Recent Advancements in the Role of Orthobiologic Therapy for Spine Surgery: A Review on Current Evidence,Outcomes,and Future Applications

Abstract

Spine surgery technology and technique have improved considerably over the past few decades, thereby widening the breadth of pathologies that may benefit from surgical intervention. However, entities such as non-radicular low back pain, degenerative disc disease, and chronic spinal cord injury have limited if any reliable surgical treatment. As the body of literature and the reliability of preparation strategies for orthobiologics such as platelet-rich plasma and mesenchymal stem cells continues to improve, so has their potential for use in the treatment of spine pathology. The number of clinical studies and high-level trials that evaluated the use of orthobiologics in fields such as degenerative disc disease/discogenic back pain, post-decompressive pain, spinal cord injuries and surgical adjuncts during discectomy or fusion procedures has grown significantly over the past decade. While this field remains immature and most works are either uncontrolled or evaluate a small number of patients, the body of evidence that supports the efficacy of orthobiologics in areas where traditional surgical intervention is less effective continues to increase. Here, we review the current body of literature that assesses the use of orthobiologics in spine surgery and provide additional discussion as to current questions and future directions for the field. Particular attention will be paid to high-quality trials that have been completed or are currently in progress, and comment as to the quality and shortcomings of current protocols will be made.

Introduction

Chronic lower back pain affects more than 1 in 3 adults, and 1 in 10 have degenerative disc disease. The combined all-cause medical costs of these two conditions over 3 months in 2019 were over $800 billion USD. ¹ Musculoskeletal disease has been repeatedly identified as the most common cause of high health service utilization in multiple countries, ² and the demand for care due to musculoskeletal pain in some regions has increased 2- to 4-fold over a 40-year period. ³ As the global lifespan and the prevalence of obesity increase and elderly patients remain active for longer, the cost burden from chronic pain and degenerative disease is expected to increase.

There has been a significant increase in the demand for services related to lower back and neck pain over the past 20 years, particularly among elderly (age 65 and older) patients. Dieleman et al ⁴ reported a 6.5% year-over-year increase in spending on back and neck pain from 1996 to 2013, costs mostly driven by demand for pharmaceuticals, inpatient care, and visits to the emergency department. Such cost trends suggest that novel pharmaceuticals that provide lasting pain relief and keep patients out of the hospital can be both patient and cost-effective.

Regenerative medicine is defined by the United States Food and Drug Administration (FDA) as the translation of biology and engineering into therapeutic approaches for the regeneration, replacement or repair of organs. ⁵ Nearly all exogenous and endogenous compounds that can be procedurally manipulated or dosed to achieve a clinical effect can be defined as orthobiologics. Targets and techniques of particular interest have focused on optimizing the healing environment, inducing and accelerating the inflammatory and regenerative cascade, and directly replacing damaged cells.

Popular applications of such “orthobiologics” include the treatment of osteoarthritis, tendinopathy, and cartilage repair. Initial supporters of orthobiologics touted their potential to improve and reverse the underlying causes of the patient’s pain and dysfunction. However, much of the basis for these therapies was in vitro. Early clinical reports that studied the benefits of orthobiologics used “proprietary,” poorly described synthesis techniques on small samples of selected patients. High-quality works were rare. ⁶ As such, the literature inconsistently supported the efficacy of orthobiologics, insurance rarely reimbursed for such therapies, and patient adoption was consequently slow.

The emergence of recent high-quality evidence has prompted reconsideration of the value of orthobiologics and regenerative medicine in the treatment of pain and degenerative disease, particularly of the spine. Areas of particular interest include generalized back and neck pain, degenerative disk disease, post-decompression neuroinflammation, spinal cord injuries, and adjuncts to spinal fusions. This review aims to introduce and define common terms in orthobiologics and regenerative medicine, and to review the current evidence on the clinical efficacy of prolotherapy, platelet-rich plasma (PRP) and mesenchymal stem cell (MSC), therapy via intradiscal, facet joint, epidural or sacroiliac joint delivery.

Types of Orthobiologics

Prolotherapy

Prolotherapy introduces a noxious substance into an area of damaged or diseased tissue to induce an inflammatory and healing response. Prolotherapy in orthopedics has two desired effects: (1) trigger the inflammatory cascade to promote inflammatory cell influx and the subsequent deposition of collagen as a healing reaction and (2) “temporary neurolysis” of peripheral pain receptors for symptomatic relief. Used since at least the 1930s, many substances have been used as irritants for prolotherapy, including phenol, chemotactic agents, and glycerin. ⁷ More recent works have used hyperosmolar dextrose to dehydrate local cells to death, triggering a focal inflammatory and healing reaction. ⁸ Bae et al ⁹ in their meta-analysis of 10 studies and 750 patients who had their chronic musculoskeletal pain treated with prolotherapy found that targeted hypertonic saline injections were superior to normal saline and exercise for chronic pain due to any musculoskeletal etiology. However, the extent of this analysis was limited by the heterogeneity of the conditions studied and the lack of long-term follow-up beyond 1 year.

