Thehttp://nro.sagepub.com/ Neuroscientist
Cell Therapy for Spinal Cord Injuries: What Is Really Going on? Nicolas Granger, Robin J.M. Franklin and Nick D. Jeffery Neuroscientist published online 10 January 2014 DOI: 10.1177/1073858413514635 The online version of this article can be found at: http://nro.sagepub.com/content/early/2014/01/09/1073858413514635
Published by: http://www.sagepublications.com
Additional services and information for The Neuroscientist can be found at: Email Alerts: http://nro.sagepub.com/cgi/alerts Subscriptions: http://nro.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Jan 10, 2014 What is This?
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
514635
research-article2014
NROXXX10.1177/1073858413514635The NeuroscientistGranger and others
Article
Cell Therapy for Spinal Cord Injuries: What Is Really Going on?
The Neuroscientist 1–16 © The Author(s) 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1073858413514635 nro.sagepub.com
Nicolas Granger1, Robin J.M. Franklin2, and Nick D. Jeffery3
Abstract During the last two decades, many experiments have examined the ability of cell transplants to ameliorate the loss of function after spinal cord injuries, with the hope of developing interventions to benefit patients. Although many reports suggest positive effects, there is growing concern over the quality of the available preclinical data. It is therefore important to ask whether this worldwide investigative process is close to defining a cell transplant protocol that could be translated into human patients with a realistic chance of success. This review systematically examines the strength of the preclinical evidence and outlines mechanisms by which transplanted cells may mediate their effects in spinal cord injuries. First, we examined changes in voluntary movements in the forelimb associated with cell transplants after partial cervical lesions. Second, we examined the efficacy of transplanted cells to restore electrophysiological conduction across a complete thoracic lesion. We postulated that cell therapies found to be successful in both models could reasonably have potential to treat human patients. We conclude that although there are data to support a beneficial effect of cell transplantation, most reports provide only weak evidence because of deficits in experimental design. The mechanisms by which transplanted cells mediate their functional effects remain unclear. Keywords transplantation, animal model, functional, forelimb reaching, transection The seminal nerve transplantation studies by Aguayo and others in the 1980s (Aguayo and others 1981; David and Aguayo 1981; Richardson and others 1984; Zhou and others 1986) spurred a new wave of interest in the effects of cell transplantation in the CNS. In part, this was motivated by scientific curiosity regarding the effects of cells on the host, and vice versa, and the precise mechanisms underlying those effects. However, another clear rationale is the hope that transplanted cells may be able to repair injuries to the nervous system and, in particular, provide benefits to patients who have suffered spinal cord injuries (SCIs). Despite three decades of laboratory studies on cell transplantation and although cell transplantation for human SCIs has now commenced in many countries, including the USA (e.g., see http://www.miamiproject.miami.edu/page.aspx?pid=339), it currently remains unclear whether any type of cell transplant can be conclusively considered likely to be of benefit in clinical SCIs. Here, we critically review the literature on cell transplantation in the spinal cord of laboratory animals with the aim of defining the strength of evidence to support translation into human patients. There is a plethora of published material describing the effects of many different cell transplants in various types of SCIs, meaning that a review can become unwieldy in the absence of clear parameters defining its
scope. Therefore, here, we restrict our review to an examination of the published evidence regarding efficacy in restoring “function” after SCIs to focus on the likelihood of translation of functional benefits to human recipients. For that reason, we considered first that a translational candidate cell type should be associated with improvement in voluntary movement. This is, of course, difficult to ascertain definitely in nonverbal species. However, we considered that improvement in animal activities such as grooming or use of the limbs to perform actions in accordance with an animal’s “wants” rather than simply reflecting changes in reflex activity (e.g., such as pelvic limb stepping) would provide a reasonable substitute. Second, we considered that it would be necessary for the same cell type to be effective in a complete lesion. This is because 1
School of Veterinary Sciences, University of Bristol, Bristol, UK Wellcome Trust-Medical Research Council Cambridge Stem Cell Institute and Department of Veterinary Medicine, University of Cambridge, Cambridge, UK 3 Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Iowa State University, Ames, IA, USA 2
Corresponding Author: Nick D. Jeffery, Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Iowa State University, 1600 South 16th Street, Ames, IA 50011, USA. Email: [email protected]
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
2
The Neuroscientist XX(X)
Figure 1. The review of the literature was divided into two parts to assess the strength of the available data on the possible benefits of cell transplantation after spinal cord injuries. (A) First, we reviewed studies which examined the recovery of goaldirected forelimb movement after partial cervical lesions. (B) Second, we reviewed whether there was evidence that cell types applied in cervical lesions were also effective in restoring electrophysiological conduction in a spinal cord transection model.
patients who are most likely to be candidates for cell therapy trials will be those with very severe lesions, as they have the least function to lose and the ability to detect beneficial effects is greater (Fawcett and others 2007). Recovery through plasticity is of course beneficial but cannot be expected to be efficacious alone in anatomically complete lesions. In an effort to address these two aspects, this systematic review is divided into two parts (Fig. 1): first, we examined publications showing evidence of improvement in goaldirected function after cells were transplanted into cervical SCIs. Cervical injuries were necessary because of the need to examine “voluntary” function. Second, limiting the review to cell types discussed in the first part, we examined experiments in which the effects of such cells have also been studied using electrophysiological methods after
transplantation in complete transection lesions in the thoracolumbar spinal cord. We focused the analysis on electrophysiological examination to allow unequivocal differentiation between regeneration and the recovery of reflex actions, which would not be possible through behavioral observations alone. This part of the review therefore pertains only to thoracolumbar lesions because complete transection of the cervical cord is incompatible with reasonable animal survival and experimental ethics. We also review the quality of the evidence provided by these studies with reference to the guidelines provided for clinical trials, the CONSORT (Consolidated Standards of Reporting Trials) statement regarding clinical trials (http:// www.consort-statement.org/consort-statement/), and those proposed by Dobkin (2010) to report clinical data and by Landis and others (2012) to report experimental results.
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
3
Granger and others
Analyzing the Efficacy of Cell Therapy Transplants in SCI Functional Outcome after Cell Transplantation in the Cervical Spinal Cord Most (n = 22) of the studies examining volitional functional outcomes after cell transplant in incomplete cervical spinal cord lesions (of a total of 25 that we identified) were carried out in rodents, and an additional three were performed in marmosets (Table 1). The reported outcomes (Table 2) focused on forepaw tasks: to reach food (n = 10 studies), remove tape from the paw (n = 4), or groom (n = 1). In 10 other studies, the analysis concentrated on tasks in which the animal had to extend the forelimb to explore the environment (vertical exploration: n = 5; climbing: n = 5). Cell Transplants Associated with Beneficial Effects. In 17 of the 25 studies, a positive effect of the transplant on the recovery of function was claimed. Five of nine studies on olfactory ensheathing cells (OECs) reported a positive effect; of these, the cells were derived from the olfactory mucosa in one study and from the olfactory bulb in four. Neural stem/ progenitor cells were associated with a positive effect in six of six reports. Mesenchymal stem cells (MSCs) were tested in four studies, of which none found a functional benefit. In the remaining six studies claiming a positive effect, the cells tested were olfactory neuroepithelial neurosphere-forming cells (n = 2 from one laboratory), BDNFexpressing fibroblasts (n = 2 from one laboratory), and Schwann cells (n = 2 from two laboratories). Functional Effect of OECs Studies in which OECs were associated with benefits. Four studies used a directed forepaw reaching task (KeyvanFouladi and others 2003; Li and others 1997; Nash and others 2002; Yamamoto and others 2009). In three studies, the dorsal corticospinal tract was unilaterally (KeyvanFouladi and others 2003; Li and others 1997; Yamamoto and others 2009) or bilaterally (Nash and others 2002) destroyed. The OEC transplants were considered to have a large beneficial effect because some transplanted animals regained near normal function (Table 2). However, the results have limitations because blinding and allocation concealment were not reported and statistical analysis was lacking, and one study (Li and others 1997) did not include a control group. Importantly, in one report (Yamamoto and others 2009), the transplants contained only 5% OECs and 95% fibroblasts, and although the recovery of forepaw reaching was observed, no axonal growth across the lesion was detected. In the fourth study using OECs and the forepaw reaching test (Nash and others 2002), a large improvement in the forepaw reaching task was
