W ilf r ie d B iirz le Department of Mechanical and Process Engineering, ETH Zurich, Zurich 8092, Switzerland; Institute for Mechanical Systems, Tannenstrasse 3, CLA H 23.2, Zurich 8092, Switzerland e-mail: [email protected]
E d o a rd o M a z z a 1 Department of Mechanical and Process Engineering, ETH Zurich, Zurich 8092, Switzerland Institute for Mechanical Systems, Leonhardstrasse 21, LEE N 210, Zurich 8092, Switzerland e-mail: [email protected]
J o h n J . M o o re Division of Neonatology, Case Western Reserve University School of Medicine, 2500 MetroHealth Drive, Cleveland, OH 44109 e-mail: [email protected]
About Puncture Testing Applied for M echanical Characterization of Fetal M em branes Puncture testing has been applied in several studies for the mechanical characterization of human fetal membrane (FM) tissue, and significant knowledge has been gained from these investigations. When comparing results of mechanical testing (puncture, inflation, and uniaxial tension), we have observed discrepancies in the rupture sequence of FM tissue and significant differences in the deformation behavior. This study was undertaken to clarify these discrepancies. Puncture experiments on FM samples were performed to reproduce previous findings, and numerical simulations were carried out to rationalize particular aspects of membrane failure. The results demonstrate that both rupture sequence and resistance to deformation depend on the samples’ fixation. Soft fixation leads to slippage in the clamping, which reduces mechanical loading of the amnion layer and results in chorion rupturing first. Conversely, the stiffer, stronger, and less extensible amnion layer fails first if tight fixation is used. The results provide a novel insight into the interpretation of ex vivo testing as well as in vivo membrane rupture. [DOI: 10.1115/1.4028446]
Introduction The FM is the thin and compliant membrane surrounding the developing fetus inside the pregnant uterus. Rupture of the mem brane is part of normal term delivery but has serious consequences when it happens prior to term. The spontaneous preterm prema ture rupture of the membrane (PPROM) is associated with 30%^t0% of all preterm deliveries and related to a higher mortal ity and morbidity of the newborn [1,2]. Understanding the defor mation behavior of FM tissue as well as its rupture characteristics is a prerequisite to identify factors leading to PPROM, to develop methods to prevent preterm rupture, and eventually to repair the lesions at the fetoscopic entry site after minimally invasive fetal surgery. The composition of the two constitutive layers amnion and cho rion and their sublayers has been extensively analyzed and described in literature, e.g., Refs. [3] and [4]. The thin amnion layer, which is mainly composed of collagen types I, III, IV, V, and VI, is attached to the thicker and more cellular chorion layer. The connection between amnion and chorion is referred to as the intermediate or the “spongy” layer, which is usually attributed to amnion and originates from the fusion of amnion and chorion at around 17-20 weeks of gestation [4]. Data characterizing the mechanical behavior of FM were mainly obtained by the use of three types of mechanical testing: uniaxial tension, inflation or burst, and puncture testing. Tensile testing is the most commonly used method to test engineering materials but does not allow characterization of FM tissue in a physiologically relevant biaxial stress state. Puncture testing is characterized by simple sample handling and preparation and it allows many samples to be tested from one membrane. Puncture testing originates from the textile industry and became more pop ular for the characterization of FM tissue following the work of 'Corresponding author. Manuscript received April 10, 2014; final manuscript received August 8, 2014; accepted manuscript posted August 29, 2014; published online September 17, 2014. Assoc. Editor: David Corr.
