Abstract BACKGROUND Human central nervous system stem cells (HuCNS-SC) are multipotent adult stem cells with successful engraftment, migration, and region-appropriate differentiation after spinal cord injury (SCI). OBJECTIVE To present data on the surgical safety profile and feasibility of multiple intramedullary perilesional injections of HuCNS-SC after SCI. METHODS Intramedullary free-hand (manual) transplantation of HuCNS-SC cells was performed in subjects with thoracic (n = 12) and cervical (n = 17) complete and sensory incomplete chronic traumatic SCI. RESULTS Intramedullary stem cell transplantation needle times in the thoracic cohort (20 M HuCNS-SC) were 19:30 min and total injection time was 42:15 min. The cervical cohort I (n = 6), demonstrated that escalating doses of HuCNS-SC up to 40 M range were well tolerated. In cohort II (40 M, n = 11), the intramedullary stem cell transplantation needle times and total injection time was 26:05 ± 1:08 and 58:14 ± 4:06 min, respectively. In the first year after injection, there were 4 serious adverse events in 4 of the 12 thoracic subjects and 15 serious adverse events in 9 of the 17 cervical patients. No safety concerns were considered related to the cells or the manual intramedullary injection. Cervical magnetic resonance images demonstrated mild increased T2 signal change in 8 of 17 transplanted subjects without motor decrements or emerging neuropathic pain. All T2 signal change resolved by 6 to 12 mo post-transplant. CONCLUSION A total cell dose of 20 M cells via 4 and up to 40 M cells via 8 perilesional intramedullary injections after thoracic and cervical SCI respectively proved safe and feasible using a manual injection technique. Cervical, Injection, Spinal cord injury, Stem cells, Thoracic ABBREVIATIONS ABBREVIATIONS AE adverse event AIS American Spinal Injury Association Impairment Scale ASIA American Spinal Injury Association CNS central nervous system DREZ dorsal root entry zone ES embryonic stem HuCNS-SC human central nervous system stem cells IRB Institutional Review Board INT intramedullary needle time IV intravascular MEP motor evoked potential MRI magnetic resonance imaging NPC neural progenitor cell OR operating room SAE serious adverse event SCI spinal cord injury SSEP somatosensory evoked potential UTI urinary tract infection The worldwide incidence of spinal cord injury (SCI) is estimated between 10 and 83 per million annually.1 Given that there is limited capacity for self-repair, cellular replacement or regenerative strategies are critical in central nervous system (CNS) injury and in particular for spinal cord trauma. There are a number of cellular therapies based on pluripotent and multipotent cell sources that show therapeutic promise. Predifferentiated human embryonic stem (ES) cells can be differentiated into neurons, oligodendrocytes, and astroglia, and have been shown in several rodent models to result in functional recovery after transplantation.2-8 However, there continues to be concerns of tumorigenesis with cells derived from ES cells (ie, pluripotent stem cells). In contrast, human neural stem cells, or neural progenitor cells (NPCs), are multipotent and have the potential to self-renew and differentiate into CNS cell types which could replace injured cell populations. NPCs can be derived from ES cells as well as from different developmental stages of the fetal or postnatal brain. NPCs derived from the fetal brain or the spinal cord are lineage restricted and demonstrate site-specific phenotypic differentiation upon transplantation.9-11 These fetal brain NPCs have the advantage of being nontumorigenic and have shown promise in promoting locomotor recovery in rodent SCI models.12-15 Human fetal CNS-derived stem cells (HuCNS-SC®, Stem Cells, Inc., Newark, California) have been utilized in the clinical trials under discussion in this manuscript. Human clinical trial data for use of HuCNS-SC has also been obtained for patients with Neuronal Ceroid Lipofuscinoses, Pelizaeus-Merzbacher Disease,16,17 and dry age-related macular degeneration. Several routes of cellular administration to treat SCI have been studied experimentally including intravascular (IV), intrathecal, and direct microinjection into the cord via an intralesional or perilesional approach. There is some experience with each of these routes using various stem cell sources for SCI (see review, Lamanna et al, 201318). Microinjections of cells into the spinal cord have the greatest potential of long-term engraftment and function, but are associated with the surgical risks and morbidity of an open “invasive” operative procedure. The optimal method of delivery of cells via injection into the spinal cord is an area of active research and consensus regarding the optimal technique remains debated. Based on preclinical studies in a rodent model of SCI, direct microinjection of HuCNS-SC into the spared parenchyma rostral and caudal to the injury epicenter resulted in cell engraftment, differentiation, extensive migration, and recovery of locomotor function. Functional improvement with improved locomotor recovery was observed after transplants performed 60 d post-SCI.13-15,19 Initial dosing and injection volume can be extrapolated from the murine model,13 using an allometric scale, so that an equivalent volume and cell dose can be calculated for injection into the human thoracic spinal cord.20 Large animal models of SCI 18,21-24 have been used to establish the safety of cell delivery by some investigators; however, full assessment of surgical safety and tolerability is inherently limited. In most animal studies, injections have been done in the setting of normal, uninjured spinal cords, and the questions regarding the safety of injection into the partially injured cord and changes introduced in the scale-up from small to large animals and ultimately into humans remain unclear. The impact of changes in human spinal cord structure and function after injury can only be determined by human trials, wherein post-transplant effects on motor and sensory function, as well as neuropathic pain, can be directly assessed using examination methods difficult to apply in experimental animal models. In addition, the safety and tolerability of cellular injections in the thoracic cord can be substantially different from that of the cervical cord, as changes in motor segments are more likely to be apparent after cervical injections. Designing a study with a dose escalation component allows for the determination of a safety window. With many possible options for cellular delivery, we sought to examine the safety of the freehand, also referred to as the manual injection, technique, which is clearly the most feasible and translatable technique for a multicenter study, particularly if the goal is to perform multiple injections around the injury epicenter. This operative technique would also be scalable for a wider clinical application. The objectives of the paper include describing the first surgical experience25 with HuCNS-SC in chronic thoracic and cervical SCI and highlighting the technical approach to the surgical administration of the cells in the thoracic study and the dose-escalation component of the cervical study. The emphasis will be on the surgical safety, feasibility, and tolerability of the neural stem cell transplantation used in both studies. Short- and long-term efficacy data for the cervical and thoracic study will be presented in a separate publication. METHODS Experimental Design The phase I/II thoracic study was conducted under regulatory authorization from Swissmedic and Health Canada, and the phase II cervical study under an Investigational New Drug application filed with the US Food and Drug Administration and Health Canada. Both trials were registered with ClinicalTrials.gov. The Western Institutional Review Board (IRB) evaluated and approved the cervical clinical trial for eligible participating centers, and all other centers relied on their respective site IRBs for approval. Patient consent was obtained for evaluation including screening and imaging, and a separate consent was obtained if a subject was enrolled as a control or in the treatment group including transplantation surgery, blood draws, immunosuppressive medications, etc. In the phase I/II open-label, single-dose safety and preliminary efficacy study of HuCNS-SC cell transplantation, subjects with a T2-T11 thoracic injury were eligible if they were at least 3 mo postinjury. Subjects, at least 4 mo postinjury, with C5-C7 cervical injury were enrolled in the phase II safety and efficacy study. The current report includes the subjects from both the chronic thoracic and cervical trial of HuCNS-SC transplantation. There were a total of 25 subjects screened for the 12 transplants in the thoracic study and 46 subjects screened to enroll the 31 patients in the cervical study that includes cohort I and II (with controls) and 2 patients nonrandomized due to early study termination. Follow-up ranged from 18 to 56 mo in the thoracic study and 1 to 12 mo in the cervical study. Patient demographics including level of injury, cell dose, and follow-up is described in Table 1 for all transplanted subjects in both the thoracic and cervical studies. Cervical cohort I included 2 subjects for each of 3 escalation dosages for a total of 6 subjects. No subjects in either study were enrolled more than 24-mo postinjury. The article was written in accordance with Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. TABLE 1. Demographic Characteristics of all Subjects Subject ID Age (years)/AIS level at Enrollment AIS Gender Race Dose HuCNS-SCs Exposure duration (mo) at study termination Thoracic SCI (Open label) 04-1001 23; T8 A M White/Caucasian 20 × 106 57 04-1002 53; T9 A M White/Caucasian 20 × 106 57 04-1003 45; T5 A M White/Caucasian 20 × 106 57 04-1004 32; T11 B M White/Caucasian 20 × 106 45 04-1005 25; T5 