The expanders had separate filling ports similar to the Hickman catheter injection "button". Some of these implants were fitted with additional accessories which required special surgical procedures. Later, devices with integrated filling ports were marketed as a fundamentally different concept. The products were sold by McGhan Corporation and had a built-in self-sealing portion of the shell which acted as a "port" (Magna-Site™). This was a magnetic locating ring placed on the flank of the implant and a central portion designed to accept direct filling. Filling took place through perforation of its septum by an hypodermic syringe. The device was accident prone in spite of its locating accessory and perforations generally occurred early.

The most common tissue expansion devices used separate filling tubes linked to remote filling ports. The manufacturing technology was similar to conventional mammary prostheses. However, the devices were more fragile and subject to damage because of the size of the items and the requirement for handling and extracting separate ports. The need to perform transcutaneous post-explantation injections of fluid with needles and cannulae greatly increased the risks for shell perforations.

The port was made removable in some designs by incorporating a sliding joint at the connecting point to the prostheses (convertible implants); a typical example includes the Becker Expander made initially by the American Heyer Schulte Corporation and produced in slightly different forms by the Mentor Corporation. Cox Uphoff International later produced other variations during the Inamed ownership period. These devices have filling ports with long extension tubes which are linked to the main prosthesis through a slip joint. This extractable silicone conduit could be removed ideally with simple traction and the port at the shell was expected to reseal.

In the late-eighties, multi-compartment expanders or "stacked" expanders made from two and sometimes three conventional expanders bonded together by adhesive were marketed by Dow Corning and by a few other manufacturers. Several of these devices involved fabric components and complex assemblies of fixation appendages. A few required multiple filling ports. As the devices increased in complexity, their failure susceptibility and their scope for injury increased. Eventually in the early-nineties, many of these devices ceased to be sold and instead multiple implantation of expanders into the same surgical site became an alternative.

These devices were virtually impossible to keep sterile. The ports and the connecting points frequently leaked after several attempts at refilling. Shell and valve failure are also common occurrences because of the complexity of the shell assembly and the stresses that tend to concentrate around its accessories.

The Radovan-Style of Tissue Expander:

The Radovan tissue expanders appeared in the late-seventies. Several versions were made by different manufacturers. All were based on separate ports with simple saline-filled shells. Originally developed under the American Heyer Schulte ownership period, the devices appeared in widespread clinical usage circa 1983 and were continued in different versions during the subsequent Mentor Corporation ownership period. The Medical Engineering Corporation produced several versions under different names such as Surgitek T-Span™ and Creat™.

The Radovan variants are fundamentally modified saline-inflatable implants coupled "permanently" to a percutaneous injection port via a small bore tube. The saline-filled Radovans are primarily intended for temporary use. They are also promoted as permanent implants or as convertible implants which, upon implementation of a simple surgical intervention, would be freed of the filling appendage. The items are failure-prone because of the complexity of the shell, the generally large size of the item and the stresses imposed on the main shell by the filling tube accessories. Convertible versions are even worse as they require a comparatively destructive procedure for extraction of the filling tube.

Expanders of this design are extremely difficult to keep sterile and liquid-tight. The ports and the connecting sleeves frequently leak after several attempts at refilling, in particular when elevated pressure is required to break or stretch thick fibrotic tissue. Rupture and fatigue failures of shells are common because the items are habitually underfilled leading to multiple sharp pleats and puckering of the shell. Connecting points and seams are associated with stresses and failure sites. Deflation with slow release of contaminated aqueous filling solutions exacerbate capsule problems and create new difficulties associated with colonization of the capsule space. Most habitually leak fluid and micro-organisms into the intracapsular space from faulty valve/ports and connecting accessories.

The complexity of the shell assembly and the stresses that tend to concentrate around the filling tubes make them particularly vulnerable. Body fluids and tissue also ingress into the shell and contamination/infection of the surgical area is habitually found within one or two years of implantation.

The insertion of ill-conceived or defective implants in a disease prone area has predictable morbidity and a potential for injury which rises with the dwell time of the product. It is well established that the implants cause major structural, physiological and biochemical changes in the breast environment. They also act as time release systems for pharmacologically active compounds. Mechanisms by which such devices can affect the patient's health, appearance and comfort are outlined below.

(1) "Normal" surgical trauma and surgical misadventures resulting in damage to functional/sensorial parts of the chest and the upper limbs;

(2) Biomechanical effects induced by the presence of large foreign objects that cause compressive trauma, excoriation, distention, atrophy and restrictive adhesion of tissues or compressive/occlusive ischemia of the vasculature within the pectoral-axillary area;

(3) Locally injurious biochemical effects from reactive dispersible substances that induce fibrotic, inflammatory or destructive tissue changes;

(4) Long term tissue remodelling and deviant repair processes leading to hyperplasia, densification, mineralization and dehydration of the implant site;

(5) Implant-capsule-oil adjuvant interactions leading to tissue degeneration or denaturation to produce host tissue-derived antigens that elicit antagonistic host-directed antibodies;

(6) Pathologic effects from bacterial, viral or fungal colonization of the capsule space leading to low grade chronic infections and toxic phenomena from microbiological metabolites.

