 01 o CO r-5<+ro C i *-* in +j 2 O in <- 03 c E h- <u cu 0) 4-> 4- 5 in c > ro <D CO -- CO CL C W ME O *- . i- o 1. p +J (fl 3 - r-- *-3-3 03 U t- >, m UP 3 OE O U 03 r- S- XP >oc r- P M +J O iapi3 -*-> >yC F- CJ (O 4-> C 4) re u m 333 O' 4-3 o r-- 03 3C o re 4J 3 O re s4-3 C 4-3 c re Q ME reu CO lO CM i-- CM CO r-- 3 a OOo in 3 rEe j- p> c 03 4r-e> in 3 4- in m UJ p 03 -- o 3 CO m in C_) 03 3 os O 0) 03 re4-3 c 4-> in re oj 4-3 3 H- 3 U UJ re re s- 03 EC c F-- c --1 < re CO re* ON r>. 5^ r-- ^ O CM ON ON r*N. CO rNo-. "oS C*oSM 0) m 4-3 4-> re m3 C3 - Ul oc *" o CoO 4-3 u cu 0> rerc"9 4-> *--< 3 i-- u UJ E 03 03 4-> C F in >N re l/> CO -- i/> r^. to n r- o <M oOooo in on co VO CO r-- CO oooo in c U 03 a> E o EU 03 3 03 l_ Z Q. 3 V 4-> 3 43 OU 3 4- t_ i-- 03 03 3 4J o O *" I- 03 |o O -- CM 03 > g ur* 03 CL fc. o> WILSNACK Denominator - number specimens tested C orrelation Between WI-38 Cell Culture C y to to x ic ity T ite rs , Systemic In je c tio n , Intracutaneous, and Intramuscular Implant Tests fo r Selected PVC P la s tic Specimens BIOCOMPATIBILITY OP BIOMATERLALS 247 oooooooooo + 0 + 0+ + oooo ooooooooooo ooooooooooo 03 n o> o oooo o o 3 oj O 4-3 ^p-r-NNCVJ + ^+ COCO r0--1 LO.f Q..O 6<0 E3 co Z _ CO o ^ rree re re renjHCai*OhC^rei*OhcreCo*i Mhireni*Ohiroi- ON o m CO CO ti V- re* cw CO o ID CinO *T re* li v- rtieon ti re- K N K fs N r*- c*. N fs. N TV5 OVJI c o O u 1 o CO u '0 toq CO CO o o Ni II 248 WILSNACK TABLE 8 Formulation-Specific Comparison of PVC Plastic Cytotoxicity Titers in WI-38 Cell Culture with Intracutaneous Tests on CSO and MEM Eluates PVC Formulation AA a BB CC DD DD DD EE FF GG HH 11 JJ AA KK LL MM HH AA UN 00 Cell Culture Cytotoxicity Titer 1 8 1 2 1 1 1 1 1 2 2 2 8 4 8 0 2 1 1 1 Intracutaneous Rabbit Test CSO Eluate MEM Eluate 00 00 00 00 00 00 00 00 00 0 *1 00 0 +1 00 00 00 00 +4 0 +3 0 +3 +4 NT aSamples with same letter designations are multiple batches of the same formulation bNT - not tested CSO - cottonseed oil culture test, WI-38 cells, detects cytotoxic moieties eluted from PVC plastic formulations that were not demonstrable with the same eluate in vivo (Table 8). E. Rubber Rubber formulations proved the most toxic compound group, in terms of cytotoxic titer as well as percentage of specimens BFG60790 BIOCOMPATIBILITY OF BIOMATERIALS 249 positive, in cell culture as well as in vivo. One hundred ninety-five (65%) of 296 specimens tested proved cytotoxic in cell culture, and 127 (65% of the cell culture-positive samples were not reactive in vivo (Table 9). The systemic injection test yielded negative results with all saline eluates (162) and 130 of 132 cottonseed oil eluates from cell culture-positive specimens (Table 9). Similarly, all 98 saline eluates were negative in the intracutaneous tests corresponding to positive cell culture tests. Intracutaneous tests with cottonseed oil eluates proved the most sensitive of the in vivo tests as evidenced by 72 (38%) positive tests of 192 cell culture-positive specimens tested (Table 9). The proportion of specimens positive with the Intracutaneous test (cottonseed oil) eluate did not vary as a function of the cyto toxic titer in cell culture (Tables 9 and 10). Intramuscular implantation proved a relatively insensitive test with 5 of 29 tests positive, on materials cytotoxic in cell cultures, without quantitative relationship to cell culture titer (Table 11). None of the specimens negative in cell culture demonstrated reactivity In any In vivo test. In an attempt to elucidate the mechanism of action causing a discrepancy in the toxic reactivity of rubber formulations in vivo and in vitro, complete MEM eluates from cell culture-positive specimens were Injected Intracutaneously into rabbits. Twenty complete MEM eluates with cytotoxic titers 16 or greater, pre pared from specimens negative in the intracutaneous