detected (success rate at reaching food pellets at six weeks after injury was 71%–78% in the cell transplant group [n = 19] v. 38%–42% in vehicles and control lesion rats). The study was blinded, but allocation concealment was not reported; moreover, 15 animals were excluded because they died within 14 days of surgery, and it is not possible to know to which group they belonged. One study (Moreno-Flores and others 2006) used OECs and found a beneficial effect during a tape removal task; this assessment was conducted blinded, but allocation concealment was not reported. Studies on OEC transplantation in which no effect was reported. In four studies, OECs had no reported effect on thoracic limb recovery (Bretzner and others 2008; Bretzner and others 2010; Collazos-Castro and others 2005; Ruitenberg and others 2005). Collazos-Castro and others (2005) reported results in the treatment of relatively mild lesions from which control animals were able to recover substantial function in the absence of any intervention, which might account for the difficulty in detecting an effect of transplanted OECs. Bretzner and others (2008) found no motor improvement during forepaw reaching or vertical exploration tests. The lesions (unilateral dorsolateral funiculus section) were much larger than those in studies in which a positive effect of OECs was found. Further, both studies used immunosuppression because the authors transplanted mice OECs into rats, which could be a confounding factor. A reduced sensory threshold in animals receiving OECs (thermal hypersensitivity) was also detected in these studies and constitutes a potential detrimental effect of OEC transplantation. The control rats used by Ruitenberg and others (2005) recovered function to preoperative levels, rendering the detection of an intervention effect difficult. It was also one of the two studies (Lu and others 2012b; Ruitenberg and others 2005) in which treatment allocation was concealed, which is known to reduce the magnitude of the treatment effect through a reduction in bias. Studies in Which a Positive Effect of Neural Stem/Progenitor Cells Was Reported. In two studies using a forepaw reaching task, one using motor neuron progenitors (Rossi and others 2010) found that transplanted rats could retrieve a mean of approximately 8 pellets versus approximately 3 in the vehicle group, and the second, using neural stem cells (NSCs) (Ogawa and others 2002), found that 13 of 15 rats ate >5 pellets versus 8 of 13 control rats who ate >5 pellets. Although both are interpreted as positive results, it is difficult to know how these two different end points can be compared. Both these two reports involve a contusive injury and cells transplanted at a chronic stage. Rossi and others (2010) did not report the results of the nontransplanted animals, allocation to experimental
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
4
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
Cell type
F Fischer 344 rats BDNF-expressing bone marrow stroma cells
Lu and others 2012a
Mice olfactory mucosa OECs
M SpragueDawley rats
F Sprague-Dawley Human motor neuron rats progenitors F Swiss albino rats Olfactory bulb OECs and ONFs F Swiss albino rats Olfactory bulb OECs and ONFs F Swiss albino rats Schwann cells and fibroblasts F Swiss albino rats Olfactory mucosa OECs F Sprague-Dawley BDNF-expressing rats fibroblasts F Sprague-Dawley BDNF-expressing rats fibroblasts F Sprague-Dawley Human bone marrow rats stroma cells F Sprague-Dawley Human olfactory rats neuroepithelial neurosphere-forming cells F Sprague-Dawley Human olfactory rats neuroepithelial neurosphere-forming cells M SpragueMice olfactory mucosa Dawley rats OECs
F Sprague-Dawley NSCs as neurospheres rats M Wistar rats Olfactory bulb OECs and ONFs F Fischer rats Schwann cells
Species
Bretzner and others 2010
Bretzner and others 2008
Xiao and others 2007
Ogawa and others 2002 Collazos-Castro and others 2005 Schaal and others 2007 Rossi and others 2010 Li and others 1997 Keyvan-Fouladi and others 2003 Keyvan-Fouladi and others 2005 Yamamoto and others 2009 Liu and others 1999 Schwartz and others 2003 Neuhuber and others 2005 Xiao and others 2005
Reference
Vertical exploration
Immediate Immediate Immediate 7 days
Immediate
Immediate
C3/C4 hemisection C3/C4 hemisection C3/C4 hemisection
C3/C4 hemisection
C4/C5 hemisection
C5 hemisection
Immediate
Immediate
Direct forepaw reaching
8 weeks
C4/C5 hemisection
Direct forepaw reaching
Immediate
Direct forepaw reaching, vertical exploration, thermal and mechanical threshold Direct forepaw reaching, vertical exploration, thermal and mechanical threshold Forelimb grooming task, grid walk, platform locomotion
Vertical exploration
Vertical exploration, BBB score, thermal threshold Vertical exploration, horizontal rope walking
Vertical exploration, BBB score
Direct forepaw reaching
8 weeks
Forearm grip, inclined plane test, BBB score, forelimb hanging Direct forepaw reaching, beam walking, forearm grip Direct forepaw reaching
Forelimb kinematics during walking
Direct forepaw reaching
Behavioral tests
C1/C2 electrolytic CST lesion C1/C2 electrolytic CST lesion C1/C2 electrolytic CST lesion C1/C2 electrolytic CST lesion C3/C4 hemisection
Immediate
Immediate
C5 contusion 7 days
Immediate
C6/C7 contusion
C5/C6 contusion
9 days
Time of transplant after lesion
C4/C5 contusion
Injury paradigm
NS
Yesa
Yes
NS
Yesa
Yes
NS
NS
Yes
Yes
NS
NS
Yesa Yes
NS
NS
NS
NS
NS
NS
NS
NS
NS
Reported allocation concealment
NS
NS
NS
NS
NS
NS
Yes
NS
NS
Blinding
Table 1. Summary of Studies that Reported Analysis of Movement after Cell Transplantation in Incomplete Cervical Lesions.
(continued)
24 cell group (± growth factor) v. 21 vehicle (of various forms ± growth factor) 48 cell group ± growth factor (4 groups of 12) v. 12 no cells but growth factor
7 cell group v. 17 vehicle or growth factor
30 cell group v. 18 vehicle v. 4 control
12 cell group v. 12 vehicle
6 cell group v. 6 cell group + growth factor v. 5 vehicle 6 cell group + growth factor v. 8 cell group v. 4 normal 32 cell group v. 8 vehicle v. 10 control
6 cell group v. 6 lesion v. 2 vehicle
18 cell group v. 12 historical cell group
12 cell group v. 2 vehicle v. 6 lesion
7 cell group v. 28 lesion
15 cell group v. 15 vehicle
14 cell group v. 13 lesion v. 16 sham
15 cell group v. 13 lesion v. 17 vehicle v. 10 control 14 cell group v. 10 vehicle
No. of animals used for functional analysis
5
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
Cell type
F common marmosets
Human induced pluripotent stem cells and NSCs
Sprague-Dawley Olfactory bulb OECs rats and ONFs F Fischer 344 rats NT-3–expressing olfactory bulb OECs and ONFs F common Human NSCs as marmosets neurospheres F common Human NSCs as marmosets neurospheres
F Fischer 344 rats Bone marrow stroma cells F Fischer 344 rats BDNF-expressing bone marrow stroma cells M Wistar rats Immortalized olfactory bulb OECs M Fischer 344 rats Neural progenitor cells
Species
Direct forepaw reaching, BBB score
Immediate
Immediate
C4 DC and CST transection C3 DC and CST transection C4 DC and CST hemisection 9 days 9 days
9 days
C5 contusion C5 contusion
C5 contusion
Immediate
Tape removal, beam walking, grid walking, wire ladder walking Direct forepaw reaching
Immediate
C3 DC crush
Forearm grip, treadmill and spontaneous locomotion (custom-designed score) Forearm grip, treadmill and spontaneous locomotion (custom-designed score)
Forelimb kinematics, forearm grip
Tape removal, beam walking
Immediate
C3 DC transection
Tape removal, horizontal ladder locomotion, rope walking Tape removal, rope walking
Behavioral tests
Immediate
Time of transplant after lesion
C4 DC transection
Injury paradigm
NS NS
NS
NS
NS
Yes
NS
NS
Yes: BBB score; no: direct forepaw reaching NS
Yes
NS
NS
NS
Yesa Yes
NS
Reported allocation concealment
Yes
Blinding
5 cell group v. 4 vehicle
7 cell group v. 7 vehicle v. 2 sham
5 cell group v. 5 vehicle
19 cell group (± drug) v. 5 drug group v. 5 sham v. 5 vehicle v. 5 lesion 21 cell group (± growth factor) v. 6 vehicle
5 groups of 12 receiving cells ± growth factor 6 cell group v. 6 cell group + growth factor 20 cell group v. 12 vehicle v. 6 sham v. 6 control 16 cell group v. 16 vehicle v. 10 sham
No. of animals used for functional analysis
The studies are grouped according to the method of injury and listed in chronological order. BBB = Basso, Beattie, and Bresnahan; CST = corticospinal tract; DC = dorsal column; DLF = dorsolateral funiculus; NS = not specified; NSCs = neural stem cells; OECs = olfactory ensheathing cells; ONFs = olfactory nerve fibroblasts. If not specified, the species of the origin of cells is rat. a. Blinded scoring from videos without mention of videos being obtained blinded.