Platelet-Rich Plasma

PRP is an ultra-concentrated dose of the patient’s platelets. The dense granules, o-granules and lysosomes inside of platelets contain growth factors and cytokines that modulate inflammation, angiogenesis, chemotaxis and cellular proliferation. ¹⁰ PRP’s exact mechanism of action is unclear, but there is some evidence that its effects are anatomy- and disease-dependent. ¹¹ In contrast with cellular delivery protocols, PRP is entirely a milieu-inducing therapy. ¹² PRP contains few if any viable cells and instead accelerates the host’s own endogenous processes to produce a therapeutic effect. ¹³ Noback et al ¹⁴ in their cross-sectional survey study reported that PRP is the most common exogenous therapy used by sports physicians in the treatment of muscle and tendinous injuries both surgically and non-surgically, in particular rotator cuff tears, tennis elbow, patellar tendinopathy and articular cartilage injuries. A recent work documenting current practices in orthopedic regenerative medicine found that PRP was significantly less costly for patients than other commonly used treatments such as bone marrow concentrate and bone marrow aspirate concentrate. ¹⁵

Mesenchymal Stem Cells (MSCs)

Stem cells are capable of self-renewal and can differentiate into multiple phenotypes. Stem cells can be adult or embryo-derived, with the latter capable of a broader range of differentiation. However, given the ethical and legal boundaries and concerns regarding the use of embryonic stem cells, translational research has largely focused on adult stem cells. The multipotent nature of adult stem cells permits their differentiation into cell types of a similar origin. Such cells are capable of being harvested from the host and readministered into target injured tissues.

The large majority of orthobiologic stem cell therapies utilize adult mesenchymal stem cells. The minimum criteria for a cell to be considered an MSC per The International Society for Cellular Therapy ¹⁶ are: (1) the cells must be plastic adherent in culture; (2) the cells must express the surface antigens (CD105, CD73, and CD90) expressed by all MSCs, and not the surface antigens expressed by differentiated cells such as leukocytes, monocytes/macrophages, and endothelial cells; and (3) the cells must be capable of differentiating into osteoblasts, chondroblasts, and adipocytes. Bone-marrow aspirate concentrate (BMAC) is a non-processed aspirate typically of the iliac crest that is considered stem cell rich. MSCs are isolated and sorted following harvest from various tissues and include bone marrow-derived MSCs (bmMCs), adipose-derived MSCs (aMSCs) from subcutaneous fat through liposuction or lipoaspiration, synovial-derived MSCs (snMSCs) from the synovial and joint tissues, and birth-tissue-derived MSCs (btMSC) from the amnion, placenta and umbilical cord.

Two newer forms of MSC therapy include cellular bone matrices (CBM) and demineralized bone matrices (DBM). CBM utilizes allogeneic mesenchymal stem cells to combine osteogenic, osteoinductive, and osteoconductive factors into one product.^17,18 A prospective study found that 95.3% of patients undergoing lumbar surgery receiving an adjunctive allograft CBM achieved successful fusion at 24 months with significant improvements in Oswestry Disability Index (ODI), Visual Analog Scale for back pain (VAS-Back), and Visual Analog Scale for neck pain (VAS-Neck). ¹⁷ While this study was limited in that it assessed collective data on several different surgical approaches rather than assessing individual procedure effectiveness, this shows a promising role of CBM in regenerative orthopedics. DBM are also formed from allograft that undergo processing to remove the protein matrix of the bone. A review that analyzed current trials on spinal fusion showed limited support for the use of DBM alone and rather suggested its use as a graft extender. ¹⁸

Clinical Applications of Orthobiologics in Spine Care

Discogenic Pain and Degenerative Disc Disease

Pain originating from the axial anatomy is common, with an annual point prevalence of 13% for low back pain and 4.9% for neck pain.^19,20 While the origin of 85% of chronic lower back pain is unidentifiable, ²¹ up to 50% of the low back pain treated in specialized pain or orthopedic clinics is discogenic, while another 33% is facet-related. ²² Degenerative disc changes including desiccation and annular tearing can be used to identify potential pain drivers, and techniques such as computed tomography (CT) discography can reproduce symptoms and confirm disc integrity to avoid leakage. ²³ While the current body of literature supports the clinical efficacy of orthobiologics in the treatment of discogenic and non-radiating low back pain, the level I evidence supporting these indications is for PRP and MSC intradiscal injections for discogenic pain, PRP injections for facet disease, epidural injections of autologous conditioned serum and prolotherapy, and prolotherapy for sacroiliac joint pain. An additional study on facet joint prolotherapy reported negative results, and at the time of this work no intervention had more than one supporting level I study.

The large body of literature that utilizes prolotherapy in the axial anatomy is in the treatment of discogenic disease. Miller et al ²⁴ performed an uncontrolled prospective study in which 76 patients with lower back pain, radicular symptoms and intact disc anatomy received up to 5 biweekly disc space prolotherapy injections of a 50% dextrose, 0.25% bupivacaine solution. Sustained improvement in pain scores of 71% were reported by 43.4% of patients 18 months after treatment. Bupivacaine was used to provide rapid pain relief and confirm treatment localization, and Ohtori et al ²⁵ found no evidence that bupivacaine accelerated disc degeneration in a 5-year comparison of discoblock (bupivacaine alone) and discography (saline). Derby et al ²⁶ explored the benefits of the addition of 0.5% chondroitin sulfate and 20% glucosamine hydrochloride to 50% hypertonic dextrose prolotherapy in a group of 35 patients with lower back pain refractory to prior conservative and surgical treatments. Significantly decreased pain as measured with visual analogue pain score (VAS) after 18 months was noted in 65.6% of patients, although 81% of participants reported at least one post-procedure pain flare-up.