A M White/Caucasian 20 × 106 36 04-1006 39; T6 A M White/Caucasian 20 × 106 33 04-1007 31; T5 A M African American 20 × 106 33 04-1008 29; T6 B M White/Caucasian 20 × 106 32 04-1009 29; T2 B M White/Caucasian 20 × 106 29 07-1002 41; T3 A F Asian 20 × 106 25 08-1001 29; T5 B M White/Caucasian 20 × 106 29 08-1003 19; T6 A M White/Caucasian 20 × 106 28 Cervical SCI—cohort I (open, dose escalation) 17-1002 24; C6 B M White/Caucasian 15 × 106 12 13-1001 23; C6 B M White/Caucasian 15 × 106 12 17-1003 30; C6 A M White/Caucasian 30 × 106 12 13-1002 27; C4 B M White/Caucasian 30 × 106 12 18-1002 19; C6 A M White/Caucasian 40 × 106 12 19-1001 23; C6 B M White/Caucasian 40 × 106 12 Cervical SCI—cohort II (single blind, randomized, controlled) 15-1001 22; C5 B M White/Caucasian 40 × 106 9 17-1005 49; C5 B M White/Caucasian 40 × 106 9 18-1001 28; C6 A F White/Caucasian 40 × 106 9 19-1003 24; C6 B M White/Caucasian 40 × 106 9 19-1004 29; C4 B M White/Caucasian 40 × 106 7 17-1008 34; C5 B M White/Caucasian 40 × 106 7 23-1004 21; C6 B M White/Caucasian 40 × 106 5 22-1001 38; C6 B M White/Caucasian 40 × 106 5 20-1001 39; C6 B M White/Caucasian 40 × 106 2 18-1006 18; C6 B M Asian 40 × 106 1 22-1002 20; C5 B M White/Caucasian 40 × 106 1 Subject ID Age (years)/AIS level at Enrollment AIS Gender Race Dose HuCNS-SCs Exposure duration (mo) at study termination Thoracic SCI (Open label) 04-1001 23; T8 A M White/Caucasian 20 × 106 57 04-1002 53; T9 A M White/Caucasian 20 × 106 57 04-1003 45; T5 A M White/Caucasian 20 × 106 57 04-1004 32; T11 B M White/Caucasian 20 × 106 45 04-1005 25; T5 A M White/Caucasian 20 × 106 36 04-1006 39; T6 A M White/Caucasian 20 × 106 33 04-1007 31; T5 A M African American 20 × 106 33 04-1008 29; T6 B M White/Caucasian 20 × 106 32 04-1009 29; T2 B M White/Caucasian 20 × 106 29 07-1002 41; T3 A F Asian 20 × 106 25 08-1001 29; T5 B M White/Caucasian 20 × 106 29 08-1003 19; T6 A M White/Caucasian 20 × 106 28 Cervical SCI—cohort I (open, dose escalation) 17-1002 24; C6 B M White/Caucasian 15 × 106 12 13-1001 23; C6 B M White/Caucasian 15 × 106 12 17-1003 30; C6 A M White/Caucasian 30 × 106 12 13-1002 27; C4 B M White/Caucasian 30 × 106 12 18-1002 19; C6 A M White/Caucasian 40 × 106 12 19-1001 23; C6 B M White/Caucasian 40 × 106 12 Cervical SCI—cohort II (single blind, randomized, controlled) 15-1001 22; C5 B M White/Caucasian 40 × 106 9 17-1005 49; C5 B M White/Caucasian 40 × 106 9 18-1001 28; C6 A F White/Caucasian 40 × 106 9 19-1003 24; C6 B M White/Caucasian 40 × 106 9 19-1004 29; C4 B M White/Caucasian 40 × 106 7 17-1008 34; C5 B M White/Caucasian 40 × 106 7 23-1004 21; C6 B M White/Caucasian 40 × 106 5 22-1001 38; C6 B M White/Caucasian 40 × 106 5 20-1001 39; C6 B M White/Caucasian 40 × 106 2 18-1006 18; C6 B M Asian 40 × 106 1 22-1002 20; C5 B M White/Caucasian 40 × 106 1 Thoracic patients were followed for 12 mo and then enrolled in a long-term follow-up study up to a maximum of and additional years (for a total of 5 yr). Twelve months was intended follow-up period for cervical study. View Large Safety assessments included collection of adverse event reports (AEs) coded using the Medical Dictionary for Regulatory Activities Terminology (MedDRA V11.0 or higher) dictionary. Any AE resulting in death, that was perceived as life-threatening, that required prolongation of existing hospitalization or hospital readmission, or that resulted in persistent or significant disability/incapacity, was considered a serious adverse event (SAE). The number and percentage of subjects with S/AEs were summarized for each treatment by maximum intensity and relationship to study treatment. Inclusion and Exclusion Criteria Two studies were undertaken that included male and female subjects 18 to 60 yr of age with a single traumatic and nonpenetrating SCI, based on magnetic resonance imaging (MRI) with American Spinal Injury Association (ASIA) Impairment Scale (AIS) grade A or B. All subjects were also required to be in generally good medical condition other than their injury, to have no contraindications for systemic immunosuppression, MRIs, or safe surgical exposure of the lesion area. The detailed inclusion and exclusion criteria are listed in Table 2. TABLE 2. Major Inclusion and Exclusion Criteria Inclusion Criteria: 1. Male or female, 18 to 60 yr. 2. Traumatic, AIS A or B (cohort I), AIS B (cohort II). a. Phase II study in cervical SCI: C5-7 (motor) b. Phase I/II study in thoracic SCI: T2-11 3. Greater than 6 wk (thoracic SCI) and 12 wk (cervical SCI) postinjury prior to screening a. AIS A subjects may be transplanted between 16 to 52 wk postinjury. b. AIS B subjects may be transplanted between 16 to 104 wk postinjury. Exclusion criteria: 1. History of penetrating SCI or MRI evidence of complete spinal cord interruption. 2. Medical contraindication to MRI and evidence of spinal instability, stenosis and/or persistent cord compression related to the initial trauma. 3. Presence of spinal instrumentation and/or fusion construct that would preclude safe exposure of the spinal cord. 4. Active conditions: pregnancy; lactation; pressure ulcers; or vaccination of live virus within 6 wk prior to initiation of immunosuppression. 5. Contraindication for immunosuppression or allergy to immunosuppression medications. 6. Prior participation in another investigational study within 90 d prior to screening. 7. Significant comorbidities and other significant medical findings that would preclude safe participation in the trial including behavioral or social conditions. Inclusion Criteria: 1. Male or female, 18 to 60 yr. 2. Traumatic, AIS A or B (cohort I), AIS B (cohort II). a. Phase II study in cervical SCI: C5-7 (motor) b. Phase I/II study in thoracic SCI: T2-11 3. Greater than 6 wk (thoracic SCI) and 12 wk (cervical SCI) postinjury prior to screening a. AIS A subjects may be transplanted between 16 to 52 wk postinjury. b. AIS B subjects may be transplanted between 16 to 104 wk postinjury. Exclusion criteria: 1. History of penetrating SCI or MRI evidence of complete spinal cord interruption. 2. Medical contraindication to MRI and evidence of spinal instability, stenosis and/or persistent cord compression related to the initial trauma. 3. Presence of spinal instrumentation and/or fusion construct that would preclude safe exposure of the spinal cord. 4. Active conditions: pregnancy; lactation; pressure ulcers; or vaccination of live virus within 6 wk prior to initiation of immunosuppression. 5. Contraindication for immunosuppression or allergy to immunosuppression medications. 6. Prior participation in another investigational study within 90 d prior to screening. 7. Significant comorbidities and other significant medical findings that would preclude safe participation in the trial including behavioral or social conditions. View Large Surgical Technique In both the thoracic and cervical studies, the surgical approach involved perilesional intramedullary injections of stem cells. Injections in the clinical trials were not specifically targeted to a designated anatomic motor tract. This was based on data from the Non-Obese diabetic-Severe Combined Immunodeficient (NoD-SCID) mouse spinal cord contusion model,12,13 where extensive survival and migration of HuCNS-SC was observed following nonspecific targeted intramedullary injections. The trajectory of the needle relative to the dorsal root entry zone (DREZ) was selected to avoid eloquent intramedullary structures, and the depth of the injection was determined by the preop axial MRI. After induction of general anesthesia and intubation, the subjects were placed in the prone position on the operating room (OR) table. Preincisional IV antibiotics were administered. Somatosensory and motor evoked potentials (SSEPs/MEPs) were used for spinal cord and nerve root monitoring.26 All patients had previously undergone decompressive laminectomy and/or spinal realignment and instrumented fusion after injury. A part of the old midline skin incision was reopened to expose the surgical transplantation site. Scar tissue and the paraspinal muscles were dissected to expose the underlying dura overlying the intended transplant sites. The injections were performed with maintenance of the posterior spinal fusion in place, apart from cross bars, if it was obstructing the transplantation area. In 2 thoracic patients, the intradural scarring was dense and a skip dural opening (ie, 2 dural openings) was performed to expose the transplantation site superior and inferior to the injury site (Figures 1A-1D). FIGURE 1. View largeDownload slide A-D, Thoracic spine anteroposterior x-rays A and C demonstrating presence of spinal instrumentation and intraoperative images B and D of same patients in thoracic SCI trial. In the first case, there is a single dural opening and in the other case skip dural openings because of dense arachnoid adhesions at injury epicenter prevented safe opening of the dura. FIGURE 1. View largeDownload slide A-D, Thoracic spine anteroposterior x-rays A and C demonstrating presence of spinal instrumentation and intraoperative images B and D of same patients in thoracic SCI trial. In the first case, there is a single dural opening and in the other case skip dural openings because of dense arachnoid adhesions at injury epicenter prevented safe opening of the dura. Intraoperative ultrasound imaging (Hitachi HI Vision Ascendus, Hitachi Medical Systems Europe Holding AG, Zug, Switzerland/12 Mhz linear array transducer on an IU22 scanner, Hitachi Aloka Medical America, Inc., Wallingford, Connecticut) was used to visualize intramedullary changes, especially cystic cavities. This was essential to define the rostral and caudal transplantation sites. The extent of dural opening was determined according to the ultrasound findings. The intramedullary echogenicity produced by the injury helped to define the site of injection. This information was used in addition to the borders of the