J Manipulative Physiol Ther 2009 (May); 32 (4): 294–302 ~ FULL TEXT
Charles N.R. Henderson, DC, PhD, Gregory D. Cramer, DC, PhD,
Qiang Zhang, MD, DC, James W. DeVocht, DC, PhD,
Randall S. Sozio, BS, LATG, Jaeson T. Fournier, DC, MPH
Palmer Center for Chiropractic Research,
Davenport, Iowa 52803, USA.
OBJECTIVE: This is the last article in a series of 3 articles introducing a new animal model, the external link model (ELM), that permits reversible, nontraumatic control of the cardinal biomechanical features of the subluxation: fixation and misalignment. A detailed description of current ELM procedures is presented and practical issues are reviewed such as expense (dollars and time) and construct failure rates during and after the surgical implant procedure.
METHODS: Descriptive report of current ELM procedures, refinements to the spinous attachment units (SAUs), and tabulation of costs and failure rates drawn from recent studies.
RESULTS: In contrast to the older, 1–piece stainless steel SAUs, new 3–piece titanium SAUs may be reimplanted many times without failure. Consequently, the cost per ELM ranges from $579 to $69, depending on whether the SAUs, links, and screws must be purchased or are already available for implanting. The SAU implant procedure requires between 0.5 and 1.25 hours, depending on the experience of the surgeon. The total construct failure rate for the ELM is 24.2% (6.6% at surgery failure + 17.8% postsurgery failures). This rate is consistent with that reported in spine implant studies with other devices. To date, more than 500 male Sprague-Dawley rats (350–450 g) have been implanted with SAUs for ELM studies at the Palmer Center for Chiropractic Research and the National University of Health Sciences.
CONCLUSION: It has been our experience that individuals with basic animal research training will become proficient at producing the ELM after observing 3 to 4 implant procedures and performing 4 to 6 procedures on their own.
From the FULL TEXT Article
In this last article of the series, we have provided a detailed description of the current implant procedure, listed required materials, and addressed model practicality issues such as cost, procedure time, and failure rate. We believe the ELM is a promising tool for examining subluxation theory. However, to be adopted by the greater research community as a research tool, the ELM must also be practical in terms of cost, procedure time, and failure rate. This consideration is particularly important for investigators at complementary and alternative medicine (CAM) institutions that may have especially limited research resources. Reproducibility of the model has been shown over the past 5 years by successfully replicating the model at 2 sites, Palmer Center for Chiropractic Research (PCCR) and National University of Health Sciences (NUHS).
Cost and Procedure Time
The cost per ELM ranges from $579 to $69 each, depending on whether the SAUs, links, and screws must be purchased or are already available for use. For any given study, the husbandry costs (feed, bedding, and facility personnel) will depend on the length of the animal survival period. At the PCCR and NUHS facilities, rat husbandry expenses averaged $100 per rat ($0.40/d × 250 days average survival in most ELM studies).
It has been our experience that individuals with basic animal research training will become proficient at producing the ELM after observing 3 to 4 implant procedures and performing 4 to 6 procedures on their own. After the new procedure becomes routine, it takes from 0.5 to 1.25 hours to complete a 3–SAU implant (initial incision to closure).
Failure rates were determined from a study of 228 male Sprague-Dawley rats (400–475 g). Failures occurring during the SAU implant surgery were minimal (6.6% of the 228 SAU implant procedures performed; Table 2). However, postsurgical failures due to loosening or breaking of the SAU/spinous construct and aseptic bone resorption at the SAU-spinous interface contributed an additional 17.8% (Table 3). This postsurgical failure rate is considerable, but it is consistent with that reported in other instrument-assisted spine fixation studies (human studies: Connolly et al, 22%; Cook et al, 17%; Dickson et al, 16%; McAfee et al, 20%; and Wetzel et al, 22%; animal studies: Foster et al, 29%; Erulkar et al, 20%; Boden et al, 18%). [11–18] In their surgical model of spine fixation (subluxation), DeBoer and McKnight  reported construct failures in excess of 50%. The total failure rate for the ELM reported here is 24.2% (6.6% at surgery failure + 17.8% postsurgery failures).
