Sterilisation considerations for implantable sensor systems
Cutting edge electronics and materials technologies in implantable sensor systems challenge available sterilisation technologies. Selecting and verifying a sterilisation method for patient safety and regulatory requirements cannot compromise sensor system performance. To make the selection, the engineer must have a comprehensive appreciation of how thermal, chemical, pressure and radiation energy effects and cycle times could affect the materials in the sensor system electronics, surface treatments and base materials of the implant. These constraints must all be accomplished within the regulatory boundaries for the device. The most important consideration is the choice of sterilisation technology being made in the context of per unit costs and system delivery.
Sterilisation methods commonly used in the medical device industry include electron-beam (E-beam) and gamma radiation, ethylene oxide gas (EO or EtO), and, less often, hydrogen peroxide gas plasma. Although steam sterilisation is very commonly used in the hospital setting, it is not as commonly used for terminal sterilisation by the medical device industry. A few less common forms of sterilisation with potential utility for sensitive implantable sensor systems, such as peracetic acid (also known as peroxyacetic acid), may be considered. Each sterilisation method has its own advantages and disadvantages for implantable sensor systems: For example, EO gas is a solvent for some plastics and adhesives; heat, moisture and vacuum can affect sensitive membranes; and radiation can alter the molecular bonds of plastics.
Sterilisation costs are a major factor in selecting the most appropriate system, but the effect of process methods of sterilisation on materials and the efforts involved in validation of the cycle all have a significant influence on selecting the cycle, particularly for a start-up manufacturer. As a general rule, a device under development may be qualified using a small-batch EO sterilisation cycle. Such small-batch sterilisation services are widely available and allow the developer to avoid the need to use dunnage (simulated product matching the density and configuration of the actual product) to fill a large chamber. Because of shipping time and costs of a large facility, a dedicated cycle can be prohibitive. Therefore, qualifying EO sterilisation cycles in a small-batch steriliser has become a popular process for new product development. However, when a device has extremely small lumens, tortuous pathways for gas evacuation or parts that are sensitive to the cycle conditions (heat, vacuum and moisture), radiation may be a better alternative. Typically, the higher the radiation dose the simpler it is to validate sterility, but the trade-off can be a greater risk of damage to materials and components (described further below). Gamma sterilisation lends itself to larger, bulkier loads while E-beam can be ideal for smaller implantable sensor systems. E-beam typically uses a ‘conveyor belt’ approach to pass product under the beam, so relatively thin packages are ideal for E-beam, in contrast to gamma radiation of bulkier products such as surgical drapes or large volume components. Radiation doses can be controlled more precisely with E-beam, which is another advantage to this process in working with high cost and radiation sensitive materials. Although E-beam has been adopted more slowly than EO sterilisation or gamma radiation, it is becoming more widely available in the global medical device arena. E-beam offers an attractive alternative to EO, particularly for intricate devices with difficult to aerate passages or designs sensitive to vacuum or humidity.
For implantable sensor systems, sterilisation must be considered a critical requirement throughout the product life cycle. The sterilisation method, protective packaging and potential risk of sterilisation failure must factor heavily in risk management. All too often product development focuses on the device, while packaging, sterilisation methods and validation requirements are neglected until further down the pipeline, with detrimental consequences for time to market and development cost. Instead, sterilisation requirements must be considered as part of the early design inputs for the design control and review process. International standards, national requirements, and geographical bias for or against certain sterilisation methods must be taken into account early in the design programme. Specific standards, described more fully below, have a great influence on the sterilisation cycle requirements, which in turn can affect the material qualification in order to use those cycles. Potential risks of:
Verification and validation testing, including biocompatibility testing and other preclinical studies, must be conducted on post-sterilisation ‘finished’ product. (If dunnage or product ‘dummies’ that are non-functional but otherwise represent mass and materials are used, the engineer must carefully justify in the protocol why this substitution for final product should be allowed as representative of final device outcomes.) Thus, if there are changes during development or after design transfer to manufacturing, the biocompatibility and material qualification studies may need to be repeated, again with obvious impact for time and cost to market. As an example, preliminary small-batch cycles may use different combinations of gas (more likely to be 100% EO), may have steeper vacuum cycles and may involve more rapid aeration processes. Although these are often ‘worst case’ conditions, larger chambers may require longer pre-cycle processes involving higher humidity, repeated vacuum and pressure cycles and combination gases. The ability to sterilise the same device with the same packaging in different sterilisation chambers is not a foregone conclusion, particularly if any of the cycle parameters change (such as volume of product to volume of chamber proportions). Compromises made for costs and timing in the early development phase, often unavoidable, may result in delays and added costs if the final cycle affects device performance. Yet, if the developer cannot use small-batch cycles to expedite early testing of implantable sensor systems the fine-tuning of the systems in the development phase may not take place at all, since test units, even for bench testing, must be sterile.
• compatibility with materials used in manufacture, for example new compounds – such as ethylene chlorohydrin – can be formed in the presence of the sterilant that may not de-gas (also known as outgas) as quickly as the sterilant ethylene oxide gas;
Because device verification testing and validation activities must be performed using ‘as marketed’ devices, including final packaging, labelling and a validated sterilisation process, it is essential that sterilisation is considered in all phases of product development. For some regulatory applications, evidence of a final validated sterilisation process may not be required; instead, prototypes used in testing may have been released under ‘small-batch cycle’ release criteria. In such circumstances it is particularly critical that at each design phase review an assessment is made concerning the sterilisation process. It is important to confirm that the process selected is still optimal and that any compromise made during development can be resolved. This may include understanding the impact that the larger capacity cycles may have on tests already conducted so that evolution to the cycle to be used on marketed product is introduced at the best time in the development program. Even after product release, continuous product improvement activities should include monitoring product and packaging performance and new and better cycles as they evolve.
Implantable sensor systems are not new in the context of medical devices. Their development preceded the modern medical device industry boom in the 1960s. Sensors were on board with the advent of the space race in the mid-1940s, used to monitor and broadcast real-time biological parameters of dogs and chimps (Gray, 1999).