Kobayashi and others 2012
Iwanami and others 2005 Yamane and others 2010
Lu and others 2004 Lu and others 2005 Moreno-Flores and others 2006 Webber and others 2007 Nash and others 2002 Ruitenberg and others 2005
Reference
Table 1. (continued)
6
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
No effect
No effect
Positive effect: >50% recovery of prelesion level in 5/6 rats v. no recovery in 8 controls No effect
Positive effect: 8 pellets taken v. 3; success rate: 22% v. 2%
OECs
OECs
NSCs
OECs
Lu and others 2004 Lu and others 2005
Reference
Results of tape removal test
Bone marrow No effect stroma cells Bone marrow No effect stroma cells
Cell type Liu and others 1999 Schwartz and others 2003
Reference
Positive effect: 13/15 rats ate >5 pellets Moreno-Flores OECs Latency of 24.7 seconds in cell group v. Neuhuber and v. 8/13 control rats ate >5 pellets and others 2006 43 seconds in vehicle rats others 2005 Xiao and others Webber and NSCs Latency of ~45 seconds in cell group Positive effect: ~8/50 pellets retrieved 2005 v. ~55 seconds in vehicle rats; only in OEC group v. ~2/50 in Schwann cell others 2007 significant at two time points; all other group (comparison of the two studies; behavioral test results were negative control group only in 2003 study with no recovery) Lu and others Bone marrow No effect (grooming task) Xiao and others 2012a stroma cells 2007
4/7 animals returned to normal reaching; no control Positive effect: success rate of 78% v. 38% in controls
Results of direct forepaw reaching test
OECs
Schwann cells
OECs
NSCs
OECs
OECs
Cell type
Bone marrow stroma cells Human olfactory neuroepithelial neurosphereforming cells Human olfactory neuroepithelial neurosphereforming cells
Fibroblasts
Fibroblasts
Cell type
8% usage of affected limb and 75% usage of affected + unaffected limb in cell group v. 10% in vehicle group 10% usage of affected limb and 80% usage of affected + unaffected limb in cell group v. 5% in vehicle group
50% use of affected limb in cell group + growth factor v. 10% in cell group No effect
Complete recovery
Results of vertical exploration test
Three tests: 1) direct forepaw reaching test for food, 2) tape removal test, and 3) vertical exploration in a cylinder. Studies are listed in chronological order. NSCs = neural stem cells; OECs = olfactory ensheathing cells.
Bretzner and others 2010 Rossi and others 2010
Ruitenberg and others 2005 Bretzner and others 2008 Yamamoto and others 2009
Keyvan-Fouladi and others 2005
Ogawa and others 2002 Keyvan-Fouladi and others 2003
Li and others 1997 Nash and others 2002
Reference
Table 2. Summary of Results of the Three Most Commonly Used Behavioral Tests.
7
Granger and others groups appeared to be nonrandomized, and observer blinding was not reported. Ogawa and others (2002) did not take account of the multiple comparisons between groups in the analysis of the data. Reanalysis of the data (using Kruskal-Wallis and post hoc tests) eliminated the reported treatment effect. Rats were also analyzed according to whether they ate more or less than five pellets, which was not a prespecified outcome in the methods. In a third study using neural progenitor cells (Webber and others 2007), a positive effect using the tape removal task was claimed. However, examination of the data showed that at only two time points (of seven) was there a significant difference between transplants and controls, neither of which was the final time point; furthermore, there was no difference in any of the three other behavioral analyses. The lesions in this study were not severe enough to produce lasting deficits, therefore rendering the detection of a treatment effect very difficult. Finally, in three other studies, NSCs were derived from human tissue and tested in marmosets in contusive and chronic lesions (Iwanami and others 2005; Kobayashi and others 2012; Yamane and others 2010), a model that is clearly clinically relevant. Although there is a claim that transplants were effective, these studies present several limitations because of the lack of reporting of blinding or allocation concealment, incorrect statistical analysis, and small animal numbers. The behavioral outcomes were specific for marmosets, and in that respect convincingly transferrable to humans, but were not used in studies in rats, rendering a direct comparison difficult. Studies Using MSCs. Four studies reported on bone marrow stromal cells, and none of them found a positive effect of the transplant. Three studies (Lu and others 2004, 2005, 2012a) were from the same laboratory. Lu and others (2012a) used one specific cell type bone marrow stromal cells (BMSCs) successively in a cervical lesion and thoracic transection and did not identify a beneficial effect of the cells in either experiment. It was also one of the two studies reporting allocation concealment. Finally, Neuhuber and others (2005) used human cells, and it is possible that the need for immunosuppression confounded the ability to detect beneficial effects. Studies Using Other Cell Types. Three other cell types (fibroblasts, olfactory neuroepithelial neurosphere-forming cells, and Schwann cells) were tested in six studies, and a positive effect was claimed for all of them. None of these specified whether the allocation of animals to the treatment group was concealed from researchers before the lesions. Two studies from the same laboratory (Liu and others 1999; Schwartz and others 2003), testing fibroblasts, used the vertical cylinder exploration test and
reported a positive effect after a moderately large lesion. Schwartz and others (2003) reported results in which all rats received fibroblasts ± BDNF, and there was no vehicle control group, rendering an assessment of the efficacy of the cells themselves difficult. However, rats receiving only fibroblasts recovered approximately 10% usage of the affected limb versus approximately 50% in rats receiving fibroblasts + BDNF, suggesting that the cell transplant in itself did not provide a large benefit. In the study by Liu and others (1999), recipients of the transplant regained complete use of the thoracic limb in the vertical cylinder exploration test, which was compared to lesion-only and vehicle groups. However, there was no blinding specified, and the number of animals was small (n = 6 per group), reducing the robustness of the conclusions. Two studies from the same laboratory (Xiao and others 2005, 2007), using olfactory neuroepithelial neurosphere-forming cells, performed the vertical cylinder exploration test after a moderately large lesion. In the 2005 study, although a statistical plan was presented in the methods, no statistics were reported in the results for the vertical exploration tests, and only a single time point (11 weeks) was presented. Transplanted animals recovered 8% usage of the affected limb, and it was only through an analysis of the combined use of affected plus unaffected forelimbs that an improvement was detected (75% usage of the limbs). In the 2007 study, transplanted animals only recovered 10 in the cell group. In the remaining 10 studies, animal numbers were 20 mV, and it is questionable whether the unit of measurement should have been “μV” as in other studies, and the MEP latencies were one sixth of the duration of those reported in all other studies; and 4) Wang and others (2011) reported MEP and SEP amplitudes only but not latencies. A large increase in the BBB score was observed in both studies (~7–8 points), but in one (Pan and others 2008), the score of the control lesioned animals was the highest (i.e., 3.8 points) of all the studies reviewed, again raising concerns regarding the completeness of the lesion. The study from Lu and others (2012b) did not suffer from these methodological concerns, except that allocation concealment was not specified, and showed that four of six transplanted animals recovered conduction across the lesion, which did not occur in any of the control lesioned animals. This was
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
9
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
Sprague-Dawley rats F Sprague-Dawley rats
Wistar Hannover rats
F Fischer rats
F Fischer rats
Chen and others 2011 Wang and others 2011
Ziegler and others 2011
Kang and others 2012
Lu and others 2012b
Human mesenchymal stem cells Rat and human NSCs
Olfactory bulb OECs
NSCs ± granulocyte colony-stimulating factor Bone marrow stroma cells NSCs and Schwann cells expressing NT-3 and TrkC
TS at T8–T9, leaving a 2-mm gap TS at T3, leaving a 2-mm gap
TS at T8–T9
Immediate
Immediate
BBB score
Kinematics, inclined grid climbing BBB score
BBB score
Immediate
Immediate
BBB score
BBB score
BBB score, inclined grid climbing BBB score
BBB score, inclined grid climbing BBB score
None
Behavioral tests
Immediate
Immediate
TS at T8–T9, leaving a 2-mm gap TS at T8, leaving a 2-mm gap TS at T10
Immediate
Immediate
TS at T9–T10, leaving a 2-mm gap TS at T8
Immediate
Immediate or at 7 days
Immediate
Time of transplant after lesion
TS at T8–T9
TS at T8–T9, leaving a 4- to 5-mm gap TS at T8–T9
Injury paradigm
Yes
Yes
NS
Electrical MEPs
SEPs (from L1 to sensory cortex) SCEPs
NS
Yes Electrical MEPs and SEPs (sciatic nerve to cortex)
NS
NS
NS
NS
NS
Yes
Electrical MEPs
Yes
NS
NS
NS
Yes
NS
NS
NS
Reported allocation concealment
Yes
Yes
Electrical MEPs and H-reflex Electrical MEPs and H-reflex Electrical MEPs and SEPs (sciatic nerve to cortex) Electrical MEPs and H-reflex Electrical MEPs
NS
Blinding
SCEPs
Electrophysiological tests
8 vehicle v. 3 groups of 8 with 3 cell concentrations 26 cells v. 6 lesion
15 cells v. 15 vehicle v. 10 lesion 4 groups of 10 with combinations of genetically modified cells v. 10 lesion v. 10 control 10 vehicle v. 11 cells v. cells + training v. 10 control
10 cells v. 10 lesion v. 10 drug v. 10 cells and drug
12 vehicle v. 9 drug v. 9 cells v. 9 cells + drug 6 groups of 10 (control, vehicle, co-graft, co-graft ± genetic modification) 10 cells v. 8 lesion
8 vehicle (acute and chronic) v. 8 cells (acute and chronic)
9 cells v. 2 vehicle v. 2 control
No. of animals used for functional analysis
BBB = Basso, Beattie, and Bresnahan; MEPs = motor evoked potentials; NS = not specified; NSCs = neural stem cells; OECs = olfactory ensheathing cells; SCEPs = spinal cord evoked potentials obtained by stimulation of the spinal cord below the lesion and recording potentials in the spinal cord above the lesion; SEPs = sensory evoked potentials; TS = transection. If not specified, the species of the origin of cells is rat.