Initial pre-clinical studies evaluating the effects of PRP on human disc cells demonstrated decreased apoptosis, improved proteoglycan synthesis, and reduced expression of inflammatory cytokines.^27-29 Levi et al ³⁰ in their prospective cohort study reported that 47% (9/19) of patients with refractory lower back pain with imaging or discographic localization reported at least a 50% decrease in pain and 30% improved function as measured with the Oswestry Disability Index (ODI) 6 months after a single PRP treatment. Tuakli-Wosomu et al ³¹ in their randomized double-blind controlled trial of 47 patients with >6 months of refractory low back pain at one or more levels localized with discography, no surgical history, and mild to moderate disc degeneration defined as a protrusion <5 mm on magnetic resonance imaging (MRI) and a grade 3 or 4 annular fissure on discography were treated with either PRP or contrast alone. PRP was delivered at a single dose to one level or divided into equal doses at multiple levels. While follow-up was only 8 weeks, patients who received PRP had significantly improved pain, patient satisfaction, and function compared with controls. The concentration of PRP appears to correlate with its treatment effect. Jain et al ³² positively correlated the beneficial effects of PRP 6 months after treatment with the platelet counts of the administered sample in a prospective study of 25 patients with chronic lower back pain (r = 0.7 for the ODI and r = 0.73 for the pre-procedure numerical rating scale). A meta-analysis of these three prospective trials concluded that intradiscal PRP significantly reduced VAS pain scores and improved function as measured with the ODI 6 months after treatment, although short-term benefits (improvements at time points earlier than 6 months) were mixed. ³³ However, the authors noted that only one of these trials was controlled, the preparation of the PRP in all three studies was heterogeneous (one work quantifying the variability in the dosages given), and long-term effects >6 months were not available. A way to standardize PRP dosing may be to use PRP releasate, which is the concentrated growth factors and cytokines within the PRP isolated from a clotted sample. Akeda et al ³⁴ injected PRP releasate into 14 patients with at least 3 months of refractory, localizable discogenic back pain and monitored participants for a mean of 10 months. Mean VAS pain score and Roland-Morris Disability Questionnaire were significantly improved from baseline measurements 1-6 months after treatment and persisted 1-year post-injection in the 9 patients who presented for follow-up. Zielinski et al ³⁵ conducted a double-blind, randomized, placebo-controlled study on a small cohort of 26 patients with discogenic pain to determine the effects of PRP on clinical improvement, ultimately finding clinical improvement in 17% of patients, but also a clinical decline in 5% of the group. More recently, however, Navani et al ³⁶ conducted a multicenter, prospective, crossover RCT with long-term follow-up of 12 months and found that PRP led to significantly lower discogenic low back pain using patient reported outcomes on the numeric rating scale. This study overcame limitations of previous work such as dosing and follow-up time. Future work should replicate this study at a larger scale and assess outcome differences by disc injury.

The initial evaluation of cellular-directed therapies in the treatment of discogenic and degenerative spine disease focused on the reimplantation of cultured autologous intervertebral disc chondrocytes (ACI). Anderson et al ³⁷ in their review of cellular therapies in the treatment of degenerative disc disease identified cell viability, source, preparation, and dosing as key considerations when evaluating the efficacy of ACI, and while identifying some reliability of using imaging to measure improvement emphasize the importance of measurable clinical outcomes. Interim and subgroup analyses of early participants in the EuroDISC randomized controlled trial that compared discectomy + ACI vs discectomy alone found that patients treated with ACI had significantly improved pain and patient-reported function and disability scores that persisted for at least 2 years.^38,39 The difference in efficacy between autologous and cell-line chondrocytes is unclear, and the advantages of commercially available chondrocyte preparations include improved regulation and dosing consistency. Couric et al ⁴⁰ in their uncontrolled prospective pilot trial injected 1-2 mL of 10⁷/mL of commercially purchased juvenile chondrocytes into 15 participants, noting a significant improvement in patient-reported outcomes 12 months after injection and radiographic improvements 6 months post-injection.

Early reports of the use of MSCs in the treatment of spine disease focused on treating degenerative disease, radiculopathy and spinal instability in patients who were poor surgical candidates. The first report of the clinical use of MSCs in degenerative disc disease by Yoshikawa et al ⁴¹ followed two patients with chronic lower back pain and lower extremity numbness who received bmMSCs impregnated into a collagen sponge and delivered via a discectomy approach into a diseased segment. Both patients reported persistent symptom relief two years after treatment, and improved water content of the target disc space was observed on the follow-up MRI. These favorable findings were replicated by the follow-up 10-patient pilot study by Orozco et al, ⁴² and the viability of the direct injection of btMSC into the disc with good 2-year response was shown by Pang et al. ⁴³ Elabd et al ⁴⁴ reported a more than 50% improvement in the strength and mobility of 3/5 patients 4-6 years after treatment with hypoxic cultured bmMSCs. The viability of aMSCs in degenerative disc disease was demonstrated by Kumar et al, ⁴⁵ who noted a marked improvement in VAS, ODI, and Short Form-36 (SF-36) scores that lasted for 12 months in 6 of the 10 patients who “responded” to MSC therapy. Six-year safety and radiographic efficacy of bmMSCs was demonstrated by Centeno et al. ⁴⁶

The efficacy of stem cell therapies appears to be dose and source dependent. The dose dependency of BMAC intradiscal injections was demonstrated by Pettine et al, ⁴⁷ who in a trial of 26 patients showed that BMAC improved symptoms at high and low concentrations, but a higher concentration of cells was necessary for functional improvement at all ages, and for pain improvement in patients > 40 years of age. A 3-year post-operative follow-up of these patients reported that only 6 (23.1%) went on to require surgical intervention for their degenerative disease, while the functional and VAS pain improvements persisted in the other 20.^48,49 The importance of delineating between responders and non-responders is best seen in the randomized controlled trial by Noriega et al, ⁵⁰ who randomized 24 patients between 25 million allogeneic MSCs injected into the disc or paraspinal infiltration of bupivacaine (n = 12 per group). The authors note that the pain and functional outcomes of the experimental group split clearly into a group of responders (n = 5) and non-responders (n = 7). The responders mirrored near-perfect therapeutic outcomes while the non-responders matched the control group almost perfectly. A mean 3.5-year follow-up study of these same patients showed that responders remain as such long term, while only a small number (3/7) of non-responders converted to responders. ⁵¹ Likewise, a randomized clinical trial by Gornet et al ⁵² evaluated the efficacy of allogeneic disc progenitor cell intradiscal injections in a cohort of 60 patients with single-level lumbar degenerative disease, separating this group into low and high-dose administrations. Ultimately, the high-dose group achieved the primary endpoint of mean visual analog scale pain improvement greater than 30% at 52 weeks (−62.8%, P = 0.005) which was maintained until week 104. On the other hand, the low-dose group showed minimal improvements, thereby exhibiting the dose-dependent nature of this intervention. Cannon et al ⁵³ conducted a pre-clinical study that showed MSC effectiveness in intervertebral disc degeneration is tissue source dependent. They found that amniotic membrane-derived MSCs were the most effective with the lowest rates of senescence and apoptosis under acidic microenvironments.