traumatic lesion based on the extent of the altered T1 and/or T2 intramedullary MRI signals/characteristics typical for SCI. The sites targeted for injection were based on the immediate adjacent spinal segment beyond the abnormal signal. The spinal cord segments targeted for transplantation may have contained a partial signal change consistent with the tapering edge of the radiographic changes secondary to the injury. The dura was opened in the midline extending past the rostral and caudal transplantation site and then tack up sutures were placed. Using the magnification and illumination of the operating microscopic, the dorsal surface anatomy was exposed by opening the arachnoid and lysis of any intradural adhesions overlying the intended transplant site. Exposure of the area lateral to the dorsal columns, medial to the DREZ was obtained above and below the lesion area. Specifically, extensive untethering or fenestration of the spinal cord cyst was avoided. Thoracic Cord Injections (n = 12) A total volume of 280 μL of HuCNS-SC was provided for the transplantation procedure in 2 equivalent vials. Gentle swirling and tapping of the vial guaranteed that no cells would remain within the cap of the cell vial. Using an 18-gauge blunt needle, the cell suspension was aspirated up and down 3 times avoiding any air bubble formation. This led to mainly a solution of small neural cell clusters as demonstrated in Vitro previously. The cell containing fluid was then drawn into the syringe. A commercially available syringe with 10 μL marks was used. A volume of 140 μL was finally aspirated into the syringe to perform 2 subsequent injections. The cells were injected using a 30-gauge needle premarked to a depth of 3 to 4 mm. After the area of injection was assessed with the ultrasound, the operating microscope was used to examine the superficial aspect of the dorsal cord surface. Special care was taken to avoid any superficial blood vessels. The needle was then slowly inserted in a slightly lateral to medial trajectory until the tip reached the desired depth calculated based on the preoperative MRI. The needle was then advanced slightly past target and withdrawn back into position to create less resistance as the initial cells were injected. The study protocol indicated that the rate of cell infusion should be 20 μL/60 s followed by an additional 1-min dwell time and slow needle withdrawal to avoid reflux of cells along the needle tract. It was essential to have an OR technologist or the study nurse control the injection time. During the injection procedure, the microscope was directed to observe the pial surface entry point and the microliter marks on the syringe. Upon withdrawal of the needle, special attention was given to the pial surface to document cell reflux through the needle tract. Stabilization of the hand-held syringe and needle is critical to avoid injury to the spinal cord. Neurosurgeons with expertise in microsurgical techniques are accustomed to using stabilizing techniques when working in highly eloquent areas. One hand was used to stabilize the syringe resting on the side of the surgical opening, the other hand was used to move the plunger in a very controlled way, also stabilized on the border of the surgical field. The stable position was maintained for the duration of the injection (3:30 min) and the 1-min dwell time. Intramedullary needle time (INT) = active injection time + dwell time with needle in place within the spinal cord was recorded for each injection. Total INT = the cumulative time for all injections (n = 4, 6, or 8) with needle in place within the spinal cord. Total injection time = the time between start of first injection and withdrawal of needle from the spinal cord upon last injection was also recorded. After the injections, an ultrasound was performed and dura was closed in a watertight fashion covered with Duragen Plus (Integra, LifeSciences Corporation, Plainsboro, New Jersey) and fibrin glue. Cervical Injections (n = 17) A vial containing a total of 1 mL of HuCNS-SC was provided for each cervical transplant. The patients in cervical cohort I formed the dose escalation study. In cohort I, a series of 4, 6, or 8 microinjections were performed depending on the dosage group: 15 (n = 2 subjects), 30 (n = 2), or 40 (n = 2) million cells. Group 1a patients received 4 microinjections for a total dose of 15 million cells delivered in 2 separate 70 μL microinjections (10 million cells) at the inferior border and 2 separate 35 μL microinjections (5 million cells) at the superior border of the injury. Group 1b received 6 injections for a total dose of 30 million cells delivered in 4 separate 70 μL microinjections at the inferior border (20 million cells) and 2 separate 70 μL microinjections (10 million cells) at the superior border of the injury. Group 1c received 8 injections for a total dose of 40 million cells delivered in 4 separate 70 μL microinjections (20 million cells) at the inferior border and 4 separate 70 μL microinjections (total of 20 million cells) at the superior border of the injury. In cohort I, intramedullary transplantations were balanced into hemicord: the pattern of 4 injections was comprised of 1 injection into each hemicord above and below the epicenter, 6 injections (1 injection into each hemicord above and 2 into each hemicord below), and 8 injections (2 hemicord injections above and below). For multiple transplants below or above the injury level, the injection sites were separated by approximately 5 to 7 mm along the long axis of the hemicord from the adjacent injection site (Figure 2A). The entry sites were located at the dorsal pial surface, half way between the dorsal intermediate sulcus and medial to the DREZ. The selected microinjection dorsal entry site and approximately 30-degree inclination of the needle (Figure 2B) trajectory toward the central canal of the spinal cord was done to avoid transgression of eloquent motor (corticospinal) tracts. Rostral injections performed within the dorsal columns, while within the spinal cord above the injury epicenter also represented a “safe” zone as the tissue had undergone Wallerian degeneration (Figure 2C). Caudal injections were below injury level. FIGURE 2. View largeDownload slide A, The dose escalation schema including cell number/volume and approximate location relative to injury epicenter in cervical cohort I is demonstrated. B, Axial light microscopic thin section of the normal cervical spinal cord demonstrating the white matter tracts and the relatively large ventral horns. Needles during the transplant procedure sit in the region of the dorsal columns. C, Demonstrates a T2-weighted axial MR rostral to the cervical injury epicenter with evidence of Wallerian degeneration seen (area within dashed lines). The needles during the injection are located for the most part in this relatively functionally silent area of the spinal cord. FIGURE 2. View largeDownload slide A, The dose escalation schema including cell number/volume and approximate location relative to injury epicenter in cervical cohort I is demonstrated. B, Axial light microscopic thin section of the normal cervical spinal cord demonstrating the white matter tracts and the relatively large ventral horns. Needles during the transplant procedure sit in the region of the dorsal columns. C, Demonstrates a T2-weighted axial MR rostral to the cervical injury epicenter with evidence of Wallerian degeneration seen (area within dashed lines). The needles during the injection are located for the most part in this relatively functionally silent area of the spinal cord. The pial surface at the point of the needle entry was sharply opened with a #11 blade tip in order to facilitate easy insertion of the needle through the superficial aspect of the dorsal cord surface. The 30-gauge needle was slowly inserted into position within the spinal cord until the tip reached the depth calculated on the cord dimensions based on the preoperative MRI and intended site of injection within the cord. The target depth was between 3 and 5 mm below the dorsal surface. A variety of techniques were used to mark the depth of the needle including scoring the needle with a marker, a surgical rongeur without narrowing the needle outflow, placing a precut silicone sheet at the required depth or placing an IV catheter hub on the needle with the desired depth exposed. The needle trajectory was directed slightly medial with a 15° to 30° angle towards the central region of the cord to access the anatomic target. Intraoperative visualization of the selected pial entry sites allowed the surgeon to avoid injury to the dorsal vessels (Figures 3A-3C). The protocol mandated that a rate of cell infusion was 20 μL/min with a maximum time of injection of 3:30 min (70 μL) followed by an additional 1-min dwell time to avoid reflux of cells along the needle tract. Stabilizing a hand-held syringe and needle was critical for the time required for each injection (2:45-4:30 min) and using a 2-hand technique with stabilization of the surgeons’ hands on the wound side walls and retractors facilitated the process. After the injections, an ultrasound was performed and dura was closed in a water tight fashion with the use of a dural graft, if required. The INT, total INT, and total injection time (see above) was recorded for each subject. FIGURE 3. View largeDownload slide A, Intraoperative image with the dura widely opened and dural tack up stiches in place demonstrating the dorsal surface of the cervical spinal cord. Surface vessels look relatively normal and the area of the dorsal columns can be seen between the dorsal rootlets. There are areas of spinal cord thinning representing the area of the post-traumatic cystic change seen in B, upper left corner. Intraoperative ultrasound is critical in determining the cyst size and perimeter (dashed lines). C, Intraoperative image of 30-gauge