F Sprague-Dawley rats Sprague-Dawley rats
Lopez-Vales and others 2007 Pan and others 2008
Olfactory bulb OECs v. FK506 NSCs and Schwann cells expressing NT-3 Olfactory bulb OECs
Olfactory bulb OECs
F Sprague-Dawley rats
F Sprague-Dawley rats F Sprague-Dawley rats
Schwann cells
Cell type
F Fischer rats
Species
Lopez-Vales and others 2006a Guo and others 2007
Pinzon and others 2001 Lopez-Vales and others 2006b
Reference
Table 3. Summary of Studies That Reported Electrophysiological Results after Spinal Cord Transection Listed in Chronological Order.
10
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
Behavioral test results
MEP amplitude results
21 ms in TP v. 0 ms in lesions
0/8 in lesions v. 6/10 in TP
NS
NS
2/9 in TP v. 0/2 in vehicles 0/15 in vehicles v. 4/6 in immediate TP v. 4/5 in chronic TP
No. of animals with MEP recovery
~6 mV in controls v. ~0.19 mV in TP + training v. ~0.05 mV in TP + no training v. 0 mV in vehicles None
4/6 in TP v. 0 in lesions
0/20 in vehicles v. 9/18 in TP + training v. 4/18 in TP + no training
~24 mV in lesions v. ~31 mV in drugs v. ~39 mV All animals had detectable MEPs in TP v. ~48 mV in TP + drug ~0.002 mV in vehicles v. ~0.25 mV in TP v. 0 NS ms in lesions 0.05 uV in lesions v. 0.4 uV in best combination All animals had detectable MEPs of cells
0.35 mV in TP v. 0 mV in lesions
None None 6 ms in controls v. ~16 ms in immediate TP 8–10 mV in controls v. ~0.7 mV in immediate v. ~17 ms in chronic TP v. 0 ms in vehicles TP v. ~0.3 mV in chronic TP v. 0 mV in vehicles 6 ms in controls v. ~16 ms in TP 28–32 mV in controls v. 0 mV in vehicles v. ~0.35 mV in TP 3.5 ms in controls v. 3.1 ms in 0.002 mV in controls v. 0.2 mV in combinatorial combinatorial cell treatment treatment
MEP latency results
BBB score: ~3.8 in lesions v. ~7.9 in drugs ~1.5 ms in lesions v. ~1.4 ms in drugs v. v. ~9.6 in TP v. ~11.7 in TP + drug ~1.4 ms in TP v. ~1.3 ms in TP + drug Bone marrow BBB score: 8 in lesions v. 8.5 in vehicles v. ~6 ms in vehicles v. ~5 ms in TP v. 0 ms stroma cells 8.9 in TP in lesions NS NSCs BBB score: 0.5 in lesions to 7.6 in best combination of cells; inclined grid climbing: no difference OECs Improved stepping abilities but identical ~8 ms in controls v. ~18 ms in TP + inclined grid climbing abilities in all groups training v. ~25 ms in TP + no training v. 0 ms in vehicles NSCs BBB score: ~1 in lesions v. ~7 in TP None
NSCs
Schwann cells None OECs BBB score: 2 in vehicles v. 4.2 in immediate TP v. 3.7 in chronic TP; inclined grid climbing: no difference OECs BBB score: 2 in vehicles v. 4.1 in TP v. 5.1 in TP + drug Schwann cells BBB score: 0.5 in lesions v. 10.7 in combinatorial cell treatment; inclined grid climbing: recovery OECs BBB score: ~1 in lesions v. 2.5 in TP
Cell type
BBB = Basso, Beattie, and Bresnahan; MEP = motor evoked potential; NS = not specified; NSCs = neural stem cells; OECs = olfactory ensheathing cells; TP = transplant group.
Lu and others 2012b
Ziegler and others 2011
Wang and others 2011
Chen and others 2011
Lopez-Vales and others 2007 Pan and others 2008
Lopez-Vales and others 2006a Guo and others 2007
Pinzon and others 2001 Lopez-Vales and others 2006b
Reference
Table 4. Summary of MEP Results and Behavioral Score Results Listed in Chronological Order.
11
Granger and others further supported by the fact that the recovered potentials were abolished after retransection. However, these evoked potentials were recorded between the C7 and T6 spinal cord segments (after a T3 transection) and therefore did not assess connections with the brain. The BBB score increased by 6 points in transplanted animals. Guo and others (2007) reported latencies in control animals that were half those of control rats from other OEC studies. An improvement of the MEP and SEP amplitudes and latencies was found, although the statistical analysis employed multiple testing with t-tests, which does not make an allowance for repeated testing. The BBB score (10.2 points) was the largest of all the studies included here. Neither blinding of data collection nor allocation concealment was reported. Neural stem cells (from rodents) were combined with Schwann cells expressing NT-3 in the study by Guo and others (2007), with Schwann cells expressing NT-3 and TrkC in the study by Wang and others (2011), and with human NSCs from spinal cord fetuses by Lu and others (2012b), rendering an analysis of the true effect of this class of cells on their own difficult to determine. Studies on MSCs. Two of two studies found a positive effect of MSCs. Kang and others (2012) used sensory evoked potentials (stimulation of the spinal cord and recording over the cortex), but they were termed MEPs. The amplitudes were reported but not latencies, although the methods specified that both amplitudes and latencies were collected. All animals (including controls) had detectable SEPs, leading to questions regarding the completeness of the transections. The increase in the BBB score in transplanted animals was minor, from 2 to 4.5 to 6 points. Chen and others (2011) claimed a positive effect based on a one-millisecond reduction of MEP latencies between transplanted and vehicle animals, and there was also an increase in MEP amplitude. However, these findings did not correlate with changes in the BBB scores, and it was not stated how many rats actually recovered MEPs. Studies on Schwann Cells. Only one study assessed Schwann cells alone (Pinzon and others 2001) and found a positive effect, but neither blinded data collection nor allocation concealment was reported. There were only two control lesioned animals (with no electrophysiological recovery) versus two of nine transplanted rats that recovered evoked potentials, and no statistics were presented. The measured potentials were from the spinal cord, and although recovery was observed, the methods did not directly assess connection with the brain. Animal Number, Blinding, and Allocation Concealment. The mean number of animals per treatment group was 11 ± 5
in thoracic studies. Eight of 11 studies used >10 animals in at least the transplant group. However, some studies had a very imbalanced number of controls. For example, Lu and others (2012b) compared 26 transplanted animals to 6 controls. Pinzon and others (2001) compared nine transplanted animals to two transected controls. Eight of 11 studies were blinded, and none reported allocation concealment.
Summary of Effects of Cell Transplants in Thoracolumbar Lesions Overall, OECs appear to have a possible effect on electrophysiological recovery and a modest effect on the BBB score (i.e., ~3–4 points) because the four examined studies converged on similar findings, although three were from the same laboratory. There were many methodological concerns in studies testing NSCs, rendering an interpretation of their effect difficult notably because of the preservation of MEPs in lesioned control animals (thereby questioning the completeness of the lesion) and the discrepancy between studies with regard to MEP latencies in control animals, which ranged from approximately 1.5 milliseconds to approximately 8 milliseconds. Efficacy of MSCs was difficult to assess because there were missing data in the study by Kang and others (2012) and a change in the BBB score of
Cell Therapy for Spinal Cord Injuries: What Is Really Going on? Nicolas Granger, Robin J.M. Franklin and Nick D. Jeffery Neuroscientist published online 10 January 2014 DOI: 10.1177/1073858413514635 The online version of this article can be found at: http://nro.sagepub.com/content/early/2014/01/09/1073858413514635
Published by: http://www.sagepublications.com
Additional services and information for The Neuroscientist can be found at: Email Alerts: http://nro.sagepub.com/cgi/alerts Subscriptions: http://nro.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav
>> OnlineFirst Version of Record - Jan 10, 2014 What is This?
