Polyurethane elastomers demonstrate high tensile strength, lubricity, good abrasion resistance, good biocompatibility, and ease of handling. These properties make polyurethanes the most commonly used elastomeric materials available for implant applications. These materials are the critical component in many vital medical devices such as compliant polyurethane vascular grafts, pacemaker leads, and ventricular assist devices. The Thoratec ventricular assist device (VAD) uses Thoralon, a thromboresistant and biocompatible material for the blood pumping sac and the cannulae which connect the VAD to the heart and circulatory system. 1 Thoralon is a proprietary polyetherurethaneurea (PEUU) blended with siloxane based surface modifying additive (SMA) during fabrication. After fabrication, siloxane migrates to the surface reducing surface free energy, platelet activation, and adhesion, thus providing thromboresistance. Thoralon is also used in the manufacture of Thoratec’s Vectra vascular access graft.
Because of failures encountered in pacemaker leads, there is prevailing concern over the use of polyether based polyurethanes in long-term medical devices. 2 There has been extensive research in understanding the mechanism of degradation of polyether urethanes. 3 A number of interrelated and diverse factors, such as polyurethane chemistry, device manufacturing processes, implant techniques, and complex biological environment of the implant site, play a role in degradation of polyurethanes. This study was undertaken to understand physicochemical changes that occur in Thoralon in the blood pumping chamber during medium and long-term duration VAD implants. Previous investigations of blood sacs that were made from other linear segmented polyether urethanes have been reported for shorter implant durations. Bedini et al. 4 investigated blood sacs made from Biomer that had been used clinically for 43 days and reported that there were no significant changes in physical properties of the polymer other than a slight reduction of percent elongation. They also reported the presence of biofilm on the surface but did not discover any pitting or cracking of the blood contacting surface. Biomer diaphragms explanted from Jarvik-7 total artificial heart after 112 days in clinical use did not show any calcification, pannus, or vegetative thrombus. 5 Wu et al. 6 investigated blood sacs made from Biolon that had been implanted in calves for up to 132 days. They reported serious deterioration of the flex regions of right ventricular assist device blood sacs after 17 weeks of use. Generally, there was an increase in molecular weight with increasing implantation time. SEM observations suggested significant deterioration of flex region. Hunter et al. 7 analyzed Biomer blood sacs from calves implanted up to 7 months. They found that there was a significant increase in molecular weight of the blood sacs.
The purpose of this study is to describe the evaluation of Thoralon VAD blood sacs that have been explanted from human subjects after up to 14 months of clinical use. The physical properties, bulk morphology, surface chemistry, and surface morphology were characterized for explanted and control (unimplanted) blood sacs.
Materials and Methods
Blood sacs from four paracorporeal LVADs removed from clinical patients were studied. The duration of LVAD pumping and the age and gender of the patients were as follows: 53 days (28/male; transplanted), 171 days (57/male; transplanted), 336 days (23/male; transplanted), and 434 days (13/male; reimplanted with another device to enable home discharge). Unimplanted blood sacs were tested as controls.
The clinical VADs were sterilized in ethylene oxide then rinsed, photographed, and disassembled, and the blood sacs were removed from the plastic VAD housing. Cannula sections were disconnected from the VAD ports. The samples were washed in isopropanol for 1 hour then immersed in 0.1N NaOH or 10% Tergazyme for 5–7 days, rinsed thoroughly in deionized water, and dried under vacuum. The 0.1 N sodium hydroxide wash has been shown to have no significant effect upon the properties of polyurethanes. 8
Physicochemical analysis was conducted on samples taken from three regions of the blood sacs based upon different stress levels expected during blood pumping actuation. In Figure 1, an explanted blood sac after 434 days of use is presented, and the regions are specified as described below. The apex of the blood sac experiences the most flexing, and it is classified as the high flex region. The circular region beneath the outlet and inlet port on the front of the blood sac is classified as the medium flex region. The inlet, outlet, and the back of the blood sac remain relatively unmoved during use, and they are classified as the low flex region.