needle in place during injection of “normal” spinal cord caudal to the cyst with HuCNS-SC is seen in the lower right corner. FIGURE 3. View largeDownload slide A, Intraoperative image with the dura widely opened and dural tack up stiches in place demonstrating the dorsal surface of the cervical spinal cord. Surface vessels look relatively normal and the area of the dorsal columns can be seen between the dorsal rootlets. There are areas of spinal cord thinning representing the area of the post-traumatic cystic change seen in B, upper left corner. Intraoperative ultrasound is critical in determining the cyst size and perimeter (dashed lines). C, Intraoperative image of 30-gauge needle in place during injection of “normal” spinal cord caudal to the cyst with HuCNS-SC is seen in the lower right corner. RESULTS Thoracic Cord Injections (n = 12) Intraoperative ultrasound observations were extremely helpful in defining injury epicenter and demonstrated a heterogeneous pathology with variable cyst size (Figures 4A-4E). Examination of the pial surface alone was not predictive of the area of injection to target or the degree of underlying cystic change within the spinal cord. Initial technical feasibility of neural cell transplantation using a hand-held syringe was performed in the 12 thoracic subjects across 3 centers and by 3 different neurosurgeons. Using appropriate microsurgical techniques with good hand positioning provided the necessary stability for the duration of the manual injections. Motion of the spinal cord was insignificant for the duration of the transplantation procedure, and the hand-held syringe injection technique allowed the surgeon to compensate for minimal systolic and/or respiratory movements encountered (Figures 5A-5C). FIGURE 4. View largeDownload slide A-E, Schematic diagram of injection procedure in thoracic spinal cord rostral to injury epicenter with hand-held syringe and needle in place. Green represents HuCNS-SC upon injection A and upon completion of all injections and allowing for migration of the cells similar to observations seen in basic science preclinical studies12,13B. C, Intraoperative image of 30-gauge needle in place during injection of thoracic spinal cord with HuCNS-SC. D, Intraoperative ultrasound performed immediately after thoracic injection and 5 min later E. These images are from one of 2 cases in which the injectate could be seen and appears to disperse along white matter tracts of the spinal cord. FIGURE 4. View largeDownload slide A-E, Schematic diagram of injection procedure in thoracic spinal cord rostral to injury epicenter with hand-held syringe and needle in place. Green represents HuCNS-SC upon injection A and upon completion of all injections and allowing for migration of the cells similar to observations seen in basic science preclinical studies12,13B. C, Intraoperative image of 30-gauge needle in place during injection of thoracic spinal cord with HuCNS-SC. D, Intraoperative ultrasound performed immediately after thoracic injection and 5 min later E. These images are from one of 2 cases in which the injectate could be seen and appears to disperse along white matter tracts of the spinal cord. FIGURE 5. View largeDownload slide A-F, MRI sagittal T2-weighted images of the thoracic spinal cord A, C, and E correlated with intraoperative images of the dorsal spinal cord after opening of the dura B, D, and F. The images demonstrate the marked heterogeneity of the appearance of the thoracic cord lesions by MRI including simple cyst, multicystic, to solid cord injuries. The dorsal surface view alone cannot discern between these injuries illustrating the importance of preoperative imaging and intraoperative ultrasound. FIGURE 5. View largeDownload slide A-F, MRI sagittal T2-weighted images of the thoracic spinal cord A, C, and E correlated with intraoperative images of the dorsal spinal cord after opening of the dura B, D, and F. The images demonstrate the marked heterogeneity of the appearance of the thoracic cord lesions by MRI including simple cyst, multicystic, to solid cord injuries. The dorsal surface view alone cannot discern between these injuries illustrating the importance of preoperative imaging and intraoperative ultrasound. Intraoperatively, no reflux of cells was observed in any of the thoracic cord procedures when using the slow injection and needle dwell technique. The time required for each of the 4 individual injections, including cell preparation, aspiration, and injection, ranged between 4 and 7 min. The average total dwell time of cell transplantation was 19:30 min and total injection time 42:15 min (Table 3). There were no pretransplant SSEPs or MEPs detected below the level of injury, nor were changes noted in recorded electrophysiological changes during the microinjections above injury. Immediate post-transplant ultrasound did not demonstrate new hemorrhages or swelling in the area of cell injection. In 2 subjects, we observed a small transient hyperintense ultrasound echo in the center of the spinal cord after cell injection; however, no hemorrhage was diagnosed in the postoperative MRI and the signal change was felt to be consistent with the cell suspension deposit. Interestingly, a craniocaudal spread of the signal alteration was observed in 2 subsequent images taken within 5 min of the injection (Figures 5D-5F). This observation could represent the early mechanical/fluid distribution of cells along white matter tracts. There were no definitive intramedullary changes observed on the postop MRI after thoracic spinal cord injections in any of the 12 subjects performed within 48 h of surgery. TABLE 3. Injection Data—Cervical/Thoracic Cohorts Thoracic (n = 12) Total INT (mins) ± SEM Total injection time (mins) ± SEM 20 M (4 inj/280 μL) 19:30 ± 1:17 42:15 ± 4:36 Cervical cohort I (n = 6) Total Intramedullary needle time (mins) ± SEM Total injection time (mins) ± SEM 15 M (n = 2/4 inj/210 μL) 10:30 ± 1:30 29:00 ± 4:00 30 M (n = 2/6 inj/420 μL) 18:30 ± 0:30 53:00 ± 7:00 40 M (n = 2/8 inj/560 μL) 25:30 ± 2:30 1:01:30 ± 8:30 CERVICAL COHORT II (n = 11) Total intramedullary needle time (mins) ± SEM Total injection time (mins) ± SEM 40 M (8 inj/560 μL) 26:05 ± 1:08 58:14 ± 4:06 Thoracic (n = 12) Total INT (mins) ± SEM Total injection time (mins) ± SEM 20 M (4 inj/280 μL) 19:30 ± 1:17 42:15 ± 4:36 Cervical cohort I (n = 6) Total Intramedullary needle time (mins) ± SEM Total injection time (mins) ± SEM 15 M (n = 2/4 inj/210 μL) 10:30 ± 1:30 29:00 ± 4:00 30 M (n = 2/6 inj/420 μL) 18:30 ± 0:30 53:00 ± 7:00 40 M (n = 2/8 inj/560 μL) 25:30 ± 2:30 1:01:30 ± 8:30 CERVICAL COHORT II (n = 11) Total intramedullary needle time (mins) ± SEM Total injection time (mins) ± SEM 40 M (8 inj/560 μL) 26:05 ± 1:08 58:14 ± 4:06 Intramedullary needle time (INT) = active injection time + dwell time with needle in place within the spinal cord. Total INT = the cumulative time for all injections (n = 4, 6, or 8) with needle in place within the spinal cord. Total Injection time = the time between start of first injection and withdrawal of needle from the spinal cord upon last injection. SEM = standard error of the mean. Inj = injections. Mins = minutes. View Large Within the first year postsurgery, there were a total of 4 SAEs and 90 AEs. The 4 SAEs included cerebrospinal fluid leak (11 d postop), pseudomeningocele (13 d postop), constipation (44 wk postop), and a urinary tract infection (UTI; 49 wk postop). There were no cases of worsening spasticity or pain syndrome (allodynia). Of the 90 AEs, the 3 most common conditions were UTI (n = 7/12; 58.3%), decubitus ulcer (n = 5/12; 41.7%), and headache (n = 5/12; 41.7%). Approximately 9.6% of the AEs were related to surgery and 6.4% related to immunosuppression. Specifically, there were no SAEs or AEs attributable to the injection procedure or the transplanted neural stem cells. Cervical Cord Injections (n = 17) The surgical approach and cell transplantation technique using a hand-held syringe implemented in the thoracic study was adapted easily for the 6 subjects in the dose escalation arm (cohort I) and the 11 accrued subjects in the randomized arm (cohort II) of the cervical study. As with the observations made in the thoracic transplantations, no significant systolic pulsations of the cervical spinal cord or respiratory motions were appreciated. No subject required a request for anesthesia to induce controlled apnea. The total INT and total injection time varied with cell dose, number of injection and volume within cervical cohort I. The range between the low- (15 M) and the high-dose (40 M) groups spanned from 10:30 to 25:30 min for INT and 29 to 61:30 min for total injection time (Table 3). In cohort II, the INT and the total injection time was 26:05 ± 1:08 min and 58:14 ± 4:06 min, respectively. The total injection time included the time required for each of the 8 transplantations, including cell preparation, mixing, aspiration, and injection (Table 3). The recorded INT and total injection times were remarkably similar (relatively low Standard Error of the Mean (SEM)) despite involving 7 individual neurosurgeons in separate institutions in cohort II. Intraoperatively, there was minimal to no reflux of cells observed in any of the cervical cord injured subjects when using the slow injection and needle withdrawal technique; average reflux reported across all centers was 0% to 10% of volume injectate. There was no detected electrophysiological activity in the lower extremities pre- or intraoperatively, even in the cervical AIS B subjects. There were no detected changes in the SSEPs, MEPs, or spontaneous Electromyogram (EMGs) detected in the upper extremities intraoperatively during the microinjections. Intraoperative ultrasound was critical in defining the rostral and caudal limits of the spinal cord cystic change and no detectable signal changes on ultrasound was observed postinjection. In 1 cervical