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
514635
research-article2014
NROXXX10.1177/1073858413514635The NeuroscientistGranger and others
Article
Cell Therapy for Spinal Cord Injuries: What Is Really Going on?
The Neuroscientist 1–16 © The Author(s) 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/1073858413514635 nro.sagepub.com
Nicolas Granger1, Robin J.M. Franklin2, and Nick D. Jeffery3
Abstract During the last two decades, many experiments have examined the ability of cell transplants to ameliorate the loss of function after spinal cord injuries, with the hope of developing interventions to benefit patients. Although many reports suggest positive effects, there is growing concern over the quality of the available preclinical data. It is therefore important to ask whether this worldwide investigative process is close to defining a cell transplant protocol that could be translated into human patients with a realistic chance of success. This review systematically examines the strength of the preclinical evidence and outlines mechanisms by which transplanted cells may mediate their effects in spinal cord injuries. First, we examined changes in voluntary movements in the forelimb associated with cell transplants after partial cervical lesions. Second, we examined the efficacy of transplanted cells to restore electrophysiological conduction across a complete thoracic lesion. We postulated that cell therapies found to be successful in both models could reasonably have potential to treat human patients. We conclude that although there are data to support a beneficial effect of cell transplantation, most reports provide only weak evidence because of deficits in experimental design. The mechanisms by which transplanted cells mediate their functional effects remain unclear. Keywords transplantation, animal model, functional, forelimb reaching, transection The seminal nerve transplantation studies by Aguayo and others in the 1980s (Aguayo and others 1981; David and Aguayo 1981; Richardson and others 1984; Zhou and others 1986) spurred a new wave of interest in the effects of cell transplantation in the CNS. In part, this was motivated by scientific curiosity regarding the effects of cells on the host, and vice versa, and the precise mechanisms underlying those effects. However, another clear rationale is the hope that transplanted cells may be able to repair injuries to the nervous system and, in particular, provide benefits to patients who have suffered spinal cord injuries (SCIs). Despite three decades of laboratory studies on cell transplantation and although cell transplantation for human SCIs has now commenced in many countries, including the USA (e.g., see http://www.miamiproject.miami.edu/page.aspx?pid=339), it currently remains unclear whether any type of cell transplant can be conclusively considered likely to be of benefit in clinical SCIs. Here, we critically review the literature on cell transplantation in the spinal cord of laboratory animals with the aim of defining the strength of evidence to support translation into human patients. There is a plethora of published material describing the effects of many different cell transplants in various types of SCIs, meaning that a review can become unwieldy in the absence of clear parameters defining its
scope. Therefore, here, we restrict our review to an examination of the published evidence regarding efficacy in restoring “function” after SCIs to focus on the likelihood of translation of functional benefits to human recipients. For that reason, we considered first that a translational candidate cell type should be associated with improvement in voluntary movement. This is, of course, difficult to ascertain definitely in nonverbal species. However, we considered that improvement in animal activities such as grooming or use of the limbs to perform actions in accordance with an animal’s “wants” rather than simply reflecting changes in reflex activity (e.g., such as pelvic limb stepping) would provide a reasonable substitute. Second, we considered that it would be necessary for the same cell type to be effective in a complete lesion. This is because 1
School of Veterinary Sciences, University of Bristol, Bristol, UK Wellcome Trust-Medical Research Council Cambridge Stem Cell Institute and Department of Veterinary Medicine, University of Cambridge, Cambridge, UK 3 Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Iowa State University, Ames, IA, USA 2
Corresponding Author: Nick D. Jeffery, Department of Veterinary Clinical Sciences, College of Veterinary Medicine, Iowa State University, 1600 South 16th Street, Ames, IA 50011, USA. Email: [email protected]
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
2
The Neuroscientist XX(X)
Figure 1. The review of the literature was divided into two parts to assess the strength of the available data on the possible benefits of cell transplantation after spinal cord injuries. (A) First, we reviewed studies which examined the recovery of goaldirected forelimb movement after partial cervical lesions. (B) Second, we reviewed whether there was evidence that cell types applied in cervical lesions were also effective in restoring electrophysiological conduction in a spinal cord transection model.
patients who are most likely to be candidates for cell therapy trials will be those with very severe lesions, as they have the least function to lose and the ability to detect beneficial effects is greater (Fawcett and others 2007). Recovery through plasticity is of course beneficial but cannot be expected to be efficacious alone in anatomically complete lesions. In an effort to address these two aspects, this systematic review is divided into two parts (Fig. 1): first, we examined publications showing evidence of improvement in goaldirected function after cells were transplanted into cervical SCIs. Cervical injuries were necessary because of the need to examine “voluntary” function. Second, limiting the review to cell types discussed in the first part, we examined experiments in which the effects of such cells have also been studied using electrophysiological methods after
transplantation in complete transection lesions in the thoracolumbar spinal cord. We focused the analysis on electrophysiological examination to allow unequivocal differentiation between regeneration and the recovery of reflex actions, which would not be possible through behavioral observations alone. This part of the review therefore pertains only to thoracolumbar lesions because complete transection of the cervical cord is incompatible with reasonable animal survival and experimental ethics. We also review the quality of the evidence provided by these studies with reference to the guidelines provided for clinical trials, the CONSORT (Consolidated Standards of Reporting Trials) statement regarding clinical trials (http:// www.consort-statement.org/consort-statement/), and those proposed by Dobkin (2010) to report clinical data and by Landis and others (2012) to report experimental results.
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
3
Granger and others
Analyzing the Efficacy of Cell Therapy Transplants in SCI Functional Outcome after Cell Transplantation in the Cervical Spinal Cord Most (n = 22) of the studies examining volitional functional outcomes after cell transplant in incomplete cervical spinal cord lesions (of a total of 25 that we identified) were carried out in rodents, and an additional three were performed in marmosets (Table 1). The reported outcomes (Table 2) focused on forepaw tasks: to reach food (n = 10 studies), remove tape from the paw (n = 4), or groom (n = 1). In 10 other studies, the analysis concentrated on tasks in which the animal had to extend the forelimb to explore the environment (vertical exploration: n = 5; climbing: n = 5). Cell Transplants Associated with Beneficial Effects. In 17 of the 25 studies, a positive effect of the transplant on the recovery of function was claimed. Five of nine studies on olfactory ensheathing cells (OECs) reported a positive effect; of these, the cells were derived from the olfactory mucosa in one study and from the olfactory bulb in four. Neural stem/ progenitor cells were associated with a positive effect in six of six reports. Mesenchymal stem cells (MSCs) were tested in four studies, of which none found a functional benefit. In the remaining six studies claiming a positive effect, the cells tested were olfactory neuroepithelial neurosphere-forming cells (n = 2 from one laboratory), BDNFexpressing fibroblasts (n = 2 from one laboratory), and Schwann cells (n = 2 from two laboratories). Functional Effect of OECs Studies in which OECs were associated with benefits. Four studies used a directed forepaw reaching task (KeyvanFouladi and others 2003; Li and others 1997; Nash and others 2002; Yamamoto and others 2009). In three studies, the dorsal corticospinal tract was unilaterally (KeyvanFouladi and others 2003; Li and others 1997; Yamamoto and others 2009) or bilaterally (Nash and others 2002) destroyed. The OEC transplants were considered to have a large beneficial effect because some transplanted animals regained near normal function (Table 2). However, the results have limitations because blinding and allocation concealment were not reported and statistical analysis was lacking, and one study (Li and others 1997) did not include a control group. Importantly, in one report (Yamamoto and others 2009), the transplants contained only 5% OECs and 95% fibroblasts, and although the recovery of forepaw reaching was observed, no axonal growth across the lesion was detected. In the fourth study using OECs and the forepaw reaching test (Nash and others 2002), a large improvement in the forepaw reaching task was
