17 Total Disc Transplantation: Current Results and Future Development
Jason Pui Yin Cheung, Dike Ruan, and Keith D.K. Luk
Total disc transplantation has been found in both animal and a preliminary human clinical trial to be successful in restoring biochemical and mechanical properties. Degeneration in the form of disc height reduction is still observed after allograft transplantation. Biological agents and nucleus pulposus (NP) cells may have a role in reducing allograft disc degeneration. A combination of cryoprotective agents with a slow cooling rate, limited incubation time, and liquid nitrogen for storage yields the best overall disc metabolic activity and elastic and viscous modulus. Future studies will utilize the goat model due to the poor availability of bipedal animal models. Future study direction includes fine-tuning of the graft harvesting and preservation procedures, understanding nutritional pathways, and novel imaging options for measuring nutrition transfer across the end plates.
Keywords: allograft, autograft, disc, intervertebral, transplantation
17.1 Historical Perspective
Similar to organ transplantation in other parts of the body, the concept of vertebral transplantation has been present for decades. The first reported transplantation of a functional spinal unit was in 1991.1 The authors transplanted a fresh frozen vertebra with adjacent intervertebral discs (IVDs) into a canine model and at 18 months, the discs survived and the vertebral body was revascularized. Mobility and stability of the spinal column were maintained in this study. The same group performed a follow-up study with a fresh autograft transplantation in a canine model.2 In contrast to previous findings, the transplanted discs had distorted morphology and metabolic function contributed by fixations that were too rigid and limited disc nutrition. Similar studies conducted by Katsuura and Hukuda,3 and Matsuzaki et al4 were performed on quadrupedal models but encountered the same problems with graft fixation.
Professor Keith D.K. Luk and investigators at the Department of Orthopaedics and Traumatology at the University of Hong Kong have been instrumental in experimenting with total disc transplantation. A similar idea regarding disc transplantation has been developed that also avoids too much constraint on the graft. Studies are based on Rhesus monkeys, which as bipedal animals, are closest to humans in terms of biomechanics. Series of research experiments conducted by the team were performed to verify the animal model and develop protocols in autografting, allografting, and fresh-frozen allografting. Further studies including processing and storage techniques were undertaken. A human clinical trial was conducted as well to test the efficacy and safety of this procedure. The following describes in detail each of these groundbreaking studies and also provides a glimpse of the future of disc transplantation based on current research themes and direction.
17.2 Bipedal Animal Model Experiments
The first autograft experiment5 was performed at the Tangdu Hospital, Affiliated Hospital of the 4th Military Medical University, Xian, China, in collaboration with Dr. Dike Ruan. Fourteen male Rhesus monkeys were used for this study because they were most similar to humans. Osteotomies were made 1.5 mm above and below each L3–4 disc without fracturing the end plates during retrieval. The graft was replaced back into the disc space and anchored to the outer annulus with thread sutures only. The animal was allowed to freely move without any fixation. Serial X-rays were used to measure the disc height and observe for any degeneration. Gradual loss of disc height was noted postoperatively with the lowest dip at 2 months (72.9% of the control disc height) but was stabilized with some recovery to 83.5% at 12-month final follow-up. Animals were euthanized at 2, 4, 6, and 12 months after surgery and the grafted discs were taken from two monkeys for biochemical and histological testing. The native discs that were not transplanted were used as controls. Autografts were retrieved from the animals and underwent biochemical, histological, and biomechanical testing. Due to the small sample size, analysis showed no significant differences between the grafted or control discs in terms of water, proteoglycan, and hydroxyproline contents. In the annulus fibrosus (AF), there was a steady drop in water content up to 6 months postoperatively, but with slight recovery up to 12 months. Hydroxyproline content in the AF was higher in operated discs at 2 to 4 months but was similar to control at 6- and 12-month follow-up. For the nucleus pulposus (NP), water content continued to decrease in both controls and operated discs with time. Both groups also had drops in proteoglycan content at 2 to 4 months postoperatively, but there was an increase at 6 months for the operated discs. Hydroxyproline content was higher in operated discs at all follow-ups.
