Table 2.

Erythropoietin

Paper	Animal model and injury model	Intervention and timing	Experimental groups	Reported outcomes: • Histologic/biochemical/physiological • Behavioral
Guízar-Sahagún Spinal cord2009	Model: Adult Female Long–Evans Rats Injury: T9 Contusion NYU Impactor 10 g × 12.5 mm	EPO IP • 2000 U/kg @ 0h PI; 12 h pre-injury; 24 h PI Methylprednisolone (MP) • 30 mg/kg @ 0h PI; 2 h pre-injury; 2 h PI Cyclosporine-A IP • 5 mg/kg@ 0h PI; 12 h pre-injury; 24 h PI	SCI + • EPO • MP • Cyclosporine-A • Saline n = 7 per group/time point	Histologic/Biochemical/Physiological: Neither drug improved the amount of spared tissue at the injury site at 6w PI. Behavioral: No differences were observed in the number of rearing events in the open field at at 5w and 6w between the rHuEPO, MP and saline groups. No differences in BBB from 1 d to 6w. The number of rears was significantly less in the group treated with Cyclosporine-A.
Huang J Int Med Res2009	Model: Male SD Rats, 210 g Injury: T10 Weight Drop 10 g × 50 mm for 20 sec	EPO IP • 1000 IU/kg @ 0 h	SCI + • EPO (n = 20) • Saline (n = 20) Sham (n = 20)	Histologic/Biochemical/Physiological: EPO increased spared tissue, decreased the cavity volume at injury site. • Some neuronal regeneration was observed in the EPO-treated group. • In the saline group, ERK phosphorylation was increased at 7d PI while mitogen-activated protein kinase phosphatase (MKP-1) was inactivated. The EPO-treated group showed significantly lower p-ERK and significantly higher MKP-1 at 7d PI. Behavioral: EPO significantly improved BBB score in comparison to the saline group (∼11 vs.∼9, p < 0.01) at 7d PI.
Kontogeorgakos Arch Orthop Trauma Surg2009	Model: Female Wistar Rats, 270 − 300 g Injury: T10 Aneurysm Clip 0.7 N × 1 min	EPO SC • Low total dose (EPO-L): 1,000 IU @ 0h and 24 h PI • High total dose (EPO-H) 1,000 IU @ 0 h, then every 2nd day × 13 doses	SCI + • Low EPO (EPO-L) • High EPO (EPO-H) • Saline n = 10/group	Histologic/Biochemical/Physiological: Not reported. Behavioral: The mean score 17.3 +/- 1.2 in the EPO-L group and 14.7 ± 1.8 in the the EPO-H vs. 8.2 ± 0.8 in the control at 5w and 6w. The difference between all three groups was significant.
Yazihan Injury2008	Model: Adult Wistar Rats, 200 - 220 g Injury: T9-11 Epidural Clip Compression 40g × 30 sec	EPO IP • 150 IU/kg @ 1 hr PI Ketamine IP • 100 mg/kg @ 30 min PI (NMDA receptor antagonist)	SCI + • EPO (n = 7) • EPO + Ketamine (n = 7) • Ketamine (n = 7) • Untreated (n = 7) • Sham-operated (n = 7)	Histologic/Biochemical/Physiological: SCI decreased the levels of antioxidant enzyme catalase and glutathione (GSH), and increased the levels of TNFα and the lipid peroxidation product malonyldialdehyde (MDA). 12h PI, EPO significantly increased the catalase and GSH levels and decreased the TNF-α and MDA levels. • Application of ketamine before EPO treatment decreased effects of EPO. Behavioral: Not reported.
Fumagalli Eur J Pharmacol2008	Model: Adult SD Rats, 240 - 260 g Injury:T9 Univ of Trieste Impactor 1N × 1sec	EPO IP • 1000 U/kg @ 30 min PI Methylprednisolone (MP) • 30 mg/kg IP	SCI + • EPO • MP • Untreated Laminectomized n = at least 8 per group	Histologic/Biochemical/Physiological: EPO, but not MP, significantly up-regulated nerve growth factor (NGF) expression both caudally and rostrally to the lesion at 3d PI. At 1w PI, both rEPO and MP increased the NGF expression in the epicenter. • No effect of either drug on BDNF or basic fibroblast growth factor-2 expression. Glial-derived neurotrophic factor expression was enhanced by both drugs at 7d. Behavioral: EPO caused significant motor function recovery at 1w to 4w (BBB scores of 13.9 ± 0.5, 9.0 ± 0.2, and 9.5 ± 0.5 in EPO, MP, and Control groups respectively).