Striving to narrow this review to implantable sensor systems brings to focus various currently available technologies described in other chapters of this book, and the future promises broader applications for implantable sensor systems as materials and processes, only now in their infancy, mature. Thus to refine our discussions, implantable sensor systems are those wherein the sensor element is invasive and communicates sensor data to an external system for processing and display, worn either on the body or located remotely. ‘Biotelemetry’ is the term coined during the space race to designate biological parameters from the test subject, which required capture of the signal from the implanted sensor and transmission to an external system, regardless of distance. In most implantable sensor systems, this requires that the ‘detector’ must transform the information from a mechanical or chemical signal to an electrical signal. Even if an electrical signal is created directly, such as with piezoelectric crystals, the signal must be modulated and amplified to a signal that can be exported from the body. In the decades since biotelemetry first evolved, it proved simpler and more reliable to send the sensor signal directly to an implanted therapeutic device. (This lowered power requirements and reduced the potential for signal interference.) For example, an implantable ventricular assist device depends upon an external communication system that can control the input of power, download data from the implanted sensors and thereby influence the programming of the device. Such high-risk, high value systems do not, yet, work independently of a pre-programmed directive or human intervention based upon the downloaded sensor data (Proven Process, 2011). Sophisticated sensors can collect data and store the information in buffers until the unit is interrogated by an external device. This ‘dumbs down’ the requirements of the biotelemetry sensors, which were expected to continuously monitor and continuously transmit their data in real-time to the earth-based scientist.
The aim of this chapter is to review the potential impact and value of various sterilisation methods for implantable sensors systems, and thus show manufacturers and developers of new devices the need to recognise the benefits and the risks of commercially available sterilisation methods and their potential impact on sensor system performance. All implantable sensor systems require environmentally controlled manufacturing systems and processes to ensure the sensor system does not introduce microorganisms or pyrogens (typically the denatured remains of bacteria) during implantation and thus reduce the potential for infection during surgery. Until now, the assumption has been that only the outside of the sensor system needed to be sterile. However as sensor technologies evolve, the limits to true ‘hermetic seals’ are better understood and it is desirable that fluids ingress and egress the device to perform the sensor function. Therefore, ‘terminal sterilisation’ processes (the device is sterilised only once, in its final configuration, and there is no attempt to ensure all internal components are also sterile) could be inadequate. Due to the widening definition of what constitutes an implantable sensor system and what one might do, the industry is now evolving the potential option of aseptic manufacturing methods. These methods typically would obviate the need for harsh sterilisation environments destructive to sensitive chemistries and mechanisms, although costs of construction could be considerably higher.
Regulatory controls for implantable sensor systems are surprisingly similar around the globe considering the significant differences for the general medical device industry. For most situations, the highest level of regulatory scrutiny is required for implantable sensor systems, across all regulatory boundaries. Globally, regulatory agencies are in agreement that implantable sensor systems are to be regulated at the highest level, but it is the specific differences in the regulations that vary around the globe.
The major reason underlying the differences between how implantable sensor systems and other medical devices are regulated is the source of the regulations. In the USA, regulations were motivated by unsafe devices and unsafe medical practices, which had to be regulated ‘against’. In contrast, throughout the majority of the other industrialised regions, ‘standardisation’ of the definition of quality for medical devices was seen as an economic stimulus. When devices could meet a commonly held quality expectation, despite where they were made, they could be sold to any country accepting that level of quality. In contrast, the ‘overseer’ regulatory scheme is not voluntary; it is experience based and usually arbitrary. Despite how these different approaches evolved, they have both evolved to extraordinary common ground for what constitutes ‘unsafe’ performance. Broadly recognised international standards have sought to make the systems compatible with one another for purposes of power and data outputs, but have not managed to achieve common acceptance criteria for safety.
Manufacturers of implantable sensor systems are obliged to comply with the requirements of the relevant European Directive(s), viz the Medical Devices Directive (MDD; Council Directive 93/42/EEC) and the Active Implantable Medical Devices Directive (AIMDD; Council Directive 90/385/EEC). Medical devices are classified according to the level of risk they present to the patient. Class I devices are the least risky and many are supplied non-sterile, for example, bandages and tongue compressors. Class IIa, IIb and III devices are subject to increasing levels of regulation, but sterility is an absolute and therefore is not influenced by classification. Implantable sensor systems are likely to fall into Class III and to require clinical studies under a notification to the country’s Competent Authority (if there is any uncertainty, the manufacturer’s Notified Body can be contacted). There is a choice of conformity assessment routes to the award of the Conformité Européenne (CE) mark and many manufacturers elect to use a Quality Management System (QMS) approach. Design verification and validation processes and activities are similar to those described below for the United States and, while there are some nuances between the regulatory requirements of the two regions, the requirements for sterilisation remain the same. Similarly, there are globally recognised methods for sterilisation and sterilisation validation.
Thus, requirements concerning sterilisation methods are implicated at all parts of the product life cycle and have significant impact within the risk management processes. For example, where an implant must be sterile inside and out, manufacturing is influenced by the need for aseptic procedures. These can limit the manufacturing processes and packaging, which in turn could influence the sterilisation method and product shelf life.
Numerous European and international standards apply to the sterilisation of medical devices. Lists of applicable sterilisation and other standards are available from the International Organisation for Standardisation (ISO),1 the American Society for Testing and Materials (ASTM)2 and the Association for the Advancement of Medical Instrumentation (AAMI).3 However, there is no legal obligation for a manufacturer to comply with any particular standard or guidance document. Rather, there is an assumption that by complying with the standards manufacturers will meet the requirements of the directives. Developers and manufacturers of novel medical devices involving cutting edge technologies can find themselves ‘ahead’ of the published and/or draft standards. It is wise to agree with the Notified Body which standards are applicable and what constitutes adequate demonstration of compliance. Medical device manufacturers using a sterilisation process for which a specific standard does not exist should refer to ISO 14937 (2009).
In practice, sterilisation procedures and processes for medical devices are conducted under the auspices of the QMS. Guidance about the requirements of the QMS for design and development of medical devices and their production, installation and servicing is provided in ISO 13485 (2009). This standard does not tell the manufacturer how to conduct sterilisation but provides guidance about conducting testing and managing data. While conformance with ISO 13485 is voluntary, the Notified Body must be satisfied with the level of ‘voluntary conformance’.
Sterilisation is an example of a process that cannot be verified by subsequent inspection and/or testing. Because demonstration of sterility involves opening the product, it cannot then be used for its clinical purpose. Instead, sterilisation processes are validated for use, the equipment is maintained, and the process is monitored routinely (see Section 8.4).
The United States Food and Drug Administration (FDA) regulates medical devices to ensure the design, clinical testing and commercialisation of implantable sensor systems meets extensive regulatory requirements prior to marketing. There is no single regulatory citation for implantable sensor systems, although there are certain areas of regulation and controls that are more applicable. It is relevant to the understanding of medical device regulations today that the Food Drug Law and Cosmetic Act has been revised many times since 1906, with the most significant regulatory change occurring in 1976 when Good Manufacturing Practices (GMP) were introduced, along with the hierarchy of named devices and associated risk categorisation. Amendments subsequently have added federal laws, which have then been interpreted into regulations. These regulations are issued by the FDA by way of notice in the Federal Register. Their interpretation and implementation are frequently described within a guidance document. Very often, the guidance document ‘recognises’ an existing national or international standard. Some device regulations have accompanying ‘recognised standards’ without specific guidance.
The level of regulatory oversight generally follows the risk classification of the device. Class I devices have the lowest level of regulation, consisting of registration and listing. Class I devices are most often exempt from pre-market notification (i.e. 510(k) clearance) but are usually non-invasive, sold non-sterile and may be exempt from certain quality system requirements. Class II devices represent the largest number of new devices that come to the market in the USA, as long as the device is ‘substantially equivalent’ to another marketed device for which certain controls are considered adequate for maintaining safety and performance. Class II devices range from implantable devices with a long history of use to electronic/software controlled devices with sophisticated performance requirements. Class III devices are typically high-risk or devices for which no prior clinical history exists and require full premarket approval (PMA) prior to market release. For the most part, a developer of an implantable sensor system should expect that the product will require a PMA. Some implantable sensor systems that incorporate biologics or pharmaceuticals in combination may also involve a dual submission coordinated by the Office of Combination Products.
Medical device oversight in the USA is multi-layered and extends to the full life cycle of the product. The FDA regulates the early development of medical devices (Class II, Class III and Class I devices with software) with requirements known as Design Control and Review (21 CFR 820.30). These procedures (briefly described above) require documentation of design inputs, design planning and design reviews with a well-ordered design history file. Design verification and validation ensure the design outputs meet the inputs and that risks have been mitigated. Preclinical animal testing is regulated through Good Laboratory Practices (21 CFR Part 58) and clinical investigations are regulated through Investigational Device Exemptions (21 CFR Part 812). Once a product is marketed, the FDA requires registration of the manufacturer and listing of the products. The FDA will inspect according to classification for conformance to the Quality System Regulations (QSR) (21 CFR Part 820). The FDA may also impose post-market surveillance, device tracking and annual reporting for devices even after market approval.
Extensive regulation of manufacturing processes, including sterilisation and packaging, ensures only devices meeting sterilisation requirements can be marketed. Implantable devices must be validated to demonstrate a Sterility Assurance Level (SAL), using methods identified in published standards (ISO 14937, 2009). Sterilisation is not viewed as an absolute, but rather as a probability. As an illustration of this principle, for radiation sterilisation, when selecting a sterilisation dose the objective is to establish the minimum permissible dose necessary to provide the required or desired SAL. SAL refers to the ‘probability of a viable microorganism being present on a product unit after sterilisation’. So, for an implantable sensor system to be considered sterile the dose must be high enough to ensure that the probability of an organism surviving the dosage is no greater than one in one million units tested (i.e. SAL 10− 6; ISO 14937, 2009).
Most manufacturers of medical devices depend upon contracted sterilisation companies and laboratories to conduct the sterilisation process regardless of the type of sterilisation, though some small sterilisation systems do exist for in-house sterilisation of product. In-house sterilisation largely depends upon the type of sterilisation employed because many commonly used systems require extensive safety and environmental controls. Thus, the least expensive and most reliable forms of medical device sterilisation are conducted in central locations with high capacity sterilisers and trained personnel. Typically, regulatory authorities inspect contract sterilisation facilities as well as the device manufacturer’s documentation to confirm that sterilisation processes have been validated and that each batch was confirmed to have functioned according to the required parameters before product release.
Facilities for medical device sterilisation can now be found in nearly every country where medical device manufacturing is conducted. Many sterilisation service companies are international corporations, thus making easier the task of exporting medical devices around the world. These international contract service providers not only subscribe to international standards for cycle validation, they may be recognised specifically by the host country, so the products sterilised there may also be sold in the same country provided the host country registration and regulatory requirements have been addressed by the original equipment manufacturer. The universal adoption of and compliance with the internationally recognised standards for facility operation and device release make the transfer of developed implantable sensor systems to production facilities around the globe feasible.
Globally, design verification and validation processes and activities are similar with some regional variation, for example in China and Japan. However, the requirements for sterilisation and the principles and processes for validating sterilisation presented above for the USA are applicable globally.
In the following section, various sterilisation methods suitable for medical devices are discussed. They include sterilisation using steam, EO and radiation as well as a few alternative methods potentially suitable for implantable sensor systems.
A sterile medical device is one that is free of viable microorganisms. International standards pertaining to sterilisation validation and routine control of sterilisation processes require that microbial contamination is minimised prior to sterilisation. Nevertheless, medical devices manufactured under the required controlled environmental conditions could still have microorganisms on them. The objective of sterilisation is to inactivate microbial contamination, thus turning a non-sterile device into a sterile one. Sterilisation must achieve a SAL of 10− 6 (see Section 8.4.5).
Sterilisation should not be confused with disinfection or sanitisation. By definition, sterilisation terminates microorganisms. On the contrary, disinfection and sanitisation reduce the number of pathogenic organisms to what are considered ‘acceptable’ levels, as for example by Pasteurisation (Guideline for Disinfection and Sterilization in HealthCare Facilities, 2008). Sterilisation can be achieved by applying the appropriate combinations of heat, chemicals, irradiation, high pressure and filtration.
Steam sterilisation is most commonly used for medical devices such as surgical instrumentation and is unlikely to be the method of choice for implantable sensor systems. Using steam is one of the most reliable sterilisation methods, but it damages many plastics, electronics, fibre optics and biological materials. Therefore, only a brief overview of steam sterilisation is presented here.
Autoclaves are widely used for heat sterilisation and commonly use steam heated to 121–134 °C (250–273 °F) with a holding time of at least 15 minutes at 121 °C or 3 minutes at 134 °C, longer for liquids and surgical instruments packed in layers of cloth. Treatment inactivates all fungi, bacteria, viruses and also bacterial spores, which can be quite resistant to some methods. The most common, and historic, steam sterilisation cycles used in the medical device industry are gravity-displacement and dynamic air removal (Perkins, 1982). In a gravity-displacement system, steam enters the sterilisation chamber and displaces the residual air through an open vent (hot air rises). However, dynamic air removal has been shown to be more efficient because the machine can pump in conditioning air (humid and warm typically), then forcefully discharge this environment and replace with subsequent cycles.
A draw down vacuum is used to remove the conditioning air cycles from the packaged product and the chamber. Ambient air is removed from the chamber in what is known as a ‘pre-vac’ cycle, which is typically a series of pressure and vacuum excursions. These serial staged cycles provide the time and conditions necessary to ensure the entrapped chamber atmosphere can be withdrawn from within the package and from within the product itself.
Another method, flash sterilisation, involves much higher temperatures being applied for a shorter time and is suitable for devices for immediate use, such as surgical instruments. Usually ‘flash’ refers to an open batch sterilisation cycle where instruments are only lightly wrapped, if at all, only high temperature steam is injected into the chamber, and little or no vacuum cycling is involved. Flash sterilisation is unlikely to have any utility for sterilising implantable sensor systems since the high temperature and humidity could harm the materials and electronics. Even if the sensors withstand the cycle, the conditions cannot force sterilant into the sensor elements to ensure sterilisation of any accessible monitoring chambers.
Gas sterilisation systems expose the product to high concentrations of very reactive gases at relatively low temperatures compared with steam sterilisation. Commonly used gases are alkylating agents such as EO and oxidising agents such as hydrogen peroxide and ozone. EO is a common choice for sterilising medical devices where good penetration into the device is required and other methods could affect material properties. EO is a strong candidate for the sterilisation of implantable sensor systems, since other methods using heat and moisture are likely to impact upon the product and radiation can affect plastics, electrics and optics.
EO is highly toxic and residuals can be harmful. EO is an irritant and sensitiser, with known mutagenic and carcinogenic potential. It is toxic by inhalation, ingestion and through skin contact (typical EO material safety information4). Therefore, product verification should include biocompatibility testing for EO and ethylene chlorohydrin residuals. ISO 10993-7 (2008) sets acceptance limits for residuals. Because implantable sensor systems are intended for long-term or permanent use, they fall into the most stringent category. ISO 10993-7 also provides guidance on the testing itself.
EO treatment is generally carried out between 30 °C and 60 °C with relative humidity above 30% and a gas concentration between 200 and 800 mg/L, typically taking at least three hours. EO is highly effective, penetrating well through paper, cloth and some plastic films. EO kills all known viruses, bacteria, including spores, and fungi. A typical process consists of a pre-conditioning phase, the actual sterilisation run and a longer period of post-sterilisation aeration to remove the toxic residues (for detail see Sections 8.3.5 and 8.4.6).
The most relevant large scale EO sterilisation method for device manufacturers is the gas chamber method. To benefit from economies of scale, EO is delivered by flooding a large chamber with a combination of EO and other gases used as dilutants (usually carbon dioxide). This method has drawbacks inherent in the use of large amounts of sterilant being released into a large space. Processing is highly regulated and, consequently, the required level of control means it is becoming less popular in hospitals.
Despite its toxicity, EO has been in common usage since the 1950s and is the most frequently used method for sterilising medical devices worldwide, with an estimated annual growth rate of 2–4% per annum. The reasons for this are because it is the most cost effective low-temperature sterilant for widespread use and it has the highest material compatibility factor of any sterilant, even when applied repeatedly. Nevertheless, care should be taken to ensure the compatibility of device materials with EO.
Methods of sterilising medical devices using radiation include X-rays, gamma rays, and electron-beam (ISO 11137-1, 2006). Their suitability for the sterilisation of implantable sensor systems depends on compatibility with the sensor materials and on the volume of product.
High-energy X-rays are a form of ionising energy suitable for irradiating large volumes of product. Their penetration is sufficient to treat multiple pallet loads of low-density packages while maintaining a uniform dose across all of the product. As well as benefitting from quantity, X-ray sterilisation is a ‘clean’ electricity-based process not requiring chemical or radioactive material. Commercial use for medical devices began in the mid-1990s, though adoption has been slow. The current availability of high-power accelerators may increase the market share (Stichelbaut et al., 2006).
Gamma rays are very penetrating and are used for the sterilisation of about one-third of medical devices, especially for disposable medical equipment, such as syringes, needles, cannulas and intravenous sets. The disadvantages of this method are that gamma radiation can affect some materials and involves an isotope, usually Cobalt-60, meaning that operators require bulky shielding and storage of the isotope presents a hazard for the facility (Technical information, Cobalt5).
Electron-beam (E-beam) processing is less frequently used for medical device sterilisation but is increasingly popular with manufacturers for whom it is suitable. Unlike gamma radiation, E-beams use an on-off technology and provide a much higher dosing rate than gamma or X-rays. Another advantage is that owing to the higher dose rate, less exposure time is needed, lowering the potential for degradation to polymers. A limitation is that electron beams are less penetrating than either gamma or X-rays.
Other less conventional, less frequently used methods may be considered as suitable alternatives for implantable sensor systems, because these systems have different performance and risk priorities compared with large volume, low cost disposable medical devices. When protecting the performance of the implantable sensor is critical, neither extra time nor cost will prevent these alternative sterilisation processes from being the method of choice.
Vaporised hydrogen peroxide (VHP) sterilisation is a low-temperature gaseous method of sterilisation. Compared to EO gas, VHP typically cannot penetrate large, dense packaging, but it can offer an alternative where material compatibility with EO is a problem. Because hydrogen peroxide breaks down to water and oxygen, aeration time is greatly reduced and any concern for toxicity is very low risk. The FDA has granted 510(k) clearance for the use of various models of vapour sterilisation systems to terminally sterilise medical devices (K083097;6 K0713857). However, when used in an industrial setting, the sterilisation equipment and support systems must undergo installation, operations and sterilisation validation which can be time-consuming and costly. To date, these constraints have retarded the widespread use of the VHP sterilisation method, but the process and chemistry are relatively more compatible with the materials, making this a suitable alternative for implantable sensor systems.
Numerous other chemical sterilisation systems that create penetrating vapours currently have limited availability or are still under development and qualification. Chemical sterilants such as ozone, nitrogen dioxide and supercritical carbon dioxide may encounter constraints for validation of sterility because the commercial sterilisation indicators are not as widely available, and fewer laboratories are trained in their use (Lambert, 2010). One very popular peracetic acid sterilisation system for endoscopes and devices that are sensitive to moisture and radiation, Steris 1, is now under mandatory recall by the FDA (Steris 1 peracetic acid sterilisation system recall8). Such large scale recalls can make manufacturers nervous about adopting new technologies, even when the systems have had FDA clearance to market.
Liquid chemical sterilants (also known as germicides) such as ortho-phthalaldehyde (OPA) are now in use in Europe and the USA and have been shown to have superior mycobactericidal activity compared with glutaraldehyde (New Disinfection and Sterilization Method: Ortho-phthalaldehyde: a New Chemical Sterilant9). These sterilants do not irritate the eyes and nasal passages as much as glutaraldehyde. Although OPA has good materials compatibility, it stains proteins grey. Colour changes in materials as a result of sterilisation are often perceived as undesirable characteristics, particularly if the colour change makes the product look dirty even if there is no effect on device function. To cover up the colour change, manufacturers sometimes add pigments, further complicating the qualification of the product in terms of biocompatibility. Furthermore, disposal of spent OPA requires special precautions, which can add to the cost of processing the product. Any sensor manufacturer hoping to use this chemical steriliser would of course have to conduct extensive testing on the implantable sensor system to rule out potential effects of the chemical on the sensor output. Without experience of the chemistries involved it could be difficult to anticipate the various modes of failure.
Sterilox is super-oxidised water, made from saline, and effective against a wide range of organisms (K07138510). Although described as not damaging to the environment, the most active component is hypochlorous acid. Its compatibility to many medical device materials has not yet been demonstrated. The sterilisation system can be corrosive to certain materials but is non-toxic to biological tissues.
Aseptic processing standards are being developed for medical devices to eventually enable greater acceptance of devices where processing procedures attempt to maintain sterility of the device throughout the manufacturing process to the point of use. Aseptic manufacturing means that the components and materials that comprise a device are pre-sterilised appropriately and all materials, equipment and support systems are used only after sterilisation. All working steps are performed in clean areas to avoid contamination. For such processing, only the highest standards of purity and cleanliness for the manufacturing room, the personnel, the equipment, and the supply of air, water, sterile gases and materials used in the working process must be maintained at all times. Maintaining such control is usually prohibitively expensive for medical devices.
Dry-heat sterilisation has been available for decades (Darmady et al., 1961), but has limited commercial uses due to the considerations of material impact from the high temperatures (170–180 °C) and time (cycles of an hour or longer). It has increasing interest for instrumentation because packaging materials that can withstand dry-heat sterilisation have become available. Its utility for implantable sensors made with polymers is unlikely due to the high temperature.
Many different alternative sterilisation methods are available. The design, materials and limited environmental functional range of implantable sensor systems may require the use of these alternative sterilisation methods despite their obvious drawbacks if conventional sterilisation methods simply do not work.
Currently, manufacturers of implantable sensor systems report using the more common sterilisation methods such as EO gas (Ohashi and Karube 1995) and E-beam irradiation of complex medical devices (2010). Sterilisation systems that use potentially corrosive chemicals such as hydrogen peroxide are avoided due to the potential impact on component electronics and batteries. For the same reason, and for potential effects on materials, hot steam is not used, although some manufacturers have investigated high heat.
EO sterilisation offers low heat, good penetration, and high throughput and is therefore likely to be the method of choice for manufacturers of implantable sensor systems. Hence, this section focuses primarily on EO technologies for the sterilisation of implantable sensor systems.
The choice of sterilisation method may be unique to the implantable sensor technology, considering how the implantable sensor is designed, the materials therein, any coating or interface technologies employed and its methods of construction. Incorporating principles of Design Control and Review, designers must evaluate how materials and processes support or hinder the performance objectives of the sensor system, and must evaluate these periodically as the program moves through each design phase of development.
The implantable sensor technology and intended use drive the decisions during development about cost, materials, electronic components, power supplies and hermetic seals. For example, it is important to appreciate that beyond the use of implantable sensor systems for medical applications in humans, some sensors are used to track and record the conditions of wild and domestic animals for bioresearch and habitat evaluation. A sensor that will be implanted in a wild animal to collect data for a single breeding season will have different design requirements than a glucose sensor implanted in a human intended to interface with an implanted insulin pump. For instance, sensors for wild animal tracking may be potted in biocompatible wax and thus may only last a few months in the animal, but the sterilisation of the system is just as crucial for this application as for human-use (Burger, 1994). An implantable glucose sensor in an insulin-dependent diabetic human must be dependable and designed to be ‘fail-safe’, which may mean giving warning to the wearer of the pump and considering risks of fluid egress if hermetic seals fail (Woodward, 1982). Implanted sensor systems in both applications face daunting environmental challenges.
Host response to implanted sensors is the primary cause of sensor failure (Bridges and Garcia, 2008). The available biomaterials able to withstand fluid egress while allowing the transfer of biochemical vectors represent the greatest challenge to the performance of implantable sensor systems. For example, hydrogel coatings have been explored as a potential surface modification to control the host reaction to the biomaterial, but hydrogels can be adversely affected by the conditions required for sterilisation, such as the heat and gases from EO or the oxidative effects from radiation sterilisation. Anti-fouling surface treatments often include active ingredients that provide anti-inflammatory agents to the surface as a means to counteract the host response, such as the use of nitric oxide (Wu and Meyerhoff 2008). As new materials and combinations of materials with coatings, drugs, biologics and mechanical treatments are developed, each must be evaluated as sterilised components within or on the surface of the device.
Once these factors are documented, they provide further input to the designer whose challenge is to select a sterilisation system that will not affect the performance requirements of the sensor system. Frequently, this will require trial and error, as various cycles may be investigated for their impact on the materials and whether the cycle mechanics (such as heat and energy) could affect sensitive electronics or delicate biomaterial properties. As an example, early pacemakers were epoxy-encapsulated and thus could be sterilised by EO gas. As hermetic seals were sought to protect the internal circuitry and thus ensure longevity, the ability to sterilise the inner-parts became less of a concern because rupture of the hermetic seal was highly improbable over the life-time of the device.
Once the product development phase can demonstrate an effective design, manufacturing must maintain strict controls to ensure the process parameters are stable and no variables creep into the process to affect outcome. Reliable material and component suppliers are vital and must be under a supplier agreement not to make changes without notifying the manufacturer, but this remains one of the most difficult aspects for medical device manufacturers to control because of the relatively small volumes of materials and components used for implantable sensors. This requires constant vigilance and a refined appreciation for the potential hazards associated with unanticipated manufacturing process changes. An unqualified change to a material or component undergoing sterilisation could create havoc within the delicate implantable sensor system.
Other than packaging failures, the most significant risk for sterilisation failure is the potential that the manufacturing process will fall outside nominal conditions for microbiological contamination. Sterilisation validation is highly dependent upon quantifying the bioburden on the product to be sterilised. If a manufacturing facility has a sudden increase in the microbiological load, the sterilisation process could be inadequate despite maintaining pre-determined sterilisation cycle parameters. To this end, manufacturers attempt to maintain controlled environments in the manufacturing and packaging areas. Frequently referred to as ‘controlled environment’, medical device manufacturers need to sustain cleanliness and sanitation of the manufacturing process to varying degrees, dependent upon the specific process in the room and the requirement of the product. Most ‘clean rooms’ do an adequate job of creating a dust-free environment by controlling particulate matter. Although the QSR, the MDD and ISO 13485:2003 require that the manufacturer provides a controlled environment for the manufacturing of the product, there is no specific requirement to meet a certain level of clean room certification (see below).
The permitted size, amount and type of particulates on a product will vary with the sensitivity of the sensor performance parameters to the presence of dirt. Manufacturers may need to appreciate the difference between hazards created by ‘sterile dirt’ and those created by microbial contamination of the sensor. For example, a highly sensitive device could be affected by dirt that responds to static electricity, by permitting an arcing between two points. Another type of sensor could be compromised by the presence of denatured proteins that compromise the sensor’s ability to detect the presence of a blood agent to be monitored.
The two most significant potential factors for product contamination (beyond the manufacturing process itself) are the personnel who enter the environment and the potential for incursion of the external environment into the controlled space. A single person working in a clean room environment can produce 100 000 particles per minute or three million particles in a thirty-minute shift. High efficiency particle arrestor (HEPA) filters remove most of the particulates that are 0.5 micron size and larger. Thus, it is unlikely that a clean room can maintain a clean environment by filtration alone, much less an aseptic environment if aseptic processing is also a requirement.11 Standard operating procedures (SOPs) and personnel training attempt to minimise the impact of personnel on the potential bioburden by gowning, hand-washing, gloves, nets and booties, but bioburden deposited by personnel from skin, hair and expiration are difficult to control even in the best facilities.
Manufacturers must proactively establish, as part of QMS conformance, specifications for the clean environment and have objective evidence that the specifications have been met. Manufacturers must not only establish adequate controls and monitoring systems to demonstrate the required level of cleanliness has been maintained, but must also understand explicitly how product performance would be affected by any failure to maintain those conditions. Environmental controls must be adequate for purpose, quantifiable, measurable and sustainable.
Standards for clean rooms and their classification are described in ISO 14644-1 (1999). The use of HEPA filters for air filtration in clean room facilities is a common method of reducing incursion of the environment into the clean room and can reduce the particulate burden from recycled air by capturing particles up to a certain size. However, HEPA filters cannot sterilise the air that passes through and are not themselves sterile. HEPA filters must be maintained and the air supply equipment (ductwork and vent screens) downstream of the filter must be well sealed. Laminar flow, highly filtered air is intended to move particulates generated by the process and operator away from the device without creating turbulence that would stir the particulates from surfaces into the air. Downward, positive pressure, laminar flow, HEPA filtered air produces the best clean room environment, but alone is not usually sufficient.
Other techniques to reduce the potential risk of biological contamination include airlocks, product pass-through windows, positive air flow (the room is maintained at a higher pressure than the surrounding facility) and sticky-mats at door thresholds. To eliminate bacteria from the clean room, sinks and water supplies must be eliminated entirely, but this depends on the processes required in the room. Workflows should be organised to maintain the best air quality at most critical manufacturing steps.
Meticulous attention to detail is required within the facility and by personnel to avoid the potential risk of introducing either unknown organisms or routine but excessive biological contamination. Most companies conduct settling plate analysis (Guidelines on Test Methods, 2004) during an entire manufacturing cycle in order to independently assess if the bioburden during manufacturing is maintained no higher than the level used in the qualification of the sterilisation cycle. Hazard analysis critical control point (HACCP) at the process level can help to identify those process steps that might most critically put the product at risk for contamination. HACCP can help the manufacturer understand which steps in the process to optimise in order to reduce unanticipated biological level excursions.
Designing and developing packaging systems for terminally sterilised medical devices is critical to bringing a new or modified product to market. Together, the packaging and contents should comprise a system that is efficient, safe and effective. ISO 11607-1 (2006) provides general guidance including quality systems, sampling, test methods and documentation, and specific guidance about materials, design and development, and information to be provided by the manufacturer.
Whatever the method of sterilisation, the choice of dose is made to assure sterility while minimising risk to patients, users and operators. Freedom from microorganisms cannot be measured for medical devices, because doing so renders the device unsterile and the numbers required are prohibitive. Therefore, a SAL is used, that is the probability that a viable microorganism remains on a medical device after sterilisation. The accepted SAL for medical devices is a probability that one in a million, or 10− 6, devices should be non-sterile (ISO 11137-2, 2006). For EO, which is likely to be the method of choice for manufacturers of implantable sensor systems, the potential hazards are the toxicity of the gas and its post-sterilisation residuals and therefore ideally the dose will be the minimum that meets the SAL.
A typical EO process is described by Isotron and paraphrased herein with permission (Isotron, 2011).
Initially, finished (packaged) product, the ‘load’, is brought to the required temperature and moisture content using a two stage pre-conditioning procedure, which lasts approximately 24 hours. Final temperatures are typically between 40 °C and 60 °C with relative humidity levels of 45–75%.
Having achieved the necessary temperature and humidity, the load is transferred into the sterilisation chamber. The stainless steel sterilisation chamber is evacuated and flushed with nitrogen to remove air, steam is injected to the chamber to re-establish the necessary moisture content and EO is then introduced until the required concentration is achieved. The chamber itself is heated by circulating hot water through a surrounding jacket, maintaining the temperature typically within ± 1 °C of the target temperature. Circulation fans within the chamber ensure rapid and uniform distribution of EO right through into the centre of the load.
Product is held under these conditions for a defined period, typically 2–4 hours. Any loss of EO as a result of absorption into product and packaging is automatically compensated for during the process. On completion of the pre-determined dwell period, EO from the chamber is exhausted to the atmosphere via a catalytic converter. This unit ensures catalytic conversion of EO to carbon dioxide and water typically with an efficiency greater than 99.9%, ensuring that atmospheric emissions from chamber to atmosphere remain within internationally accepted environmental limits.
The sterilisation chamber and its contents are then repeatedly flushed with nitrogen or air to remove the remaining EO from the chamber. Air pulsing and vacuum hold can also be used to accelerate the removal of the EO.
After the post-sterilisation flushing is completed, product is transferred to an aeration cell, where it is subjected to high rates of air change at temperatures close to the sterilisation temperature for approximately 12 hours This phase of the process serves to draw further residual EO from product and packaging. The exhaust from this phase is also treated via the catalytic converter.
On completion of initial aeration, optionally, product can be transferred to a secondary aeration area where an additional period of elevated temperature storage can be used to further reduce residual EO levels. This takes place at 25–35 °C, typically for up to seven days. Alternatively, product can be transferred directly to storage.
Cycle parameters should be individually controllable and generate separately identifiable time-stamped records. This allows pre-conditioning time and temperature to be optimised to meet the requirements of the specific medical device as well as a paper trail of the cycle parameters. In the sterilisation cycle, the parameters of temperature, humidification, gas concentration, evacuation rate and level, and gas exposure period can all be controlled independently to develop a sterilisation cycle meeting individual requirements. Both the device manufacturer and the laboratory personnel are responsible for confirming that the documented cycle parameters met the approved cycle requirements according to the sterilisation validation. Strict conformance to calibration of cycle monitoring is required to ensure the manufacturer has an accurate documentation of the cycle parameters, which is included in the device history record.
Prior to routine sterilisation, a complex validation study is undertaken to ensure that the chosen sterilisation cycle is effective in producing sterile product routinely. For routine processing, the validated cycle is monitored by including a series of Biological Indicators (BIs), which contain a known number of an indicator organism. These are placed at various positions across the sterilisation load. A ‘no growth’ result from the BIs indicates successful processing. Prior to release of a batch of product, all process records, BI test records, and cycle parameters should be reviewed against the pre-set specifications, recorded and included in the batch record.
The ISO 11135 Standard and the guidance in TIR 28 define the principles of validation and routine control for EO sterilisation. Prior to routine sterilisation, a complex validation study is undertaken to ensure that the chosen cycle is effective in routinely producing sterile product. Optionally, TIR 28 facilitates the development of sterilisation cycles allowing smaller pallet quantities of product to be used for validation.
Sterilisation validation for a new implantable sensor system can be especially challenging for a manufacturer, because the most commonly used cycles may not work for the new product. The heat might be too high, the vacuum too low, the aeration too long, the humidity may compromise the sensor or the gas could affect the biochemistry. Chemical and biological indicators, such as spore strips, are usually too bulky to be placed directly on a sensor system and certainly are seldom ‘inserted’, as with more traditional product, because the sensor systems are often sealed, quite small or could be damaged by insertion or removal of the indicator. Even a simple test for cycle residuals (EO and ethylene chlorohydrin, ISO 10993-7, 2008) is a problem for a sensor system due to the cost of making product that is sacrificed during testing. Typical testing for cycle residuals using extraction methods calls for huge surface areas (ISO 10993-7, 2008) when viewed from the perspective of the size of a sensor, and the sensors usually represent a major financial investment by the time the device has gone through the manufacturing process. Rare and expensive materials are often required for sensors but a residual test for EO requires extraction.
All of the conventional methods for validation of an EO sterilisation process involve destruction of an expensive, small, sealed device. Some companies will therefore attempt to validate a sterilisation cycle with ‘dunnage’ (dummy devices). Ideally, the dunnage will have the same properties as the device in terms of volume, weight, air-penetration, absorptive properties, etc. Sterilisation validation procedures usually require sterilisation of three lots of a product, representing three complete manufacturing builds.
Unanticipated harm to the product, only discovered when sterilisation is first attempted can represent a major economic setback for a small company. Therefore, pre-qualification of materials, manufactured semi-finished goods, packaging and labels prior to conducting sterilisation validation is an absolute must for manufacturers of implantable sensor systems. A comprehensive sterilisation development and validation plan, early in product development to qualify the impact of the sterilisation cycle on the materials and subcomponents can serve as the foundation for further testing. It is sensible for manufacturers to test the intended sterilisation method during development, using a staged approach, to avoid any unanticipated effects being discovered after the design and materials have been locked. All design verification and validation studies, as well as stability testing, must be performed using implantable sensor systems that have been sterilised using either the validated method or the method that will be validated later in product development.
Sterilisation validation is costly and time-consuming. Therefore, some medical device manufacturers have developed a sophisticated strategy to justify not repeating full sterilisation validation, based upon the concept of ‘product families’ (Mohanty and Kougianos, 2006). This strategy is not recommended for manufacturers of implantable sensor systems, even if they do have a range of similar devices. The designation of a ‘product family’ for the purpose of ‘adopting’ sterilisation validation data requires demonstration of similar bioburden data and characterisation of organism types. Any change in manufacturing (materials suppliers, process changes, etc.) may prove too variable to ever establish a ‘family’, leaving the manufacturer without a validated sterilisation cycle.
Unless in-house facilities are available, sterilisation validation and services require a qualified, accredited, experienced supplier with a good track record. Look for a good QMS (e.g. ISO 90001, ISO 13485), compliance with standards that are appropriate to your device, your needs and market regions (e.g. ISO 11135, ISO 11137, 21 CFR 820 [GMP], ISO 17025) and whether they have facilities conveniently located for your manufacturing sites and target markets. Always audit potential facilities, and re-audit as required. As well as auditing the QMS, audit the processes and data at both the manufacturing site and the sterilisation site. It is critical to confirm that the parties can demonstrate a chain of custody and that processes have been conducted according to the agreed protocol. Data must be available for auditing by regulatory authorities and must demonstrate that the cycle achieved sterility.
Sterilisation of implantable sensor systems is not for the faint hearted. No single sterilisation process exists to suit all sensors and nearly all existing processes will have trade-offs. All solvents can prove corrosive to the components or compromise material properties of the outer casing. Terminal sterilisation may not be satisfactory if a sensor communicates with body fluids. Considerations for the impact of the sterilisation systems on sensitive biomaterials and surface treatments will further limit options. Even when an ideal process can be selected, the sterility of the manufacture lots must be demonstrated, with data that can be audited, based upon process validations that recognise the potential risks.
Without question, implantable sensor systems will expand the link between real-time diagnostics and real-time therapeutics, optimising the use of medicines, tailoring treatments, and reducing the risk of life-threatening conditions. The interfacial sensory requirements have already pushed known biomaterials to the limits, and successful implantable sensor systems will continue to push material research and technologies, thus enabling extremely sensitive sensors and miniaturisation. In the midst of delicate sensor technologies and materials, it is hard to imagine that, for the future, sterilisation technologies such as gamma radiation and EO sterilisation will adequately serve some parts of this sector of the medical device industry. Finely tuned E-beam radiation or alternative chemical sterilants may serve these needs, but aseptic processing is likely to be required throughout the manufacturing process to satisfy the requirements of a truly real-time, dynamic, implantable sensor system for continuous operation.
Sources for further reading are available from industry groups and regulatory and standards agencies. High-level sources are recommended below, through which more specific product related detail can be sourced:
American National Standards Institute, http://www.ansi.org/
Applied Regulatory Consulting European Regulatory Intelligence, http://reg-info.com/europe
ASTM International Standards, http://hosted.verticalresponse.com/256708/7a9a5fb95c/1399024779/5df7145dd2/http://www.astm.org/Standard/index.shtml
European Commission, Medicinal products for human use, http://ec.europa.eu/health/human-use/index_en.htm
FDA guidance on sterilisation, http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/GuidanceDocuments/ucm252999.htm
Global Sterilisation: Making the Standards Standard, http://pharmaceuticalvalidation.blogspot.com/2009/12/global-sterilization-making-standards.html
ISO/IEC 17025:2005. General requirements for the competence of testing and calibration laboratories, http://www.iso.org/iso/catalogue_detail.htm?csnumber=39883
World Medical Device Organisation, http://www.wmdo.org/
Darmady, E.M., Hughes, K.E.A., Jones, J.D., Prince, D., Tuke, W., Sterilization by dry heat. Journal of Clinical Pathology 1961; 14:38–44 Available from. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC480155/pdf/jclinpath00060-0042.pdf
E-beam irradiation of complex medical devices. 2010. Available from. http://www.mddionline.com/article/beamone-expands-meet-complex-device-demand?quicktabs_2=0
Example of a protocol for clean room operation and maintenance, with general information, http://www.liberty-ind.com/pdf/Maint_Protocol.pdf
Example of a 510(k) cleared vapour sterilization systems to terminally sterilize medical devices: K083097, http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfPMN/pmn.cfm?ID=29290
Example of a 510(k) cleared vapour sterilization systems to terminally sterilize medical devices: K071385, http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfPMN/pmn.cfm?ID=24889
Guideline for DisinfectionSterilization in Health Care Facilities, Center for Disease Control and Prevention. 2008. http://www.cdc.gov/ncidod/dhqp/pdf/guidelines/Disinfection_Nov_2008.pdf
Guidelines on Test Methods for Environmental Monitoring for Aspectic Dispensing Facilities, Produced by a Working Group Scottish Quality Assurance Specialist Interest Group. 2nd edit. 2004. http://www.astcp.scot.nhs.uk/QASIG/environmental%20monitoring.pdf
ISO 14937, Sterilization of health care products – General requirements for characterization of a sterilizing agent and the development, validation and routine control of a sterilization process for medical devices. 2009.
Isotron, Sterilisation services for medical devices and diagnostics. 2011. http://www.isotron.com/
Mohanty, S., Kougianos, E., Biosensors: a tutorial review. 2006. http://isites.harvard.edu/fs/docs/icb.topic860667.files/Biosensor_Tutorial.pdf
Process, Proven, Devices, Medical, DSP based telemetry system to communicate with implantable defibrillators and pacemakers. 2011. http://www.provenprocess.com/case_studies.html
Stichelbaut, F., Bol, J.-L., Cleland, M.R., Herer, A.S., Hubeau, J.P., Mullier, B. A high-performance X-ray system for medical device irradiation. Radiation Physics and Chemistry. 2006; 76(11–12):1775–1778. [International Meeting on Radiation Processing 2006].