European Journal of Obstetrics & Gynecology and Reproductive Biology, 34 (1990) 247-255 Elsevier 247 EUROBS 00894 Biomechanical analysis of human chorioamniotic membranes H. Oxlund ‘, R. Helmig 2, J.T. Halaburt 2 and N. Uldbjerg *
`
`’ Department of Connective Tissue Biology, Institute of Anatomy, University of Arhus and ’ Research Laboratory, Department of Gynecology and Obstetrics, A’rhus Kommunehospital, t&hus C, Denmark Accepted for publication 20 June 1989
`
`Summary
`
`The biomechanical properties of human fetal membranes were analyzed by means of a materials testing machine. Special attention was paid to the biomechani- cal properties of the intact chorioamniotic membrane and the amniotic and chorionic components, separately, and thickness and storage of membrane samples. The load-strain and stress-strain relationships, and parameters calculated from the curves: extensibility, strength, elastic stiffness and failure energy, express the visco-elastic behavior of these membranes. The mechanical properties of the chorioamniotic membranes are determined by the interaction between the amniotic and chorionic components of the membrane. The strength of the intact chorioamniotic membrane, however, is primarily determined by the amniotic com- ponent, because the amniotic component is much less extensible. Thus, the chorionic component only contributes 10-U% of the strength when the amniotic component breaks. The chorionic component is twice as extensible as the amniotic component. Samples of fetal membranes can be stored at - 70
`’ C, with no significant changes in the biomechanical properties. No significant differences were found between speci- mens, which were oriented in parallel with and at right angles to the placental edge. Small samples can be analysed and the localization of samples in relation to the placental edge and rupture site of the membranes can be standardized. The method is well suited for studies of premature rupture of fetal membranes. Fetal membrane; Chorioamniotic membrane; Amnion; Chorion; Biomechanics Correspondence: H. Oxlund, Department of Connective Tissue Biology, Institute of Anatomy, University of Arhus, 8000 Arhus C, Denmark. 0028-2243/90/$03.50 0 1990 Elsevier Science Publishers B.V. (Biomedical Division)
`
`MTF Ex. 1041, pg. 1
`
`

`

`248
`
`Introduction
`
`Materials and methods
`
`The mechanical strength of the fetal membranes is of obvious interest when discussing the aetiology of premature rupture of fetal membranes (PROM), which is a significant clinical problem contributing to perinatal morbidity and death [12]. Several investigators have studied the strength and/or elasticity of fetal membranes. Artal et al. [2] found a reduced modulus of elasticity of membranes from PROM pregnancies, whereas none of the investigators have demonstrated reduced strength of the fetal membranes in PROM pregnancies [15,17]. On the contrary, the very elaborate studies of Lavery and Miller [8] showed a greater ability of preterm membranes to tolerate stress application compared to term membranes. As pointed out by Lonky and Hayashi [9], however, the membranes investigated in these studies had been obtained after delivery, and they might have undergone proteolysis and mechanical deformation during labor. Polishuk et al. [16] included fetal membranes from patients undergoing Cesarean section in their studies. The tensile strength of these membranes did not differ from that of the term membranes, and the integrity and mechanical properties of fetal membranes obtained after delivery seem to be preserved, allowing subsequent studies of such membranes. The aim of the present study was to design a method for uniaxial analysis of the mechanical properties of fetal membranes with special attention to the thickness of the membranes, mechanical properties of the amniotic and chorionic components, separately, and storage of membrane samples.
`15.0 mm, were punched out through the membrane and the supporting filter paper, using a cutting instrument of razor blades in parallel. The membrane specimens were then mounted in a materials testing machine (Lorenzen Wettre, Stockholm, Sweden). Specimen thicknesses were determined by placing them without filter paper between glass plates and using a micrometer to measure the resulting ‘sandwiches’ with and without the specimens [8], the difference giving the membrane thickness. The preparation of separate specimens of amnion and chorion is a problem, because the amnion and chorion have to be separated by force and consequently stretched. Both membranes will crumple afterwards, and the dimensions of both membranes will end up far from the dimensions of the original chorioamniotic membrane, and the subsequent mechanical analysis will be of little value. Therefore, a special procedure was followed. The original dimensions of the chorioanmiotic sample was kept during the separation of the amniotic and chorionic components by sticking another piece of wet filter paper to the other side of the chorioamniotic sample. The amniotic and chorionic membranes were then carefully separated, each of them supported and protected against deformation by the filter papers. There-
`
`Samples of chorioamniotic membranes from 10 intact placentas with membranes delivered at term after spontaneous and uncomplicated labor were obtained within 20 min of delivery. The samples measuring 40
`40 mm were taken halfway between the placental edge and the rupture margin, rinsed in Ringer’s solution (pH 7.4 at 4’C) and taken up on a piece of wet filter paper. Strip specimens, 4.0
`
`X
`
`X
`
`MTF Ex. 1041, pg. 2
`
`