Mann, Exp Neurol2008	Model: Male SD Rats, 320 - 340 g Injury: T9/T10 Contustion OSU Impactor 200-260 kdyn	EPO IV • 5000 IU/kg @ 1 h PI Darbepoetin IV • 10 mg/kg @ 1 h PI	SCI + • EPO (n = 12) • Darbepoetin (n = 11) • Vehicle (n = 11)	Histologic/Biochemical/Physiological: Neither EPO nor darbepoetin improved white or grey matter sparing (EC staining) over controls. Behavioral: Neither EPO nor darbepoetin yielded any behavioral improvements on BBB scoring, footfall analysis, footprint analysis, or hindpaw sensation.
Pinzon, Exp Neurol2008	Model: • Adult female Wistar Rats, 220 - 280 g • Adult Female SD Rats, 240 - 260 g Injury: • T3 Aneurysm Clip, 50 g × 60 sec or 20 g × 10 s • T9 Contusion NYU Impactor 10 g × 12.5 mm	EPO IP • 1000 IU/kg @ 0, 24, or 48 h PI for aneurysm clip SCI • 5000 IU/kg @ 0 h PI, then q24 h × 7d for contusion SCI	SCI (Aneurysm Clip) + • EPO • Saline control SCI (Contusion) • EPO • Saline IP once • Saline IP, q24 h × 7 days n = 10-15/compression group n = 11/contusion group	Histologic/Biochemical/Physiological: In the aneurysm clip model, EPO did not result in improved preservation of tissue at the injury epicenter. In the contusion injury model, EPO resulted in better preservation of tissue at the epicenter and reduced cavitation, but the differences were not statistically significant compared to control animals. Behavioral: In the aneurysm clip model, EPO did not improve BBB scores up to 4w post-injury. In the contusion model, EPO did not improve BBB scores up to 7 weeks post-injury.
Vitellaro-Zuccarello, Neuroscience2008	Model: Male SD Rats, 240 - 270 g Injury: T9 Contusion with a modified Univ of Trieste Impactor	EPO IP • 5000 IU/kg @ 30 min PI	SCI + • EPO • Saline n = 14/group	Histologic/Biochemical/Physiological: EPO treatment increased AQP4 immunoreactivity and decreased GFAP immunoreactivity. EPO induces an increase in the total volume occupied by microvessels in the injured cord. Behavioral: 4w PI, the BBB locomotor scores were ∼15 and ∼8 for the EPO and control groups respectively (statistically significant).
King, Eur J Neurosci2007	Model: Male Wistar Rats, 200 - 220 g Injury: T10/T11 Unilateral Hemisection	EPO or cEPO IP • 40 μg/kg twice @ 30 min and 24 h PI	SCI + • EPO • cEPO • Vehicle n = 6/group/timepoint	Histologic/Biochemical/Physiological: Lesion size was significantly reduced by both cEPO and EPO. In addition, TUNEL (apoptosis) and b-APP (damaged axons) staining was decreased around the lesion site in response to either EPO or cEPO treatment. Schwann cell infiltration was increased with either treatment, however macrophage infiltration remained unchanged. Behavioral: Not reported.
Okutan, J Clin Neurosci2007	Model: Female WistarRats, 210 - 250 g Injury: T8 Weight Drop Contusion 40 g × cm	EPO IP • 1000 IU/kg @ 0 h PI Methylprednisolone (MP) IP • 30 mg/kg @ 0 h PI	SCI + • EPO • MP • Vehicle n = 8/group	Histologic/Biochemical/Physiological: 24h PI, MPO activity (neutrophil infiltration) was equally decreased by MPSS or EPO treatment. Activated caspase3 activity (apoptosis) was also similarly decreased by either drug. Behavioral: BBB scores were reported 24h PI. EPO treatment increased locomotor recovery (∼8) when compared to MP (∼4) or vehicle control (∼3).
Vitellaro-Zuccarello, Neuroscience2007	Model: Male SD Rats, 240 - 270 g Injury: T9 Contusion with a modified Univ of Trieste Impactor	EPO IP • 5000 IU/kg @ 30 min PI	SCI + • EPO • Saline n = 7/group	Histologic/Biochemical/Physiological: 4w PI, EPO preserved ventral white matter compared to saline, increased the number of NG2 expressing OPCs in the cord, increased 5-HT immunoreactivity, and reduced phosphacan immunoreactivity. Behavioral: 4w PI, the BBB locomotor scores were ∼15 and ∼8 for the EPO and control groups respectively (statistically significant).
Arishima, Spine2006	Model: Male Wistar Rats, 380 − 500 g Injury: T8-T9 Weight Compression 120 g × 2 min	▪ EPO IP • 5000 IU/kg @ 15 min and 24 h PI, sacrificed at 0, 6, 12 h, 1, 3, 5, or 7d PI	SCI + • EPO • Saline n = 4/group/sacrifice time	Histologic/Biochemical/Physiological: The number of apoptotic cells (mainly believed to be oligodendrocytes and motor neurons) was significantly reduced with EPO treatment (assessed by TUNEL or activated caspase3 positive cells). These anti-apoptotic effects were observed up to 7d PI. Cavitation volume was also reduced at the injury epicenter (H&E stain). Behavioral: Not reported.
Cetin, Eur Spine J2006	Model: SD Rats, 200 - 300 g Injury: T3 Aneurysm Clip 0.6 N × 1 min	EPO IP • 1000 IU/kg @ 5 min PI • @ 5 min PI, then q24h × 3d. • @ 40 min PI • @ 40 min PI, then q24h × 3d Methylprednisolone (MP) IP • 30 mg/kg @ 5 min PI, then 5.4 mg/kg/h q8h starting @ 50 min PI	SCI +  • EPO • Saline n = 8/group	Histologic/Biochemical/Physiological: Histological scores indicated that a daily dose of 1000 IU/kg for 3d preserved tissue better than a single 1000 IU/kg treatment. However, the combination of three 1000 IU/kg doses with methylprednisolone provided the best results. Behavioral: Locomotor recovery was analyzed using the swimming test 24h and 72h after injury. A daily dose of 1000 IU/kg for 3d promoted better behavioral recovery than a single 1000 IU/kg treatment. Contrary to Gorio (2005), the combination of three 1000 IU/kg doses with methylprednisolone provided the best results.
Grasso, J Neurosurg Spine2006	Model: SD Rats, 275 - 300 g Injury: T3 Clip Compression 58g × 1 min	EPO or asialoEPO IV • 10 μg/kg @ 24h PRIOR to SCI • 10 μg/kg @ 0 h PI • 10 μg/kg 3x/d, then twice per week	SCI + • EPO • asialoEPO • Saline n = 6/group	Histologic/Biochemical/Physiological: EPO increased GFAP expression (Western Blot) and the GFAP positive cells (immunoreactivity). Serum analysis confirmed that EPO does not need to be present within systemic circulation at the time of injury. Behavioral: BBB analysis at 6w PI revealed that EPO given at the time of injury was superior to EPO pretreatment (BBB of ∼16 vs ∼13), but both were better than saline control (BBB of ∼9). No benefit was observed from administering multiple doses of EPO as opposed to just once at the time of injury.
Boran Restor Neurol and Neurosci2005	Model: Male Wistar Rats, 180-220 g Injury: T6-T7 Weight Drop Contusion 50 g × cm	EPO IP • 5000 IU/kg @ 1 h PI Methylprednisolone (MP) IP • 30 mg/kg	SCI + • EPO • MP • Vehicle n = 20/group	Histologic/Biochemical/Physiological: Not reported Behavioral: At 14d PI (experimental endpoint), only EPO increased motor recovery as measured by a swimming test. Significant improvements began at day 4 and were observed until day 14.
Gorio, PNAS2005	Model: SD rats (gender unspecified), 240 - 260 g Injury: T9 modified Univ of Trieste Impactor	EPO IV • 100, 500, or 5000 IU/kg within 30 min PI EPO IP • 100, 500, or 5000 IU/kg within 30 min PI • 5000 IU/kg @ 24 h PI • 5000 IU/kg @ 48 h PI Methylprednisolone (MP) IP • 30 mg/kg	SCI + • EPO • MP • Saline IP n ≥ 18/group	Histologic/Biochemical/Physiological: EPO at 5000 IU/kg IP significantly increased spared tissue at and around the injury epicenter, an effect that was negated by co-treatment with MP. Both EPO and MP given alone both reduced pro-inflammatory cytokines MIP-2, TNF-α, IL-1β, and IL-6. Behavioral: Optimal effects on BBB locomotor scores were achieved with EPO 5000 IU/kg IP or 500 IU/kg IV given by 30min post injury. BBB scores at 28d PI were: 13 for EPO and 9 for saline. Delay in treatment by 24h or 48 h PI yielded a small benefit that lasted only for the first 3 weeks. Co- treatment with MP negated the positive behavioral effects of EPO.
Brines, PNAS2004	Model: • C57/BL6 WT and bcR KO Mice; 8-16 weeks of age Injury: T3 Steel Rod Compression 4 min	EPO IP • 10 μl/kg @ 0 h PI cEPO IP • 10 μl/kg @ 0 h PI	SCI + • EPO • cEPO • PBS n = 10/strain/group	Histologic/Biochemical/Physiological: The KO vs WT experiment confirmed that the bc-Receptor (bcR) mediates EPOs protective effect. Behavioral: The BMS locomotor scale was used to evaluate locomotion. EPO or cEPO mediated complete locomotor recovery in WT mice; however this was not seen in the bcR KO mice or PBS controls.
Kaptanoglu, Neurosurg Rev2004	Model: Male Wistar Rats, 215 - 260 g Injury: T7-T8 Weight Drop Contusion 50 g × cm	EPO IP • 100, 1000, or 5000 IU/kg @ 0 h PI Methylprednisolone (MP) IP • 30 mg/kg	SCI + • EPO • MP • Vehicle n = 8/group	Histologic/Biochemical/Physiological: 2h post injury – thiobarbituric acid-reactive substances were used to estimate lipid peroxidation. Out of all the treatments, EPO (5000 IU/kg) reduced peroxidation and protected spinal cord ultrastructure the greatest. Behavioral: Not reported.
Gorio, PNAS2002	Model: Female SD Rats, 180-300 g Injury: • T3, Aneurysm Clip 0.6 N × 1 min • T9 modified Univ of Trieste Impactor 1 N × 1 s	EPO IP 1st Model (aneurysm clip) • 1,000 IU/kg @ 0 h PI • 1,000 IU/kg/d @ 0 h PI, then q24h for 3d 2nd Model (contusion) • 500 IU/kg/d IP @ 1 h PI, then q24 h for 7d • 5000 IU/kg/d IP @ 1 h PI, then q24 h for 7d	SCI (Clip) + • EPO • Saline @ 0 h PI SCI (Contusion) + • EPO • Saline @1 h PI n = 14/treatment group n = 6/saline group	Histologic/Biochemical/Physiological: 7d after contusion injury, EPO resulted in a significantly smaller cavitation volume (∼25%) at injury site and significantly fewer apoptotic (TUNEL+) cells rostral to the injury site. Behavioral: After clip compression SCI, EPO as single or 3d dose significantly improved motor function at 7d PI. The final motor scores were near normal at 28d in treatment groups, but stay quite poor (10) in the saline group. After contusion SCI, a single 5000 IU/kg dose was just as effective as the 7d dosing, and both were more effective than a daily 500 IU/kg injection. Day 28 BBB scores were roughly 18 for the EPO treated (5000 IU/kg) and 10 for saline.

SCI: spinal cord injury; d: day, days; h: hour, hours; w: week, weeks; PI: post-injury; q24h: interval 24 hours;

T8: thoracic vertebra 8; C5: cervical vertebra 5; IP: intraperitoneal; IV: intravenous; SC: subcutaneous.

BBB: Basso, Beattie and Bresnahan locomotor test; BDNF: brain-derived neurotrophic factor; EPO: Erythropoietin; GFAP: glial fibrillary acidic protein; KO: knockout; IL: interleukin; MDA: malondialdehyde; MIP-2: macrophage inflammatory protein-2; MP(SS): Methylprednisolone (Sodium Succinate); MPO: Myeloperoxidase; SD rats: Sprague-Dawley rats; TNF: tumor necrosis factor; WT: wild type.