Journal of Biomechanical Engineering
Ref. [5], A series of important studies reporting results from punc ture tests on FM are summarized in the review by Moore et al. [6]. Using puncture testing El Khwad et al. [7] provided evidence of mechanical changes in the zone of altered morphology (ZAM), which overlays the cervix and is assumed to be the natural FM rupture site. Arikat et al. [8] studied the biophysical properties of individual components (amnion and choriodecidua) and compared them with intact FM. The authors observed that the separation of amnion from choriodecidua, which happens prior to rupture, con stitutes a significant component of the work required to rupture the sample. Further puncture studies analyzed membrane strength of sam ples exposed to bacterial protease, the effect of labor and preterm delivery, as well as the relationship between puncture force and gestational age [9-12], One specific question raised in previous investigations con cerned the rupture sequence of FM components (amnion or cho rion to rapture first). Analyzing the rupture sequence is not only important for better understanding the failure mechanisms of FM in vitro and physiological and pathological in vivo rupture behav ior, but also to study factors affecting the toughness of the mem brane. It seems that the answer might depend on specific details of the experimental configuration used for testing the membranes. Helming et al. [13] (uniaxial) and Schober et al. [5] (puncture) found that amnion ruptures first, whereas Artal et al. [14] (uniax ial), Lavery and Miller [15] (inflation), and Oyen et al. [10] (punc ture) state that chorion raptures first. The work by Arikat et al. [8] provided evidence of the rupture sequence of FM by video docu mentation of puncture tests, showing that chorion ruptures first. This is in contrast to observations from recent inflation and uniaxial tension experiments on intact FM as well as on separate amnion and chorion layers [16-18], With the aim of rationalizing the different findings, we have performed mechanical analysis based on a constitutive model recently demonstrated to well represent the mechanical response of FM tissue in a wide range of loading conditions [18], The model was calibrated on separate layer tensile test data under
Copyright © 2014 by ASME
NOVEMBER 2014, Vol. 136 / 111009-1
consideration of their unique in-plane contraction behavior, implemented as user material in a finite element code, and vali dated with data from inflation tests on separate layers and intact FM [19]. Numerical simulations of puncture experiments using this model led to results which consistently overestimated the puncture force for a given plunger displacement in contrast to experimental results reported in literature. This study aims to rationalize these discrepancies in terms of deformation behavior (more compliant response in puncture tests) and rupture sequence (chorion ruptures first in puncture, amnion in inflation). We focus here on puncture and inflation experiments, since these test configurations provide physiologically relevant biaxial loading conditions. Puncture tests are performed with the objective of reproducing previous findings, and numerical simula tions of these tests are carried out to investigate particular aspects related to clamping conditions and rupture sequence. The insights gained from the present study enable clarification of the rupture sequence of FM tissue in different experimental configurations and allow a new interpretation of existing data relevant to under standing in vivo rupture behavior of FMs.
Materials and Methods Puncture Experiments. In order to perform puncture tests on FM samples, a dedicated setup was built reproducing the model used in the studies summarized in Ref. [6]. The setup consists mainly of a vertical tension test machine (Stentor II, Andilog Industries, Vitrolles, France) with a vertical travel distance of 200 mm (7.87 in.). Attached to the traverse is a force gauge with 50N (11.2lbf) capacity, see Fig. 1(a). The tension test machine has a position accuracy of 0.01 mm (0.39 mil) and the force sensor a resolution of 5 mN (0.5 gf). The device is controlled by the soft ware califort (Andilog Industries, Vitrolles, France) which allows sequential programming of the test protocol. The clamping was designed to allow fixation of the FM sample, representative of the one used in Refs. [7], [8], and [20]. Clamping
dimensions were read out of the images (Fig. 1 in Ref. [8]). In detail, the inner diameter of the clamping is 25 mm (0.98 in.) and a 05 mm (0.19 in.) sealing ring of inner diameter 40 mm (1.57 in.) is inserted in the lower part of the clamping, see Fig. 1(b). A spherical steel plunger with diameter 10 mm (0.39 in.) has been used. In addition, a digital camera (Microsoft, LifeCam Cinema, 720p) is placed below the clamping to monitor the rupture sequence. The light conditions and the membrane surface struc ture were not suitable to determine the local state of deformation of the membrane from camera images. The tests are performed with a constant plunger speed of 25mm/min (0.98 in./min). The experimental data consist of the force and displacement data and the recorded movie from the lower view of the sample. Sample Collection and Preparation. FM samples were col lected from patients with single child pregnancies who underwent elective Cesarean sections between 37 and 40 weeks of gestation. Patients were recruited for this study with informed written con sent according to the protocol approved by the Ethical Committee of the District of Zurich (study Stv22/2006). The patients were randomly selected for this study after negative testing for HIV, hepatitis B, and streptococcus B, as well as chlamydia and cytomegaly. The selected pregnancies had no history of diabetes, connective tissue disorders, or chromosomal abnormalities. The membranes were cut approximately 2 cm (0.78 in.) away from the placental border and stored in saline solution until mechanical testing which was carried out within few hours after delivery. It should be noted that the FM samples included parts of the mater nal decidua. Thus, the main cellular layer might be referred to as choriodecidua. Note also that the investigated samples may have included tissue originating from the ZAM. In order to create samples for mechanical testing, the FM is spread with the chorion side downward on a lubricated mat. A piece of plastic-coated paper with the dimensions of approxi mately 50 x 50 mm (1.97 x 1.97 in.) is placed on the amnion side and the sample cut along the borders of the paper. The paper is used to support the membrane during transfer into the clamping. The membrane sample is placed on the fixture with the chorion side downward and is clamped by the cover plate. Saline solution is frequently sprayed on the sample to avoid dehydration and to reduce possible friction effects. A total of 26 samples originating from five different membranes underwent mechanical testing according to the protocol described before. Analysis of Test Data. A force threshold of 25 mN (2.5 gf) was used for the definition of the reference configuration, which is characterized by an initial apex displacement Uo, see Fig. 2(a). Thus, a deflected reference configuration was considered for fur ther analysis. Characterization of the mechanical response of FM tissue in puncture testing is done by the evaluation of maximum values of puncture force and plunger displacement similar to Refs. [7] and [8]. Here, the maximum value of membrane tension in the apex region is determined in order to characterize the membrane’s rupture strength. Membrane tension is evaluated by considering axial equilibrium of forces inside the contact region (Laplace’s law, see Fig. 2(b))
2nRsin2(Pc )
Fig. 1 Illustration of the puncture test setup (a) and sam ple fix ation ( b). The sam ple fixation was designed to be sim ilar to the one used in Refs. [7] and [8].