The most common cause for revision surgery after either peripheral or spinal implant surgery in humans and animals is aseptic loosening with periprosthetic osteolysis. [20–21] Aseptic bone loss at the implant-bone interface is thought to have several contributory causes. Micromovement (1–100 µm) may develop between component parts of the implant or at the implant/bone interface. Movement between component parts of an implanted device produces particulate debris through a process known as “fretting corrosion.” Fretting corrosion occurs when cyclic movement between metallic parts of an implanted device produces repeated fracture and reformation of the protective oxide layer (passivation layer) on its contiguous surfaces. Metallic debris particles are produced and subsequently phagocytized by local tissue macrophages. These macrophages release chemical mediators that initiate a complex local inflammatory reaction and induce a widening zone of soft tissue damage, bone resorption, and inflammation around the implanted device. Implant movement also induces interstitial fluid pressure fluctuations that can stimulate osteocytes to die via apoptosis or release cytokines that modulate osteoclast and osteoblast activities. [9, 22, 23, 24] Ever present movement in biological systems makes aseptic bone resorption an ongoing concern when devices must be implanted in either humans or animals. Finally, recent studies have shown that barely detectible endotoxin residues left behind by killed bacteria on implanted devices can induce a localized inflammatory response with resulting bone resorption at the implant-bone interface.  Aseptic bone resorption therefore has a multifactorial cause that must be considered to reduce the postsurgical failure rate in the ELM.
Spinous Attachment Unit Modifications
The saddle of the new 3–piece SAU design features one fixed leg and one free leg, providing a more uniformly distributed compression grip on the spinous process and eliminating the stress fatigue problem associated with the living hinge of the 1–piece design (Fig 1). Also, the independent movement of the free saddle leg reduces the likelihood that the SAU legs may be compressed asymmetrically on the spinous process, which has sometimes caused the stem of the implanted 1–piece SAU to list to one side.
The new 3–piece titanium SAUs offer additional technical advantages. Spinous attachment units with stems that may be removed from the saddles reduce the likelihood of spinous fractures occurring due to contact between the SAU stems and the animal's cage during the post-link period. Moreover, removable stems allow investigators to design and use various procedure-specific stems. The removable stem of the 3–piece design also allows delayed exteriorization of the stem until after the postsurgical recovery period. The older 1–piece design required that the nonremovable stem be passed through the skin just before closing the skin incision at the conclusion of the surgical procedure. The consequence of this older design was that the skin would tug on the SAU stems as the rat moved around in its cage during the postsurgical recovery period and the stems were also vulnerable to striking cage structures during this sensitive time. Many investigators have stressed the importance of a motion-free, early postsurgical period to establish a stable implant. [9, 21,25, 26, 27]
Consequently, we have delayed exteriorization of the SAU saddle and stem attachment until after the postsurgical recovery period. This should reduce potentially disruptive stresses and enhance interface integration of bone and implant.
Titanium has become the material of choice for metal implants because it is twice as strong as steel with only half the weight and is also more biocompatible. [28, 29, 30]
Unlike steel implants, which the body envelopes with a thin fibrous capsule, bone will attach directly to titanium implants, a process known as “osseointegration.”  The stability of any bone implant is determined largely by the competing activities of osteoclasts and osteoblasts that are constantly remodeling living bone.  Bone resorption at the implant site will weaken the SAU/spinous construct, whereas bone growth around the saddle legs can strengthen it. The SAU saddle screw torque applied at surgery substantially affects bone remodeling associated with the implant. If the SAU saddle is tightened too much, it may cause pressure atrophy of the underlying bone. [9, 22, 23]
By contrast, if the SAU saddle is loose, implant movement can stimulate osteoclast activity with subsequent bone resorption and implant failure. Moreover, although bone deposition around the SAU saddle legs may stabilize the implanted device, it may become excessive and impede normal joint motion. Therefore, both bone resorption and excessive reactive hyperostosis must be avoided.
In the new 3–piece SAU, holes have been drilled through the SAU saddle legs to permit bone ingrowth that will increase the stability of the implant. To permit bone ingrowth into and around the SAU legs, the animal must have an adequate postsurgery recovery period. In the original ELM protocols, we allowed a 1–week postsurgery recovery period. But, over time, the length of the recovery period has progressively been increased, such that we now use a 6–week postsurgery recovery period.
