GEMINI-supported spinal cord transplantation for the treatment of chronic spinal paralysis: Overview and initial clinical translation

Introduction

There is at present no biological cure for chronic paralysis following spinal cord injury (SCI). A medley of experimental approaches have been tried over the years, and, despite some improvement in some patients, no full recovery has been observed ( ). As underscored by Illis,

“it would be difficult to find any other branch of science with over a century of such sterile endeavour. In effect, there has been repetition of the same idea, albeit with different techniques, that is, looking at the lesion site. Are we sentenced to repeating the same experiments in the hope of expecting a different result?” ( ).

Work conducted mainly by US neurosurgeon L. Walter Freeman more than half a century ago suggests that a permanent, biological cure is possible, at least in several cases of chronic SCI, by removing en bloc the “ lesion site ,” i.e., the most damaged portion of the spinal cord, and connecting the two free undamaged ends, after spinal (vertebral) shortening ( ; see also ). Transplantation of a healthy segment of cord is an alternative (see below).

In both scenarios, a technology that allows the functional reconnection of the severed ends is necessary. In 2013, the GEMINI spinal cord fusion (SCF) protocol was proposed with this goal in mind ( ).

Here, we will briefly review the GEMINI SCF protocol and lay out the various alternatives for its ongoing deployment in the clinical setting.

The GEMINI spinal cord fusion protocol: Overview

The GEMINI spinal cord fusion protocol is predicated on four pillars ( ; ; ):

1.

Sharp severance of the spinal cord with an ultra-thin blade, with its attendant minimal tissue damage, as compared to clinical SCI (< 10 N vs 26,000 N)
2.

Exploitation of the gray matter internuncial sensori-motor “highway” (so-called cortico-trunco-reticulo-propiospinal pathway; C-TRPS) re-bridged by sprouting connections between the two re-apposed cord stumps. This “highway” runs in parallel to corticospinal descending fibers and is co-responsible for sensori-motor transmission through the cord (see also ; ).
3.

Application of “fusogens/sealants”: these “seal” the thin layer of injured cells in the gray matter, both neuronal, glial and vascular; simultaneously they fuse a certain number of axons in the white matter. The pro-regenerative early scar is not inhibited, but necrosis/cavitation is.
4.

Acceleration of axonal outgrowth by electrical spinal cord stimulation straddling the fusion point and concomitant motor cortex stimulation.

We will briefly review the technology involved. An in-depth discussion can be found elsewhere ( ; ).

GEMINI: Fusogens

Cell fusion is the process by which the membranes and intracellular contents of any two cells are permanently fused. There are multiple situations where this occurs naturally—fertilization is one very fundamental example, but there is also increasing evidence that cell fusion can be induced for clinical benefit. The most striking clinical application of this is the potential for reconstitution of axon integrity and electrical conductivity through the fusion of axon membranes; this can be performed in order to reconnect the two cut ends of a severed axon, or to create membranous continuity between the ends of two axons that have been newly placed in apposition (such as in a allogeneic transplant). While close spatial proximity is required, and usually achieved through some sort of physical (often micro-surgical) manipulation, the scale at which this occurs is small and relies on a chemical manipulation to complete the fusion at the level of the cellular membrane. The substances that are capable of performing this function are heterogeneous in composition and function, and often referred to as “fusogens” ( ).

Axon membrane fusion is valuable in any attempt to manage nerve injury because it allows for the possibility of immediate reconstitution of electrical conductivity and therefore nerve function. An outcome on such a timeline is impossible given the current standard methods for managing nerve injury (approximation via epineurial repair), which accept distal Wallerian degeneration and the need to wait for axon regeneration from the point of injury to the end organ (which is traditionally thought to progress at the rate of approximately 1 mm per day). Therefore, fusogens represent a quantum leap in the way that nerves are managed, as they allow for the possibility of true axon repair , as opposed to the traditional methods, which at best harness axon regeneration (and accordingly the outcomes after this treatment approach are partial at best, and measured on the order of months to years).