111009-2 / Vol. 136, NOVEMBER 2014
which provides a relationship between the puncture force F and the membrane tension T. The only remaining unknown, the con tact angle pc , can be estimated by the use of a simple analytical model (see Appendix) so that pc can be expressed solely as a function of the plunger displacement u, radius of the plunger R, and half diameter of the clamping a. The second derivative of the force-displacement curve provides information about stiffness reduction, for example, from slippage Transactions of the ASME
Plunger Amnion ylJpper clamping io
/F
o tr
X ? /^ F ric tio n
u,
©
' Chorion
®
025 mm
Lower clamping
045 mm
-------- N
Fig. 3 Illustration of the finite elem ent model used for the two num erical sim ulations perform ed, i.e., case 1 sim ulates tight sam ple fixation at the outer boundary, and case 2 allows for sam ple sliding within the clam ping assum ing a friction coefficient p
b)
Fig. 2 Schem atic drawing of the deflected m em brane sample with indication of geom etrical quantities (a) as well as illustra tion of the axial equilibrium of forces (b ). T he reference configu ration obtained by the use of a force threshold is characterized by the initial apex displacem ent U0.
in the clamping or weakening of the material prior to failure. In fact, monotonic stiffening during elongation is expected for an intact material as a consequence of geometry and material nonlinearity. This implies that the first derivative of the forceelongation curve is monotonically increasing. An indication of deviation from ideal conditions is therefore provided when the second derivative at a certain time t* becomes negative d2F{t*)
E d o a rd o M a z z a 1 Department of Mechanical and Process Engineering, ETH Zurich, Zurich 8092, Switzerland Institute for Mechanical Systems, Leonhardstrasse 21, LEE N 210, Zurich 8092, Switzerland e-mail: [email protected]
J o h n J . M o o re Division of Neonatology, Case Western Reserve University School of Medicine, 2500 MetroHealth Drive, Cleveland, OH 44109 e-mail: [email protected]
About Puncture Testing Applied for M echanical Characterization of Fetal M em branes Puncture testing has been applied in several studies for the mechanical characterization of human fetal membrane (FM) tissue, and significant knowledge has been gained from these investigations. When comparing results of mechanical testing (puncture, inflation, and uniaxial tension), we have observed discrepancies in the rupture sequence of FM tissue and significant differences in the deformation behavior. This study was undertaken to clarify these discrepancies. Puncture experiments on FM samples were performed to reproduce previous findings, and numerical simulations were carried out to rationalize particular aspects of membrane failure. The results demonstrate that both rupture sequence and resistance to deformation depend on the samples’ fixation. Soft fixation leads to slippage in the clamping, which reduces mechanical loading of the amnion layer and results in chorion rupturing first. Conversely, the stiffer, stronger, and less extensible amnion layer fails first if tight fixation is used. The results provide a novel insight into the interpretation of ex vivo testing as well as in vivo membrane rupture. [DOI: 10.1115/1.4028446]
Introduction The FM is the thin and compliant membrane surrounding the developing fetus inside the pregnant uterus. Rupture of the mem brane is part of normal term delivery but has serious consequences when it happens prior to term. The spontaneous preterm prema ture rupture of the membrane (PPROM) is associated with 30%^t0% of all preterm deliveries and related to a higher mortal ity and morbidity of the newborn [1,2]. Understanding the defor mation behavior of FM tissue as well as its rupture characteristics is a prerequisite to identify factors leading to PPROM, to develop methods to prevent preterm rupture, and eventually to repair the lesions at the fetoscopic entry site after minimally invasive fetal surgery. The composition of the two constitutive layers amnion and cho rion and their sublayers has been extensively analyzed and described in literature, e.g., Refs. [3] and [4]. The thin amnion layer, which is mainly composed of collagen types I, III, IV, V, and VI, is attached to the thicker and more cellular chorion layer. The connection between amnion and chorion is referred to as the intermediate or the “spongy” layer, which is usually attributed to amnion and originates from the fusion of amnion and chorion at around 17-20 weeks of gestation [4]. Data characterizing the mechanical behavior of FM were mainly obtained by the use of three types of mechanical testing: uniaxial tension, inflation or burst, and puncture testing. Tensile testing is the most commonly used method to test engineering materials but does not allow characterization of FM tissue in a physiologically relevant biaxial stress state. Puncture testing is characterized by simple sample handling and preparation and it allows many samples to be tested from one membrane. Puncture testing originates from the textile industry and became more pop ular for the characterization of FM tissue following the work of 'Corresponding author. Manuscript received April 10, 2014; final manuscript received August 8, 2014; accepted manuscript posted August 29, 2014; published online September 17, 2014. Assoc. Editor: David Corr.
