The Discovery of a New Failure Mechanism Environmental Stress Cracking

On May 15, 1981, we received a Model 6991U arterial lead that had been removed from a human after only 5 months of implantation. The lead had about a 1in. gap in the insulation at the base of the "J," similar to that shown in Fig. 4. Thorough analysis produced no evidence of chemical degradation of the device.970 For example, we found no changes in the surface or bulk infrared spectra. There was no change in the molecular weight of the sample. However, optical microscopic analysis revealed an interesting occurrence. The edges of the breach, even though separated by about an inch, matched perfectly (see Fig. 4). In fact, it was clear that no material was missing and that the polymer had somehow pulled apart. This led us to suspect that the insulation had failed because of some kind of stress-cracking mechanism.

The insulation failure was completely unanticipated. We had seen no such cracking in 12-weeks canine implants, in 2-year rat implants, or during premarket clinical studies. Stress-cracking mechanisms were not among the known possible degradation mechanisms for polyether polyurethane elastomers. No insulation failures of the first neurologic leads were known, even after 6 years in service. A thorough review of the literature by both us and independent sources found no explanation for this. Thus, we had discovered a previous unknown failure mechanism.

Fig. 4. An illustration of the first polyurethane insulated atrial "J" lead marketed in 1980 is shown on the left. An example of environmental stress cracking (ESC) failure at the base of the "J" is shown on the right at about X30. Note that the edges of the breach match closely. The insulation has cracked and pulled apart as a result of ESC in the presence of unusual tension. No material is missing.

Fig. 4. An illustration of the first polyurethane insulated atrial "J" lead marketed in 1980 is shown on the left. An example of environmental stress cracking (ESC) failure at the base of the "J" is shown on the right at about X30. Note that the edges of the breach match closely. The insulation has cracked and pulled apart as a result of ESC in the presence of unusual tension. No material is missing.

Stress cracking is defined as cracking or crazing of a material in the presence of stress (strain) and a chemical environment.77 Many forms of stress cracking are known in rigid plastics. For example, polyethylene will crack when bent and exposed to a detergent. Polycarbonate can stress crack in ethylene oxide because of the residual molded-in stresses. Oxidative stress cracking is known for many rigid plastics but, with the exception of natural rubber in the presence of ozone, was not known for elastomers. Indeed, it did not appear to be possible in vivo based on current knowledge. Although we conducted tests on strained polyurethane and many different chemical agents, we could not duplicate the mechanism in vitro. We could not address any chemical component of the mechanism, except that it required exposure to the in vivo mammalian environment. Therefore, the mechanism was labeled environmental stress cracking (ESC). We did discover that during manufacture the insulation of the Model 6991U lead was occasionally and inadvertently stretched at the approximate point where the failure had occurred in the returned lead. Although we could not control the environmental portion of the mechanism, we could control residual strain in the manufacturing process. Manufacturing techniques were changed to assure that no residual stresses remained in the device as shipped.72 These changes appeared to be completely effective. Some patches of shallow cracks were still found in the tissue-exposed surfaces of some explanted and returned leads, although cracks through the insulation causing

Fig. 5. An electron microscopic view of environmental stress cracking at a tight ligature at about X500. The cracks become deeper and wider closer to the ligature (to the right), and decrease in depth and width away from the source of stress (to the left).

clinical failure no longer occurred in the Model 6991U manufactured after the change date. It appeared that the problem had been identified and corrected.

Later in 1981, a few explanted bipolar ventricular leads (Model 6972) were returned with cracks in the insulation around the fixating ligature (see Fig. 5). Until this time, anchoring sleeves had always been supplied separately in the lead packages. The instruction manual indicated that the sleeve should be placed on the lead before ligation to prevent damage to the device; however, it was common clinical practice to ignore the anchoring sleeve and simply ligate the lead directly in the vein. Now that we had identified that the polyurethane insulation was susceptible to a form of stress cracking, it became apparent that ligating the lead directly was no longer acceptable. Therefore, in February 1982, the factory began placing an anchoring sleeve on each lead. The sleeves could not be ignored, because they would have to be cut off to be removed. The instruction manual was also changed to state that the use of the anchoring sleeve was mandatory. Based on the analysis of returned products, only a small fraction of a percentage of the leads that were sold had failed by this or any other mechanism, whereas the next best silicone rubber bipolar ventricular lead had a 5% reoperation rate (see Table 1). We were satisfied that the aforementioned changes had solved a problem that had affected very few leads with otherwise superior clinical performance.73

0 0

Post a comment