Abstract
Granulocyte colony-stimulating factor (G-CSF) is a major growth factor in the activation and differentiation of granulocytes. This cytokine has been widely and safely employed in different conditions over many years. In this translational study, G-CSF is administered to 19 patients with chronic motor complete spinal cord injury, and outcomes are reported. All 19 patients received subcutaneous G-CSF (5 μg/kg per day) for 5 days and were followed for at least 6 months. The American Spinal Injury Association (ASIA) scale was used for motor and sensory assessment, and the International Association of Neurorestoratology-Spinal Cord Injury Functional Rating Scale (IANR-SCIFRS) and the Spinal Cord Independence Measure (SCIM) III were used to assess improvements in the ability to perform basic daily tasks. At the 6-month follow-up, upper extremity motor scores improved by 10, which was statistically significant (p = 0.007), whereas there were no significant changes in lower extremity motor scores. Also, the median of light touch sensory scores improved by 5 (p = 0.001). Pinprick sensory scores significantly improved (p = 0.002). The median increment in SCIM III total score was 7 (p = 0.001). The improvements in bladder and bowel management as well as moderate distance mobility subscales were also significant (p < 0.05). Total IANR-SCIFRS scores changed from 17 to 32, which was statistically significant (p = 0.001); again the bladder and bowel management subscale improvements were statistically significant (p < 0.05). Mild side effects of the G-CSF treatment such as bone pain, rash, fever, neuropathic pain, and spasticity were noted in a few patients; all of them resolved after 1 week. Our results indicate that G-CSF administration is a safe process and is associated with neurological as well as functional improvement. This manuscript is published as part of the International Association of Neurorestoratology (IANR) supplement issue of Cell Transplantation.
Keywords
Introduction
Spinal cord injuries (SCIs) remain among the most devastating human ailments with ominous complications (7,38) and care costs totaling more than US$200,000 within the first 2 years after injury (28). Additional therapies may be warranted to reduce disability-adjusted life years. Achievement of this goal may require enormous scientific effort to achieve a clinically applicable solution worldwide for millions of patients who suffer each year from SCI (12). Clinicians, researchers, and patients should be optimistic, and the universal recommendations should be followed in conducting SCI clinical trials (19), with adherence to high ethical standards by researchers, clinicians, and journalists, to ensure that the results are communicated to the general public in a manner that reflects honesty, safety, and efficacy for the potential therapy (1,55). The Spinal Cord Independence Measure (SCIM) III as a scale for evaluating outcomes of rehabilitation (8) even under optimal conditions may be negatively affected by any delay in starting the program (31); therefore, neuroregenerative strategies should never interrupt optimal rehabilitation programs. In this study, we planned to restart rehabilitation programs after termination of treatment. Clinical trials need to be focused, safe, ethical, and backed up by robust translationally relevant preclinical research strategies (14).
Among common proposed cellular treatments, many experiments have used autologous bone marrow stromal cells (4); the results were relatively safe but the effect was limited. Olfactory ensheathing cells (25), Schwann cells (46), and activated macrophages have also been transplanted into injured spinal cord. The cell source and processing method not only should be feasible and safe but also be carried out with the least invasive and/ or noninvasive procedures (International Campaign to Cure Paralysis Guidelines) (13,34,51,56). Systemic drug administration with neuroprotective and/or cell activation effects has been considered as a treatment modality. Various drugs have been speculated for neurorestorative purposes for acute and chronic SCI, including steroid therapy, monosialotetrahexosyl ganglioside (GM-1 ganglioside), cethrin (rho pathway antagonists), riluzole (sodium channel blocker), nimodipine, minocycline (26), erythropoietin (3), and granulocyte colony-stimulating factor (G-CSF) (32).
The use of G-CSF has been suggested to be a viable alternative to bone marrow stem cell autografts in patients with SCI (15). Therefore G-CSF is a candidate drug for SCI management that can be cautiously translated into the clinic (10), taking into consideration that G-CSF is one of the few growth factors that have been shown to be safe for clinical use (6). Accordingly, two uncontrolled clinical studies (47,48) and one phase I/IIa clinical trial (54) with G-CSF have recently been reported with promising results. In the same way, in this study, safety concerns as well as possible effects of G-CSF administration for postrehabilitated motor complete SCI have been reported.
Patients and Methods
This study was conducted in the Brain and Spinal Injury Repair Research Center at Tehran University of Medical Sciences and was approved by their ethical committee. Nineteen patients with motor complete SCI, American Spinal Injury Association (ASIA) Impairment Scale (AIS) A or B, were chosen and enrolled after obtaining written informed consent for the study. The inclusion criteria included being between the ages of 18 and 50 years, accomplishing at least 3 months of active rehabilitation according to the consortium of spinal cord medicine guidelines, and at least 3 months' duration of SCI. The lesions were located at the lower cervical and/or thoracic levels. There was neither history of blood dyscrasia hematological problems on hematologist consultation or drug administration, nor evidence of systemic infection or complications of SCI (untreated pressure sores, urinary tract infection, severe constipation, and prominent depressive symptoms) in the included patients. Treatment consisted of 5 μg/kg per day of subcutaneous G-CSF (Filgrastim, Neupogen® Amgen, Thousand Oaks, CA, USA) for 5 days. Basal complete blood counts were performed with hematology analyzer Sysmex XT-1800i (Norderstedt, Germany) and were repeated daily for the next 5 days. An effort was made to ensure that the white blood cell (WBC) count did not exceed 50,000/mm3 in the samples; in any such case the drug was discontinued and resumed after a 24-h delay. The test was requested again at 3 and 6 months after treatment to assess chronic hematological effects. Also, any sign of hypersensitivity reactions was an indication for discontinuation of the treatment. After completion of the treatment, the patients resumed previous rehabilitation programs.