There are two general mechanisms by which fusogens are thought to function: cell aggregation and membrane modification ( ). Substances that function only via cell aggregation lead to increasingly closer apposition of the two axon ends (often by extraction of intervening water molecules). This group includes chitosan, dextran sulfate, small organic molecules, and lipids. Substances that function only via membrane modification alter membrane charges to create a more favorable electrochemical environment for apposition and ultimately fusion to occur. This group includes cations such as Ca ^{2 +} and Mg ^{2 +} , sodium nitrate, and H-alpha-7. Polyethylene glycol (PEG) is a substance that is thought to function by a combination of both mechanisms, and for that reason is it is the fusogen that, to date, has been used most widely and most successfully in both research and clinical settings.

PEG is a hydrophilic polymer that is inexpensive and commonly encountered, as it has proven to play a role in many different medical, biological, and industrial scenarios. Its potential utility as a means of promoting cellular fusion, however, was first recognized in the 1970s after it was shown to be able to induce immortalized hybridoma cells and fused mammalian erythrocytes ( ). The ability to induce axon membrane fusion (either in repair of cut ends of the same axon, or the fusion to ends of a different axons) was first realized and progressively tested beginning in the 1980s, starting with invertebrate models ( ), and gradually progressing into rodents and then large mammals (fully reviewed in ). It has successfully restored complete digital nerve function after traumatic injury in small human studies ( ; see also ) and has been effective at restoring neurologic function at the level of the spinal cord in multiple animal models ( ; fully reviewed in ).

In addition to the cell fusion mechanisms of cell aggregation and membrane modification, PEG has been shown to have other effects that are beneficial for maximizing nerve function. First, it is neuroprotective due to its ability to inhibit the formation of the mitochondrial permeability transition pore; second, it may reduce the neuronal membrane tension and thereby improve membrane fluidity; third, it causes the release of neurotrophin-3, which attracts neural stem cells and promotes their differentiation into axons. However, PEG does not lead to the fusion or repair of connective tissue and therefore does not provide any structural strength and in very rare cases may be associated with allergic reactions (see references in ).

GEMINI: Electrical stimulation

The GEMINI SCF protocol includes the positioning of an extradural stimulating 16-contact paddle straddling the point(s) of fusion followed by continuous stimulation (15–60 Hz, 5–9 V) during the entire rehabilitation period ( ; ; ). Spinal cord stimulation (SCS) is paired to non-invasive transcranial magnetic stimulation (TMS) of the motor cortex (M1) ( ).

Electrical stimulation (ES) serves two primary purposes:

1.

Promotion/Acceleration of axonal outgrowth and plasticity across the fusion interface. Shirres first emphasized the ability of electricity to stimulate spinal cord regeneration (see full details in ). ES has been shown to accelerate axonal outgrowth (neural sprouting) at sites of injury, including the cord, in animal studies (e.g., ; see also , ). Of note, ES is co-additive to PEG in SCI ( ).
2.

Activation/Facilitation of propriospinal neurons caudal to the fusion interface.

Following complete spinal transection in man, the neuronal networks responsible for locomotion (i.e., central pattern generators, CPG) residing in the spinal cord are intact, but fail to produce limb movements: the excitability of the spinal cord is too depressed to enable the coordinated recruitment of motor neuron pools. Consequently, enabling robust levels of activity during rehabilitation is critical to steer activity-dependent plasticity in the trained circuitry. This is achieved by ES. ES replaces missing sources of excitation to reactivate spinal circuits, and thus enable motor control. In particular, following chemical fusion, sprouting across the fusional interface restarts the flow of electrical transmission between the fused cords. ES amplifies voluntary commands from the brain and thus supports their propagation to the CPGs at cervical and lumbar levels ( ; ; ). In turn, these integrate supraspinal and sensory information into the execution of purposeful movements.

Over the past 20 years, on the heels of Dimitrijevic’s pioneering work in the last decades of the XX century ( ), multiple independent laboratories have shown that the delivery of continuous electrical stimulation (tonic) over the lumbar spinal cord immediately reestablishes intentional control over the activity of previously paralyzed leg muscles, even more than a decade after the occurrence of the SCI. Continuous ES also restores full weight-bearing standing and facilitates stepping (see reviews in ; ; ; ; see also chapters in this volume). ES of the spinal cord (SCS) is known to elicit activity locally and rostrally up to the cortex (as shown in chronic pain: see , pp. 465–474).