`

`249
`
`7 strain at maximum load, i.e., extensibility; eFbK&
`
`after, the specimens were punched out through the filter paper and mounted in the materials testing machine, supported by the filter paper as explained above for the chorioamniotic specimens. The mechanical analysis was performed as described elsewhere [13,14]. The membrane specimens supported by their strip of filter paper were mounted in the clamps of the materials testing machine with a jaw space of 7.0 mm. During the testing procedure the clamps with the specimen were immersed in Ringer’s solution (pH 7.4 at room temp.). The filter paper strip was cut through, and the distance between the clamps was increased, stretching the specimen at a constant deforma- tion rate of 10 mm per min until rupture. During this procedure, the tension of the strip (load values) and distance between the clamps (deformation values) were continuously recorded by transducers coupled to measuring bridges, and the signals were fed to an x-y recorder. The resulting load-deformation curves were read by a digitizer into a computer and transformed into load-strain and stress-strain curves. Strain values were obtained by expressing deformation values in units of original specimen length. Stress values were calculated from the load values, by normalizing load values to the cross-sectional area of each sample. From the load-strain curves (examples in Fig. 1) the following parameters were calculated: “FIX%
`’ F,,,,,; E o,break, strain at breaking stress = E Fbreak; u max, maximum stress value, i.e., tensile strength; tan (Y, slope of the stress-strain curve, i.e., elastic modulus; Es, area under the stress-strain curve, i.e., relative failure energy of the stress-strain curve. Samples from seven placentas were used for studies of fresh and stored speci- mens. From each placenta samples of chorioamniotic membrane, amniotic and chorionic membranes separately were prepared. From each of these membranes, ten specimens were prepared as described above. Five of these specimens were chosen at random, and the mechanical analysis was performed within an hour, i.e., fresh specimens. The remaining five specimens from the membrane were stored in air-tight plastic tubes at - 70” C for 2 days. Then they were thawed to room temperature, mounted in the materials testing machine and analysed as above. During the specimen handling and analysis, care was taken to keep the specimens soaked with Ringer’s solution (pH 7.4 at room temp.). Samples from three placentas were used for studies of specimens which were punched out when oriented in parallel with and at right angles to the placental edge. Comparisons between parameters were performed by the non-paired Wilcoxon two-sample test [19]. Differences were regarded significant if 2P < 0.05.
`
`9 strain at breaking load; F max, maximum load value i.e. strength; tan p, slope of the approximately linear part of the curve, i.e., stiffness; E,, area under the load-strain curve, i.e., relative failure energy. From the stress-strain curves the following parameters were calculated: E o,max, strain at maximum stress =
`
`MTF Ex. 1041, pg. 3
`
`