Physical Property Testing.
A total of eight dog bones were cut (one from the high flex region, five from the medium flex region, and two from the low flex region) using a microtensile die (JISK 6251–7, ODC, Ontario, Canada) from each of the explants. A total of 14 unimplanted blood sacs served as controls. Physical property testing was done on an Instron 4464. Ultimate tensile strength (UTS), stress at 300% elongation, Young’s modulus, and elongation at break of each sample were acquired using a crosshead speed of 5.0 inches/min.
Molecular Weight: Gel Permeation Chromatography.
The molecular weight of the polymer samples was determined with a gel permeation chromatography (GPC) system containing Waters 600 pump, 717 Autosampler, and 410 RI detector in conjunction with a Wyatt miniDawn laser light scattering detector. Dimethylacetamide containing 0.05 M LiBr was used as the mobile phase. The explants were prepared for GPC by washing in 0.1N NaOH to remove any biological debris on the surface. The samples were then rinsed in reverse osmosed water and dried before being dissolved in dimethylacetamide. Eleven unimplanted blood sacs served as controls, and samples from them were dissolved and analyzed without subjecting them to washing protocol.
Bulk Calcium Analysis.
Bulk calcium analysis was performed at Galbraith Laboratories, Knoxville, TN. The samples were digested with a mixture of sulfuric acid, nitric acid and hydrochloric acid and the digest was analyzed by ICP-atomic emission method.
Surface analysis was performed using ATR-FTIR and ESCA to understand any surface changes that took place during implantation.
The three different regions as specified previously for the blood contacting side of the control and implanted samples were analyzed by ATR-FTIR spectra using an Inspect IR Plus microscope connected to a Nicolet 5ZDX FT-IR. Three spectra from each of the regions of each blood sac were taken using Ge and ZnSe crystals, which have reported depth of analysis of approximately 1.5 μ and 4 μ, respectively. Peak height analysis quantified peak heights using local baselines. Each spectrum was normalized to the aromatic ν(C = C) 1,591 cm−1 band, which is considered to be an internal reference that is reportedly unaffected by implantation. 9 Loss of ether peak was studied by comparing the intensities of 1,367 cm−1 (α-CH2 wagging) of the polyether soft segment and 1,591 cm−1 in both unimplanted control and explanted samples.
Scanning Electron Microscopy.
Scanning electron microscopy (SEM) was used to characterize the surface morphology of the control sacs and explants in the three different regions. SEM analysis was performed with a Hitachi S4700 instrument (Charles Evans Associates, Santa Clara, CA). The samples were sputter coated with Pd before analysis. Electron dispersive x-ray (EDX) analysis of blood contacting surface of long-term explants was also performed.
ESCA was obtained for the high flex regions of the blood sac. Analyses were performed on the blood contacting surface using Surface Science Instruments (SSI) S-Probe ESCA instrument (University of Washington, Surface Analysis Recharge Center, Seattle, WA). An aluminum Kα1,2 monochromatized x-ray source was used to stimulate photoemission. The spectra were collected with the analyzer at 55° with respect to the surface normal of the sample, allowing for a data collection depth of approximately 50 –80Å. An electron flood gun set at 4 eV was used to minimize surface charging of the samples. Typical pressures in the analysis chamber during spectral acquisition were 10−9 torr.
Statistical analysis was conducted using Sigma Stat 2.0. The physical property and molecular weight raw data were analyzed using one way and two way ANOVA. The data were considered statistically significant when the p < 0.05.
No apparent defects were visibly evident upon close inspection of the explanted blood sacs (Figure 1; an example of 434 day explanted blood sac).
Physical Property Testing
Tensile strength, Young’s modulus, stress at 300% elongation, and percent elongation at break of different flex regions and average of all the regions of the test specimens are presented in Table 1 along with the physical property data for the unimplanted control sacs. Tensile strength of all of the blood sacs analyzed showed approximately 20% increase at 53 days over the unimplanted control blood sacs but no further change for longer implant duration. Percent elongation showed a slight decrease in the 336 day and 434 day samples. The physical property data for the control blood sacs showed statistically significant difference within the flex regions. However, none of the explanted blood sacs showed a difference among different flex regions in all physical properties. Two way ANOVA showed that all properties are statistically significant when compared with the control blood sacs, with respect to time. With respect to flex region, there was statistically significant difference only in stress at 300%; other properties showed no statistically significant differences among different flex regions at all time points.
