Novel Use of Intraarticular Granulocyte Colony Stimulating Factor (hG-CSF) Combined with Activated Autologous Peripheral Blood Stem Cells Mobilized with Systemic hG-CSF: Safe and Efficient in Early Osteoarthritis

Abstract

Osteoarthritis (OA) tends to occur in older individuals frequently burdened with comorbidities and diverse pharmacological interactions. As articular cartilage has low regenerative power, potent local tissue engineering approaches are needed to support chondrogenic differentiation. Acellular preparation methods as well as approaches to coax endogenous reparative cells into the joint space appear to have limited success. Supported by our in-vitro and clinical studies, we propose that our novel intra-articular administration of human granulocyte colony stimulating factor (IA-hG-CSF) combined with autologous activated peripheral blood stem cells (AAPBSC) is safe and offers treatment advantages not seen with other cellular interventions in early osteoarthritis.

Osteoarthritis (OA) tends to occur in older individuals frequently burdened with comorbidities and diverse pharmacological interactions. ¹ Articular cartilage is avascular and alymphatic with low regenerative power. Potent local tissue engineering approaches are needed to create a stable, local regenerative environment with the necessary growth factors and signals, adequate nutritional supply, and available operative cells, all supporting chondrogenic differentiation. When this regenerative ensemble is orchestrated and assembled in place, it will exert its effects in unison, whether that is fusion with resident cells, stimulation and differentiation of local chondrocyte precursors, paracrine support, or a combination of all the above. Methods that do not include the aforementioned tissue engineering components may be at a disadvantage for successful cartilage regeneration. Supported by our in vitro and clinical studies, we propose the use of intraarticular granulocyte colony stimulating factor hG-CSF (IA-hG-CSF) as an adjuvant in cellular interventions in early OA.^2-4

Acellular preparation methods as well as approaches to coax endogenous reparative cells into the joint space appear to have limited success. ⁵ We and others have presumed that bone marrow and/or peripheral blood stem cells and mesenchymal stromal cells (MSCs), mobilized by systemic hG-CSF (S-hG-CSF), will “home” to the articular cavity and aid cartilage regeneration.^2,4-7 Okano et al. hypothesized that the significantly higher (P < 0.05) tissue repair rate 4 weeks postsurgery was induced by S-hG-CSF mobilized peripheral blood stem cell (PBSC) homing at the target cartilage defect. ⁶ S-hG-CSF’s short half-life of 3.5 hours is not supportive of local actions with such time lag and duration. ⁸ Similarly, the greatest limitation in Sasaki’s study was that even though they demonstrated an in vitro chondrogenic potential of MSC with hG-CSF added in the culture media, they did not evaluate whether MSCs actually proliferated at the site of the cartilage defect. ⁷ Their results are essentially supportive of the use of local hG-CSF as its effects were observed in the MSC culture. ⁷ In our 2014 in vitro study, we evaluated activated autologous peripheral blood stem cell (AAPBSC) attachment and cell proliferation on human early osteoarthritic cancellous tibial bone scaffold. ³ We observed that regardless of the preparation applied (AAPBSC-only, AAPBSC and hyaluronic acid-HA, AAPBSC, HA, platelet rich plasma [PRP], and hG-CSF) only approximately half of the seeded cells attached. ³ Cell proliferation was PRP and IA-hG-CSF dependent and significantly increased 2-fold compared to AAPBSC-only (P < 0.001) and 3-fold compared to AAPBSC plus HA (P = 0.03). ³ Furthermore, the local application of the preparation containing AAPBSC, HA, PRP, and IA-hG-CSF on the bone scaffold resulted in statistically significant and sequential increases in key chondrogenic factors such as SOX-9 (P < 0.05), COL-2 (P < 0.001), and Aggrecan (P < 0.01), with histological evidence showing newly formed hyaline cartilage. ³ Simply injecting s-hG-CSF begs the question of what percentage of mobilized PBSC would attach in the damaged cartilage in vivo, and of these, what percentage will subsequently induce chondrogenic differentiation in the absence of local IA-hG-CSF and other growth factors present in PRP. Apparently, as witnessed in the study of Ono et al., the limited effectiveness of bone marrow stimulation (BMS) did not improve after a 3-day S-hG-CSF administration. ⁵ Moreover, S-hG-CSF-induced leukocytosis might cause more leukocytes than PBSC ultimately homing in the knee joint from the peripheral circulation. ⁶ Proteases and reactive oxygen species released from white blood cells might create a proinflammatory intraarticular environment, not conducive to chondrogenesis. ⁹ With the addition of PRP, however, regardless of leukocyte content, the coexisting high concentration of growth factors, including endogenous G-CSF, may balance the proinflammatory components that in S-hG-CSF-induced leukocytosis are unopposed. ⁹ Autologous leukapheresis procures a leukocyte-poor concentrated volume of AAPBSCs. ¹⁰ To our knowledge, there are no human studies in early knee OA comparing the use of S-hG-CSF-only versus isolated intraarticular injections of bone marrow, peripheral, mesenchymal, or adipose stem cells. Adjunctive isolated hG-CSF treatment in diabetes does not appear to increase the likelihood of resolution of infection or healing of the foot ulcer. ¹¹