Delivery medium may be key to a higher rate of treatment response. PRP and aMSCs delivered as stromal vascular fraction after mini-lipoaspirate into one or more degenerative discs were well tolerated (no severe adverse effects) and significantly improved flexion, VAS pain ratings and quality of life scores as measured with the Short Form-12 (SF-12) and the Short Form McGill Pain Questionnaire 6 months after treatment. ⁵⁴ There may also be a role for structural rather than substantive augmentation of the degenerative disc. The Viable Allograft Supplemented Disc Regeneration in the Treatment of Patients with Low Back Pain With or Without Disc Herniation (VAST) prospective randomized multicenter trial randomized 220 subjects at 15 clinical sites to percutaneous fluoroscopy-guided injection of commercial viable allograft vs saline into a diseased intervertebral disc space, reporting that allograft-treated patients had improved ODI and VAS scores 6 and 12 months post-injection (VAS pre-op 54.8, 6 months 16.0, 12 months 12.3; ODI pre-op 53.7, 6 months 18.5, 12 months 15.7) vs controls (VAS pre-op 55.3, 6 months 41.0, 12 months 19.7; P-values not provided). ⁵⁵

Post-Decompression Pain

Surgical decompression is indicated in the setting of radicular pain caused by a structural abnormality, such as a disc herniation following the failure of an attempted course of non-surgical management. However, pain can persist after decompression. This post-operative radiculitis is defined as transient radicular pain after initial symptoms and resolving within 6 months of surgery. ⁵⁶ Boakye et al ⁵⁷ reported in a cross-sectional analysis of laminectomy patients that 77% of those surveyed had clinically significant short- or long-term “post-decompressive neuropathy,” which the authors defined as the development of unrelated neuropathic pain following a lumbar laminectomy for chronic degenerative disease. While that work was limited to central decompressions, this phenomenon appears to exist after peripheral nerve decompressions as well. Multiple reports have linked post-decompressive neuropathic pain with a response similar to that of an ischemic-reperfusion injury, which is manifest by monocyte infiltration into the dorsal root ganglia and peripheral nerves, microglia and astrocyte activation and the increased expression of acute phase cytokines such as TNF-α, IL-1β, and IL-6, which in turn mimics and triggers a neuro-inflammatory response.^58-61

The large body of data on the characterization and treatment of post decompression pain is pre-clinical. Musolino et al ⁶² injected bmMSCs into the 4th lumbar dorsal root ganglia of rats who underwent single ligature nerve constriction of their sciatic nerve. Animals treated with bmMSCs had a reduced number of allodynic responses to cold stimuli, but treatments did not resolve either mechanical or thermal hypersensitivity 56 days after treatment. However, central treatment via intrathecal injection of bmMSCs by Chen et al ⁶³ in a similar sciatic nerve constriction model observed both long term (4-5 weeks) relief of neuropathic pain and evidence of migration to the affected dorsal root ganglia via CXCL12-mediated chemotaxis. The authors speculated that the relative efficacy of their intrathecal injection model vs more direct delivery mechanisms may be due to the local trauma caused by the injection. MSCs may also have a role in the treatment of more mechanical forms of chronic pain. Guo et al ⁶⁴ performed a long-term (22-week) observational study of intravenous vs local delivery of bmSCs following experimental ligation of the tendon of the anterior superficial part of the rat masseter muscle, which is a known cause of chronic orofacial pain in this model. Both delivery methods resulted in lasting, long-term pain relief via central and peripheral nerve receptor downregulation.

A single high-level study evaluated the use of orthobiologics in the mediation of post-decompressive symptoms. A randomized controlled trial of 269 patients evaluated the use of amniotic-derived products (ADP)—which have been found to reduce the expression of MMP- and TGF-β-mediated scarring and contraction in vitro ⁶⁵ —and BMAC at the surgery site following endoscopic lumbar decompression. ⁶⁶ Patients received either ADP, BMAC, both, or no supplementation. The authors reported that patients treated with ADP or BMAC had significantly improved VAS leg pain and back scores over 9 months of follow-up, although this translated into modest, short-lived improvements in function (ODI improvement only 2 weeks post-op) and SF-36 (up to 4 = month improvement) compared with untreated patients. No synergy between ADP and BMAC was observed.

Spinal Cord Injury

Spinal cord injuries (SCI) represent financial and personal devastation for the patient and a substantial loss of quality-adjusted life years. Occurring in approximately 180000 people per year, ⁶⁷ mortality from SCIs is twice as likely in developed countries, ⁶⁸ and outcomes suffer in regions with poor infrastructure and immature early resuscitation and rapid airway management strategies during initial management and stabilization. ⁶⁹