test by cottonseed oil elution, yielded 14 positive intracutaneous tests (Table 12). Complete MEM medium leached toxic moieties from rubber formulations that were biologically unavailable by saline or cottonseed oil extraction. In addition, the complete MEM extractions were performed at 37C as compared to 50C for the saline and cottonseed oil extractions. F. Polyethylene Although 36 of 108 polyethylene samples tested were cyto toxic in cell culture, none of the positive specimens repre- tn o to tt> o o QD 250 WILSNACK co m oo*~oo -- coo JJLaO3>. > O3C c o i.o u co CO I2C o r- 4-> esc ip ai to 40(3J1pO4<>O/lJr--QE-. Os LU CO eo i*/-i --&fo. r QJ 3 CO *+--iafJ,l *4rJ-- C(3e/>J UL. 4>-1 fLO. O **" 4-> U J c 44t-t-> -O=MS r- C_) 4cf30J qt3-j O' 4-> rO3- 4. Od A O On O VO O O) <t 9 ^ (O N p- C\J ^ r- ^ IO r- O a\ \n r-- 00 -- i-- n n cvj No No No *oS No orcorops- i-- CM r- CM OOO O CM > O <8 m Ol W M r-- CM-- ^-COU0CT><M OOOOOOOO tA ^0va3j Ca> *ae-- 3 0) Z CL IS} -o^oiinrsNfO O O r* i- ^ to r- fr-- u0) 3 4-> L> f- O r-- CM ^ 00 lO CJ Tt r-- CO LO o Numerator = number o f specimens p o s itiv e Denominator = number o f specimens tested C y to to x ic ity P ro file o f Rubber Formulations BIOCOMPATIBILITY OF BIOMATERIALS 251 4J C r- rn J3 QJ QJ JO s: C Ql O QJ Cl/Ll 3t/1 +|Q- <u o 3 Q1 Q1 r- CJ fCO LU 3 +-> O wV) u3(u0 hr-- <i0 4- -M C 10 CM r-- O O O r- CO CM O O O O 3QJ LtoO 4-> 0) 5QJ > QJ tA o E 0O9 *o I<u L. 3 O3 OQJ ?> <j >< ofj cc<c0 o > U c o U 4- QJ t0 JO #-- JO 3 MflWOp-Olfl uo co o CM -- CM VO ^ CM ^ CM CO r-- LO CO CSJ VO CO IIIII O CM r-- ^ O VO CO ^ CM CM VO CO (M CO CM CO r-- CO r- VO CM ^ 3<03UIxJQIl C3Ih<o^ js * o to o o4-> +- *" o CO QJ <o0 to <0 c co E o w CD iA tO JD C fO QJ E O <J a 4J <d QJ Q> tO Ec J= O *4i"-j tao. QJ to c o L. QJ +J II It ou N> cc <001 0JJ s: s: J JQ VI Oooo o s o 3a*'" "=a WILSNACK 252 < w o OoV --J VO K> C %A OC S(J *1i" <f-uy uQj *C- <Q/>. U r- 0) S J=i t<a>Mo>/>l iU.*) -CQ3C -*<aoa->>u m<L> ra--i ^o X 00 o a O2S tfZo a> a. UE rO3- *Lt3~O. r- u r-- </l OQJ 3EC 00 fc I cCO -*- caa2>> *coa * m in 01 3 CD O QJ O to M 3 tO O r-- tO Q> SL- -*- uc o *- o + 000 + 00 + 0 oooo + oo + ooo + ooooo oooooooo ooooooooo s a) f- N M 't CO lO ID lO w ^ r* r- r* fO iO <U 1. r- Q) 0.-0 EE t/(O> z3 N VO O' co c\i CO to in r iO co o\ CO to o> CO CM r~ o\ in r-- III t H >-- H* co CO CO >-- CO r- r>. r*. h-. BIOCOMPATIBILITY OF BIOMATERIALS TABLE 12 Formulation-Specific Comparison of Rubber Cytotoxicity Titers In WI-38 Cell Culture with Intracutaneous Tests on CSO and MEM Eluates 253 Rubber Formulation Ca F C C I F U A C F F F C C C C A K I I Cell Culture Cytotoxicity Titer 16 16 16 16 16 16 32 16 16 16 16 16 16 16 16 16 16 32 32 32 Intracutaneous Rabbit Test CSO Eluate MEM Eluate 00 00 0 +3 0 +2 0 +2 0 +2 0 +2 0 +3 0 +3 0 +2 00 00 0 +2 00 0 +2 0 +3 0 +2 0 +3 00 0 +3 aSamples with the same letter designations are multiple batches of the same formulation CSO - cottonseed oil sented virgin polymer. All of the toxic responses were attrib utable to pigments, adhesives, or other additives (Table 13). Saline eluates from polyethylene samples were negative in sys temic Injection tests with the exception of one specimen containing a formaldehyde residue (Table 13). Positive responses were evoked by cottonseed oil eluates in 1 of 108 systemic injec tion tests and 22 of 108 intracutaneous tests (Table 13). The degree of inflammatory or parenteral response observed in vivo CSO - cottonseed o il 0T08S0S2J 254' TABLE 13 Correlation Between M^ers WILSNACK Cell Culture Cell Culture TUer Polyethylene 0 1 2 4 8 15 32 less* 0 2 Acetal Q 1 2 leflon 9 1 Number Specimen* 72 17 9 4 1 1 A 16 1 19 2 \ 10 1 19 1 Animal Test Results Systemic Injection Saline Eluate CSO Eluate Intracutaneous Tests Saline Eluate CSO Eluate 0/72 0/17 0/9 0/4 0/1 0/1 1/4 0/16 0/1 0/19 0/Z 0/1 0/10 0/1 