Journal of Biomechanical Engineering
Ref. [5], A series of important studies reporting results from punc ture tests on FM are summarized in the review by Moore et al. [6]. Using puncture testing El Khwad et al. [7] provided evidence of mechanical changes in the zone of altered morphology (ZAM), which overlays the cervix and is assumed to be the natural FM rupture site. Arikat et al. [8] studied the biophysical properties of individual components (amnion and choriodecidua) and compared them with intact FM. The authors observed that the separation of amnion from choriodecidua, which happens prior to rupture, con stitutes a significant component of the work required to rupture the sample. Further puncture studies analyzed membrane strength of sam ples exposed to bacterial protease, the effect of labor and preterm delivery, as well as the relationship between puncture force and gestational age [9-12], One specific question raised in previous investigations con cerned the rupture sequence of FM components (amnion or cho rion to rapture first). Analyzing the rupture sequence is not only important for better understanding the failure mechanisms of FM in vitro and physiological and pathological in vivo rupture behav ior, but also to study factors affecting the toughness of the mem brane. It seems that the answer might depend on specific details of the experimental configuration used for testing the membranes. Helming et al. [13] (uniaxial) and Schober et al. [5] (puncture) found that amnion ruptures first, whereas Artal et al. [14] (uniax ial), Lavery and Miller [15] (inflation), and Oyen et al. [10] (punc ture) state that chorion raptures first. The work by Arikat et al. [8] provided evidence of the rupture sequence of FM by video docu mentation of puncture tests, showing that chorion ruptures first. This is in contrast to observations from recent inflation and uniaxial tension experiments on intact FM as well as on separate amnion and chorion layers [16-18], With the aim of rationalizing the different findings, we have performed mechanical analysis based on a constitutive model recently demonstrated to well represent the mechanical response of FM tissue in a wide range of loading conditions [18], The model was calibrated on separate layer tensile test data under
Copyright © 2014 by ASME
NOVEMBER 2014, Vol. 136 / 111009-1
consideration of their unique in-plane contraction behavior, implemented as user material in a finite element code, and vali dated with data from inflation tests on separate layers and intact FM [19]. Numerical simulations of puncture experiments using this model led to results which consistently overestimated the puncture force for a given plunger displacement in contrast to experimental results reported in literature. This study aims to rationalize these discrepancies in terms of deformation behavior (more compliant response in puncture tests) and rupture sequence (chorion ruptures first in puncture, amnion in inflation). We focus here on puncture and inflation experiments, since these test configurations provide physiologically relevant biaxial loading conditions. Puncture tests are performed with the objective of reproducing previous findings, and numerical simula tions of these tests are carried out to investigate particular aspects related to clamping conditions and rupture sequence. The insights gained from the present study enable clarification of the rupture sequence of FM tissue in different experimental configurations and allow a new interpretation of existing data relevant to under standing in vivo rupture behavior of FMs.