Early management of SCIs largely focuses on reducing inflammation, minimizing ischemia, and stabilizing/decompressing critical vertebral injuries. ⁶⁹ Minimizing inflammation while maintaining spinal cord perfusion is prognostic of neurologic recovery. ⁷⁰ However, while evidence supporting suppressing the inflammatory response with intravenous steroids during the acute injury period is controversial at best, ⁷¹ pre-clinical works support the potential use of MSCs to halt and potentially reverse the Wallerian degeneration and scarring that lead to permanent neurologic injury. Yousefifard et al ⁷² injected 1 million bmMSCs or btMSCs into the injury site of a rat SCI model. Both MSC groups improved motor recovery and neuropathic pain relief, and btMSCs had superior post-injection survival compared with bmMSCs. Additional pre-clinical works utilizing MSCs reported improved locomotor function, decreased glial scarring, ⁷³ reduced autophagy at the injury site, ⁷⁴ and improved local angiogenesis^75,76 and myelination. ⁷⁷ These benefits have largely been rendered through reduced inflammatory and degradative markers such as matrix metalloproteinases and hypoxia inducible factor, ⁷⁸ and increased anabolic factors such as transforming growth factor-beta. The Notch pathway may also play a role in neurologic recovery after severe spinal cord compression. Cheng et al ⁷⁹ reported increased levels of vascular endothelial growth factor (VEGF), von Willebrand factor (vWF), and Notch-1 1 week after decompression of the injured spine, and that Notch-1 inhibition inhibited post-decompression motor recovery. Questions regarding the utility of pre-clinical therapies relate to localization, survival and vector-attributable functional improvement. The final of these questions is hardest to answer, and largely relies on the creation of a reliable “steady-state” SCI, which is easiest to achieve in a chronic injury model. Ryu et al ⁸⁰ in a retrieval analysis of aMSCs transplanted into a canine subacute (1 week) SCI model noted increased neural differentiation of the stem cells in the setting of quantitative improvements in nerve conduction velocity and Olby pelvic limb functional scores. Localization was key to efficacy, and therapeutic efficacy can be enhanced with a viable scaffold that maximizes therapeutic localization and duration. The meta-analysis by Yousefifard et al ⁷² of pre-clinical works evaluating scaffolds and MSCs in the treatment of acute SCIs found that the combination of scaffolds with either btMSCs or bmMSCs improved motor function recovery compared to the use of a scaffold (standardized mean difference 2.00, P < 0.0001) or vector alone (standardized mean difference 1.58, P < 0.0001). Phase II testing of commercially produced MSC-impregnated scaffolds is ongoing (NCT02688049). Moreover, Mu et al ⁸¹ observed the effects of human umbilical cord mesenchymal stem cells (hUCMSCs), neural stem cells (NSCs), and epidural electrical stimulation (EES) on enhancing recovery in murine spinal cord injury models. In vitro, NSCs showcased pathways associated with increased differentiation and expression of growth factors and in vivo, combined cell therapy along with EES were shown to decrease inflammation and enhance neuron repair more efficaciously than either intervention alone. This pre-clinical study further provides evidence for the utilization of orthobiologics in the treatment of spinal cord injury in combination with other methodologies such as neuromodulation.

While pre-clinical studies are largely optimistic regarding the efficacy of MSCs on acute SCI outcomes, current clinical evidence is limited primarily due to our inability to predict how acute injury patients will recover with supportive interventions alone. Geffner et al ⁸² reported that bmMSC delivery via direct spinal cord, spinal canal, or intravenous injection into patients with an acute SCI was safe, without systemic or oncologic side effects. Xiao et al ⁸³ presented two-patients, one with a complete T11 SCI and another with a complete C4 injury, into whom they placed a commercially produced umbilical btMSC graft during the decompressive surgery. The authors reported that both patients improved markedly over a 1-year follow-up, although could not identify the MSC scaffolds as the reason why. A phase I/II clinical trial (NCT02917291) is currently enrolling patients with acute SCI for treatment with allogeneic aBMC, and another seeks patients for bmMSC therapy (NCT04528550).

Glial scarring represents a physical block to axonal regeneration in the setting of a chronic SCI. Scar resection has been discussed, but the efficacy of this strategy is predicated on optimizing the microenvironment for axonal healing and providing a scaffold to direct neurons across the resected scar. ⁸⁴ The safety of MSC transplantation via intravenous or local site delivery has been established in small phase I clinical trials.^30,85-91 Cristante et al ⁹² reinfused autologous cryopreserved bmMSCs into 39 randomly selected patients with a complete cervical or thoracic SCI at least 2 years after injury. The authors reported that 26/39 patients (66.7%) had an objective increase in lower extremity somatosensory evoked potentials 9 months after stem cell treatment, although no factors predictive of treatment response were identified. A phase III trial of a single intradural MSC administration in patients with a chronic SCI (minimal 12 months post-injury) was terminated due to poor efficacy (2/16 patients with motor improvement), which was attributable to the lack of multiple MSC administrations. ⁹³ Park et al ⁹⁴ had somewhat better outcomes with three separate administrations of autologous bmMSCs 4 weeks apart into patients with ASIA A or B cervical SCIs at least 1 month after injury. The authors reported that 6 months after treatment 6/10 (60%) patients had significant improvements in upper extremity motor function while 3/10 (30%) had improved completion of activities of daily living as determined by a rehabilitation specialist. The largest completed trial of intrathecal MSCs was published by El-Kheir et al, ⁹⁵ who reported that bmMSCs injected into patients with chronic SCIs (minimum 12 months post-injury) resulted in functional improvement (as measured with the functional rating mean score) in 23/50 (46%) of cases 18 months post-injection, with high response rates in those with thoracic SCIs and smaller cord lesions. Of note, the comparison group used in this trial was historical, and the bmMSCs were filtered to include only those that adhered to a poly-l-lysine coated dish, which represents a departure from the preparation techniques used in earlier works. While currently enrolling phase II/III clinical trials (NCT02687672, NCT03505034, NCT03521336, NCT03521323, NCT02574585, NCT04520373, NCT03225625, NCT04213131, and NCT04288934) are designed to be adequately powered to report on the efficacy of MSCs in the treatment of acute or chronic SCIs, MSCs may also have some role in the management of sequelae related to SCIs. Sarasua et al ⁹⁶ reported that autologous bmMSCs injected into a closed cavity created during pressure ulcer debridement led to uncomplicated healing without additional surgeries within 21 days of surgery in 15/22 (70.4%), and 19/22 (86.4%) healed after a single repeat treatment.