`

`Results
`
`A typical load-strain curve of the specimen prepared from a chorioamniotic membrane is given in Fig. la. The first part represents the intact chorioamniotic membrane and has a toe part which is convex towards the x-axis, followed by an approximately linear part leading to the point of maximum load, where the amniotic component breaks. After rupture of the amniotic component of the specimen, the rest of the curve represents the chorionic component, which typically possesses less strength and higher extensibility compared to the amniotic component. The maxi- mum load of the chorioamniotic specimen (F,,), extensibility at the maximum load (“F,,,) and ultimate extensibility of the chorioamniotic specimen (EFbreak) are marked on the curve. In Fig. la, the approximately linear part of the curve is marked by a broken line, and the slope of this line is the stiffness of the intact chorioamniotic specimen (tan p). The hatched area between the curve and the x-axis gives the relative failure energy of the chorioamniotic specimen (EL). All preparations of chorioamniotic membranes showed this two-component behavior of an amniotic component with lesser extensibility and a chorionic component with higher extensibility.
`
`CHORIOAMNIOTIC
`MEMBRANE
`
`b.
`
`LOAD
`(N)
`
`10.
`
`Fmax ____________
`
`I
`
`/
`I
`
`AMNIOTIC
`MEMBRANE
`
`E Fmax
`
`c Fbreak
`
`= lojLPND
`
`CHORIONIC
`MEMBRANE
`
`02
`
`0.6
`1 0.4
`FFmax=‘Fbreak
`
`W’”
`
`0.6
`
`05
`max_
`F
`
`_____________-
`
`------_I
`
`I’
`
`/’
`
`L n
`
`0’2
`
`04
`
`0.6
`
`I
`
`0.6
`
`S;&hIN
`
`Fig. 1. Samples of human fetal membranes were analysed by means of a materials testing machine. A typical load-strain curve is given for (a) intact chorioamniotic membrane; (b) the amniotic component of this membrane and (c) the chorionic component. The load-strain relationship of the intact chorioamniotic membrane is that of a two component material, composed of the lesser extensible and stronger amniotic component, and more extensible chorionic component. The following parameters were calculated from the curves: ‘F,,, = extensibility, i.e., strain at the maximum load value; ‘Fbrcak = strain at the breaking point; F,,,,, = maximum load value; tan j3 = stiffness; and E, = relative failure energy, i.e., area between the curve until the breaking point and the x-axis. N = newton.
`
`MTF Ex. 1041, pg. 4
`
`250
`$$$&
`

`

`TABLE I Biomechanical properties of fresh and stored membranes Number of patients (-) Thickness ‘F_ (am) (-) Load-strain parameters ;? F,,, TmP E, (N) (N) (N) Stress-strain parameters %lmax Talla (N/mm2) (N/mm2 ) Es (N/mm2 ) Fresh chorio-
`
`7
`
`0.35 (0.03) 0.32 (0.02) 0.30 (0.02) 0.59 (0.06) 0.49 (0.03) 0.83 (0.05) 0.75 (0.06) 0.32 (0.02) 0.31 (0.02) 0.60 (0.06) 0.49 (0.03) 0.95 (0.08) 1.01 (0.09) 0.72 (0.07) 0.70 (0.15) 0.45 (0.03) 0.50 (0.07) 4.4 14.5 (0.7) (1.1) 4.7 11.7 (0.5) (1.0) 3.5 8.6 (0.4) (0.8) (Z) 8.0 (1.9) (t;) 7.9 (0.5) 1.8 9.1 (0.3) (1.0) 1.03 (0.09) 1.06 (0.10) 3.94 (0.24) 3.93 (0.74) 0.62 (0.06) 0.65 (0.09) 4.1 (0.7) 5.1 (0.6) 19.0 (1.0) 21.0 (3.0) 2.2 (0.3) (G) 15.0 (0.9) 12.8 (1.1) 47.3 (3.0) 44.8 (10.1) 10.9 (0.8) 11.9 (1.2) Mean values with SEM in parentheses. N, Newton.
`
`243 amniotic specimens (13) Stored chorio- 7 231 amniotic specimens (6) Fresh amniotic 6 41 specimens (4) Stored amniotic 6 44 specimens (2) Fresh chorionic 6 185 specimens (15) Stored chorionic 6 192 specimens (9)
`
`(0.W
`0.40
`
`MTF Ex. 1041, pg. 5
`
`