Molecular Weight Analysis
The absolute number average molecular weight (Mn) and weight average molecular weight (Mw) results of control blood sacs and different flex regions of all explanted blood sacs are presented in Table 2. There was a highly significant (p < 0.001) increase in molecular weight of all explanted bloods sacs compared with controls. Similar to the physical properties, there were no significant differences in the flex regions of the blood sacs, with one exception. Similar to the tensile strength measurements, the 53 day implant showed molecular weight of Mn = 59,000 and Mw of 114,500, significantly greater than the molecular weight of control blood sacs of Mn = 41,650 and Mw = 79,700. However, there was no further change in molecular weight for longer duration samples, and there was no change in the polydispersity (Mw/Mn) of the explanted samples compared with the control samples.
Surface chemical analysis of the bloods sacs using ATR-FTIR spectroscopy indicated that there were very little changes on blood contacting surfaces of the sacs using both Ge and ZnSe crystals. A representative overlay of spectra collected using Ge crystal from the control blood sac and high flex region of the 434 day explanted blood sac is provided in Figure 2, including a subtraction result. Subtraction results indicate no detectable decrease in 1,367 cm−1 peak indicating no degradation of Thoralon in this long-term application. Surface ether concentration measured at the blood contacting side of the blood sacs using Ge and ZnSe crystals showed no significant reduction up to 434 days of use. Almost all of the original ether content remains on the blood contacting surface except for the high flex region of the 336 day sac, where some minimal depletion of ether is noticed.
Scanning Electron Microscopy Analysis.
SEM analysis was conducted on the blood contacting side of the different flex regions of the blood sacs. The results of the analysis of the control and high flex region of 434 day explanted blood sac are presented (Figure 3, A–C). All of the explanted blood sacs exhibited somewhat irregular surfaces. However, there was no evidence of cracking or pitting on any of the blood sac surfaces noted at ×500. At ×2,000, there was some pitting observed on almost all of the blood sacs. The size of the pitting was in the order of 1–2μ. In the 336 days blood sacs, some deposit was noticed in a very small area on the lower posterior region (high flex region) of the sacs.
ESCA analysis was conducted on the low flex regions of the samples of the blood sacs. The surface elemental composition such as C, O, N, and Si and C1s2 concentration using high resolution Carbon spectra are presented in Table 3. A significant amount of silicon could be detected on the surface of the blood sacs, and there appeared to be slight decrease in Si and increase in N. The C1s2 peak did not show any change compared with control and no trend with respect to implant duration.
VADs as a bridge to transplantation are generally used for clinical applications lasting up to 6 months in most cases and much longer in others. However, because of the limited supply of donor hearts for transplant patients and new indications for destination therapy for nontransplant patients, there is an increasing trend toward use of these devices for longer durations. It is important that the various blood contacting, polymer based components of the VAD, specifically the blood pumping sac, display superior high flex endurance, high strength, and continued thromboresistant properties over the life of the intended application of the device.