Adjuncts in Spine Surgery

Few clinical works have evaluated the impact of orthobiologics on surgical outcomes, as therapies are largely considered preventative measures. Anderson et al ⁹⁷ randomized 80 subjects who underwent an elective lumbar microdiscectomy to receive either amniotic stem cells or no treatment to the microdiscectomy site following the procedures. The authors found that ADM improved ODI and SF-12 scores 24 months after surgery (ODI: 6.6 ± 1.3 for ADM, 14.4 ± 3.3 controls, P = 0.02; SF-12 scores graphed but not stated), although it was underpowered to show if ADM treatment could reduce the rate of recurrent disc herniation (0/40, 0% with ADM vs 3/40, 7.5% in the control group, P > 0.05). Kim et al ⁹⁸ conducted a prospective, post-market study on Vibone Viable Bone Matrix (VBM), a cellular matrix comprising osteogenic, osteoinductive, and osteoconductive factors, and its effects on clinical and radiographic outcomes in 104 cervical and lumbar spinal fusion surgeries. At 12 months, they found improvements of 51.5% in the VAS for pain, 42.6% in the Neck Disability Index, and 55.4% in the ODI, all of which were significant. All patients achieved clinically significant improvements with fusion rates of 88.1% per patient in cervical and 97.6% in lumbar patients. Per-level fusion rates were 98.5% for cervical and 100% for lumbar surgeries. However, these findings are limited in that CBM was not compared to a control group for these reported pain outcomes.

It is challenging to measure the effects of orthobiologics on spine fusions due to the low incidence of non-unions, making such works difficult to power. Blanco et al ⁹⁹ embedded BMAC into tricalcium phosphate to augment a single-level spine fusion, reporting an 80% fusion rate without adverse effects. However, this work was uncontrolled, so the actual benefit of this strategy could not be quantified. Mochida et al ¹⁰⁰ injected 1 million activated nucleus pulposus cells into a diseased disc adjacent to a level indicated for fusion, reporting that the technique itself was safe and did not appear to accelerate degeneration, but conceded that they were unable to show that the cell therapy had any ability to prevent adjacent segment disease.

Limitations and Future Directions

As we are early on Rogers’ bell curve in the adoption of orthobiologics in spine care, it is unsurprising that current works are presently methodologically limited. Patients paid out of pocket to participate in some trials, ³⁰ a factor that is likely to skew patient-reported outcomes given their “buy-in” to that technology. Protocols for cell harvesting, processing, and administration are often “proprietary” and incompletely described, guaranteeing that the preparation strategies used for PRP and MSCs are heterogeneous between works and difficult to validate. In some cases, this is due to commercial and competitive concerns. ¹⁰¹ While this selective opacity can limit our mechanistic understanding of more effective therapies, the benefits of commercial preparations include more consistent dosing and preparation, and potential price efficacy when compared with the need for a second adjunct intervention and the cost of the in-house processing of autologous preparations. Allogeneic MSCs (6 million) mixed with hyaluronic acid were well tolerated and yielded clinical benefit up to 3 years after treatment in a randomized controlled trial of 100 patients. ¹⁰² However, the randomized controlled trial of 25 million allogeneic MSCs vs bupivacaine by Noriega et al^50,51 with a slow uptick in treatment response years after therapy may question the clinical efficacy of allogeneic preparations. Further works are required to achieve clarity regarding the relative efficacy of allogeneic vs autologous stem cell therapies. Our overall understanding of what predisposes to a clinical response to orthobiologics therapy must also improve so that informed, cost-effective, patient-specific decisions can be made between autologous and allogeneic cell preparations (Tables 1–4).

Table 1.