In response to studies reporting implant failure due to trace endotoxin residues,  we recently modified our SAU cleaning procedure to minimize residual bacterial endotoxins. Spinous attachment unit implant components are not allowed to dry after removal from the spine. They are disassembled, brushed clean of large debris, and placed in a sonic bath containing a 0.525% aqueous NaOCl solution (diluted household bleach) for 15 minutes. The components are then sonicated in 70% aqueous ethanol for 15 minutes and rinsed 3 times in deionized water. After air drying in a clean enclosure, all SAU implant components are stored in a sterile envelope and autoclaved before use.
In pleasant contrast to the 1–piece stainless steel SAUs, the new 3–piece titanium SAUs may be reimplanted many times without device failure. To date, we have reimplanted 60 of these 3–piece SAUs a total of 3 times each with no sign of failure (unpublished data). We are confident that the new 3–piece titanium SAUs may be implanted numerous times without device failure.
Finally, we now use a subcutaneously implanted electronic chip identification system (Avid ID Systems Power Tracker III, Avid Identification Systems, Inc, Norco, Calif) to ensure integrity of group assignments for all rats in these long-survival studies.
External Link Model Applications and Limitations
In the first article of this series, we noted that:
[c]hiropractic researchers have long appreciated the need for an animal model of subluxation because such a model would permit evaluation of the effects predicted by the widely held conceptual model of this lesion.… Because the ELM is a long-term survival model, with the rats studied for many months after experimental spine fixation, we believe that it may be used to study the putative chronic effects of subluxation as well as the effects of various therapeutic interventions.
These statements underscore the promise of the ELM as a research tool for directly examining and developing theories related to the subluxation, a foundational concept for much of chiropractic practice. Recent ELM publications have, in fact, reported degenerative spine changes as well as apparent neuroplastic changes in the central nervous system that are predicted by current subluxation theory. [1, 3, 8, 32]
However, the limitations of the ELM as a research tool must also be acknowledged. One limitation imposed by using rats is that the rat spine does not permit ready fixation of the thoracic or cervical regions because the spinous processes in these areas are too small for attachment of our implant devices, the SAUs. This limitation may be overcome by future application of the model in larger animals such as sheep or miniature pigs. Challenges presented by the anatomic and biomechanical differences between biped and quadrupeds, and the experiments currently underway to address these issues, were addressed in the first 2 articles in this series. [2, 33]
Moreover, in the second article, we stressed that further study is needed to determine whether the experimental hypomobility and misalignment observed in the ELM have a homologue in either the rat or the human population.
The ELM may be used to examine theoretical linkages between chronic spine fixation and its biological outcomes (eg, spine degeneration, somatic and visceral neural function, behavioral changes), as well as the putative benefit of spinal manipulation to restore normal spine mobility and, consequently, normalize these biological outcomes. Although validation of the various mechanisms that underpin subluxation theory strengthens the logical argument behind the theory, it does not establish subluxation as a mechanical entity that exists in either rats or people. Claims for the reality of this lesion currently rest on “mosaic evidence,”  whereby advocates selectively collect scientific studies and cobble them together to fit an established perspective. The body of evidence is thus tailored to the perspective rather than to reconciling the perspective with the full body of evidence. To establish subluxation as more than a theoretically plausible biological entity, measurable characteristics must be identified and observed in the general population (rats or humans). The ELM provides a means to identify such potential characteristics.
We believe that this 3–article series has shown the value, use, and practicality of the ELM. Essential equipment, supplies, and procedures are fully described. It has been our experience that individuals with basic animal research training will become proficient at producing the ELM after observing 3 to 4 implant procedures and performing 4 to 6 procedures on their own. We hope that this information will encourage and assist other researchers to use the ELM in their studies of spine fixation and spinal manipulation.
The great promise of basic chiropractic “subluxation” research is that it will clarify for clinical researchers the mechanisms by which spine fixation or malposition may cause harm and show or suggest effective therapeutic remedies. Answers are needed to pressing and fundamental questions such as: Does chiropractic subluxation actually occur? If so, does chiropractic spinal subluxation significantly threaten a patient's health? Are there features that will allow researchers and clinicians to determine its accurate and precise location as well as its specific nature? Can spinal manipulative therapy prevent, stop the progression, or reverse adverse health effects related to chiropractic subluxation? Are there “time windows” that might influence the outcome of treatment? When these questions are answered, clinicians will be able to more objectively match the unique features of a patient's presentation to the diversity of chiropractic techniques, treatment frequency, number of visits, and treatment duration.