detected (success rate at reaching food pellets at six weeks after injury was 71%–78% in the cell transplant group [n = 19] v. 38%–42% in vehicles and control lesion rats). The study was blinded, but allocation concealment was not reported; moreover, 15 animals were excluded because they died within 14 days of surgery, and it is not possible to know to which group they belonged. One study (Moreno-Flores and others 2006) used OECs and found a beneficial effect during a tape removal task; this assessment was conducted blinded, but allocation concealment was not reported. Studies on OEC transplantation in which no effect was reported. In four studies, OECs had no reported effect on thoracic limb recovery (Bretzner and others 2008; Bretzner and others 2010; Collazos-Castro and others 2005; Ruitenberg and others 2005). Collazos-Castro and others (2005) reported results in the treatment of relatively mild lesions from which control animals were able to recover substantial function in the absence of any intervention, which might account for the difficulty in detecting an effect of transplanted OECs. Bretzner and others (2008) found no motor improvement during forepaw reaching or vertical exploration tests. The lesions (unilateral dorsolateral funiculus section) were much larger than those in studies in which a positive effect of OECs was found. Further, both studies used immunosuppression because the authors transplanted mice OECs into rats, which could be a confounding factor. A reduced sensory threshold in animals receiving OECs (thermal hypersensitivity) was also detected in these studies and constitutes a potential detrimental effect of OEC transplantation. The control rats used by Ruitenberg and others (2005) recovered function to preoperative levels, rendering the detection of an intervention effect difficult. It was also one of the two studies (Lu and others 2012b; Ruitenberg and others 2005) in which treatment allocation was concealed, which is known to reduce the magnitude of the treatment effect through a reduction in bias. Studies in Which a Positive Effect of Neural Stem/Progenitor Cells Was Reported. In two studies using a forepaw reaching task, one using motor neuron progenitors (Rossi and others 2010) found that transplanted rats could retrieve a mean of approximately 8 pellets versus approximately 3 in the vehicle group, and the second, using neural stem cells (NSCs) (Ogawa and others 2002), found that 13 of 15 rats ate >5 pellets versus 8 of 13 control rats who ate >5 pellets. Although both are interpreted as positive results, it is difficult to know how these two different end points can be compared. Both these two reports involve a contusive injury and cells transplanted at a chronic stage. Rossi and others (2010) did not report the results of the nontransplanted animals, allocation to experimental
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
4
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
Cell type
F Fischer 344 rats BDNF-expressing bone marrow stroma cells
Lu and others 2012a
Mice olfactory mucosa OECs
M SpragueDawley rats
F Sprague-Dawley Human motor neuron rats progenitors F Swiss albino rats Olfactory bulb OECs and ONFs F Swiss albino rats Olfactory bulb OECs and ONFs F Swiss albino rats Schwann cells and fibroblasts F Swiss albino rats Olfactory mucosa OECs F Sprague-Dawley BDNF-expressing rats fibroblasts F Sprague-Dawley BDNF-expressing rats fibroblasts F Sprague-Dawley Human bone marrow rats stroma cells F Sprague-Dawley Human olfactory rats neuroepithelial neurosphere-forming cells F Sprague-Dawley Human olfactory rats neuroepithelial neurosphere-forming cells M SpragueMice olfactory mucosa Dawley rats OECs
F Sprague-Dawley NSCs as neurospheres rats M Wistar rats Olfactory bulb OECs and ONFs F Fischer rats Schwann cells
Species
Bretzner and others 2010
Bretzner and others 2008
Xiao and others 2007
Ogawa and others 2002 Collazos-Castro and others 2005 Schaal and others 2007 Rossi and others 2010 Li and others 1997 Keyvan-Fouladi and others 2003 Keyvan-Fouladi and others 2005 Yamamoto and others 2009 Liu and others 1999 Schwartz and others 2003 Neuhuber and others 2005 Xiao and others 2005
Reference
Vertical exploration
Immediate Immediate Immediate 7 days
Immediate
Immediate
C3/C4 hemisection C3/C4 hemisection C3/C4 hemisection
C3/C4 hemisection
C4/C5 hemisection
C5 hemisection
Immediate
Immediate
Direct forepaw reaching
8 weeks
C4/C5 hemisection
Direct forepaw reaching
Immediate
Direct forepaw reaching, vertical exploration, thermal and mechanical threshold Direct forepaw reaching, vertical exploration, thermal and mechanical threshold Forelimb grooming task, grid walk, platform locomotion
Vertical exploration
Vertical exploration, BBB score, thermal threshold Vertical exploration, horizontal rope walking
Vertical exploration, BBB score
Direct forepaw reaching
8 weeks
Forearm grip, inclined plane test, BBB score, forelimb hanging Direct forepaw reaching, beam walking, forearm grip Direct forepaw reaching
Forelimb kinematics during walking
Direct forepaw reaching
Behavioral tests
C1/C2 electrolytic CST lesion C1/C2 electrolytic CST lesion C1/C2 electrolytic CST lesion C1/C2 electrolytic CST lesion C3/C4 hemisection
Immediate
Immediate
C5 contusion 7 days
Immediate
C6/C7 contusion
C5/C6 contusion
9 days
Time of transplant after lesion
C4/C5 contusion
Injury paradigm
NS
Yesa
Yes
NS
Yesa
Yes
NS
NS
Yes
Yes
NS
NS
Yesa Yes
NS
NS
NS
NS
NS
NS
NS
NS
NS
Reported allocation concealment
NS
NS
NS
NS
NS
NS
Yes
NS
NS
Blinding
Table 1. Summary of Studies that Reported Analysis of Movement after Cell Transplantation in Incomplete Cervical Lesions.
(continued)
24 cell group (± growth factor) v. 21 vehicle (of various forms ± growth factor) 48 cell group ± growth factor (4 groups of 12) v. 12 no cells but growth factor
7 cell group v. 17 vehicle or growth factor
30 cell group v. 18 vehicle v. 4 control
12 cell group v. 12 vehicle
6 cell group v. 6 cell group + growth factor v. 5 vehicle 6 cell group + growth factor v. 8 cell group v. 4 normal 32 cell group v. 8 vehicle v. 10 control
6 cell group v. 6 lesion v. 2 vehicle
18 cell group v. 12 historical cell group
12 cell group v. 2 vehicle v. 6 lesion
7 cell group v. 28 lesion
15 cell group v. 15 vehicle
14 cell group v. 13 lesion v. 16 sham
15 cell group v. 13 lesion v. 17 vehicle v. 10 control 14 cell group v. 10 vehicle
No. of animals used for functional analysis
5
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
Cell type
F common marmosets
Human induced pluripotent stem cells and NSCs
Sprague-Dawley Olfactory bulb OECs rats and ONFs F Fischer 344 rats NT-3–expressing olfactory bulb OECs and ONFs F common Human NSCs as marmosets neurospheres F common Human NSCs as marmosets neurospheres
F Fischer 344 rats Bone marrow stroma cells F Fischer 344 rats BDNF-expressing bone marrow stroma cells M Wistar rats Immortalized olfactory bulb OECs M Fischer 344 rats Neural progenitor cells
Species
Direct forepaw reaching, BBB score
Immediate
Immediate
C4 DC and CST transection C3 DC and CST transection C4 DC and CST hemisection 9 days 9 days
9 days
C5 contusion C5 contusion
C5 contusion
Immediate
Tape removal, beam walking, grid walking, wire ladder walking Direct forepaw reaching
Immediate
C3 DC crush
Forearm grip, treadmill and spontaneous locomotion (custom-designed score) Forearm grip, treadmill and spontaneous locomotion (custom-designed score)
Forelimb kinematics, forearm grip
Tape removal, beam walking
Immediate
C3 DC transection
Tape removal, horizontal ladder locomotion, rope walking Tape removal, rope walking
Behavioral tests
Immediate
Time of transplant after lesion
C4 DC transection
Injury paradigm
NS NS
NS
NS
NS
Yes
NS
NS
Yes: BBB score; no: direct forepaw reaching NS
Yes
NS
NS
NS
Yesa Yes
NS
Reported allocation concealment
Yes
Blinding
5 cell group v. 4 vehicle
7 cell group v. 7 vehicle v. 2 sham
5 cell group v. 5 vehicle
19 cell group (± drug) v. 5 drug group v. 5 sham v. 5 vehicle v. 5 lesion 21 cell group (± growth factor) v. 6 vehicle
5 groups of 12 receiving cells ± growth factor 6 cell group v. 6 cell group + growth factor 20 cell group v. 12 vehicle v. 6 sham v. 6 control 16 cell group v. 16 vehicle v. 10 sham
No. of animals used for functional analysis
The studies are grouped according to the method of injury and listed in chronological order. BBB = Basso, Beattie, and Bresnahan; CST = corticospinal tract; DC = dorsal column; DLF = dorsolateral funiculus; NS = not specified; NSCs = neural stem cells; OECs = olfactory ensheathing cells; ONFs = olfactory nerve fibroblasts. If not specified, the species of the origin of cells is rat. a. Blinded scoring from videos without mention of videos being obtained blinded.
Kobayashi and others 2012
Iwanami and others 2005 Yamane and others 2010
Lu and others 2004 Lu and others 2005 Moreno-Flores and others 2006 Webber and others 2007 Nash and others 2002 Ruitenberg and others 2005
Reference
Table 1. (continued)
6
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
No effect
No effect
Positive effect: >50% recovery of prelesion level in 5/6 rats v. no recovery in 8 controls No effect
Positive effect: 8 pellets taken v. 3; success rate: 22% v. 2%
OECs
OECs
NSCs
OECs
Lu and others 2004 Lu and others 2005
Reference
Results of tape removal test
Bone marrow No effect stroma cells Bone marrow No effect stroma cells
Cell type Liu and others 1999 Schwartz and others 2003
Reference
Positive effect: 13/15 rats ate >5 pellets Moreno-Flores OECs Latency of 24.7 seconds in cell group v. Neuhuber and v. 8/13 control rats ate >5 pellets and others 2006 43 seconds in vehicle rats others 2005 Xiao and others Webber and NSCs Latency of ~45 seconds in cell group Positive effect: ~8/50 pellets retrieved 2005 v. ~55 seconds in vehicle rats; only in OEC group v. ~2/50 in Schwann cell others 2007 significant at two time points; all other group (comparison of the two studies; behavioral test results were negative control group only in 2003 study with no recovery) Lu and others Bone marrow No effect (grooming task) Xiao and others 2012a stroma cells 2007
4/7 animals returned to normal reaching; no control Positive effect: success rate of 78% v. 38% in controls
Results of direct forepaw reaching test
OECs
Schwann cells
OECs
NSCs
OECs
OECs
Cell type
Bone marrow stroma cells Human olfactory neuroepithelial neurosphereforming cells Human olfactory neuroepithelial neurosphereforming cells
Fibroblasts
Fibroblasts
Cell type
8% usage of affected limb and 75% usage of affected + unaffected limb in cell group v. 10% in vehicle group 10% usage of affected limb and 80% usage of affected + unaffected limb in cell group v. 5% in vehicle group
50% use of affected limb in cell group + growth factor v. 10% in cell group No effect
Complete recovery
Results of vertical exploration test
Three tests: 1) direct forepaw reaching test for food, 2) tape removal test, and 3) vertical exploration in a cylinder. Studies are listed in chronological order. NSCs = neural stem cells; OECs = olfactory ensheathing cells.
Bretzner and others 2010 Rossi and others 2010
Ruitenberg and others 2005 Bretzner and others 2008 Yamamoto and others 2009
Keyvan-Fouladi and others 2005
Ogawa and others 2002 Keyvan-Fouladi and others 2003
Li and others 1997 Nash and others 2002
Reference
Table 2. Summary of Results of the Three Most Commonly Used Behavioral Tests.
7
Granger and others groups appeared to be nonrandomized, and observer blinding was not reported. Ogawa and others (2002) did not take account of the multiple comparisons between groups in the analysis of the data. Reanalysis of the data (using Kruskal-Wallis and post hoc tests) eliminated the reported treatment effect. Rats were also analyzed according to whether they ate more or less than five pellets, which was not a prespecified outcome in the methods. In a third study using neural progenitor cells (Webber and others 2007), a positive effect using the tape removal task was claimed. However, examination of the data showed that at only two time points (of seven) was there a significant difference between transplants and controls, neither of which was the final time point; furthermore, there was no difference in any of the three other behavioral analyses. The lesions in this study were not severe enough to produce lasting deficits, therefore rendering the detection of a treatment effect very difficult. Finally, in three other studies, NSCs were derived from human tissue and tested in marmosets in contusive and chronic lesions (Iwanami and others 2005; Kobayashi and others 2012; Yamane and others 2010), a model that is clearly clinically relevant. Although there is a claim that transplants were effective, these studies present several limitations because of the lack of reporting of blinding or allocation concealment, incorrect statistical analysis, and small animal numbers. The behavioral outcomes were specific for marmosets, and in that respect convincingly transferrable to humans, but were not used in studies in rats, rendering a direct comparison difficult. Studies Using MSCs. Four studies reported on bone marrow stromal cells, and none of them found a positive effect of the transplant. Three studies (Lu and others 2004, 2005, 2012a) were from the same laboratory. Lu and others (2012a) used one specific cell type bone marrow stromal cells (BMSCs) successively in a cervical lesion and thoracic transection and did not identify a beneficial effect of the cells in either experiment. It was also one of the two studies reporting allocation concealment. Finally, Neuhuber and others (2005) used human cells, and it is possible that the need for immunosuppression confounded the ability to detect beneficial effects. Studies Using Other Cell Types. Three other cell types (fibroblasts, olfactory neuroepithelial neurosphere-forming cells, and Schwann cells) were tested in six studies, and a positive effect was claimed for all of them. None of these specified whether the allocation of animals to the treatment group was concealed from researchers before the lesions. Two studies from the same laboratory (Liu and others 1999; Schwartz and others 2003), testing fibroblasts, used the vertical cylinder exploration test and
reported a positive effect after a moderately large lesion. Schwartz and others (2003) reported results in which all rats received fibroblasts ± BDNF, and there was no vehicle control group, rendering an assessment of the efficacy of the cells themselves difficult. However, rats receiving only fibroblasts recovered approximately 10% usage of the affected limb versus approximately 50% in rats receiving fibroblasts + BDNF, suggesting that the cell transplant in itself did not provide a large benefit. In the study by Liu and others (1999), recipients of the transplant regained complete use of the thoracic limb in the vertical cylinder exploration test, which was compared to lesion-only and vehicle groups. However, there was no blinding specified, and the number of animals was small (n = 6 per group), reducing the robustness of the conclusions. Two studies from the same laboratory (Xiao and others 2005, 2007), using olfactory neuroepithelial neurosphere-forming cells, performed the vertical cylinder exploration test after a moderately large lesion. In the 2005 study, although a statistical plan was presented in the methods, no statistics were reported in the results for the vertical exploration tests, and only a single time point (11 weeks) was presented. Transplanted animals recovered 8% usage of the affected limb, and it was only through an analysis of the combined use of affected plus unaffected forelimbs that an improvement was detected (75% usage of the limbs). In the 2007 study, transplanted animals only recovered 10 in the cell group. In the remaining 10 studies, animal numbers were 20 mV, and it is questionable whether the unit of measurement should have been “μV” as in other studies, and the MEP latencies were one sixth of the duration of those reported in all other studies; and 4) Wang and others (2011) reported MEP and SEP amplitudes only but not latencies. A large increase in the BBB score was observed in both studies (~7–8 points), but in one (Pan and others 2008), the score of the control lesioned animals was the highest (i.e., 3.8 points) of all the studies reviewed, again raising concerns regarding the completeness of the lesion. The study from Lu and others (2012b) did not suffer from these methodological concerns, except that allocation concealment was not specified, and showed that four of six transplanted animals recovered conduction across the lesion, which did not occur in any of the control lesioned animals. This was
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
9
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
Sprague-Dawley rats F Sprague-Dawley rats
Wistar Hannover rats
F Fischer rats
F Fischer rats
Chen and others 2011 Wang and others 2011
Ziegler and others 2011
Kang and others 2012
Lu and others 2012b
Human mesenchymal stem cells Rat and human NSCs
Olfactory bulb OECs
NSCs ± granulocyte colony-stimulating factor Bone marrow stroma cells NSCs and Schwann cells expressing NT-3 and TrkC
TS at T8–T9, leaving a 2-mm gap TS at T3, leaving a 2-mm gap
TS at T8–T9
Immediate
Immediate
BBB score
Kinematics, inclined grid climbing BBB score
BBB score
Immediate
Immediate
BBB score
BBB score
BBB score, inclined grid climbing BBB score
BBB score, inclined grid climbing BBB score
None
Behavioral tests
Immediate
Immediate
TS at T8–T9, leaving a 2-mm gap TS at T8, leaving a 2-mm gap TS at T10
Immediate
Immediate
TS at T9–T10, leaving a 2-mm gap TS at T8
Immediate
Immediate or at 7 days
Immediate
Time of transplant after lesion
TS at T8–T9
TS at T8–T9, leaving a 4- to 5-mm gap TS at T8–T9
Injury paradigm
Yes
Yes
NS
Electrical MEPs
SEPs (from L1 to sensory cortex) SCEPs
NS
Yes Electrical MEPs and SEPs (sciatic nerve to cortex)
NS
NS
NS
NS
NS
Yes
Electrical MEPs
Yes
NS
NS
NS
Yes
NS
NS
NS
Reported allocation concealment
Yes
Yes
Electrical MEPs and H-reflex Electrical MEPs and H-reflex Electrical MEPs and SEPs (sciatic nerve to cortex) Electrical MEPs and H-reflex Electrical MEPs
NS
Blinding
SCEPs
Electrophysiological tests
8 vehicle v. 3 groups of 8 with 3 cell concentrations 26 cells v. 6 lesion
15 cells v. 15 vehicle v. 10 lesion 4 groups of 10 with combinations of genetically modified cells v. 10 lesion v. 10 control 10 vehicle v. 11 cells v. cells + training v. 10 control
10 cells v. 10 lesion v. 10 drug v. 10 cells and drug
12 vehicle v. 9 drug v. 9 cells v. 9 cells + drug 6 groups of 10 (control, vehicle, co-graft, co-graft ± genetic modification) 10 cells v. 8 lesion
8 vehicle (acute and chronic) v. 8 cells (acute and chronic)
9 cells v. 2 vehicle v. 2 control
No. of animals used for functional analysis
BBB = Basso, Beattie, and Bresnahan; MEPs = motor evoked potentials; NS = not specified; NSCs = neural stem cells; OECs = olfactory ensheathing cells; SCEPs = spinal cord evoked potentials obtained by stimulation of the spinal cord below the lesion and recording potentials in the spinal cord above the lesion; SEPs = sensory evoked potentials; TS = transection. If not specified, the species of the origin of cells is rat.