Key Works Studying Discogenic Pain and Degenerative Disc Disease

Authors	Study design	Level of evidence	Experimental groups	Outcomes	Significant findings
Miller et al	Prospective, uncontrolled	II	5 biweekly disc space prolotherapy injections of 50% dextrose, 0.25% bupivacaine (n = 76)	Sustained improvement in VAS pain scores (>20% decrease)	43.4% of patients were responders, with a mean improvement in VAS 18 months post-intervention of 71%
Levi et al	Prospective, uncontrolled	II	Single injection of PRP into one or more intervertebral discs for discogenic pain (n = 22)	Success (50% decrease in VAS, 30% decrease in ODI 1, 2 and 6 months post-treatment	14% success at 1 month, 32% success at 2 months, and 47% success at 6 months
Tuakli-Wosomu et al	Prospective, double-blind randomized controlled trial	I	Intradiscal PRP (n = 29) vs contrast (n = 18)	FRI, NRS, SF-36, NASS Outcome Questionnaire 8 weeks post-treatment	FRI significantly decreased after PRP (38.0 vs 44.5 in controls, P = 0.3), as did best pain score (2.0 after PRP vs 2.7 in controls, P = 0.02). All other outcomes equivocal
Jain et al	Prospective, uncontrolled	II	Single injection of PRP into one or more levels with discogenic pain (n = 20)	NRS and ODI	11/20 patients had a > 50% decrease in NRS, and 13/20 had a > 50% decrease in ODI
Akeda et al	Prospective, uncontrolled	II	PRP releasate injected into a single disc in the setting of discogenic pain (n = 14)	VAS, RDQ, Disc Height Index (DHI), MRI disc signal measurements	Persistent improvement in VAS 12 months after treatment. No change in DHI or MRI disc signal
Bertagnoli et al	Prospective, multicenter, double-blind randomized controlled trial (EuroDISC Trial)	I	Autologous disc-derived chondrocyte transplantation (ACI, n = 27) vs controls (sequestrectomy only, n = 26)	ODI, QBPD, VAS, X-rays and MRI	2 years post-implantation the ACI group had sig decreased ODI (2 vs 6.5 in controls), OPDQ (4 vs 13 in controls) and VAS (9 after ACI vs 15 in controls)
Couric et al	Prospective, uncontrolled	II	Single injection of allogeneic juvenile chondrocytes into one degenerative disc from L3-S1 (n = 15)	SF-36, ODI, NRS	Significant improvement in ODI (20.3 vs 53.3 at baseline, P < 0.0001), NRS (3.1 vs 5.7 at baseline, P = 0.003) and SF-36 physical subscore (50.5 vs 48.5 at baseline, P = 0.0002) 1 year after treatment
Kumar et al	Prospective, uncontrolled	II	Single injection of HA/adipose-derived MSCs into single level of patients with chronic back pain (n = 10)	VAS, ODI, SF-36	Decreased VAS (2.9 ± 1.7 at 1 year vs 6.5 ± 1.3 pre-treatment, P = 0.004) and ODI (16.8 ± 9.8 at q year vs 42.8 ± 15.0 pre-treatment, P = 0.01)
Centeno et al	Prospective, uncontrolled	II	Single injection of bone marrow-derived MSCs into degenerative disc with posterior bulge (n = 33)	NPS, SANE rating, FRI	Significant improvement 5 years post-treatment in NPS (3.2 from 5.2, P < 0.05) and FRI (35 from 61, P < 0.05)
Pettine et al	Prospective, uncontrolled	II	Single injection of autologous bone marrow concentrate into patients with degenerative disc disease (n = 26)	ODI, VAS, MRI Pfirrmann grade	6/26 patients went on to require surgery 3 years post-injection. Non-operative patients had significant improvement in ODI (17.5 ± 3.2 from 56.7 ± 3.6, P < 0.001) and VAS (21.9 ± 4.4 from 82.1 ± 2.6, P < 0.001). No degeneration on MRI
Noriega et al	Prospective, double-blind randomized controlled trial	I	Single injection of allogeneic bone marrow-derived MSCs (n = 12) vs controls (n = 12)	VAS, ODI, MRI Pfirrmann grade	Therapeutic efficiency 0.6 for pain relief, 0.71 for disability. Decrease in > 0.6 in Pfirrmann grade (P < 0.01)
Beall et al	Prospective, double-blind randomized controlled trial (VAST trial)	I	Single injection of nucleus pulposus allograft (n = 12) vs saline controls (n = 12) into 1-2 degenerative disc spaces (L1-S1)	VAS, ODI, MRI	12 months post-treatment VAS in the treatment group improved from 54.8 to 12.3 vs 55.3 to 19.7 in controls, and ODI in the treatment group improved from 53.7 to 15.7 vs 49.3 to 9.3 at 12 months. Improved MRI signal in treated disc spaces.
Tschugg et al	Prospective, single-blind randomized controlled trial	I	24 patients randomized to receiving the NDisc (proprietary autologous cell compound for chondrocyte transplant, n = 12) vs carrier alone (n = 12)	Adverse events only	No severe side effects or persistent laboratory abnormalities
Amirdelfan et al	Prospective, multicenter randomized controlled trial	I	100 patients with chronic low back pain and degenerative lumbar disease for > 6 months and failing > 3 months of conservative management randomized to receive 6-million allogeneic mesenchymal precursor cells (MPCs, n = 30), 18-million allogeneic MPCs (n = 30), HA vehicle alone (n = 20), and saline (n = 20)	VAS, ODI, SF-36, Work Productivity and Activity Index	VAS improvement from 6-million MPCs superior to saline and HA-vehicle at 6 months, 18-million MPCs superior at 12 months. 6 and 18 million MPCs superior to saline and HA 36 months after treatment. ODI at 36 months significantly improved after both 6 and 18 million MPCs vs HA and saline. SF-36 physical component significantly improved in 18 million MPC group vs saline and HA controls.
Park et al	Prospective, single-arm, multicenter trial	II	86 patients undergoing posterolateral or interbody lumbar fusion surgery with CBM allograft.	ODI, VAS-Back pain, VAS-Leg pain, Fusion success	Clinical outcomes showed statistically significant improvements in ODI (46.3%), VAS-Back pain (75.5%), and VAS-Leg pain (85.5%) (P < 0.01) scores at Month 24.
Navani et al	Multicenter, prospective, crossover, randomized, controlled trial	I	40 patients randomized into intradiscal PRP or BMC.	NRS for pain	Both PRP and BMC demonstrated statistically significantly improvement in pain and function.
Kim et al	Multicenter, prospective, post-market study	II	104 patients undergoing ACDF surgery using CBM allograft at 1-3 contiguous levels between C2 and C7.	VAS-Pain, NDI, ODI	Clinical outcomes evaluated with VAS-pain, NDI, and ODI scales were improved significantly at 6 and 12 months. At 12 months, per-level fusion was 98.5% for cervical and 100% for lumbar segments.
Zielinski et al	Prospective, multicenter, randomized, double-blind study	I	26 patients, ages 25 to 71 with chronic lumbar discogenic pain randomized to PRP or saline (control) groups	ODI, NRS for pain	Clinically significant improvement in 17% of PRP patients decline in 5% (1 patient) of the active group.
Gornet et al	Randomized clinical trial	I	60 patients with single-level lumbar degenerative disc disease randomized into single intradiscal injections of low-dose cells, high dose, cells, vehicle alone, or placebo.	VAS pain improvement > 30% at 52 weeks and adverse events at 4, 12, 26, 52, 78, and 104 weeks post-treatment.	High-dose intradiscal injections of allogeneic disc progenitor cells produced statistically significant improvements in pain.