`

`252
`
`Fig. lb gives the amniotic component alone. It has a toe part and an approxi- mately linear part leading to the breaking point. The maximum load value (F,,,,,) and ultimate extensibility (“F,,,,) are marked. For amniotic specimens as well as for chorionic specimens, the extensibility at the maximum load and extensibility at the breaking point are similar. In Fig. lc the load-strain relationship of the chorionic component alone is given, and the breaking load (F,,) and ultimate extensibility (“F,,) are marked. From these curves it appears that the mechanical properties (the extensibility, strength, stiffness and failure energy) of the chorioamniotic membrane are determined by the interaction of the amniotic and chorionic components. Load-strain parameters for analyses of fresh chorioamniotic membranes, amniotic and chorionic membranes separately, and membranes, which had been stored at - 70 o C, followed by quick thawing to room temperature, are given in Table I. No significant differences were found between fresh membranes and membranes which had been stored at - 70 o C. Furthermore, stress-strain parameters, i.e., load-strain parameters divided by the cross-sectional areas of the specimens and thus corrected for differences in the thickness and width, are given in Table I. No significant differences were found between the stress-strain parameters of fresh and stored specimens (level of significance: 2P c 0.05). Likewise, no significant differences were found between specimens which were oriented at right angles to each other. The coefficients of variation for the mechanical parameters including both variation of biological sampling and methodological variation were 20-27s. The present study delineates a standardized method for sampling, storage and mechanical analysis of fetal membranes. The load-strain and stress-strain relation- ships of the intact chorioamniotic membrane are presented and its characteristic two-component behavior is illustrated (Fig. 1). The method allows biomechanical analysis of the amnion and chorion separately. The load-strain and stress-strain curves express the viscoelastic behavior of these membranes. The parameters de- scribing the biomechanical properties are calculated from the curves: extensibility, stiffness, breaking strength and failure energy. The load-strain relationship gives the data for the membranes as such, and the stress-strain relationship the biomech- anical characteristics corrected for differences in the thickness and width of the membranes. The amniotic component is the stronger, but less extensible and therefore, it breaks first during the stretching procedure. The chorionic component is twice as extensible as the amniotic component and possesses around 60% of the strength of the amniotic component. When the load-strain parameters are corrected for differences in thickness and width of the membrane specimens, giving the tensile strength values (a,,,,), it appears that the amniotic component as a material is around 6-times as strong and 9-times as stiff as the chorionic component. The mechanical properties of the chorioamniotic membrane, therefore, are determined by the interaction between the amniotic and chorionic components of the mem- brane. The strength of the intact chorioamniotic membrane, however, is primarily determined by the amniotic component, because it is much less extensible. Thus, the chorionic component only contributes 10-15’S of the strength when the amniotic
`
`MTF Ex. 1041, pg. 6
`
`

`

`253
`
`component breaks. In Fig. lc, the isolated chorionic membranes possess less strength than the chorionic component of the chorioamniotic membranes in Fig. la. This might be caused by the preparation procedure of separating the amniotic membranes from the chorionic membranes, where fibers binding the amniotic and chorionic components together are torn. Furthermore, the study shows that samples of fetal membranes can be stored at -70” C, with no significant changes in the mechanical properties, which permits collection and storage of samples of fetal membranes. Concerning the orientation of specimens for the uniaxial analysis, no significant differences were found between specimens which were oriented differently to each other, permitting preparation of specimens of random orientation. The location of specimens at the placental edge, at the rupture margin or in between, has been shown by McLachlan [lo] to play a role. Samples close to the placental edge possess the highest bursting strength. Likewise, Halaburt et al. [6] demonstrated reduced thickness of both membranes towards the site of rupture, reduced collagen content and increased collagenolytic activity. Therefore, the location of the sampling should be standardized. Most studies of the strength of fetal membranes have been concerned with measurements of their bursting strength. A piece of the membranes has been clamped between O-rings in a chamber, and the pressure in the chamber was increased until the membranes burst [1,4,5,7,8,10,11,15,17,18]. The advantage of this method is that the membranes are strained approximately as they are in vivo during parturition. The values of bursting strength, however, depend on the construction and dimensions of each individual testing instrument, and comparisons between the investigations are therefore very difficult. The uniaxial method of biomechanical tests on strip specimens has been used by Artal et al. [2] and in the present study. It allows calculation of tensile strength, as well as important biomechanical parameters, such as extensibility, stiffness and failure energy. Small samples can be analysed, and thus, the number of observations can be increased with statistical advantages and localization of the samples stan- dardized. Furthermore, the small size of samples allows biomechanical analyses of fetal membranes from rabbits and rats. Corrections of the biomechanical parame- ters for membrane thickness or collagen content can easily be made, improving the analysis. Our data on the extensibility of the chorioamniotic membrane is consistent with the values found by Artal et al. [2]. The view held by Embrey [5] and Artal et al. [2] that the chorionic component of the chorioamniotic membrane is the lesser extensible and therefore ruptures first is not in agreement with out data. On the contrary, our data show that the chorion is around twice as extensible as the amnion which breaks first. The lesser extensibility and higher tensile strength of the amnion is consistent with histologic descriptions of the two membranes [3,11]. The amnion is rich in collagenous fibers and would be expected to possess relatively high tensile strength and low extensibility when compared to the chorion which is composed of loose connective tissue rich in capillaries and cells. The method is well suited for studies of premature rupture of fetal membranes (PROM). The aetiology of PROM is unclear. Some of the risk factors have been reviewed by Lonky and Hayashi [9]: chorioamniotic infection, proteases, protease inhibitors, maternal smoking, coitus etc. At present the studies of the strength of
`
`MTF Ex. 1041, pg. 7
`
`