The results of the study show that Thoralon exhibited no deterioration of physical properties during clinical use for up to 14 months. There was a slight but statistically significant increase in tensile strength and decrease in elongation at break, which suggests that the polymer upon implantation undergoes slight stiffening. The changes that occurred in the physical properties occurred in all flex regions of the blood sac with no significant differences between regions. There was also no trend with respect to implant duration after an initial change was seen at the shortest duration. In addition to slight increase in physical properties, there were significant increases in both number average molecular weight and weight average molecular weight. The molecular weight increase was observed after the shortest duration of implantation (53 days) with no further change up to the longest time point studied of 434 days of clinical use. Similar to physical properties, the molecular weight also did not exhibit any relationship with different flex regions. All of the regions uniformly exhibited increase in molecular weight. Other groups have also reported increase in molecular weight of polyetherureaurethanes during in vivo use. For example, Hunter et al. 7 reported an increase in the molecular weight of Biomer blood sacs after 7 months in calves. The increase in molecular weight was attributed to leaching of the low molecular weight species 10 or crosslinking reactions caused by residual isocyanate. 11 In our case, the later reason can be ruled out based upon the method of preparation of the material as well as the processing of the blood sacs. The formation of higher molecular weight species in our case may be attributed to reactions of unknown origin as discussed by Wu et al. 6 However, others have seen a reduction in molecular weight experienced in the high flex region of the Biolon blood sacs, 6 but Thoralon blood sacs did not show any reduction in the molecular weight of the high flex region.
The SEM images of the surface of the explanted blood sacs had a slightly irregular appearance, but there was no evidence of cracking observed on the surface of any of the blood sacs. In general, there were no major differences on the explanted blood sac surfaces and the controls. In some instances, the surface showed micropitting under high magnifications (×2,000). There was no evidence of cracking observed on the surface of any of the blood sacs.
Previously, ATR-FTIR spectroscopy has been used to observe oxidative degradation of polyurethane surfaces by monitoring the ether content of the polymer surface as a result of in vivo or in vitro environments. 11 Surface chemical analysis of the bloods sacs using ATR-FTIR spectroscopy indicated that there were few changes on blood contacting surfaces of the sacs. In some cases, there was a minor loss in surface ether (20%) content in the 336 day high flex region; this could have been an instance of surface oxidation. The peak occurs at 1,367 cm−1 (-CH 2 -O), α-carbon to oxygen and 1,110 cm−1C-O-C stretch in the polyether soft segment can be analyzed for changes. The 1,110 cm−1 C-O-C stretch in the soft segment usually shows the most changes after implantation, but these bands often have several other absorbencies overlapping in this region; therefore the band at 1,367 cm−1 (-CH 2-O) was used in this analysis. All absorbencies were normalized to 1,591 cm−1, which is the aromatic band that is believed to remain constant during implantation. 9 Reduction in the intensity of 1,367 cm−1 peak compared with 1,591 cm−1 peak is typically attributed to oxidation of polyether urethanes. This reduction in ether content was evident only when using the Ge crystal in ATR FT-IR and only on the region where the biologic deposition was present. It is possible that biologic deposition could have masked the ether content on the surface, making it appear as though there is depletion of ether content. As pointed out earlier, this particular region of the blood sac showed no decrease in molecular weight or physical properties, suggesting no significant degradation. The studies performed using ZnSe crystal did not show any decrease in ether content of the interior surface of the blood sacs regardless of the location and implant duration. McCarthy et al. 12 advocate the use of subtraction method to determine the extent of degradation. The subtraction method that was used also did not show any depletion of ether content on the explants. The more surface sensitive technique like ESCA did not show any significant changes upon implantation. The decrease in C1s2 peak is indicative of oxidation of ether soft segments 11 and showed no change at all over implant duration. The quantity of C1s2 peak was used to assess the oxidation of polyether soft segments by ESCA. 13 Except for the 171 day explant, all of the explants showed a peak at 288.4 eV because of urea carbonyl in the high resolution Carbon region at varying concentration. It is hypothesized at this point that it could be caused by some biologic debris that did not get removed from the surface completely upon washing or possible migration of hard segments to the surface of the blood sacs upon implantation.
Previous studies have shown that polyurethanes are prone to calcification in vivo such as in explanted trileaflet valves 14 and total artificial hearts. 15 A closer look at the two long-term explants, such as the 336 and 434 day explants, by electron dispersive x-ray analysis (EDX) as well as bulk calcium analysis indicated no sign of calcification. It is also worth mentioning that the 336 day explant and the 434 day explant are from very young patients, 23 and 13 years old respectively. These experiments indicate no calcification of Thoralon in its use as blood sac material over the time frame studied, even in young patients.