patient, a dural graft was added to achieve a water-tight dural closure. The postoperative cervical spinal MRIs (Figures 6A-6D) obtained within 24 to 48 h demonstrated mild increased diffuse T2 signal in 8 of 17 transplanted subjects (47%). When specifically evaluating the high-dose group (40 M cells), the incidence of postoperative signal change remained the same: 6 of 13 subjects (46%). The location of the signal change, generally observed both above and below the lesion, did not extend over more than 1 to 2 spinal segments and tended to be located primarily in the dorsal column. There was no evidence on MRI of intramedullary hemorrhage or reactive edema and the postoperative signal change resolved in a majority of the subjects by 6 mo and in all subjects by 12 mo post-transplant (Figures 6E-6G), based on available follow-up. No decrements in motor function as measured by 2 different validated strength instruments or the emergence of a new pain syndrome associated with the cervical injections, including those cases where T2 signal change was seen postoperatively. FIGURE 6. View largeDownload slide A-G, MR images of pretransplant appearance A of the cervical spinal cord of a patient in cohort II with an AIS B—chronic motor complete injury. MRI done 24 h after injection of 40 M cells (560 μL) in 8 divided doses above and below the lesion B-D demonstrates high signal intensity with minor amount of cord swelling. The majority of the signal change is seen in the dorsal columns. The neurological exam is unchanged with no new complaints of parasthesias or pain syndrome. At 12 mo post-transplant, the signal change has disappeared both above and below the injury epicenter. FIGURE 6. View largeDownload slide A-G, MR images of pretransplant appearance A of the cervical spinal cord of a patient in cohort II with an AIS B—chronic motor complete injury. MRI done 24 h after injection of 40 M cells (560 μL) in 8 divided doses above and below the lesion B-D demonstrates high signal intensity with minor amount of cord swelling. The majority of the signal change is seen in the dorsal columns. The neurological exam is unchanged with no new complaints of parasthesias or pain syndrome. At 12 mo post-transplant, the signal change has disappeared both above and below the injury epicenter. Four SAEs were reported in 3 of 12 patients who underwent HuCNS-SC cell transplantation in the open-label thoracic study. Two SAEs were described as cerebrospinal fluid leakage in separate patients, 10 d postoperatively with no signs of spinal cord compromise or postoperative hemorrhage. The 2 remaining SAEs were reported at 10 and 12 mo postop, respectively, in 1 patient hospitalized for separate occurrences of constipation and UTI, and both were attributed to the underlying SCI. For the 17 subjects who underwent HuCNS-SC cell transplantation in the cervical study, a total of 18 SAEs occurred in 12 subjects in the cervical study. In cohort I, there were 2 SAEs involving a prolonged hospitalization related to a staph epidermidis wound infection that required incision and drainage 8 d postsurgery and a new hospitalization for severe constipation 3 mo postsurgery. In cohort II, there were 13 SAEs recorded in 7 transplanted subjects (4/16 cohort II SAEs were recorded among 3 control subjects). These events were considered serious because they either required extending an existing hospital stay, required new hospital admission, or were considered an important medical event. The SAEs reported in cohort II included autonomic dysreflexia (n = 3), postprocedural sepsis, posterior reversible encephalopathy syndrome, constipation, seizure, UTI, wound hematoma, and aphasia. By way of personal communication amongst the coauthors, no additional SAEs were recorded poststudy termination to date. Additional AEs, common to SCI and not attributed to the stem cells, were recorded in this period via informal telephone follow-up since no study termination visits were provided by the Sponsor. The most common nonserious AEs attributed to the surgical procedures included neck and musculoskeletal pain. Two cases of sensory changes were reported postoperatively. One report of left fifth metacarpal hypoesthesia was described immediately after surgery which resolved after 14 d. One report of paresthesias was described as a 50% decline in light touch but no change in pin prick (per International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI)) with marked improvement by 28 d post-transplant and complete resolution by 3 mo. There were no measurable decreases in upper extremity motor strength (per ISNCSCI) immediately postop and no patient had a decrease in their motor score 3 mo post-transplant compared with their presurgery motor score. None of the SAEs or AEs of a permanent nature were considered to be attributable to the surgical exposure, injection procedure, or the HuCNS-SC cells themselves. All SAE and AEs in the cervical study, except for a decubitus ulcer noted preoperatively, resolved completely by 3 mo postsurgery. There were no instances in which the laminectomy procedure was felt intraoperatively to have destabilized the spine and no delayed presentation of postprocedural spinal instability. While a more detailed analysis of clinical outcomes will be published at a later date, the magnitude of improvement in cohort I at 1 yr and an interim analysis of cohort II at 6 mo fell below the required clinical efficacy threshold set by the sponsor to support further development resulting in early study termination. DISCUSSION Human Central Nervous System Stem Cell HuCNS-SC can be propagated, cryopreserved, and banked, while retaining critical biological activity of self-renewal, engraftment, paracrine effects from secreted factors to enhance neural plasticity, migration, and tri-lineage differentiation (neurons, oligodendroctyes, and astrocytes). The cells have undergone rigorous in Vivo analysis in multiple animal models of CNS disorders, including experimental testing in NOD-SCID mice after SCI.12,13 Results in NOD-SCID mice with subacute and chronic contusive thoracic SCI have demonstrated engraftment, long-term survival, extensive migration, differentiation, and improvement in locomotor function.15,19 Phase I clinical trials with HuCNS-SC have been conducted in a variety of CNS diseases including neuronal ceroid lipofuscinosis (Batten disease), Pelizaeus-Merzbacher myelin disorder,16,17 and dry age-related macular degeneration. Safety and tolerability with evidence of postmortem cell survival were observed up to 6 yr after cessation of immune suppression (manuscript in preparation). Clinical trial experience with HuCNS-SC in the above indications has demonstrated an emerging safety profile with a range of cell doses and transplantation sites. Critical concepts to consider in translating stem cell therapies for SCI include injury severity, cell type, spinal level (ie, cervical vs thoracic), cell delivery system, as well as epicenter vs perilesional injections. Preclinical safety data can be gathered from small and large animal studies, including tolerable cell volumes based on the gross neurological exam and neuropathology, but ultimately data from clinical trials will demonstrate data supporting emerging safety and tolerability. The feasibility of the surgical approach to the spinal cord after injury and cell transplantation using a hand-held syringe was applied to all 29 SCI subjects (12 thoracic and 17 cervical) across both clinical trials. Preclinical Experiments in SCI Using Hand-Held Injections vs Anchored Devices In principle, freehand, or manual, cell injection is differentiated from stabilized methods where the injection syringe is either fixed by device to the animal or the surgical table. In a mouse contusive SCI, neural stem/progenitor cell engraftment was compared between intralesional vs intrathecal vs IV injections.25 Engraftment is highest in the intralesional hand-held injection group and after about 6 wk post-transplantation, cell luminescence decreases to about 10% of its initial level at the site of injury. In the intrathecal group, grafted cell luminescence was distributed throughout the subarachnoid space soon after transplantation. It was detected at the injury site pial surface 1 wk later, and by 6 wk had decreased to about 0.3% of the initial level. In the IV group, no grafted luminescent cells were detected at the injury site, but all of these mice demonstrated cell deposition in the chest, suggesting pulmonary embolism.27 Specifically looking at intralesional injection in murine SCI—cell types include neural stem cells, Schwann cells, olfactory ensheathing glia, and others—in acute, subacute, and chronic injuries.10-17 Outlining the details of dose, volume, and cell type goes beyond the scope of this paper. However, numerous of these studies were performed using a free-hand technique, but unfortunately no studies have provided a head to head comparison between the hand-held technique and the approach using a syringe stabilizing device. Device-stabilized approaches to delivery of cells provide the ability to precisely target single or multiple sites within the spinal cord such as in cranial stereotactic techniques. Preclinical cell transplantation studies in rodents have mainly used the stereotactic frame adapted as a syringe and needle holder or free-hand injections.27-29 Results of both techniques have shown atraumatic cell delivery to the spinal cord. A special emphasis has been put on scalable cell delivery techniques in preparation for clinical trials.24,30 The majority of published large animal studies have analyzed lesion size, engraftment, and neurological outcomes of injection into a mini-pig model using anchored devices, including a stereotactic stabilized microinjection platform that has been shown to precisely deliver cells.20 The magnitude of the incision as well as the muscle dissection required