FRI = functional rating index; HA = hyaluronic acid; MSC = mesenchymal stem cells; NDI = Neck Disability Index; NPS = numeric pain score; NRS = numeric rating scale; ODI: Oswestry Disability Index; PRP = platelet rich plasma; SANE = single assessment numeric evaluation; VAS = visual analogue pain score.

Table 2.

Key Works Studying the Use of Orthobiologics in Post-decompression Pain

Authors	Study design	Level of evidence	Experimental groups	Outcomes	Significant findings
Kamson et al	Prospective, single-blind (patient) randomized controlled trial	I	Cryopreserved amniotic-derived products (n = 57), bone marrow aspirate (n = 19), both (n = 10), controls (n = 183)	VAS, ODI, SF-36	Significant improvements in the BMA and ADP groups vs controls in VAS leg pain at 9 months (2.2 vs 3.6, P = 0.01), VAS back at 12 months (2.2 vs 3.6, P = 0.01). No long-term ODI or SF-36 improvements

Table 3.

Key Works Studying the Use of Orthobiologics in Spinal Cord Injury

Authors	Study design	Level of evidence	Experimental groups	Outcomes	Significant findings
Cristante et al	Prospective, uncontrolled	II	39 patients with chronic (>2 years) spinal cord injury at cervical or thoracic level administered bone marrow-derived MSCs via angiography	SSEPs 2.5 years after treatment	26/39 (66.7%) showed improved SSEPs at follow-up. No difference between paraplegic vs quadriplegic patients. Mean time of 9 months between infusion and lower limb response, and 10.4 months between infusion and upper limb response
Oh et al	Prospective, uncontrolled	II	16 patients with ASIA B quadriplegia due to traumatic cervical SCI > 1 year prior	MEP, SSEP, MRI measurements of spinal cord diameter and character	2/16 patients showed improved motor grade in their upper extremities (12.5%), and 4/16 (25%) had improved SSEPs. Increased spinal cord diameter in 5/16 patients, evidence of new tract continuity in 2/16.
El-Kheir et al	Prospective, randomized unblinded controlled trial	I	70 patients with chronic SCI between C3 and T12 > 1 year after injury randomized into receiving adherent bone marrow-derived MSCs (n = 50) and physical therapy only (n = 20)	Functional Rating Mean Score	Sustained functional improvement 18 months post-treatment in 23/50 (46%) bone marrow-derived MSC-treated patients, with better effects in patients with thoracic SCI with more recent injuries and smaller cord lesions.
Mu et al	Pre-clinical in vivo experimental study	V	Conducted on a mouse models of spinal cord injury with arms including control, hUCMSCs alone, NSCs alone, EES alone, combined hUCMSCs NSCs, and combined hUCMSCs + NSCs + EES	Histological changes, behavioral assessments, cell persistence	The combined therapy of hUCMSCs, NSCs, and EES led to superior functional recovery compared to individual treatments.

MEP = motor-evoked potential; MSC = mesenchymal stem cells; SCI = spinal cord injury; SSEP = somatosensory evoked potentials; human umbilical cord mesenchymal stem cells = hUCMSC; neural stem cell = NSC; epidural electrical stimulation = EES.

Table 4.

Key Works Studying the Use of Orthobiologics as Adjuncts During Spine Surgery

Authors	Study design	Level of evidence	Experimental groups	Outcomes	Significant findings
Anderson et al	Prospective, single-blind (patient), randomized controlled trial	I	80 patients randomized to receive cryopreserved amniotic tissue (n = 40) vs no tissue (n = 40) during lumbar microdiscectomy	ODI, SF-12, VAS back and leg pain	Significantly greater improvement in mean ODI (6.6 vs 14.4 in controls, P = 0.02) and SF-12 physical scores (P = 0.05, data not written) in amniotic tissue group 24 months post-treatment. VAS scores equivocal.
Blanco et al	Prospective, uncontrolled	II	11 patients with a single degenerative lumbar disc undergoing posterolateral spinal fusion with autologous bone marrow-derived MSCs	VAS, ODI, SF-36	Results evaluated 1 year after surgery. Radicular VAS improved from 8 to 3, back VAS from 8 to 5. Reduction in severe limitation as measured by ODI (56% pre-op to 31% post-op). Ten of 11 patients went back to work.
Mochida et al	Prospective, uncontrolled	II	9 patients with degenerative changes at a lumbar level adjacent to a planned fusion, receiving an injection of nucleus pulposus cells co-cultured with bone marrow-derived MSCs	Clinical and X-ray follow-up	All 9 patients went back to their prior jobs. No transplanted IVD worsened on MRI over a 3-year follow-up.

Despite the numerous limitations to current works, the potential of orthobiologics is undeniable. Silverman et al ¹⁰³ in their value proposition for cell-based therapies for the degenerative lumbar spine remind us that lower back pain is a leading cause of missed time from work and the most common non-oncologic indication for chronic opioid therapy in the United States. Even a measured symptomatic improvement could have a substantial impact on societal and absolute cost of care and represent a significant sparing of quality-adjusted life years that would have otherwise been lost. While cost-benefit analyses of any proposed therapy, including decision curve analysis, ¹⁰⁴ are paramount in our current age of value-based care, it is anticipated that further advancements in orthobiologics will yield improved efficacy at a cost that is markedly lower than that of surgery.

Conclusion

The use and diversity of orthobiologics continues to expand. Spine applications are generally encouraging, although there remain few high level-of-evidence studies and current works rely on small cohorts. It is anticipated that the acceptance of orthobiologics as clinical standard of care will continue to broaden as the body of literature on these techniques improves and preparation strategies become more standardized.