Spatiotemporal SCS protocols have been touted as an improvement over continuous stimulation (reviewed in ). However, no trials exist that compare continuous with spatiotemporal stimulation in homogeneous patients. Serotonergic agonists may further potentiate the effects of ES ( ; ).

ES can be employed in conjunction with specialized multidirectional robotic body weight support systems to steer activity-dependent plasticity in response to training ( ).

Although clinical experience is limited to ES of the lumbar CPG, similar results are expected for the cervical cord ( ; ).

Importantly, since recovery of supraspinal control over limb movements is directly correlated with the amount of spared tissues , ES in the present context can only be deployed after chemical fusion.

As mentioned, M1 is concomitantly stimulated to promote corticospinal tract axonal outgrowth and plasticity; timing of paired stimuli is leveraged to produce plasticity during the rehabilitative phase ( ; ; ; ). Stimulation of M1 can be carried out both invasively and noninvasively ( ). In the present context, noninvasive cortical stimulation (NICS) by means of TMS is the preferred modality. M1 ES promotes sprouting of corticofugal fibers and modulates post-injury neuroplasticity ( ). M1 orchestrates recovery after SCI, developing new routes through the C-TRPS ( ; see also ).

Clinical translation

Based on experimental evidence (reviewed in ), PEG turns out to be the fusogen of choice: it is most effective when applied locally and acutely after cord transection and stump apposition. Cord necrosis and cavitation are greatly reduced or curbed. Studies show that the molecular weight of PEG greatly affects the final outcome, with PEG 600 and PEG1500 as the ones evincing the best results in animal models of cord transection (see in ).

Gemini hydrogelation

In the simplest approach, the damaged cord segment (or at least half of it) is removed, up to its border with rostral and caudal healthy tissue and the gap is filled with a PEG hydrogel. PEG can be cross-linked to form porous hydrogels, which can serve as biocompatible matrices that can closely mimic the Extra-Cellular Matrix. PEG hydrogels possess both high water content/porosity and solidity. Injectable PEG, by in situ gelling, can conform geometrically to the defect (see ). Regenerating propriospinal fibers course through this hydrogel “superhighway,” as demonstrated in rodent studies after many months ( ). Certainly, the use of PEG alone cannot completely mimic the three-dimensional porous structure of the spinal cord and would allow the upper and lower fiber bundles to grow in mismatched or even misplaced channels or pores, which is not the case with the next two approaches (see next section). Moreover, this approach would need many months for recovery. However, microspheres loaded with neurotrophins (e.g., BDNF and GDNF) could be embedded for slow release to accelerate this regrowth (see ).

Resection-apposition (Freeman-GEMINI approach)

As mentioned, Walter Freeman suggested the severance-reapposition approach to chronic SCI: he surgically removed the damaged segment of the cord in dogs creating a gap, performed a complete en bloc vertebrectomy thus shortening the spine, brought the two fresh cord stumps in contact with fresh plasma and sutured the dura tightly: walking animals resulted after several months ( ; see also ). Notably, he observed direct electrophysiological conductance across the apposed stumps and provided histological evidence of axonal regeneration across the sectional interface (see also and ; review in ). There is experimental evidence that the transected spinal cord in rabbits too can regenerate with concomitant return of motor, sensory and sphincter function when adequate apposition is obtained ( ). Others carried out Freeman’s surgery in dogs either immediately ( ) or 7 days after full section of the cord ( ). However, carried out Freeman’s procedure in paraplegic patients without success.

Spine-shortening vertebral osteotomy (a.k.a. vertebral column resection) is a surgical technique for correcting severe spinal deformities, treating congenital spinal anomalies, e.g., cord tethering, traumatic spine dislocations, and spine tumors at both cervical and thoraco-lumbar levels ( ; ; ; ; ). In the Freeman-GEMINI variant ( ; ), section of the damaged segment of the cord is performed at the moment of removing the vertebral body; the two ends are further trimmed so that no undue pressure is exerted on either stump by pressure vectors (too much as opposed to modest-pressure would lead to squeezing and local ischemia, jeopardizing the result). PEG is applied at this moment. The vertebra is removed and stabilization carried out simultaneously. The two spinal cord ends can be kept in apposition by some kind of micro-connector system: one incorporating a micro-channel system, through which PEG circulates, has been proposed ( ). Systems like this allow a tension-free, precise apposition of sharply transected nerve spinal cord stumps, as required by GEMINI. Spinal cord stimulators are epidurally positioned.