Based upon the physicochemical analysis of explanted VAD blood pumping sacs from human subjects, there were no major changes to the Thoralon material, with the exception of a slight increase in tensile and molecular weight after implantation, without further changes with respect to implant duration. No significant changes in its surface properties for durations of up to 434 days were observed. This study provides evidence in support of the biodurability of Thoralon for long-term clinical use in Thoratec VADs.
1. Farrar DJ, Litwak P, Lawson JH, et al: In vivo
evaluations of a new thromboresistant polyurethane for artificial heart blood pumps. J Thorac Cardiovasc Surg
95: 1–20, 1988.
2. Pinchuck L: A review of the biostability and carcinogenecity of polyurethanes in medicine and the new generation of biostable polyurethanes. J Biomater Sci Polym Ed
6: 225–267, 1994.
3. Anderson JM, Hiltner A, Zhao QH, et al
: Cell/Polymer Interactions in the Biodegradation of Polyurethanes, in Vert M, Feijen J, Albertson A, Scott G, Chiellini E, eds, Biodegradable Polymers and Plastics
Cambridge, England: Royal Society of Chemistry, 1992: pp. 122–136.
4. Bedini R, Chistolini P, De Angelis G, Formisano G, Caiazza S: VAD Biomer blood sacs: Mechanical tests and ultrastructural observations. Med Prog Technol
19: 83–88, 1993.
5. Coleman DL, Meuzelaar HLC, Kessler TR, McClennen WH, Richards JM, Gregonis DE:. Retrieval and analysis of clinical total artificial heart. J Biomed Mater Res
20: 417–431, 1986.
6. Wu L, Weisberg D, Runt J, Felder G, Snyder AJ, Rosenberg G: An investigation of the in vivo stability of poly (ether urethaneurea) blood sacs. J Biomed Mater Res
44: 371–380, 1999.
7. Hunter SK, Gregonis DE, Coleman DL, Andrade JD, Kessler T: Molecular weight characterization of pre- and post-implant artificial heart polyurethane materials. Trans Am Soc Artif Intern Org
28: 473–477, 1982.
8. Zhang Z, King MW, How TV, Laroche G, Guidoin R: Chemical and morphological analysis of explanted polyurethane vascular prostheses: The challenge of removing fixed adhering tissue. Biomaterials
17: 1843–1848, 1996.
9. Wu Y, Sellitti C, Anderson JM, Hiltner A, Lodoen G, Payet CR: An ATR-FTIR investigation of in vivo poly(etherurethane) degradation. J Appl Poly Sci
46: 201–211, 1992.
10. Xi T, Sato M, Nakamura A, Umemura T, Tsuda M, Korukawa Y: Degradation of polyetherurethane by subcutaneous implantation into rats: I. Molecular weight change and surface morphology. J Biomed Mater Res
28: 483–490, 1994.
11. Phua SK, Castillo E, Hiltner A, Anderson JM: Biodegradation of polyurethane in vitro. J Biomed Mater Res
21: 231–246, 1987.
12. McCarthy SJ, Meijs GF, Mitchell N, et al
: In vivo degradation of polyurethanes: Transmission–FTIR microscopic characterization of polyurethanes sectioned by cryomicrotomy. Biomaterials
18: 1387–1409, 1997.
13. Marchant RE, Zhao Q, Anderson JM, Hiltner A: Degradation of a poly(ether urethane urea) elastomer: Infra-red and XPS studies. Polymer
28: 2032–2039, 1987.
14. Hilbert SL, Ferrans VJ, Tomita Y, Eidbo EE, Jones M: Evaluation of explanted polyurethane trileaflet cardiac valve prostheses. J Thorac Cardiovasc Surg
94: 419–429, 1987.
15. Dostal M, Vasku J, Sotolova O, Vasku A, Dolezel S, Hartmannova B: Mineralization of polyurethane membranes in the total artificial heart (TAH): A retrospective study from long term animal experiments. Int J Artif Organs
13: 498–502, 1990.