to place the stereotactic apparatus can be significant.20 Clinical Experience with Hand-Held Injections in Acute and Chronic SCI A true understanding of the practical aspects of a spinal cord injection procedure in humans including consequences of scale-up and subtle segmental cord changes may only be acquired by testing the surgical technique in clinical studies. The optimal cell delivery device and the necessity of “anchoring” the syringe during injection is subject to debate, perhaps in part dependent on the duration of the injection. The motion of the spinal cord both in the thoracic and cervical spinal regions was minimal and no apnea was required during the microinjection procedure. Moreover, the hand-held syringe injection technique allowed the surgeon to compensate for minimal systolic and/or respiratory movements. For an overview of technical considerations for clinical application of cellular therapeutic delivery to the spinal cord, see table 1 in Lamanna et al 2013.18 There have only been a few published trials reported on the results of hand-held intramedullary cellular or tissue injections after human SCI.29-32 Clinical trials that used hand-held cellular injections after acute, subacute, and chronic SCI involved bone marrow stem cells,33 activated macrophages (Proneuron study),34 and autologous Olfactory Ensheathing Glia (OEG) tissues,32 respectively. In the acute and subacute trials, injections occurred in the setting where proinflammatory cascades are in process and may harm the transplanted cells, and thus the immediate safety of the injection is difficult to determine especially when some neurological recovery is the rule. In the Proneuron phase II trial,35 free-hand cell injection of activated autologous macrophages was performed in 26 acute C5 to T11 SCI subjects. Six 20 μL injections, each containing 250 000 autologous incubated macrophages (total dose of 1.5 × 106 cells in 120 μL), were performed with a single hand-held syringe at the caudal boundary of the spinal cord contusion only. In this clinical trial, the conversion rate from AIS A to a higher grade (B or C) was higher in the control group (58.8%) than the treatment group (26.9%). Although there was a trend for greater overall improvement in the control group, it was not significantly different. No clinical or safety concerns were raised with the study and early termination was related to failure to achieve the a priori primary hypothesis neurological improvement in the treatment group (http://www.proneuron.com/ClinicalStudies/index.html). In the study from Lisbon,32 the spinal cord (C4-T6) was trimmed of scar tissue, and olfactory mucosa autografts were placed within the defect, with claims of motor improvement in each transplanted subject.32 Long-term follow-up in at least 1 subject demonstrated the presence of a mass containing mucoid cysts produced from respiratory epithelium 8 yr post-transplant.35 This emphasizes the complication of failing to transplant a purified population and the consequence of contamination from undesirable cells (ie, the respiratory epithelium), as well as the importance of long-term follow-up. Clinical Experience With Syringe Positioning Device Injections in Subacute to Chronic SCI The options of syringe stabilization include table-mounted syringe positioning devices, or patient anchored, retractor-based syringe positioning devices. Each of the techniques can be rigid or attached to a floating cannula. The potential disadvantages of the retractor-mounted or patient-anchored devices include the amount of tissue dissection required to anchor the device. This potentially could lead to spinal instability such as kyphosis. While preclinical large animal data are useful in establishing safe injection volumes, the assumptions of scaling upwards to the human spinal cord remain debatable. Cervical and thoracic injection into the normal or partially injured human spinal cord surrounding an SCI can give critical clues of safety that are not available with large animal studies. In particular, subtle changes in sensory function, radicular motor changes, and importantly pain, cannot be adequately interpreted in animal studies. Published trials reporting on the surgical experience and results of intramedullary injections testing a variety of cell types after human SCI have been performed. Autologous OEG cells prepared in cell culture (4-10 wk) were delivered into the thoracic spinal cord via multiple injections using a syringe positioning device.31,33 The Geron GRNOPC1 human ES cell-based trial utilized a table stabilized, single injection for 4 subjects, and did not report any AEs related to the surgery, injection, or injectate (http://ir.geron.com/phoenix.zhtml%3Fc=67323%26p=irol-newsArticle%26ID=1635760%26highlight). Transplantation of autologous human Schwann cells targeting the epicenter of a subacute thoracic SCI was successfully completed using a dose escalation strategy of 5, 10, and 15 M cells in 50, 100, and 150 μL with the Geron table mounted syringe positioning device (Anderson et al in press).36 The neural stem cell trial for amyotrophic lateral sclerosis used a patient anchored retractor based device24,37,38 with great precision for multiple 0.1 M neural stem cell (10 μL; 100 000 cell) injections targeting the anterior horn. Trial design for HuCNS-SC therapies for SCI In the HuCNS-SC thoracic and cervical trials, free-hand injection was selected for several reasons. Neurosurgeons with microsurgical training and experience can stabilize the syringe during the injection procedure. In the current clinical trial, 29 subjects with a total of 150 separate injections in the thoracic (48) or cervical (102) spinal cord were performed using hand stabilization using microsurgical technique. Given that no serious clinically or imaging detected-AEs related to the cell injection technique were observed, this represents a safe and feasible method of cell delivery in the future. A second important reason for choosing a free-hand injection technique without device is clinical scalability. This was established in the current protocol as demonstrated by the number of participating stem cell transplant centers and their lead surgeon (9) with very consistent intramedullary injection times. If cell transplantation for SCI becomes a widely used therapeutic option, the injection technique would need to be simple and practical to scale up among treating centers. The introduction of complex devices requiring approval from local agencies and extensive training would delay the deployment of such a therapeutic approach. In that, free-hand injections do not require additional muscle dissection of the spine to anchor a platform, minimal morbidity is added to a patient population who are prone to postlaminectomy instability given their history of prior instrumentation and paraspinal muscle atrophy as a result of the SCI. Furthermore, the free-hand injection allows for compensation of potential patient movements and respiratory variations during surgery. While spinal cord motion was no concern in the 29 cases, subtle patient position changes can be compensated by the surgeon's hand resting on the patient. An important aspect is the shorter operative time requirements to complete multiple (up to 8) injections by obviating the anchoring of spine- or table-mounted devices and positioning the needle in 3 planes over the spinal cord adds inherent safety concerns to the procedure. Finally, in all cases, stem cell injections were possible without removal or the addition of spinal instrumentation and in no cases were spinal deformities such as kyphosis observed. Safety data gathered from small and large animal studies including tolerable level of cell dose, injection location and number, and cell suspension volumes based on neurological exam and neuropathology have guided the dose and delivery method for first-in-human studies. Allometric scale-up to humans can be calculated and an additional measure of safety can be provided by initial studies targeting reduced cell volumes below the equivalent tested in animals. The human cell dose in the thoracic study was based on the experimental murine cord contusion model.13 The extrapolated calculation of the average volume of cord injected in the midthoracic murine model (total cord injection volume of 6.8 mm3) to the equivalent of the human cord (total cord injection volume of 3140 mm3) yielded a translated human dose of 35 million cells. A further margin of safety in the first in-human administration was created by targeting a 60% dose administered in 4 equally divided microinjections totaling 20 million cells in 280 μL, for the thoracic study. Based on the experience from the phase I/II thoracic study, the overall risk related to HuCNS-SC injection was considered low, and the neurosurgical feedback indicated that the spinal cord had the capacity to tolerate higher cell doses and volume. This was the rationale for a predicted therapeutic target dose of 40 million cells in the phase II cervical study. The first arm of the phase II study allowed for incremental dose escalation starting at 15 million cells in 2 cervical subjects followed by 30 and 40 million cells in 2 additional subjects at each dose. Each dose level was divided into 4, 6, and 8 microinjections, respectively, administered bilaterally in a pattern rostral and caudal to the epicenter of the spinal cord lesion. The distribution over 8 injections was also intended to enhance intramedullary administration by increasing the anatomic distribution of the cells at the injury margins. The final 40 million cell dose was determined based on a priori dose stopping and reduction rules for safety and tolerability. The observed safety summary with HuCNS-SC and SCI reveals that the AEs are consistent with expected similar surgeries and immunosuppression regimens. No AEs were attributed to the injection technique or