Spinal cord transplantation (Shirres-GEMINI approach)

In 1905, Shirres reported his attempt to graft a segment of healthy canine cord in a human paraplegic patient: clear signs of neuroregeneration were observed on autoptic material, with initial sensory recuperation at 3 months; autopsy showed signs of neuroregeneration (see ). Forty years later, operated on a 16-year-old male with complete loss of motor and sensory function after he was shot in his right shoulder with the bullet reaching the superior border of T4. Following laminectomy, the injured spinal cord was completely transected and replaced with a cadaveric spinal cord that had been fixed in 10% formalin for 12 days, and cleaned and sterilized with running and distilled water and 70% alcohol. No improvement in the patient’s condition was noted, and the patient died almost 4 months after the surgery. Autopsy showed exceptional preservation of the transplanted graft, although with restricted regeneration and limited tissue reaction. The preservation was attributed to the preoperative use of formalin, and no explanations or related conclusion on the microscopic findings could be made.

Today, human cord grafts/transplants are once again a consideration given the availability of an effective spinal fusion protocol, i.e., GEMINI, which would be key to interfacing the donor segment with the recipient’s cord stumps ( ). In this variant, PEG would neuroprotect the graft/transplant until vascularization from the healthy ends of the patient. PEG hydrogels support the formation of vascularized tissue in vivo and PEG has been shown to promote angiogenesis in an SCI model (see ). The donor segment is harvested from a brain-dead organ donor at the same level of injury. All the roots would be reanastomosed (a process akin to reimplantation of avulsed nerve roots: , see ( )). However, until tolerogenic protocols are developed for this contingency, immunosuppressants would be likely needed and this adds toxicity.

Depending on the length of the segment, surgical restoration of the vascular supply on one side may be required, but remains challenging with current techniques ( ).

Alternatively, a vascularized segment of the anatomically undamaged healthy cord caudal or rostral to the injury epicenter represents a suitable, interim alternative to fill the gap, as are bundles of peripheral nerves, as suggested by past animal experiments ( ; ) (modified Shirres-GEMINI variant).

Three points should be emphasized:

1.

the hemicord used as a bridge in chronic SCI has undergone years of plastic readjustments (as the entire CNS: ), some maladaptive (e.g., in the case of cord central pain: ). This might interfere with the process of recovery in some cases. However, it is well known that so-called entrenched plasticity is in fact reversible (discussed in ).
2.

Roots from the hemigraft have to be cut, which causes sensorimotor disruption in their respective territories of innervation. Studies of dorsal rhizotomies in human chronic pain patients verify that nearby roots take over and compensate for the resulting deficits (discussed in ). Motor deficits would affect chest and abdominal wall muscles in a limited area.
3.

The entire anatomical layout is disrupted. However, as mentioned (see above), it is the gray matter that supports locomotor recovery, even if white matter bundles are misaligned and not in continuity. This bridging segment is basically a gray matter bridge through which sensorimotor transmission is restored.

In all cases, PEG and electrical stimulation are deployed. PEG alone leads to remarkable recovery, but this is expected to be boosted by spinal cord and motor cortex stimulation (see above; reviewed in ; ).

Current human trial

CANINE TRIAL : A preliminary canine study of the modified Shirres-GEMINI approach has been conducted on four female Beagles in China. In this proof-of-principle canine trial, paralysis following removal of an entire 3 cm long segment of the cord was at least partially reversed by a pedicled autograft of the animal’s hemicord treated with PEG at the contact interfaces ( Fig. 1 ). At 6 months the final average motor score (canine 0–19 Basso-Bresnahan-Beatty scale, cBBB) was roughly in line with our previous canine PEG studies of a complete spinal cord transection (medians: respectively 11.5 vs 11: Table 1 ) (reviewed in ). The main differences were a slower onset of symptom reversal (17 days vs 3 days) and lower final top scores (12 vs 18). This is likely due to the need for re-growing fibers to cross two, instead of one, sectional levels and more prolonged neuroplastic rearrangement than a simple section. Sphincter control was not regained at 6 months. DTI and histology data verified a vigorous regrowth of fibers across the lesional interfaces ( Figs. 2 and 3 ).