0/6 0/1 0/1 0/19 0/1 0/72 0/17 0/9 0/4 0/1 0/1 1/4 0/16 0/1 0/19 0/2 0/1 0/10 0/1 0/6 0/1 0/1 0/19 0/1 1/2 0/1 0/1 0/1 0/72 12/17 5/9 0/4 0/1 1/1 4/4 0/16 0/1 0/19 0/2 0/1 1/10 l/l 0/6 0/1 0/1 0/19 0/1 "Numerator Denominator > nunber number specimens specimens positive testeo CSO - cottonseed oil did not correlate to cytotoxic titer in cell culture, a phenom enon previously observed with PVC plastic and rubber formulations. No toxic in vivo tests were noted among the group of specimens negative in the cell culture test, BFG60793 BIOCOMPATIBILIHf OF BIOMATERIALS 255 G. Epoxies The epoxy compounds tested were biologically inert with the exception of a single specimen positive in cell culture and nega tive in all in vivo tests (Table 13). H. Acetals Twenty-two acetals specimens were tested and found to be devoid of toxicity in vivo as assessed by saline and cottonseed oil extraction (Table 13). Three of the 22 specimens evoked low level cytotoxicity in cell culture, and all of the positive speci mens were samples of separate lots of the same material formula tion. I. Teflons Consistent with past reports of biologic inertness (17), 2 of 11 Teflon specimens elicited toxic responses in cell culture or animal tests (Table 13). A single specimen was cytotoxic In cell culture and reactive in the Intracutaneous test when assessed by cottonseed oil extraction. The second biologically active speci men exhibited a toxicity pattern aberrant to the entire study; a cottonseed oil extract evoked an inflammatory response upon intracutaneous injection; and the cell culture test was negative. This specimen was not representative of the virgin polymer and contained pigmented additives. J. Polyurethane Two of eight polyurethane specimens, uncategorized as to polyether or polyester type, were cytotoxic In cell culture; and one of the two was recorded positive by virtue of cottonseed oil extraction and intracutaneous injection (Table 13). All systemic injection tests were negative. None of the specimens negative in cell culture were positive in vivo. XT08OS2 | 1 BFGf 94 256 WILSNACK K. Polystyrene Bio-Incompatibility was noted In a single specimen out of a test group of 20 (Table 13). The single positive response was reported in cell culture, and all in vivo tests were devoid of a toxic response. L. Miscellaneous and Formulations of Unknown Composition This specimen category Included those formulations for which an Inadequate number of lots were tested for meaningful analysis and a more diverse polymer group including those specimens sub mitted without definitive Identification as to formulation. Sixty-two of 144 miscellaneous and unidentified specimens were cytotoxic in cell culture, some attaining cytotoxic titer values of 32 (Table 4). Nine of the 62 cell culture positive specimens were reactive in In vivo tests, all by cottonseed oil extraction and Intracutaneous Inoculation. M. Silicone Rubbers Silicone rubbers, undelineated as to heat cure or room temperature cure, proved a diverse specimen group In terms of biocompatlbllity. Twenty out of 24 silicone rubber specimens, representative of more than one basic material manufacturer, were cytotoxic In cell culture; and 9 of this cytotoxic group were toxic in vivo (Table 14). Intramuscular Implantation was the most definitive of the in vivo tests for silicone rubber, detect ing toxicity in 6 specimens negative by cottonseed oil extraction and Intracutaneous Injection (Table 14). This relationship between toxic response detected by intramuscular Implantation and intracutaneous tests is unique to silicone rubber and has not been observed with any other class of polymers. The severity of the Intracutaneous test response, as measured by the cottonseed oil extraction, was not directly proportional to the cytotoxic titers as evidenced by the fact that the 