Materials and Methods Puncture Experiments. In order to perform puncture tests on FM samples, a dedicated setup was built reproducing the model used in the studies summarized in Ref. [6]. The setup consists mainly of a vertical tension test machine (Stentor II, Andilog Industries, Vitrolles, France) with a vertical travel distance of 200 mm (7.87 in.). Attached to the traverse is a force gauge with 50N (11.2lbf) capacity, see Fig. 1(a). The tension test machine has a position accuracy of 0.01 mm (0.39 mil) and the force sensor a resolution of 5 mN (0.5 gf). The device is controlled by the soft ware califort (Andilog Industries, Vitrolles, France) which allows sequential programming of the test protocol. The clamping was designed to allow fixation of the FM sample, representative of the one used in Refs. [7], [8], and [20]. Clamping
dimensions were read out of the images (Fig. 1 in Ref. [8]). In detail, the inner diameter of the clamping is 25 mm (0.98 in.) and a 05 mm (0.19 in.) sealing ring of inner diameter 40 mm (1.57 in.) is inserted in the lower part of the clamping, see Fig. 1(b). A spherical steel plunger with diameter 10 mm (0.39 in.) has been used. In addition, a digital camera (Microsoft, LifeCam Cinema, 720p) is placed below the clamping to monitor the rupture sequence. The light conditions and the membrane surface struc ture were not suitable to determine the local state of deformation of the membrane from camera images. The tests are performed with a constant plunger speed of 25mm/min (0.98 in./min). The experimental data consist of the force and displacement data and the recorded movie from the lower view of the sample. Sample Collection and Preparation. FM samples were col lected from patients with single child pregnancies who underwent elective Cesarean sections between 37 and 40 weeks of gestation. Patients were recruited for this study with informed written con sent according to the protocol approved by the Ethical Committee of the District of Zurich (study Stv22/2006). The patients were randomly selected for this study after negative testing for HIV, hepatitis B, and streptococcus B, as well as chlamydia and cytomegaly. The selected pregnancies had no history of diabetes, connective tissue disorders, or chromosomal abnormalities. The membranes were cut approximately 2 cm (0.78 in.) away from the placental border and stored in saline solution until mechanical testing which was carried out within few hours after delivery. It should be noted that the FM samples included parts of the mater nal decidua. Thus, the main cellular layer might be referred to as choriodecidua. Note also that the investigated samples may have included tissue originating from the ZAM. In order to create samples for mechanical testing, the FM is spread with the chorion side downward on a lubricated mat. A piece of plastic-coated paper with the dimensions of approxi mately 50 x 50 mm (1.97 x 1.97 in.) is placed on the amnion side and the sample cut along the borders of the paper. The paper is used to support the membrane during transfer into the clamping. The membrane sample is placed on the fixture with the chorion side downward and is clamped by the cover plate. Saline solution is frequently sprayed on the sample to avoid dehydration and to reduce possible friction effects. A total of 26 samples originating from five different membranes underwent mechanical testing according to the protocol described before. Analysis of Test Data. A force threshold of 25 mN (2.5 gf) was used for the definition of the reference configuration, which is characterized by an initial apex displacement Uo, see Fig. 2(a). Thus, a deflected reference configuration was considered for fur ther analysis. Characterization of the mechanical response of FM tissue in puncture testing is done by the evaluation of maximum values of puncture force and plunger displacement similar to Refs. [7] and [8]. Here, the maximum value of membrane tension in the apex region is determined in order to characterize the membrane’s rupture strength. Membrane tension is evaluated by considering axial equilibrium of forces inside the contact region (Laplace’s law, see Fig. 2(b))
2nRsin2(Pc )
Fig. 1 Illustration of the puncture test setup (a) and sam ple fix ation ( b). The sam ple fixation was designed to be sim ilar to the one used in Refs. [7] and [8].
111009-2 / Vol. 136, NOVEMBER 2014
which provides a relationship between the puncture force F and the membrane tension T. The only remaining unknown, the con tact angle pc , can be estimated by the use of a simple analytical model (see Appendix) so that pc can be expressed solely as a function of the plunger displacement u, radius of the plunger R, and half diameter of the clamping a. The second derivative of the force-displacement curve provides information about stiffness reduction, for example, from slippage Transactions of the ASME
Plunger Amnion ylJpper clamping io
/F
o tr
X ? /^ F ric tio n
u,
©
' Chorion
®
025 mm
Lower clamping
045 mm
-------- N
Fig. 3 Illustration of the finite elem ent model used for the two num erical sim ulations perform ed, i.e., case 1 sim ulates tight sam ple fixation at the outer boundary, and case 2 allows for sam ple sliding within the clam ping assum ing a friction coefficient p
b)
Fig. 2 Schem atic drawing of the deflected m em brane sample with indication of geom etrical quantities (a) as well as illustra tion of the axial equilibrium of forces (b ). T he reference configu ration obtained by the use of a force threshold is characterized by the initial apex displacem ent U0.
in the clamping or weakening of the material prior to failure. In fact, monotonic stiffening during elongation is expected for an intact material as a consequence of geometry and material nonlinearity. This implies that the first derivative of the forceelongation curve is monotonically increasing. An indication of deviation from ideal conditions is therefore provided when the second derivative at a certain time t* becomes negative d2F{t*)