`

`254
`
`fetal membranes obtained from pregnancies associated with PROM, compared with term membranes seem to give little information about the aetiology of PROM. The reason could be that no substantial differences exist between PROM and term membranes, because the fetal membranes in both cases have been subjected to proteolytic degradation. Alternative research techniques might be to obtain mem- branes from patients undergoing elective surgery [9] and in vitro analyses of the influence of chorioanmiotic infections, proteases, protease inhibitors etc. on the mechanical properties of fetal membranes. Appropriate animal experiments would enable studies of intact fetal membranes which had not been subjected to proteoly- sis and mechanical strain during delivery and answer the questions: how the process of delivery influences the mechanical properties of fetal membranes, and whether the fetal membranes are subjected to a proteolytic process, e.g., in parallel to the cervical ripening. Studies on PROM considering the questions above are in progress in our laboratories.
`
`Acknowledgements
`
`This investigation was supported by the Danish Medical Research Council and University of Arhus. The skilled technical assistance of Mr. N. Kappel and linguistic revision of Mrs. Aa. Young are gratefully acknowledged.
`
`References
`
`1 Al-Zaid NS, BowResli, Goldspink G. Bursting pressure and collagen content of fetal membranes and their relation to premature rupture of the membranes. Br J Obstet Gynecol 1980;87:227-229. 2 Artal R, Sokol RJ, Neuman M, Burstein AH, Stojkov J. The mechanical properties of prematurely and non-prematurely ruptured membranes. Am J Obstet Gynecol 1976;125:655-659. 3 Boume GL. The microscopic anatomy of the human amnion and chorion. Am J Obstet Gynecol 1960;79:1070-1086. 4 Danforth DN, McElin TW, States MN. Studies on fetal membranes. Am J Obstet Gynecol 1953;65:480-490. 5 Embrey MP. On the strength of the fetal membranes. J Obstet Gynecol 1954;61:793-796. 6 Halaburt JT, Uldbjerg N, Helmig R, Ohlsson K. The concentration of collagen and the collagenolytic activity in the amnion and the chorion. Eur J Obstet Gynecol Reprod Biol 1989;31:75-82. 7 Lavery JP, Miller CE. The viscoelastic nature of chorioamniotic membranes. Obstet Gynecol 1977;50:467-472. 8 Lavery JP, Miller CE. Deformation and creep in the human chorioamniotic sac. Am J Obstet Gynecol 1979;134:366-375. 9 Lonky NM, Hayashi RH. A proposed mechanism for premature rupture of membranes. Obstet Gynecol Surv 1988;43:22-28. 10 McLachlan TB. A method for the investigation of the strength of fetal membranes. Am J Obstet Gynecol 1965;91:309-313. 11 Meudt R, Meudt E. Rupture of the fetal membranes. Am J Obstet Gynecol 1967;99:562-568. 12 Naeye RL. Causes of perinatal mortality in the U.S. Collaborative Perinatal Project. J Am Med Assoc 1977;237:228-237. 13 Oxlund H, Andreassen TT. The roles of hyaluronic acid, collagen and elastin in the mechanical properties of connective tissues. J Anat 1980;131:611-620. 14 Oxlund H. Changes in connective tissues during corticotrophin and corticosteroid treatment. Dan Med Bull 1984;31:187-206.
`
`MTF Ex. 1041, pg. 8
`
`

`

`255 15 Parry-Jones E, Priya S. A study of the elasticity and tension of fetal membranes and of the relation of the area of the gestational sac to the area of the uterine cavity. Br J Obstet Gynecol 1976;83:205-212. 16 PoIishuk WZ, Kohane S, Peranio A. The physical properties of fetal membranes. Obstet Gynecol 1962;20:204210. 17 Polishuk WZ, Kohane S, Hadar A. Fetal weight and membrane tensile strength. Am J Obstet Gynecol 1964:88:247-250. 18 Sbarra AJ, Selvaraj RJ, Cetrulo CL, Feingold M, Newton E, Thomas GB. Infection and phagocytosis as possible mechanisms of rupture in premature rupture of the membranes. Am J Obstet Gynecol 1985;153:38-43. 19 SokaI RR, Rholf J. Biometry. San Francisco: WH Freeman and Company, 1981.
`
`MTF Ex. 1041, pg. 9
`
`