5 specimens evoking the highest cell culture titer were negative by the intra cutaneous test with cottonseed oil eluates (Table 14). BIOCOMPATIBILITY OF BIOMATERIALS 257 i TABLE 14 Quantitative Correlation Between WI-38 Cell Culture Cytotoxicity Titers, Intracutaneous, , and Intramuscular Implant Tests for Silicone Rubber Cell Culture Cell Culture Number Titer Specimens Animal Test Results Intracutaneous Test Intramuscular CSO Eluate Implant 04 21 45 85 16 3 0/4 a 0/1 0/5 2/5 1/3 0/2 0/1 4/5 3/4 2/3 Numerator = number of specimens positive Denominator = number of specimens tested CSO - cottonseed oil Analysis of test results for the silicone rubber group was impeded by the inability to identify specific formulations. It is reasonable to assume that silicone rubber formulations are not monolithic in terms of biocompatlbllity, but exhibit formulationspecific toxicity profiles as previously demonstrated for rubber and PVC plastic. IV. DISCUSSION Cell culture methods have been utilized and reported for over a decade, in a wide variety of modes, to assess the bioincompati bility of medical devices and materials. Data presented in this report is intended to extend previous reports and to accommodate those variables most critical to the extrapolation of in vitro data to the clinical environment. Conceptual elements of the study include: (A) use of a normal human cell, unmodulated by virus (30) I or mycoplasma contamination (31), as a target system since the I products tested are intended for human use; (B) use of large number 25036012 258 WILSNACK of specimens to measure biologic compatibility of a variety of formulations and multiple lots of specific formulations; (C) quanti tative assessment of cytotoxicity to allow correlation studies; (D) use of physiologically normal elution menstrua and temperatures to permit an evaluation of the bioavailability of toxic moieties; and (E) provide for the correlation of all in vitro data with accepted in vivo systems In U.S.P. animal tests. The WI-38 cell culture test employed in this study (26) de fined a larger proportion of toxic specimens than any of the animal tests. Cytotoxic profiles were unique to classes of materials; and in many instances, were quantitatively and qualitatively character istic of specific material formulations. If a small number of form ulations had been tested for each material class, an array of con trasting conclusions on test methods could have been tendered depending on the particular formulation tested. Rubber specimens yielded the highest percentage of cytotoxic samples (65%) in vitro and fewest discrepancies (43%) between in vitro and in vivo test results. Disparities in cell culture and animal test results were attributable, in some instances, to an elution differential; wherein complete MEM medium eluted toxic moieties at 37C not manifested by elution with cottonseed oil or saline at 50-60C. Toxic molecules in rubber were the most readily leachable as evidenced by the cytotoxic titer range. Cytotoxicity profiles, in vitro and in vivo, were formulation-specific and re producible from lot to lot. PVC plastics manifested a high percentage (45%) of cell cul ture-positive samples with a concomitantly low proportion (11%) of these specimens positive in animal tests. The differential in test results is circumstantially ascribable to the target phase (cell cultures or animals) of the test system since efforts to delineate differences in the elution phases were negative. Positivity in animal tests was not proportional to cytotoxic titers in cell cul ture; however, none of the materials negative in cell culture were positive in animals. $FG60195 BIOCOMPATIBILITY OF BIOMATERIALS 259 Polypropylenes, polyethylenes, most epoxies, nylons, most Teflons, and most polystyrenes were biologically inert, in vitro and in vivo, under