injectate by the investigators. In the thoracic study, moderate back and mild postoperative pain were described as musculoskeletal or complications of the surgical procedure. Of the AEs that were related to the CNS (increased spasticity, headache, dizziness, and meningocele), the injection technique or injectate might have the potential to have an impact on spasticity given that spasticity results from upper motor neuron lesions.39 However, changes in muscle properties, increased supraspinal descending inputs (eg, postsurgical pain, swelling, and/or other intercurrent issues such as UTI, constipation, skin irritations, or physical discomfort) can also contribute to spasticity. Given the mild nature and quick resolution of this instance of increased spasticity and lack of any other associated sensorimotor sequelae, it was likely caused by factors not related to the injection technique or injectate. Similarly, pain reported in subjects in the cervical study was musculoskeletal or procedural in origin. There was an instance of paresthesia described in the fifth metacarpal on the left hand immediately after surgery, which resolved 14 d later. In comparison to other completed cell-based SCI studies, the current trial addresses safety of perilesional injections, including delivery of a larger cell suspension volume in an area above the injury epicenter and in the chronically injured patient where the neurological status is stable. MRIs of chronic SCI reveal Wallerian degeneration of the dorsal column above40 and less so in the lateral column below the epicenter of the primary injury site over time points up to 4 yr postinjury.40,41 The dorsal column target point in the setting of SCI pathology provides the safety rationale for the current injection technique and supports the favorable AE profile as exhibited in both thoracic and cervical studies (Figure 6). The migratory properties of HuCNS-SC cells theoretically preclude the surgical requirement of a precise and more eloquent intramedullary anatomic target (ie, the lateral corticospinal tracts responsible for descending motor control). The Importance of Intraoperative Ultrasound Intraoperative ultrasound has been an important tool in neurosurgery ORs for more than 50 yr.42 In spinal cord surgery, its use is well recognized in the identification of intradural tumors, determining the adequacy of decompression after removal of bone from trauma and in the treatment of post-traumatic cysts or syringomyelia. Intraoperative ultrasonography has been used to determine the extent of tumor resection and found to be 92% sensitive.43 In addition, the average time of 7 min spent for intraoperative ultrasonography assessment,43 makes it reliable, practical, and highly sensitive for spinal cord surgery.43-45 In the current study, the use of ultrasound assisted in confirming the amount of lamina needed to be removed to adequately expose the post-traumatic cyst, and was particularly helpful in defining the caudal and rostral regions of the cyst so that the appropriate location of injection could be determined, particularly because external or pial surface evidence of the injury epicenter was subtle and indistinct. The variability in lesion cyst size and appearance was very heterogeneous in both the cervical and thoracic trials. In 2 of the 12 cases, in which pericyst thoracic injections were performed, a small area of high signal was seen in the thoracic spinal cord that was felt to represent the cell bolus. The presumed rapid spread of the cell bolus, as suggested in 2 subsequent images taken 5 min apart, indicates that cell distribution can happen early after injection along white matter tracts. In postop MRIs in the cervical study, one could visualize the cell injectate above and below the lesion epicenter in at least half of the cases. The T2-weighted signal changes were felt to represent the relatively large volume cell injectate as opposed to reactive edema or spinal cord infarction. The signal after cell injections seen in the cervical spinal cord and its absence in the thoracic injections is likely multifactorial and includes (1) larger cell dose/volumes in the cervical study, (2) better visualization of the cervical spinal cord due its larger size, (3) proximity to the magnet, and (4) relatively smaller magnitude of metal instrumentation used in cervical fusion cases. CONCLUSION In conclusion, we have demonstrated a surgical approach and free-hand technique for HuCNS-SC administration after chronic SCI. Injection of HuCNS-SC into perilesional tissues above and below the thoracic and cervical SCI using a free-hand technique demonstrates an excellent safety profile. While various methods for cell delivery are under study, our surgical experience with a free-hand technique appears to be well-tolerated, feasible, and scalable for larger clinical trials. Further studies will be needed to investigate and confirm biological activity and clinical efficacy. Disclosures This study was funded by Stem Cells, Inc and the respective Academic Institutions. Dr Allan D. Levi receives teaching honorarium from Medtronic and grant support from the Department of Defense. Dr Paul Park is a consultant with Globus, Medtronic, Zimmer, and NuVasive. He receives royalties from Globus and grant support from Pfizer. Dr Michael Fehlings is a consultant with In Vivo Therapeutics. Dr Kim Anderson is a consultant for Vertex Inc. Drs Allyson Gage and Stephen Huhn are former employees of Stem Cells, Inc, and Jane Hseih is a former consultant with Stem Cells Inc. REFERENCES 1. Wyndaele M, Wyndaele JJ. Incidence, prevalence and epidemiology of spinal cord injury: what learns a worldwide literature survey? Spinal Cord . 2006; 44( 9): 523- 529. Google Scholar CrossRef Search ADS PubMed 2. Erceg S, Ronaghi M, Oria M et al. Transplanted oligodendrocytes and motoneuron progenitors generated from human embryonic stem cells promote locomotor recovery after spinal cord transection. Stem Cells . 2010; 28( 9): 1541- 1549. Google Scholar CrossRef Search ADS PubMed 3. Harper JM, Krishnan C, Darman JS et al. Axonal growth of embryonic stem cell-derived motoneurons in vitro and in motoneuron-injured adult rats. Proc Natl Acad Sci USA . 2004; 101( 18): 7123- 7128. Google Scholar CrossRef Search ADS PubMed 4. Kerr CL, Letzen BS, Hill CM et al. Efficient differentiation of human embryonic stem cells into oligodendrocyte progenitors for application in a rat contusion model of spinal cord injury. Int J Neurosci . 2010; 120( 4): 305- 313. Google Scholar CrossRef Search ADS PubMed 5. Marques SA, Almeida FM, Fernandes AM et al. Predifferentiated embryonic stem cells promote functional recovery after spinal cord compressive injury. Brain Res . 2010; 1349 (Aug 19): 115- 128. Google Scholar CrossRef Search ADS PubMed 6. Nistor GI, Totoiu MO, Haque N, Carpenter MK, Keirstead HS. Human embryonic stem cells differentiate into oligodendrocytes in high purity and myelinate after spinal cord transplantation. Glia . 2005; 49( 3): 385- 396. Google Scholar CrossRef Search ADS PubMed 7. Rossi SL, Nistor G, Wyatt T et al. Histological and functional benefit following transplantation of motor neuron progenitors to the injured rat spinal cord. PLoS One . 2010; 5( 7): e11852. Google Scholar CrossRef Search ADS PubMed 8. Salehi M, Pasbakhsh P, Soleimani M et al. Repair of spinal cord injury by co-transplantation of embryonic stem cell-derived motor neuron and olfactory ensheathing cell. Iran Biomed J . 2009; 13( 3): 125- 135. Google Scholar PubMed 9. Hall A, Giese NA, Richardson WD. Spinal cord oligodendrocytes develop from ventrally derived progenitor cells that express PDGF alpha-receptors. Development . 1996; 122( 12): 4085- 4094. Google Scholar PubMed 10. Pringle NP, Guthrie S, Lumsden A, Richardson WD. Dorsal spinal cord neuroepithelium generates astrocytes but not oligodendrocytes. Neuron . 1998; 20( 5): 883- 893. Google Scholar CrossRef Search ADS PubMed 11. Woodruff RH, Tekki-Kessaris N, Stiles CD, Rowitch DH, Richardson WD. Oligodendrocyte development in the spinal cord and telencephalon: common themes and new perspectives. Int J Dev Neurosci . 2001; 19( 4): 379- 385. Google Scholar CrossRef Search ADS PubMed 12. Cummings BJ, Uchida N, Tamaki SJ, Anderson AJ. Human neural stem cell differentiation following transplantation into spinal cord injured mice: association with recovery of locomotor function. Neurol Res . 2006; 28( 5): 474- 481. Google Scholar CrossRef Search ADS PubMed 13. Cummings BJ, Uchida N, Tamaki SJ et al. Human neural stem cells differentiate and promote locomotor recovery in spinal cord-injured mice. Proc Natl Acad Sci USA . 2005; 102( 39): 14069- 14074. Google Scholar CrossRef Search ADS PubMed 14. Hooshmand MJ, Sontag CJ, Uchida N, Tamaki S, Anderson AJ, Cummings BJ. Analysis of host-mediated repair mechanisms after human CNS-stem cell transplantation for spinal cord injury: correlation of engraftment with recovery. PLoS One . 2009; 4( 6): e5871. Google Scholar CrossRef Search ADS PubMed 15. Salazar DL, Uchida N, Hamers FP, Cummings BJ, Anderson AJ. Human neural stem cells differentiate and promote locomotor recovery in an early chronic spinal cord injury NOD-scid mouse model. PLoS One . 2010; 5( 8): e12272. Google Scholar CrossRef Search ADS PubMed 16. Selden NR, Al-Uzri A, Huhn SL et al. Central nervous system stem cell transplantation for children with neuronal ceroid lipofuscinosis. J Neurosurg Pediatr . 2013; 11( 6): 643- 652. Google Scholar CrossRef Search ADS PubMed 17. Gupta N, Henry RG, Strober J et al. Neural stem cell engraftment and myelination in the human brain. Sci Transl Med . 2012; 4( 155): 155ra137. Google Scholar CrossRef Search ADS PubMed 18. Lamanna JJ, Miller JH, Riley JP, Hurtig CV, Boulis NM. Cellular therapeutics delivery to the spinal cord: technical considerations for clinical application. Ther Deliv . 