For long term users, all of these effects are present concurrently and the severity increases with use, in particular for users of tissue expanders. The early occurrence of intracapsular infection, seromas/hematomas are strong promoting factors for implant adverse reactions.

Users of tissue expanders are exposed to risks from aqueous media implants in addition to new ones including periprosthetic infections initiated through repeated perforations of the skin and the capsule wall. The filling port must be repeatedly breached during filling. Some tissue expanders such as the Becker have compartments filled with silicone gel and embody risks from gel leakage. They habitually leak releasing the substance into the capsule space. Transcutaneous filling procedures virtually guarantee transmission of skin flora to the periprosthetic space and the saline compartment. The implants offer a nearly ideal environment for micro-organisms. Infective sequelae are widespread and difficult to control. Early removal and debridement are mandatory. Intracapsular infections and chronic colonization enhance capsular fibrosis. Micro-organisms in saline compartments resist antibiotics. Formation of pharmacologically significant quantities of toxins are associated with incubation processes within the devices. Calcification takes place following necrosis of tissue after prolonged usage. Calcification of this kind is unlike naturally occurring "dystrophic" calcification. It is a more severe and injurious phenomenon which culminates in formation of sharp, abrasive structures which traumatize the periprosthetic tissue and eventually erode the shell of the prosthesis leading to dispersion of fine silica-containing elastomer, gel extravasation and additional complications.

General Considerations:

A noteworthy characteristic of saline-filled tissue expanders is that they share the problems of saline inflatable prostheses in addition to other problems associated with the increased complexity of the items because of the need to increase volume post-operatively. Their ability to sustain microbiological activity within the saline-containing compartment is a dominant negative aspect of their use. Another is their susceptibility to rapidly lose their saline charge because of incompetent valves or adventitious valve failure and/or shell deterioration culminating in leakage. Such characteristics have long term safety implications. Atypical infections and their sequelae account for a significant part of the morbidity. The diagnostic of emerging problems and the clinical management of sequelae are difficult. The area requires additional investigation to establish the optimum treatment options.

Design Issues for Saline-Containing Devices:

Saline mammary implants were introduced commercially in the mid-sixties. Early models were simple and robust. All were sold non-sterile. They required individual hospital sterilization a well as meticulously clean intra-operative procedures. Many gave excellent service for more than two decades and a few are still in use in original patients. As demand grew, the quality diminished and many unsatisfactory designs were introduced. Most were promoted on the basis of lower prices, faster surgical implantation and more gratifying immediate appearance. By the mid-eighties, the quality and performance had degraded to the point where seasoned clinicians avoided their use. Saline-filled implants were deflating after several months and most perforated within 1-2 years following implantation. Expanders shared the same problems.

Many expanders were explanted in grossly contaminated states with viable and non-viable entities, sometimes visible to the naked eye. Intraoperative failure and early leakage during attempted pressurization convinced many clinicians to abandon the products in the belief that problems were intractable. They resorted to the use of 'permanent' implants and periodically changed the devices to progressively larger sizes.

Gel-Containing Tissue Expanders - The Becker:

The double-lumen (gel-saline) convertible or "combination tissue expanders" were introduced in the late-seventies without supportable technical or clinical rationale. They are rarely recovered with unperforated saline compartments and account for an above average failure rate of the gel compartments. Wall failure of the outer gel compartment of such devices does not lead to grossly visible deflation. The saline-filled core does not immediately deflate and thus limits the change in volume of the breast. In the absence of severe adverse reaction or grossly unsatisfactory cosmetic results, users unknowingly retain the perforated devices.

Risk Factors for Saline-Containing Systems:

Virtually all devices have been sold as sterile products since the late-seventies but saline chambers with small amounts of viable micro-organisms have been found at various times. They may be explained simply as improperly sterilized products. This problem beset the industry for many years and forced the introduction of specialized sterilization technology in the eighties. Because of the closed internal configuration of expanders, the validity of current sterilization techniques is still a contentious issue in some circles.

Non-sterile compartments may reflect an improperly sterilized device. Post manufacturing contamination can also account for lack of sterility. It can take the form of surgically introduced pathogens during in-situ filling as opposed to sterile intraoperative filling. Given the non-sterile character of the mammary gland, the possibility of introducing viable entities from contact with extracellular fluids and breast tissue is not negligible.

Additional pitfalls include resterilization attempts following unsanctioned device reuse or habitual filling with non-sterile substances. Deviations from established filling recommendations were commonplace in the past. Some of these procedures have been published and were once very popular. A few remain in use by some surgeons to this day. Such procedures typically include filling with non-parenteral grades of electrolytes or colloids and extemporaneously prepared solutions of oral pharmaceuticals such as anti-infectives and anti-inflammatory steroids. The long term fate of these degraded mixtures in a closed environment is also worthy of consideration as nutrients to late-arriving micro-organisms and in the context of chemically-mediated adverse reactions after shell rupture.