the ascribed test parameters. Positive speci mens in the polyethylene group were artifactual in the sense that none were virgin polymer and toxicities were ascribable to addi tives. A Teflon sample proved the aberrant specimen of the entire study by virtue of a positive Inflammatory response (intracutaneous) to a cottonseed oil extract and a negative cell culture response. Overall, the WI-38 cell culture test proved a sensitive and quantitatively reproducible test. In comparison, animal tests were relatively insensitive, particularly systemic Injection tests (6 of 944 positive); whereas the intracutaneous test with cotton seed oil extraction proved the most definitive in vivo test. Intramuscular Implantation proved uniquely sensitive (among in vivo tests) for the bioassessment of silicone rubbers, detecting more positive specimens than the other animal tests. The biologi cal significance of the cell culture-positive tests, unconfirmed in animals, is dependent in part on the Intended use of the tested product. Realistically, these cell culture-positive and animal negative tests cannot be shunted as biologically false since they represent toxic moieties eluted by physiologically normal menstrua under physiologically normal environments, physical and biochemi cal, that destroy normal human cells during the 24-hour exposure. REFERENCES 1. J. Autian, Ann. N.V. Acad. Sc*., 146, 251 (1968). 2. J. H. Brewer and H. H. Bryant, J. Am. PhaAm. Aaa., 49, 652 (1960). -- 3. S. A. Rosenbluth, G. R. Weddington, W. L. Guess, and J. Autian, J. PhaAm. Sca... 54, 156 (1965). 4. W. L. Guess, S. A. Rosenbluth, B. Schmidt, and J. Autian, J. PhaAm. Sc*., 54, 1545 (1965). I 260 WILSNACK 5. P. 0. Vasington, Bull. PaAent. VAug An., 2T_, 7 (1967). 6. C. N. D. Cruickshank, C. Hooper, H. B. M. Lewis, and J. D. B. MacDougall, J. Clin. Path., _13. 42 (I960). 7. J. D. B. MacDougall, Natalle, 172, 124 (1953). 8. P. J. Vasington, H. D. Piersma, J. J. Corbett, and J. L. Bittle, J. Pham. Set., 56, 1276 (1967). 9. F. W. Gunz, Ba. J. CanceA, 2, 29 (1948). 10. D. M. Conning, and J. Firth, Food & Comet. Toxicol., 7_, 461 (1969). 11. D. Powell, W. H. Lawrence, J. Turner, and J. Autian, J. Eiomed. Mat. Rea., 4, 583 (1970). 12. L. H. Eklund, C. A. Grant, and V. Winners, Aeta Phanma. Suectca, 2, 349 (1965). 13. T. Karisanova, Aeta Biol. Med. Gem., 21_, 79 (1968). 14. C. W. Smith, S. H. Wender, and C. A. Nau, Am. Jnd. Hyg. Aaa. J., 30, 402 (1969). 15. A. Edwards, J. Marish, J. D. Gupta, and J. D. Hartley, Exp. Eye Rea., 10, 288 (1970). 16. J. Autian, Can. J. Pham. Set., 3, 23 (1965). 17. W. H. Lawrence, J. E. Turner, and J. Autian, J. Btomed. Mat. Pei., 3, 291 (1969). 18. L. Spanberg, Odentol. Revy., 20, 133 (1969). 19. L. Spanberg, Odentol. Revy., 20, 123 (1969). 20. B. D. Ratner, T. Horbett, and A. S. Hoffman, J. Biomed. Mat. Rea., 9, 407 (1975). 21. P. Grasso, J. Gaydon, and R. J. Hendy, Food 6 Comet. Toxtcot., 1!_, 255 (1973). 22. M. Kasuya-, N. Sugawara, and A. Okada, Ball. EnvtAon. Contam. S Toxicol., 12, 535 (1974). 23. M. Kasuya, Bull. Envhon. Contam. 6 Toxtcol., 12., 167 (1974). BFG60796 BIOCOMPATIBILITY OF BIOMATERIALS 261 24. C. A. Homsy, K. 0. Ansevin, W. 0'Bannon, S. A. Thompson, R. Hodge, and M. A. Estrella, J. Macnomol. Scl-Chem, A4(3), 615 25. A. F. Hegyeli, J. Btomed. Mat. Rea., 7, 205 (1973). 26. R. E. Wilsnack, F. J. Meyer, and J. G. Smith, Btomat. Med. Vev., S Ant. (kg., 1(3), 543 (1973). 27. The United Statei Phanmacopela. 19th Revision; Mack Publlshing Co.; Easton, Pennsylvania; p. 644 (1975). 28. L. Hayflick, Exp. Cell Rea., 25, 585 (1961). 29. R. A. DelGuidice, N. F. Roblllard, and T. R. Carski, J. Bact., 93, 1205 (1967). 30. J. C. Led ere, J. P. Levy, B. Varet, S. Oppenheim, and A. Senik, CanceA Res., 30, 2073 (1970). 31. K. K. Sethi and M. Teschner, Klin. WochenschA., 50, 226, (1972). 25038014