Patient Characteristics
No patients were lost during follow-up and none withdrew from the study. Patient compliance was assessed based on self-report and physician estimation according to laboratory findings. Among 19 patients in this study, there were 13 (68.4%) men and 6 (31.6%) women. The median age of patients was 29.0 years (range, 18–44 years), and the median duration of SCI after trauma was 34 months. The most common cause of SCI was a motor vehicle accident (68.5%). There were 10 (52.6%) patients with cervical and 9 (47.4%) with thoracic injuries (Table 1). The rostrocaudal length of the signal change area on T1-weighted magnetic resonance imaging (MRI) was between 6 and 20 mm, with the median value being 20 mm. The median of WBC count reached 30,000 (range, 10,000–55,400) after treatment.
Demographics and Clinical Features of the Patients
Unless otherwise indicated, values represent the number of patients with percentages in parentheses. SCI, spinal cord injury; WBC, white blood cell.
Assessments
Assessments were performed by independent observers composed of neurosurgeons, urologists, and occupational therapists. The tools employed included ASIA impairment scale (37), SCIM III (27), and the International Association of Neurorestoratology-Spinal Cord Injury Functional Rating Scale (IANR-SCIFRS) (25) forms. The assessments were performed at baseline and at 6-month follow-up, while any evidence of complications were recorded regularly and reported by the observers.
Statistical Methods
The median and the interquartile range (IQR) were used to describe the sample. For statistical analysis, the nonparametric Wilcoxon signed-rank test was applied to compare ASIA, IANR-SCIFRS, and SCIM III scores pre- and postintervention. In addition, subgroup analyses were performed based on the site of the lesion (cervical and thoracic). A p value < 0.05 was considered statistically significant.
Results
Findings Based on the ASIA Scale
The ASIA motor score increased significantly after G-CSF administration in all patients (10 scores, p = 0.007). The median of the obtained motor scores improved by 3 in the cervical subgroup (p = 0.011), but changes in motor scores were not statistically significant in the thoracic patients (Table 2). The median of the ASIA light touch sensory score improved by 5.0 (p = 0.001), whereas improvement in light touch was more significant in cervical patients (8 scores, p = 0.007) compared to thoracic patients. Pinprick sensory scores also significantly improved in all patients (p = 0.002), but the median did not change. After G-CSF administration, five ASIA grade A patients improved to grade B (41.7%) and seven patients remained at the ASIA grade A (58.3%); ASIA grade B patients had no improvement in AIS.
Comparison of Patients' Motor, Sensory, IANR-SCIFRS, and SCIM III Scores at Baseline and After G-CSF Administration
Ranges for score categories are as follows. Motor: minimum 0, maximum 100; light touch: minimum 0, maximum 112; pinprick: minimum 0, maximum 112. IANR-SCIFRS, International Association of Neurorestoratology-Spinal Cord Injury Functional Rating Scale; SCIM III Spinal Cord Independence Measure III; G-CSF: granulocyte colony-stimulating factor.
Functional Assessment
The IANR-SCIFRS overall score increased significantly (p < 0.001). A significant improvement was also observed for the SCIM III total score (p = 0.001) (Table 2). Comparison of baseline and post-G-CSF administration IANR-SCIFRS subscores was performed. We found that patients had significant improvement in the following subscales (p < 0.05): Standing Without Brace, Sitting, Turning Body Over, Transfers: Bed to Chair/ Wheelchair, Bathing, Dressing, Bladder Control, and Bowel Control. The highest rate of improvement was observed for Bowel Control (78.9%) and Bladder Control (63.2%) subscales (Fig. 1).
Functional rating scale before and after G-CSF administration.
Comparison between the changes in SCIM III subscores after G-CSF administration was performed based on SCIM III assessments. Our patients showed significant improvement in Bladder Management, Bowel Management, and Moderate Distance Mobility (p < 0.05). Similar to IANR-SCIFRS assessment scores, higher rates of improvement were observed for Bowel Management (47.4%) and Bladder Management (36.8%) (Fig. 2). Also, two female cases reported improved menstrual sensation, and two cases reported healing of chronic pressure sores.
Spinal cord independence measure before and after G-CSF administration.
White Blood Cell Count
Complete blood count was performed before, as well as daily after, each injection. The median WBC count was 6000/mm3 (4500–7000), whereas postinjection median value was 30,000/mm3 (10,000–55,400).