The first clinical study reporting the use of both systemic and intraarticular hG-CSF was our 2013 pilot study in early knee OA that failed conservative treatment. ² We combined S-hG-CSF mobilized AAPBSC with PRP, HA, and hG-CSF in the same preparation, and delivered it as 3 weekly consecutive IA knee injections, with the first injection of fresh cell preparation and growth factors approximately 6 hours post-leukapheresis. ² We reported significant quality-of-life improvements measured by Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score, Knee Injury and Osteoarthritis Outcome (KOO) scores, and histologically confirmed articular cartilage regeneration. ² In 2017, we enrolled 60 patients below the age of 60 years with early knee OA, in the largest clinical randomized controlled study (RCT) to date. ⁴ Other inclusion criteria were Kellgren-Lawrence (KL) stages 1-3, International Cartilage Repair Society (ICRS) grade III and IV lesions that failed conservative treatment for more than 6 months, not more than 3° varus or valgus deformity, and visual analog scale (VAS) scores of more than 40. Forty patients were randomized in 2 stem cell groups of 20 patients each. One group received IA-AAPBSC, HA, PRP, and IA-hG-CSF, while the other received IA-AAPBSC, HA, PRP, but no IA-hG-CSF, in both as 3 weekly consecutive injections. The outcomes were compared with a 20-patient non–stem cell group that received HA-only as 3 weekly consecutive injections. ⁴ The comparisons between the AAPBSC groups were double blinded but open when compared with the non–stem cell, HA-only control group, as leukapheresis cannot be a sham procedure. ⁴ Our RCT showed absence of notable adverse events and significant avoidance of total knee arthroplasty (TKA) in the AAPBSC groups at 12 months with no TKA reported in the cellular groups. ⁴ The AAPBSC groups also improved significantly versus the HA-only group at all WOMAC subscales, at all times. ⁴ Importantly, significant and differential IA-hG-CSF effects with an early onset (at 6 months) of total WOMAC score improvement, pain alleviation, and function amelioration were noted in the patients that received the IA-hG-GSF addition. ⁴ Stiffness, as it possibly depends more on connective tissue rearrangements and new cartilage formation, resolved later at 12 months. ⁴

Apart from the local IA-hG-CSF delivery, our cellular choice may account for the successful cartilage regeneration observed in our studies.^2,4 It is unknown which cell faction(s) are the most appropriate in regenerating cartilage. Studies using purified cell factions in vascular regenerative cell therapies appeared less successful compared with bone marrow or PBSC. ¹² Moreover, only autologous PBSC significantly decreased the amputation rate. ¹² The use of S-hG-CSF attenuates the impact of aging on bone marrow stem cells and bypasses purported age-related limitations and objections in using autologous PBSC, making AAPBSC an attractive cell source in early OA. ⁴ Our administered AAPBSC populations contain a mixture of endothelial progenitor cells (EPC) and MSC that after 3 weekly injections did result in histologically confirmed cartilage resurfacing.^2,4 Our final AAPBSC harvest sample contains 4.56% and 8.84% CD90+ and CD105+ cells, respectively, while CD34+ hematopoietic markers were seen in 0.05%. ³ EPC from both the CD34+ and the CD105+ populations promote early angiogenesis by upregulating vascular endothelial growth factor that appears instrumental in maintaining physiological cartilage oxygenation, ultimately improving chondrocyte survival rate. ¹³

Several nonhematological effects have been reported for hG-CSF and its receptor is present in numerous target organs (including brain) and cells (neuronal, immune, and endothelial cells). ¹⁴ hG-CSF’s immunomodulating, neuroprotective, and neuroregenerative potential, as well as its anti-apoptotic and anti-inflammatory properties, have been acknowledged and demonstrated repeatedly in diverse clinical settings.^15,16 Its systemic hematological effects are short-lived and in line with its short half-life. ⁸ Repeated systemic injections ¹⁵ or local delivery systems (osmotic pumps or injectable gels) are being evaluated for sustained release to exert desired nonhematological, regenerative actions. ¹⁶

Additional beneficial effects of local IA-hG-CSG could be mediated by its immunomodulating actions as an inducer of Fas ligand (FasL) on chondrocytes. ⁴ hG-CSF-treated tissue-engineered cartilage showed less infiltration of macrophages, with increased formation of cartilage after transplantation. ⁴ Furthermore, local IA-hG-CSF effects on skeletal myocyte development and regeneration have been generally overlooked, but increase of the satellite cell population and preservation of the satellite stem cell pool supporting long-term muscle regeneration and functional maintenance may be behind improvements in function and stiffness readily observed in our RCT. ⁴ In contrast, attenuation of neuropathic pain after peripheral nerve injury, swift pain relief in diabetic foot gangrene in humans, and a dramatic therapeutic effect on a rodent model of diabetic neuropathy have been reported with S-hG-CS and are mediated via BM/PBSC. ⁴ The combined systemic and repetitive, local IA-hG-CSF administration in our RCT may have resulted in the significant, early, and sustained pain relief we noted. ⁴

Concluding, we believe that our novel intraarticular knee administration of hG-CSF combined with AAPBSC is safe and offers treatment advantages not seen with other preparations. ¹⁷ IA-hG-CSF resulted in enhanced proliferation of AAPBSC and increases in key chondrogenic factors along with histological evidence of new hyaline cartilage formation. ³ Clinically, avoidance of TKA and early onset of improvements were statistically significant. ⁴ Local growth factor release is actively being been researched in other orthopedic conditions such as in rotator cuff tendon repair and anterior cruciate ligament graft. ¹⁶ The use of generic biosimilars significantly reduces collection and treatment costs even when local IA hG-CSF is employed. ¹⁰ Autologous leukapheresis procured AAPBSC can be cryogenically stored for future use in the same or contralateral knee, or even other joints, while the use of local IA hG-CSF will effectively compensate for its systemic absence. Further studies delineating the effect of IA hG-CSF in the absence of or with different cellular populations in early knee OA are needed.