Viable cells were seen in both the AF and NP on histology. Similarly, these were found in the cartilaginous end plates, and there were no signs of fissuring or cracks in the AF or abnormal vascular infiltration. NF had increased cellularity as compared with controls. Under electron microscopy, notochordal cells were found in the NP but were degenerated with loss of nucleus and abnormally shrunken in size. Mobility was seen across all discs during biomechanical testing in both flexion and extension in the early postoperative period. However, increased deformation was found in the operated discs which indicated increased mobility. At 6 months, loss of mobility was noted at the operated discs.
From this bipedal animal study, it was shown that autograft disc cells are able to recover their biochemical and biomechanical function despite transient ischemia. Viable chondrocytes and notochordal cells could still be observed after transplantation in the cartilaginous end plate and NP.
17.2.2 Allograft Experiment
In a natural follow-up to the experiments on disc autografting, allogenic transplantation was required to assess its applicability clinically because living or cadaveric donors are the main sources of transplanted organs. An allograft experiment was performed by switching the L3–4 discs between two monkeys. The age and size of the monkeys were similar and the operations were conducted by two surgical teams simultaneously to attempt standardization of the operating time and blood loss. Problems of subluxation and dislocation were encountered due to graft size mismatch; thus host and recipient size matching of the grafts with press-fit insertion is important to allow for allograft stability without fixation. The issue of immunogenicity was also raised from this study. No preoperative compatibility was performed for the animals and no immunosuppressants were used in this study.
17.2.3 Fresh Frozen Allograft6
As a feasibility study to resolve the issues of preservation, immunocompatibility, and size mismatching, further experimentation was performed with fresh-frozen allografting. Seventeen Rhesus monkeys were used in this study. T10–L7 discs were harvested from two monkeys via osteotomies made 1 to 2 mm above and below each disc. This included a vertebra and its adjacent end plates. For storage, grafts were measured and placed in a dimethyl sulphoxide (DMSO) solution and cooled stepwise to –196°C using liquid nitrogen. Three monkeys were used as controls and the rest of the animals were graft recipients. After removing the disc from the recipient, an appropriately sized graft was thawed and was placed snug-fit into the defect. No immunosuppressant was used in these animals. Bony union of the end plates was obtained successfully without graft subluxation or dislocation. The recipient monkeys were euthanized at 2 to 8 weeks, 6 months, and 24 months for analysis. Radiological, biochemical, and biomechanical testing was performed.
Unlike the autograft experiment, the disc height decreased progressively up to 24 months of follow-up down to 64.9% of control disc height. Secondary degenerative changes with traction osteophytes (▶ Fig. 17.1) were also observed. On pathology, disc degeneration was also evident with indistinct boundaries between the AF and NP. Chronic inflammatory changes were noted early and at 6 months only at the osteotomy site indicating that lymphocyte infiltration did not reach the grafted disc, and no immunoreactivity was noted in the cartilaginous end plate. However, the growth layer of the cartilaginous end plate in all specimens showed partial or complete disappearance and electron microscopy showed nuclear disc cell degeneration with irregular nuclei, karyopyknosis, mitochondrial swelling, and cytolysis. Steady decreases in water and proteoglycan content were also observed from postimplantation 6 to 24 months. For mechanical testing, similar mobility and stability was observed for both grafted and control spinal units.
Fig. 17.1 Progressive degenerative changes noted in the allograft disc on radiographs. (a) Prior to transplantation; (b) immediately after transplantation; (c) 6-month follow-up; (d) 12-month follow-up; (e) 22 month-follow-up. (Modified with permission from Spine.6)
This study confirmed cryopreserved allografts could retain their cell viability and mechanical properties despite degenerative changes. In addition, minimal immunoreaction was found only at the bone interface; thus the bony end plate should be washed prior to cryopreservation. The survivability of cells after cryopreservation requires further investigation especially after the freezing and drying procedures.7 Due to the loss of water content, cell shrinkage can damage cellular membranes and intracellular molecules. The optimal cryopreservation technique can allow maximum equilibration of water intra-and extracellularly. Refining the preservation protocol is necessary to maintain normal cellular function and delay the degeneration process. These steps are important for establishing a long-term allograft bank.