Complications
There were three individuals with transient bone pain and another patient with increased spasticity. Three other patients showed increased neuropathic pain that was amenable to transient conservative treatment. Transient accentuation of autonomic dysreflexia, palpitation, nausea, chills, fever, and rash happened in four different cases. All complications subsided spontaneously 1 week after the last dose of G-CSF.
Discussion
So far, three mechanisms have been proposed for G-CSF action: stimulation of stem cells released in peripheral blood to potentially mediate repopulation of the injured region with neural progenitor cells, inhibition of the proapoptotic environment due to neuronal injury, and reduction of the inflammatory response to traumatic injury (11). The biological mechanism for the beneficial effect of G-CSF on spinal cord ultrastructure (49) has been explained on the basis of white matter sparing (32), stabilization of presynaptic terminals in the spinal cord (9), upregulation of glutathione present in the central nervous system (CNS) (10,35,57), counteraction of apoptosis, enhancing connectivity in the injured spinal cord (40,45), and promotion of angiogenesis (29). This effect has also been confirmed in the peripheral nervous system (PNS) (43). On the other hand, G-CSF has also been used for nontraumatic motor neuron disease, especially in the case of amyotrophic lateral sclerosis, a disease involving mainly motor function (21,22,58). All of these reports form a credible body of experimental evidence that G-CSF has a robust effect on SCI in different experimental models and species.
In recent studies, G-CSF has been administered clinically for compressive myelopathy (47), myocardial regeneration (18), stroke (50), and amyotrophic lateral sclerosis (39,60). According to the first study involving G-CSF, administration at 10 μg/kg per day has been safe for patients with worsening symptoms of compressive myelopathy and has effectively improved the neurological status of the patients (47). Although in this study the model was nontraumatic myelopathies, it was conducted on patients with chronic cord lesion, so it still resembles our study in disease chronicity. In another multicenter clinical trial, G-CSF was administered to patients with worsening symptoms of thoracic compressive myelopathy and was associated with neurological recovery (48).
In some studies, G-CSF has been employed in the acute trauma setting, and it has been shown that initiation of G-CSF administration within 48 h after injury has had neuroprotective effects (54). Again in their study the authors have suggested that G-CSF possibly may be administered for more chronic cases; however, the time interval suggested is about 48 h after injury, which is much earlier than the time table of our patients (for whom the mean duration of disease is 34 months).
The proof of local homing of mobilized stem cells on MRI, obtained with iron oxide nanoparticle-labeled cells, has been a useful method for evaluating cellular migration toward a lesion site (33,52,53). We have not conducted any paraclinical evaluation for the establishment of cell homing and/or microenvironmental alterations; however, modern MRI and MRI spectroscopy may be of help for this problem, which may be tackled in future studies.
Various forms of G-CSF administration in addition to intravenous routes have been tested experimentally especially via the intrathecal route. Intrathecal G-CSF has produced significant motor neuron protection and exhibited beneficial effects on spinal cord ischemia-induced neurological defects (5).
Augmentation of the neuroprotective efficacy of G-CSF, when administered in combination with other neurorestorative modalities, has been reported. For example, combined therapy with neural stem cells and G-CSF yields better results in the injured spinal cord than treatment with either of the component alone (42). Also, cotreatment by bone marrow stem cell transplantation and G-CSF administration was a partially effective method for treating severe SCI (36). This combination not only has improved motor function recovery, but has also induced the accumulation of intrinsic microglia and the active proliferation of intrinsic oligodendrocyte precursor cells (41).
A close derivative of G-CSF called granulocytemonocyte colony-stimulating factor (GM-CSF) has also been frequently employed for experimental SCI models, with similar mechanisms of action (2,16,17,20,23,24,30,44); however, it has higher adverse drug reactions than G-CSF.
It has been reported that stem cell transplantation increases the likelihood of neuropathic pain in animal studies attributed to the aberrant regeneration of damaged axons (59). However, this phenomenon was observed only transiently in one of our patients and was amenable to simple analgesics. The observed functional and neurological changes in our patients may in part be due to the placebo effect of the drug and/or a remote possibility of autorecovery in chronic cases; therefore, a double-blind randomized clinical trial may be warranted to establish efficacy in the future. The majority of previous clinical studies were not double-blind randomized clinical trials and they possessed a small sample size. The absence of placebo controls makes this study and the others difficult to interpret as this is essential for a valid SCI clinical trial.
Conclusion
Our study suggests that G-CSF administration in SCI patients appears to be safe for 6 months postinjection, whereas longer follow-up may be necessary. The observed neurological improvements warrant further establishment by a randomized double-blind clinical trial.
Footnotes
Aknowledgments
The authors are grateful to Dr. Abbas Norouzi Javidan, Dr. Gholam Reza Tugeh, Miss Mahrokh Nik Mohammadi, Mrs. Maryam Kavoosi, and Miss Asal Derakhshan Rad for their kind contributions in the manuscript preparation. We would also like to thank the IANR organizing committee for their kind support. The authors declare no conflict of interest.