F Sprague-Dawley rats Sprague-Dawley rats
Lopez-Vales and others 2007 Pan and others 2008
Olfactory bulb OECs v. FK506 NSCs and Schwann cells expressing NT-3 Olfactory bulb OECs
Olfactory bulb OECs
F Sprague-Dawley rats
F Sprague-Dawley rats F Sprague-Dawley rats
Schwann cells
Cell type
F Fischer rats
Species
Lopez-Vales and others 2006a Guo and others 2007
Pinzon and others 2001 Lopez-Vales and others 2006b
Reference
Table 3. Summary of Studies That Reported Electrophysiological Results after Spinal Cord Transection Listed in Chronological Order.
10
Downloaded from nro.sagepub.com at UNIV OF VIRGINIA on June 4, 2014
Behavioral test results
MEP amplitude results
21 ms in TP v. 0 ms in lesions
0/8 in lesions v. 6/10 in TP
NS
NS
2/9 in TP v. 0/2 in vehicles 0/15 in vehicles v. 4/6 in immediate TP v. 4/5 in chronic TP
No. of animals with MEP recovery
~6 mV in controls v. ~0.19 mV in TP + training v. ~0.05 mV in TP + no training v. 0 mV in vehicles None
4/6 in TP v. 0 in lesions
0/20 in vehicles v. 9/18 in TP + training v. 4/18 in TP + no training
~24 mV in lesions v. ~31 mV in drugs v. ~39 mV All animals had detectable MEPs in TP v. ~48 mV in TP + drug ~0.002 mV in vehicles v. ~0.25 mV in TP v. 0 NS ms in lesions 0.05 uV in lesions v. 0.4 uV in best combination All animals had detectable MEPs of cells
0.35 mV in TP v. 0 mV in lesions
None None 6 ms in controls v. ~16 ms in immediate TP 8–10 mV in controls v. ~0.7 mV in immediate v. ~17 ms in chronic TP v. 0 ms in vehicles TP v. ~0.3 mV in chronic TP v. 0 mV in vehicles 6 ms in controls v. ~16 ms in TP 28–32 mV in controls v. 0 mV in vehicles v. ~0.35 mV in TP 3.5 ms in controls v. 3.1 ms in 0.002 mV in controls v. 0.2 mV in combinatorial combinatorial cell treatment treatment
MEP latency results
BBB score: ~3.8 in lesions v. ~7.9 in drugs ~1.5 ms in lesions v. ~1.4 ms in drugs v. v. ~9.6 in TP v. ~11.7 in TP + drug ~1.4 ms in TP v. ~1.3 ms in TP + drug Bone marrow BBB score: 8 in lesions v. 8.5 in vehicles v. ~6 ms in vehicles v. ~5 ms in TP v. 0 ms stroma cells 8.9 in TP in lesions NS NSCs BBB score: 0.5 in lesions to 7.6 in best combination of cells; inclined grid climbing: no difference OECs Improved stepping abilities but identical ~8 ms in controls v. ~18 ms in TP + inclined grid climbing abilities in all groups training v. ~25 ms in TP + no training v. 0 ms in vehicles NSCs BBB score: ~1 in lesions v. ~7 in TP None
NSCs
Schwann cells None OECs BBB score: 2 in vehicles v. 4.2 in immediate TP v. 3.7 in chronic TP; inclined grid climbing: no difference OECs BBB score: 2 in vehicles v. 4.1 in TP v. 5.1 in TP + drug Schwann cells BBB score: 0.5 in lesions v. 10.7 in combinatorial cell treatment; inclined grid climbing: recovery OECs BBB score: ~1 in lesions v. 2.5 in TP
Cell type
BBB = Basso, Beattie, and Bresnahan; MEP = motor evoked potential; NS = not specified; NSCs = neural stem cells; OECs = olfactory ensheathing cells; TP = transplant group.
Lu and others 2012b
Ziegler and others 2011
Wang and others 2011
Chen and others 2011
Lopez-Vales and others 2007 Pan and others 2008
Lopez-Vales and others 2006a Guo and others 2007
Pinzon and others 2001 Lopez-Vales and others 2006b
Reference
Table 4. Summary of MEP Results and Behavioral Score Results Listed in Chronological Order.
11
Granger and others further supported by the fact that the recovered potentials were abolished after retransection. However, these evoked potentials were recorded between the C7 and T6 spinal cord segments (after a T3 transection) and therefore did not assess connections with the brain. The BBB score increased by 6 points in transplanted animals. Guo and others (2007) reported latencies in control animals that were half those of control rats from other OEC studies. An improvement of the MEP and SEP amplitudes and latencies was found, although the statistical analysis employed multiple testing with t-tests, which does not make an allowance for repeated testing. The BBB score (10.2 points) was the largest of all the studies included here. Neither blinding of data collection nor allocation concealment was reported. Neural stem cells (from rodents) were combined with Schwann cells expressing NT-3 in the study by Guo and others (2007), with Schwann cells expressing NT-3 and TrkC in the study by Wang and others (2011), and with human NSCs from spinal cord fetuses by Lu and others (2012b), rendering an analysis of the true effect of this class of cells on their own difficult to determine. Studies on MSCs. Two of two studies found a positive effect of MSCs. Kang and others (2012) used sensory evoked potentials (stimulation of the spinal cord and recording over the cortex), but they were termed MEPs. The amplitudes were reported but not latencies, although the methods specified that both amplitudes and latencies were collected. All animals (including controls) had detectable SEPs, leading to questions regarding the completeness of the transections. The increase in the BBB score in transplanted animals was minor, from 2 to 4.5 to 6 points. Chen and others (2011) claimed a positive effect based on a one-millisecond reduction of MEP latencies between transplanted and vehicle animals, and there was also an increase in MEP amplitude. However, these findings did not correlate with changes in the BBB scores, and it was not stated how many rats actually recovered MEPs. Studies on Schwann Cells. Only one study assessed Schwann cells alone (Pinzon and others 2001) and found a positive effect, but neither blinded data collection nor allocation concealment was reported. There were only two control lesioned animals (with no electrophysiological recovery) versus two of nine transplanted rats that recovered evoked potentials, and no statistics were presented. The measured potentials were from the spinal cord, and although recovery was observed, the methods did not directly assess connection with the brain. Animal Number, Blinding, and Allocation Concealment. The mean number of animals per treatment group was 11 ± 5
in thoracic studies. Eight of 11 studies used >10 animals in at least the transplant group. However, some studies had a very imbalanced number of controls. For example, Lu and others (2012b) compared 26 transplanted animals to 6 controls. Pinzon and others (2001) compared nine transplanted animals to two transected controls. Eight of 11 studies were blinded, and none reported allocation concealment.
Summary of Effects of Cell Transplants in Thoracolumbar Lesions Overall, OECs appear to have a possible effect on electrophysiological recovery and a modest effect on the BBB score (i.e., ~3–4 points) because the four examined studies converged on similar findings, although three were from the same laboratory. There were many methodological concerns in studies testing NSCs, rendering an interpretation of their effect difficult notably because of the preservation of MEPs in lesioned control animals (thereby questioning the completeness of the lesion) and the discrepancy between studies with regard to MEP latencies in control animals, which ranged from approximately 1.5 milliseconds to approximately 8 milliseconds. Efficacy of MSCs was difficult to assess because there were missing data in the study by Kang and others (2012) and a change in the BBB score of