17.3 Clinical Trial8
With the success of the experimental studies, an ethics committee approved a small-scale clinical trial, which was initiated in 2000. One female and four male subjects received disc transplantations at the Navy General Hospital in Beijing, China. Four of the patients had cervical spondylotic myelopathy caused by single-level degenerative disc herniation, whereas one patient had a traumatic disc herniation with incomplete paraplegia. Four patients had surgery performed on C5–6 and one on C4–5. Grafts were obtained from three previously healthy young female trauma victims within 2 hours from their death. After retrieval, the grafts were placed in a similar solution of 10% DMSO and 10% calf serum (Gibco BRL; Invitrogen, Carlsbard, CA) and were stored at −196°C with liquid nitrogen. Prior to storage, all grafts were sized and these measurements were used in the transplantation for choosing the matched graft size. Peripheral blood and bone specimens were taken and underwent microbial culture and were screened for transmissible diseases including hepatitis B and C, tuberculosis, and human immunodeficiency virus (HIV). No other immunological matching was performed.
Complete discectomy and removal of the posterior longitudinal ligament were performed for cord decompression and preparation of recipient graft bed. Intraoperative measurement of the defect translated to the allograft size chosen for transplantation. The allograft discs were thawed in a 37°C water bath for 30 minutes prior to insertion. Similar to the animal experiments, no internal fixation was required. A neck collar was used as immobilization for up to 6 weeks postoperatively. No immunosuppressing agents were used. Interval flexion-extension radiographs and magnetic resonance imaging (MRI) were performed for assessment.
No complications of graft subluxation or dislocation, or immunoreaction were encountered. One graft was malpositioned too anteriorly but underwent complete remodeling (▶ Fig. 17.2) by the 5-year follow-up. Restoration of the AF and NP was observed indicating disc viability and ability to regenerate. All patients improved with transplantation. The Japanese Orthopaedic Association (JOA) score improved by an average of at least 3 points at the last follow-up. None of the patients experienced significant neck pain and only one patient had loss of disc height upon follow-up radiographs.
Fig. 17.2 Remodeling of the allograft disc with time. (a) Lateral view of the cervical spine showing a malpositioned C4–5 allograft immediately after surgery. (b) Lateral view of the same disc showing complete remodeling 61 months after surgery. (Modified with permission from the Lancet.8)
Similar to the animal studies, the mean disc height dropped from 5.88 mm to 4.33 mm at 5-year follow-up. Mobility arc of 7 to 11.3 degrees was observed in all but one patient. Spontaneous fusion occurred in the remaining patient and required a revision posterior foraminotomy procedure for residual radiculopathy. The overall arc of motion was reduced compared with the preoperative investigations, and it had also shifted more toward the extension arc. This indicated a reduction in the flexion range and an increase in the extension range. This could be contributed to disc space distraction during graft insertion or over-restoration of the disc space with extension of the spinal segment. The grafts were also placed slightly anterior with a space of 1 to 3 mm posteriorly to avoid recompression of the spinal cord, which might also have contributed to the change in motion arc. On MRI (▶ Fig. 17.3), two patients retained high T2 signal at the NP showing hydrated discs despite the possible cell damage triggered by the freezing process. The remaining three patients with dark discs showed no deterioration or worsening compared with the adjacent levels. No evidence of accelerated degeneration was observed in the other levels.
Fig. 17.3 Maintenance of magnetic resonance imaging (MRI) T2 signal with time. (a) Sagittal T2-weighted MRI preoperatively showing a C5–6 disc herniation with cord compression. (b) Sagittal T2-weighted MRI immediately after implantation showing a similar nucleus pulposus (NP) T2 signal as the adjacent C6–7 disc. (c) Sagittal T2-weighted MRI 14 months after surgery showing narrowing of the allograft disc but similar T2 signal as the adjacent C6–7 disc and more hyperintense than the other levels. (d) Sagittal T2-weighted MRI 68 months after surgery showing narrowing of disc height, but the T2 signal in the NP remains similar to C6–7 and more hyperintense than the other levels. (Modified with permission from the Lancet.8)