An inoculated prosthesis may remain in a biologically dormant state until the viable entities are provided with nutrients. If the filling substance contains no nutrient, no biological activity will take place. Alternately, the inocula may die spontaneously. However, on balance of probability, some proteinaceous matter will eventually enter the compartment through valves, shell defects or incidental to post-operative filling. Paradoxically, very few valve and filling port designs reflect concern about this issue. Few are secure even when new and definitive sealing of entry ports is rarely used, even when available.

Most valves and filling ports of the eighties are variants of unidirectional flow control devices or drug infusion access ports. They were developed for hydrocephalus drainage shunts, administration of anti-tumor drugs and non-medical applications. They allow inward flow and frequently will release parts of their charge as pressure is increased beyond the capacity of valves and couplings. Perforated ports also leak at certain threshold pressures when incorrect types of needles are used. There is therefore no reliable means of preventing nutrients from reaching an inoculum present in these systems.

The Prosthesis as a Fermentation Reactor:

Inoculated devices with nutrients may not present an overt risk until the biological processes involve sufficient mass to produce large amounts of pathogenic material. In the early post-surgical period, the processes may be no more than survival of the most hardy entities and may not be sufficient to cause problems. However, with faulty valves/port systems and late perforation and development of porosity in the shell, conditions for colonization of the compartment will be met. Even if sterile at the onset, leaky saline compartments constitute nearly idealized incubation zones for adventitious pathogens that can find their way in the capsular space.

Adjuvant-like impurities from leaky gel compartments further complicate the chemistry of these pockets. They add to the denatured proteins, decaying tissue, fermenting pharmaceuticals and active micro-organisms. Scar tissue may temporarily occlude perforations thus further improving the environment for less competitive micro-organisms.

The cul-de-sac or "blind pocket" geometry of the area forbids irrigation by physiologic fluids or administered antibiotics. The limited oxygen permeability of the silicone wall favors anaerobic processes. Egress of this contamination eventually takes place in response to movement and pressure applied to the breast area. Dispersion of this material may periodically invade the prosthetic capsular space and may establish secondary colonies outside of the prosthesis. Symptoms of infection should appear at this stage but may resolve temporarily.

Typical Sequence of Events:

The adverse affects in saline solution filled mammary implants usually take place according to the following sequence: (1) Prostheses sometime containing viable micro-organisms within the saline compartment are filled according to a procedure which further contaminate the solution with new organisms and nutrient. (2) Colonization of the interior and the valves of the prostheses by bacteria, mycobacteria and fungi takes place slowly . (3) Pannus tissue grows into the unprotected valve and allows more nutrient into the chamber and egress of some of the colonizing organisms. (4) Silent infection sets in the periprosthetic capsule space. (5) Under stimulus of natural chest movement, breast compression (closed capsulotomy) and/or tissue contracture, the colonizing organisms are redistributed around the prosthesis and secondary colonization takes place. (6) Recurring fleeting episodes of infection, discomfort , fever and swelling are noted; antibiotic medication may be prescribed by treating physicians yet the symptoms recur. (7) Early displacement and "hardening" of the prostheses may take place as the intracapsular colonization progresses to fibrosis and capsular contracture become obvious. (8) There may be attempts at breaking the capsule via compression (capsulotomy) but there is only temporary relief. (9) the outer lumen then perforates or ruptures providing pressure relief but fevers and discomfort worsens as the colonized interior is discharged into the intracapsular space. (10) Additional procedures may be performed with limited success until the devices are finally explanted and the site is disinfected.

Clinical Impact of Implant Colonization:

Valve and saline compartment (outer lumen) contamination by micro-organisms, scar tissue and decaying proteinaceous debris, as well as colonization of the interior of the prostheses by atypical flora, is a major long term safety issue for aqueous media filled devices because the durability of the shell is very limited and release of the content is inevitable.

The eventual loss of the fluid containment system because of shell perforation and valve leakage allows contaminated fluids to egress from the prosthesis to the capsule space and thus reach the patient. The very long incubation periods and the unusual growth environment favor the presence of atypical flora and micro-organisms that are not commonly encountered in clinical practice and about which there is limited treatment data.

If secondary colonies become established outside of the periprosthetic space, for example in the proximal lymph nodes, the problems may persist even with the complete removal of the offending devices and their tissue capsules. Additional surgery or systemic treatments may be indicated if the identity of the flora is known or suspected through biopsy or antibody assays.

Clinical Management Strategy:

Patients suffering adverse reactions from prostheses incorporating saline and multi-compartment, aqueous solution-filled cores and perforable transcutaneous filling ports may present with problems different from individuals affected by direct inert oil injections and/or silicone gel implant misadventures or other instances that lead to systemic dissemination of oils such as paraffins and silicones. Expander users often require novel diagnostic and infection control methods that, in addition to seeking the usual nosocomial pathogens, also address anaerobes, mycobacteria, fungi and the sequelae of long term infections in spaces protected by scar tissue membranes.

Individuals with chronic exposure to prosthetic debris, denatured proteins and toxins from atypical micro-organisms that thrive in poorly irrigated environments such as periprosthetic scar tissue, granulomas and occluded lymph structures may also have immunologic and neurologic implication worthy of re-examination.