2013; 4( 11): 1397- 1410. Google Scholar CrossRef Search ADS PubMed 19. Piltti KM, Salazar DL, Uchida N, Cummings BJ, Anderson AJ. Safety of human neural stem cell transplantation in chronic spinal cord injury. Stem Cells Transl Med . 2013; 2( 12): 961- 974. Google Scholar CrossRef Search ADS PubMed 20. Parent A. Spinal Cord. In: Coryell P, ed. Carpenter's Human Neuroanatomy . 9th ed. Pennsylvania: Williams & Wilkins; 1996: 325- 405. 21. Federici T, Hurtig CV, Burks KL et al. Surgical technique for spinal cord delivery of therapies: demonstration of procedure in gottingen minipigs. J Vis Exp . 2012; 70 (Dec 7): e4371. 22. Kwon BK, Streijger F, Hill CE et al. Large animal and primate models of spinal cord injury for the testing of novel therapies. Exp Neurol . 2015; 269 (Jul): 154- 168. Google Scholar CrossRef Search ADS PubMed 23. Raore B, Federici T, Taub J et al. Cervical multilevel intraspinal stem cell therapy: assessment of surgical risks in Gottingen minipigs. Spine (Phila Pa 1976) . 2011; 36( 3): E164- E171. Google Scholar CrossRef Search ADS PubMed 24. Riley J, Federici T, Park J et al. Cervical spinal cord therapeutics delivery: preclinical safety validation of a stabilized microinjection platform. Neurosurgery . 2009; 65( 4): 754- 761; discussion 761-752. Google Scholar CrossRef Search ADS PubMed 25. Myers SA, Bankston AN, Burke DA, Ohri SS, Whittemore SR. Does the preclinical evidence for functional remyelination following myelinating cell engraftment into the injured spinal cord support progression to clinical trials? Exp Neurol . 2016; 283( pt B): 560- 572. Google Scholar CrossRef Search ADS PubMed 26. Cabraja M, Stockhammer F, Mularski S, Suess O, Kombos T, Vajkoczy P. Neurophysiological intraoperative monitoring in neurosurgery: aid or handicap? An international survey. Neurosurg Focus . 2009; 27( 4): E2. Google Scholar CrossRef Search ADS PubMed 27. Takahashi Y, Tsuji O, Kumagai G et al. Comparative study of methods for administering neural stem/progenitor cells to treat spinal cord injury in mice. Cell Transplant . 2011; 20( 5): 727- 739. Google Scholar CrossRef Search ADS PubMed 28. Parr AM, Tator CH, Keating A. Bone marrow-derived mesenchymal stromal cells for the repair of central nervous system injury. Bone Marrow Transplant . 2007; 40( 7): 609- 619. Google Scholar CrossRef Search ADS PubMed 29. Pearse DD, Pereira FC, Marcillo AE et al. cAMP and Schwann cells promote axonal growth and functional recovery after spinal cord injury. Nat Med . 2004; 10( 6): 610- 616. Google Scholar CrossRef Search ADS PubMed 30. Guest J, Benavides F, Padgett K, Mendez E, Tovar D. Technical aspects of spinal cord injections for cell transplantation. Clinical and translational considerations. Brain Res Bull . 2011; 84( 4-5): 267- 279. Google Scholar CrossRef Search ADS PubMed 31. Lima C, Pratas-Vital J, Escada P, Hasse-Ferreira A, Capucho C, Peduzzi JD. Olfactory mucosa autografts in human spinal cord injury: a pilot clinical study. J Spinal Cord Med . 2006; 29( 3): 191- 203; discussion 204-196. Google Scholar CrossRef Search ADS PubMed 32. Mackay-Sim A, Feron F, Cochrane J et al. Autologous olfactory ensheathing cell transplantation in human paraplegia: a 3-year clinical trial. Brain . 2008; 131( pt 9): 2376- 2386. Google Scholar CrossRef Search ADS PubMed 33. Yoon SH, Shim YS, Park YH et al. Complete spinal cord injury treatment using autologous bone marrow cell transplantation and bone marrow stimulation with granulocyte macrophage-colony stimulating factor: phase I/II clinical trial. Stem Cells . 2007; 25( 8): 2066- 2073. Google Scholar CrossRef Search ADS PubMed 34. Lammertse DP, Jones LA, Charlifue SB et al. Autologous incubated macrophage therapy in acute, complete spinal cord injury: results of the phase 2 randomized controlled multicenter trial. Spinal Cord . 2012; 50( 9): 661- 671. Google Scholar CrossRef Search ADS PubMed 35. Dlouhy BJ, Awe O, Rao RC, Kirby PA, Hitchon PW. Autograft-derived spinal cord mass following olfactory mucosal cell transplantation in a spinal cord injury patient: Case report. J Neurosurg Spine . 2014; 21( 4): 618- 622. Google Scholar CrossRef Search ADS PubMed 36. Anderson KD, Guest JD, Dietrich WD et al. Safety of autologous human schwann cell transplantation in sub-acute thoracic spinal cord injury. J Neurotrauma . Mar 21, 2017. In press. 37. Glass JD, Boulis NM, Johe K et al. Lumbar intraspinal injection of neural stem cells in patients with amyotrophic lateral sclerosis: results of a phase I trial in 12 patients. Stem Cells . 2012; 30( 6): 1144- 1151. Google Scholar CrossRef Search ADS PubMed 38. Mazzini L, Gelati M, Profico DC et al. Human neural stem cell transplantation in ALS: initial results from a phase I trial. J Transl Med . 2015; 13( 1): 17. Google Scholar CrossRef Search ADS PubMed 39. Pandyan AD, Gregoric M, Barnes MP et al. Spasticity: clinical perceptions, neurological realities and meaningful measurement. Disabil Rehabil . 2005; 27( 1-2): 2- 6. Google Scholar CrossRef Search ADS PubMed 40. Lee TK, Yoon SH, KIM Y et al. MRI finding of spinal cord injury with wallerian degeneration. Eur Cong Radiol . 2015. 41. Terae S, Taneichi H, Abumi K. MRI of wallerian degeneration of the injured spinal cord. J Comput Assist Tomogr . 1993; 17( 5): 700- 703. Google Scholar CrossRef Search ADS PubMed 42. Dohrmann GJ, Rubin JM. History of intraoperative ultrasound in neurosurgery. Neurosurg Clin N Am . 2001; 12( 1): 155- 166, ix. Google Scholar PubMed 43. Parisini P, Bettini N, Palmisani M et al. Intraoperative ultrasonography imaging in spinal surgery (technique and indications). Chir Organi Mov . 1992; 77( 2): 187- 194. Google Scholar PubMed 44. Quencer RM, Montalvo BM, Eismont FJ, Green BA. Intraoperative spinal sonography in thoracic and lumbar fractures: evaluation of Harrington rod instrumentation. AJR Am J Roentgenol . 1985; 145( 2): 343- 349. Google Scholar CrossRef Search ADS PubMed 45. Toktas ZO, Sahin S, Koban O, Sorar M, Konya D. Is intraoperative ultrasound required in cervical spinal tumors? A prospective study. Turk Neurosurg . 2013; 23( 5): 600- 606. Google Scholar PubMed Acknowledgments We sincerely thank the multiple clinical care coordinators, rehabilitation, neuroradiology, immune suppression, anesthesia, back up and blinded physicians and nurses who participated in the study. In addition, we are grateful to the SCI patients around the world who participated in the current trials. We are greatly indebted to Linda Alberga for her editorial assistance and Katie Gant for her assistance with tables. COMMENTS This study details the safety and feasibility of multiple intramedullary, peri-lesional injections of human central nervous system stem cells after chronic spinal cord injury. The authors focus on safety, adverse events, and standardized technique, and allude to separate publications about the clinical outcomes and effectiveness of the injections. They have made a valuable addition to the burgeoning field of therapeutic stem cell-approaches to spinal cord injury. Focusing on a reproducible free-hand method to inject the cells has clear advantages such as simplicity of delivery and wider adoption, including in resource-limited regions. The report is a necessary step in developing a protocol for cell delivery in a large scale clinical trial for patients with few available therapeutic options. Alexander E. Ropper Houston, Texas This manuscript is important since it presents a safe, reproducible, and easy method for the free-hand delivery of cells to the chronically injured spinal cord. Using a standardized protocol, from preoperative delineation of injection site, surgical approach, intraoperative adjuncts (ultrasonography, electrophysiology, dural substitutes, and sealants) and postoperative imaging, the authors were able to demonstrate safety of an escalating dose of injected HuCNS-SC to the injured cervical and thoracic spinal cords of patients with chronic SCI, ASIA A/B. Although the authors argue that the use of intraoperative ultrasound is critical, it's not clear if this modality is necessary or helpful for localization and safety of the procedure. Other clinical trials of cell transplantation for thoracic and cervical subacute SCI and degenerative CNS disease utilize only preoperative MR guidance, with a similar safety profile. It's also not known if this free-hand technique is preferable to other approaches currently being used clinically, which employ a specially designed positioning and stabilization device attached to the patient's bed or anchored to the patient's body, but still with a hand injection. As the authors point out, there are still many unanswered questions including the best cell type, number of cells, number of injections, location of intraparenchymal stem cell injection, and timing of transplant following SCI to achieve both safety and optimal efficacy. It's unfortunate that the cervical SCI study and future development of these cells was terminated when the magnitude of improvement in cohort I at 1 year and an interim analysis of cohort 2 at 6 months fell below the required clinical efficacy threshold set by the sponsor. This is a problem in the emerging cellular therapy field, when companies make business decisions sometimes based primarily on financial issues, and prematurely abandon a cell that might have promise if evaluated more extensively. Given the potential for future cell-based therapies, establishing a consistent and safe protocol for delivery of such cells is paramount. Furthermore, an efficient and scalable way to deliver these cells free-handed may allow for wider-use of cell-based therapies for those with SCI. The authors are to be congratulated for taking a leadership role in pioneering stem cell therapy for the treatment of SCI. Maziyar Kalani Gary K. Steinberg Stanford, California Copyright © 2017 by the Congress of Neurological Surgeons
Neurosurgery – Oxford University Press
Published: Apr 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera